Validasi Antro Nurhidayah Tiasya Sanas-Compress0

Nurhidayah Tiasya Sanas
Nur Rahma
Memo rio Cakra Adi Mulia
Azizah Sudirman
Nama Peneliti Kesimpulan

Tujuan Metode apa Bagaimana cara
No dan Judul Hasil Validasi Analisis
Penelitian yg divalidasi validasinya
Peneliti Kelompok
Lingkar
Penggunaan perut
Penentuan titik Penelitian ini
Penilaian validitas kriteria SSM dengan merupakan
potong skor bertujuan
diagnostik titik potong untuk parameter
sindroma untuk
parameter laki-laki = 2,40 antropome
metabolik -Mengetahui
antropometri (sensitivitas=86%, tri yang
remaja dan titik potong
untuk memprediksi spesifi sitas=82%) terbaik
penilaian skor sindroma
skor sindroma dan untuk dalam
validitas metabolik Pengukuran
metabolik perempuan mendeteksi
diagnostik Remaja sindroma
1 berupa nilai = 2,02 Sindrom
parameter -Mengetahui metabolic
sensitivitas dan (sensitivitas=84%, Metabolik
antropometri parameter remaja
spesifi sitas setiap spesifi sitas=85) karena
(Zahra Anggita antropometri
parameter dapat digunakan memiliki
Pratiwi, terbaik untuk
antropometri yaitu untuk menentukan nilai AUC
Mubasysyir mendeteksi
IMT, lingkar perut, remaja yang berisiko paling
Hasanbasri, sindroma
dan rasio lingkar sindroma besar
Emy metabolik
perut-tinggi badan metabolik. dbanding
Huriyati,2013) pada remaja.
IMT dan
LP/TB.
Hasil uji validasi
pengukuran
antropometri dan
PLT model prediksi
dengan
cut off point
Penelitian ini Pengukuran persen
Pengukuran referensi jika dilihat
bertujuan lemak tubuh model
Model Prediksi persen dari nilai sensitivitas,
untuk prediksi Metode
Persen Lemak lemak tubuh spesifisitas, PVP,
mendapatkan dibandingkan pengukura
Tubuh Remaja dilakukan PVN, LR+,
model prediksi dengan BIA sebagai n model
Putri: dengan dan LR-nya, maka
yang memiliki gold standard. Cut prediksi
2 Studi Cross menggunaka pengukuran
validitas off point kelebihan memiliki
Sectional ( n BIA merk antropometri atau
optimal untuk berat badan validitas
Nurul Khairani, Omron model prediksi yang
memperkiraka (overweight) tertinggi
Trini Sudiarti, HBF 302 memiliki
n persen berdasarkan cut off yaitu IMT
2020) sebagai gold validitas tertinggi
lemak tubuh point
standard dari peringkat
remaja putri referensi.
tertinggi sampai
terendah adalah IMT
(kg/m2),
PLT Lee, PLT
Khairani, IMT WHO
(Z Score).
Penelitia ini
bertujuan
Studi Validasi untuk Membandingkan
Model Prediksi IMT
Pengukuran membuat pengukuran Pengukura
WHO memilki
Antropometri model prediksi antropometri (IMT n IMT WHO
sensitivitas paling
Dan Model persen lemak WHO, skinfold memiliki
tinggi yaitu 94%,
Prediksi tubuh dengan thickness, dan korelasi
Pengukuran diikuti dengan model
Terhadap melakukan lingkar pinggang) paling kuat
persen prediksi IMT WHO
Persen Lemak studi validasi dan persen lemak dengan
3 lemak tubuh dan umur (94%) dan
Tubuh Bia Pada pengukuran tubuh model persen
bia, TB dan model prediksi
Siswa Mts Dan antropometri prediksi (Slaughter, lemak
BB Sitoayu. Seluruh
Ma Multiteknik dan model Deurenberg, Lee tubuh BIA
variabel memiliki
Yayasan Asih Prediksi dan Chan) dengan disbanding
korelasi yang
Putera Cimahi (Slaughter, hasil pengukuran antropome
signifikan dengan
( Laras Sitoayu, Deurenberg, persen lemak tri lainnya
persen lemak tubuh
Trini Sudiarti) Lee dan Chan) tubuh BIA.
BIA (p < 0,0005).
terhadap
persen lemak
tubuh BIA.
Hasil validasi
Berat badan dan
Penelitian ini kesepakatan yang
tinggi badan yang
bertujuan baik yang diamati
Validity Of Self- dilaporkan sendiri
menyelidiki dari plot B & A
Reported dan diukur Berat dan
keakuratan menunjukkan bahwa
Weight, Height langsung tinggi
tinggi dan perbedaan antara
And Body Mass dikumpulkan badan yang
berat badan kedua variabel
Index Among melalui panggilan dilaporkan
remaja akhir pengukuran
Malaysian Late Berat badan, telepon dan sendiri
4 Malaysia yang independen kecil
Adolescents ( Tinggi Badan. instrumen studi. sesuai
dilaporkan dan sebagian besar
Kong Pun Pun, Nilai berat dan dengan
sendiri perbedaan berada
Nithiah tinggi yang metode
dibandingkan dalam area batas
Thangiah1, diperoleh pengukura
dengan kesepakatan pada
Hazreen Abdul digunakan untuk n langsung
metode tingkat populasi
Majid, 2021) menghitung BMI
pengukuran untuk nilai berat
dan menentukan
langsung badan, tinggi badan
klasifikasi
dan BMI
Penelitian ini Berat badan,
Anthropometry Prevalensi
bertujuan tinggi badan, Pengukuran SFT
and Body kelebihan berat
untuk pinggang, harus digunakan Penjumlaha
Composition badan dan
mengetahui dan sebagai alat skrining n 4 SFT
of Adolescents obesitas yang
tingkat lingkar yang disukai, karena dan WHtR
in Macedonia ditentukan oleh
adipositas dan pinggul (WC, mudah dilakukan memiliki
(Nazim Myrtaj referensi
obesitas pada HC) serta dan merupakan daya
5 ,dkk pertumbuhan
remaja trisep, betis, prediktor yang lebih pembeda
2018)/Antropo anak-anak IOTF
Makedonia subscapular, baik untuk terbaik
metri dan dihitung dan kurva
dan dan kegemukan untuk
Komposisi persentil smoothed
membandingk ketebalan tubuh orang dewasa mempredik
Tubuh yang bergantung
an lipatan kulit daripada BMI si obesitas.
Remaja di pada usia dan jenis
hasilnya suprailiaka remaja.
Makedonia kelamin
dengan (SFT) diukur.
penelitian untuk kurva BMI
sebelumnya dan ROC
pada populasi dihasilkan.
ini, serta yang
dilakukan
pada populasi
lain.
Data massa tubuh
antropometri – Tidak
berat dan tinggi hanya
badan, lingkar (IMT)
pinggang, dapat
Anthropometri Penelitian ini ketebalan lipatan digunakan
c indicators of bertujuan kulit trisep (TSFT) BMI, WHtR dan WC pada
obesity in the Untuk dan ketebalan memiliki remaja
prediction of mengetahui lipatan kulit betis kemampuan yang untuk
high body fat in indikator medial (MCST) lebih besar untuk mendiagno
adolescents/In antropometri BMI, WHtR diukur menurut membedakan lemak sis lemak
6
dikator obesitas dan WC prosedur standar. tubuh pada kedua tubuh yang
antropometri dalam prediksi Indeks massa jenis kelamin tinggi,
obesitas dalam tinggi lemak tubuh (BMI) dibandingkan tetapi juga
prediksi tinggi tubuh pada dihitung dan diberi dengan C-Index. indikator
lemak tubuh remaja dari peringkat obesitas
pada remaja ) Negara Brazil berdasarkan titik sentral
potong untuk (WC,
remaja , yang WHtR)
bervariasi menurut dapat
usia dan jenis digunakan
kelamin
Tangan diukur
secara manual
(menggunakan
antropometrik). (ca
3D Hand liper dan
Anthropometry Tujuan dari pita pengukur) dan Menggunakan
of Korean penelitian ini menggunakan metode pengukuran
Pengukuran
Teenager’s and yaitu untuk pindaian tangan 3D langsung ada
tangan
Comparison mengukur resolusi tinggi keterbatasan jumlah Metode 3D
kanan
with Manual antropometri (NEXHAND H-100, dimensi yang dapat dapt
(panjang,
7 Method/Antrop tangan 3D Knitech, Korea diukur. Namun menggantik
lebar, dan
ometri Tangan dan Selatan) dengan demikian, data an metode
lingkar
3D Remaja membandingk akurasi pemindaian pindaian tangan 3D langsung
tangan dan
Korea dan annya dengan ± 0. mm. Dari data mengandung
jari)
Perbandingan metode yang dipindai, dimensi tangan yang
dengan manual pengukuran tangan berbeda.
Manual metode diekstraksi
menggunakan
perangkat lunak
pemindaian
(Enhand,
Knitech, Korea
Selatan). Rata-rata
dan standar deviasi
untuk setiap
pengukuran tangan
dihitung.
Study of
Penelitian ini
anthropometric
bertujuan
profile of
untuk
working Membandingkan Data
mengetahui
(labour) Tinggi badan, Berat antropome
profil
adolescent girls Badan dan indeks trik tinggi
antropometrik
of massa tubuh dari Selisih rerata BMI badan,
meliputi berat
urban slums of remaja yang penelitian ini dengan berat
badan, tinggi Tinggi Badan,
India ( Yadav S bekerja dengan Standar NCHS untuk badan dan
badan dan Berat Badan
8 Bharti R. Standar kelompok umur yang IMT remaja
indeks massa dan indeks
2019)/Studi internasional- berbeda berkisar lebih
tubuh remaja massa tubuh
profil National Center for antara 2,15 hingga rendah dari
putri bekerja
antropometrik Health Statistik 3,57. standar
usia 10-19
gadis remaja (NCHS) dan India
tahun di
yang bekerja populasi referensi maupun
kawasan
(pekerja) di India NCHS
kumuh
daerah kumuh
perkotaan
perkotaan
Bilaspur
India
Reliability and
validity of
inexpensive
and easily
Penelitian ini Validitas kriteria
administered
bertujuan yang dilaporkan dari
anthropometric
untuk melihat Menilai dua Validitas dan skoliometer dan foto
clinical
pengukuran jenis aplikasi reliabilitas alat 2D, jika
evaluation Hasil
kelengkungan iPhone penilaian kritis dibandingkan
methods of fotografi
tulang sebagai yang terdiri dari 13 dengan sudut Cobb
postural 2D memiliki
belakang yang ukuran item, diselesaikan dinilai dari radiografi,
asymmetry korelasi
akurat dan skoliosis, oleh dua pemirsa berkisar dari rendah
measurement sedang
dapat termasuk buta (AA dan AP) hingga sangat tinggi.
9 in adolescent hingga
direproduksi penelitian untuk menilai Pengukuran iPhone
idiopathic tinggi
sangat dengan kualitas berkorelasi baik
scoliosis: dengan
penting dalam pelekatan metodologis setiap dengan pengukuran
a systematic hasil
pemeriksaan lengan akrilik artikel yang skoliometer. Hasil
review topografi
pasien yang disertakan dan fotografi 2D memiliki
(Ashleigh 3D.
dengan menyerupai hasilnya korelasi sedang
Prowse,Rodney
skoliosis skoliometer ditabulasikan hingga
Pope, Paul
idiopatik tinggi dengan hasil
Gerdhem,Allan
remaja (AIS) topografi 3D.
Abbott,2015)/K
eandalan dan
validitas
metode
evaluasi klinis
antropometrik
yang murah
dan mudah
diberikan untuk
pengukuran
asimetri
postural pada
skoliosis
idiopatik
remaja:
tinjauan
sistematis.
Persamaan dibuat
berdasarkan
kombinasi dari
sembilan variabel
antropometri dan
disajikan koefisien
Equations
determinasi (r2)
based on Validitas
sama dengan atau
anthropometry persamaan yang
lebih besar dari
to predict body diterbitkan
92,4% untuk anak
fat sebelumnya
Penelitian ini laki-laki dan 85,8%
measured by Tinggi badan, (misalnya,
bertujuan untuk anak
absorptiometry tinggi duduk, Slaughter) juga pengukura
untuk perempuan. Dalam
in berat badan, dievaluasi. %BC n
mengembangk sampel validasi,
schoolchildren lingkar diperkirakan antropome
an dan persamaan yang
and pinggang dan dengan tri tepat
memvalidasi dikembangkan
adolescents lengan, absorptiometry mempredik
persamaan menyajikan nilai r2
(Luis Ortiz- lipatan kulit sinar-X energi si% BF
untuk yang tinggi (ÿ
Hernández, (trisep, ganda (sinar-X dalam
10 memperkiraka 85,6% pada anak
dkk, 2017) / bisep, energi kelompok
n persentase laki-laki dan 78,1%
Persamaan subskapular, ganda (DXA)). heterogen
lemak tubuh pada anak
berdasarkan suprailiaka, Model regresi linier anak
anakanak dan perempuan) di
antropometri dan betis), diperkirakan sekolah
remaja dari semua kelompok,
untuk serta lebar dengan %BF dan
Meksiko tingkat kesalahan
memprediksi siku dan sebagai variabel remaja
menggunakan standar yang rendah
lemak tubuh bitrokanterik dependen dan Meksiko.
pengukuran (SE 3, 05% pada anak
yang diukur a. pengukuran
antropometri laki-laki dan 3,52%
dengan antropometrik
pada anak
absorptiometry sebagai variabel
perempuan) dan
pada anak independen.
penyadapan
sekolah dan
tidak berbeda dari
remaja
asal (p > 0,050).
Dengan
menggunakan
persamaan yang
diterbitkan
sebelumnya,
koefisien
determinasi lebih
kecil dan/atau titik
potongnya berbeda
dari titik asal.
View metadata, citation and similar papers at core.ac.uk brought to you by CORE
provided by Jurnal Gizi Klinik Indonesia
Jurnal Gizi Klinik Indonesia, Vol. 14, No. 2, Oktober 2017: 80-89
Jurnal Gizi Klinik Indonesia
Vol 14 No 2 - Oktober 2017 (80-89)
ISSN 1693-900X (Print), ISSN 2502-4140 (Online)
Online sejak Januari 2016 di https://jurnal.ugm.ac.id/jgki
Penentuan titik potong skor sindroma metabolik remaja dan penilaian

validitas diagnostik parameter antropometri: analisis Riskesdas
2013
Determination of cutoff points for metabolic syndrome scores in Indonesian adolescents and assessment of the
diagnostic validity of anthropometric parameters
Zahra Anggita Pratiwi1, Mubasysyir Hasanbasri1, Emy Huriyati2
1
Departemen Biostatistik, Epidemiologi, dan Kesehatan Populasi, Fakultas Kedokteran Universitas Gadjah Mada
2
Departemen Gizi dan Kesehatan, Fakultas Kedokteran Universitas Gadjah Mada
ABSTRACT
Background: The risk of death caused by non-communicable diseases is related to metabolic syndrome. Metabolic syndrome not
only occurs in adults, but also occurs in adolescents. The problem of metabolic syndrome in adolescents shows the importance
of early detection and management. Early detection of metabolic syndrome in adolescents can be done through non-invasive
approaches such as anthropometric measurements. However, the definition of metabolic syndrome has so far not reached an
agreement. Objective: This study aims 1) To know the intersection points of adolescent metabolic syndrome; 2) To know the best
anthropometry parameters for detecting metabolic syndrome in adolescents. Method: This study used cross sectional design, using
Riskesdas 2013 survey data. The sample size of this study was 3273 adolescents aged 15-24 years. The analysis using receiver
operating characteristic curve (ROC) indicated the accuracy of the score to diagnose metabolic syndrome, supported by area
under the curve (AUC) results. The best parameters were seen from the largest AUC values, taking into account the sensitivity
and specificity values. Results: The metabolic syndrome scores in general for Indonesian adolescents=2.21 (sensitivity=83%,
specificity=84%). Specific cut off point for women=2.02 (sensitivity=84%, specificity=85%), and for males=2.40 (sensitivity=86%,
specificity=82%). The best anthropometric parameters for detecting metabolic syndrome in adolescents are abdominal circumference
(AUC=0.77; sensitivity=71%, specificity=67%). Conclusion: Abdominal circumference has the best validity and can be used for
early detection of the risk of metabolic syndrome in adolescents.
KEY WORDS: anthropometric parameter; cut off point; Indonesian adolescent; metabolic syndrome score
ABSTRAK
Latar belakang: Risiko kematian yang disebabkan oleh penyakit tidak menular berkaitan dengan sindroma metabolik. Sindroma
metabolik tidak hanya terjadi pada orang dewasa, tetapi juga terjadi pada remaja. Permasalahan sindroma metabolik pada remaja
menunjukkan pentingnya deteksi dan manajemen dini. Upaya deteksi dini sindroma metabolik pada remaja dapat dilakukan
melalui pendekatan non-invasive seperti melakukan pengukuran antropometri. Namun, definisi sindroma metabolik sampai saat ini
masih belum mencapai kesepakatan. Tujuan: Penelitian ini bertujuan untuk 1) Mengetahui titik potong skor sindroma metabolik
remaja; 2) Mengetahui parameter antropometri terbaik untuk mendeteksi sindroma metabolik pada remaja. Metode: Penelitian
dengan desain cross sectional menggunakan data survei Riskesdas 2013. Besar sampel penelitian sebesar 3.273 remaja usia 15-
24 tahun. Analisis menggunakan kurva receiver operating characteristic (ROC) dalam mengindikasikan keakuratan skor untuk
mendiagnosis sindroma metabolik, didukung dengan hasil area under the curve (AUC). Parameter terbaik dilihat dari nilai AUC
terbesar dengan mempertimbangkan nilai sensitivitas dan spesifisitas. Hasil: Titik potong skor sindroma metabolik secara umum
untuk remaja Indonesia=2,21 (sensitivitas=83%, spesifitas=84%). Titik potong spesifik untuk perempuan=2,02 (sensitivitas=84%,
spesifisitas=85%) dan untuk laki-laki=2,40 (sensitivitas=86%, spesifisitas=82%). Parameter antropometri terbaik untuk mendeteksi
sindroma metabolik pada remaja adalah lingkar perut (AUC=0,77; sensitivitas= 71%, spesifisitas=67%). Simpulan: Lingkar perut
memiliki validitas yang paling baik dan dapat digunakan untuk deteksi dini risiko sindroma metabolik pada remaja.
KATA KUNCI: parameter antropometri; titik potong; remaja Indonesia; skor sindroma metabolik
Korespondensi: Zahra Anggita Pratiwi, Departemen Biostatistik, Epidemiologi, dan Kesehatan Populasi, Fakultas Kedokteran Universitas Gadjah Mada, Jl. Farmako
Sekip Utara, Yogyakarta 55281, e-mail: zahraanggita91@gmail.com
80
Zahra Anggita Pratiwi, dkk: Penentuan titik potong skor sindroma metabolik remaja dan penilaian validitas diagnostik parameter antropometri
PENDAHULUAN perhitungan z-score. Penggunaan skor dalam mendeteksi

sindroma metabolik membutuhkan nilai titik potong sebagai
Sindroma metabolik merupakan kumpulan gejala
dasar diagnosis. Nilai titik potong didefiniskan sebagai nilai
kelainan metabolik tubuh yang mencakup dislipidemia
antara normal dan abnormal atau nilai batas hasil uji positif
(peningkatan kadar trigliserida dan penurunan high
dan negatif (10). Di beberapa negara sudah memiliki titik
density lipoprotein/HDL), hiperglikemia, hipertensi, dan
potong skor sindroma metabolik seperti di India, Iran, dan
obesitas sentral (1). Sindroma metabolik bukan merupakan
Brazil (11–13). Indonesia belum memiliki nilai titik potong
penyakit, tetapi lebih menggambarkan kumpulan faktor
skor sindroma metabolik khususnya pada remaja sehingga
risiko metabolik yang berhubungan langsung dengan
diperlukan penelitian terkait penentuan nilai titik potong
penyakit tidak menular, terutama penyakit kardiovaskuler
skor sindroma metabolik di Indonesia.
arterosklerotik (1). Penderita sindroma metabolik berisiko
Permasalahan sindroma metabolik pada remaja
mengalami penyakit kardiovaskular dan komplikasi
menunjukkan pentingnya dilakukan deteksi dan
diabetes mellitus. Sindroma metabolik berkaitan dengan
manajemen dini (6). Diagnosis sindroma metabolik
mortalitas dan morbiditas penyakit kardiovaskular (2).
dilakukan dengan tes darah (invansive) untuk mengetahui
Penelitian menemukan sindroma metabolik tidak
kondisi metabolik tubuh. Beberapa penelitian menemukan
hanya terjadi pada kelompok usia dewasa, tetapi dapat
deteksi dini sindroma metabolik dapat dilakukan melalui
pula mulai terjadi pada usia muda. Data menunjukkan
pendekatan non-invasive yang murah dan sederhana
prevalensi sindroma metabolik pada remaja Amerika
seperti pengukuran antropometri. Hasil penelitian
mencapai 12,7% (3) dan 13% pada remaja Korea (4).
sebelumnya memaparkan bahwa pengukuran tekanan
Prevalensi sindroma metabolik pada usia lebih dari atau
darah dan indeks massa tubuh (IMT) dapat dilakukan
sama dengan 15 tahun di Indonesia sebesar 12,5% (5).
untuk memprediksi sindroma metabolik tahap awal
Kondisi pubertas merupakan sumber potensial terjadinya
ketika tes darah tidak dapat dilakukan (14). Hal tersebut
sindroma metabolik pada remaja karena terjadi perubahan
didukung dengan penelitian lain yang menyatakan bahwa
regulasi hormon dan distribusi lemak yang dapat
pengukuran antropometri yang teratur, penting untuk
menyebabkan pertambahan berat badan (6).
mendeteksi risiko sindroma metabolik pada kelompok
Beberapa ahli telah merumuskan definisi sindroma
eksekutif di Jakarta (15).
metabolik. Saat ini terdapat tiga definisi sindroma
Pengukuran antropometri menggambarkan
metabolik yang telah dirumuskan dan sering digunakan
pengukuran massa tubuh, ukuran, bentuk, dan tingkat
pada penelitian yaitu definisi World Health Organization
kegemukan. Pengukuran antropometri yang sering
(WHO); National Cholesterol Education Program Expert
dilakukan untuk memperkirakan komposisi tubuh adalah
Panel on Detection, Evaluation and Treatment of High
indeks massa tubuh (IMT), lingkar perut (LP), dan
Blood Cholesterol In Adults Treatment Panel III (NCEP
rasio lingkar perut terhadap tinggi badan (LP/TB) (2).
ATP-III); dan International Diabetes Federeation (IDF).
Penting adanya penelitian diagnostik terkait parameter
Ketiga definisi tersebut memiliki komponen utama sama
antropometri mana yang lebih sensitif dan spesifik dalam
dengan penentuan kriteria yang berbeda (7). Belum
mendeteksi risiko sindroma metabolik pada remaja.
adanya kesepakatan dalam mendefinisikan sindroma
Tujuan penelitian ini adalah untuk mengetahui nilai
metabolik memungkinkan penemuan yang berbeda pada
titik potong skor sindroma metabolik pada remaja dan
penelitian epidemiologi dan rendahnya prevalensi pada
mengetahui parameter antropometri yang paling baik
beberapa populasi. Alasan itu yang mendorong American
dalam mendeteksi sindroma metabolik pada remaja.
Diabetes Association dan European Association
for the Study of Diabetes merekomendasikan untuk
menggunakan continuous value of metabolic syndrome BAHAN DAN METODE
atau skor sindroma metabolik (SSM) (8,9).
Penelitian kuantitatif ini menggunakan data
Skor sindroma metabolik adalah penilaian semua
sekunder Riset Kesehatan Dasar (Riskesdas) 2013.
komponen sindroma metabolik dengan menggunakan
81
Riskesdas merupakan survei skala nasional berkala yang serta mengukur berat badan, tinggi badan, lingkar perut,
dilakukan oleh Badan Penelitian dan Pengembangan dan tekanan darah. Kriteria eksklusi penelitian ini adalah
Kesehatan Kementerian RI untuk memantau indikator subjek dengan data yang tidak lengkap pada variabel
kesehatan seluruh wilayah Indonesia. Penelitian yang diteliti. Alur pemilihan subjek penelitian tercantum
Riskesdas 2013 dilakukan di 33 provinsi, 497 kabupaten/ pada Gambar 1.
kota yang mencakup 12.000 blok sensus terpilih. Penentuan titik potong skor sindroma metabolik
Pemeriksaan biomedis yang dilakukan mewakili tingkat menggunakan analisis kurva Receiver Operating
nasional yang merupakan sub-sample provinsi (1000 blok Characteristic (ROC) dan Area Under The Curve (AUC).
sensus). Pengumpulan data (wawancara, pengukuran, Area di bawah kurva ROC digunakan untuk menilai
pemeriksaan, pengambilan spesimen darah, data entry, keakuratan suatu diagnosis. Semakin luas AUC, maka
dan validasi) dilakukan pada bulan Mei sampai Juni menunjukkan tes atau uji terbaik. Dalam menentukan
2013 (16). titik potong skor sindroma metabolik, membutuhkan
Unit analisis penelitian ini adalah remaja berusia pembanding standar. Penelitian ini menggunakan baku
15 sampai 24 tahun yang mengikuti survei Riskesdas standar kriteria sindroma metabolik NCEP ATP-III
2013 dan melakukan pemeriksaan sampel darah (kadar modifikasi untuk remaja. Definisi sindroma metabolik
glukosa darah puasa, kolesterol HDL, dan trigliserida), berdasarkan NCEP ATP-III adalah seseorang menderita
Jumlah remaja usia 15 – 24 tahun (unit analisis) yang mengikuti survei

Riskesdas 2013 = 3568 orang
Tidak terdapat data berat badan dan

atau tinggi badan, sehingga tidak
terdapat data IMT = 16
Jumlah remaja, setelah drop missing value pertama

= 3552 orang
Tidak terdapat data lingkar perut

= 80
Jumlah remaja, setelah drop missing value kedua

= 3472 orang
Tidak terdapat data tekanan

darah = 104
Jumlah remaja, setelah drop missing value ketiga = 3368 orang
Nilai ekstrim (oulier) pada

variabel SSM = 95
Jumlah remaja, setelah drop outlier
= 3273 orang
Gambar 1. Alur pemilihan subjek penelitian
82
sindroma metabolik jika menderita minimal tiga dari HASIL

lima kriteria. Lima kriteria tersebut adalah 1) Lingkar
Deskripsi subjek penelitian
perut ≥90th persentil; 2) HDL ≤40 mg/dL; 3) Trigliserida
≥110 mg/dL; 4) Tekanan darah ≥90th persentil; dan 5) Subjek penelitian merupakan remaja Indonesia
Glukosa darah puasa (GDP) ≥110 mg/dL (17). Alasan berusia 15-24 tahun yang merupakan subjek penelitian
kriteria NCEP ATP-III digunakan sebagai standar baku Riskesdas 2013. Rerata usia subjek adalah 19±2,87
karena lebih sederhana dan reliable untuk digunakan tahun. Total sampel penelitian ini sebanyak 3.273
dibandingkan definisi WHO, European Group for the subjek, laki-laki (43,63%) dan perempuan (56,37%).
Study of Insulin Resistance (EGIR), dan International Sebagian besar subjek penelitian memiliki status gizi
Diabetes Federation (IDF) (18,19). normal. Sebanyak 6,90% subjek penelitian ini memiliki
Perhitungan skor sindroma metabolik (SSM) kadar GDP yang tidak normal. Subjek yang memiliki
dilakukan untuk semua komponen sindrom metabolik kadar HDL kolesterol rendah (tidak normal) sebanyak
(lingkar perut, tekanan darah, trigliserida, HDL, dan 43,29%. Kadar trigliserida subjek yang tidak normal
glukosa darah puasa) dengan menggunakan z-score sebanyak 21,51% serta subjek yang mengalami obesitas
yang mengacu pada penelitian sebelumnya (20). abdominal sebanyak 5,04% dan hipertensi sebanyak
Skor sindroma metabolik (SSM) merupakan total 21,08% (Tabel 1.).
z-score semua komponen sindroma metabolik. Sistem
skoring tekanan darah menggunakan mean arterial Tabel 1. Karakteristik subjek penelitian
pressure (MAP)=[(tekanan darah sistolik-tekanan
darah diastolik/3)+tekanan darah diastolik] (13,21). Frekuensi Persentase
Variabel
(n=3.273) (%)
Rumus perhitungan z-score untuk lingkar perut, MAP, Jenis kelamin
trigliserida, dan glukosa darah puasa: Laki-laki 1.428 43,63
Perempuan 1.845 56,37
Z-score = Kategori usia
Remaja awal 1.546 47,78
Remaja akhir 1.709 52,22
Perhitungan z-score HDL, digunakan rumus Status Gizi
berikut: Kurus 517 15,80
Normal 2.425 74,09
Z-score =
Overweight 190 5,81
Obesitas 141 4,31
Interpretasi skor sindroma metabolik (SSM) adalah Glukosa darah puasa
semakin rendah skor yang dimiliki seseorang, maka Normal 3.047 93,10
Tidak normal 226 6,90
semakin baik kondisi metabolime (20,22).
HDL
Penilaian validitas diagnostik parameter Normal 1.856 55,71
antropometri untuk memprediksi skor sindroma metabolik Tidak normal 1.417 43,29
berupa nilai sensitivitas dan spesifisitas setiap parameter Trigliserida
Normal 2.569 78,49
antropometri yaitu IMT, lingkar perut, dan rasio lingkar Tidak normal 704 21,51
perut-tinggi badan. Analisis yang dilakukan adalah Lingkar pinggang
analisis kurva ROC yang didukung dengan hasil AUC. Obesitas abdominal 165 5,04
AUC dibagi menjadi tiga kategori yaitu, 1) Akurasi rendah Tidak obesitas abdominal 3.108 94,96
Tekanan darah
jika 0,5<AUC≤0,7; 2) Akurasi sedang jika 0,7<AUC≤0,9; Hipertensi 690 21,08
3) Akurasi tinggi jika 0,9<AUC≤ 1,0. Parameter terbaik Tidak hipertensi 2.583 78,92
adalah yang memiliki nilai AUC terbesar (13).
83
Tabel 2. Distribusi ukuran komponen sindroma metabolik berdasarkan jenis

kelamin
Komponen sindroma Rerata

p1
metabolik Laki-laki Perempuan
Lingkar perut (cm) 70,35 ± 7,61 71,04 ± 8,82 0,0192
Tekanan sistolik (mmHg) 115,93±11,44 109,42±10,82 <0,0001
Tekanan diastolik (mmHg) 74,08 ± 8,92 75,67 ± 8,54 <0,0001
GDP2 (mg/dL) 95,95 ± 9,16 94,51 ± 9,51 <0,0001
HDL3 (mg/dL) 45,09 ± 9,78 51,50 ± 11,77 <0,0001
Trigliserida (mg/dL) 96,09 ± 42,34 86,95 ± 41,05 <0,0001
1
uji Chi-Square, signifikan secara statistik (p<0,05); 2GDP=glukosa darah puasa; 3HDL=high density
lipoprotein
perempuan memiliki kadar kolesterol HDL lebih tinggi

1.00
daripada laki-laki dan bermakna secara statistik (Tabel

2).
0.75
Sensitivity
Penentuan titik potong skor sindroma metabolik

0.50
Penentuan titik potong secara statistik melalui dua

0.25
tahap yaitu pertama, dengan melakukan analisis regresi

logistik; kedua, dengan membuat grafik sensitivitas-
0.00
0.00 0.25 0.50 0.75 1.00

spesifisitas untuk menentukan titik potong. Pada hasil
1 - Specificity analisis regresi logistik didapatkan nilai p <0,0001 (95%
Area under ROC curve = 0.9119
CI:0,72 – 0,88), artinya SSM secara statistik berhubungan

Gambar 2. Nilai AUC dengan kurva ROC dengan kriteria sindroma metabolik NCEP ATP-III.
Kurva ROC yang menggambarkan nilai AUC
disajikan pada Gambar 2. Terlihat nilai AUC SSM
1.00
terhadap sindroma metabolik adalah sebesar 0,9119 yang

0.75
menunjukkan bahwa SSM memiliki akurasi yang tinggi

Sensitivity/Specificity
secara statistik untuk mendiagnosis sindroma metabolik

0.50
pada remaja. Langkah selanjutnya adalah membuat grafik

sensitivitas dan spesifisitas untuk menentukan nilai titik
0.25
potong. Titik potong probabilitas SSM diketahui dari

kurva perpotongan antara sensitivitas dan spesifisitas
0.00
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00
pada Gambar 3.
Probability cutoff
Sensitivity Specificity Nilai titik potong optimal SSM yang
tergambarkan dari titik potong probabilitas adalah
Gambar 3. Kurva ROC dalam penentuan titik potong
SSM 0,065 (sensitivitas=83,1%; spesifisitas=84,2%). Hal ini
menunjukkan probabilitas 6,5% merupakan nilai batas
Remaja perempuan memiliki rerata lingkar perut optimal untuk mendeteksi dan membedakan antara
lebih tinggi dibandingkan remaja laki-laki dan signifikan remaja yang menderita sindroma metabolik dan orang
secara statistik (p=0,0192; p>0,05). Sementara itu, sehat. Titik potong probabilitas SSM terhadap SM (NCEP
tekanan sistolik dan diastolik, kadar glukosa darah puasa ATP III) tersebut bertepatan dengan skor sindroma
(GDP), HDL, dan trigliserida remaja laki-laki lebih tinggi metabolik sebesar 2,21. Dapat disimpulkan secara umum
daripada remaja perempuan. Penelitian ini menemukan remaja usia 15-24 tahun di Indonesia yang memiliki SSM
84
1.00
1.00
0.75
0.75
0.50
0.50
0.25
0.25
0.00
0.00
A B
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00
Probability cutoff Probability cutoff
Sensitivity Specificity Sensitivity Specificity
Keterangan: A: Kurva ROC untuk laki-laki; B: Kurva ROC untuk perempuan

Gambar 4. Kurva ROC penentuan titik potong berdasarkan jenis kelamin
A B
Keterangan: A: Prevalensi sindroma metabolik berdasarkan kriteria NCEP ATP-III; B: Prevalensi

sindroma metabolik dengan skoring (SSM)
Gambar 5. Prevalensi sindroma metabolik remaja
Tabel 3. Perbandingan titik potong SSM berdasarkan jenis kelamin
n AUC 95% Conf. Interval Titik potong Sensitivitas Spesifisitas

Laki-laki 1428 0,92 0,76 - 1,90 2,40 86% 82%
Perempuan 1845 0,92 0,76 - 1,07 2,02 84% 85%
Umum 3273 0,91 0,79 - 1,01 2,21 83% 84%
di atas 2,21 didiagnosis menderita sindroma metabolik. titik potong SSM laki-laki lebih tinggi dibandingkan titik
Titik potong SSM dibedakan berdasarkan jenis kelamin potong perempuan.
karena perempuan dan laki-laki memiliki perbedaan Hasil penelitian ini menunjukkan prevalensi
kecenderungan untuk menderita SM. Kurva ROC dalam sindroma metabolik dengan metode skoring atau yang
penentuan titik potong SSM berdasarkan jenis kelamin dikenal dengan SSM pada remaja usia 15-24 tahun
ditampilkan pada Gambar 4. sebesar 19,89%. Di sisi lain, prevalensi sindroma
Lebih lanjut, Tabel 3. menunjukkan perbandingan metabolik berdasarkan kriteria NCEP ATP-III lebih
titik potong SSM pada laki-laki dan perempuan, serta rendah dibandingkan dengan prevalensi menggunakan
titik potong SSM secara umum. Berdasarkan analisis, metode skoring yaitu sebesar 5,93% (Gambar 5.).
85
Persen (%)
Gambar 6. Perbedaan prevalensi sindroma metabolik berdasarkan jenis kelamin

1.00
1.00
0.75
0.75
0.50
0.50
0.25
0.25
0.00
0.00
0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00
1 - Specificity 1 - Specificity
A Area under ROC curve = 0.7744
B Area under ROC curve = 0.7455
1.00
Keterangan:
0.75
A: AUC lingkar perut = 0,77

B: AUC LP/TB = 0,74
0.50
C: AUC IMT = 0,73

0.25
0.00
0.00 0.25 0.50 0.75 1.00

1 - Specificity
C Area under ROC curve = 0.7310
Gambar 7. Area Under Curve (AUC) parameter antropometri untuk diagnosis kategori SSM
Prevalensi sindroma metabolik baik berdasarkan Penilaian diagnostik parameter antropometri

kriteria NCEP ATP-III maupun metode skoring (SSM)
Parameter antropometri yang digunakan untuk
menunjukkan bahwa remaja laki-laki memiliki prevalensi
mendiagnosis SSM adalah lingkar perut (LP), indeks
lebih tinggi untuk menderita sindroma metabolik daripada
massa tubuh (IMT), dan lingkar perut berdasarkan tinggi
remaja perempuan (Gambar 6.). Prevalensi remaja laki-
badan (LP/TB). Parameter antropometri yang paling
laki menderita sindroma metabolik berdasarkan kriteria
baik untuk mendiagnosis sindroma metabolik (dengan
NCEP ATP-III sebesar 5,74% sedangkan prevalensi
nilai titik potong SSM = 2,21) adalah lingkar perut (titik
sindroma metabolik pada perempuan sebesar 5,2%.
86
Tabel 4. Perbandingan uji diagnosis parameter antropometri terhadap kategori SSM
Parameter Sensitivitas Spesifisitas AUC 95% CI p

Lingkar perut 71% 67% 0,77 0,75 – 0,79
LP/TB 71% 64% 0,74 0,72 – 0,77 <0,0001
IMT 69% 64% 0,73 0,71 – 0,75
potong=71,50; sensitivitas=71%; spesifisitas=67%). remaja perempuan, baik menggunakan kriteria NCEP

Hasil analisis menggunakan kurva ROC disajikan pada ATP-III maupun metode skoring. Penemuan ini sejalan
Gambar 7. dengan hasil penelitian yang menggunakan data
Lingkar perut merupakan parameter antropometri Epidemiological Study on the Insulin Resistance
yang terbaik dalam mendeteksi SM karena memiliki Syndrome (D.E.S.I.R) pada kelompok dewasa usia 30-65
nilai AUC paling besar dibandingkan dua parameter tahun, bahwa laki-laki (12,4%) memiliki prevalensi lebih
antropometri lainnya. Lingkar perut memiliki nilai tinggi menderita sindroma metabolik (NCEP ATP-III)
sensitivitas dan spesifisitas tertinggi dibandingkan dibandingkan perempuan (7,2%) (24). Remaja Algeria
sensitivitas dan spesifisitas IMT dan LP/TB. Perbedaan dan Korea juga memiliki prevalensi sindroma metabolik
AUC ketiga parameter antropometri tersebut bermakna (NCEP ATP-III) lebih tinggi pada remaja laki-laki (4% dan
secara statistik (Tabel 4.). 15,4%) daripada remaja perempuan (2% dan 10%) (4,25).
Namun, penelitian lain menunjukkan bahwa perempuan
justru memiliki prevalensi lebih tinggi dibandingkan
BAHASAN
laki-laki. Prevalensi sindroma metabolik pada kelompok
Penelitian ini menemukan hampir sebagian subjek usia 45-54 tahun lebih tinggi pada perempuan (20,2%)
penelitian memiliki kadar kolesterol HDL yang rendah. Hal daripada laki-laki (14,2%) (26).
ini dapat disebabkan oleh adanya gangguan faktor-faktor Penelitian lain yang dilakukan di Brazil pada anak-
biogenesis atau akibat kondisi subjek yang mengalami anak usia 8 dan 9 tahun menemukan sebanyak 24% anak
obesitas, merokok, aktivitas fisik yang rendah, dan kurang yang didiagnosis menderita sindroma metabolik dengan
mengkonsumsi serat. Konsumsi serat yang tidak mencukupi menggunakan metode skoring dan prevalensinya turun
kebutuhan dapat menyebabkan naiknya hidrolisis HDL dan menjadi 8,9% apabila menggunakan kriteria NCEP
meningkatkan sintesis trigliserida yang berdampak pada ATP-III modifikasi usia (13). Perbedaan yang besar
peningkatan kadar trigliserida (23). antara prevalensi menggunakan metode skoring atau
Kriteria sindroma metabolik sampai saat ini SSM dengan kriteria NCEP ATP-III disebabkan oleh
masih belum mencapai kesepakatan, terutama pada pengukuran menggunakan SSM memiliki statistical
anak-anak dan remaja. Terdapat perbedaan prevalensi power lebih besar daripada menggunakan klasifikasi
sindroma metabolik dari berbagai hasil penelitian. dikotomi (11,13).
Perbedaan tersebut selain disebabkan karena perbedaan Pada penelitian ini, titik potong SSM dibedakan
demografis, dapat juga disebabkan oleh tidak konsistennya berdasarkan jenis kelamin karena pertimbangan
penggunaan kriteria dalam mendiagnosis sindroma perbedaan biologis dan fisiologis antara laki-laki dan
metabolik. Penelitian ini menemukan prevalensi perempuan. Titik potong SSM untuk remaja laki-laki
sindroma metabolik lebih tinggi bila menggunakan lebih tinggi dibandingkan perempuan. Hasil analisis
skoring daripada menggunakan kriteria NCEP ATP-III menunjukkan validitas diagnostik titik potong SSM,
modifikasi. Sindroma metabolik lebih direkomendasikan baik laki-laki maupun perempuan sama-sama memiliki
menggunakan tipe kontinyu (skoring) dibandingkan nilai sensitivitas dan spesifisitas yang baik. Sementara
dengan dikotomi atau biner (“ya” dan “tidak”) (11). itu, penelitian sebelumnya juga menemukan titik
Penelitian ini menemukan prevalensi sindroma potong SSM lebih tinggi pada laki-laki dibandingkan
metabolik lebih tinggi pada remaja laki-laki daripada perempuan. Hal ini dapat disebabkan adanya perbedaan
87
karakteristik komposisi tubuh dan distribusi lemak. dapat mencegah perkembangan sindroma metabolik
Laki-laki lebih cenderung mengalami obesitas sentral menjadi penyakit kardiovaskular atau diabetes melitus
(berbentuk apel), sementara perempuan lebih cenderung di masa depan.
mengalami obesitas dengan akumulasi lemak pada
panggul (berbentuk pear). Hal tersebut terjadi sejak masa
SIMPULAN DAN SARAN
pubertas (27,28).
Penentuan titik potong SSM yang dilakukan di Sindroma metabolik tidak hanya terjadi pada usia
Brazil menemukan titik potong optimal SSM sebesar dewasa, tetapi juga terjadi pada remaja. Penggunaan
1,86 (sensitivitas 96,7%; spesitifitas 82,7%) dengan kriteria SSM dengan titik potong untuk laki-laki = 2,40
nilai AUC sebesar 0,96 (13). Sementara itu, penelitian (sensitivitas=86%, spesifisitas=82%) dan untuk perempuan
serupa yang dilakukan di Iran menemukan titik potong = 2,02 (sensitivitas=84%, spesifisitas=85) dapat digunakan
optimal SSM pada anak-anak usia 8 dan 9 tahun sebesar untuk menentukan remaja yang berisiko sindroma
-1,15 (sensitivitas 89%; spesifisitas 88,2%). Perbedaan metabolik. Lingkar perut memiliki validitas yang paling
titik potong pada penelitian yang dilakukan di Indonesia, baik dan dapat digunakan untuk deteksi dini risiko sindroma
Brazil, dan Iran dapat disebabkan oleh perbedaan usia, metabolik pada remaja. Program deteksi dini pada remaja
lingkungan, gaya hidup, dan ras (29). khususnya remaja awal dapat melibatkan pihak sekolah
Hasil penelitian ini menunjukkan lingkar perut adalah maupun dinas kesehatan setempat. Program deteksi dini
parameter antropometri yang paling baik dalam mendeteksi dapat dilakukan sebagai sistem monitoring kesehatan yang
sindroma metabolik pada remaja karena memiliki nilai dilakukan berkesinambungan dan terus menerus.
AUC tertinggi daripada IMT dan LP/TB. Sejalan dengan
penelitian yang di lakukan di Iran, meskipun penelitian UCAPAN TERIMAKASIH
tersebut juga melibatkan subjek lansia (30). Berbeda dengan
Peneliti mengucapkan terima kasih kepada Kepala
hasil penelitian lain yang justru menemukan lingkar perut
Badan Penelitian dan Pengembangan RI yang telah
tidak berhubungan dengan sindroma metabolik pada orang
memberikan izin menggunakan data survey Riskesdas
obesitas dan overweight. Parameter antropometri yang
2013 sehingga penelitian ini dapat terlaksana.
paling baik untuk memprediksi adiposit total pada anak-
anak dan remaja usia 8-18 tahun adalah rasio lingkar perut
Pernyataan konflik kepentingan
terhadap tinggi badan (LP/TB) (31). Rasio LP/TB mampu
mengidentifikasi persen lemak tubuh pada usia muda di Tidak ada konflik kepentingan dalam penelitian ini.
Australia (32).
Pengukuran lingkar perut memiliki keunggulan
dari segi proses yang cepat, mudah, murah, dan lebih DAFTAR PUSTAKA
sensitif dan spesifik. Pengukuran lingkar perut dapat 1. Alberti KGMM, Zimmet P, Shaw J. The metabolic
dijadikan sebagai alat deteksi dini sindroma metabolik syndrome - a new worldwide definition. Lancet
yang non-invansive pada remaja. Program deteksi dini 2005;366(9491):1059–62.
2. Bray GA, Ryan D. Overweight and the metabolic
merupakan salah satu bentuk aktivitas dari promosi
syndrome: from bench to bedside. USA: Springer; 2006.
kesehatan yang berfokus pada pencegahan penyakit (ill
3. De Ferranti SD, Gauvreau K, Ludwig DS, Newburger JW,
health prevention). Deteksi dini termasuk dalam kategori Rifai N. Inflammation and changes in metabolic syndrome
secondary prevention yang bertujuan untuk mengurangi abnormalities in US adolescents: findings from the 1988-
prevalensi penyakit dengan memperpendek durasi 1994 and 1999-2000 National Health and Nutrition
penyakit (33). Contoh, ketika seseorang diindikasikan Examination Surveys. Clin Chem 2006;52(7):1325–30.
4. You M-A, Son Y-J. Prevalence of metabolic syndrome and
berisiko sindroma metabolik dari pengukuran lingkar
associated risk factors among Korean adolescents: analysis
perut, maka dapat dilakukan penanganan dini seperti
from the Korean national survey. Asia-Pacific J Public Heal
merubah gaya hidup secara bertahap sehingga diharapkan 2012;24(3):464–71.
88
5. Bantas K. Perbedaan gender pada kejadian sindrom surveillance 2006. Acta Med Indones 2010;42(4):199–
metabolik pada penduduk perkotaan di Indonesia. Kesmas, 203.
Jurnal Kesehatan Masyarakat Nasional 2012;7(5):219– 20. Eisenmann JC. On the use of a continuous metabolic
26. syndrome score in pediatric research. Cardiovasc Diabetol
6. Wang J, Zhu Y, Cai L, Jing J, Chen Y, Mai J, et al. Metabolic 2008;7:17.
syndrome and its associated early-life factors in children 21. Eisenmann JC, Laurson KR, DuBose KD, Smith BK,
and adolescents: a cross-sectional study in Guangzhou, Donnelly JE. Construct validity of a continuous metabolic
China. Public Health Nutr 2015;19(13):1–8. syndrome score in children. Diabetol Metab Syndr
7. Rini S. Sindrom metabolik. J Major 2015;4(4):88–93. 2010;2:8.
8. Kahn R, Buse J, Ferrannini E, Stern M. The metabolic 22. Stabelini Neto A, de Campos W, Dos Santos GC,
syndrome: time for a critical appraisal. Diabetes Cares Mazzardo Junior O. Metabolic syndrome risk score and
2005;28(9):2289–304. time expended in moderate to vigorous physical activity
9. Okosun IS, Boltri JM, Lyn R, Davis-Smith M. Continuous in adolescents. BMC Pediatr 2014;14:42.
metabolic syndrome risk score, body mass index percentile, 23. Velásquez-villa M, Gómez-ocampo L, Bermúdez-cardona J.
and leisure time physical activity in American children. J Abdominal obesity and low physical activity are associated
Clin Hypertens (Greenwich) 2010;12(8):636–44. with insulin resistance in overweight adolescents : a cross-
10. Sastroasmoro S, Ismael S. Dasar-dasar metodologi sectional study. BMC Pediatr 2014;14(52):1–9.
penelitian klinis. edisi ke-4. Jakarta: Sagung seto; 2011. 24. Hillier TA, Rousseau A, Lange C, Lépinay P, Cailleau
11. Hosseini M, Sarrafzadegan N, Kelishadi R, Monajemi M, M, Balkau B, et al. Practical way to assess metabolic
Asgary S, Vardanjani HM. Population-based metabolic syndrome using a continuous score obtained from principal
syndrome risk score and its determinants: The Isfahan components analysis. Diabetologia 2006;49(7):1528–35.
Healthy Heart Program. J Res Med Sci 2014;9(12):1167- 25. Benmohammed K, Valensi P, Benlatreche M, Nguyen MT,
74. Benmohammed F, Pariès J, et al. Anthropometric markers
12. Pandit D, Chiplonkar S, Khadilkar A, Kinare A, Khadilkar for detection of the metabolic syndrome in adolescents.
V. Efficacy of a continuous metabolic syndrome score in Diabetes Metab 2015;41(2):138–44.
Indian children for detecting subclinical atherosclerotic 26. Sihombing M, Tjandrarini DH. Faktor risiko sindrom
risk. Int J Obes (Lond) 2011;35(10):1318–24. metabolik pada orang dewasa di Kota Bogor. Penel Gizi
13. Villa JKD, e Silva AR, Santos TSS, Ribeiro AQ, Sant’Ana dan Makanan 2015;38(1):21–30.
LFDR. Metabolic syndrome risk assessment in children: 27. Kelsey MM, Zeitler PS. Insulin resistance of puberty. Curr
use of a single score. Rev Paul Pediatr 2015;33(2):187– Diab Rep 2016;16(7):64.
93. 28. Neinstein LS. Handbook of adolescent health care. United
14. Hsiung D-Y, Liu C-W, Cheng P-C, Ma W-F. Using States: Lippincott Williams & Wilkins; 2009.
non-invasive assessment methods to predict the risk of 29. Susilawati MD, Bantas K, Jahari AB. Nilai batas dan
metabolic syndrome. Appl Nurs Res 2014;28(2):72–7. indikator obesitas terhadap terjadinya diabetes mellitus
15. Kamso S, Dharmayati P, Lubis U, Juwita R, Kurnia Y, tipe 2. Penel Gizi Makanan 2014;2(1):11–20.
Besral R. Prevalensi dan determinan sindrom metabolik 30. Gharipour M, Sadeghi M, Dianatkhah M, Bidmeshgi
pada kelompok eksekutif di Jakarta dan sekitarnya. Kesmas, S, Ahmadi A, Tahri M, et al. The cut-off values of
Jurnal Kesehatan Masyarakat Nasional 2011;6(2):85–90. anthropometric indices for identifying subjects at risk
16. Kementerian Kesehatan. Riset Kesehatan Dasar 2013. for metabolic syndrome in Iranian elderly men. J Obes
Jakarta; Kemenkes RI; 2013. 2014;2014.
17. Kassi E, Pervanidou P, Kaltsas G, Chrousos G. Metabolic 31. Brambilla P, Bedogni G, Heo M, Pietrobelli A. Waist
syndrome: definitions and controversies. BMC Med circumference-to-height ratio predicts adiposity better than
2011;9(1):48. body mass index in children and adolescents. Int J Obes
18. Moy FM, Bulgiba A. The modified NCEP ATP III criteria 2013;37(7):943–6.
maybe better than the IDF criteria in diagnosing metabolic 32. Nambiar S, Hughes I, Davies PS. Developing waist-
syndrome among Malays in Kuala Lumpur. BMC Public to-height ratio cut-offs to define overweight and
Health 2010;10(678):2–7. obesity in children and adolescents. Public Health Nutr
19. Soewondo P, Purnamasari D, Oemardi M, Waspadji 2010;13(10):1566–74.
S, Soegondo S. Prevalence of metabolic syndrome 33. Carr SNUTP-M. An introduction to public health and
using NCEP/ATP III criteria in Jakarta, Indonesia: the epidemiology. second edition. New York: Open University
Jakarta primary non-communicable disease risk factors Press; 2007.
89
See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/358707021
Nutri-Sains Model Prediksi Persen Lemak Tubuh Remaja Putri: Studi Cross
Sectional
Negative Results · January 2020

DOI: 10.21580/ns.2020.4.1.4367
CITATIONS READS
0 88
1 author:
Trini Sudiarti
University of Indonesia
32 PUBLICATIONS 119 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Publication Preview Source Effects of Restrained, External, and Emotional Eating Styles on Weight Gain Among Female Students at Faculty of Public Health, Universitas
Indonesia View project
stunting View project
All content following this page was uploaded by Trini Sudiarti on 18 February 2022.
The user has requested enhancement of the downloaded file.

Nutri-Sains: Jurnal Gizi, Pangan dan Aplikasinya, Vol 4, No 1(2020): 51-66
DOI: 10.21580/ns.2020.4.1.4367
Copyright © 2020 Nutri-Sains: Jurnal Gizi, Pangan dan Aplikasinya
Model Prediksi Persen Lemak Tubuh Remaja Putri:

Studi Cross Sectional
Nurul Khairani1, Trini Sudiarti2

1Program Studi S1 Kesehatan Masyarakat, STIKES Tri Mandiri Sakti, Bengkulu, Indonesia
2Departemen Gizi Kesehatan Masyarakat, Universitas Indonesia, Depok, Indonesia
Email : nurul.sulaksono@gmail.com
Abstract
The purpose of this study was to get the prediction model which had optimum validity for
estimating body fat percentage of adolescent girls. The design of this study was a Cross Sectional.
The sampels were 110 school girls which taken by Stratified Proportional technique.
Anthropometric measurements consisted of measurements of age, weight, and height (to obtain the
BMI (kg/ m2) and BMI (WHO Z Score)), waist circumference, and skinfold thickness. Body fat
percentage was measured using various predictive models. Bivariate analysis used was the correlation
test. Multivariate analysis used was multiple linear regression test. The ROC test was used for the
validation test to determine the Area Under Curve (AUC) value, sensitivity, specificity, Positive
Predictive Value (PVP), Negative Predictive Value (PVN), Likelihood Ratio + or (LR +),
and LR-. The result of the study showed that the average of body fat percentage of samples was
26,51% ± 5,48%. The prediction model which obtained from multivariate analysis was BFP=
0,991 BMI +0,069 ST+0,249 A -1,703. Based on the validation test, the prediction model of
this study had optimum validity in compared with other prediction models.
Keywords : Adolescent girls, body fat percentage, nutritional status, prediction model
Abstrak
Tujuan dari penelitian ini adalah untuk mendapatkan model prediksi yang
memiliki validitas optimal untuk memperkirakan persen lemak tubuhremaja putri.
Desain penelitian adalah Cross Sectional. Sampel yang diambil sebanyak 110 siswi
dengan menggunakan teknik Stratifikasi Proporsi. Pengukuran antropometri terdiri
dari pengukuran usia, berat badan dan tinggi badan (untuk mendapatkan nilai IMT
(kg/m2) dan IMT WHO Z Score), lingkar pinggang, dan skinfold thickness. Persen
lemak tubuh diukur dengan menggunakan berbagai model prediksi. Analisis bivariat
menggunakan uji korelasi. Analisis multivariat menggunakan uji regresi linier ganda.
Uji ROC digunakan untuk uji validasi untuk mengetahui nilai Area Under Curve
(AUC), sensitivitas, spesifisitas, Predictive Value Positif (PVP), Predictive Value Negative
(PVN), Likelihood Ratio + atau (LR+), dan LR-. Hasil penelitian menunjukkan
bahwa rerata persen lemak tubuh responden adalah 26,51 % ± 5,48 %. Model
prediksi yang didapatkan dari hasil multivariat adalah PLT= 0,991 IMT + 0,069
ST + 0,249 U -1,703. Berdasarkan hasil uji validasi, model prediksi tersebut
memiliki validitas optimal jika dibandingkan dengan model prediksi lainnya.
Kata Kunci : Model prediksi, persen lemak tubuh, remaja putri, status gizi
ISSN 2528-3170 (printed media); ISSN 2541-5921 (online media) Nutri-Sains

http://journal.walisongo.ac.id/index.php/Nutri-Sains
51
PENDAHULUAN
Anak yang mengalami kegemukan atau obesitas sudah menjadi masalah
kesehatan masyarakat, baik di negara maju maupun negara berkembang, terutama di
daerah perkotaan (WHO, 2016). Obesitas pada anak-anak dan remaja berdampak
kesehatan jangka pendek dan jangka panjang. Dampak kesehatan yang paling
signifikan dari kegemukan dan obesitas pada masa kanak-kanak, yang seringkali
tidak terlihat sampai dewasa, meliputi penyakit kardiovaskular, diabetes mellitus,
gangguan muskuloskeletal, dan kanker. Setidaknya 2,6 juta orang setiap tahun
meninggal akibat kegemukan atau obesitas (WHO, 2020a).
Lebih dari 340 juta anak dan remaja berusia 5-19 tahun mengalami
kegemukan atau obesitas pada tahun 2016 (WHO, 2020b). Prevalensi malnutrisi
pada anak masih tinggi di Indonesia (Hardiansyah et al., 2017). Berdasarkan hasil
Riskesdas tahun 2018, proporsi obesitas sentral pada penduduk usia ≥ 15 tahun di
Indonesia sebesar 31,0%. Anak berusia 5-12 tahun yang mengalami kegemukan
sebesar 10,8% dan obesitas sebesar 9,2%. Prevalensi kegemukan pada remaja juga
dilaporkan masih tinggi. Menurut data Riskesdas, sebanyak 11,2% remaja usia 13-15
tahun kegemukan dan 4,8% obesitas. Adapun prevalensi kegemukan dan obesitas
pada remaja usia 16-18 tahun masing-masing sebesar 9,5% dan 4,0% (Kemenkes RI,
2018).
Saat ini, persentase lemak tubuh telah dianggap sebagai standar akurat untuk
menentukan kegemukan atau obesitas karena dapat mengukur lemak tubuh secara
langsung. Penggunaan persentase lemak tubuh sebagai untuk mengukur kegemukan
atau obesitas semakin meningkat (Trang et al., 2019). Persen lemak tubuh dapat
dinilai dengan akurat menggunakan metode seperti Dual-energy X-ray absorptiometry
(DXA) (Lohman & Chen, 2005; Thomas et al., 2005). Namun, DXA tidak dapat
digunakan untuk pengukuran di lapangan dan lebih sulit untuk digunakan pada
orang berusia muda daripada orang dewasa karena serangkaian protokol yang harus
dipatuhi. Teknik dengan akurasi tinggi seperti DXA, air displacement plethysmography
(ADP), computed tomography, dan MRI membutuhkan biaya operasional dan pelatihan
yang mahal (dos Santos Cavalcanti et al., 2009). Oleh karena itu, metode alternatif
seperti indikator antropometri yang dapat membedakan lemak tubuh dengan biaya
operasional yang rendah diperlukan dalam praktek klinis (Silva et al., 2013).
Model prediksi persen lemak tubuh sering digunakan untuk menentukan
prevalensi obesitas pada populasi atau program manajemen obesitas. Hal ini
dikarenakan tidak memerlukan biaya yang besar dan relatif mudah digunakan
untuk populasi yang besar (Deurenberg et al., 2000).Variabel yang termasuk dalam
persamaan (model) prediksi untuk anak muda bervariasi, tetapi telah terbukti bahwa
skinfold layak untuk digunakan (Stevens et al., 2007; Stomfai et al., 2011). Menurut
Lee & Nieman (2010), pengukuran skinfold thickness memiliki koefisien korelasi
52 Nutri-Sains: Jurnal Gizi, Pangan dan Aplikasinya, Vol 4 No 1

DOI: 10.21580/ns.2020.4.1.4367
sebaik hasil pengukuran dengan menggunakan hydrostatic weighing jika pengukuran

dilakukan dengan benar. Hidrostatic weighing merupakan metode yang paling
banyak digunakan untuk menentukan komposisi tubuh di laboratorium.
Indeks massa tubuh (IMT) merupakan diskriminator lemak tubuh yang
sangat baik pada laki-laki dan perempuan. Indikator ini bisa digunakan oleh tenaga
kesehatan profesional untuk menilai lemak tubuh dengan cepat pada anak-anak dan
remaja dengan biaya operasional rendah (Junior et al., 2017). IMT memiliki korelasi
yang kuat dengan persen lemak tubuh yang diukur dengan menggunakan
densitometry, magnetic resonance imaging atau DXA. Berdasarkan penelitian Taylor et al.
(2000), Lingkar Pinggang (LP) memiliki korelasi yang kuat dengan jumlah lemak
tubuh bagian atas yang diukur dengan menggunakan DXA (r = 0,92).
Penggunaan model prediksi Slaughter untuk remaja di Indonesia
direkomendasikan oleh Isjwara et al. (2007) karena saat itu model prediksi tersebut
dianggap sebagai metode antropometri terbaik untuk memperkirakan persen lemak
tubuh sembari menunggu pengembangan model prediksi khusus untuk orang
Indonesia. Model prediksi ini sebenarnya berlaku untuk etnis Kaukasian, berjenis
kelamin laki-laki dan perempuan dengan rentang usia 8 sampai 29 tahun
(Slaughter et al., 1988). Model prediksi persen lemak tubuh yang ada di Indonesia
sampai saat ini adalah model prediksi untuk remaja putra (Sitoayu & Sudiarti, 2016),
sedangkan untuk remaja putri belum dikembangkan di Indonesia. Oleh karena
itu, perlu dikembangkan model prediksi untuk remaja putri di Indonesia.
Prevalensi kegemukan dan obesitas pada siswa-siswi MTs dan MA
Multiteknik Yayasan Asih Putera Kota Cimahi Tahun 2012 berdasarkan data Unit
Kesehatan Sekolah (UKS) bulan Februari sampai Maret masing-masing sebesar
18,8% dan 9,8%. Prevalensi ini sangat besar jika dibandingkan dengan prevalensi
kelebihan berat badan dan obesitas di Indonesia, Propinsi Jawa Barat dan Kota
Cimahi. Penelitian ini bertujuan untuk mendapatkan model prediksi yang memiliki
validitas optimal sebagai alternatif alat ukur persen lemak tubuh untuk menentukan
status gizi lebih pada siswi MTs dan MA Multiteknik Yayasan Asih Putera Kota
Cimahi.
METODE
Desain, Waktu, dan Tempat
Desain penelitian adalah Cross Sectional dengan pendekatan Kuantitatif
Observational. Penelitian dilakukan pada tanggal 19-30 November 2012 pada pukul
7.30 – 12.00 WIB. Adapun interpretasi terhadap model dan penulisan artikel
dilakukan pada tahun 2019.

53
Populasi dan Sampel
Sampel penelitian adalah siswi MTS dan MA Multiteknik Yayasan Asih
Putera tahun ajaran 2012/2013 yang berusia 12-18 tahun sebanyak 110 siswi dengan
menggunakan teknik Stratifikasi Proporsi. Penghitungan besar sampel sebelumnya
menggunakan uji hipotesis koefisien korelasi yang didasarkan kepada perhitungan
nilai transformasi Fisher (Ariawan, 1998). Karena jumlah sampel yang didapatkan
sangat kecil, maka perhitungan jumlah sampel ditentukan dengan menggunakan
ketentuan yang dikemukakan oleh Hastono (2007) yang mengemukakan bahwa
untuk analisis multivariat, untuk setiap variabel minimal memerlukan 10 responden.
Karena penelitian memiliki 9 variabel, maka jumlah sampel minimal adalah 90
variabel. Aktual subjek sebanyak 110 karena ada 20 siswi yang bersedia menjadi
responden tambahan.
Kriteria inklusi untuk sampel adalah siswi MTS dan MA tahun ajaran
2012/2013 yang aktif secara administrasi. Kriteria eksklusi untuk sampel adalah
siswi terlibat narkoba, tidak mematuhi peraturan sekolah, dan menderita penyakit
berat dan kronis. Sampel harus memenuhi syarat-syarat untuk pengukuran PLT
BIA, yaitu tidak makan dan minum, khususnya kopi empat jam sebelum
pengukuran, tidak boleh mengonsumsi alkohol 48 jam sebelum pengukuran,
menghindari aktivitas fisik yang berat 12 jam sebelum pengukuran, jika
memungkinkan, tidak mengonsumsi obat atau supplemen yang bersifat diuretik
tujuh hari sebelumnya, mengosongkan kandung kemih 30 menit sebelum
pengukuran, tidak menggunakan alas kaki ketika pengukuran (Bozkirli et al.,
2007), dan tidak sedang menstruasi (Heyward & Stolarczyk, 1996). Syarat-syarat
tersebut diinformasikan kepada siswi minimal satu hari sebelum pengukuran.
Siswi yang mengikuti pengukuran diberikan informed consent. Orang tua siswi juga
diinformasikan tentang keikutsertaan anaknya dalam penelitian.
Jenis dan Cara Pengambilan Data

Studi pendahuluan dilakukan untuk mengetahui prevalensi gizi lebih di MTs
dan MA Yayasan Asih Putera tahun ajaran 2012/2013. Pada penelitian utama,
pengukuran persen lemak tubuh dilakukan dengan menggunakan BIA merk Omron
HBF 302 sebagai gold standard. Pengukuran antropometri terdiri dari pengukuran
usia, berat badan dan tinggi badan (untuk mendapatkan nilai IMT (kg/m2) dan
IMT WHO Z Score), lingkar pinggang, dan skinfold thickness. Skinfold thickness
didapatkan dari penjumlahan tricep, bicep, subscapular, dan suprailiaca (Gibson, 2005;
Roche, Heymsfield, & Lohman, 1996). Pengukuran dilakukan sebanyak dua kali
dan diambil nilai rata-ratanya. Pengukuran berat badan dan tinggi badan masing-
masing dilakukan dengan menggunakan timbangan digital merk SECA dan
stadiometer. Pengukuran lingkar pinggang dilakukan dengan menggunakan pita meter
non-elastis merk Butterfly. Pengukuran skinfold thickness dilakukan dengan

DOI: 10.21580/ns.2020.4.1.4367
menggunakan skinfold caliper Harpenden, London. IMT WHO (Z Score) diukur

dengan menggunakan software WHO AnthroPlus.
Persen lemak tubuh diukur dengan menggunakan berbagai model prediksi,
yaitu Slaughter, Deurenberg, Lee, Chan (Tabel 1) dan hasil analisis multivariat. Hasil
pengukuran persen lemak tubuh model prediksi dibandingkan dengan BIA sebagai
gold standard. Cut off point kelebihan berat badan (overweight) berdasarkan cut off point
referensi. Cut off point IMT sebesar 23,00 kg/m2 (WHO Expert Consultation,
2004), IMT WHO (Z Score) sebesar 1 SD (WHO, 2013), skinfold thickness sebesar 82
mm (Nieman, 2003), dan LP sebesar 80 cm (Gibson, 2005). Semua model prediksi
mempunyai cut off point sebesar 32 % (Nieman, 2003).
Analisis Data
Analisis data yang dilakukan pada penelitian ini meliputi univariat, bivariat,
multivariat, dan uji asumsi model prediksi yang dihasilkan dari analisis multivariat.
Analisis bivariat dilakukan dengan menggunakan uji korelasi. Analisis multivariat
dilakukan dengan menggunakan uji regresi linier ganda. Uji ROC digunakan untuk
uji validasi, yaitu untuk mengetahui nilai Area Under Curve (AUC), sensitivitas,
spesifisitas, Predictive Value Positif (PVP), Predictive Value Negative (PVN), Likelihood
Ratio + atau (LR+), dan LR- (Park, Goo, & Jo, 2004). Uji asumsi model prediksi
terdiri dari uji asumsi eksistensi (variabel random), asumsi indepedensi, asumsi
linieritas, asumsi homoscedascity, asumsi normalitas, dan asumsi diagnostic multicollinearity
(Hastono, 2007). Uji statistik dilakukan secara parametrik dan kenormalan data
dilihat dari histogram variabel penelitian. Analisis data dilakukan dengan
menggunakan SPSS 16.
Tabel 1. Model prediksi

No. Peneliti Model Prediksi
1. Slaughter et al. (1988) P LT = 1,33 (tricep skinfold +
subscapular skinfold) –
0,013 (tricep skinfold + subscapular
skinfold) 2 – 2,5
2. Deurenberg et al. (1991) P LT = 1,20 IMT + 0,23 U – 10,8 JK

– 5,4
3. Lee et al. (2007) PLT = 7,596 + 0,060 IMT 2 – 0,460 U
+ 2,445 JK –
0,002 IMT 2 U JK
4. Chan et al. (2009) PLT = 28,725 + 0,089 ST
Keterangan :
P LT : Persen Lemak Tubuh JK = Jenis Kelamin U : Usia
ST : Skinfold Thickness ( jumlah tricep skinfold, bicep skinfold, subscapular skinfold,
suprailiaca skinfold)

55
HASIL DAN PEMBAHASAN
Jumlah responden dalam penelitian ini adalah 110 orang siswi dari MTs dan
MA Yayasan Asih Putera, yaitu sebanyak 63 orang dari MTs dan 47 orang dari MA.
Rerata usia responden adalah 14,85 tahun dengan rentang usia mulai dari 12 - 18
tahun (95% CI : 14,52 – 15,18). Standar deviasi usia adalah 1,75 tahun. Persentase
siswi MTS dan MA Yayasan Asih Putera yang mempunyai status gizi kurus dan
normal masing-masing sebesar 1,82% dan 78,18%. Persentase siswi MTS dan MA
Yayasan Asih Putera yang mempunyai mempunyai status gizi kegemukan (overweight)
dan obesitas masing-masing sebesar 11,82% dan 8,18% (Tabel 2).
Tabel 2. Status gizi responden berdasarkan IMT WHO (Z Score)

Status Gizi Jumlah Persentase (%)
Kurus 2 1,82
Normal 86 78,18
Kegemukan (overweight) 13 11,82
Obesitas 9 8,18
Total 110 100
Tabel 3. Hasil pengukuran persen lemak tubuh dan antropometri

Variabel Rerata Standar Deviasi (SD)
PLT BIA (%) 26,51 5,48
BB (kg) 48,76 10,66
TB (m) 152,26 5,37
IMT (kg/m2) 20,96 4,11
IMT WHO (Z Score) 0,13 1,16
Tricep (mm) 16,13 5,50
Bicep (mm) 8,94 3,82
Subscapular (mm) 12,66 3,60
Suprailiaca (mm) 16,77 6,30
ST (mm) 54,51 17,23
LP (cm) 67,46 8,53
PLT Slaughter (%) 24,08 4,27
PLT Deurenberg (%) 23,51 5,90
PLT Lee (%) 28,12 0,12
PLT Chan (%) 33,58 1,53
Keterangan : BB : Berat Badan TB : Tinggi Badan ST : Skinfold Thickness
Hasil pengukuran persen lemak tubuh dengan menggunakan BIA pada

responden menunjukkan bahwa rerata persen lemak tubuh secara keseluruhan
adalah 26,51%±5,48%. Nilai tersebut tidak terlalu jauh dengan hasil penelitian
Murbawani & Firiana (2017) yaitu sebesar 26,40%±9,95% untuk remaja usia 11-15
tahun. Rata-rata persen lemak tubuh remaja putri berusia 10-15 tahun di desa
23,14%±4,74% dan rata-rata persen lemak tubuh remaja di kota 24,00%±5,28%
(Handayani, Dwiriani, & Riyadi, 2013). Hasil penelitian Rahmayanti & Sudiarti
(2019) menunjukkan bahwa persen lemak tubuh remaja putri usia 15-17 tahun

DOI: 10.21580/ns.2020.4.1.4367
tergolong cukup atau good fair (19-28%). Ketika masa puber, remaja putri memiliki
lemak tubuh sebanyak 15,0% dan ketika kematangan fisiknya berakhir, maka
persentase lemak tubuhnya menjadi sebesar 28,0% (Krummel & Kris-Etherton,
1996).
Hasil uji korelasi hasil pengukuran antropometri dan PLT model prediksi
dengan PLT BIA terdapat pada Tabel 4. Variabel yang memiliki korelasi paling
kuat dengan PLT BIA adalah IMT, yaitu nilai r= 0,938. Hal ini sejalan dengan
penelitian yang dilakukan oleh Isjwara et al., (2007) terhadap 102 remaja perempuan
Indonesia yang berusia 11-15 tahun. Berdasarkan hasil penelitian tersebut,
didapatkan koefisien korelasi antara IMT dengan persen lemak tubuh BIA sebesar
0,782 dan dengan BIA Tanita sebesar 0,916.
Tabel 4. Koefisien korelasi antarvariabel penelitian

Variabel r p value
PLT BIA vs Usia 0,072 0,4570
PLT BIA vs IMT 0,938 0,0001
PLT BIA vs IMT WHO (Z Score) 0,925 0,0001
PLT BIA vs ST 0,878 0,0001
PLT BIA vs LP 0,895 0,0001
PLT BIA vs PLT Slaughter 0,853 0,0001
PLT BIA vs PLT Deurenberg 0,922 0,0001
PLT BIA vs PLT Lee 0,889 0,0001
PLT BIA vs PLT Chan 0,878 0,0001
Keterangan : PLT : Persen Lemak Tubuh ST : Skinfold Thickness
Variabel yang paling lemah korelasinya dengan PLT BIA adalah usia, yaitu
nilai r= 0,072. Semua variabel berkorelasi secara signifikan, kecuali usia. Hasil
penelitian ini tidak sejalan dengan teori yang dikemukakan oleh Daniels et al. (2000)
yang menyatakan bahwa terdapat penyimpanan lemak pada pusat tubuh yang relatif
lebih besar dengan semakin meningkatnya usia.
Hasil penelitian ini sejalan dengan penelitian Abraham & O’Dea (2001) yang
menyatakan bahwa remaja putri yang telah mengalami menarche secara signifikan
lebih mungkin untuk menurunkan berat badan dan melakukan pengaturan pola
makan (diet) jika dibandingkan dengan remaja putri yang belum menarche. Menurut
Santrock (2007), seiring dengan berlangsungnya perubahan di masa pubertas, remaja
putri sering merasa tidak puas dengan tubuhnya sehubungan dengan meningkatnya
jumlah lemak. Remaja putri memiliki motivasi untuk memiliki tubuh yang sangat
kurus. Keinginan ini dipengaruhi oleh kecenderungan untuk mengidentikkan tubuh
yang sangat kurus dengan kecantikan.
Hasil akhir dari analisis multivariat yang melibatkan variabel usia, IMT, dan
skinfold thickness dapat dilihat pada Tabel 5. IMT merupakan variabel yang dominan
berkorelasi dengan PLT BIA yang dikontrol oleh variabel skinfold thickness dan usia.

57
Tabel 5. Hasil akhir analisis multivariat
Variabel B Std. Error Beta T p value
Konstanta -1,703 1,760 -0,968 0,335
IMT 0,991 0,099 0,743 10,006 0,0005
ST 0,069 0,024 0,216 2,905 0,004
Usia 0,249 0,102 0,079 2,435 0,017
Keterangan : ST : Skinfold Thickness
Berdasarkan hasil analisis multivariat akhir, maka dapat dibentuk model

prediksi sebagai berikut :
PLT = 0,991 IMT + 0,069 ST + 0,249 U-1,703
Keterangan :
PLT : Persen Lemak Tubuh IMT : Indeks Massa tubuh
ST : Skinfold Thickness U : Usia
Berdasarkan Tabel 6. terlihat bahwa PLT BIA memiliki nilai perbedaan rerata
persen lemak tubuh tertinggi dengan PLT Chan, yaitu sebesar 7,07% (overestimed) dan
perbedaan terendah dengan PLT Khairani, yaitu sebesar 0,06%. PLT Deurenberg
memiliki nilai underestimated tertinggi terhadap PLT BIA, yaitu dengan perbedaan
sebesar -3,00%.
Tabel 6. Perbedaan persen lemak tubuh

Model Prediksi PLT (%) Perbedaan dengan PLT BIA (%)
Slaughter 24,08 -2,43
Deurenberg 23,51 -3,00
Lee 28,12 1,61
Chan 33,58 7,07
Khairani 26,57 0,01
Keterangan : PLT : Persen Lemak Tubuh p value = 0,0001
Hasil Uji Validasi

Area Under Curve (AUC) adalah ukuran keseluruhan kinerja tes diagnostik
yang diinterpretasikan sebagai rerata nilai sensitivitas untuk semua nilai spesifisitas
yang mungkin. Hasil uji validasi pengukuran antropometri dan PLT model prediksi
dengan cut off point referensi dan hasil analisis ROC masing-masing dapat dilihat pada
Tabel 7 dan Tabel 8.
Cut off point kelebihan berat badan (overweight) pada Tabel 7 berdasarkan cut
off point referensi. Cut off point IMT sebesar 23,00 kg/m2 (WHO Expert Consultation,
2004), IMT WHO (Z Score) sebesar 1 SD (WHO, 2013), skinfold thickness sebesar 82

DOI: 10.21580/ns.2020.4.1.4367
mm (Nieman, 2003), dan LP sebesar 80 cm (Gibson, 2005). Semua model

prediksi mempunyai cut off point sebesar 32% (Nieman, 2003).
Hasil uji validasi dapat dilihat dari nilai r, AUC, sensitivitas, spesifisitas, PVP,
PVN, LR+, dan LR-. Hasil uji validasi yang menggunakan cut off point referensi
dibandingkan dengan cut off point hasil analisis ROC.
Berdasarkan nilai koefisien r yang terdapat pada Tabel 7 dan Tabel 8, semua
hasil pengukuran antropometri dan model prediksi memiliki nilai r yang besar yang
menandakan hubungan yang sangat kuat dengan PLT BIA. Nilai tertinggi dimiliki
oleh PLT Khairani sebesar 0,944, sedangkan yang terendah dimiliki oleh PLT
Slaughter sebesar 0,853.
Berdasarkan nilai AUC yang terdapat pada Tabel 7 dan Tabel 8, semua hasil
pengukuran antropometri dan model prediksi memiliki nilai AUC di atas garis
diagonal dan memiliki nilai AUC yang bagus. Nilai AUC tertinggi dimiliki oleh IMT
dan PLT Lee sebesar 1,000, sedangkan nilai terendah dimiliki oleh PLT Slaughter
sebesar 0,957. PLT Khairani memiliki nilai AUC yang bagus, walaupun masih di
bawah IMT (kg/m2) dan PLT Lee.
Tabel 7. Hasil uji validasi dengan cut off point referensi

Cut
Variabel Nilai r AUC Se (%) Sp (%) PVP (%) PVN (%) LR+ LR-
Off
IMT (kg/m2) 0,938 1,000 23,00 100 92,47 70,83 100 13,28 0,00
IMT WHO 0,925 0,994 1,00 88,24 92,47 68,18 97,73 11,72 0,13
(Z Score)
ST (mm) 0,878 0,993 82,00 58,82 100 100 93,00 ∞ 0,41
LP (cm) 0,895 0,990 80,00 47,05 100 100 91,18 ∞ 0,53
PLT Khairani 0,944 0,989 32,00 82,35 100 100 96,88 ∞ 0,18
PLT Slaughter 0,853 0,957 32,00 0,00 100 0,00 84,55 0,00 1,00
PLT 0,922 0,999 32,00 58,82 100 100 93,00 ∞ 0,41
Deurenberg
PLT Lee 0,889 1,000 32,00 94,12 91,40 66,67 98,83 10,94 0,06
PLT Chan 0,878 0,993 32,00 100 12,90 17,35 100 1,15 0,00
Keterangan : Se : Sensitivitas Sp : Spesifisitas ST : Skinfold Thickness p value = 0,0005
Berdasarkan nilai Se yang terdapat pada Tabel 7, nilai Se tertinggi dimiliki oleh
IMT dan PLT Chan, yaitu sebesar 100,00%, sedangkan nilai terendah dimiliki oleh
PLT Slaughter sebesar 0,00%. Berdasarkan nilai Sp, nilai tertinggi dimiliki oleh ST,
LP, PLT Khairani, PLT Slaughter, dan PLT Deurenberg sebesar 100,00%,
sedangkan nilai terendah dimiliki oleh PLT Chan sebesar 12,90%. Berdasarkan nilai
PVP, nilai tertinggi dimiliki oleh ST, LP, PLT Khairani, dan PLT Deurenberg
sebesar 100,00%, sedangkan nilai terendah dimiliki oleh PLT Slaughter sebesar
0,00%.
Berdasarkan nilai PVN, nilai PVN tertinggi dimiliki oleh IMT (kg/m2) dan
PLT Chan sebesar 100,00%, sedangkan nilai terendah dimiliki oleh PLT Slaughter

59
sebesar 84,55%. Berdasarkan nilai LR+, nilai tertinggi dimiliki oleh ST, LP, PLT
Khairani, dan PLT Deurenberg sebesar ∞, sedangkan nilai terendah dimiliki oleh
PLT Chan sebesar 1,15. Berdasarkan nilai LR-, nilai tertinggi dimiliki oleh PLT
Slaughter sebesar 1,00 sedangkan nilai terendah dimiliki oleh IMT (kg/m2) dan PLT
Chan sebesar 0,00.
Hasil uji validasi pengukuran antropometri dan PLT model prediksi dengan
cut off point referensi jika dilihat dari nilai sensitivitas, spesifisitas, PVP, PVN, LR+,
dan LR-nya, maka pengukuran antropometri atau model prediksi yang memiliki
validitas tertinggi dari peringkat tertinggi sampai terendah adalah IMT (kg/m2),
PLT Lee, PLT Khairani, IMT WHO (Z Score).
Tabel 8. Hasil uji validasi dengan cut off point hasil analisis ROC
Cut Se Sp PVP PVN
Variabel Nilai r AUC LR+ LR-
Off (%) (%) (%) (%)
IMT (kg/m2) 0,938 1,000 24,93 88,24 98,92 93,75 97,87 81,70 0,12
IMT WHO (Z 0,925 0,994 1,78 70,59 100 100 94,90 ∞ 0,29
Score)
ST (mm) 0,878 0,993 72,33 76,47 100 100 95,88 ∞ 0,24
LP (cm) 0,895 0,990 74,50 76,47 98,92 92,86 95,83 70,80 0,24
PLT Khairani 0,944 0,989 32,00 82,35 100 100 96,88 ∞ 0,18
PLT Slaughter 0,853 0,957 30,19 64,70 100 100 93,94 ∞ 0,35
PLT 0,922 0,999 30,49 76,47 100 100 95,88 ∞ 0,24
Deurenberg
PLT Lee 0,889 1,000 40,53 76,47 100 100 95,88 ∞ 0,24
PLT Chan 0,878 0,993 35,16 76,47 100 100 95,88 ∞ 0,24
Keterangan : Se : Sensitivitas Sp : Spesifisitas ST : Skinfold Thickness p value = 0,0005
Berdasarkan nilai Se yang terdapat pada Tabel 8, nilai Se tertinggi dimiliki

oleh IMT sebesar 88,24%, sedangkan nilai terendah dimiliki oleh PLT Slaughter
sebesar 64,70%. Berdasarkan nilai Sp, semua pengukuran antropometri dan model
prediksi bernilai 100,00%, kecuali IMT (kg/m2) dan LP. Berdasarkan nilai PVP,
semua pengukuran antropometri dan model prediksi bernilai 100,00%, kecuali IMT
(kg/m2) dan LP.
Berdasarkan nilai PVN, nilai PVN tertinggi dimiliki oleh IMT (kg/m2)
sebesar 97,87%, sedangkan nilai terendah dimiliki oleh PLT Slaughter sebesar
93,94%. Berdasarkan nilai LR+, semua pengukuran antropometri dan model
prediksi bernilai ∞, kecuali IMT (kg/m2) dan LP. Berdasarkan nilai LR-, nilai
tertinggi dimiliki oleh PLT Slaughter sebesar 0,35, sedangkan nilai terendah dimiliki
oleh IMT (kg/m2) sebesar 0,12.
Hasil uji validasi pengukuran antropometri dan PLT model prediksi dengan
cut off point hasil analisis ROC Jika dilihat dari nilai sensitivitas, spesifisitas, PVP,
PVN, LR+, dan LR-nya, maka pengukuran antropometri atau model prediksi yang
memiliki validitas tertinggi adalah IMT (kg/m2) dan PLT Khairani. Skinfold
thickness berada pada peringkat kedua.

DOI: 10.21580/ns.2020.4.1.4367
Berdasarkan hasil uji validasi, model prediksi yang memiliki validitas optimal
adalah model prediksi Khairani jika hanya dibandingkan dengan model prediksi
lainnya. Model prediksi Khairani memiliki nilai sensitivitas, spesifisitas, PVP, PVN,
dan LR+ yang selalu tinggi dan stabil jika menggunakan cut off point referensi
maupun hasil analisis ROC. Nilai r dan AUC-nya juga tinggi dan nilai LR-
rendah. Tidak semua hasil uji validasinya bernilai paling tinggi, namun secara
keseluruhan selalu tinggi dan stabil jika dibandingkan dengan model prediksi lainnya.
Menurut Isjiswara et al. (2007), model prediksi Slaughter dapat dijadikan
sebagai model prediksi untuk menentukan persen lemak tubuh pada remaja putri di
Indonesia. Berdasarkan hasil penelitian ini, sensitivitas, dan PVP PLT Slaughter
bernilai 0,00% jika menggunakan cut off point referensi dan bernilai 64,70% dan
100% jika menggunakan cut off point hasil analisis ROC. Nilai ini menunjukkan bahwa
validitas model prediksi Slaughter kurang baik.
Model prediksi Khairani dapat digunakan sebagai pengganti model prediksi
Slaughter khusus untuk siswi pada sekolah yang memiliki karakteristik yang sama
dengan MTs dan MA Multiteknik Yayasan Asih Putera. Sebelum digunakan untuk
populasi lain yang berbeda karakteristiknya, model prediksi ini harus diuji validasi
terlebih dahulu pada populasi tersebut.
Berdasarkan hasil uji validasi secara keseluruhan, IMT merupakan alat ukur
yang memiliki validitas yang optimal dan juga memenuhi syarat sebagai alat ukur
untuk menentukan status gizi lebih di lapangan. Hasil ini sejalan dengan hasil
penelitian Junior et al. (2017) yang mengungkapkan bahwa IMT adalah penanda
lemak tubuh yang sangat baik pada anak dan remaja, baik jenis kelamin laki-laki dan
perempuan. Indikator ini bisa digunakan oleh tenaga kesehatan profesional untuk
mengukur lemak tubuh dengan cepat pada anak-anak dan remaja dengan biaya
operasional rendah.
IMT adalah proxy yang paling banyak digunakan untuk menilai kegemukan
dan obesitas anak-anak dan orang dewasa. Walaupun IMT mudah diukur dan
berbiaya rendah, IMT tidak membedakan antara lemak dan jaringan massa tanpa
lemak dan dengan demikian dapat menyebabkan kesalahan klasifikasi (Dulloo et al.,
2010). Penelitian Costa-Urrutia et al. (2019) memvalidasi IMT sebagai prediktor
persen lemak tubuh pada anak sekolah dan remaja (6-17 tahun) karena hubungan
linier dan nilai AUC yang tinggi, serta nilai sensitivitas yang ditunjukkannya.

61
KESIMPULAN
Persentase siswi MTS dan MA Yayasan Asih Putera yang mempunyai status
gizi kurus dan normal masing-masing sebesar 1,82% dan 78,18%. Persentase siswi
MTS dan MA Yayasan Asih Putera yang mempunyai status gizi kegemukan
(overweight) dan obesitas masing-masing sebesar 11,82 % dan 8,18 %. Rerata persen
lemak tubuh siswi adalah 26,51% ± 5,48%. Model prediksi yang dihasilkan dari hasil
pengukuran antropometri adalah PLT = 0,991 IMT + 0,069 ST + 0,249 U -1,703.
Model prediksi tersebut memiliki validitas yang optimal jika dibandingkan dengan
model prediksi lainnya. Untuk studi validasi selanjutnya dapat menggunakan metode
Bland Altman dan pengukuran persen lemak tubuh dengan menggunakan DXA
sebagai gold standard. Status pubertas dapat memengaruhi persentase lemak tubuh
remaja.
UCAPAN TERIMA KASIH

Terima kasih diucapkan kepada STIKES Tri Mandiri Sakti Bengkulu,
Departmen Gizi Kesehatan Masyarakat Universitas Indonesia, dan MTs dan MA
Yayasan Asih Putera Kota Cimahi Propinsi Jawa Barat.
DAFTAR PUSTAKA
Abraham, S., & O’Dea, J. A. (2001). Body mass index, menarche, and perception of
dieting among peripubertal adolescent females. International Journal of Eating
Disorders, 29(1), 23–28. https://doi.org/10.1002/1098-
108X(200101)29:1<23::AID-EAT4>3.0.CO;2-Z
Ariawan, I. (1998). Besar dan metode sampel pada penelitian kesehatan. Jakarta: Fakultas
Kesehatan Masyarakat, universitas Indonesia.
Bozkirli, E., Ertorer, M. E., Bakiner, O., Tutuncu, N. B., & Demirag, N. G. (2007).
The validity of the World Health Organisation’s obesity body mass index
criteria in a Turkish population: A hospital-based study. Asia Pacific Journal of
Clinical Nutrition, 16(3), 443–447.
Chan, D., Li, A., So, H., Yin, J., & Nelson, E. (2009). New skinfold-thickness
equation for predicting percentage body fat in Chinese obese children. HK J
Paediatr, 14(96–102).
Costa-Urrutia, P., Vizuet-Gámez, A., Ramirez-Alcántara, M., Guillen-González, M.
Á., Medina-Contreras, O., Valdes-Moreno, M., … Rodriguez-Arellano, M. E.
(2019). Obesity measured as percent body fat, relationship with body mass
index, and percentile curves for Mexican pediatric population. PLOS ONE,
14(2), e0212792. https://doi.org/10.1371/journal.pone.0212792
Daniels, S. R., Khoury, P. R., & Morrison, J. A. (2000). Utility of different measures
of body fat distribution in children and adolescents. American Journal of
Epidemiology, 152(12), 1179–1184. https://doi.org/10.1093/aje/152.12.1179
Deurenberg, P., Deurenberg-Yap, M., Wang, J., Lin, F. P., & Schmidt, G. (2000).
Prediction of percentage body fat from anthropometry and bioelectrical

DOI: 10.21580/ns.2020.4.1.4367
impedance in Singaporean and Beijing Chinese. Asia Pacific Journal of Clinical

Nutrition, 9(2), 93–98. https://doi.org/10.1046/j.1440-6047.2000.00149.x
Deurenberg, P., Weststrate, J. A., & Seidell, J. C. (1991). Body mass index as a
measure of body fatness: Age- and sex-specific prediction formulas. British
Journal of Nutrition, 65(2), 105–114. https://doi.org/10.1079/BJN19910073
dos Santos Cavalcanti, C., do Egito Carvalho, S., & de Barros, M. (2009).
Indicadores antropométricos de obesidade abdominal: Revisão dos artigos
indexados na biblioteca SciELO. Braz J Kinathrop Hum Perform, 11(2).
https://doi.org/10.1590/1980-0037.2009v11n2p217
Dulloo, A. G., Jacquet, J., Solinas, G., Montani, J.-P., & Schutz, Y. (2010). Body
composition phenotypes in pathways to obesity and the metabolic syndrome.
International Journal of Obesity, 34(S2), S4–S17.
https://doi.org/10.1038/ijo.2010.234
Gibson, R. (2005). Nutritional assessmen (2nd ed.). USA: Oxford University Press.
Handayani, M. S., Dwiriani, C. M., & Riyadi, H. (2013). Hubungan komposisi tubuh
dan status gizi dengan perkembangan seksual pada remaja putri di perkotaan
dan perdesaan. Jurnal Gizi Dan Pangan, 8(3), 181–186.
https://doi.org/10.25182/jgp.2013.8.3.181-186
Hardiansyah, A., Hardinsyah, & Sukandar, D. (2017). Kesesuaian Konsumsi Pangan
Anak Indonesia dengan Pedoman Gizi Seimbang. Nutri-Sains : Jurnal Gizi,
Pangan dan Aplikasinya, 1 (2), 35-45. https://doi.org/10.21580/ns.2017.1.2.2452.
Hastono, S. (2007). Analisis data kesehatan. Depok: FKM UI.
Heyward, V. H., & Stolarczyk, L. M. (1996). Applied body composition assassment. USA:
Human Kinetics.
Isjwara, R. I., Lukito, W., & Schultink, J. W. (2007). Comparison of body
compositional indices assessed by underwater weighing, bioelectrical
impedance and anthropometry in Indonesian adolescent girls. Asia Pacific
Journal of Clinical Nutrition, 16(4), 641–648.
Junior, C. A., Mocellin, M. C., Gonçalves, E. C. A., Silva, D. A., & Trindade, E. B.
(2017). Anthropometric Indicators as Body Fat Discriminators in Children and
Adolescents: A Systematic Review and Meta-Analysis. Advances in Nutrition: An
International Review Journal, 8(5), 718–727.
https://doi.org/10.3945/an.117.015446
Kemenkes RI. (2018). Hasil Utama Riskesdas 2018. Jakarta (ID): Kemenkes RI.
Krummel, D., & Kris-Etherton, P. (1996). Nutrition in women’s health. Maryland: An
Aspen Publication Inc.
Lee, K., Lee, S., Kim, S. Y., Kim, S. J., & Kim, Y. J. (2007). Percent body fat cutoff
values for classifying overweight and obesity recommended by the
International Obesity Task Force (IOTF) in Korean children. Asia Pacific
Journal of Clinical Nutrition, 16(4), 649–655.
Lee, R., & Nieman, D. (2010). Nutritional assessment. New York: McGraw Hill.
Lohman, T., & Chen, Z. (2005). Dual-energy X-ray absorptiometry. In S.
Heymsfield, T. Lohman, & et al (Eds.), Human Body Composition (2nd ed.).
Champaign: Human Kinetics.
Murbawani, E. A., & Firiana, L. (2017). Hubungan persen lemak tubuh dan aktifitas

63
fisik dengan tingkat kesegaran jasmani remaja putri. JNH (Journal of Nutrition
and Health), 5(2), 69–84. https://doi.org/10.14710/jnh.5.2.2017.69-84
Nieman, D. (2003). Exercise testing and prescription : A health-related approach. Boston:
McGraw-Hill Higher Education.
Park, S. H., Goo, J. M., & Jo, C.-H. (2004). Receiver Operating Characteristic
(ROC) curve: Practical review for radiologists. Korean Journal of Radiology, 5(1),
11–18. https://doi.org/10.3348/kjr.2004.5.1.11.
Rahmayanti, K., & Sudiarti, T. (2019). Hubungan tinggi badan dan persen lemak
tubuh dengan kebugaran muskoskeletal pada remaja. Nutri-Sains : Jurnal Gizi,
Pangan dan Aplikasinya, 3 (1), 30-42.
https://doi.org/ 10.21580/ns.2019.3.1.3326.
Roche, A., Heymsfield, S., & Lohman, T. (1996). Human body composition. Champaign:
Human Kinetics.
Santrock, J. (2007). Adolescent (11th ed.). New York: Mc Graw Hill Company.
Silva, D. R. P., Ribeiro, A. S., Pavão, F. H., Ronque, E. R. V., Avelar, A., Silva, A.
M., & Cyrino, E. S. (2013). Validade dos métodos para avaliação da gordura
corporal em crianças e adolescentes por meio de modelos
multicompartimentais: Uma revisão sistemática. Revista Da Associação Médica
Brasileira, 59(5), 475–486. https://doi.org/10.1016/j.ramb.2013.03.006
Sitoayu, L., & Sudiarti, T. (2016). studi validasi pengukuran antropometri dan model
prediksi terhadap persen lemak tubuh bia pada siswa MTs dan MA multiteknik
yayasan Asih putera Cimahi tahun 2012. Forum Ilmiah, 13(2), 67–75.
Slaughter, M. H., Lohman, T. G., Boileau, R. A., Horswill, C. A., Stillman, R. J., Van
Loan, M. D., & Bemben, D. A. (1988). Skinfold equations for estimation of
body fatness in children and youth. Human Biology, 60(5), 709–723.
Stevens, J., Taber, D. R., Murray, D. M., & Ward, D. S. (2007). Advances and
controversies in the design of obesity prevention trials. Obesity, 15(9), 2163–
2170. https://doi.org/10.1038/oby.2007.257
Stomfai, S., Ahrens, W., Bammann, K., Kovács, É., Mårild, S., Michels, N., …
Molnár, D. (2011). Intra- and inter-observer reliability in anthropometric
measurements in children. International Journal of Obesity, 35(S1), S45–S51.
https://doi.org/10.1038/ijo.2011.34
Taylor, R. W., Jones, I. E., Williams, S. M., & Goulding, A. (2000). Evaluation of
waist circumference, waist-to-hip ratio, and the conicity index as screening
tools for high trunk fat mass, as measured by dual-energy X-ray
absorptiometry, in children aged 3–19 y. The American Journal of Clinical
Nutrition, 72(2), 490–495. https://doi.org/10.1093/ajcn/72.2.490
Thomas, S. R., Kalkwarf, H. J., Buckley, D. D., & Heubi, J. E. (2005). Effective dose
of dual-energy x-ray absorptiometry scans in children as a function of age.
Journal of Clinical Densitometry, 8(4), 415–422.
https://doi.org/10.1385/JCD:8:4:415
Trang, L. T., Trung, N. N., Chu, D.-T., & Hanh, N. T. H. (2019). Percentage body
fat is as a good indicator for determining adolescents who are overweight or
obese: A cross-sectional study in Vietnam. Osong Public Health and Research
Perspectives, 10(2), 108–114. https://doi.org/10.24171/j.phrp.2019.10.2.10

DOI: 10.21580/ns.2020.4.1.4367
WHO. (2020a). Noncommunicable diseases: Childhood overweight and obesity. Geneva.

WHO. (2020b). Obesity and overweight. Geneva.
WHO. (2016). Childhood overweight and obesity. Geneva.
WHO. (2013). Growth references 5-19 years : BMI – for - age (5 – 19 years).
WHO Expert Consultation. (2004). Appropriate body-mass index for Asian
populations and its implications for policy and intervention strategies. The
Lancet, 363(9403), 157–163. https://doi.org/10.1016/S0140-6736(03)15268-3

65
Halaman ini sengaja dikosongkan.
View publication stats

Studi Validasi Pengukuran Antropometri dan Model Prediksi Terhadap Persen Lemak Tubuh BIA pada Siswa
MTS dan MA Multiteknik Yayasan Asih Putera Cimahi Tahun 2012
STUDI VALIDASI PENGUKURAN ANTROPOMETRI DAN

MODEL PREDIKSI TERHADAP PERSEN LEMAK TUBUH
BIA PADA SISWA MTS DAN MA MULTITEKNIK YAYASAN
ASIH PUTERA CIMAHI TAHUN 2012
Laras Sitoayu1, Trini Sudiarti1
1
Departemen Gizi Kesehatan Masyarakat, Universitas Indonesia
Universitas Indonesia, Depok 16424, Indonesia
laraz_one@yahoo.co.id
Abstract
The primary purpose of this cross-sectional study to develop percentage body fat
prediction model for boys aged 12-19 years (n = 111), by conducting a validation
study anthropometric measurements and predictions model of body fat percentage
(Slaughter, Deurenberg, Lee and Chan) to percent body fat BIA. In this study also
examined the correlation between anthropometric measurements (WHO BMI,
skinfold thickness and waist circumference) and predictions model (Slaughter,
Deurenberg, Lee and Chan) with percent body fat BIA with stratified proportional
design. The research was carried out on students MTs and MA. Multiteknik
Yayasan Asih Putera grade 7-12. Bivariat analysis showed BMI WHO has the
strongest correlation with percent body fat BIA (r = 0.804) between the
anthropometric measurements were used. The Prediction model IMT WHO also
has the best sensitivity (94%), the second is IMT WHO and Age (94%) and the
third is Sitoayu. All variables have a significant correlation with percent body fat
BIA (p < 0,0005). Multiregresi analysis results indicate that the dominant variable
is the WHO BMI, skinfold thickness and age with the predictions model of
percent body fat Sitoayu = 23,28 + 1.56 *BMI WHO + 0.13*ST - 0.62*Age. This
prediction model has AUC 0,937 and 84% sensitivity value.
Keywords : model prediction of percent body fat, validation study, boys
Abstrak
Penelitian ini merupakan cross sectional yang bertujuan membuat model prediksi
persen lemak tubuh untuk remaja laki-laki usia 12-19 tahun (n = 111), dengan
melakukan studi validasi pengukuran antropometri dan model Prediksi (Slaughter,
Deurenberg, Lee dan Chan) terhadap persen lemak tubuh BIA. Pada penelitian ini
juga menjelaskan korelasi antara pengukuran antropometri (IMT WHO, skinfold
thickness dan lingkar pinggang) serta model prediksi (Slaughter, Deurenberg, Lee
dan Chan) dengan persen lemak tubuh BIA. Pengambilan sampel dilakukan
dengan teknik stratified proportional. Penelitian ini dilakukan pada siswa MTs
dan MA. Multiteknik Yayasan Asih Putera kelas 7-12. Hasil penelitian
menunjukkan IMT WHO memiliki korelasi paling kuat dengan persen lemak
tubuh BIA (r = 0,804) diantara pengukuran antropometri yang digunakan. Model
Prediksi IMT WHO memilki sensitivitas paling tinggi yaitu 94%, diikuti dengan
model prediksi IMT WHO dan umur (94%) dan model prediksi Sitoayu. Seluruh
variabel memiliki korelasi yang signifikan dengan persen lemak tubuh BIA (p <
0,0005). Hasil analisis multiregresi menunjukkan variabel yang dominan adalah
IMT WHO, skinfold thickness, dan umur dengan model prediksi persen lemak
Forum Ilmiah Volume 13 Nomor 2, Mei 2016 64

tubuh baru (Sitoayu) = 23,28 + 1,56*IMT WHO + 0,13*ST - 0,62*U. Model

prediksi ini memiliki AUC 0,937 dan nilai sensitivitas yaitu 84%.
Kata kunci : model prediksi lemak tubuh, validasi penelitian, anak laki-laki
Pendahuluan Banyak metode yang digunakan

Penanggulangan terhadap gizi untuk mengukur komposisi tubuh namun
lebih dapat dilakukan dengan baik bila, harganya sangat mahal dan tidak dapat
dalam penentuan status gizi dilakukan digunakan untuk populasi yang besar (Mei,
dengan menggunakan pengukuran yang Z:2007). BIA merupakan metode yang
tepat, yang dapat menggambarkan keadaan direkomendasikan untuk mengukur persen
gizi sesungguhnya terutama untuk lemak tubuh pada anak dengan populasi
mengetahui besarnya lemak tubuh. yang besar. Pengukuran lemak tubuh
Peningkatan kejadian gizi lebih yang menggunakan BIA sangat baik karena
sangat cepat secara global, merupakan tidak melukai bagian tubuh, lebih cepat
suatu ancaman bagi kesehatan masyarakat digunakan, lebih mudah, lebih sederhana
(Bozkirli: 2007). Peningkatan jumlah gizi dan murah dibandingkan alat ukur lainnya.
lebih pada remaja perlu mendapatkan Pengukuran lain yang direkomendasikan
perhatian. Hal ini dikarenakan akan untuk mengukur persen lemak tubuh
berdampak pada kesehatan di masa dewasa adalah berat badan menurut tinggi badan
dan tua. Di tengah epidemi gizi lebih atau IMT sering digunakan sebagai alat
seperti ini dibutuhkan informasi mengenai untuk menentukan status gizi karena
pengukuran komposisi tubuh terutama sederhana dan murah untuk digunakan
persen lemak tubuh yang tepat. serta praktis untuk dibongkar dan dipasang
Pengukuran persen lemak tubuh yang tidak di lapangan (Gibson: 2005; Mei:2002).
tepat akan menyebabkan tidak Lemak tubuh juga dapat diukur dengan
terdeteksinya jumlah persen lemak tubuh menggunakan skinfold thickness. Skinfold
yang sesungguhnya dengan akurat thickness memiliki korelasi kuat dengan
terutama dampak negatif yang ditimbulkan berbagai macam hasil pemeriksaan
jika berlebih di dalam tubuh laboratorium terhadap lemak tubuh(3).
(Sopher:2004). Lingkar pinggang juga merupakan salah
Di Indonesia, informasi mengenai satu pengukuran yang dapat digunakan
kevalidan alat ukur antropometri untuk untuk mengukur total lemak tubuh dan
mengukur persen lemak tubuh remaja laki- memiliki korelasi yang kuat dibandingkan
laki masih sangat jarang. Akibatnya total lemak tubuh yang diukur dengan
penentuan status gizi lebih pada remaja densitometry (Gibson: 2005; Gillum:
laki-laki menjadi tidak valid, tidak ada 1999).
kepastian alat ukur yang baik untuk Yayasan Asih Putera merupakan
mengukur status gizi lebih pada remaja salah satu sekolah swasta favorit di Kota
dengan jumlah responden yang banyak, Cimahi Jawa Barat, sebagian besar siswa
terutama untuk persen lemak tubuh yang Madrasah Tsanawiyah (MTs) dan
dapat digunakan sebagai pencegahan Madrasah Aliyah (MA) Multiteknik
(preventif) pada masyarakat dan hal yang Yayasan Asih Putera berasal dari keluarga
terpenting adalah terjadi kesalahan dalam ekonomi menengah ke atas. Berdasarkan
intervensi terhadap responden karena data Unit Kesehatan Sekolah (UKS) MTs.
ketidakvalidan alat ukur dalam Yayasan Asih Putera kelas 7-9 pada bulan
menentukan status gizi lebih. Februari-Maret 2012 terdapat 18,8% siswa
mengalami gizi lebih, sedangkan pada n= +3

MA. Multiteknik Yayasan Asih Putera
terdapat 9,8% siswa mengalami gizi lebih. Analisis data menggunakan perangkat
Angka ini cukup besar bila dibandingkan lunak program statistik dan program WHO
dengan data evaluasi penjaringan tahun AnthroPlus.
2011 Kota Cimahi pada siswa MTs/SMP
dan MA/SMA yang hanya sebesar 0,8% Hasil dan Pembahasan
pada siswa MTs/SMP dan 0,1% pada Karakteristik Responden
siswa MA/SMA. Penelitian ini dilakukan Responden dalam penelitian ini
untuk mengetahui alat ukur antropometri berjumlah 112 siswa, namun setelah
(IMT WHO, skinfold thickness, dan dikeluarkan outlier, menjadi 111 siswa
lingkar pinggang) dan model prediksi yang terdiri dari 73 siswa MTs (65,77%)
(Slaughter, Deurenberg, Lee dan Chan) dan 38 siswa MA Multiteknik (34,23%)
yang memiliki validitas optimal jika mulai dari kelas 7 hingga kelas 12. Umur
dibandingkan dengan persen lemak tubuh responden yang mengikuti penelitian ini
BIA sebagai Golden Standard, untuk memiliki rentang mulai dari 12 tahun
menapis gizi lebih pada remaja laki-laki. sampai 19 tahun, dengan umur terbanyak
adalah 14 tahun dan umur yang paling
Metode Penelitian sedikit adalah 19 tahun. Jika status gizi
Penelitian ini merupakan berdasarkan IMT WHO (Z-Score)
penelitian kuantitatif observasional yang responden dikelompokkan maka diperoleh
menggunakan data primer yaitu hasil sebanyak 15 siswa (13,50%) gemuk
membandingkan pengukuran antropometri dan 17 siswa (15,30%) obesitas. Siswa
(IMT WHO, skinfold thickness, dan dengan status gizi normal sebanyak 75
lingkar pinggang) dan persen lemak tubuh siswa (67,60%).
model prediksi (Slaughter, Deurenberg,
Lee dan Chan) dengan hasil pengukuran Hasil Pengukuran Persen Lemak Tubuh
persen lemak tubuh BIA. Rancangan dan Antropometri pada Siswa MTS. dan
penelitian ini menggunakan desain cross MA Yayasan Asih Putera
sectiona. Penelitian ini dilakukan di MTs. Persen lemak tubuh responden
dan MA. Multiteknik Yayasan Asih yang diukur menggunakan BIA dijadikan
Putera, Kota Cimahi. Waktu penelitian sebagai Golden Standard. Berdasarkan
pada tanggal 19-30 November 2012. hasil pengukuran persen lemak tubuh BIA
Populasi pada penelitian ini adalah responden, persen lemak tubuh responden
semua siswa MTs. dan MA. Multiteknik cukup bervariasi, mulai 8,05 sampai
Yayasan Asih Putera tahun ajaran 36,15%. Persen lemak tubuh yang besar
2012/2013 kelas 7-12. Responden pada ternyata tidak hanya dimiliki oleh
penelitian ini memenuhi kriteria inklusi responden MA. Multiteknik tetapi
antara lain siswa MTs. dan MA. sebagian besar dimiliki oleh responden
Multiteknik Yayasan Asih Putera tahun MTs. Yayasan Asih Putera. Berdasarkan
ajaran 2012/2013, bersedia untuk hasil pengamatan yang dilakukan selama
berpartisipasi dalam penelitian ini dan aktif pengambilan data berlangsung, kegiatan
secara administratif. Pengambilan sampel yang dilakukan oleh siswa MA
dilakukan dengan teknik stratified Multiteknik lebih banyak dibandingkan
propotional. Besar sampel ditentukan siswa MTs. Lebih tingginya aktifitas fisik
dengan rumus : dapat mempengaruhi besarnya persen
lemak tubuh yang dimiliki oleh seseorang akibat dari meningkatnya massa otot,
(4)
. Remaja laki-laki, berat badan tanpa sehingga persen lemak tubuh semakin
lemak akan meningkat dari 80% menjadi sedikit (Santrock :2007; Rogol:2002).
85-90% pada saat maturasi. Keadaan ini
Tabel 1
Rata-rata Hasil Pengukuran Responden
Pada penelitian ini, status gizi merupakan salah satu metode pengukuran
responden diukur dengan menggunakan persen lemak tubuh terbaik(Isjwara:2007).
IMT WHO (Z-score). Berdasarkan nilai Z Berdasarkan hasil observasi pada
score yang dimiliki responden berkisar responden, banyak responden yang
antara sangat kurus hingga obesitas. Status memiliki lingkar pinggang lebih dari yang
gizi sangat kurus banyak dimiliki siswa direkomendasikan IOTF 90 cm, yaitu
MA Multiteknik dibandingkan siswa MTs. sekitar 12 siswa, artinya memiliki resiko
Yayasan Asih Putera. untuk menderita penyakit degeneratif yang
Nilai Z score dengan klasifikasi diakibatkan oleh gizi lebih atau obesitas.
sangat kurus dimiliki oleh siswa MA, hal Besarnya lingkar pinggang memiliki
ini dikarenakan siswa tersebut memiliki korelasi yang kuat terhadap distribusi
berat badan yang kurang sekitar 50 kg dan lemak tubuh (Umi,F:2007).
tinggi badan 180 cm, begitu juga siswa Persen lemak tubuh responden
dengan status gizi kurus banyak dimiliki selain dihitung menggunakan BIA juga
siswa MA. dihitung dengan model prediksi Slaughter,
Beberapa siswa MTs dengan status Deurenberg, Lee dan Chan yaitu dengan
gizi obesitas, dikarenakan siswa tersebut cara memasukkan hasil pengukuran
bertubuh kecil dan gemuk. Responden antropometri seperti IMT, total skinfold
dengan rentang umur 12-19 tahun thickness, umur serta jenis kelamin ke
mempengaruhi nilai skinfold thickness dalam model prediksi. Rata-rata persen
yang bervariasi pula yaitu 19,40-111,75 lemak tubuh responden yang dihitung
mm. Skinfold thickness bervariasi menurut menggunakan model prediksi Slaughter
umur, jenis kelamin dan etnis adalah 19,97% (95% CI 18,84-21,11),
(Daniels:2000). Skinfold thickness sedangkan dengan model prediksi
Deurenberg sebesar 18,18% (95% CI BIA adalah persen lemak tubuh yang
16,86-19,50). Persen lemak tubuh dihitung menggunakan model prediksi Lee
responden yang dihitung menggunakan dengan r = 0,805. Bila dilihat dari alat
model prediksi Lee memiliki rata-rata ukur antropometri, IMT WHO memiliki
sebesar 17,73% (95% CI 16,47-18,98) korelasi paling kuat diantara alat ukur
sedangkan yang dihitung dengan model antropometri lainnya (r = 0,804). Variabel
prediksi Chan sebesar 25,53% (95% CI yang memiliki korelasi paling lemah
24,91-26,15). adalah umur yaitu dengan r = -0, 371
namun keduanya memiliki p value 0,0005.
Korelasi Antar Variabel
Variabel yang memiliki korelasi
paling kuat dengan persen lemak tubuh
Tabel 2
Koefisien Korelasi Antar Variabel Penelitian
putera lebih tinggi daripada remaja puteri.

Kuatnya korelasi ini karena dalam Korelasi antara IMT dengan persen lemak
model prediksinya, Lee melibatkan IMT tubuh yang digambarkan dengan kurva
sebagai salah satu indikator linier berdasarkan jenis kelamin
perhitungannya. IMT memiliki korelasi menunjukkan bahwa dengan IMT yang
yang cukup kuat pula dengan persen lemak sama, remaja putera memiliki persen
tubuh BIA, kedua setelah persen lemak lemak tubuh yang lebih rendah daripada
tubuh model prediksi Lee. Hal ini remaja puteri.
diperkuat pula oleh sasaran yang
dikembangkan dalam model prediksi ini Analisis Regresi Linier Sederhana dan
meliputi usia responden yaitu 7-18 tahun di Ganda untuk Mendapatkan Model
wilayah Asia yaitu Korea sehingga cukup Prediksi
mewakili responden. IMT berkorelasi Dari hasil analisis regresi linier
dengan persen lemak tubuh secara sederhana diperoleh bahwa IMT WHO
signifikan menurut jenis kelamin dan etnis yang memiliki korelasi dan nilai R Square
(Wang, J1994). Korelasi IMT terhadap paling besar dengan persen lemak tubuh
persen lemak tubuh menggunakan BIA BIA yaitu r = 0,804 dan R Square 0,647,
sangat kuat yaitu r = 0,825 untuk laki- artinya persamaan regresi IMT WHO
laki(16). Korelasi IMT WHO (Z score) dapat menerangkan 64,7% variasi persen
dengan persen lemak tubuh BIA sangat lemak tubuh BIA atau persamaan garis
kuat dengan r = 0,946 untuk siswa laki-laki IMT WHO yang diperoleh cukup baik
sekolah dasar (Arini:2010)(17). Lee K et al untuk menjelaskan variabel persen lemak
(2007) menyatakan bahwa IMT remaja tubuh BIA.
Tabel 3
Analisis Regresi IMT WHO, Skinfold Thickness dan Lingkar Pinggang dengan Persen
Lemak tubuh BIA
dikeluarkan ternyata tidak terjadi

*PLT = Persen Lemak Tubuh perubahan R Adjusted hingga lebih dari
Keterangan : 10%, namun perubahan nilai B sebelum
IMT : Indeks MassaTubuh WHO dan sesudah variabel umur dihilangkan
WHO (Z Score) untuk variabel lingkar pinggang lebih dari
ST : Skinfold Thickness (mm) 10% yaitu 18,48% sehingga variabel umur
LP : Lingkar pinggang (cm) tetap masuk dalam model multivariat.
Model prediksi akan lebih valid jika
memenuhi asumsi/persyaratan yang
Berdasarkan hasil bivariat antar ditetapkan. Model prediksi akan lebih valid
variabel dapat disimpulkan bahwa semua jika memenuhi asumsi/persyaratan yang
variabel independen memenuhi syarat ditetapkan. Ada satu asumsi yang tidak
yaitu p value < 0,25, ini berarti variabel terpenuhi dari 6 asumsi yang ada, yaitu
dapat diikutkan dalam model multivariat. asumsi multicollinearity untuk variabel
Berdasarkan hasil uji, nilai p value semua lingkar pinggang. Artinya lingkar
variabel independen < 0,05 namun untuk pinggang berkorelasi dengan variabel
variabel umur >0,005 oleh karena itu umur lainnya, dan harus dikeluarkan dari model
harus dikeluarkan dari model prediksi multivariat. Hasil uji regresi linier setelah
dengan pertimbangan perubahan R lingkar pinggang dikeluarkan dari model
Adjusted dan nilai B. Setelah umur pada tabel 4 :
Tabel 4
Permodelan Prediksi Persen Lemak Tubuh

Tabel 5
Analisis Regresi IMT WHO dan Umur Skinfold Thickness dan Umur, Lingkar Pinggang
dan Umur dengan Persen Lemak Tubuh BIA
Berdasarkan hasil analisis maka sehingga tidak dapat digunakan untuk

diperoleh Model prediksi persen lemak keseluruhan populasi remaja Jawa Barat.
tubuh baru yaitu : Model prediksi ini dapat digunakan pada
Model Prediksi Sitoayu = populasi yang memiliki karakteristik yang
23,28+1,56*IMT WHO+0,13*ST-0,62*U. sama dengan responden pada penelitian
ini. Berbagai penelitian yang dilakukan
Hasil analisis multivariat antara untuk mengembangkan model prediksi
IMT WHO, skinfold thickness dan lingkar dengan indeks antropometri sudah banyak
pinggang dengan umur pada tabel 5 dapat dilakukan seperti model prediksi Slaughter
dilihat bahwa model prediksi skinfold (1988), Deurenberg (1991), Lee (2007)
thickness dan umur memiliki korelasi dan dan Chan (2009). Namun model prediksi
R Square paling besar dengan persen tersebut belum mewakili remaja di
lemak tubuh BIA yaitu r = 0,831 dan R Indonesia khususnya pada remaja laki-laki.
Square 0,691. Pengembangan model prediksi pada
remaja di Indonesia sudah dilakukan oleh
Analisis Cut-Off Point, Sensitivitas dan Isjwara RI., Widjaja L & Jan WS tahun
Spesivisitas 2007, namun untuk remaja puteri.
Model prediksi Sitoayu mewakili Penelitian ini mencoba mengembangkan
populasi remaja laki-laki usia 12-19 tahun model prediksi untuk menghitung persen
di Kota Cimahi Jawa Barat. Model lemak tubuh pada remaja laki-laki.
prediksi tersebut dapat digunakan untuk Hasil uji korelasi menyatakan
memperkirakan persen lemak tubuh bahwa model prediksi Sitoayu memiliki
dengan menggunakan IMT WHO, skinfold korelasi yang paling kuat dengan persen
thickness dan umur yang hasilnya sesuai lemak tubuh BIA dengan nilai (r = 0,855).
dengan persen lemak tubuh yang diukur Berdasarkan hasil tersebut dapat
menggunakan BIA. Berdasarkan nilai R disimpulkan bahwa model prediksi Sitoayu
Square, model prediksi ini dapat dikatakan dapat direkomendasikan untuk menghitung
cukup baik jika digunakan untuk besarnya persen lemak tubuh remaja laki-
mengukur persen lemak tubuh karena laki wilayah Jawa Barat khususnya Kota
sudah dapat menjelaskan 72,7% nilai Cimahi. Meskipun persen lemak tubuh
persen lemak tubuh. model prediksi skinfold thickness dan umur
Penggunaan model prediksi ini di memiliki korelasi kuat kedua setelah
wilayah yang lain mungkin akan persen lemak tubuh model prediksi Sitoayu
menyebabkan terjadi pergeseran nilai dari dengan nilai (r = 0,831).
hasil yang dihitung sebab pada penyusunan
model ini hanya melibatkan responden
yang berasal dari Kota Cimahi, Jawa Barat
Tabel 6
Korelasi Persen Lemak Tubuh Model Prediksi dengan Persen Lemak Tubuh BIA
Persen lemak tubuh yang dihitung 0,13% dan setiap bertambahnya umur
menggunakan model prediksi Lee under (tahun) persen lemak tubuh akan
estimate, tidak mendekati hasil persen mengalami penurunan sebesar 0,62%.
lemak tubuh BIA, hal ini dikarenakan rata- Penggunaan model prediksi persen lemak
rata persen lemak tubuh anak laki-laki tubuh ini diharapkan akan memenuhi
pada penelitian Lee cenderung memiliki kebutuhan ilmu pengetahuan untuk
persen lemak tubuh yang rendah yaitu 17- mengukur persen lemak tubuh secara lebih
22% sedangkan pada penelitian ini persen efisien dan efektif tanpa membutuhkan
lemak tubuh hingga 36%. Begitu pula biaya yang besar. Persen lemak tubuh yang
dengan persen lemak tubuh model prediksi dihitung menggunakan model prediksi ini
Deurenberg pada populasi Kaukasia (7-83 diharapkan dapat mendeteksi secara akurat
tahun), hal ini dikarenakan sebagian besar besarnya lemak tubuh yang dimiliki
responden pada penelitian Deurenberg seseorang, sekaligus dapat mendeteksi
berusia lebih dari 16 tahun yaitu 749 orang kemungkinan terjadinya gizi lebih terkait
sehingga memiliki persen lemak tubuh besarnya lemak tubuh yang dimiliki
berbeda dengan remaja yang cenderung seseorang.
tinggi.
Dari konstanta pada model Kurva ROC, Sensitivitas dan
prediksi Sitoayu dapat diketahui bahwa Spesifisitas
variabel yang memiliki pengaruh paling Penelitian ini menemukan bahwa
besar terhadap persen lemak tubuh adalah area di bawah kurva yang terbesar dicapai
IMT WHO, semakin besar kostanta maka oleh persen lemak tubuh yang dihitung
semakin besar pula pengaruhnya terhadap menggunakan model prediksi Sitoayu
penentuan persen lemak tubuh. Setiap (93,7%) dan di urutan kedua adalah persen
kenaikan 1 SD IMT WHO akan menaikkan lemak tubuh yang dihitung dengan model
persen lemak tubuh sebesar 1,56%, setiap prediksi skinfold thickness dan umur
kenaikan 1 mm skinfold thickness akan (92,9%). Kurva ROC merupakan ringkasan
menaikkan persen lemak tubuh sebesar antara nilai sensitivitas dan spesifisitas

pada berbagai cut-off point. Semakin standar yang telah ada tetap lebih baik
mendekati nilai 1 maka dikatakan bahwa untuk menapis gizi lebih berdasarkan
uji diagnostik tersebut semakin baik persen lemak tubuh karena lebih sensitif.
(Park:2004). Namun, untuk model prediksi IMT WHO,
Meskipun nilai AUC antara persen IMT WHO dan umur dan Sitoayu memiliki
lemak tubuh dengan model prediksi nilai sensitivitas yang tetap baik meskipun
Sitoayu dan skinfold thickness dan umur mengalami penurunan yaitu untuk IMT
tinggi namun berdasarkan nilai sensitivitas WHO memiliki sensitivitas 86% dan
diperoleh model prediksi dengan spesifisitas 79%, IMT WHO dan umur
sensitivitas tiga terbaik yaitu IMT WHO, memilki sensitivitas sebesar 82% dan
IMT WHO dan umur dan Sitoayu. spesifisitas 82%, sedangkan untuk model
Berdasarkan cut off point standar (19,5%), prediksi Sitoayu memiliki sensitivitas 82%
persen lemak tubuh yang dihitung dengan dan spesifisitas 89%.
model prediksi IMT WHO memiliki Nilai prediksi positif yang rendah
sensitivitas yang lebih baik dibandingkan akan mengakibatkan tingginya jumlah
dengan persen lemak tubuh yang dihitung kasus false positive, artinya akan ada
menggunakan model prediksi Sitoayu. banyak kasus yang dideteksi positif
Persen lemak tubuh yang dihitung memiliki persen lemak tubuh tinggi oleh
menggunakan model prediksi IMT WHO pengukuran persen lemak tubuh namun
memiliki sensitivitas yang paling tinggi sebenarnya tidak. Sebaliknya nilai prediksi
yaitu 94% dan spesifisitas sebesar 71%, negatif yang tinggi akan mengakibatkan
sedangkan persen lemak tubuh yang rendahnya kasus false negative, artinya
dihitung menggunakan model prediksi hampir seluruh kasus yang dideteksi
IMT WHO dan umur juga memiliki memiliki persen lemak tubuh normal
sensitivitas yang tinggi yaitu 94% dengan benar-benar memiliki persen lemak tubuh
spesifisitas 73%. Persen lemak tubuh yang yang tidak tinggi. Dalam hal ini persen
dihitung menggunakan model prediksi lemak tubuh yang dihitung dengan model
Sitoayu berada diurutan ketiga dengan prediksi IMT WHO dan IMT WHO dan
sensitivitas 84% dan spesifisitas sebesar umur memiliki nilai prediksi negatif yang
85%. Uji diagnostik dengan sensitivitas paling tinggi yaitu 92,2% dan 92,4%.
yang tinggi diperlukan pada kondisi Sedangkan persen lemak tubuh
dimana kasus yang ingin dideteksi yang dihitung menggunakan model
merupakan kasus yang serius walaupun prediksi Sitoayu memilki nilai prediksi
nilai prediksi positif uji tersebut relatif positif paling tinggi yaitu 85% dengan cut
rendah. Persen lemak tubuh untuk off berdasarkan standar. Jika dibandingkan
mendeteksi status gizi lebih memerlukan dengan cut off baru (tidak berdasarkan
nilai sensitivitas yang tinggi, karena standar), nilai prediksi positif yang paling
kekeliruan hasil uji negatif akan berakibat tinggi dimiliki oleh model prediksi Sitoayu
serius (Park:2004) . sebesar 88% dan nilai prediksi negatif
Perhitungan persen lemak tubuh yang paling tinggi dimiliki oleh model
menggunakan cut off baru (bukan prediksi IMT WHO sebesar 84,9%.
berdasarkan standar yang telah ditetapkan) Persen lemak tubuh yang dihitung
cenderung mengalami penurunan nilai menggunakan model prediksi Sitoayu
sensitivitas dibandingkan dengan meng- memiliki nilai likelihood ratio positive
gunakan standar, hal ini membuktikan yang paling tinggi baik yang dihitung
bahwa cut off point dengan menggunakan mengginakan cut off standar maupun cut

off baru, artinya remaja yang hasil semakin kecil nilai LR- menunjukkan
perhitungan persen lemak tubuhnya positif bahwa uji diagnostik yang digunakan
memiliki kemungkinan untuk benar-benar semakin baik. Nilai LR+ ≥ 10 dan LR- ≤
memiliki persen lemak tubuh yang tinggi. 0,1 merupakan batasan yang digunakan
Sedangkan persen lemak tubuh yang untuk menilai performa sebuah uji
dihitung menggunakan model prediksi diagnostik.
IMT WHO dan IMT WHO dan umur Cut-off point baru yang diperoleh
memiliki nilai likelihood ratio negative pada penelitian ini untuk persen lemak
yang paling baik (0,08), artinya remaja tubuh yang dihitung menggunakan model
yang dinyatakan memiliki persen lemak prediksi Sitoayu adalah 20,08%. Ber-
tubuh rendah dengan perhitungan model dasarkan hasil analisis dapat diketahui
prediksi IMT WHO dan IMT WHO dan bahwa model prediksi persen lemak tubuh
umur, kemungkinan 12,5 kali untuk benar- yang memiliki sensitivitas paling tinggi
benar tidak memiliki persen lemak tubuh adalah model prediksi IMT WHO, IMT
yang tinggi. Semakin besar nilai LR+ dan WHO dan umur dan Sitoayu.
Tabel 7
Nilai Cut-off Point, Sensitivitas dan Spesifisitas Berbagai Titik Pengukuran Berdasarkan
Standar dan Bukan Standar
Model Prediksi Sitoayu = 23,28 +

Kesimpulan 1,56*IMT WHO + 0,13*ST - 0,62*U
Telah ditemukan model prediksi Model prediksi IMT WHO dan IMT WHO
baru untuk remaja laki-laki sebagai Umur memiliki sensitivitas paling baik
alternatif menghitung persen lemak tubuh untuk menapis gizi lebih pada remaja laki-
dengan sensitivitas 84%, spesifisitas 85%, laki yaitu 94% dan dapat digunakan
nilai prediksi positif 85% dan nilai prediksi sebagai alternatif model prediksi dalam
negatif 84%, yaitu model prediksi Sitoayu menghitung persen lemak tubuh.
yang dapat digunakan pada populasi Diharapkan adanya penelitian lain yang
dengan karakteristik sama: serupa untuk memvalidasi temuan
penelitian ini terkait penggunaan model
prediksi untuk menghitung persen lemak Journal of Clinical Nutrition, 55,

tubuh pada remaja di masing-masing 973-979.
daerah di Indonesia, dengan menggunakan
DEXA atau Under Water Weighing Gibson, R.S. (2005). Principles of
sebagai golden standard dan metode Bland Nutrition Assessment, Second
dan Altman untuk validasi. Edition. New York. USA: Oxford
University Press, Inc.
Daftar Pustaka
Arini, F.A. (2010). Pengukuran Gillum, R.F. (1999). Distribution of Waist-
Antropometri dan Korelasinya To-Hip Ratio, Other Indices of
dengan “Golden Standard” Body Fat Distribution and Obesity
Persen Lemak Tubuh Bioelectrical and Associations with HDL
Impedance Analysis : Studi Cholesterol in Children and Young
Validasi Pada Anak SD tahun Adults Aged 4-19 Years: The Third
2010, [Tesis]. Fakultas Kesehatan National Health and Nutrition
Masyarakat Universitas Indonesia, Examination Survey. International
Depok. Journal of Obesity and Related
Metabolic Disorders, 23, 556-563.
Bozkirli, E. et al. (2007). The Validity of
The World Health Organisation’s Isjwara, R.I., Widjaja, L. & Jan, W.S.
Obesity Body Mass Index Criteria (2007). Comparison of Body
in a Turkish Population: a Compositional Indices Assesed by
Hospital-Based Study. Asia Pacific Underwater Weighing,
Journal Clinical Nutrition, 16(3), Bioelectrical Impedance and
443-447. Anthropometry in Indonesian
Adolescent Girls. Asia Pasific
Chan, D.F.Y. et al. (2009). New Skinfold Journal of Clinical Nutrition,
Thickness Equation For Predicting 16(4), 641 – 648.
Percentage Body Fat in Chinese
Obese Children. Hong Kong Lee, K. et al. (2007). Percent Body Fat Cut
Journal Pediatric, 14, 96-102. off Values For Classifying
Overweight and Obesity
Daniels, S.R., Philip R.K., & John A.M. Recommended by The
(2000). Utility of Different International Obesity Task Force
Measures of Body Fat Distribution (IOTF) in Korean Children. Asia
in Chidren and Adolescents. Pasific Journal of Clinical
American Journal of Nutrition, 16(4), 649-655.
Epidemiology, 152(12), 1179-
1183. Mei, Z. et al. (2002). Validity of Body
Mass Index Compared with Other
Deurenberg, P. et al. (2001). Original Body Composition Screening
Communication: The Validity of Indexes for The Assesement of
Predicted Body Fat Percentage Body Fatness in Children and
from Body Mass Index and From Adolescents. American Journal of
Impedance in Samples of Five Clinical Nutrition, 75, 978-985.
European Populations. European

Mei, Z. et al. (2007). Do Skinfold

Measurements Provide Additional Wang, J. et. al. (1994). Asian Have Lower
Information to Body Mass Index in Body Mass Index (BMI) But
The Assesments of Body Fatness Higher Percent Body Fat Than Do
Among Children and Adolescents?. Whites: Comparisons of
Pediatrics, 119(6), e1306-e1314. Antropometric Measurement.
American Journal of Clinical
Park, S.H., Jin, M.G., & Chan-Hee, J. Nutrition, 60, 23-28.
(2004). Receiver Operating
Characteristic (ROC) Curve: Yusnita, E. (2005). Studi Validasi
Practical Review for Radiologists. Pengukuran Persen Lemak Tubuh
Korean Journal Radiologists, 5(1), dengan Bioelectrical Impedance
11-18. Analysis (BIA) dan Menggunakan
IMT, RLPP, Skinfold Thickness
Rogol, A.D. et al. (2002). Growth at pada Mahasiswa Program
Puberty. Journal of Adolescent Magister FKM UI 2005, [Skripsi].
Health, 31(6S), 192-200. Fakultas Kesehatan Masyarakat
Universitas Indonesia, Depok.
Santrock, J.W. (2007). Remaja Edisi
Kesebelas. Jakarta: Penerbit
Erlangga.
Sopher, et al. (2004). Measurement of

Percentage of Body Fat in 411
Children and Adolescents: A
Comparison of Dual Energy X Ray
Absorptiometry with a Four
Compartment Model. Pediatric,
113(5), 1285-1290.
Sopher, et al. (2004). Measurement of

Percentage of Body Fat in 411
Children and Adolescents: A
Comparison of Dual Energy X Ray
Absorptiometry with a Four
Compartment Model. Pediatric,
113(5), 1285-1290.
Umi, F. & Drupadi, HS.D. (2007).

Handbook Nutritional Assessment.
Jakarta: South East Asian of
Education Organization Tropical
Medicine and Public Health
Regional Center for Community
Nutrition (SEAMEO-TROPMED
RCCN), Universitas Indonesia.

Pun, et.al. Amerta Nutr (2021). 377-386. 377
DOI: 10.20473/amnt.v5i4.2021. 377-386.
RESEARCH STUDY Open Access
Validitas Berat Badan, Tinggi Badan, dan Indeks Massa Tubuh Diri Sendiri
diantara Remaja Akhir Malaysia
Validity Of Self-Reported Weight, Height And Body Mass Index Among

Malaysian Late Adolescents
Kong Pun Pun1, Nithiah Thangiah1*, Hazreen Abdul Majid1,2
Abstrak
Latar belakang: Malaysia menderita masalah obesitas dengan hampir seperempat dari anak-anak dan remaja mengalami
kelebihan berat badan atau obesitas. Metode yang divalidasi diperlukan untuk menilai tren kelebihan berat badan dan
obesitas dengan mudah.
Tujuan: Penelitian ini menyelidiki keakuratan tinggi dan berat badan remaja akhir Malaysia yang dilaporkan sendiri
dibandingkan dengan metode pengukuran langsung.
Metode: Studi potong lintang ini dilakukan pada tahun 2019 pada 114 partisipan berusia 19-20 tahun. Berat dan tinggi badan
yang dilaporkan sendiri dan diukur langsung dikumpulkan melalui panggilan telepon dan instrumen studi. Nilai berat dan
tinggi yang diperoleh digunakan untuk menghitung BMI dan menentukan klasifikasi.
Hasil: Nilai ICC sangat baik antara berat badan yang dilaporkan dan diukur (ICC= 0.94; 95% CI=0.91-0,96), tinggi badan (ICC=
0.95; 95% CI=0.83-0.98) dan BMI (ICC = 0.89; 95% CI =0,78-0,94) ditemukan. Selain itu, kesepakatan yang baik yang diamati
dari plot B & A menunjukkan bahwa perbedaan antara kedua variabel pengukuran independen kecil dan sebagian besar
perbedaan berada dalam area batas kesepakatan pada tingkat populasi untuk nilai berat badan, tinggi badan dan BMI.
Selanjutnya, Kappa Cohen menunjukkan kesepakatan substansial antara status BMI yang diukur dan dikategorikan dari berat
dan tinggi badan dari kedua metode, semua (ҡ = 0,61; 95% CI = 0,48-0,74), anak laki-laki (ҡ = 0,67; 95% CI = 0,43-0,91 ) dan
anak perempuan (ҡ = 0,58; 95% CI=0,43-0,73).
Kesimpulan: Studi validasi ini menyimpulkan bahwa berat dan tinggi badan yang dilaporkan sendiri sesuai dengan metode
pengukuran langsung. Teknik ini dapat digunakan untuk menilai status antropometrik remaja akhir Malaysia untuk studi
populasi.
Kata kunci: Validitas, Lapor Diri, Ukur Langsung, Berat Badan, Tinggi Badan, IMT, Remaja Akhir
ABSTRACT
Introduction: Malaysia suffers from obesity problems with nearly a quarter of the children and adolescents being overweight
or obese. Validated methods are needed to measure the trend of overweight and obesity easily.
Objective: The study investigates the precision of Malaysian late adolescents' self-reported height and weight compared to
direct-measured methods.
Methods: About 114 young adults aged 19-20 years old participated in this cross-sectional study in 2019. Self-reported and
direct-measured anthropometrics including height and weight were gathered through phone calls and study instruments.
The obtained height and weight values calculates BMI and determines classifications.
Results: The ICC value were excellent between reported and measured weight (ICC= 0.94; 95% CI=0.91-0.96), height (ICC=
0.95; 95% CI=0.83-0.98) and BMI (ICC = 0.89; 95% CI=0.78-0.94) were found. Additionally, good agreement observed from B
& A plots indicated that the differences between both independent measurement variables were minor and most of the
differences were within the area of the limit of agreement at the population level for their anthropometric measurements.
The Cohen’s Kappa showed substantial agreement of BMI calculated from reported weight and height (ҡ = 0.61; 95% CI=
0.48-0.74), boys (ҡ = 0.67; 95% CI=0.43-0.91) and girls (ҡ = 0.58; 95% CI=0.43-0.73).
Conclusion: This validation study concluded that self-reported height and weight were in agreement with direct-measured
methods. This technique can be utilized to assess the anthropometric status of Malaysian late adolescents for population
studies.
Keywords: Validity, Self-Reported, Direct-Measured, Weight, Height, BMI, Late Adolescents
*Corresponding
Author:
nithiah@um.edu.my / nithiah_6@yahoo.com
©2021. Pun, et.al. Open Access under CC BY – SA license.

Received: 17-07-2021, Accepted: 24-11-2021, Published online: 25-11-2021.
doi: 10.20473/amnt.v5i3.2021. 377-386. Joinly Published by IAGIKMI & Universitas Airlangga
DOI: 10.20473/amnt.v5i4.2021. 377-386.
Nithiah Thangiah
1Centre for Population Health, Department of Social & Preventive Medicine, Faculty of Medicine, University of Malaya, Kuala
Lumpur, Malaysia
2Department of Nutrition, Faculty of Public Health, Universitas Airlangga. Indonesia
INTRODUCTION correlations have been observed between the self-

A rapidly growing global epidemic of reported and direct-measured height and weight in many
overweight and obesity is occupying many parts of the validities research but most were in western countries
world. According to the current facts from World Health 10,15-17. Limited information exists regarding the
Organization (WHO), obesity has almost tripled globally rationality of these measures in the Asian context and
since 1975. Malaysia has been leading within Asian especially even rare among the adolescent population.
countries with overweight/obesity problems with nearly Only one self-reported weight and height validation study
half of the population (aged ≥18) are overweight or obese was conducted among 663 adolescents aged 13-17 years
1. Based on the “Tackling obesity in ASEAN” report old in Kelantan, Malaysia 18 but no study has been done
produced by Economist Intelligence Unit (EIU) revealed on late adolescents (aged 19-20) who completed
that Malaysia achieved the highest obesity (13.3%) and secondary schools. Therefore, validation studies of self-
overweight (38.5%) across all six countries studied 2. The reported height, weight and BMI status are to be used as
most recent National Health and Morbidity Survey an alternative method in assessing the weight of the
(NHMS) in 2019 reported overweight and obese adolescents in the community.
prevalence of 30.4% and 19.7% respectively among
Malaysian adults3. Childhood and adolescent obesity are METHODS
another big issue because it contributes to the escalating A cross-sectional study in which the data of self-
prevalence of non-communicable diseases among reported and direct-measured height, weight and BMI
youngsters and in turn become another burden to the and social demographic characteristics were collected
country 1,4. In Malaysia, the obesity prevalence among from 114 adolescents aged 19-20 years old in 2019.
children (aged < 18) is escalating 1,4,5. School-going students who participated in the Malaysian
Overweight children and adolescents are at many Health and Adolescents Longitudinal Research
folds at risk of becoming overweight adults as negative (MyHeARTs) study from 2012-2016 were contacted. The
impacts of childhood obesity continue through adulthood6-8. MyHeARTs study included schools from three states in
Consequently, adolescents become at risk of developing Peninsular Malaysia including Kuala Lumpur and Selangor
numerous medical conditions later in life9. Health conditions and Northern Perak19. A telephone interview was
may lead to chronic diseases that result in death and a loss of conducted to obtain sociodemographic characteristics,
working population that impacts the nation's economy. In self-reported weight and height. Subsequently,
addition, obesity also affects children's cognitive & mental appointments were made to measure their height and
health by developing stress, sadness and low self-esteem. weight. Data collection was conducted between January
Obese children often face stigmatization and social and March 2019 and participants who completed both
discrimination that cause them to lack participation in social phases of self-reported and direct measures were
activities. Their habit of keeping away from the community included in the study. Individuals’ self-reported values of
and less social interaction results in fewer friends and more weight and height were recorded to the nearest kilogram
time spent in sedentary activities9. These alarming (kg) and nearest centimeter 10 respectively. Subsequently,
adolescent obesity issues require a need for a quick and the self-reported BMI was calculated and recorded in
valid method to evaluate the trend of overweight and kilogram per meter square (kg/m2). Other than that,
obesity among them. several socio-demographic characteristics such as marital
Generally, directly measured height and weight status, socioeconomic status, place of residence as well as
are the first preference in calculating BMI 1, however self- past medical history were assessed. During the second
reported height and weight are most commonly applied phase, trained enumerators took anthropometric
in surveillance systems and wide-ranging studies 10 but measures according to the standard protocols.
due to the restriction of time, manpower 11 and cost, an Student height was measured with a SECA 217
alternative way is worth to be considered. Nevertheless, portable calibrated stadiometer (Seca Portable 217, Seca,
inaccuracy of this self-reported weight and height UK) and student weight was assessed using calibrated
especially among overweight or obese adolescents 12 SECA 813 high-capacity digital flat scale (Seca 813, Seca,
potentially may impact on misclassification of BMI 13, UK). Individuals’ BMI cut-off points of 25 kg/m2 defines
leads to difficulty in implementing and evaluating health overweight and 30 kg/m2 defines obesity20. These cut-off
programs 14. points are meaningful for comparisons between or within
Self-reported methods are utilized when there populations.
are limitations in terms of budget or manpower The data management and analysis were
constraints to obtain measured anthropometric values. conducted with IBM SPSS statistical software package
Thus, the credibility of such self-reported measures is version 25.0 (SPSS Inc., Chicago, IL, USA) for Windows.
critical to ensure the quality of data that circumvent Statistical significance for all the data analyses performed
biases due to misclassification. Apart from that, high in the study was set at p < 0.05. A paired t-test was applied

DOI: 10.20473/amnt.v5i4.2021. 377-386.
for normally distributed data to determine the mean good and excellent reliability, respectively. Cohen’s Kappa
differences between the self-reported and direct statistics was applied to assess the correct categorization
measured height for all late adolescents and then of body mass index (BMI) status calculated from reported
stratified by sex. In contrast, the Wilcoxon Signed-Rank and measured values. The range for kappa values is from
test (non-parametric) was used considering the non- -1.00 to 1.00 indicating perfect disagreement to
normally distributed data of some variables (weight and agreement.
BMI values). Independent T-test was applied to compare Ethical approval was obtained by the MyHeART
the differences of variables (height) bias according to study team from Medical Ethics Committee, University
gender as height was distributed normally. In opposition, Malaya Medical Centre (MEC Ref. No: 896.34) and from
the Mann-Whitney U test was used for weight and BMI as Medical Research & Ethics Committee; Ministry of Health
they were not normally distributed. Bland Altman plot (MOH) Malaysia (Reference number: NMRR-14-376-
(limits of agreement) 21 was used to display and measure 20486). Informed consent was collected from the study
the agreement between self-reported and direct participants before data collection.
measured weight, height and BMI and to analyze the
differences between the measurements by the two RESULTS AND DISCUSSION
methods on each subject 22. The agreement is The selected socio-demographic characteristics
represented by the mean/median difference between of 114 healthy late adolescents without chronic diseases
self-reported and direct-measured data. The intraclass were displayed in Table 1. More than half of the
correlation coefficient (ICC) acts as a reliability index that participants were female (68.4%) and males were 31%.
measures the correlation and agreement between The study subjects comprised of the three main ethnic
measurements. In this study, ICCs were calculated to groups in Malaysia which were Malay (61.4%), Chinese
measure the association between self-reported and (33.3%) and Indian (5.3%). Most of the study participants
direct measured methods. Reliability ranges between 0.5- (76.3%) were full-time students who furthered their
0.75, 0.75-0.9 and greater than 0.9 indicating moderate, studies in colleges, polytechnic schools and universities.
Table 1: Sociodemographic characteristics of study participants

Socio-demographic n (%)
characteristics
Age 19 102 (89.5)
20 12 (10.5)
Sex Male 36 (31.6)
Female 78 (68.4)
Ethnicity Malays 70 (61.4)
Chinese 38 (33.3)
Indian 6 (5.3)
Marital Status Single 114 (100)
Socioeconomic Status Full time study 87 (76.3)
Full-time work 19 (16.7)
Part-time work 5 (4.4)
Work and Study 2 (1.8)
At home 1 (0.9)
Place of Residence Urban 111 (97.4)
Rural 3 (2.6)
Figure 1 shows the Bland-Altman plot of the zero. Out of 114, 110 (96.5 %) of the differences were
differences of self-reported and direct-measured weight, within the area of limits of agreement, while there were
height and BMI versus mean weight, height and BMI, only four data (3.5%) that were equally distributed below
respectively from both methods. Bland-Altman plot was the lower and above the upper limit. Overall, there was
constructed to study agreement between self-reported good agreement between self-reported and direct
and measured anthropometrics. Y-axis displays the measured weight with a minimal 0.5 kg differences below
difference between the self-reported and direct- zero only.
measured data and the X-axis was the mean of total self-
reported and direct-measured values divided by two. In
this study the median weight difference (bias) was -0.5 kg,
this indicates that the self-reported method measures 0.5
kg lesser than the direct-measured and this value formed
the middle line of the 95% limits of the agreement below

DOI: 10.20473/amnt.v5i4.2021. 377-386.
Bland-Altman Pot
/cm
/kg
/kg
/cm
Figure 1 Bland-Altman plot of overall weight, height and BMI differences versus mean weight, height and BMI of self-
reported and direct-measured method
*Difference= Reported – Measured
*Mean = (reported + measured)/2
The mean height difference was 1.54 cm that self-reported and direct-measured BMI values with a
indicates on average the self-reported method measures minimal difference of 0.5 kg/m2.
1.54 cm more than the direct-measured and this value Table 2 shows the descriptive of self-reported
formed the middle line of the 95% LOA above zero. Out of and direct-measured weight (median, inter-quartile
114, 109 of the differences were within the area of limits range), height (mean, standard deviation) and BMI
of agreement, while there were only five participants (median, inter-quartile range) by sex. The magnitude of
(4.4%) were found to be out of the LOA. Approximately differences between both methods was also compared.
96% of the height differences were within the area of LOA Late adolescents significantly under-reported their
between 5.47 and -2.38 cm with a mean bias of 1.54 cm. weights with a minimal median difference of 0.5 kg (IQR
The median difference of BMI was -0.5 kg/m2, =3.7, range=-27.9 to 9.2, d=0.23, p<0.05) and significantly
this indicates that the self-reported method measures 0.5 over-reported their heights by 1.5 cm (SD = 2.0, 95% CI
kg/m2 less than the direct-measured and this value =1.2-1.9, d=0.19, p<0.05) but both the effect size for the
formed the middle line of the 95% limits of the agreement difference was small. The self-reported BMI values were
below zero which were -6.1 and 1.7 kg/m2. Out of 114, quantified from self-reported weight and height whereas
110 of the differences were within the area of limits of the direct measured BMI values were calculated from
agreement, while there were only four readings (3.5%) direct measured weight and height. Overall, the median
were equally found located out of the LOA area. Around of the self-reported BMI (21.6 kg/m2, IQR=5.3,
96.5% of the adolescents' BMI difference values fell range=14.6-37.1) was significantly lower than direct-
between 1.7 kg/m2 and -6.1 kg/m2. In summary, the width measured BMI (22.9 kg/m2, IQR=6.0, range=14.0-47.8)
of LOA for BMI (7.8 kg/m2) among all participants was with a median difference of -0.5 kg/m2 (IQR=1.8, range=-
greater than one IQR of the measured weight values 11.4- 4.3, p<0.05), but the effect size of difference was
(IQR=6.0 kg/m2). However, the width of LOA was less than small (d=0.38).
two IQR, therefore showing good agreement between

DOI: 10.20473/amnt.v5i4.2021. 377-386.
For males, the median differences between females, self-reported BMI was 21.6 kg/m2 (IQR=5.7,
self-reported and direct-measured values were 0.7 kg range=14.6-37.1) which was significantly lower than the
(IQR = 5.2, range=-10.4-3.6, d=0.19, p<0.05) for weight direct-measured BMI, 22.3 kg/m2 (IQR=7.3, range=14.0-
with small magnitude of difference, and 1.8 cm (SD=2.2, 47.8). Wilcoxon’s Sign ranked test shows all the
95% CI= 1.1-2.5, d=0.36, p<0.05) for height. For females, differences for both sexes between two methods were
the corresponding values were −0.4 kg (IQR = 3.0, range=- significant and self-reported was lower than direct
27.9-9.2, d=0.21, p<0.05) for weight with small magnitude measured BMI, but the magnitude of differences was
of difference and 1.4 cm (SD = 1.9, 95% CI=1.0-1.9, small ranged between 0.34 to 0.38 and the difference
d=0.23, p<0.05) for height. In male, the median self- values were also small (-0.5 kg/m2) which were near to
reported BMI was 22.0 kg/m2 (SD=3.4, 95% CI=20.8-23.1) zero. Besides that, by gender, Mann Whitney U test
while the direct-measured BMI was 23.2 kg/m2 (SD=3.6, analysis presented no significant differences discovered
95% CI=22.0-24.4), this means there was a significant between boys and girls proven that there were no gender
underestimate of BMI values with a median difference of differences in the underestimation of BMI values.
-0.5 kg/m2 (IQR=2.0, range=-6.1 to 1.7, p<0.05). Then in
Table 2: Descriptives of weight, height and BMI based on self-reported and direct measurement
All/n=114 Male/n=36 Female/n=78
Median Median Median
Weight (kg)
(IQR) (IQR) (IQR)
Self-reported 57.5(17.7) 64.9 (11.1) 55.0(18.3)
Range 36-101 45-85 36-101
Direct measured 58.4(21.1) 67.0 (11.2) 56.2 (20.5)
Range 34.8-126.9 49.5-89.4 34.8-126.9
**Difference -0.5 (3.7)* -0.7 (5.2)* -0.4(3.0)*
Range -27.9-9.2 -10.4-3.6 -27.9-9.2
Cohen’s d 0.23 0.19 0.21
Mann Whitney U test#
Difference -0.3
p 0.6
Height Mean Mean Mean
Self-reported 163.0 (8.3) 171.7 (4.9) 158.9 (6.2)
95% CI§ 161.4-164.5 170.1-173.4 157.5-160.3
Direct measured 161.4 (8.2) 169.9 (5.1) 157.5 (6.2)
95% CI§ 159.9-163.0 168.2-171.6 156.1-158.9
Paired T test
Mean difference* 1.5 (2.0)* 1.8 (2.2)* 1.4 (1.9)*
95% CI§ 1.2-1.9 1.1-2.5 1.0-1.9
Cohen’s d 0.19 0.36 0.23
Independent T Test#
Mean difference 0.4
p 0.3
Median Median
BMI values Median (IQR)
(IQR) (IQR)
Self-reported 21.6 (5.3) 22.0 (3.4) 21.6(5.7)
Range 14.6-37.1 20.8-23.1 14.6-37.1
Direct measured 22.9(6.0) 23.2 (3.6) 22.3(7.3)
Range 14-47.8 16.1-30.6 14-47.8
**Difference -0.5(1.8)* -0.5 (2.0)* -0.5(1.7)*
Range -11.4-4.3 -6.1-1.7 -11.4-4.3
Cohen’s d 0.38 0.34 0.35

DOI: 10.20473/amnt.v5i4.2021. 377-386.
Mann Whitney U Test#

Difference# 0
p 0.6
*Significant at p<0.05
§ 95% CI= 95% Confidence Interval
**Difference=Self-reported data – directly measured data (Wilcoxon signed-rank test and Paired
Sample t-test compare the differences)
#Mann Whitney U test and Independent T-test compare between sexes
Intraclass Correlation Coefficient height. Males (0.85, CI=0.47-0.94) had a lower ICC value
Table 3 shows results of intraclass correlation which falls under the good reliability category as
coefficient (ICC) between self-reported and direct compared to females (0.93, CI=0.79-0.98) were
measured weight, height and BMI of all participants, by categorized under excellent reliability. The ICC values for
males and females. The overall ICC analysis shows BMI proved good reliability (0.89, CI=0.78-0.94) between
excellent reliability (0.94, CI=0.91-0.96) between self- the two measurements methods. Males had an ICC value
reported and direct measured weight. Both males and of 0.83 (CI=0.54-0.93) for BMI which falls under the good
females also show excellent reliability, which was 0.92 reliability category as well as females 0.90 (CI=0.81-0.94)
(CI=0.82-0.98) and 0.94 (CI=0.85-0.98) respectively. The were also categorized under good reliability.
ICC values demonstrate excellent reliability (0.95,
CI=0.83-0.98) between self-reported and direct measured
Table 3: Results of ICC between Self-Reported and Direct Measured Weight, Height and BMI
Weight (kg) Height BMI (kg/m2)
ICC (95% CI)
0.94 0.95 0.89
All (n=114)
(0.91-0.96) (0.83-0.98) (0.78-0.94)
0.92 0.85 0.83
Male (n=36)
(0.82-0.98) (0.47-0.94) (0.54-0.93)
0.94 0.93 0.9
Female(n=78)
(0.85-0.98) (0.76-0.98) (0.81-0.94)
Percentage of Agreement (Kappa) values (ҡ = 0.61; 95% CI=0.48-0.74; p<0.05). A higher

Table 4 shows the number of late adolescents agreement among boys were found compared to girls. A
categorized by BMI status based on self-reported and significant substantial agreement found between male
direct measurements of all participants and between BMI status categorized based on self-reported and the
boys and girls. Based on the direct-measured BMI status direct measurements (ҡ = 0.67, 95% CI=0.43-0.91; p<0.05)
calculated from direct-measured weight and height, and only moderate agreement for female BMI status
approximately half of the adolescents were under normal categorized based on self-reported and the direct
weight (n=60), almost one-third of the total participants measurements (ҡ = 0.58, 95% CI=0.43-0.73; p<0.05).
were overweight and obese (n=36) while the remaining However, observed from the 95% confidence interval
participants were underweight (n=18) category. Among between males and females, there will be a chance that
all 114 respondents, most adolescents were of normal the Cohen’s kappa statistic value will be the same,
weight, 53 out of 60 (88%) and classified correctly, then therefore this gives evidence that there was no difference
the second-highest was underweight adolescents, when comparing the agreement for BMI status between
approximately 78% of them classified correctly. males and female. In summary, overall as well as for both
Furthermore, 56% of overweight and 54.5% of obese males and females, a substantial agreement between
adolescents were classified correctly. From direct- self-reported and measured BMI status was discovered.
measured BMI status, higher proportions of female
adolescents were under the overweight and obese
category as compared to males.
Cohen’s Kappa statistics shows a significant and
substantial agreement between the BMI categorization
quantified from self-reported and directly measured

DOI: 10.20473/amnt.v5i4.2021. 377-386.
Table 4: Number of late adolescents categorized by BMI status based on self-reported and direct measured weight and
height
Direct Measured (n)
Cohen’s Kappa
Underweight Normal Overweight Obesity Total
(95% CI)
All 0.61*
Underweight 14 5 0 0 19
(0.48, 0.74)
Normal 4 53 11 2 70
Overweight 0 2 14 3 19
Obesity 0 0 0 6 6
Total 18 60 25 11 114
Males 0.67*
Self-reported (n)
(0.43, 0.91)
Normal 0 21 3 1 25
Obesity 0 0 0 0 0
Total 3 23 9 1 36
Females 0.58*
(0.43, 0.73)
Normal 4 32 8 1 45
Obesity 0 0 0 6 6
Total 15 37 16 10 78
*Significant at p<0.05
sex for ICC and Bland Altman plot where both boys and
This study investigated the validity of self- girls had a good correlation and fair agreement between
reported weights and heights to analyze the accuracy of both weight and height measurement methods. As such,
self-reported BMI to identify overweight and obesity in a many studies concluded that self-reported weight and
sample of Malaysian late adolescents. The mean height are appropriate to be used in large-scale surveys
difference between self-reported and direct measured and epidemiology studies at the population level18,36,38 to
weight values was significantly under-reported and calculate BMI and determine weight status.
between self-reported and direct-measured height values The median difference of self-reported BMI
was significantly over-reported as expected. Findings values was significantly under-reported10,18,26,29,31,36,38,40.
stated that although most adolescents and adults will In particular, a study in Thailand found parallel findings
under-report weight 17,23,24, the differences between self- with lower self-reported BMI35. This was perhaps due to
reported and direct measured values are greater among similarities in culture, education standards and
females 10,12,25-34. This general finding was in contrast with knowledge in health. However, in this study, the BMI
the current study results as the presented study values between males and females showed no gender
discovered boys had a larger weight difference (-0.7 kg) differences. In addition, the ICC and B&A plots showed
while the weight difference for girls was -0.4 kg. The main good correlation and fair agreement in all adolescents as
reason for the disparities might be due to the differences well as separately by gender. When self-reported and
in norms, culture and ethnicity which may affect the direct-measured methods of BMI values calculated were
perception and satisfaction of an adolescent. The study classified into BMI status (underweight, normal weight,
results echo the findings from Asian studies18,35. In terms overweight and obese), substantial agreement was found
of height, there were mixed findings that involved over- between both measurement methods. Therefore, self-
reporting of height 31,36-38 and under-reporting of height reported weight, height and corresponding BMI values
18,26,39,40 thus producing partial consistency with the are valid proxy measurements10,40. Nevertheless, self-
present findings. reported BMI should be utilized cautiously and better to
The ICC values16,26,27,38 and Bland-Altman be used in the form of a continuous variable rather than
plots 12,26,38,41 showed good correlation and fair to be categorized into BMI status due to the possibility of
agreement for weight and height measurements between BMI misclassification.
self-reported and direct-measured methods and this In summary, self-reported weight and height
concurred with findings from many other countries. can be used as an alternative method in large-scale
Similar results were also found in a separate analysis by surveys and epidemiological to quantify self-reported
DOI: 10.20473/amnt.v5i4.2021. 377-386.
BMI. A beneficial situation to consider is especially during 1 IPH. National Health and Morbidity Survey 2015
the current pandemic of Covid-19 where it is much easier (NHMS 2015).Vol. II: Non-Communicable
to obtain self-reported measurements via a telephone Diseases, Risk Factors & Other Health Problems.
interview rather than meeting face to face. 49 (Institute for Public Health, National
Strengths and Limitations Institutes of Health, Ministry of Health,
This study is one of the first to validate self- Malaysia 2015).
reported weight, height and resultant BMI amongst late
adolescents entering early adulthood in Peninsular 2 ARoFIIN. Obesity In Asean: A Call To Action,
Malaysia. Apart from that, self-reporting as an alternative <http://www.arofiin.org/News/tid/57/Obesity-
method can be used in the future in large epidemiological in-ASEAN-A-Call-to-Action> (2017).
surveys and surveillance studies to calculate the resultant
BMI values and monitor the trend of nutritional status or 3 IPH. National Health and Morbidity Survey 2019
to assess the effectiveness of a health-related program. (NHMS 2019). 49 (Institute for Public Health,
This method will save time, manpower and cost. This is National Institutes of Health, Ministry of Health,
considered a cost-effective method with minimum Malaysia 2019).
resources to yield maximum benefits. In addition, these
results were beneficial in planning and developing new 4 IPH. National Health and Morbidity Survey 2011
programs, activities, studies or surveys focused on health (NHMS 2011). Vol. II: Non-Communicable
and anthropometric data that will increase compliance Diseases. 188 (Institute for Public Health,
and participation of the population in the program or National Institutes of Health, Ministry of Health,
large-scale surveys. Malaysia, 2011).
The study had several limitations. Some parents
were not willing to participate and did not allow 5 IPH. The Third National Health and Morbidity
researchers to reach/contact their kids. The far distance Survey (NHMS III) 2006, Nutritional Status. 130
(a few hundred kilometers) of some schools made it (National Institutes of Health, Ministry of
challenging to approach the participants. Apart from that, Health, Malaysia, 2008).
not all participants were comfortable allowing unknown
enumerators into their living or working place for 6 Thangiah, N. et al. Clustering and tracking the
anthropometric measurements. Some participants also stability of biological CVD risk factors in
furthered their studies or worked in another region that adolescents: the Malaysian health and
made a hindrance in data collection. adolescents longitudinal research team study
(MyHeARTs). 8, 69 (2020).
CONCLUSION
In general, the adolescents were able to report 7 Mohammadi, S. et al. Determinants of diet and
their height and weight precisely. Hence, this study physical activity in Malaysian adolescents: a
provides evidence that self-reported data could be systematic review. 16, 603 (2019).
considered for use in large epidemiology surveys or
studies as well as surveillance systems. The late
8 Mohammadi, S. et al. Dietary and physical
adolescent's height that was measured before them
activity patterns related to cardio-metabolic
leaving schools at the age17-18 years may remain the
health among Malaysian adolescents: a
same throughout adulthood. The current study has
systematic review. 19, 1-19 (2019).
shown self-reported weight and height are suggested to
be used at population level among Malaysian late
9 Sahoo, K. et al. Childhood obesity: causes and
adolescents and also during adulthood. This is important
consequences. 4, 187 (2015).
to assist with future large studies in monitoring
population weight and height.
10 Brener, N. D., McManus, T., Galuska, D. A.,
ACKNOWLEDGEMENT Lowry, R. & Wechsler, H. Reliability and validity
Hazreen Abdul Majid conceptualized the study of self-reported height and weight among high
and secured the funding. Kong Pun Pun conceptualized school students. The Journal of adolescent
the study and went for data collection, called the health : official publication of the Society for
participants and analyzed the data. The manuscript was Adolescent Medicine 32, 281-287 (2003).
written by Kong Pun Pun and Nithiah Thangiah.
11 Himes, J. H. Challenges of accurately measuring
CONFLICT OF INTEREST AND FUNDING DISCLOSURE and using BMI and other indicators of obesity in
This research was funded by IF017-2017. The authors children. Pediatrics 124 Suppl 1, S3-22,
declared no conflict of interest. doi:10.1542/peds.2008-3586D (2009).
REFERENCES 12 Wang, Z., Patterson, C. M. & Hills, A. P. A

comparison of self-reported and measured

DOI: 10.20473/amnt.v5i4.2021. 377-386.
height, weight and BMI in Australian 22 Giavarina, D. Understanding Bland Altman

adolescents. Australian and New Zealand analysis. Biochemia medica 25, 141-151,
journal of public health 26, 473-478 (2002). doi:10.11613/BM.2015.015 (2015).
13 Galfo, M., Censi, L., D'Addezio, L., Martone, D. 23 Connor Gorber, S., Tremblay, M., Moher, D. &
& Roccaldo, R. Validity of self-reported weight, Gorber, B. A comparison of direct vs. self-report
height and BMI in Italian adolescents for measures for assessing height, weight and body
assessing prevalence of overweight/obesity. mass index: a systematic review. Obesity
(2018). reviews : an official journal of the International
Association for the Study of Obesity 8, 307-326,
14 Ng, C. D. Biases in self-reported height and doi:10.1111/j.1467-789X.2007.00347.x (2007).
weight measurements and their effects on
modeling health outcomes. SSM - population 24 Seijo, M. et al. Comparison of self-reported and
health 7, 100405, directly measured weight and height among
doi:10.1016/j.ssmph.2019.100405 (2019). women of reproductive age: a systematic
review and meta-analysis. Acta obstetricia et
15 Hauck, F. R., White, L., Cao, G., Woolf, N. & gynecologica Scandinavica 97, 429-439,
Strauss, K. Inaccuracy of self-reported weights doi:10.1111/aogs.13326 (2018).
and heights among American Indian
adolescents. Annals of Epidemiology 5, 386- 25 Strauss, R. S. Comparison of measured and self-
392, doi:https://doi.org/10.1016/1047- reported weight and height in a cross-sectional
2797(95)00036-7 (1995). sample of young adolescents. International
journal of obesity and related metabolic
16 Fonseca, H. et al. Validity of BMI based on self- disorders : journal of the International
reported weight and height in adolescents. Acta Association for the Study of Obesity 23, 904-908
paediatrica (Oslo, Norway : 1992) 99, 83-88, (1999).
doi:10.1111/j.1651-2227.2009.01518.x (2010).
26 Aasvee, K. et al. Validity of self-reported height
17 Sherry, B., Jefferds, M. E. & Grummer-Strawn, L. and weight for estimating prevalence of
M. Accuracy of adolescent self-report of height overweight among Estonian adolescents: the
and weight in assessing overweight status: a Health Behaviour in School-aged Children
literature review. Archives of pediatrics & study. BMC research notes 8, 606,
adolescent medicine 161, 1154-1161, doi:10.1186/s13104-015-1587-9 (2015).
doi:10.1001/archpedi.161.12.1154 (2007).
27 Štefan, L., Baić, M. & Pekas, D. Validity of
18 Kee, C. C. et al. Validity of self-reported weight Measured vs. Self-Reported Height, Weight and
and height: a cross-sectional study among Body-Mass Index in Urban Croatian
Malaysian adolescents. BMC medical research Adolescents. 2, e89627,
methodology 17, 85, doi:10.1186/s12874-017- doi:10.5812/intjssh.89627 (2019).
0362-0 (2017).
28 Himes, J. H., Hannan, P., Wall, M. & Neumark-
19 Hazreen, M. A. et al. An exploratory study on Sztainer, D. Factors associated with errors in
risk factors for chronic non-communicable self-reports of stature, weight, and body mass
diseases among adolescents in Malaysia: index in Minnesota adolescents. Ann Epidemiol
overview of the Malaysian Health and 15, 272-278,
Adolescents Longitudinal Research Team study doi:10.1016/j.annepidem.2004.08.010 (2005).
(The MyHeART study). 14, 1-10 (2014).
29 Brettschneider, A. K., Rosario, A. S. & Ellert, U.
20 WHO. Obesity: preventing and managing the Validity and predictors of BMI derived from self-
global epidemic. Report of a WHO consultation. reported height and weight among 11- to 17-
World Health Organization technical report year-old German adolescents from the KiGGS
series 894, i-xii, 1-253 (2000). study. BMC research notes 4, 414,
doi:10.1186/1756-0500-4-414 (2011).
21 Altman, D. G. Practical Statistics For Medical
Research. 2 edn, Vol. Volume 12 of Texts in 30 Giacchi, M., Mattei, R. & Rossi, S. Correction of
statistical science series (Chapman and the self-reported BMI in a teenage population.
Hall,2015, 1991). International journal of obesity and related
metabolic disorders : journal of the
International Association for the Study of
Obesity 22, 673-677 (1998).
DOI: 10.20473/amnt.v5i4.2021. 377-386.
31 Tokmakidis, S. P., Christodoulos, A. D. & from individuals in the city of Sao Paulo. Revista
Mantzouranis, N. I. Validity of self-reported brasileira de epidemiologia = Brazilian journal
anthropometric values used to assess body of epidemiology 17, 735-746 (2014).
mass index and estimate obesity in Greek
school children. The Journal of adolescent 40 De Vriendt, T., Huybrechts, I., Ottevaere, C., Van
health : official publication of the Society for Trimpont, I. & De Henauw, S. Validity of self-
Adolescent Medicine 40, 305-310, reported weight and height of adolescents, its
doi:10.1016/j.jadohealth.2006.10.001 (2007). impact on classification into BMI-categories and
the association with weighing behaviour.
32 Engstrom, J. L., Paterson, S. A., Doherty, A., International journal of environmental research
Trabulsi, M. & Speer, K. L. Accuracy of self- and public health 6, 2696-2711,
reported height and weight in women: an doi:10.3390/ijerph6102696 (2009).
integrative review of the literature. Journal of
Midwifery & Women's Health 48, 338-345, 41 Elgar, F. J., Roberts, C., Tudor-Smith, C. &
doi:https://doi.org/10.1016/S1526- Moore, L. Validity of self-reported height and
9523(03)00281-2 (2003). weight and predictors of bias in adolescents.
The Journal of adolescent health : official
33 Bowring, A. L. et al. Measuring the accuracy of publication of the Society for Adolescent
self-reported height and weight in a Medicine 37, 371-375,
community-based sample of young people. doi:10.1016/j.jadohealth.2004.07.014 (2005).
BMC medical research methodology 12, 175,
doi:10.1186/1471-2288-12-175 (2012).
34 Olfert, M. D. et al. Self-Reported vs. Measured

Height, Weight, and BMI in Young Adults.
International journal of environmental research
and public health 15,
doi:10.3390/ijerph15102216 (2018).
35 Lim, L. L., Seubsman, S.-a. & Sleigh, A. J. P. H. M.

Validity of self-reported weight, height, and
body mass index among university students in
Thailand: Implications for population studies of
obesity in developing countries. 7, 1-8 (2009).
36 Kintziou, E., Nikolaidis, P. T., Kefala, V.,

Rosemann, T. & Knechtle, B. Validity of Self-
Reported Body Mass, Height, and Body Mass
Index in Female Students: The Role of Physical
Activity Level, Menstrual Cycle Phase, and Time
of Day. International journal of environmental
research and public health 16,
doi:10.3390/ijerph16071192 (2019).
37 Linhart, Y., Romano-Zelekha, O. & Shohat, T.

Validity of self-reported weight and height
among 13-14 year old schoolchildren in Israel.
The Israel Medical Association journal : IMAJ 12,
603-605 (2010).
38 Zhou, X., Dibley, M. J., Cheng, Y., Ouyang, X. &

Yan, H. Validity of self-reported weight, height
and resultant body mass index in Chinese
adolescents and factors associated with errors
in self-reports. BMC public health 10, 190,
doi:10.1186/1471-2458-10-190 (2010).
39 Carvalho, A. M., Piovezan, L. G., Selem, S. S.,

Fisberg, R. M. & Marchioni, D. M. Validation and
calibration of self-reported weight and height
Int. J. Morphol.,
36(4):1398-1406, 2018.
Anthropometry and Body Composition

of Adolescents in Macedonia
Antropometría y Composición Corporal de Adolescentes en Macedonia
Nazim Myrtaj1; Arben Maliqi1; Seryozha Gontarev2; Ruzdija Kalac2; Georgi Georgiev2 & Biljana Bojadzieva Stojanoska3
MYRTAJ, N.; MALIQI, A.; GONTAREV, S.; KALAC, R.; GEORGIEV, G. & STOJANOSKA, B. B. Anthropometry and body
composition of adolescents in Macedonia. Int. J. Morphol., 36(4):1398-1406, 2018.
SUMMARY: The aim of this research was to determine the level of adiposity and obesity among Macedonian adolescents and
to compare the results with previous studies conducted in this population, as well as those conducted in other populations. The sample
included 2390 adolescents from four urban different regions of R. Macedonia aged between 11 to 18 years; 1238 males and 1152
females. Weight, height, waist, and hip circumference (WC, HC) as well as triceps, calf, subscapular, and suprailiac skinfold thickness
(SFT) were measured. Body mass index (BMI), waist-to-hip ratio (WHR), waist-to-height ratio (WHtR), subscapular/triceps skinfold
ratio (STR), and percentage body fat were computed. The prevalence of overweight and obesity defined by the IOTF children growth
reference were calculated and age-dependent and gender-specific smoothed percentile curves for BMI and ROC curves were generated.
The boys have statistically significantly higher values of WC, WHR and WHtR in all adult categories (except WHtR at 18 years old),
greater body weight at the age of 12 to 18, and body weight 13 to 18 years (p<0.001). Weight, height and BMI are increasing with age in
both, boys and girls, and decreases in girls. The level of adiposity of Macedonian adolescents has increased over the past 20 years and
has reached the level of developed countries that face an obesity epidemic.
KEY WORDS: Macedonian adolescents; Skinfold thickness; Waist circumference; Waist-to-height ratio; Waist-to-hip ratio.
INTRODUCTION
The increased adiposity in children and adolescents various tools and indicators. These indicators can be used to
becomes a global health problem both in development and describe several types of obesity, but they also have different
the developing countries (Li et al., 2006; Senbanjo et al., limitations on the sensitivity and reliability of the data
2013). In 2010, around 43 million children worldwide were obtained. Therefore, care must be taken in selecting adequate
considered obese - 35 million in developing countries (de tools to be used and measure the obesity of the population
Onis et al., 2010). The global prevalence of overweight and level. The most commonly used tool is the Body Mass Index
obesity in children has increased from 4.2 % 1990 to 6.7 % (BMI) used by the World Health Organization as a standard
in 2010 and is expected to reach 9.1 % until 2020 (de Onis for recording obesity and constructing growth curves in
et al., 2010). If obesity occurs, especially in adolescence, it children and adolescents (de Onis et al., 2007). Despite the
remains in adulthood, which affects the overall health of the many advantages of this index, the BMI has some limitations.
individual by increasing the risk of developing diabetes It does not take into account many factors such as the
mellitus, hypertension, coronary artery disease and metabolic percentage of muscle tissue (Moreno et al., 2003), bone
syndrome (Lee et al., 2012). Furthermore, the increased level density, different proportions of fat, bone tissue water in the
of adiposity is associated with depression suggests that body. Also, a very anthropometric measure for assessing
psychological morbidity is also an important consequence central obesity in children and adolescents is the waist
(Murabito et al., 2013). Therefore, it is necessary to moni- circumference (WC). Conversely, the waist to hip ratio
tor the body parameters in children and adolescents in order (WHR) is not an accurate indicator of abdominal obesity in
to prevent the accumulation of excess body fat and its youth because of the weak correlation with central adiposity
associated morbidities. Body fat can be measured through (Taylor et al., 2000). Skin folds measurements (SFT) can be
1
Faculty of Physical Culture and Sports, AAB College, Prishtinë, Kosovo.
2
Faculty of Physical Education, Sport, and Health, Ss. Cyril and Methodius University, Skopje, Macedonia.
3
Institute of Anatomy, Ss. Cyril and Methodius University, Skopje, Macedonia.
1398
MYRTAJ, N.; MALIQI, A.; GONTAREV, S.; KALAC, R.; GEORGIEV, G. & STOJANOSKA, B. B. Anthropometry and body composition of adolescents in Macedonia.
Int. J. Morphol., 36(4):1398-1406, 2018.
used as specific indexes for estimating obesity due to their Anthropometry measurement. For evaluation
high sensitivity. Because of the latter, it is recommended to anthropometric variables were selected and measured
monitor obesity in children and adolescents (Hughes et al., according to the International Biological Program (IBP):
1997) and to allow an indirect assessment of total body fat body height, body weight, four skinfolds (triceps,
that can be used as a screening tool. According to Noyens et subscapular, supraspinale and calf), and two
al. (2007), the best indicator of obesity in body is subcapular circumferences (waist and hip). BMI was calculated as a
SFT, and in girls biceps SFT. In this study, the two measures ratio of body weight and height to square (kg/m2), WHR
were better predictors than the sum of the four SFTs. In was calculated as a ratio of the circumference of the waist
addition, obesity can be effectively evaluated by using the and hips and WHtR to the ratio of the circumference of
ratio of subscapular and triceps – SFT (Moreno et al., 2007). the waist and height %BF was calculated using the
Therefore, obesity in adulthood is better predicted by equations of Slaughter et al. (1988) - from the triceps and
adolescent skinfold thickness than by adolescent BMI. It is subscapular sites (SL1); and triceps and calf sites (SL2).
therefore important to check how other indexes such as The following standard anthropometric instruments were
skinfold thickness, might be applied to the Macedonian used: for measuring body height. Anthropometer by
population. Martin, with 1 mm reading accuracy; decimal weight scale;
“John Bull” caliper square for determination of skin-folds
The purpose of this study was to predict the level of with pressure of 10 g/cm2 and precision of 0.1 mm; elastic
adiposity and obesity in Macedonian adolescents. The band, also with 1 mm reading accuracy, for measuring
collected data can be used to document and describe the circumferences; and caliper square for measuring of
changes in obesity in children and adolescents in Macedo- diameters with reading precision of 1 mm and weight scale
nia. Monitoring the anthropometric changes at the population for measuring body weight. Anthropometric measurements
level can be crucial in prevention of negative health outcomes were made during school hours, not interrupting the
related to adiposity. The regular anthropometric estimates lessons. Subjects were standing, facing ahead, and body
of the population can provide the key information that will height was measured as maximum distance from the floor
allow health institutions to determine the points at which to the highest point on the head. Shoes were off, both feet
they should begin mounting interventions. together, and arms at the sides. Heels, buttocks and upper
back were in contact with the wall. Body height
measurement can vary throughout the day, usually being
MATERIAL AND METHOD higher in the morning, so to ensure reliability we measured
height at the same time of the day.
Participants. The sample included 2390 adolescents from Statistical analysis. Data was presented as the mean and
four urban different regions of R. Macedonia aged between standard deviation (SD). Sex differences in anthropometric
11 to 18 years; 1238 males and 1152 females. Participants characteristics were analyzed by one-way analysis of
from selected schools and classes were informed about the variance (ANOVA). Correlation coefficients were
objective of the study and their parents signed the informed calculated to measure statistical dependence between
consent for participation in the research. All adolescents were anthropometric indices and age. The relation between
healthy at the time of this study. In order to avoid errors in WHtR, four skinfolds (triceps, subscapular, supraspinale
the selection of the sample, volunteer students were not and calf), sum of four skinfolds, body fat % and
included. overweight⁄obesity as defined by the IOTF was
investigated with ROC analysis. The discriminating power
In the sample entered all students for whom the was expressed as area under the curve (AUC). Smoothed
parents gave their consent to participate in the project and age-and sex-specific table and graph percentiles were
were psycho-physically healthy and regularly attended constructed for BMI by the LMS method. This estimates
classes for physical and health education. The respondents the measurement centiles in terms of three age-sex-specific
were treated in accordance with the Helsinki Declaration. cubic spline curves: the L curve (Box-Cox power to remove
skewness), M curve (median) and S curve (coefficient of
The measurement was realized in March, April and variation). For the construction of the percentile curves,
May 2016, in standard school conditions at regular classes data were imported into the Lms Chart Maker software (v.
in physical and health education. Measurement was 2.3; by Tim Cole and Huiqi Pan) and the L, M and S curves
realized by experts in the field of kinesiology and medici- estimated. Except for the LMS method calculations, we
ne, who were trained for measuring certain anthropometric used SPSS v. 22.0 software for Windows (SPSS, Chicago,
measures. Illinois, USA).
1399
Int. J. Morphol., 36(4):1398-1406, 2018.
RESULTS
The research was realized on a sample of 2390 SFT as well as the percentage of fat tissue (BF%) decreases
adolescents at the age of 11 to 18 years. The average age with age for boys. On the other hand, the values of
of the sample (±sd) was 14,4 (±2,27) years. Descriptive Subscapular, Supraspinale SFT, sum of 4 skinfolds and
statistical parameters for varaiables BW, Ht, BMI, WC, %BF increases with age in girls. Girls have statistically
HC, WHR and WHtR, in terms of sex and age, are shown significantly higher values of triceps, calf SFT and sum of
in Table I. The weight, height and BMI increase with age 4 skinfolds at the age of 13 to 18 years and higher values
for boys (respectively: r=.740; r=.604; r=.331. All p<0.001 of subscapular supraspinale SFT at the age of 13 to 17
and for girls (respectively: r=.468; r=.435; r=.286. All years. Also, girls have higher values of %BF in all age
p<0.001). WHR remains constant for boys with age, and categories.
decreases for girls (r=.111, p<0.001). WHtR NS remain
constant with age for boys and girls. Boys have statistically In Table III and Figure 1 the smoothed sex and age-
significantly higher values of WC, WHR, WHtR in all age specific percentile values are shown at the 5rd, 10th, 25th,
categories (except WHtR at 18 years old), a greater body 50 th, 75 th, 75th, 85th and 95 th percentiles, which were
weight at the age of 12 to 18 years and a body height of 13 developed and smoothed by the LMS method. According
to 18 years (p<0.001). Also, boys have statistically to the IOTF study BMI cut-off norms there were 362 (28.3
significant BMI values at the age of 15, 16 and 18 years %) males and 207 (17.9 %) females considered overweight
(p<0.01). (including obesity), which represents 23.8 % of all
participants. Obese status was determined in 89 (7.2 %)
Skinfolds measurements values and BF% are boys and 28 (2.4 %) girls. These values are presented in
presented in Table II. Values of triceps, subscapular, calf the Table IV.
Table I. Mean values (±SD) of body measurements in male and female Macedonian adolescents aged 11–18.
Age
Number BW (kg) Ht (cm) BMI (kg/m2) WC (cm) HC (cm) WHR WHtR
(y)
Boys
11 167 44.02 11.19 151.13 7.31 19.08 3.65 69.01 9.68 82.32 8.71 0.84 0.05 0.46 0.06
12 150 51.62 14.43 158.23 9.83 20.34 4.14 71.79 10.97 87.28 10.69 0.82 0.06 0.45 0.06
13 140 54.69 13.72 164.46 10.35 20.00 3.55 71.26 9.38 88.16 8.75 0.81 0.06 0.43 0.05
14 159 62.20 14.69 170.02 8.11 21.34 4.11 75.45 11.37 92.55 9.76 0.81 0.07 0.44 0.06
15 164 67.66 13.63 173.63 7.14 22.41 4.24 78.04 11.22 94.01 9.53 0.83 0.06 0.45 0.06
16 161 69.26 11.81 175.52 6.31 22.50 3.91 77.94 9.35 92.96 8.39 0.84 0.06 0.44 0.06
17 160 71.74 13.72 177.89 7.08 22.62 3.87 79.49 11.00 94.90 8.10 0.84 0.07 0.45 0.06
18 137 75.12 11.56 179.24 7.04 23.35 3.14 79.93 8.50 97.54 6.74 0.81 0.09 0.45 0.05
Total 1238 62.04 13.09 168.76 7.89 21.45 3.83 75.36 10.18 91.22 8.83 0.82 0.06 0.45 0.06
Girls
11 148 42.70 10.87 151.58 7.09 18.41 3.54 64.23 8.55 83.26 9.48 0.77 0.05 0.42 0.05
12 151 48.72 9.67 157.24 7.50 19.64 3.38 67.41 7.50 88.45 7.83 0.76 0.05 0.43 0.05
13 146 51.69 9.19 160.78 6.54 19.94 3.04 67.80 7.91 90.75 7.61 0.75 0.06 0.42 0.05
14 146 56.13 9.24 163.14 6.54 21.07 3.08 69.94 7.63 94.38 7.06 0.74 0.05 0.43 0.05
15 148 56.21 9.21 163.08 5.96 21.12 3.23 70.69 7.75 93.46 7.45 0.75 0.08 0.43 0.05
16 150 56.29 8.21 163.33 6.34 21.09 2.71 70.15 7.61 93.28 6.50 0.75 0.06 0.43 0.05
17 131 58.66 9.32 163.45 5.53 21.93 3.16 71.98 8.60 94.51 7.20 0.76 0.07 0.44 0.05
18 132 57.52 7.84 164.78 6.02 21.19 2.76 69.48 6.03 94.45 5.88 0.74 0.04 0.42 0.04
Total 1152 53.49 9.19 160.92 6.44 20.55 3.11 68.96 7.70 91.57 7.38 0.75 0.06 0.43 0.05
1400
Table II. Mean values (±SD) of skinfold thickness and percentage body fat in male and female Macedonian adolescents aged 11–18.
Subscapular Body Body
Age(y) Num Triceps Subscapular S upraspinale Calf ∑ Skinfolds
/triceps Fat1 % Fat2 %
Boys
11 167 15.57 6.86 10.42 6.57 12.15 7.94 13.49 6.61 51.63 26.51 0.65 0.18 24.30 9.83 18.57 9.27
12 150 15.70 7.85 11.10 6.99 13.46 9.31 14.49 7.51 54.74 29.98 0.70 0.19 22.59 11.47 20.17 9.68
13 140 12.91 6.27 9.13 5.59 10.68 7.67 11.80 5.98 44.52 24.12 0.72 0.21 18.75 9.22 16.79 7.79
14 159 12.62 6.61 10.17 6.02 11.89 8.16 11.76 5.98 46.45 25.44 0.83 0.23 19.37 9.84 17.12 8.41
15 164 12.88 7.44 10.81 6.06 11.94 7.90 12.67 6.57 48.30 26.68 0.89 0.24 18.36 11.06 18.25 8.96
16 161 12.15 5.85 10.66 4.39 10.97 6.09 11.36 5.22 45.13 19.61 0.93 0.25 17.71 8.42 17.18 6.55
17 160 12.45 7.36 11.55 5.49 12.05 7.49 11.56 5.69 47.61 24.12 1.03 0.35 18.67 10.39 17.98 7.68
18 137 12.09 4.84 11.84 4.22 12.65 6.36 10.70 4.61 47.29 17.81 1.03 0.27 18.63 7.16 17.57 5.91
Total 1238 13.32 6.85 10.71 5.80 11.97 7.70 12.25 6.19 48.25 24.78 0.85 0.28 19.83 10.03 17.97 8.20
Girls
11 148 15.58 6.47 10.70 5.50 12.13 6.69 13.84 5.97 52.25 23.30 0.69 0.17 26.76 5.84 20.07 6.67
12 151 16.24 6.19 11.24 5.56 13.47 6.49 14.58 5.73 55.53 22.48 0.70 0.18 27.93 5.63 20.63 6.80
13 146 16.18 6.40 11.85 4.81 13.62 6.06 13.92 4.96 55.58 20.43 0.75 0.16 28.26 5.47 20.61 5.94
14 146 18.11 5.90 13.08 5.09 15.01 5.98 15.41 5.27 61.61 19.62 0.73 0.17 29.81 4.65 22.48 5.59
15 148 17.50 5.30 13.04 4.70 14.46 5.47 14.72 5.05 59.72 17.93 0.76 0.19 29.70 3.90 22.04 5.45
16 150 16.87 4.88 12.77 4.35 13.99 5.24 14.95 4.67 58.57 16.13 0.77 0.16 29.24 3.70 22.01 4.69
17 131 17.72 5.78 14.23 5.28 15.09 5.62 16.30 5.64 63.34 19.41 0.83 0.25 30.33 4.26 23.72 5.96
18 132 16.18 4.65 12.63 4.46 14.00 5.11 13.94 4.82 56.75 16.02 0.80 0.24 29.33 3.81 21.31 5.12
Total 1152 16.79 5.79 12.41 5.09 13.95 5.92 14.70 5.32 57.85 19.85 0.75 0.20 28.89 4.86 21.58 5.91
Int. J. Morphol., 36(4):1398-1406, 2018.
obesity.
Fig. 1. Smoothed percentiles of BMI for Macedonian adolescents Boys (A), Girls (B).
indicators for predicting obesity on the basis of IOTF
supraspinale SFT (AUC: 0.879± 0.012; 95 % CI

(AUC: 0.936± 0.007; 95 % CI 0.922-0.955) and
0.898± 0.012; 95 % CI 0.875-0.927) and then the

(area under the curve [AUC]: 0.941± 0.007; 95 % CI
study BMI cut-offs are shown in Figure 2 and Table
The ROC curves of measured and computed
measure sum of 4 skinfolds (AUC: 0.907± 0.010; 95

WHtR (AUC: 0.929±0.009; 95 % CI 0.912-0.946).
skinfolds for male respondents have the greatest
V. ROC analysis show that the measure sum of 4
0.855-0.903) values were the best predictors of

% CI 0.887-0.927), followed by the WHtR (AUC:
For girls, the greatest discriminatory power is the
0.927-0.955), followed by the supraspinale SFT
discriminating power in predicting IOTF obesity
1401
Int. J. Morphol., 36(4):1398-1406, 2018.
Table III. Percentile values of body mass index (BMI) in male and female Macedonian adolescents aged 11–18 for the 85th and 95th
percentiles.
Age (yr)
L M S 5 th 10th 25th 50th 75th 85th 90th 95th
Boys
11 -0.70 18.61 0.18 14.17 14.98 16.53 18.61 21.17 22.81 24.04 26.09
12 -0.99 19.32 0.18 14.96 15.75 17.26 19.32 21.93 23.64 24.97 27.22
13 -1.19 19.70 0.17 15.44 16.20 17.67 19.70 22.32 24.07 25.43 27.81
14 -1.43 20.48 0.16 16.30 17.04 18.48 20.48 23.11 24.89 26.31 28.82
15 -1.57 21.49 0.15 17.37 18.10 19.52 21.49 24.07 25.82 27.21 29.67
16 -1.55 21.79 0.15 17.71 18.44 19.85 21.79 24.29 25.98 27.31 29.63
17 -1.50 22.02 0.14 17.98 18.71 20.11 22.02 24.47 26.09 27.35 29.53
18 -1.28 22.87 0.13 18.86 19.60 21.01 22.87 25.15 26.60 27.70 29.53
Girls
11 -1.10 17.83 0.17 13.93 14.63 15.98 17.83 20.19 21.76 22.97 25.05
12 -0.57 19.14 0.16 14.94 15.73 17.21 19.14 21.42 22.82 23.86 25.54
13 -0.53 19.73 0.15 15.63 16.42 17.87 19.73 21.91 23.23 24.19 25.74
14 -0.69 20.56 0.14 16.59 17.35 18.76 20.56 22.69 23.98 24.93 26.46
15 -0.80 20.81 0.14 16.94 17.68 19.05 20.81 22.89 24.17 25.11 26.63
16 -0.88 202.95 0.13 17.16 17.88 19.22 20.95 23.00 24.26 25.19 26.70
17 -0.97 21.06 0.13 17.35 18.06 19.37 21.06 23.07 24.31 25.23 26.73
18 -1.07 21.08 0.13 17.49 18.17 19.44 21.08 23.04 24.26 25.16 26.63
Table IV. The prevalence of overweight and obesity in Macedonian

boys and girls aged 11–18
n Normal Overweight Obese
Boys 1238 876 273 89
70,8% 22,1% 7,2%
Girls 1152 945 179 28
82,0% 15,5% 2,4%
Total 2390 1821 452 117
76,2% 18,9% 4,9%
c2 =50,94; p<0.001
Table V. Cut-Points of anthropometric variables in Adolescents According to Sex.

Sensitivity Specificity
AUC (95% CI) Cut-point
(95% CI)
Male
Triceps 0,893 0,874 to 0,909 >16,8 66,85 92,92
Subscapular 0,928 0,912 to 0,941 >12,0 77,07 92,58
Supraspinale 0,936 0,921 to 0,949 >14,2 81,22 90,06
Calf 0,894 0,875 to 0,910 >15,8 65,19 93,84
∑ Skinfolds 0,941 0,926 to 0,953 >56,4 81,49 92,81
Body Fat1 % 0,921 0,904 to 0,935 >24,5 70,99 93,38
Body Fat2 % 0,935 0,920 to 0,948 >20,3 79,28 93,72
WHtR 0,929 0,913 to 0,943 >0,50 78,67 92,12
WC 0,906 0,888 to 0,922 >81,0 69,25 93,38
Female
Triceps 0,866 0,845 to 0,885 >25,4 32,37 97,35
Subscapular 0,865 0,843 to 0,884 >17,8 51,21 94,39
Supraspinale 0,879 0,859 to 0,897 >20,2 49,28 95,13
Calf 0,858 0,836 to 0,877 >23,4 26,09 98,20
∑ Skinfolds 0,907 0,889 to 0,923 >82,4 47,83 97,14
Body Fat1 % 0,851 0,828 to 0,870 >33,7 38,65 97,88
Body Fat2 % 0,888 0,868 to 0,905 >28,2 54,11 95,45
WHtR 0,898 0,879 to 0,915 >0,5 64,25 92,69
WC 0,866 0,845 to 0,885 >81,0 29,95 98,62
1402
Int. J. Morphol., 36(4):1398-1406, 2018.
Fig. 2. ROC curves (boys - A, girls - B) for predicting obesity according to the growth reference figure obtained for waist-to-height ratio
(WHtR), percentage body fat (%BF), calf, triceps, subscapular, suprailiac skinfolds and their sum.
DISCUSSION
The aim of this study was to determine the level of Furthermore, the result of the research indicate that
adiposity and obesity among Macedonian adolescents using 22.1 % of boys and 15.5 % of girls have overweight
several anthropometric measures and to compare the (according to the BMI classification), which represents 18.9
obtained results with the results obtained from adolescents % of the total number of adolescents. On the other hand 7.2
from different geographical regions and ethnic backgrounds. % of boys and 2.4 % of girls are classified as obese
BMI cut-offs from the IOTF study were used to establish (BMI>p95), which represents 4.9 % of the total number of
overweight and obesity status. Comparing our results with adolescents. The cumulative percentage of obesity and
global collected data over the past 17 years has shown many overweight status was 29.2 % for boys and 18.0 % for girls
differences in the measured parameters. The BMI values of (23.8 % for all adolescents). Because of the different
the Macedonian boys in all age groups were higher than in methodology used to define overweight and obesity in
the boys from Nigeria (Senbanjo et al.), Bahrain (Gharib & adolescents, caution should be exercised when comparing
Shah, 2009), Bolivia (Baya Botti et al., 2009), Hong Kong - data from different studies. Therefore, the comparison was
except at the age of 14 years (Sung et al., 2008). The limited to those studies using a similar methodology as in
Macedonian boys showed higher values in BMI than German this study. Using the cut-off points recommended by IOTF,
boys at the age of 14 to 18 years, the Bulgarian (Galcheva et Lobstein & Frelut (2003) present the percentage of
al., 2009) and Koreans at the age of 15 to 18 years and lower overweight (overweight includes obese) adolescents aged
in relation to the Cyprus (Savva et al., 2001) and Kuwait boys around 11–18 years in various countries: Slovakia (8 %),
(Department of Food and Nutrition Administration, 2005) in Russia (9 %), Czech Republic (9 %), Netherlands (11 %),
all age categories. In girls, the BMI values were lower than Poland (12 %), Germany (13 %), Denmark (17 %),
girls from Bahrain, Bolivia and Kuwait in all age categories, Bulgaria (17 %), Croatia and Zagreb (20 %), Great Britain
and lower than the Bulgarian adolescents at the age of 11 to (21 %), Spain (21 %), Greece and Thessaloniki (22 %) and
15 years, German at the age of 11, 13 and 16 years. Cyprus (23 %). Compared to the above countries
Macedonian girls showed higher BMI values in all age Macedonian adolescents in this study seem to fall near the
categories than adolescents from Hong Kong, Nigeria and middle-high of the 10-34 % range, presented (23.8 %).
Norway (except at the age of 16 and 18 years). In most cases Comparison of the results of this study with those of the
these differences were small and dependent on age and sex. Arab counties (Musaiger et al., 2016) reveals a very high
1403
Int. J. Morphol., 36(4):1398-1406, 2018.
prevalence of overweight as shown by the high combined cm in girls. In Hong Kong Chinese from 74.0 to 81.6 cm in
percentage of overweight and obese individuals in Kuwait boys and from 68.4 to 72.6 cm in girls, in German adolescents
(60.4 %; 41.4 %), United Arab Emirates, the city Sharjah from 80.4 to 91.8 cm in boys and from 79.4 to 85.4 cm in
(38.9 %; 20.2 %) and Jordan, Haman (31.8 %; 22.1 %) both girls, in Norwegian adolescents from 71.3 to 86.7 cm in boys
for boys and girls. Our result comparison %BF with the and from 69.3 to 76.8 cm in girls, in Poland adolescents
German ones (Haas et al., 2011) indicate a higher percentage from 75.5 to 86.5 cm in boys and from 71.9 to 78.8 cm in
of fatty tissue among Macedonian adolescents (12-18 years). girls, in Kuwait adolescents from 87.1 to 106.7 cm in boys
and from 86.7 to 107.0 cm in girls.
Also, the results of our study suggests that boys tend
to have a greater prevalence of obesity compared to girls ROC analysis show that the measures sum of 4
(29.2 % for boys and 18.0 % for girls). This is probably due skinfolds have the greatest discriminating power in predicting
to the fact that in this adult period girls are more concerned the IOTF obesity for Macedonian adolescents of both sexes.
about their appearance and more attentive to the diet. On Good descriptive power also shows the variables WHtR and
the other hand, boys are less engaged in spontaneous physical supraspinale SFT. The researches conducted in German (Haas
activities and spend most of their free time in sedentary et al.) and Portugal (Sardinha et al., 1999) adolescents showed
activities (compared work, watching television, etc.). similar relations. The variable WHtR has shown the best
descriptive power to predict abdominal obesity among
The results of the study indicate that the highest German boys and girls. Then were followed %BF, subscapular
percentage of overweight including obesity have boys at the and sum of 4 skinfolds in boys and WHR, sum of 4 skinfolds
age of 12 years (34.0 %) and girls at the age of 14 years and %BF in girls (Haas et al.). In older adolescents at the age
(24.0 %), while the smallest percentage has 13-year-old boys of 12 to 15 years sum of 4 skinfolds was moving from
(20.7 %) and 16-year-old girls (10.0 %). Over 24.0 % of the 0.94±0.045 to 0.86±0.087 in boys and from 0.94±0.034 to
respondents have sum of 4 skinfolds greater than cut-off 0.95±0.036 in girls. The anthropometric measure triceps SFT
values. This high percent of body fat is associated with an had the best descriptive predictive power in obesity for boys
increased risk of acute and chronic diseases, particularly and girls of this age (Sardinha et al.).
osteoarthritis, increased blood pressure, diabetes mellitus and
cardiovascular disease, which can lead to poor quality of On the basis of the results of ROC analysis and the
life, increased personal and financial burden for the indivi- comparisons presented above, can be concluded that the sum
dual, society and shortening the life span (Dugan, 2008). of 4 skinfolds and WHtR should be used as the preferred
The determined differences in the degree of obesity may be screening tools. Sum of 4 skinfolds is easy to perform
due to the ethnic and genetic differences between the predicts later adult body fatness better than adolescent BMI
surveyed populations, the geographical and socio-economic does. Stomfai et al. (2011) determined that the intra-observer
conditions, the variations in the composition of the food, reliability for SFT measures (triceps, subscapular, biceps,
the forms of food intake and physical activity. suprailiac) was 97.7 %, and inter-observer technical error
of measurement for skinfold thicknesses was between 0.13
Because there are significant differences in the four and 0.97 mm. Also, because %BF can be measured on the
most commonly used anatomical parts for WC measures (de basis of SFT measurements these measurements seem to
Onis et al., 2010), it is difficult to make international be very useful screening tools along with WHtR measures.
comparisons of WC, and to assess the prevalence of
metabolic syndrome. For example, the prevalence of The advantages of this study is that this is one of the
metabolic syndrome as predicted by WC measurements first researches realized on Macedonian adolescents.
differed by 3 % among the measurement sites in men Anthropometric measures were collected by appropriately
between the umbilicus and minimal waist, whereas in trained health workers who used the same anatomical
women, the prevalence ranged from 15.1 % (umbilicus) to measurement points. In addition, the results are
14.4 % (iliac crest), 14.1 % (midpoint between iliac crest representative of today’s adolescents, as data on ITM were
and the lowest rib), and 13.1 % (minimal waist) (Go?ab collected in 2016. The limitations of the study consist in the
& Chrzanowska, 2002). Considering the predictive role of lack of information on the impact on sexual maturation sta-
WC in the IDF criteria for metabolic syndrome, and the tus on the anthropometric indices, as well as cross-sectional
differences in mean WC among different ethnicities, design of the study. Although, references curves are obtained
(Nooyens et al., 2007) international standardization of the on contemporary data that are likely to be representative of
protocol for measurement of WC is warranted. From 11 to the current situation in Macedonia, the validity of the
18 years in 90th percentile, WC for Macedonian adolescents obtained percentile curves should be confirmed in future
ranges from 80.8 to 90.2 cm for boys and from 73.4 to 78.5 research using longitudinal approach.
1404
Int. J. Morphol., 36(4):1398-1406, 2018.
CONCLUSIONS REFERENCES
The level of adiposity of Macedonian adolescents Baya Botti, A.; Pérez-Cueto, F. J.; Vasquez Monllor, P. A. & Kolsteren, P. W.
Anthropometry of height, weight, arm, wrist, abdominal circumference
has increased over the past 20 years and has reached the
and body mass index, for Bolivian adolescents 12 to 18 years: Bolivian
level of developed countries that face an obesity epidemic. adolescent percentile values from the MESA study. Nutr. Hosp., 24(3):304-
The obtained data from the research can be used for logical 11, 2009.
monitoring of the trends in body fat and obesity among de Onis, M.; Blössner, M. & Borghi, E. Global prevalence and trends of
overweight and obesity among preschool children. Am. J. Clin. Nutr.,
Macedonian adolescents and the spread of anthropometric
92(5):1257-64, 2010.
indicators with other populations. For the proper de Onis, M.; Onyango, A. W.; Borghi, E.; Siyam, A.; Nishida, C. & Siekmann,
monitoring of changes in the level of adulthood, it is J. Development of a WHO growth reference for school-aged children
necessary to repeat such researches every ten years. We and adolescents. Bull. World Health Organ., 85(9):660-7, 2007.
Department of Food and Nutrition Administration. Kuwait Nutritional
conclude that SFT measurements should be used as the
Surveillance System. Shuwaikh Port, Ministry of Health, 2005.
preferred screening tool, because they are simple to perform Dugan, S. A. Exercise for preventing childhood obesity. Phys. Med. Rehabil.
and are better predictors of adult body fatness than Clin. N. Am., 19(2):205-16, 2008.
adolescent BMI. On the basis of ROC analysis can be Galcheva, S. V.; Iotova, V. M.; Yotov, Y. T.; Grozdeva, K. P.; Stratev, V. K. &
Tzaneva, V. I. Waist circumference percentile curves for Bulgarian
concluded that the sum of 4 SFTs and WHtR had the best
children and adolescents aged 6-18 years. Int. J. Pediatr. Obes., 4(4):381-
discriminatory power to predict obesity. 8, 2009.
Gharib, N. M. & Shah, P. Anthropometry and body composition of school
children in Bahrain. Ann. Saudi Med., 29(4):258-69, 2009.
Goab, S. & Chrzanowska, M. Dziecko Krakowskie 2000. Poziom rozwoju
biologicznego dzieci i modziezy miasta Krakowa. Kraków, AWF, 2002.
MYRTAJ, N.; MALIQI, A.; GONTAREV, S.; KALAC, R.; Haas, G. M.; Liepold, E. & Schwandt, P. Percentile curves for fat patterning
GEORGIEV, G. & BOJADZIEVA, S. B. Antropometría y com- in German adolescents. World J. Pediatr., 7(1):16-23, 2011.
posición corporal de adolescentes en Macedonia. Int. J. Morphol., Hughes, J.; Li, L.; Chinn, S. & Rona, R. Trends in growth in England and
36(4):1398-1406, 2018. Scotland, 1972 to 1994. Arch. Dis. Child., 76(3):182-9, 1997.
Lee, J.; Chung, D. S.; Kang, J. H. & Yu, B. Y. Comparison of visceral fat and
RESUMEN: El objetivo de esta investigación consistió liver fat as risk factors of metabolic syndrome. J. Korean Med. Sci.,
en determinar el nivel de adiposidad y obesidad entre los adoles- 27(2):184-9, 2012.
Li, C.; Ford, E. S.; Mokdad, A. H. & Cook, S. Recent trends in waist
centes macedonios y comparar los resultados con estudios pre-
circumference and waist-height ratio among US children and adolescents.
vios realizados en esta población, así como aquellos realizados Pediatrics, 118(5):e1390-8, 2006.
en otras poblaciones. La muestra incluyó a 2390 adolescentes de Lobstein, T. & Frelut, M. L. Prevalence of overweight among children in
cuatro regiones urbanas diferentes de R. Macedonia con edades Europe. Obes. Rev., 4(4):195-200, 2003.
comprendidas entre 11 y 18 años (1238 hombres y 1152 muje- Moreno, L. A.; Joyanes, M.; Mesana, M. I.; González-Gross, M.; Gil, C. M.;
res). Se midieron el peso, la altura, la cintura y la circunferencia Sarría, A.; Gutierrez, A.; Garaulet, M.; Perez-Prieto, R.; Bueno, M.; Mar-
de la cadera (WC, HC), así como el grosor del pliegue cutáneo cos, A. & AVENA Study Group. Harmonization of anthropometric
tríceps, pantorrilla, subescapular y suprailíaco (SFT). Se calcu- measurements for a multicenter nutrition survey in Spanish adolescents.
Nutrition, 19(6):481-6, 2003.
laron el índice de masa corporal (IMC), la relación cintura-cade-
Moreno, L. A.; Mesana, M. I.; González-Gross, M.; Gil, C. M.; Ortega, F. B.;
ra (WHR), la relación cintura-altura (WHtR), la relación de plie- Fleta, J.; Wärnberg, J.; León, J.; Marcos, A. & Bueno, M. Body fat
gue subcutáneo / tríceps (STR) y el porcentaje de grasa corporal. distribution reference standards in Spanish adolescents: the AVENA Study.
Se calcularon las prevalencias de sobrepeso y obesidad defini- Int. J. Obes. (Lond.), 31(12):1798-805, 2007.
das por la referencia de crecimiento de niños IOTF y se genera- Murabito, J. M.; Massaro, J. M.; Clifford, B.; Hoffmann, U. & Fox, C. S.
ron curvas percentiles suavizadas dependientes de la edad y de Depressive symptoms are associated with visceral adiposity in a
sexo para las curvas BMI y ROC. Los niños tuvieron valores community-based sample of middle-aged women and men. Obesity (Silver
estadísticamente significativamente más altos de WC, WHR y Spring), 21(8):1713-9, 2013.
Musaiger, A. O.; Al-Mannai, M.; Al-Haifi, A. R.; Nabag, F.; Elati, J.;
WHtR en todas las categorías de adultos (excepto WHtR a los
Abahussain, N.; Tayyem, R.; Jalambo, M.; Benhamad, M. & Al-Mufty,
18 años), mayor peso corporal a la edad de 12 a 18 años y peso B. Prevalence of overweight and obesity among adolescents in eight Arab
corporal de 13 a 18 años (p <0,001). El peso, la estatura y el IMC countries: comparison between two international standards (ARABEAT-
aumentan con la edad tanto en niños como en niñas y disminu- 2). Nutr. Hosp., 33(5):567, 2016.
yen en las niñas. El nivel de adiposidad de los adolescentes Nooyens, A. C.; Koppes, L. L.; Visscher, T. L.; Twisk, J. W.; Kemper, H. C.;
macedonios ha aumentado en los últimos 20 años y ha alcanzado Schuit, A. J.; van Mechelen, W. & Seidell, J. C. Adolescent skinfold
el nivel de los países desarrollados que enfrentan una epidemia thickness is a better predictor of high body fatness in adults than is body
de obesidad. mass index: the Amsterdam Growth and Health Longitudinal Study. Am.
J. Clin. Nutr., 85(6):1533-9, 2007.
Sardinha, L. B.; Going, S. B.; Teixeira, P. J. & Lohman, T. G. Receiver
PALABRAS CLAVE: Adolescentes macedonios; Es- operating characteristic analysis of body mass index, triceps skinfold
pesor del pliegue cutáneo; Circunferencia de la cintura; Re- thickness, and arm girth for obesity screening in children and adolescents.
lación cintura-altura; Relación cintura-cadera. Am. J. Clin. Nutr., 70(6):1090-5, 1999.
1405
Int. J. Morphol., 36(4):1398-1406, 2018.
Savva, S. C.; Kourides, Y.; Tornaritis, M.; Epiphaniou-Savva, M.; Tafouna, Correspondence to:
P. & Kafatos, A. Reference growth curves for cypriot children 6 to 17 Seryozha Gontarev
years of age. Obes. Res., 9(12):754-62, 2001. University “Ss. Cyril and Methodius”
Senbanjo, I. O.; Oshikoya, K. A.; Olutekunbi, O. A. & Njokanma, O. F.
Faculty of Physical Education
Body fat distribution of children and adolescents in Abeokuta,
Southwest Nigeria. Am. J. Phys. Anthropol., 150(4):647-54, 2013.
Sport and Health
Slaughter, M. H.; Lohman, T. G.; Boileau, R. A.; Horswill, C. A.; Stillman, Skopje
R. J.; Van Loan, M. D. & Bemben, D. A. Skinfold equations for REPUBLIC OF MACEDONIA
estimation of body fatness in children and youth. Hum. Biol., 60(5):709-
23, 1988.
Stomfai, S.; Ahrens, W.; Bammann, K.; Kovács, E.; Mårild, S.; Michels, Email:gontarevserjoza@gmail.com
N.; Moreno, L. A.; Pohlabeln, H.; Siani, A.; Tornaritis, M.; Veidebaum,
T.; Molnár, D. & IDEFICS Consortium. Intra- and inter-observer
reliability in anthropometric measurements in children. Int. J. Obes.
(Lond.), 35 Suppl. 1:S45-51, 2011.
Received: 17-02-2018
Sung, R. Y.; So, H. K.; Choi, K. C.; Nelson, E. A.; Li, A. M.; Yin, J. A.; Accepted: 14-06-2018
Kwok, C. W.; Ng, P. C. & Fok, T. F. Waist circumference and waist-to-
height ratio of Hong Kong Chinese children. B. M. C. Public Health,
8:324, 2008.
Taylor, R. W.; Jones, I. E.; Williams, S. M. & Goulding, A. Evaluation of
waist circumference, waist-to-hip ratio, and the conicity index as
screening tools for high trunk fat mass, as measured by dual-energy X-
ray absorptiometry, in children aged 3-19 y. Am. J. Clin. Nutr.,
72(2):490-5, 2000.
1406
Rev Paul Pediatr. 2015;33(1):56–62
REVISTA PAULISTA
DE PEDIATRIA
www.rpped.com.br
ORIGINAL ARTICLE
Anthropometric indicators of obesity in the prediction

of high body fat in adolescents
Andreia Pelegrinia,*, Diego Augusto Santos Silvab,

João Marcos Ferreira de Lima Silvac, Leoberto Grigollod,
Edio Luiz Petroskib
a
Universidade do Estado de Santa Catarina (UDESC), Florianópolis, SC, Brazil
b
Universidade Federal de Santa Catarina (UFSC), Florianópolis, SC, Brazil
c
Faculdade Leão Sampaio (FALS), Juazeiro do Norte, CE, Brazil
d
Universidade do Oeste de Santa Catarina (UNOESC), Joaçaba, SC, Brazil
Received 18 April 2014; accepted 6 August 2014
KEYWORDS Abstract
Anthropometry; Objective: To determine the anthropometric indicators of obesity in the prediction of
Body fat distribution; high body fat in adolescents from a Brazilian State.
Students Methods: The study included 1,197 adolescents (15-17 years old). The following anthro-
pometric measurements were collected: body mass (weight and height), waist circumfe-
rence and skinfolds (triceps and medial calf). The anthropometric indicators analyzed
were: body mass index (BMI), waist circumference (WC), waist-to-height ratio (WHtR)
and conicity index (C-Index). Body fat percentage, estimated by the Slaughter et al equa-
tion, was used as the reference method. Descriptive statistics, U Mann-Whitney test, and
ROC curve were used for data analysis.
Results: Of the four anthropometric indicators studied, BMI, WHtR and WC had the lar-
gest areas under the ROC curve in relation to relative high body fat in both genders. The
cutoffs for boys and girls, respectively, associated with high body fat were BMI 22.7 and
20.1kg/m², WHtR 0.43 and 0.41, WC 75.7 and 67.7cm and C-Index 1.12 and 1.06.
Conclusions: Anthropometric indicators can be used in screening for identification of
body fat in adolescents, because they are simple, have low cost and are non-invasive.
© 2014 Sociedade de Pediatria de São Paulo. Published by Elsevier Editora Ltda. All rights
reserved.
*Corresponding author.
E-mail: andreia.pelegrini@udesc.br (A. Pelegrini).
DOI of refers to article: http://dx.doi.org/10.1016/j.rpped.2014.06.007
1984-1462/© 2014 Sociedade de Pediatria de São Paulo. Published by Elsevier Editora Ltda. All rights reserved.
PALAVRAS-CHAVE Indicadores antropométricos de obesidade na predição de gordura corporal elevada
Antropometria; em adolescentes
Distribuição de gordura
corporal; Resumo
Estudantes Objetivo: Determinar os indicadores antropométricos de obesidade na predição da
gordura corporal elevada em adolescentes de um estado brasileiro.
Métodos: O estudo incluiu 1.197 adolescentes (15-17 anos). Foram coletadas medidas
antropométricas: massa corporal e estatura, perímetro da cintura e dobras cutâneas
(tríceps e perna medial). Os indicadores antropométricos analisados foram: índice de
massa corporal (IMC), perímetro da cintura (PC), razão cintura-estatura (RCE) e índice de
conicidade (ÍndiceC). A gordura corporal elevada, estimada pela equação de Slaughter et
al que foi usada como método de referência. Estatística descritiva, teste U de Mann-
Whitney e curva ROC foram utilizadas para a análise dos dados.
Resultados: Dos indicadores antropométricos estudados, o IMC e RCE e o PC tiveram as
maiores áreas sob a curva ROC em relação ao gordura corporal relativa elevada em ambos
os sexos. Os pontos de corte para os rapazes e moças, respectivamente, associados com
gordura corporal elevada foram IMC 22,7 e 20,1kg/m², RCE 0,43 e 0,41 e PC de 75,7 e
67,7cm e ÍndiceC de 1,12 e 1,06.
Conclusões: Os indicadores antropométricos podem ser usados como ferramenta para
identificação da gordura corporal em adolescentes, por ser um método simples, de baixo
custo e não invasivo.
© 2014 Sociedade de Pediatria de São Paulo. Publicado por Elsevier Editora Ltda. Todos os
direitos reservados.
Introduction excess body fat.11-14 In Brazil, few studies have investigated

the ability of each indicator to detect excess body fat in
Overweight has been considered an important public health adolescents,15,16 however, studies using anthropometric
problem worldwide.1 Evidence consistently reports that indicators to predict high blood pressure17 and hypertension
there is a greater likelihood of overweight and obese ado- stand out.18 Both in Brazil and in other countries, no studies
lescents to become obese adults.2 In this context, obesity investigating the C-Index for the prediction of high body fat
in childhood and in adolescence is considered an indepen- were found. In this sense, there are discussions about what
dent risk factor in the development of cardiovascular dis- would be the best anthropometric index for predicting high
eases in adulthood.3 body fat, regardless of sex, age and total body fat.
Numerous methods have been used to assess body com- Therefore, more empirical evidence is needed, especially
position.4 Among indirect methods, hydrostatic weighing in adolescence. Thus, this study aims to verify the diagnos-
and dual energy X-ray absorptiometry (DEXA) stand out; tic performance of anthropometric indicators of obesity in
however, they are more difficult to be used in large sam- the prediction of high body fat in adolescents.
ples due to the high cost and the need for a qualified tech-
nical team for assessing the measurements.5 Among double
indirect methods, anthropometry is considered a simple, Methods
rapid, inexpensive method that can be applied to a great
number of individuals.6 This cross-sectional epidemiological study included school-
Many anthropometric indicators have been proposed to children aged 15-17 years enrolled in public schools (state
diagnose the health risks taking into account the increased and federal) in the Brazilian state of Santa Catarina. The
body fat.7 The most widely used is still the body mass index study was approved by the Ethics Committee on Human
(BMI), but it has some limitations.8 However, other indica- Research of the Federal University of Santa Catarina (pro-
tors have been recommended. Waist circumference (WC) is tocol number 372/2006) and University of Western Santa
one of the measures proposed to achieve results closer to Catarina (protocol number 079/08).
reality, since abdominal fat deposits also cause, alone, var- To conduct the survey, two regions were considered: 1) a
ious health problems.9 The waist-to-height ratio (WHtR)10 survey was conducted in 2007 in Florianópolis, capital of
and the conicity index (C-Index) have also been used as the state of Santa Catarina, located in southern Brazil.
indicators to diagnose body fat. Florianópolis has a population of approximately 420,000
Some studies have been conducted with children and inhabitants,19 and is considered one of the Brazilian cities
adolescents in order to analyze the performance of anthro- with the highest human development index (HDI=0.875).20
pometric indicators (BMI, WC, WHtR) in the diagnosis of The other region considered was the western region of
58 Pelegrini A et al
Santa Catarina, one of the mesoregions of the state.19 The (weight and height) and waist circumference, using the
western region of Santa Catarina has an HDI of 0.807.20 Valdez mathematical equation.23
Among the top 20 cities in quality of life in Brazil, five are Body fat was verified by the relative body fat – % BF,24 for
from the western region of Santa Catarina, which has an boys and girls, using the sum (Σ2DC) of TSFT and MCST, as
estimated population of 25,322 inhabitants.19 shown below:
The sample was calculated separately for each region.
The following parameters were used: prevalence for the Boys Girls
outcome of 50% (unknown prevalence), tolerable error of
five percentage points, confidence level of 95%, and a %G=0.735*(TSFT+MCST)+1.0 %G=0.610*(TSFT+MCST)+5.1
delimitation effect of 1.5, adding 10% for possible losses/ %G, relative body fat; TSFT, triceps skinfold thickness; MCST,
refusals. Thus, 634 adolescents in each region were evalu- medial calf skinfold thickness.
ated, composing a total sample of 1,268 adolescents.
In Florianópolis, the sampling process was determined in The cutoff points used for the classification of body fat
two stages: stratified by geographic region and conglomer- were those recommended by Lohman,25 according to gen-
ate groups. In the first stage, the city was divided into five der and age, in which values higher than 20 for boys and 25
geographical regions: center, continent, east, north and for girls were considered high.
south. The school with the largest number of students from Mean and standard deviation were used in the descrip-
each region was selected, and in each school, classes were tive analysis of variables. The Kolmogorov-Smirnov test
randomly selected to represent a sample representative of was used to verify data normality. Differences in the aver-
the geographic area. In the second stage, all adolescents ages of variables between genders were analyzed by the
who were present in classroom on the day of data collec- Mann Whitney test. Association between anthropometric
tion were invited to participate in the study. indicators and gender was assessed by the chi-square test.
In the Midwestern region of Santa Catarina, the sampling To evaluate the diagnostic performance of BMI, WHtR and
process was determined in two stages: stratified by public C-Index in detecting excess body fat, the ROC curve anal-
high schools and classes conglomerates. In the first stage, ysis was applied. The diagnostic accuracy refers to the
only schools with over 150 students were considered. ability of BMI, WHtR and C-Index to discriminate adoles-
Moreover, in cities with more than one teaching unit, we cents with excess body fat from those without excess body
chose the one with the highest number of students. In the fat. Areas under the ROC curve and confidence intervals
second stage, all adolescents who were present in class- were determined. To better determine the optimal critical
room on the day of data collection were invited to partici- values of anthropometric indicators with greater accuracy
pate in the study. in the overweight detection, sensitivity and specificity
For this investigation, we defined as eligible the students were considered for each gender. The significance level
enrolled in public state schools, those present in the class- was set at p<0.05. Analyses were performed using SPSS
room on the day of data collection and those aged 15-17 (Statistical Package for Social Sciences) 20.0 version and
years. The exclusion criteria were: (a) students either <15 MedCalc.
or >17 years old; (b) students who did not bring the Free
and Informed Consent Form (FICF) signed by parents and/
or guardian; (c) students who refused to participate; (d) Results
students who did not perform anthropometric measure-
ments. The study showed a response rate of 94.4% (n=1,197), with
Fieldwork was conducted by Physical Education teachers 478 male and 719 female adolescents aged 15-17 years.
and students, trained to carry out all the necessary proce- The sample characteristics are presented in Table 1. Boys
dures in order to standardize data collection. School stu- had higher body mass, height, WC, WHtR and C-Index,
dents were instructed on evaluations at least five days in while girls had higher averages of TSFT, MCST, sum of two
advance. At that time, the FICF was presented and they skinfolds (Σ2DC) and fat percentage (BF%) (p<0.05).
were informed about the procedures for the tests. The data The values of the area under the ROC curve, cutoff
collection team was trained in order to standardize the points, sensitivity and specificity are presented (Table 2)
anthropometric measurements. The technical error of mea- for all anthropometric indicators as discriminators of high
surement was not calculated, but the researcher responsi- relative body fat. All anthropometric indicators analyzed
ble for the survey had extensive experience in anthropo- showed predictive ability to identify subjects with high
metric measurements and routinely performed the quality body fat (i.e. lower limit of CI95% of the area under the
control of the team of evaluators. ROC curve >0.50). BMI, WHtR and WC had greater ability to
Anthropometric body mass data – weight and height, discriminate body fat in both genders compared to the
waist circumference, triceps skinfold thickness (TSFT) and C-Index (Table 2).
medial calf skinfold thickness (MCST) were measured The areas under the ROC curve of anthropometric indi-
according to standardized procedures.21 Body mass index cators in the prediction of body fat in adolescents can be
(BMI) was calculated and ranked according to cutoff points observed in Figure 1. Significant differences were observed
for adolescents, which vary according to age and gender.22 between the ROC curves in both genders, which show that
Abdominal obesity was verified by measuring waist circum- the ROC curve for the C-Index has the lowest percentage
ference. WHtR was assessed by the waist x height ratio in under the curve when compared to BMI, WC and WHtR
cm. C-Index was determined by measuring body mass (p<0,05).
Anthropometric indicators of obesity in the prediction of high body fat in adolescents 59
Table 1 General characteristics of the sample ±.
Male (n=478) Female (n=719) p value

± ±
Body Mass (kg) 64.52 (11.69) 55.27 (9.71) <0.001
Height (cm) 173.81 (7.50) 162.33 (6.10) <0.001
BMI (kg/m²) 21.32 (3.45) 20.95 (3.33) 0.056
WC (cm) 72.72 (7.68) 67.20 (7.23) <0.001
WHtR 0.42 (0.04) 0.41 (0.05) 0.030
TSFT (mm) 10.14 (4.57) 16.39 (5.68) <0.001
MCST (mm) 11.28 (5.53) 17.92 (6.47) <0.001
Σ2DC (mm) 21.42 (9.57) 34.27 (11.28) <0.001
%BF 16.74 (7.04) 26.00 (6.88) <0.001
C-Index 1.10 (0.05) 1.06 (0.06) <0.001
BMI, body mass index; WC, waist circumference; WHtR, waist/height ratio; TSFT, triceps skinfold thickness; MCST, medial calf skinfold
thickness; Σ2DC, sum of two skinfolds; % BF, relative body fat; C-Index, conicity index
Table 2 Diagnostic properties of anthropometric indicators of obesity to detect high body fat percentage in adolescents
according to gender.
Curve ROC (CI95%) Cutoff point Sensitivity Specificity

% (CI95%) % (CI95%)
Male
BMI 0.84 (0.81-0.87)* 22.7 63.0 (53.7-71.7) 89.5 (85.8-92.4)
WHtR 0.83 (0.79-0.86)* 0.43 68.9 (59.8-77.1) 81.7 (77.3-85.6)
WC 0.81 (0.77-0.85)* 75.7 60.5 (51.1-69.3) 88.1 (84.3-91.2)
C-Index 0.65 (0.60-0.69)* 1.12 52.1 (42.8-61.3) 74.0 (69.1-78.4)
Female
BMI 0.79 (0.76-0.82)* 20.1 78.5 (73.9-82.1) 64.9 (59.8-69.8)
WHtR 0.77 (0.74-0.80)* 0.41 65.0 (59.8-69.9) 76.4 (71.7-80.7)
WC 0.77 (0.74-0.80)* 67.7 57.6 (52.3-62.8) 81.1 (76.7-85.0)
C-Index 0.62 (0.58-0.66)* 1.06 51.7 (46.4-57.0) 68.5 (63.5-73.2)
CI95%, confidence interval; BMI, body mass index; WC, waist circumference; WHtR, waist/height ratio; C-Index, conicity index.
*: area under the ROC curve demonstrating discriminatory power for body fat (lower limit of CI95%>0.50).
Discussion ilarity to predict blood hypertension.26 The similarity

between these indicators lies in the fact that both deal
All anthropometric indicators were able to diagnose excess with fat located in the central region.10 This study also
body fat, as they showed the lowest limit of 95% of the area reported that BMI was similar to WC and WHtR to detect
under the ROC curve up to 0.50. However, BMI, WHtR and the adiposity anthropometric indicator, which shows that
WC had greater ability to discriminate body fat in both during adolescence this measure may be useful for diagnos-
genders compared to the C-Index. These results show that ing obesity.10
not only indicators of general obesity (BMI), but also indi- The findings of this study have vital implications for the
cators of central obesity (WC, WHtR) can be used in adoles- assessment of obesity among adolescents, since it reinforc-
cents to diagnose high body fat. es the use of anthropometric indicators of obesity, which
These results were similar to those presented by Brazilian are relatively simple to be evaluated, as a discriminator of
adults in relation to the C-Index, which is an anthropomet- body fat. There is no doubt that the assessment of body
ric indicator with low discriminatory power for health prob- composition by skinfold is more accurate than using anthro-
lems compared to other anthropometric indicators.26 The pometric indicators, as shown by Nooyens et al.28 However,
C-index was a good predictor for chronic non-communica- the measurement of skinfolds requires trained evaluators
ble diseases.27 to provide accurate measurements. Thus, the World Health
WC and WHtR had enough similarity to discriminate body Organization29 recommends the use of simpler anthropo-
fat in this study. A study conducted in southern Brazil also metric indicators of obesity to evaluate possible health
revealed that these anthropometric indicators showed sim- risks.
BMI Waist-to height ratio

Waist circumference Conicity index
100 100
80 80
60 60
Sensitivity
Sensitivity
40 40
20 20
0 0
0 20 40 60 80 100 0 20 40 60 80 100
100-Specificity 100-Specificity
Female Male
Figure 1 Area under the ROC curve of anthropometric indicators in predicting body fat in adolescents.
Research conducted with Spanish children and adoles- the cutoff point for BMI for males in this study is in the
cents revealed that BMI, triceps skinfold thickness and WC overweight range of other studies.22,31 Moreover, the cutoff
were good anthropometric indicators in the diagnosis of point for BMI in this study for females is below those found
total body fat assessed by the doubly labeled water meth- in literature to detect overweight. One possible explana-
od.13 In North American children and adolescents (5-18 tion for these discrepancies may be related to ethnic and
years old), it was shown that both BMI and fat percentage cultural differences in Brazilian adolescents that may influ-
(derived from skinfolds) are low cost, viable indicators that ence BMI.
can be used for screening excess body adiposity.30 BMI and As for WC, it was observed that the best cutoff point to
WC provided better diagnostic in screening obesity (mea- detect the emergence of high body fat was 75.7cm and
sured by plethysmography) in adolescents than the waist- 67.7cm for boys and girls, respectively. Fernandez et al,32
hip circumference ratio (WHtR) in Swedish adolescents.14 when developing cutoff points for WC in a sample represen-
Based on the results of this study and those found in liter- tative of children and adolescents of different ethnicities
ature, it could be inferred that for adolescents anthropo- (African Americans, European Americans and Mexican
metric indicators of general obesity and central obesity are Americans) found that, in the age group of this study, WC
both good predictors of high body fat. ranged from 79.4cm to 87.0cm for males and from 78.3cm
The best cutoff point for BMI to detect the emergence of to 85.5cm for females. It was also observed that the cutoff
high body fat was 22.7kg/m² for boys and 20.1kg/m² for points found for adolescents of this investigation are lower
girls. Usually, the cutoff points for BMI in adolescents vary than those of other studies.32 Evidence shows that, among
according to gender and age.22,31 A study that developed the anthropometric indicators, WC had the best perfor-
cutoff points for BMI in a sample of Brazilian adolescents mance in the diagnosis of obesity in children and adoles-
reported that in the age group of this study (14-17 years cents.11,14
old), BMI for overweight ranged from 21.7kg/m² to 23.6kg/ Regarding WHtR, the best cutoff point to detect the
m² for males and from 22.8kg/m² to 24.8kg/m² for females. emergence of high body fat was 0.43cm and 0.41cm for
For obesity, the mentioned study reported that for males males and females, respectively. Studies conducted with
the cutoff point for BMI ranged from 27.5kg/m² to 28.7kg/ Italian33 and African adolescents34 found that the best diag-
m² and for females the cuttoff point ranged from 27.5kg/ nostic value of WHtR for metabolic risk was 0.41, which is
m² to 29.6kg/m². In the study by Cole et al,22 who devel- similar to the findings of this study, and lower than what is
oped cutoff points for BMI in a sample of children and ado- internationally proposed (0.50). Moreover, this indicator
lescents from six countries (Brazil, Great Britain, Hong has been considered one of the best in the evaluation of
Kong, the Netherlands, Singapore and the United States), central fat distribution, and it is associated with various
BMI for overweight ranged from 22.6kg/m² to 24.5kg/m² cardiovascular risk factors.10 As for predicting high body
for males and 23.3kg/m² to 24.7kg/m² for females. For fat, it is possible to observe that WHtR has been considered
obesity, the cutoff point for BMI ranged from 27.6kg/m² to a simple, easy-to-use, accurate indicator, with high appli-
29.4kg/m² for males and the cuttoff point for females cability in screening overweight and obesity in children and
ranged from 28.6kg/m² to 29.7kg/m². It was observed that adolescents.12
Anthropometric indicators of obesity in the prediction of high body fat in adolescents 61
The best cutoff point for the C-Index was 1.12 for boys 10. Flegal KM, Shepherd JA, Looker AC, Graubard BI, Borrud LG,
and 1.06 for girls. Publications on the prediction of high Ogden CL, et al. Comparisons of percentage body fat, body
body fat through the C-Index were not found, which makes mass index, waist circumference, and waist-stature ratio in
it difficult to compare the results found in this study. adults. Am J Clin Nutr. 2009;89:500-8.
However, cutoff points for the C-Index were developed to 11. Hubert H, Guinhouya CB, Allard L, Durocher A. Comparison of
the diagnostic quality of body mass index, waist circumference
detect high blood pressure (boys=1.13 and girls=1.14), high
and waist-to-height ratio in screening skinfold-determined
levels of total cholesterol (boys=1.10) and low levels of obesity among children. J Sci Med Sport. 2009;12:449-51.
HDL-c (girls=1.10).35 12. Weili Y, He B, Yao H, Dai J, Cui J, Ge D, et al. Waist-to-height ratio
Among the limitations of the study, the use of double is an accurate and easier index for evaluating obesity in children
indirect measures (skinfold) to establish the criterion mea- and adolescents. Obesity (Silver Spring). 2007;15:748-52.
sure of body fat can be highlighted; however, in the assess- 13. Sarría A, Moreno LA, García-Llop LA, Fleta J, Morellón MP,
ment of nutritional status and body composition in children Bueno M. Body mass index, triceps skinfold and waist
and adolescents, such measures are commonly used and circumference in screening for adiposity in male children and
recommended by health agencies.29 adolescents. Acta Paediatr. 2001;90:387-92.
According to the findings of this study, it could be con- 14. Neovius M, Linné Y, Rossner S. BMI, waist-circumference and
waist-hip ratio as diagnostic tests for fatness in adolescents.
cluded that anthropometric indicators can be used in
Int J Obes (Lond). 2005;29:163-9.
screening to identify high body fat in adolescents for being 15. Lunardi CC, Petroski EL. Body mass index, waist circumference
a simple, inexpensive and non-invasive method. These find- and skinfolds for predicting lipid abnormalities in 11 years old
ings reinforce the possibility of using anthropometric indi- children. Arq Bras Endocrinol Metab. 2008;52:1009-14.
cators as an alternative to evaluate adolescents, through 16. Campagnolo PD, Hoffman DJ, Vitolo MR. Waist-to-height ratio
simple, replicable and reliable criteria, with high sensitivi- as a screening tool for children with risk factors for
ty and specificity at low cost, which allows greater range in cardiovascular disease. Ann Hum Biol. 2011;38:265-70.
the scope of monitoring nutritional and health status among 17. Fernandes RA, Christofaro DG, Buonani C, Monteriro HL, Cardoso
adolescents. JR, Freitas IF Jr, et al. Performance of body fat and body mass
index cutoffs in elevated blood pressure screening among male
children and adolescents. Hypertens Res. 2011;34:963-7.
18. Christofaro DG, Ritti-Dias RM, Fernandes RA, Polito MD, Andrade
Conflicts of interest SM, Cardoso JR, et al. High Blood pressure detection in
adolescents by clustering overall and abdominal adiposity
The authors declare no conflicts of interest. markers. Arq Bras Cardiol. 2011;96:465-70.
19. Brasil – Instituto Brasileiro de Geografia e Estatística [homepage
on the Internet]. IBGE cidades [accessed 22 October 2010].
Available from: http://www.ibge.gov.br/cidadesat.
References 20. Organização das Nações Unidas [homepage on the Internet].
Ranking IDHM Municípios 2010 [accessed 10 June 2010].
1. Bambra CL, Hillier FC, Moore HJ, Summerbell CD. Tackling Available from: http://www.pnud.org.br/atlas/ranking/
inequalities in obesity: a protocol for a systematic review of Ranking-IDHM-Municipios-2010.aspx.
the effectiveness of public health interventions at reducing 21. Canadian Society for Exercise Physiology (CSEP). The Canadian
socioeconomic inequalities in obesity amongst children. Syst physical activity, fitness and lifestyle ppraisal: CSEP’s guide to
Rev. 2012;1:16. health active living. 2nd ed. Ottawa: CSEP; 2003.
2. Goldhaber-Fiebert JD, Rubinfeld RE, Bhattacharya J, Robinson 22. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a
TN, Wise PH. The utility of childhood and adolescent obesity standard definition for child overweight and obesity worldwide:
assessment in relation to adult health. Med Decis Making. international survey. BMJ. 2000;320:1240-3.
2013;33:163-75. 23. Valdez R. A simple model-based index of abdominal adiposity. J
3. Lloyd LJ, Langley-Evans SC, McMullen S. Childhood obesity and Clin Epidemiol. 1991;44:955-6.
risk of the adult metabolic syndrome: a systematic review. Int 24. Slaughter MH, Lohman TG, Boileau RA, Horswill CA, Stillman
J Obes. 2012;36:1-11. RJ, Van Loan MD, et al. Skinfold equations for estimation of
4. Sant’Anna MS, Priore SE, Franceschini SC. Methods of body body fatness in children and youth. Hum Biol. 1988;60:709-23.
composition evaluation in children. Rev Paul Pediatr. 25. Lohman TG. Applicability of body composition techniques and
2009;27:315-21. constants for children and youths. Exerc Sport Sci Rev.
5. Ellis KJ. Human body composition: in vivo methods. Physiol Rev. 1986;14:537-57.
2000;80:649-80. 26. Silva DA, Petroski EL, Peres MA. Accuracy and measures of
6. Carvalho AB, Pires-Neto CS. Body composition by underwater association of anthropometric indexes of obesity to identify the
weighing and bioelectrical impedance methods in college presence of hypertension in adults: a population-based study in
students. Rev Bras Cineantropom Des Hum. 1999;1:18-23. Southern Brazil. Eur J Nutr. 2013;52:237-46.
7. Sigulem DM, Devincenzi MU, Lessa AC. Diagnosis of child and 27. Pitanga FJ, Lessa I. Anthropometric indexes of obesity as an
adolescent nutritional status. J Pediatr (Rio J). 2000;76 (Suppl instrument of screening for high coronary risk in adults in the
3):S275-84. city of Salvador- Bahia. Arq Bras Cardiol. 2005;85:26-31.
8. Neovius M, Linné Y, Barkeling B, Rössner S. Discrepancies 28. Nooyens AC, Koppes LL, Visscher TL, Twisk JW, Kemper HC,
between classification systems of childhood obesity. Obes Rev. Schuit AJ, et al. Adolescent skinfold thickness is a better
2004;5:105-14. predictor of high body fatness in adults than is body mass
9. Imai A, Komatsu S, Ohara T, Kamata T, Yoshida J, Miyaji K, et al. index: the Amsterdam growth and health longitudinal study. Am
Visceral abdominal fat accumulation predicts the progression J Clin Nutr. 2007;85:1533-9.
of noncalcified coronary plaque. Atherosclerosis. 2012;222:524- 29. World Health Organization [homepage on the Internet].
9. Physical status: the use and interpretation of anthropometry.
[accessed 24 June 2014]. Available from: http://www.who.int/ 33. Kruger HS, Faber M, Schutte AE, Ellis SM. A proposed cutoff
childgrowth/publications/physical_status/en/. point of waist-to-height ratio for metabolic risk in African
30. Laurson KR, Eisenmann JC, Welk GJ. Body Mass Index standards township adolescents. Nutrition. 2013;29:502-7.
based on agreement with health-related body fat. Am J Prev 34. Papalia T, Greco R, Lofaro D, Mollica A, Roberti R, Bonofiglio R.
Med. 2011;41 (4 Suppl 2):S100-5. Anthropometric measures can better predict high blood
31. Conde WL, Monteiro CA. Body mass index cutoff points for pressure in adolescents. J Nephrol. 2013;26:899-905.
evaluation of nutritional status in Brazilian children and 35. Beck CC, Lopes AS, Pitanga FJ. Anthropometric indicators as
adolescents. J Pediatr (Rio J). 2006;82:266-72. predictors of high blood pressure in adolescents. Arq Bras
32. Fernández JR, Redden DT, Pietrobelli A, Allison DB. Waist Cardiol. 2011;96:126-33.
circumference percentiles in nationally representative samples
of African-American, European-American, and Mexican-American
children and adolescents. J Pediatr. 2004;145:439-44.
See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/280239230
3D Hand Anthropometry of Korean Teenager's and Comparison with Manual

Method
Conference Paper in Communications in Computer and Information Science · June 2014

DOI: 10.1007/978-3-319-07854-0_85
CITATIONS READS
4 2,088
5 authors, including:
Se Jin Park Seungnam Min

SEWON Intelligence Shinsung University
115 PUBLICATIONS 778 CITATIONS 51 PUBLICATIONS 151 CITATIONS
SEE PROFILE SEE PROFILE
Heeran Lee Murali Subramaniyam

SRM Institute of Science and Techonolgy
SEE PROFILE
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Micro Factory View project
Anthropometry View project
All content following this page was uploaded by Seungnam Min on 22 July 2015.
The user has requested enhancement of the downloaded file.

3D Hand Anthropometry of Korean Teenager’s
and Comparison with Manual Method
Se Jin Park1, Seung Nam Min2, Heeran Lee1,

Murali Subramaniyam2, and Sang Jae Ahn3
1
Division of Convergence Technology,
Korea Research Institute of Standards and Science, Daejeon 305-340, Korea
{Sjpark,heeranlee}@kriss.re.kr
2
Center for Medical Metrology, Division of Convergence Technology,
Korea Research Institute of Standards and Science, Daejeon 305-340, Korea
msnijn12@kriss.re.kr, murali.subramaniyam@gmail.com
3
Department of Electrical and Computer Engineering, Korea Aerospace University,
Goyang 412-791, Korea
sangjae2006@naver.com
Abstract. The requirements of wearing products fitting comfort was conti-

nuously increasing and considerable attentions had been paid for a long time.
The assessment of the physical dimensions of the human hand provided a me-
tric description to establish human-machine compatibility. Higher accuracy in
hand anthropometric measurements could be achieved with the aids of an image
analysis system. Scanning of hand surfaces either 2D or 3D was an alternative
method for manual measurements. Three-dimensional anthropometry may lead
to significant improvement in fitting comfort of wearing products. The pur-
pose of this study was to measure 3D hand anthropometry and compared it with
manual methods. For that purpose, 10 hand measurements of the right hand
(lengths, breadths, and circumference of hand and fingers) were taken from
1,700 middle and high school students by age ranged from 13 to 19 years old.
The hand was measured by manual (using anthropometric sliding, spreading ca-
lipers and measuring tape) and using a high-resolution 3D hand scanner
(NEXHAND H-100, Knitech, South Korea) with the scanning accuracy ± 0.5
mm. From the scanned data, the hand measurements were extracted using scan-
ning software (Enhand, Knitech, South Korea). Mean and standard deviation for
each hand measurements were calculated. T-test statistical test on the data re-
vealed that there was no significant difference between the manual and 3D hand
measurements (p > 0.05). Therefore, 3D anthropometry can be replaced with
manual methods. The data gathered may be used for ergo-design applications of
hand tools and devices. And also it can provide a great help to develop a hand
anthropometry database for hand wearing products.
1 Introduction
Anthropometric data are one of significant factors in designing machines and devices.
Incorporating anthropometric data would yield more effective designs, ones that are
C. Stephanidis (Ed.): HCII 2014 Posters, Part II, CCIS 435, pp. 491–495, 2014.
© Springer International Publishing Switzerland 2014
492 S.J. Park et al.
more user friendly, safer, and enable higher performance and productivity. The as-
sessment of the human hand physical dimensions provides a metric description to
establish human-machine compatibility [1, 2, 3]. To protect hands from the hazardous
work environments workers are commonly used the gloves [4]. The experts in the
certain sports used the custom-made gloves to their personal specification require-
ments. Pressure therapy gloves were mainly found in the area of pressure therapy,
those gloves are designed to apply acceptable pressure to hand and fingers with the
purpose to increase the rate of scar maturation, prevent contracture formation, and
enhance cosmetic appearance without impairing circulation [5, 6, 7, 8]. There are
many types of gloves available in the market to protect against a wide variety of ha-
zards. An accurate and effective measurement of hand anthropometric dimensions is
crucial to optimize the effectiveness and practical use of the gloves [4]. There are two
hand measurement methods namely direct and indirect. Direct measurement method
used the tools such as flexible measuring tape, calipers, martin anthropometry device,
measuring boards and rulers are traditionally used to obtain hand dimensions [9]. The
indirect methods such as three-dimensional (3D) image analyses have been widely
adopted for taking body dimensions in the design of various products including medi-
cal, garments, safety instruments, etc., [10, 11, 12, 13, 14]. Along with the 3D image
analysis, multi-camera photogrammetric systems based on two-dimensional (2D)
images for body measuring have been developed in various studies [15, 16].
The objective of the present study was Korean teenagers’ hand anthropometric
measurements by using indirect method using 3D scanner and compare against the
direct measurement method. This research work aims to provide a useful reference for
the development of hand anthropometry database.
2 Method
2.1 Participants
Ten measurements (lengths, breadths, and circumference of hand and fingers) of the
right hand from 1,700 middle and high school students were measured and their age
ranged from 13 to 19 years old. Table 1 showed participants demographic information.
Table 1. Demographic information for participants
Height (mm) Weight (kg) BMI

N = 1,700 1593.37±52.26 52.72±8.26 20.73±2.88
2.2 Methodology of Measurement

The instruments used for the manual measurements were the anthropometric sliding
and spreading calipers, and measuring tape (martin anthropometer). The instrument
used for the indirect measurements was a high-resolution 3D hand scanner
(NEXHAND H-100, Knitech, South Korea) with the scanning accuracy ± 0.5 mm.
The 3D scanner generates a 3D digital hand, and then computer software (Enhand,
3D Hand Anthropometry of Korean Teenager’s and Comparison with Manual Method 493
Knitech, South Korea) measures 3D hand dimensions from the digital hand (Fig. 1).
The generated 3D digital hand with its several dimensions can be directly applied to
design of product shape. Mean and standard deviation for each hand measurement
were calculated and compared.
Fig. 1. Measurement tools used (a: Martin anthropometer, b: NEXHAND H-100 hand 3D
scanner)
The selected hand dimensions measured from the direct and indirect methods were
showed in Fig. 2
Fig. 2. Hand items measured from the direct and indirect methods
3 Results and Discussion
The t-test was performed to compare the significance between the direct and indirect
measurement methods (Table 2) for the measured items. The t-test results showed
that there were no significant differences between two measurement methods for
whole items measured.
494 S.J. Park et al.
Table 2. Right Hand Anthropometric Measurements Comparison between Direct and Indirect
Method using T-Test
Measured Items Mean (SD) in mm p-value

D 169.21 7.30
Hand length 0.182
I 168.93 4.36
D 65.32 3.61
Index finger Length 0.655
I 64.90 1.61
D 72.75 3.59
Medius finger length 0.846
I 72.54 1.57
D 68.01 3.88
Ring finger length 0.246
I 68.05 2.14
D 53.63 3.98
Little finger length 0.54
I 53.13 1.28
D 97.05 5.09
Palm length perpendicular 0.854
I 97.02 3.15
D 76.91 3.69
Hand breadth with thumb 0.68
I 75.84 2.54
D 52.08 3.16
Hand breadth with wrist 0.15
I 51.87 2.58
D 26.29 2.15
Hand thickness 0.14
I 26.14 1.24
D 177.73 9.0
Hand circumference 0.94
I 1776.24 2.1
D: Direct measurement method, I: Indirect measurement method.
Garrett, 1971 [9] performed the direct measurement using traditionally available
tools to make the glove design and pattern development. However, the direct mea-
surements are time consuming; the accuracy of the measurement depends on the
person who is measuring those dimensions. Another important point need to be hig-
hlighted that using direct measurement methods there are limited number of hand
dimensions can be measured. Nevertheless, the 3D hand scanned data contains nu-
merous hand dimensions. From the 3D scanned data, many dimensions can be
extracted using custom build software tools for example Rapidform. Also the repeata-
bility of the indirect measurement is higher than the direct measurement method [4].
However, there are still in many developed countries both measurement methods are
employed to construct national anthropometry database. The existence of an anthro-
pometry database is essential in every society and this data should be up-to-date. As
the size of the some body parts may alter during years [16]. With the indirect mea-
surement methods it would be easier to update the anthropometry database.
3D Hand Anthropometry of Korean Teenager’s and Comparison with Manual Method 495
4 Conclusion
The study performed Korean teenagers’ hand anthropometric measurements by indi-
rect measurement method using 3D hand scanner and compared the dimensions
against the direct measurement method. The findings of this study imply that the av-
erage of hand dimensions has no significant difference in the two methods (p > 0.05).
The statistical analyses showed that indirect measurement methods (using 3D scan-
ner) can be replaced with direct measurement methods. It can provide an extended
help to develop an anthropometric database for gloves manufactures and also to de-
velop national anthropometric database.
References
1. Fraser, T.: Ergonomic Principles in the Design of Hand Tools. In Occupational Safety and
Health Series No. 44. International Labour Office, Geneva (1980)
2. Freivalds, A.: The Ergonomics of Tools. International Reviews of Ergonomics 1, 43–75
(1987)
3. Nag, A., Nag, P.K., Desai, H.: Hand Anthropometry of Indian Women. Indian Journal of
Medical Research 117, 260–269 (2003)
4. Yu, A., Yick, K.L., Ng, S.P., Yip, J.: 2D and 3D Anatomical Analyses of Hand Dimen-
sions for Custom-Made Gloves. Applied Ergonomics 44, 381–392 (2013)
5. Leung, W.Y., Yuen, D.W., Ng, S.P., Shi, S.Q.: Pressure Prediction Model for Compres-
sion Garment Design. Journal of Burn Care & Research 31, 716–727 (2010)
6. Fricke, N.B., Omnell, M.L., Dutcher, K.A., Hollender, L.G., Odon, T., Engrav, L.H.: Ske-
letal and Dental Disturbances in Children after Facial Burns and Pressure Garment use: A
4-Year Follow-up. Journal of Burn Care & Research 20, 239–249 (1999)
7. Rappoport, K., Müller, R., Flores-Mir, C.: Dental and Skeletal Changes during Pressure
Garment use in Facial Burns: A Systematic Review. Burns 34, 18–23 (2008)
8. Silfen, R., Amir, A., Hauben, D.J., Calderon, S.: Effect of Facial Pressure Garments for
Burn Injury in Adult Patients after Orthodontic Rreatment. Burns 27, 409–412 (2001)
9. Garrett, J.W.: The Adult Human Hand: Some Anthropometric and Biomechanical Consid-
erations. Human Factors: The Journal of the Human Factors and Ergonomics Society 13,
117–131 (1971)
10. Yu, C.Y., Tu, H.H.: Foot Surface Area Database and Estimation Formula. Applied Ergo-
nomics 40, 767–774 (2009)
11. Yu, C.Y., Lo, Y.H., Chiou, W.K.: The 3D Scanner for Measuring Body Surface Area: A
Simplified Calculation in the Chinese Adult. Applied Ergonomics 34, 273–278 (2003)
12. Lu, J.M., Wang, M.J.J., Mollard, R.: The Effect of Arm Posture on the Scan-derived Mea-
surements. Applied Ergonomics 41, 236–241 (2010)
13. Hsu, Y.W., Yu, C.Y.: Hand Surface Area Estimation Formula using 3D Anthropometry.
Journal of Occupational and Environmental Hygiene 7, 633–639 (2010)
14. Lu, J.M., Wang, M.J.J.: Automated Anthropometric Data Collection using 3D Whole
Body Scanners. Expert Systems with Applications 35, 407–414 (2008)
15. Lin, Y.L., Wang, M.J.J.: Automated Body Feature Extraction from 2D images. Expert Sys-
tems with Applications 38, 2585–2591 (2011)
16. Habibi, E., Soury, S., Hasan Zadeh, A.: Evaluation of Accuracy and Precision of Two-
Dimensional Image Processing Anthropometry Software of Hand in Comparison with Ma-
nual Method. Journal of Medical Signals and Sensors 3 (2013)
View publication stats

August 2019/ Vol 6/ Issue 08 Print ISSN: 2349-5499, Online ISSN: 2349-3267
Original Research Article
Study of anthropometric profile of working (labour) adolescent girls of

urban slums of India
Yadav S.1, Bharti R.2
1
Dr. Swati Yadav, Assistant Professor, Department of Pediatrics, Chhattisgarh Institute of Medical Sciences, Bilaspur,
Chhattisgarh, India, 2Dr. Rajesh Bharti, Consultant Psychiatrist, Department of Psychiatry, Sadar Hospital, Purnea, Bihar,
India
Corresponding Author: Dr. Swati Yadav, Assistant Professor, Department of Pediatrics, Chhattisgarh Institute of
Medical Sciences, Bilaspur, Chhattisgarh, India. E-mail: yadavswati93@yahoo.com
………………………………………………………………………………………………………………………………...
Abstract
Introduction: Adolescence is often described as a phase of life that begins in biology and ends in society. Adolescence is
the developmental phase between childhood and adulthood with specific psychological attributes. In India, the adolescent
population constitute 22.8% of the total population. There are around 239 million adolescents in India in the age group of
10-19 years presently. Our aim is to assess the anthropometric profile of working adolescent girls of slum area. Method:
Anthropometric data were collected using predesigned proforma of 696 working adolescent girls aged 10-19years by
door to door survey in three randomly selected slums of Bilaspur, Chhattisgarh, India. The height, weight and body mass
index of the working adolescent were compared with NCHS and Indian reference populations. Result: Analysis of the
cases revealed that the maximum number (45.40%) of adolescents were in 10-13years age group and minimum (27.01%)
were in 17-19 years age group. In 14-16years 27.58% girls. The difference of mean weight of present study with NCHS
standard 50th percentile of weight in various age group ranges from 11.16 to 18.5 kgs. Height of all the girls were below
50th percentile. Difference of present study’s mean height and NCHS 50th percentile in different age group ranges
from11.7cms to 20.8cms, the average body mass index was below 15 in 10-12 years age group but was above 15 in 13-19
years age group. Difference of mean BMI of present study with NCHS Standard for different age group ranges from2.15
to 3.57. Discussion: The anthropometric height, weight and BMI data of working adolescent of slum area were lower
than Indian as well as NCHS standards. All the girls in the present study had weight and height less than 50th percentile
of the National Center for Health Statistics (NCHS), Department of Health, Education and Welfare (DHEW), U.S.A.
Standards.
Key words: Anthropometric, NCHS, BMI data

………………………………………………………………………………………………………………………………...
Introduction
Adolescence is often described as a phase of life that India in the age group of 10-19 years presently. Over
begins in biology and ends in society. Adolescence is the next two decades the number of adolescents is likely
the developmental phase between childhood and to increase further but their share to population will
adulthood with specific psychological attributes [1]. decrease marginally as per the projections [3]. WHO
Adolescence starts with a period of very rapid physical described the adolescence as the period of sexual
growth accompanied by the gradual development development from the initial appearance of secondary
reproductive organs, secondary sex characteristics and sex characteristics to sexual maturity, psychological
menarche in girl. Not only will they soon bear the future development from child to adult identification, and
generation they are also the breeding ground for new socioeconomic development from dependence to
ideas, languages, values and careers [2]. In India, the relative independence [4].
adolescent population constitute 22.8% of the total
population. There are around 239 million adolescents in In India, 28% of the total population was living in urban
Manuscript received: 10th August 2019
areas in 2001, with future projection of about 50%, with
Reviewed: 20th August 2019 605-618 million by 2021-2025. Demographic trends
Author Corrected: 27th August 2019
Accepted for Publication: 31st August 2019
show that while urban average growth rate stabilized at
Pediatric Review: International Journal of Pediatric Research Available online at: www.medresearch.in 418|P a g e
3% over past decade (1991-2001), the slum growth rate various aspects like age, sex, religion, address, mother,
was doubled at 5-6% [5]. Urban adolescents from lower father, occupation. It is followed by detailed general
class have to struggle for survival and grow in examination regarding state of health and later complete
impoverished, disadvantaged environment making them systemic examination was done. Anthropometric
vulnerable to several risks, malnutrition, risk of poor measurement, general physical was recorded on
health, becoming victim of antisocial activities brewing predesigned pro forma. Consent from their selves in
and sale of illicit liquor/sex exploitation/prostitution and those who are above 18 years and from parents in case
drug peddling were reported threats for adolescents of minor.
from slums in multi indicator survey [6].
For evaluation of nutritional status of adolescents,
Aims and Objectives anthropometric measurements like weight, height
recorded for each adolescent.
To study the anthropometric profile including weight,
height and body mass index of working adolescent girls
Age- Age of each child was calculated to nearest 6
between 10-19 years of age of urban slum area of
month (e.g.-10 years 4 month and 11 years 8 month as
Bilaspur (C.G)
12 years). Age of most of the children were calculated
and cross checked with reference to the events such as
Material and Methods some important festival, storm, flood etc. The age was
The Study is mainly focusing on health status of recorded in complete years.
working adolescent girls in urban slums of Bilaspur
(C.G.) Weight- A bathroom scale was used. It was caliberated
against known weight regularly. Zero error was checked
Design of study-Cross Sectional study for and removed if present every day, with a standard
minimal clothing and without shoes. Weight was
Place of study: Various Urban Slums of Bilaspur city
recorded to nearest 500 gms.
Duration of study: September 2010-July 2011
Height- Height in cms was marked on a wall with the
Selection of slum: Out of 167 slums of Bilaspur city,3 help of a measuring tape. All children were measured
slums were chosen for study purpose by systematic against the wall. The children were asked to remove the
random sampling with the help of table of random footwear and stand with heels together and head
number. positioned so that the line of vision was perpendicular
to the body. A glass scale was brought down to the
Sample size-The approach adopted for this community- topmost point on the head, Height was recorded to the
based study was cross sectional type. Sample size is nearest 1 cm.
based on the extent of malnutrition in adolescent. As
per literature the problem of malnutrition in school After taking the various measurements, average mean
children in India ranged from 10.0-98% and taking the was obtained and standard variation calculated. The
middle course a prevalence rate of 50% is assumed. By result obtained in various age group was compared with
taking this prevalence and permissible level of error at international standards-National Center for Health
10% sample size is computed as 696. Statistics (NCHS) and other workers reported in
literatures.
Inclusion criteria-All working adolescent females of
Slum Area between 10-19 years. Body Mass Index: BMI was calculated from measured
height and weight of various age to assess the
Exclusion criteria-Non-working adolescent female
nutritional status. BMI of 15 or below was considered
-Females having any major systemic illness, major
specific indicator to assess under nutrition and give
surgery
earliest indication of malnutrition
Method of data collection- With the help of local
Analysis of data- The data collected from the various
health worker a cross sectional survey was done by door
aspect in present study was analyzed. The results are
to door visit. Examination conducted at local
tabulated and data also subjected to necessary statistical
Anganbadi Centre. Data was collected on predesigned
test whenever required. For statistical analysis Chi-
pro forma with the answer to questionnaire regarding
square, Student T test, one-way ANOVA was applied
Result
For study purpose girls were divided into early (10-13years), mid (14-16 years) and late (17-19 years) adolescent.
Analysis of the cases revealed (Table-1) that the maximum number (45.40%) of adolescents were in 10-13 years age
group and minimum (27.01%) were in 17-19 years age group. In 14-16 years 27.58% girls.
All the girls in the present study had weight and height less than 50th percentile of the National Center for Health
Statistics (NCHS), Department of Health, Education and Welfare (DHEW), U.S.A. Standards (Table-2 and 3).
The difference of mean weight of present study with NCHS standard 50th percentile of weight in various age group
ranges from 11.16 to 18.5 kgs (Table-2).
When present study was compared with the anthropometric data of study conducted by Agrawal DK et al (1992) [7, 8],
they were coming below 50th percentile.
The average body mass index (Table-3) was below 15 in 10-12 years age group but was above 15 in 13-19 years age
group.
In Present study BMI of girls were below 50th percentile of NCHS standard, difference of mean BMI of present study
with NCHS Standard for different age group ranges from 2.15 to 3.57.
Table-1: Age wise distribution.
Age group Present study

10-13 45.40%
14-16 27.5%
17-19 27.01%
Table-2: Age wise distribution of mean weight and range of weight.
Age No. of cases Weight Range Mean Weight

(yrs) (kgs) (kg)
10 86 15.2-23.6 21.34
11 56 18.6-26.8 22.8
12 94 22.3-32.1 26.4
13 80 25.1-34.4 28.78
14 64 32.4-36.1 33.8
15 76 30.8-42.6 35.8
16 52 35.4-40.4 37.4
17 72 36.2-42.3 39.6
18 52 37.8-42.3 40.1
19 64 36.4-46.2 42.4
Table-3: Age wise distribution of mean and range of height and body mass index
Age No. of cases Range of height Mean Height Body Mass Index
(cms)
10 86 120.6-130.2 124.7 13.76
11 56 120.7-132.4 126.8 14.11
12 94 129.7-136.1 132.6 14.9
13 80 131.6-142.1 136.3 15.4
14 64 136.2144.3 139.6 17.2
15 76 141.7148.2 144 17.2
16 52 145.2-152.2 148.7 16.88
17 72 146.9-152.7 150.2 17.5
18 52 146.9-154.2 150.6 17.63
19 64 154.3-146.1 150.5 18.6
Discussion
This study is cross sectional study of anthropometric When height compared with the National Center for
profile of working adolescent girls of urban slum area. Health Statistics (NCHS), Department of Health,
All the girls in the present study were weighing less Education and Welfare (DHEW), U.S.A. Standards, all
than 50th percentile of the NCHS Standards. The girls were below 50th percentile. Difference of present
difference of mean weight of present study with NCHS study’s mean height and NCHS 50th percentile in
standard 50th percentile of weight in various age group different age group ranges from 11.7 cms to 20.8cms,
ranges from 11.16 to 18.5 kgs. the difference in height is quite obvious as the NCHS
standards are the figure of healthy children of highly
Even though all the Indian studies on high developed country. The adolescents in present study
socioeconomic group show a higher growth rate as failed to reach the American Standards because in the
compared to the growth rate of general population of present study they belonged to lower socioeconomic
children, the results vary considerably. group from slums with high prevalence of malnutrition,
infection and other morbidities.
When present study was compared with the
anthropometric data of study conducted by Agrawal DK When present study is compared with the anthro-
et al (1992) [7], they were coming below 50th percentile. pometric data of study conducted by Agrawal DK et al
The difference of mean weight of present study with (1992) [7], they were coming below 50th percentile with
50th percentile mean weight ranges from 5.56 to 10.72 difference of mean height in various age group ranging
kgs, it may be due to relative improved status of from 6.4 to 15.1 cms.
nutrition in the adolescents as Agrawal D K et al (1992)
[7, 8] done on affluent society. Singh J et al (2006) [9] These variation in the results of different workers from
and Kaul K K et al (1976) [10] study is more or less the present study may be due to difference in the
similar to present study. Rao S et al (1998) (11) and socioeconomic status, cultural and environmental
Tripathy A M et al (1976) [12] do not corresponds with factors and variation in number of children included in
present study. the studies in various samples of different age groups.
Study of Singh J et al (2006) [9] is similar to present
In study of Tripathy AM et al (1976) [12]-difference of study but Medhi G et al (2006) [14], Rao S et al (1998)
mean weight with present study’s mean weight for [13], Kaul K K et al (1976) [10] is not similar to present
different age group ranges from 0.86 to 12.2 kgs and study with range of mean height difference from 0.35 to
with Rao S et al (1998) [13] study mean weight 7.93cms,0.6 to 10.0, 0.02 to 10.3 cms respectively. It
difference ranges from 0.2 kgs to 6.12 kgs. may be due to the fact that girls in present study are
living in slum area which is not properly sanitated, nutrition programs and child health services,
unhygienic, chances of subclinical infections are more. fortification of food items and nutritional education of
As they are working girls requirement of calories and family.
protein are more but they are consuming lesser calories
and protein which is utilized in their work instead of Author’s contribution
utilization for growth and development.
Dr. Rajesh Bharti: Conceived the idea of study,
participated in study design, analysis of data and
Body mass index was applied in present study to assess
interpretation of results.
the nutritional status. The body mass index of less than
15 is considered to be the earliest indicator of protein
Dr. Swati Yadav: Supervised the study, reviewed
calorie malnutrition. Difference of mean BMI of present
related literature, data collection and write up of
study with NCHS Standard for different age group
manuscript. All authors have read and approved the
ranges from 2.15 to 3.57.
final manuscript.
The average body mass index was below 15 in 10-12 Funding: Nil, Conflict of interest: None initiated,
years age group but was above 15 in 13-19 years age Permission from IRB: Yes
group. Study showed linear increase in BMI with age
from 10-13 years age group to 14-16years age group, References
this gain might be due to attainment of menarche and
1. UNFPA: Adolescent in India. A profile. December
onset of thelarche and pubarche in these age group.
2003.
Divya rani et al (2017) [15] found 60.3% were under
2. De Silva WI. Emerging reproduction health issues
nourished, among them 20.3%in grade III, 15.1% in
among adolescent in India. 1998;139.
grade II and 24.9% in grade I under nutrition. In
contrast shahid et al [16] found that 20% of the girls are
3. Govt of India: Planning Commission-Report of
underweight (BMI <18.5%),77% within normal limit,
working group on adolescents for the tenth five year
and 3% were obese
plan. June 2001.
Conclusion
4. De Silva WI. Emerging reproductive health issues
All the anthropometrics parameter of the adolescents in among adolescents in Asia. Boston, MA: Harvard
the present study were less than the International School of Public Health. 1998.
(NCHS) standards and Indian standards. This might be
due to the fact that adolescents of NCHS (17) standard 5. Agarwal S, Taneja S. All slums are not equal: child
and other Indian studies were from affluent families health conditions among the urban poor. Indian
getting better nutrition leading to better growth. Pediatrics. 2005;42(3):233-244.
Difference with NCHS may be due to racial variation
and difference in the environment in which American 6. UNFPA: Annual Report. December 2003 Available
children are brought up at https://www.unfpa.org/ sites/default/ files/pub-pdf/
annual_ report03_eng.pdf.
What the study adds to the existing
knowledge? 7. Agarwal DK, Agarwal KN, Upadhyay SK, Mittal R,
Prakash R, Rai S. Physical and sexual growth pattern of
As there are only few studies on anthropometric profile affluent Indian children from 5 to 18 years of age.
of working (labour) adolescent girls of urban slum area, Indian Pediat. 1992;29(10):1203-1282.
the present study will be helpful in knowing the
magnitude of under nutrition in them as well will be 8. Agarwal KN, Sen S, Tripathi AM, Katiyar GP.
helpful in planning and implementation of various Physical growth characteristics in relation to sexual
program for enhancement of their nutritional status. growth. Indian Pediat. 1974;12(2):99-105.
Thus, it can be recommended that any attempt for the 9. Singh J, Singh JV, Srivastava AK. Health status of
improvement of nutritional status of slums children adolescent girls in slums of Lucknow. Indian J Comm
must enhanced like rigorous implementation of Med. 2006;31(2):102.
10. Kual KK, Taskar AD, Madhavan S, Mukerji B, 14. Medhi GK, Hazarika NC, Mahanta J. Nutritional
Parekh P, Sawhney K, Goel RK, Lamba IM. Growth in status of adolescents among tea garden workers. The
height and weight of urban Madhya Pradesh Indian J Pediat. 2007;74(4):343-347.
adolescents. Indian Pediat. 1976;13(1):31-39.
15. Rani D, Singh JK, Srivastava M, Verma P,
11. Rao S, Joshi S, Kanade A. Height velocity, body fat Srivastava D, Singh SP. Assessment of nutritional status
and menarcheal age of Indian girls. Indian Pediat. of teenage adolescent girls in urban slum of Varanasi.
1998;35(7):619-628. International journal of current research and review
2018;10(20):1-10. doi: http/dx.doi.org/10.31782/IJCRR.
12. Tripathi AM, Agarwal DK, Sen S, Agarwal KN. 2018.10202.
Physical growth during adolescence in Delhi school
children. Indian Pediat. 1976;13(3):191-199. 16. Shahid A, Siddiqui FR, Bhatti MA, Ahmed M,
Khan MW. Assessment of nutritional status of
13. Rao S, Joshi S, Kanade A. Height velocity, body fat adolescent college girls at Rawalpindi. Ann King
and menarcheal age of Indian girls. Indian Pediat. Edward Med Univers. 2009;15(1):11. doi: https://doi.
1998;35(7):619-628. org/10. 21649/akemu.v15i1.24
…………………………..
How to cite this article?
Yadav S, Bharti R. Study of anthropometric profile of working (labour) adolescent girls of urban slums of India. Int J
Pediatr Res. 2019;6(08):418-423.doi:10.17511/ijpr. 2019.i08.05
………………………………………………………………………………………………………………………...............
Eur Spine J
DOI 10.1007/s00586-015-3961-7
REVIEW ARTICLE
Reliability and validity of inexpensive and easily administered

anthropometric clinical evaluation methods of postural
asymmetry measurement in adolescent idiopathic scoliosis:
a systematic review
Ashleigh Prowse1 • Rodney Pope1 • Paul Gerdhem2,3 • Allan Abbott4,5
Received: 14 January 2015 / Revised: 13 April 2015 / Accepted: 13 April 2015

Ó Springer-Verlag Berlin Heidelberg 2015
Abstract assessed from radiographs, ranged from low to very high.

Purpose As accurate and reproducible measurements of iPhone measurements correlated well with scoliometer
spinal curvature are crucial in the examination of patients measurements. 2D photography results had a moderate to
with adolescent idiopathic scoliosis (AIS), this systematic high correlation with 3D topography results.
review aims to report on the reliability and validity of a Conclusions Overall, strong levels of evidence exist for
range of inexpensive and easily administered anthropo- iPhone and scoliometer measurements, with a high to very
metric methods of postural asymmetry measurement in an high reliability and moderate to very high validity.
AIS population, to inform practice in a clinical setting. Moderate levels of evidence exist for scoliometer with
Methods A systematic search of health research databases mathematical formula and clinical examination with
located studies assessing reliability and validity of inex- moderate and low validity, respectively. Limited evidence
pensive and easily administered anthropometric measures. exists for aesthetic tools TRACE and AI and 2D photog-
Results Fourteen studies satisfied eligibility criteria. The raphy. These results indicate there are accurate and re-
methodological quality of included studies ranged from producible anthropometric measures that are inexpensive
low to high. Validity studies were of moderate to high and applicable in therapy settings to assess postural
quality. In total, nine clinically applicable, inexpensive and asymmetry; however, these only exist for measurement in
easily administered anthropometric methods were identi- the transverse plane, despite 3D characteristics of AIS.
fied, for assessing AIS curvature. All methods demon- Further research is required into an inexpensive and easily
strated high to very high inter-observer and intra-observer administered method that can assess postural asymmetry in
reliability. Reported criterion validity of the scoliometer all anatomical planes.
and 2D photographs, when compared to Cobb angle
Keywords Reliability Validity Adolescent idiopathic
scoliosis Postural asymmetry Measurement
& Ashleigh Prowse Anthropometric
ashleigh.prowse@student.bond.edu.au
1
Department of Physiotherapy, Faculty of Health Science and
Medicine, Bond University, Robina, QLD, Australia Introduction
2
Department of Orthopaedics, Karolinska University Hospital,
Stockholm, Sweden Adolescent idiopathic scoliosis (AIS) is a three-dimen-
3
Department of Clinical Science, Intervention and Technology
sional (3D) developmental deformity of the spine and trunk
(CLINTEC), Karolinska Institutet, Stockholm, Sweden that appears during pubertal growth, commonly in ado-
4 lescence [1, 2]. AIS is defined as a coronal plane Cobb
Department of Physiotherapy, Karolinska University
Hospital, Huddinge, Sweden angle [10° [3, 4] and is characterised as: (1) a lateral
5 spinal curvature in the frontal plane, (2) a disturbance of
Division of Physiotherapy, Department of Medical and
Health Sciences, Faculty of Health Sciences, Linköping physiological spinal curvatures in the sagittal plane, and (3)
University, Linköping, Sweden an axial rotation of vertebrae in the transverse plane [4].
123
Eur Spine J
The prevalence of AIS in the general population has been clinically applicable inexpensive and easily administered
reported to be between 0.47 and 5.2 % [5]. anthropometric measures of postural asymmetry. We
AIS is commonly detected with screening, that may be aimed in this systematic review to identify and assess the
performed with manual anthropometric measures such as research evidence pertaining to inter-observer and intra-
the scoliometer [6–9]. Researchers have reported that early observer reliability and criterion validity of clinically ap-
detection of scoliosis resulted in a threefold increase in the plicable inexpensive and easily administered anthropo-
number of patients who could be conservatively treated for metric methods to measure postural deformity in AIS. The
scoliosis, thus decreasing the percentage of patients who review will provide important information that is funda-
required surgery [10–12], and further, that it is also im- mental in clinical evaluation and will inform decisions
portant in reducing the side effects of AIS [13]. Consid- regarding treatment in AIS.
ering the improvements gained from early detection of
AIS, accurate, non-invasive and reliable tools are crucial
for screening within clinical settings [14, 15]. Methods
Many scoliosis evaluation devices are described in the
literature; however, devices generally provide only two- Search strategy
dimensional (2D) measures, despite the 3D characteristics
of AIS [15]. The principal screening test for scoliosis is Relevant studies involving human participants that were
measuring angle of trunk rotation with a scoliometer. The published between January 1990 and October 2014 were
inter- and intra-observer reliability of scoliometer assess- identified by searching relevant health, science and phys-
ments has in several studies ranged from very good to iotherapy electronic databases and subsequent hand
excellent and the tool is reportedly useful as a screening searching of reference lists from identified articles. To avoid
device [8, 11]. Further, the criterion validity of scoliometer publication bias and ensure as many studies were captured
measurements, when correlated to the gold standard Cobb as possible, the search included online books, unpublished
angle from radiographs, has been found to be fair to very papers, conference abstracts, and seminar and reference
good [16]. Aesthetic evaluation is also important to be used searching, to identify clusters of highly related papers.
in clinical evaluation of AIS as it is considered a goal of Search terms (Fig. 1) were kept broad due to the high va-
conservative treatment by SOSORT experts and is the most riety of synonyms used in the literature for key terms.
relevant indication for surgery [17]. Digital photography is
becoming increasingly popular in the assessment of trunk Screening and selection
deformity [17].
Various 3D posture analysis systems such as Optotrak One reviewer (A.P) screened the titles and abstracts iden-
(Northern Digital Inc., Waterloo, Canada), Vicon (Vicon tified by the literature search, with the full text of all arti-
Motion Systems, Oxford, UK), VIVID (Konica Minolta cles that potentially met the inclusion criteria obtained for
Sensing Inc., Ramsey, NJ, USA), Motion Analysis (Motion further evaluation. Two reviewers (A.A and A.P) then
Analysis Corporation, Santa Rosa, CA, USA) and surface screened the full text of these articles using inclusion and
topography systems have been used to quantitatively assess exclusion criteria. Inclusion criteria were: (1) humans with
posture of subjects with AIS [18–21]. Recent reviews have primary diagnosis of idiopathic scoliosis, (2) male or fe-
been completed, appraising the literature of camera and male adolescents, (3) easy to administer and inexpensive
computer-based topography systems [22], suggesting they anthropometric measures, and (4) observational, cross-
offer a better 3D description of the morphologic deformity sectional or controlled trial research design. Exclusion
associated with AIS [16, 23]. However, these systems are criteria were: (1) studies of spinal deformity primarily at-
not accessible for most clinicians as they are expensive, tributed to a non-idiopathic aetiology (e.g. neuromuscular,
require specialized trained technicians and the data pro- neurological, congenital malformation, trauma-related co-
cessing is complex. New, simpler and more accessible tools morbidity), (2) non-healthy subjects (other significant co-
to measure Cobb angle are emerging, including the use of morbidity of neuromuscular, neurological, congenital or
iPhone applications [24]. However, the reliability and va- traumatic nature), (3) studies published prior to 1990, (4)
lidity of these methods have only begun to be researched. papers not available in English language, (5) studies
Grivas [17] suggested that we now have a set of clinical assessing only imaging (MRI and CT scan) or comput-
tools in research to potentially evaluate AIS rehabilitation, erised or camera-based topography, (6) validity other than
which when combined, go beyond radiographs to evaluate criterion related, and (7) reliability other than inter-ob-
and monitor aspects of AIS relevant to patients and clin- server or intra-observer. A PRISMA flow diagram was
icians. However, no recent systematic review has been used to document the study screening and selection process
performed comparing reliability and validity of the [25] (Fig. 1).
123
Eur Spine J
Critical appraisal inclusion and exclusion criteria. After abstract and title
review and removal of duplicates, 20 articles were re-
The validity and reliability critical appraisal tool [26], trieved in full text (Fig. 1). After full text review by two
consisting of 13 items, was completed by two blinded re- authors (A.A and A.P), 14 articles meeting the inclusion
viewers (A.A and A.P) in order to assess the method- criteria were retained for critical appraisal, and these 14
ological quality of each included article and the results articles were included in the review (Fig. 1). Four of the 20
were tabulated. The CAT has been designed to improve the articles retrieved in full text were excluded as participants
quality of reporting of studies of the validity and reliability did not have an AIS diagnosis [18, 30–32], another as the
of objective clinical tools [26]. In item 13, statistical power study included those with congenital deformity [33] and an
was considered in conjunction with the methods of statis- unpublished paper was excluded as an English translation
tical analysis [27]. The CAT does not incorporate a quality was not available [34]. The two reviewers reached agree-
score. A scoring system was developed by the researchers ment on inclusion or exclusion of all articles, after dis-
to assist in categorising the quality of studies. The cussion in these five cases, with no third reviewer required.
methodological quality of validity studies was scored out
of nine items, and that of reliability studies out of eight Critical appraisal of methodological quality
items. One point was awarded for meeting each criterion
item on the CAT. The total methodological quality ratings The 14 papers consisted of 30 studies examining one or
employed were: 0–2 poor, 3–5 fair, 6 or 7 moderate, and 8 more of intra-observer reliability, inter-observer reliability
or 9 high for validity studies, and 0–2 poor, 3 or 4 fair, 5 or and/or criterion validity. Tables 1, 2 and 3 provide the
6 moderate and 7 or 8 high for reliability studies. results of the critical appraisals conducted of these studies
of the various measurement methods. The two reviewers
Data analysis initially disagreed on the critical appraisal results of two
studies; however, agreement was met between reviewers.
Meta-analysis was not attempted due to the heterogeneity Of the eight studies that examined the criterion validity of
of measurement methods examined and statistical analyses measurement methods, three were of high methodological
employed in the included studies. Sub-group analysis was quality for this purpose [35–37]. Of the ten studies that
also not performed due to the limited number of studies examined intra-observer reliability of measurement meth-
assessing each measurement method. A descriptive analy- ods, two were of high methodological quality for this
sis was therefore conducted and data were synthesised purpose [38, 39]. Of the 12 studies that examined inter-
using a level of evidence approach [28]. Each measurement observer reliability, one was of high methodological
method was assigned a ‘strong’, ‘moderate’, ‘limited’ or quality for this purpose [39].
‘conflicting’ level of evidence based on the number of
studies available and the quality of the studies. Consistent Study participants
findings from C3 high quality studies indicate a strong
level of evidence, 1–2 high quality studies or 1 high quality The number of participants in each included study ranged
study and C1 low quality studies indicate moderate level from 8 to 442 with a mean age of 14.9 years. Females were
evidence, and C1 low quality studies or only 1 study more frequently studied than males. All participants had a
available indicates limited evidence. Inconsistent evidence primary diagnosis of AIS and were untreated, conserva-
in multiple studies irrespective of quality was considered tively treated or surgically treated. Mean Cobb angles were
conflicting. To assess reliability and validity, the intraclass between 21° and 26°.
correlation coefficient (ICC) and Pearson’s correlation
coefficient (r) were interpreted as follows: 0.00–0.29 as Identified measurement methods for postural
very low correlation, 0.30–0.49 as low, 0.50–0.69 as deformity in AIS
moderate, 0.70–0.89 as high correlation and 0.90–1.00 as
very high correlation [29]. A total of nine inexpensive and easily administered meth-
ods for measurement of postural deformity in AIS in a
clinical setting were examined in the included studies.
Results Three studies assessed two types of iPhone applications as
a measure of scoliosis, including a study with the attach-
Literature search and selection ment of an acrylic sleeve resembling a scoliometer [37, 39,
40]. Eight studies assessed the scoliometer [11, 36, 37, 39–
Electronic database and reference list searches yielded 43], which is a fluid filled inclinometer that measures the
1294 potential articles after the use of filters to reflect degrees of axial rotation in the spine in the transverse
123
Eur Spine J
plane. Two studies assessed use of a scoliometer plus a anthropometric methods of measuring postural deformity
mathematical formula [38, 44]. Fortin et al. [35] assessed in AIS, for clinical use. Results were supportive of the
postural indices obtained from 2D photographic interven- iPhone, scoliometer, and scoliometer with mathematical
tions in the frontal and sagittal planes. Zaina et al. [45] formula methods for the measurement of postural defor-
assessed aesthetics using the trunk aesthetic clinical eval- mity in AIS. All of these measurement methods were as-
uation tool (TRACE) and aesthetic index (AI), which sociated with moderate or strong levels of evidence and
consists of frontal and sagittal photographs depicting dif- overall demonstrated moderate to high reliability and
ferent severities of four aspects of trunk deformity; generally high criterion validity when measurements were
shoulder, scapulae, hemithorax and waist. Grosso et al. [46] compared to Cobb angle measured from radiographs or, in
assessed clinical examination in frontal, sagittal and the case of the iPhone measurements, with results of sco-
transverse planes using plumb lines, distance from plumb liometry. TRACE, AI and 2D photography methods were
lines, rib hump and axial trunk rotation. The scoliometer associated with limited levels of evidence, and in the
was the most commonly studied anthropometric measure, limited studies available were reported to assess curvature
in terms of both criterion validity and reliability. in multiple anatomical planes, providing aesthetic infor-
mation with moderate reliability. The criterion validity of
Reliability and validity of the identified these techniques was less commonly studied and ranged
measurement methods from low to high in included studies.
Tables 4, 5 and 6 provide details of the reported levels of Issues in assessment of reliability of measurement
inter-observer reliability, intra-observer reliability and cri- methods
terion validity of the measurement methods used to assess
postural deformity in AIS. Overall, high to very high re- The variable nature of AIS poses a potential challenge to
liability and moderate to very high criterion validity were studies of the reliability of measurement methods. Studies
reported for the iPhone, scoliometer measurement. included in this review aimed to reduce variance due to
Moderate criterion validity was reported for the fatigue from repeated measures by providing rest periods
scoliometer with mathematical formula and very low to between measurements [37, 40]. Studies also attempted to
moderate criterion validity evident for the clinical ex- control variance due to patient flexibility, body mass index
amination method. TRACE, AI and 2D photography (BMI) or occupational or sporting activity by re-testing
methods had moderate reliability and low to high validity. within the same session [11, 32, 36, 37, 40, 43]. High to
very high reliability supports the notion that these controls
Level of evidence were sufficient. Other potential challenges to assessing
reliability included the accuracy of palpation of spinal
Table 7 describes the levels of evidence provided by the landmarks, with poor validity of palpation of the lumbar
included articles on the reliability and criterion validity of spine reported in the literature [14]. Across available
each anthropometric clinically applicable evaluation studies the level of experience of assessors ranges from in-
method of postural asymmetry measurement in AIS. Strong training, to self-taught, to experienced. In one study [40], a
levels of evidence exist for iPhone and scoliometer mea- parent was as accurate as a nurse and surgeon in using the
surement methods. Moderate levels of evidence exist for scoliometer. Thus, it is unclear if the relative levels of
the scoliometer with mathematical formula and Clinical experience may have contributed to results obtained. Var-
Examination with plumb line, hump height and ATR with iation in subject positioning is another challenge discussed
scoliometer. Evidence for TRACE, AI and 2D photography in the literature. One study [39] used plaster casts of AIS
methods was limited for these measurement methods. patients, thus removing patient positioning influence during
the assessment of measurement reliability. However, this
resulted in no significant difference in levels of assessed
Discussion reliability when compared with other studies using live
subjects, suggesting variation in patient positioning may
Main findings not unduly influence reliability.
This systematic review identified 14 studies that provided Issues in assessment of criterion validity
evidence of the reliability or criterion validity of available of measurement methods
clinically applicable evaluation methods of AIS curvature.
The evidence provided by these studies assists in deter- Significant barriers to assessment of criterion validity of
mining the best inexpensive and easily administered inexpensive and easily administered clinical evaluation
123
Eur Spine J
measurement methods for postural deformity in AIS are the between observers, which allowed for knowledge of pre-
limited accessibility, and ethical issues regarding radiation, vious results [11]. Further, a common theme found in the
of spinal radiographs. This could explain the lack of cri- included criterion validity studies was that they lacked in
terion validity testing of scoliosis measurement methods. explanation [38, 44, 48] or did not document bias [51].
3D computerised methods have been suggested as an al-
ternative reference method. However, their own assessed Limitations of the systematic review
criterion validity is still variable within studies [15]. The
large ranges found in reported criterion validity of specific Multiple reviewers selected studies for inclusion, appraised
measurement methods within the included studies could be the methodological quality of included studies, and anal-
explained by the differences in anatomical landmarks used ysed and reported the results. The raters were blinded to
[47], including spinous process [35, 37] versus vertebral each other’s findings and disagreements were resolved by
pedicle [36], and the adipose tissue or muscle overlaying discussion of the bases for differences in opinions. An
the landmarks. Difficulty in palpation could explain the low extensive range of relevant health electronic databases was
level of criterion validity found for measurement methods searched and pre-specified inclusion and exclusion criteria
when they were applied to the lumbar region when com- were applied equally to titles and abstracts of all identified
pared to the thoracic region, as the thoracic region is articles and then to the full text of potentially relevant
considered more easily palpable [35, 47]. The level of articles. Only one reviewer completed the database
examiner competence in the use of the measurement device searches, which may have inadvertently introduced sam-
and anthropometrics also affects the validity of results. pling bias into the review despite the best efforts of the
Studies using self-taught examiners reported low criterion author to ensure all relevant articles were selected.
validity [36], and those training in the use of the device Through this review, we were able to explore potential
reported moderate criterion validity [48], while studies reasons behind observed variability in assessed reliability
where the examiners were experienced in the use of the and validity of measurement methods for assessing postural
device and anthropometrics reported high [11] and very deformity in AIS. One study minimised subject positioning
high criterion validity [37]. variance by using plaster casts, some studies lacked in
blinding of assessors to landmarks used by others for
Methodological considerations for included studies anatomical palpation, and studies using assessors with
varying levels of competence with specific measurement
The participants included within the studies fitted the dis- methods. The heterogeneity in measurement methods ex-
tribution of genders in an AIS population. A limitation of amined by the included studies strengthens the external
the study was that only one of the validity studies explored validity of the findings of this review. It is important to
the BMI of participants [48]. Many reliability studies failed recognise that outcomes of this review are situated, partial
to provide the range or mean ages of the adolescents [39, and perspectival.
41, 45]. Despite findings of high to very high inter-observer
and intra-observer reliability of measurement methods re- Clinical and research implications
ported within the included studies, many used inappropri-
ate statistical methods for reliability data analysis. Kappa Five out of nine measurement methods identified in this
statistics [45], rho [41] and Pearson’s correlation coeffi- systematic review were associated with strong levels of
cient [36, 42, 48] are suggested in literature to overestimate evidence regarding reliability and criterion validity. The
or underestimate reliability [26, 29]. The ICC with 95 % overall high to very high levels of inter-observer and intra-
confidence interval (CI) is considered the best statistical observer reliability, and moderate to high level of criterion
method for analysing reliability data [49]. Further, it is validity determined across the included studies for scolio-
recommended in literature to analyse the standard error of metry, scoliometry plus mathematical formula, two differ-
measurement (SEM) in studies of reliability of measure- ent iPhone methods and clinical examination of spinal curve
ment methods [50]. Many studies failed to assess either the measurement in AIS, indicate that there are accurate, re-
SEM or the 95 % CI. A lack of blinding between observers producible, inexpensive and easily administered radiation
may also have improved the inter-observer reliability. Two free anthropometric measurement methods, that detect AIS
studies did not remove markings of palpated landmarks and monitor curve progression. Further, the reliability
123
Eur Spine J
values also indicate these measurement methods are reliable coronal planes through clinical examination of plumb lines,
when used by the same therapist on separate occasions or by ATR and hump height [46]. Methods such as TRACE, AI
different therapists. To improve the criterion validity and and 2D photography provide frontal and sagittal plane in-
reliability of these clinical evaluation methods of measuring formation that is of significant aesthetic importance; how-
postural deformity in AIS, physicians or physiotherapists ever, more research is required into accuracy and
with appropriate training and experience in palpation of reproducibility of these methods, due to inconsistent results
anthropometric spinal landmarks should be assigned to use across included studies. These measures could provide
the methods to assess the postural asymmetry. additional aesthetic information to supplement ATR mea-
The scoliometer is not sufficient to use independently as sures such as the scoliometer or iPhone.
a basis for treatment decisions such as bracing or surgical
intervention, as despite its good level of criterion validity it
is not as accurate in assessing spinal curvature as the gold Conclusion
standard Cobb angle measurement on radiographs, par-
ticularly in the lumbar region. The iPhone applications, The results of this systematic review indicate there are
particularly those with acrylic sleeve, were shown to pro- several anthropometric measures that are inexpensive, re-
vide accurate measures of ATR when compared with the liable, valid and useful in a therapy setting to detect AIS
scoliometer and thus could replace the scoliometer as a and monitor curve progression. Overall, strong levels of
more accessible measurement method of similar cost, to evidence exist for iPhone and scoliometer measurements,
monitor curve progression or screen for AIS in a clinical with a high to very high reliability and moderate to very
setting. Further, the review found, based on the limited high validity. Moderate levels of evidence exist for
available evidence, that 2D photographs appeared to have a scoliometer with mathematical formula and clinical ex-
high correlation with 3D computer system measures and amination, with moderate validity. Evidence for TRACE,
Cobb angle measured from radiographs, and therefore AI and 2D photography methods is limited, with poor to
could provide similar information with less expense, fair reliability and poor to high validity. Considering AIS is
though further research is required to confirm this possi- a 3D developmental deformity of the spine and trunk and
bility. Clinical examination using plumb lines and rib there is limited evidence regarding methods measuring
hump in combination with ATR can add important repro- postural deformity in multiple anatomical planes, further
ducible information in multiple anatomical planes that research is warranted to assess methods of measuring AIS
could be easily applied in a clinical setting. curvature in three anatomical planes using inexpensive,
Few methods measured postural asymmetry in three easy to administer methods that will ensure comprehensive
anatomical planes. The scoliometer provides information assessment of postural asymmetry.
regarding the axial rotation of vertebrae in the transverse
plane. Of the nine measurement methods considered in the Conflict of interest None.
included studies, two examined important information re-
garding lateral spinal curvature, pelvic tilt and shoulder tilt
in the frontal plane, and disturbance of physiological spinal Appendix
curvatures in the sagittal plane [35, 45]. Only one study
assessed postural deformity in frontal, transverse and See Fig. 1 and Tables 1, 2, 3, 4, 5, 6 and 7.
123
Eur Spine J
Fig. 1 PRISMA flow diagram
123
123
Table 1 Critical appraisal inter-observer reliability studies (CAT)
Study Method Study population Testing circumstances Data analysis Execution Quality
Reference, Anatomical Study Sample Between- Adequate Order exam Time Description Appropriate Description Score
country plane and sample characteristics observer description varied between index test Statistical results with
method (mean ± SD/ blinding observer and repeated procedure method explanation
range) competence measures withdrawals
Coelho et al. Transverse 42 Y N Y N Y Y Y Y 6/8

[13], Orthopaedics 31 F 18.2 ± 3.9 years E1 spine 2 examiners Not varied 15–20 min Description ICC 95 % CI No Moderate
Brazil Systems Inc. 1M marked, E2 2 months and withdrawals
Idiopathic
scoliometer scoliosis Cobb used training prior reference
26° ± 18° markings study
Bonagamba Transverse 24 Y N Y N Y Y Y Y 6/8
et al. [43], Orthopaedics 2M 18 ± 4 years E2 marked SP 2 examiners Not varied 15–20 min Description ICC 95 % CI No Moderate
Brazil Systems Inc. 22 F and withdrawals
IS Cobb E1 used 10 h Mean (SEM)
scoliometer 25° ± 13° markings reference
Driscoll Transverse 39 Y Y Y Y Y Y Y Y 6/8
et al. [40], Scolioscreen 8M 16 ± 1.4 years Blinding Nurse, surgeon, Order varied 10 min Description ICC 95 % CI No Moderate
Canada (spinologics) 31 F parent trained by and withdrawals
Cobb Tx
iPhone 21° ± 15° nurse reference
inclinometer Lx 10° ± 9°
Zaina et al. Frontal, 160 N Y Y N Y Y N Y 5/8
[45], Italy sagittal and AIS: insufficient Blinding 20 years Insufficient 1 week Description Kappa No Moderate
transverse information experience with information and statistics withdrawals
TRACE and trunk aesthetics reference 95 % CI
AI
Murrel et al. Transverse 22 N N Y N N Y N Y 3/8
[41], USA Orthopaedics FAIS Insufficient Not stated 2 orthopaedic Not stated 4 occasions Description ICC not used No Fair
Systems Inc. information surgeons, 2 not: stable and RHO instead withdrawals
scoliometer residents reference
Grosso et al. Transverse 116 Y Y Y N N Y N Y 5/8
[46], Italy and frontal 78 F F 2 independent Rehab physician Insufficient 1 month: Description ICC but not No Moderate
Hump height, 38 M 13.3 ± 2.3 years examiners and physio information time for and 95 %CI withdrawals
ATR, M change reference
distance 14.2 ± 2.7 years
from plumb
line
Amendt Transverse 65 Y Y Y Y N Y N Y 6/8
et al. [36] Orthopaedics 57 F AIS, 14.8 years Blinding Therapist self- Randomised Time Description Pearson’s No Moderate
Systems Inc. 7M taught between and r correlation withdrawals
Eur Spine J
35: 1 curve 21°
scoliometer not stated reference coefficient
31: double curve
29°
Eur Spine J
Table 1 continued
Study Method Study population Testing circumstances Data analysis Execution Quality
Reference, Anatomical Study Sample Between- Adequate Order exam Time Description Appropriate Description Score
country plane and sample characteristics observer description varied between index test Statistical results with
method (mean ± SD/ blinding observer and repeated procedure method explanation
range) competence measures withdrawals
Galatz et al. Transverse 50 Y Y N N Y Y N Y 5/8

[42] Orthopaedics 39 F Average age: 3 blinded Insufficient Not varied Short time Description Pearson’s No Moderate
Systems Inc. 11 M 12.6 years, Cobb information frame: and r correlation withdrawals
scoliometer 30° stable reference coefficient
condition
Korovessis Transverse 65 Y Y Y N Y Y Y Y 6/8
et al. [38], Scoliometer AIS 3 blinded 2 MD’s Insufficient 2 weeks Description Kappa No Moderate
Greece and math 13.1 ± 1.4 years Scoliometer, 2 information and coefficient withdrawals
formula Range 8–16 years radiologists reference 95 % CI
Cobb angle
Balg et al. Transverse 34 Y Y Y N Y Y Y Y 6/8
[37], Scolioguage 25 F Cobb 24 ± 14°, Blinding Orthopaedic Not varied Short time Description ICC with No Moderate
Germany app iPhone 9M AIS, Mean age surgeon and frame: and Bland– withdrawals
14.7 ± 2.9yrs physical therapy stable reference Altman plot
student condition with 95 %
CI
Izatt et al. Transverse 8 N/A plaster casts Y Y Y Y Y Y Y 7/7
[39] iPhone with AIS Randomised 4 spinal surgeons, Order varied Plaster casts Description ICC CI 95 % No Very
acrylic plaster physical stable and withdrawals high
sleeve models therapist, spinal condition reference
orthotist, 2
registrars and
new-graduate
physical
therapist
Pearsall Transverse 3 Y N Y N N Y N Y 4/8
et al. [48], Scoliometer, AIS, BMI 53.3 kg Insufficient 1 experienced Insufficient 6 months: Description Pearson’s No Fair
Canada back contour information examiner in information unstable and r coefficient withdrawals
13.9 ± 3.3 years
device anthropometric reference
measurement
Y met CAT criteria, N did not meet, Lx lumbar, Tx thoracic, CAT criteria, F female, M male, BMI body mass index, HH hump height, ICC intraclass correlation coefficient, q Spearman’s rank
correlation coefficient, CI confidence interval, SEM standard error measurement
123
123
Table 2 Critical appraisal intra-observer reliability studies
Study Method Study population Testing circumstances Execution Data analysis Quality
Reference, Anatomical Study Sample Within- Adequate Order exam Time between Description Appropriate Description Score
country plane and sample characteristics observer description of varied testing and index test statistical results with
method (mean ± SD/ blinding observer and theoretical procedure method explanation
range) competence stability withdrawals
Coelho et al. Transverse 42 Y N N N Y Y ICC 95 % CI Y 5/8

[13], Orthopaedics 31 F 18.2 ± 3.9 years E1 marked 2 examiners, Not varied Short time Description Y No Moderate
Brazil Systems Inc. 1M spine, E2 2 months frame: stable and withdrawals
AIS
scoliometer used same training prior condition reference
Cobb 26° ± 18°
markings study
Bonagamba Transverse 24 Y N Y N Y Y Y Y 6/8
et al. [43], Orthopaedics 2M 18 ± 4 years E2 marked 10 h Not varied 15–20 min Description ICC 95 % CI No Moderate
Brazil Systems Inc. 22 F SP and withdrawals
AIS Cobb Mean (SEM)
scoliometer 25° ± 13° E1 used same reference
markings
Zaina et al. Frontal, 160 N N Y N Y Y N Y 4/8
[45], Italy sagittal and Insufficient Insufficient 20 years Insufficient 1 week: stable Description Kappa No Moderate
transverse information experience information and statistics withdrawals
information
TRACE and with trunk reference 95 % CI
AI aesthetics
Driscoll Transverse 39 Y Y Y Y Y Y N Y 7/8
et al. [40], Scolioscreen 8M 16 ± 1.4 years Scoliometer Surgeon, nurse, Randomised 10 min: Stable Description ICC but no No Moderate
Canada (spinologics) 31 F facing patient condition and 95 % CI withdrawals
Cobb Tx
iPhone 21° ± 15° away reference
inclinometer assistant
Lx 20° ± 9°
reported
Murrel et al. Transverse 22 N N Y N N Y N Y 3/8
[41], USA Orthopaedics Insufficient Not stated 2 orthopaedic Not stated 4 separate Description ICC RHO No Fair
Systems Inc. information surgeon, 2 occasions and withdrawals
scoliometer resident reference
Amendt Transverse 65 Y N Y Y Y Y N Y 6/8
et al. [36] Orthopaedics 57 F 14.8 years No blinding Therapist self- Randomised 3 measurements Description Pearson’s No Moderate
Systems Inc. 7M 34 single curve taught directly after and r correlation withdrawals
scoliometer 31 double each other reference coefficient
curve AIS
Korovessis Transverse 65 Y Y Y N Y Y Y Y 7/8
et al. [38], Scoliometer 13.1 ± 1.4 years 1 observer 2MD’s Insufficient 2 weeks: stable Description Kappa No High
Greece and math two weeks Scoliometer, Information condition and coefficient withdrawals
range
Eur Spine J
formula 8–16 years apart 2 radiologists reference 95 % CI
Cobb angle
AIS
Eur Spine J
Table 2 continued
Study Method Study population Testing circumstances Execution Data analysis Quality
Reference, Anatomical Study Sample Within- Adequate description Order exam Time Description Appropriate Description Score
country plane and sample characteristics observer of observer and varied between index test statistical results with
method (mean ± SD/ blinding competence testing and procedure method explanation
range) theoretical withdrawals
stability
Balg et al. Transverse 34 Y N Y N Y Y Y Y 6/8

[37] Scolioguage 25 F Cobb 24° ± 14° No blinding Orthopaedic surgeon Not varied 10 min: Description ICC with No Moderate
app iPhone 9M AIS and physical therapy stable and p \ 0.001 withdrawals
Mean age student condition reference Bland–
14.7 ± 2.9 years Altman plot
95 % CI
Izatt et al. Transverse 8 N/A Y Y Y Y Y Y Y 7/8
[39] iPhone with AIS plaster casts Blinding 4 spinal orthopaedic Randomised Plaster casts Description ICC 95 % CI No High
acrylic surgeon, physical stable and withdrawals
sleeve therapist, spinal condition reference
orthotist, 2 training
registrars and new-
graduate physical
therapist
Pearsall Transverse 3 Y N Y N N Y N Y 4/8
et al. Scoliometer, AIS Insufficient Single experienced Insufficient Insufficient Description Pearson’s No Moderate
[48], back information examiner in information information and r coefficient withdrawals
13.9 ± 3.3 years
Canada contour anthropometric reference
device measurement
Y met CAT criteria, N did not meet CAT criteria, F female, M male, ICC intraclass correlation coefficient, q Spearman’s correlation coefficient, CI confidence interval, SEM standard error
measurement, BMI body mass index
123
Eur Spine J
Table 3 Critical appraisal criterion validity studies

Study Method Population Testing circumstances and execution
Reference, Anatomical plane and Study Sample Index test Independency o F Adequate Description
country method sample characteristics description reference standard explanation reference
(mean, range, SD) from index test reference standard
standard procedure
Coelho et al. Transverse 64 Y Y Y Y Y

[13], Orthopaedics Systems 18.2 ± 3.9 years Description Independent Cobb angle Description
Brazil Inc. scoliometer and
Idiopathic
scoliosis reference
cobb 26° ± 18°
Sapkas et al. Transverse 291 Y Y Y Y Y
[44], Orthopaedic Systems 47 M 14.1 ± 2.67 years Description Independent Cobb angle Description
Greece Inc. scoliometer with 244 and
AIS
math formula F reference
Fortin et al. Frontal and sagittal 70 Y Y Y Y Y
[35] 2D photo with X-RAY 60 F AIS Description Independent 3D Topography Description
and 3D system 10 M 15.7 ± 2.5 years and and Cobb angle
reference
Pearsall Transverse 14 Y Y Y Y Y
et al. [48], Scoliometer, back 10 F AIS Description Independent for lateral 3 tests: Cobb Description
Canada contour device 4M and deformity with ATR angle used as
13.9 ± 3.3 years
reference reference
BMI 53.3 kg
Grosso et al. Transverse, frontal and 116 Y Y Y Y Y
[46], Italy sagittal 78 F F 13.3 ± 2.3 Description Independent Cobb angle Description
Hump height, ATR, 38 M M 14.2 ± 2.7 and
distance from plumb reference
line
Amendt Transverse 65 Y Y Y Y Y
et al. [36] Orthopaedics Systems 57 F AIS Description Independent Cobb angle Description
Inc. scoliometer 7M 14.8 years and
reference
35: 1 curve 21°
31: double curve
29°
Korovessis Transverse 442 Y Y Y Y Y
et al. [38], Scoliometer and math 13.1 ± 14 years Description Independent Cobb angle Description
Greece formula and
range 8–16 years
reference
AIS
Balg et al. Transverse 34 Y Y Y Y Y
[37] ATR scoliometer 25 F Cobb 24° ± 14° Description Independent Scoliometer Description
Scolioguage app iPhone 9M AIS and
reference
14.7 ± 29 years
Study Method Testing circumstances and execution Data Quality
analysis
Reference, Anatomical plane Adequate Time btw Description Outcome tested Appropriate Score
country and method description reference results with statistical
observers and standard and explanation method
competence index test withdrawals
Coelho Transverse Y N Y Correlation btw Y 8/9

et al. Orthopaedics 2 examiners 2/2 X-ray directly No withdrawals Scoliometer and Pearson’s Moderate
[13], Systems Inc. training prior after A.P X-ray
R coefficient
Brazil scoliometer study Scoliometer
123
Eur Spine J
Table 3 continued
Study Method Testing circumstances and execution Data Quality
analysis
Reference, Anatomical plane Adequate Time btw Description Outcome tested Appropriate Score
country and method description reference results with statistical
observers and standard and explanation method
competence index test withdrawals
Sapkas Transverse N N N Correlation btw Y 6/9

et al. Orthopaedic Insufficient Insufficient Insufficient Scoliometer with Pearson’s R Moderate
[44], Systems Inc. information information information math formula and coefficient
Greece scoliometer with A.P X-ray
math formula
Fortin et al. Frontal and sagittal Y Y Y Correlation btw 2D Y 9/9
[35] 2D photo with Trained physio Measures No withdrawals photographs and Pearson’s R High
X-RAY and 3D taken close X-ray and 3D coefficient
system to each other system
Pearsall Transverse Y N Y Correlation back N 7/9
et al. Scoliometer, back Single experienced Insufficient No withdrawals contour device and Sample too Moderate
[48], contour device examiner in information scoliometer with small
Canada anthropometric X-ray
Pearson’s R
measurement coefficient
Grosso Transverse, frontal Y N Y Correlation X-RAY Y 8/9
et al. and sagittal Rehab physician and 1 month: No withdrawals with ATR, DP and Linear Moderate
[46], Italy Hump height, physical therapist Potential HH regression
ATR, distance change R2
from plumb line
Amendt Transverse Y Y Y Correlation Y 9/9
et al. [36] Orthopaedics Therapist self-taught X-ray directly No withdrawals scoliometer with Pearson’s R High
Systems Inc. after A.P X-ray coefficient
scoliometer scoliometer
Korovessis Transverse Y N Y Correlation Y 8/9
et al. Scoliometer and 29 MD’s Insufficient No withdrawals Scoliometer with Kappa Moderate
[38], math formula scoliometer, 29 information Cobb angle coefficient
Greece radiologists Cobb 95 % CI
angle
Balg et al. Transverse Y Y Y Correlation Y 9/9
[37] ATR scoliometer Orthopaedic Measures No withdrawals scoliometer with ICC Bland– High
surgeon and taken close scolioguage app Altman
Scolioguage app
iPhone physical therapy to each other plot with
student 95 % CI
Y met CAT criteria, N did not meet CAT criteria, F female, M male, BMI body mass index, ICC intraclass correlation coefficient, q Spearman’s
correlation coefficient, CI confidence interval, SEM standard error measurement, HH hump height, ATR axial trunk rotation, D distance plumb
line
Table 4 Inter-observer reliability data

Study Anatomical plane and method Inter-observer reliability Score
Coelho et al. [13], Brazil Transverse ICC = Tx 0.89 High

Orthopaedics Systems Inc. scoliometer Lx 0.89
Bonagamba et al. [43], Brazil Transverse ICC = Upper Tx 0.57 Upper Tx—moderate
Orthopaedics Systems Inc. scoliometer Mid Tx 0.89 Mid Tx—high
Lower Tx 0.95 Lower Tx—very high
Lx 0.84 Lx high
SEM 2.5–5.9°
123
Eur Spine J
Table 4 continued
Study Anatomical plane and method Inter-observer reliability Score
Driscoll et al. [40], Canada Transverse ICC = scoliometer 0.89 High

Orthopaedics Systems Inc. scoliometer
Driscoll et al. [40], Canada Transverse ICC = scolioscreen 0.89 Scolioscreen high
Scolioscreen (spinologics) Smartphone 0.75 iPhone high
iPhone inclinometer
Zaina et al. [45], Italy Frontal, sagittal and transverse k 0.17–0.28 Poor–fair
Aesthetic index 92.5–99.4 % level agreement
Zaina et al. [45], Italy Frontal, sagittal and transverse TRACE: k 0.09–0.14 Poor
TRACE 95–100 % level agreement
Murrel et al. [41], USA Transverse Tx q = 0.81 SD 1.9° Tx high
Orthopaedics Systems Inc. scoliometer Lx 0.76 SD 2.3° Lx high
Grosso et al. [46], Italy Transverse and frontal ICC = 0.90–0.91 HH HH very high
Hump height, ATR, distance from plumb line 0.86–0.89 ATR ATR high
0.76–0.86 DPL PL: mod–high
Amendt et al. [36] Transverse r = DC, 0.90 upper Very high upper spine
Orthopaedics Systems Inc. scoliometer 0.88 lower High lower spine
SC, 0.92 upper Very high single curve
0.96 lower
Galatz et al. [42] Transverse r = 0.86 error SD = 1.28° ± 0.81° High
Korovessis et al. [38], Greece Transverse k = 0.64–0.92 High–very high
Scoliometer and mathematical formula
Balg et al. [37], Germany Transverse 0.9 Very high
Scolioguage app iPhone
Izatt et al. [39] Transverse ICC = 0.95 scoliometer Very high
iPhone with acrylic sleeve 0.92 iPhone
MD = 1.5° ± 2.1° and 1.1° ± 1.6°
DC double curve, SC single curve, Lx lumbar, Tx thoracic, Cx cervical, HH hump height, ATR angle trunk rotation, DPL distance from plum line,
SEM standard error of measurement, SD standard deviation, MD mean difference, ICC intraclass correlation coefficient, r Pearson’s correlation
coefficient, k kappa correlation coefficient, q Spearman’s rank correlation coefficient
Table 5 Intra-observer reliability data

Study Anatomical plane and method Intra-observer reliability Score
Coelho et al. [13], Brazil Transverse ICC = 0.92 Very high

Orthopaedic Systems Inc. scoliometer
Bonagamba et al. [43], Brazil Transverse ICC = upper Tx 0.74 Up Tx—high
Orthopaedic Systems Inc. scoliometer Mid Tx 0.92 Mid Tx—very high
Lower Tx 0.92 Lower Tx: high–very high
Lx 0.87
Zaina et al. [45], Italy Frontal, sagittal and transverse k = 0.28–0.41 Fair
Aesthetic index 98.7–99.4 % level agreement
Zaina et al. [45], Italy Frontal, sagittal and transverse k = 0.16–0.24 Poor–fair
TRACE 99.4–96.9 % level agreement
Driscoll et al. [40], Canada Transverse ICC = scoliometer 0.95 SD = 4.59° Very high
123
Eur Spine J
Table 5 continued
Study Anatomical plane and method Intra-observer reliability Score
Driscoll et al. [40], Canada Transverse ICC = scolioscreen 0.94 Scolioscreen: very high
Scolioscreen (spinologics) iPhone 0.89 iPhone: high
iPhone inclinometer
Murrel et al. [41], USA Transverse Tx q = 0.995 SD = 1.2° Very high
Orthopaedics Systems Inc. scoliometer Lx 0.998
SD = 1.6°
Amendt et al. [36] Transverse r = DC Tx 0.92–0.97 Very high
Orthopaedics Systems Inc. scoliometer Lx 0.91–0.92
SC Tx 0.95–094
Lx 0.90–0.86
Korovessis et al. [38], Greece Transverse k = 0.69–0.93 High–very high
Scoliometer and mathematical formula
Balg et al. [37] Transverse ICC = 0.96 MD = 0.0° ± 2.7° Very high
Scolioguage app iPhone
Izatt et al. [39] Transverse ICC = 0.939 scoliometer Very high
iPhone with acrylic sleeve 0.924 iPhone
MD = 0.9° ± 0.92° and 0.2° ± 1.6°
DC double curve, SC single curve, Lx lumbar, Tx thoracic, Cx cervical, HH hump height, ATR angle trunk rotation, DPL distance from plum line,
SEM standard error of measurement, SD standard deviation, MD mean difference, ICC intraclass correlation coefficient, r Pearson’s correlation
coefficient, k kappa correlation coefficient, q Spearman’s rank correlation coefficient
Table 6 Criterion Validity data

Study Anatomical plane and method Criterion related validity Score
Coelho et al. [13], Transverse r = 0.7 High

Brazil Orthopaedic Systems Inc. scoliometer p = 0.05
Sapkas et al. [44], Transverse r = 0.685 Moderate
Greece Orthopaedic Systems Inc. Scoliometer with p = 0.01
mathematical formulas
Fortin et al. [35] Frontal (anterior and posterior) and sagittal r = Tx 0.80 Tx, Tx kyphosis and trunk list
2D photography with X-RAY T-Lx 0.33 High,
Tx kyphosis 0.77 Lx Lordosis low
Lx lordosis 0.48
Trunk list 0.76
Fortin et al. [35] Frontal (anterior and posterior) and sagittal r = Tx 0.81–0.97 Tx high Tx-Lx, kyphosis and Lx
2D photography with 3D system T-Lx, Tx K, Lx lordosis lordosis low
0.30–0.56
Pearsall et al. [48], Transverse r = scoliometer 0.59 Moderate
Canada Scoliometer, back contour device p \ 0.25
BCD 0.70
p \ 0.005
Grosso et al. [46], Italy Transverse, frontal and sagittal r = HH 0.2–0.65 HH very low–mod
Hump height, ATR, distance from plumb line ATR = 0.1–0.5 ATR very low–mod
DPL = 0.2–0.6 PL very low–mod
Amendt et al. [36] Transverse r = 0.32–0.46 Low
Korovessis et al. [38], Transverse r = Tx p = 0.0254 N/A
Greece Scoliometer and mathematical formula Lx p = 0.0015
123
Eur Spine J
Table 6 continued
Study Anatomical plane and method Criterion related validity Score
Balg et al. [37] Transverse ICC = Tx 0.96 Very high

ATR scoliometer Lx 0.93
Scolioguage app iPhone MD = 0.4° ± 3°
DC double curve, SC single curve, Lx lumbar, Tx thoracic, Cx cervical, T-Lx Thoraco-lumbar, HH hump height, ATR angle trunk rotation, DPL
distance from plum line, SEM standard error of measurement, SD standard deviation, MD mean difference, ICC intraclass correlation coefficient,
r Pearson’s correlation coefficient, k kappa correlation coefficient, q Spearman’s rank correlation coefficient
MEH, Woolf SH (1993) Screening for adolescent idiopathic

Table 7 Levels of evidence relating to reliability and criterion va-
scoliosis: review article. JAMA 269(20):2667–2672. doi:10.
lidity of specific measurement methods
1001/jama.1993.03500200081038
Level of evidence Method 9. Bunnell WP (1984) An objective criterion for scoliosis screening.
J Bone Joint Surg 66(9):1381–1387
Strong Scolioguage app 10. Torell G, Nordwall A, Nachemson A (1981) The changing pat-
iPhone inclinometer tern of scoliosis treatment due to effective screening. J Bone Joint
Surg 63(3):337–341
Scolioscreen app
11. Coelho DM, Bonagamba GH, Oliveira AS (2013) Scoliometer
Orthopaedic Systems Inc. scoliometer measurements of patients with idiopathic scoliosis. Braz J Phys
Moderate Scoliometer with mathematical formula Ther 17(2):179–184. doi:10.1590/S1413-35552012005000081
Clinical examination 12. Kiely PJ, Grevitt MP (2008) Recent developments in scoliosis
surgery. Orthop Trauma 22(1):42–47. doi:10.1016/j.cuor.2007.
Limited TRACE 04.011
AI 13. Weinstein SL, Dolan LA, Spratt KF, Peterson KK, Spoonamore
2D photography MJ, Ponseti IV (2003) Health and function of patients with un-
treated idiopathic scoliosis: a 50-year natural history study.
Conflicting
JAMA 289(5):559–567
No evidence 14. Kotwicki T, Negrini S, Grivas TB, Rigo M, Maruyama T, Dur-
mala J, Zaina F, Members of the International Society on
Scoliosis Orthopaedic Rehabilitation and Treatment (2009)
Methodology of evaluation of morphology of the spine and the
trunk in idiopathic scoliosis and other spinal deformities—6th
SOSORT consensus paper. Scoliosis 4:26. doi:10.1186/1748-
References 7161-4-26
15. Knott P, Pappo E, Cameron M, Demauroy J, Rivard C, Kotwicki
1. Weinstein SL, Dolan LA, Cheng JCY, Danielsson A, Morcueride T, Zaina F, Wynne J, Stikeleather L, Bettany-Saltikov J, Grivas
JA (2008) Adolescent idiopathic scoliosis. Lancet TB, Durmala J, Maruyama T, Negrini S, O’Brien JP, Rigo M
371(9623):1527–1537. doi:10.1016/S0140-6736(08)60658-3 (2014) SOSORT 2012 consensus paper: reducing x-ray exposure
2. Wang WJ, Yeung HY, Chu WC, Tang NL, Lee KM, Qiu Y, in pediatric patients with scoliosis. Scoliosis 9:4. doi:10.1186/
Burwell RG, Cheng JC (2011) Top theories for the etiopatho- 1748-7161-9-4
genesis of adolescent idiopathic scoliosis. J Pediatric Orthop 31(1 16. Vidal C, Ilharreborde B, Azoulay R, Sebag G, Mazda K (2013)
Suppl):S14–S27. doi:10.1097/BPO.0b013e3181f73c12 Reliability of cervical lordosis and global sagittal spinal balance
3. Hawes MC, O’Brien JP (2006) The transformation of spinal measurements in adolescent idiopathic scoliosis. Eur Spine J
curvature into spinal deformity: pathological processes and im- 22(6):1362–1367. doi:10.1007/s00586-013-2752-2
plications for treatment. Scoliosis 1(1):3. doi:10.1186/1748-7161- 17. Grivas TB (2008) Conservative scoliosis treatment: section V
1-3 clinical evaluation and classification. Studies in health technology
4. Rigo M (2011) Patient evaluation in idiopathic scoliosis: radio- and informatics, vol 135. IOS Press, Amsterdam
graphic assessment, trunk deformity and back asymmetry. 18. Kowalski IM, Protasiewicz-Faldowska H, Dwornik M, Pier-
Physiother Theory Pract 27(1):7–25. doi:10.3109/09593985. o_zyński B, Raistenskis J, Kiebzak W (2014) Objective parallel-
2010.503990 forms reliability assessment of 3 dimension real-time body pos-
5. Konieczny MR, Senyurt H, Krauspe R (2013) Epidemiology of ture screening tests. BMC Pediatr 14:221. doi:10.1186/1471-
adolescent idiopathic scoliosis. J Child Orthop 7(1):3–9. doi:10. 2431-14-221
1007/s11832-012-0457-4 19. Zabjek KF, Leroux MA, Coillard C, Rivard CH, Prince F
6. Fong DY, Lee CF, Cheung KM, Cheng JC, Ng BK, Lam TP, Mak (2005) Evaluation of segmental postural characteristics during
KH, Yip PS, Luk KD (2010) A meta-analysis of the clinical quiet standing in control and idiopathic scoliosis patients.
effectiveness of school scoliosis screening. Spine Clin Biomech 20(5):483–490. doi:10.1016/j.clinbiomech.2005.
35(10):1061–1071. doi:10.1097/BRS.0b013e3181bcc835 01.003
7. Sabirin J, Bakri R, Buang SN, Abdullah AT, Shapie A (2010) 20. Somoskeoy S, Tunyogi-Csapo M, Bogyo C, Illes T (2012) Ac-
School scoliosis screening programme-a systematic review. Med curacy and reliability of coronal and sagittal spinal curvature data
J Malays 65(4):261–267 based on patient-specific three-dimensional models created by the
8. Sox HC Jr, Berwick DM, Berg AO, Frame PS, Fryback DG, EOS 2D/3D imaging system. Spine J Off J North Am Spine Soc
Grimes DA, Lawrence RS, Wallace RB, Washington AE, Wilson 12(11):1052–1059. doi:10.1016/j.spinee.2012.10.002
123
Eur Spine J
21. Komeili A, Westover LM, Parent EC, Moreau M, El-Rich M, 36. Amendt LE, Ause-Ellias KL, Eybers JL, Wadsworth CT, Nielsen
Adeeb S (2014) Surface topography asymmetry maps categoriz- DH, Weinstein SL (1990) Validity and reliability testing of the
ing external deformity in scoliosis. Spine scoliometer. Phys Ther 70(2):108–117
14(6):973.e972–983.e972. doi:10.1016/j.spinee.2013.09.032 37. Balg F, Juteau M, Theoret C, Svotelis A, Grenier G (2014) Va-
22. Brink Y, Louw Q, Grimmer-Somers K (2011) The quality of lidity and reliability of the iPhone to measure rib hump in
evidence of psychometric properties of three-dimensional spinal scoliosis. J Pediatr Orthop 34(8):774–779. doi:10.1097/BPO.
posture-measuring instruments. BMC Musculoskelet Disord 0000000000000195
12:93. doi:10.1186/1471-2474-12-93 38. Korovessis PG, Stamatakis MV (1996) Prediction of Scoliotic
23. Fortin C, Feldman DE, Cheriet F, Gravel D, Gauthier F, Labelle cobb angle with the use of the scoliometer. Spine
H (2012) Reliability of a quantitative clinical posture assessment 21(14):1661–1666
tool among persons with idiopathic scoliosis. Physiotherapy 39. Izatt MT, Bateman GR, Adam CJ (2012) Evaluation of the
98(1):64–75. doi:10.1016/j.physio.2010.12.006 iPhone with an acrylic sleeve versus the scoliometer for rib hump
24. Shaw M, Adam CJ, Izatt MT, Licina P, Askin GN (2012) Use of measurement in scoliosis. Scoliosis 7(1):14. doi:10.1186/1748-
the iPhone for Cobb angle measurement in scoliosis. Eur Spine J 7161-7-14
21(6):1062–1068. doi:10.1007/s00586-011-2059-0 40. Driscoll M, Fortier-Tougas F, Labelle H, Parent S, Mac-Thong J
25. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, (2014) Evaluation of an apparatus to be combined with a
Ioannidis JP, Clarke M, Devereaux PJ, Kleijnen J, Moher D smartphone for the early detection of spinal deformities. Scoliosis
(2009) The PRISMA statement for reporting systematic reviews 25(9):10. doi:10.1186/1748-7161-9-10
and meta-analyses of studies that evaluate health care interven- 41. Murrell GA, Coonrad RW, Moorman CT 3rd, Fitch RD (1993)
tions: explanation and elaboration. Pub Lib Sci Med Coll An assessment of the reliability of the scoliometer. Spine
6(7):e1000100. doi:10.1371/journal.pmed.1000100 18(6):709–712
26. Brink Y, Louw QA (2012) Clinical instruments: reliability and 42. Galatz LM, Sturm PF, Bomze S, Simmens S (1999) Interobserver
validity critical appraisal. Int J Eval Clin Pract 18(6):1126–1132. reliability of scoliometer measurements of rotational deformity in
doi:10.1111/j.1365-2753.2011.01707.x adolescent idiopathic scoliosis. In: Reasearch into spinal defor-
27. Faul F, Erdfelder E, Lang AG, Buchner A (2007) G*Power 3: a mities 2, vol 59. ST HEAL T, pp 242–245
flexible statistical power analysis program for the social, behav- 43. Bonagamba GH, Coelho DM, Oliveira AS (2010) Inter and intra-
ioral, and biomedical sciences. Behav Res Methods rater reliability of the scoliometer. Revista brasileira de fisioter-
39(2):175–191 apia 14(5):432–438
28. van Tulder M, Furlan A, Bombardier C, Bouter L, Editorial 44. Sapkas G, Papagelopoulos PJ, Kateros K, Koundis GL, Boscainos
Board of the Cochrane Collaboration Back Review Group (2003) PJ, Koukou UI, Katonis P (2003) Prediction of Cobb angle in
Updated method guidelines for systematic reviews in the idiopathic adolescent scoliosis. Clin Orthop Relat Res 411:32–39.
cochrane collaboration back review group. Spine doi:10.1097/01.blo.0000068360.47147.30
28(12):1290–1299. doi:10.1097/01.BRS.0000065484.95996.AF 45. Zaina F, Negrini S, Atanasio S (2009) TRACE (trunk aesthetic
29. Munro BH, Visintainer MA (2005) Statistical methods for health clinical evaluation), a routine clinical tool to evaluate aesthetics
care research, vol 1, 5th edn. Lippincott, Philadelphia in Scoliosis patients: development from the aesthetic index (AI)
30. Chowanska J, Kotwicki T, Rosadzinski K, Sliwinski Z (2012) and repeatability. Scoliosis 4:3. doi:10.1186/1748-7161-4-3
School screening for scoliosis: can surface topography replace 46. Grosso C, Negrini S, Boniolo A, Negrini AA (2002) The validity
examination with scoliometer? Scoliosis 7:9. doi:10.1186/1748- of clinical examination in adolescent spinal deformities. Stud
7161-7-9 Health Technol Inform 91:123–125
31. DeWilde L, Plasschaert F, Cattoir H, Uyttendaele D (1998) Ex- 47. Patias P, Grivas TB, Kaspiris A, Aggouris C, Drakoutos E (2010)
amination of the back using the Bunnell scoliometer in a Belgian A review of the tunk surface metrics used as scoliosis and other
school population. Acta Orthop Belg 64(2):136–143 deformity evaluation indices. Scoliosis 5:12. doi:10.1186/1748-
32. deOliveira TS, Candotti CT, LaTorre M, Pelinson PP, Furlanetto 7161-5-12
TS, Kutchak FM, Loss JF (2012) Validity and reproducibility of 48. Pearsall DJ, Reid JG, Hedden DM (1992) Comparison of three
the measurements obtained using the flexicurve instrument to noninvasive methods for measuring scoliosis. Phys Ther
evaluate the angles of thoracic and lumbar curvatures of the spine 72(9):648–657
in the sagittal plane. Rehabil Res Pract 2012:186156. doi:10. 49. Lexell J, Downham D (2005) How to assess the reliability of
1155/2012/186156 measurements in rehabilitation. Am J Phys Med Rehab
33. Cote P, Kreitz BG, Cassidy JD, Dzus AK, Martel J (1998) A 84:719–723
study of the diagnostic accuracy and reliability of the scoliometer 50. Rankin G, Stokes M (1998) Reliability of assessment tools in
and Adam’s forward bend test. Spine 23(7):796–802 (discussion rehabilitation: an illustration of appropriate statistical analyses.
803) Clin Rehab 12(3):187–199
34. Fortin C (2010) Developpement et validation d’un outil clinique 51. Tricco AC, Tetzlaff J, Sampson M, Fergusson D, Cogo E,
pour l’analyse quantitative de la posture aupres de personnes Horsley T, Moher D (2008) Few systematic reviews exist
atteintes d’une scoliose idiopathique. Universite de Montreal, documenting the extent of bias: a systematic review. J Clin
Montreal Epidemiol 61(5):422–434. doi:10.1016/j.jclinepi.2007.10.017
35. Fortin C, Feldman DE, Cheriet F, Labelle H (2010) Validity of a
quantitative clinical measurement tool of trunk posture in idio-
pathic scoliosis. Spine 35(19):E988–E994. doi:10.1097/BRS.
0b013e3181cd2cd2
123
J Pediatr (Rio J). 2017;93(4):365---373
www.jped.com.br
ORIGINAL ARTICLE
Equations based on anthropometry to predict body fat

measured by absorptiometry in schoolchildren
and adolescents夽
Luis Ortiz-Hernández ∗ , A. Valeria Vega López, Norma Ramos-Ibáñez,
L. Joana Cázares Lara, R. Joab Medina Gómez, Diana Pérez-Salgado
Universidad Autónoma Metropolitana (UAM), Departamento de Salud, Mexico City, Mexico
Received 17 April 2016; accepted 24 August 2016

Available online 26 January 2017
KEYWORDS Abstract
Body fat; Objective: To develop and validate equations to estimate the percentage of body fat of children
Children; and adolescents from Mexico using anthropometric measurements.
Adolescents; Methods: A cross-sectional study was carried out with 601 children and adolescents from Mex-
Mexico; ico aged 5---19 years. The participants were randomly divided into the following two groups: the
Dual X-ray development sample (n = 398) and the validation sample (n = 203). The validity of previously pub-
absorptiometry; lished equations (e.g., Slaughter) was also assessed. The percentage of body fat was estimated
Anthropometry by dual-energy X-ray absorptiometry. The anthropometric measurements included height, sit-
ting height, weight, waist and arm circumferences, skinfolds (triceps, biceps, subscapular,
supra-iliac, and calf), and elbow and bitrochanteric breadth. Linear regression models were
estimated with the percentage of body fat as the dependent variable and the anthropometric
measurements as the independent variables.
Results: Equations were created based on combinations of six to nine anthropometric variables
and had coefficients of determination (r2 ) equal to or higher than 92.4% for boys and 85.8% for
girls. In the validation sample, the developed equations had high r2 values (≥85.6% in boys and
≥78.1% in girls) in all age groups, low standard errors (SE ≤ 3.05% in boys and ≤3.52% in girls),
and the intercepts were not different from the origin (p > 0.050). Using the previously published
equations, the coefficients of determination were lower, and/or the intercepts were different
from the origin.
Conclusions: The equations developed in this study can be used to assess the percentage of
body fat of Mexican schoolchildren and adolescents, as they demonstrate greater validity and
lower error compared with previously published equations.
© 2017 Sociedade Brasileira de Pediatria. Published by Elsevier Editora Ltda. This is an open
access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/
4.0/).
夽 Please cite this article as: Ortiz-Hernández L, Vega López AV, Ramos-Ibáñez N, Cázares Lara LJ, Medina Gómez RJ, Pérez-Salgado D.
Equations based on anthropometry to predict body fat measured by absorptiometry in schoolchildren and adolescents. J Pediatr (Rio J).
2017;93:365---73.
∗ Corresponding author.
E-mail: lortiz@correo.xoc.uam.mx (L. Ortiz-Hernández).
http://dx.doi.org/10.1016/j.jped.2016.08.008
0021-7557/© 2017 Sociedade Brasileira de Pediatria. Published by Elsevier Editora Ltda. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
366 Ortiz-Hernández L et al.
PALAVRAS-CHAVE Equações com base na antropometria para prever a gordura corporal medida
Gordura corporal; por absorciometria em crianças em idade escolar e adolescentes
Crianças;
Resumo
Adolescentes;
Objetivo: Desenvolver e validar equações para estimar o percentual de gordura corporal (%GC)
México;
de crianças e adolescentes do México utilizando medidas antropométricas.
Absorciometria de
Métodos: Foi realizado um estudo transversal com 601 crianças e adolescentes do México com
dupla energia de
idades entre 5 e 19 anos. Os participantes foram divididos aleatoriamente nos seguintes dois
raios-X;
grupos: a amostra de desenvolvimento (n = 398) e a amostra de validação (n = 203). A validade
Antropometria
das equações publicadas anteriormente (por exemplo, Slaughter) também foi avaliada. O %GC
foi estimado por absorciometria de dupla energia de raios-X (raio-X de dupla energia (DXA)). As
medidas antropométricas incluíram estatura, altura sentado, peso, circunferências da cintura
e do braço, dobras cutâneas (tríceps, bíceps, subescapular, suprailíaca e panturrilha) e larguras
do cotovelo e bitrocantérica. Os modelos de regressão linear foram estimados com o %GC sendo
a variável dependente e as medidas antropométricas como as variáveis independentes.
Resultados: As equações foram criadas com base nas combinações de seus a nove variáveis
antropométricas e apresentaram coeficientes de determinação (r2 ) iguais ou superiores a 92,4%
para meninos e 85,8% para meninas. Na amostra de validação, as equações desenvolvidas apre-
sentaram altos valores de r2 (≥ 85,6% em meninos e ≥ 78,1% em meninas) em todos os grupos,
baixo nível de erros padrão (EP ≤ 3,05% em meninos e ≤ 3,52% em meninas) e os interceptos não
foram diferentes da origem (p > 0,050). Utilizando as equações publicadas anteriormente, os
coeficientes de determinação foram menores e/ou os interceptos foram diferentes da origem.
Conclusões: As equações desenvolvidas neste estudo podem ser utilizadas para avaliar o %GC
das crianças em idade escolar e adolescentes mexicanos, pois elas possuem uma maior validade
e menor erro em comparação às equações publicadas anteriormente.
© 2017 Sociedade Brasileira de Pediatria. Publicado por Elsevier Editora Ltda. Este é um artigo
Open Access sob uma licença CC BY-NC-ND (http://creativecommons.org/licenses/by-nc-nd/4.
0/).
Introduction subjects. Slaughter equations5 are the most frequently

used in the evaluation of body composition in children
Overweight and obesity are important public health prob- and adolescents.6---9 These equations were developed in
lems because of their magnitude and impact on health. 310 Afro-American and Euro-American people from 8 to 29
In 2012, prevalence of overweight and obesity in Mexi- years of age.5 However, these equations tend to overes-
can schoolchildren 5---11 years old was 34.4% (19.8% and timate body fat.8,9 In recent years, Dezenberg equations
14.6%, respectively) and, in adolescents 12---19 years old, have been used,7 which were developed in a sample of
it was 34.9%.1 In schoolchildren, this value was slightly 202 Afro- and Euro-American children aged 4---10 years
higher (34.8%) in 2006, whereas in adolescents, it was lower of age. In Latino children from the USA, these equations
(33.2%). inaccurately estimate body fat.10 These results show that
Children with obesity tend to become adults with obe- equations developed in an ethnic group can be used in
sity, which increases their risk for the development of other populations, but the obtained estimation could be
chronic diseases.2 Obesity represents an economic prob- inaccurate. In the case of the Latino population, it has
lem because the treatment of its co-morbidities carries high been recognized that the present-day Mexican population
costs.3 Hence, it is necessary that obesity is accurately diag- is an admixture among Amerindian, European, and African
nosed. ancestries.11
Different methods for assessing body fat include hydro- Few studies in Latin America have explored the validity
densitometry, dual energy X-ray absorptiometry (DXA), and of prediction equations to estimate body fat in children
deuterium dilution. These measures are used mainly in using anthropometric measurements.12,13 It is possible that
research settings because the required equipment is costly equations developed in populations of European or African
and must be operated by specialized technicians. For these ancestry could not be applied to the Latin American popu-
reasons, these methods are unsuitable for population stud- lation due to ethnic differences in amount and distribution
ies, clinical practice, and screening activities. On the of body fat. For example, in comparison with European-
contrary, anthropometric measurements --- although indirect or African-descended children, those with Latino ancestry
indicators of adiposity --- are economical, non-intrusive, and have higher waist circumference.14 In addition, pediatric
highly reproducible.4 Latin American populations have a high weight in relation to
Various equations based on anthropometry have been their height (an indirect indicator of adiposity), which is not
used to estimate body fat in children, and these meth- always is due to excess body fat.15,16 Other reasons for high
ods have been developed in Euro- and Afro-descendent weight without excess adiposity could be that in comparison
Prediction of body fat in Mexican population 367
with the reference population, Latino populations have USA) Calibration, scanning, and image analysis were com-
lower heights15,16 but higher values of muscle mass,16 pleted following the manufacturer procedures and those
fat-free mass hydration,17 trunk length,15 body frame (i.e., described by the International Society for Clinical Densit-
bone thickness),15 and thorax circumference.18 Therefore, ometry (ISCD).24 The images were analyzed using the whole
the objectives of this study were (1) to evaluate the validity body option for Hispanic populations, model APEX version
®
of previously published equations to estimate %BF in Latino 3.3.0.1 (Hologic , MA, USA). The coefficients of variation
children, and (2) to develop and validate new equations in for the technician were 1.27% for total body fat (TBF) in
Latino children to predict body fat using DXA as the gold kilograms, 0.66% for lean body mass, and 1.04% for %TBF,
standard. which are acceptable according to the ISCD.24
The distribution of anthropometric variables was
assessed using the Kolmogorov---Smirnov test (Table 1). When
Methods the variable had a biased distribution, it was transformed
using the base-10 logarithm to achieve a closer symmetric
A cross-sectional study was carried out with a convenience distribution. Linear regression models were estimated using
sample of schoolchildren and adolescents aged 5---19 years DXA derived %BF as a dependent variable and anthropomet-
from Mexico City. The participants were recruited from ele- ric measurement as independent variables. Separate models
mentary (n = 7), junior high (n = 8), and high schools (n = 4) in were estimated for each sex. The equations were developed
Mexico City and a community center in the State of Mexico. using the following two procedures:
The research team presented the project to the principal
of each school, and all students were invited to participate. (a) Different combinations of independent variables were
Additionally, children of employees from the Metropolitan manually tested. First, the capacity of each anthro-
Autonomous University at Xochimilco (UAM-X) participated pometric variable to estimate %BF was evaluated
in the study. Children with a plaster cast or motor disabili- (Supplementary Table S1). In addition, scatter plots
ties were excluded from the study. The study was approved were graphed to verify the linear relationship among
by the Division of Biological and Health Sciences from UAM- the variables. A curvilinear relationship of %BF with age
X. The participants and their parents or guardians were was evident in boys; meanwhile, among girls, this rela-
informed about the research procedures, and the latter pro- tionship was linear. For this reason, in models for boys,
vided informed consent. the quadratic term of age was incorporated. Because
Following the procedures used in other studies,7,10,19 age and maturity are related, separate models with
the total sample (n = 601) was randomly divided into two each variable were estimated, and no differences were
groups. Data from two-thirds of the participants were used observed. For simplicity, only models with age were
to develop equations (n = 398), and data from the remaining reported. The coefficient of determination (r2 ) and
one-third of the children were used to validate new and standard error (SE) were used as criteria to identify the
previously published equations (n = 203). The SPSS software variables that better predicted %BF. A variable was con-
package (IBM Corp. 2010. IBM SPSS Statistics for Windows, sidered an adequate predictor when the p-value was
version 20.0. NY, USA) was used to generate these samples <0.050.
(option ‘‘Random sample of cases’’ in the command ‘‘Select (b) The automatic backward and forward procedures of the
cases’’). linear regression command of SPSS software were used
Most anthropometric measurements were performed fol- (Table 2). In the backward option, all variables were
lowing the techniques described by Lohman et al.20 Waist introduced in the models, and then those that had a low
circumference was measured according to the technique partial correlation with the dependent variable were
described by Fernández et al.14 Observers were standardized excluded (elimination criteria: p > 0.100). In the for-
following the Habicht protocol.21 Participants were mea- ward option, the variables were introduced one-by-one
sured with light clothes and without shoes. Stature was according to greater correlation with the dependent
®
measured with a portable stadiometer, model 214 (SECA , variable. These variables were introduced and kept
SP, Brazil) and body weight with a digital scale, model 813 if they significantly predicted the dependent variable
®
(SECA , SP, Brazil). Waist and arm circumferences were (p < 0.050).
®
obtained with a metallic tape (Lufkin , MD, USA). Skinfolds
(triceps, bicipital, subscapular, suprailiac, and calf) were The assessment of the validity of the previously published
®
measured with a compass (Harpenden , Mediflex Produ- equations was carried out using age-matched subjects from
tos Cirúrgicos, NY, USA). Each skinfold was measured three the population that was used to develop them (e.g., for the
times, and the average value was analyzed. Bitrochanteric Slaughter equations,25 participants data from 8 to 18 years
breadth and sitting stature were measured with an anthro- old were used and for the Deurenberg equations26 the range
®
pometer (Harpenden , Mediflex Produtos Cirúrgicos, NY, was 7---19 years). The validity of Stevens equations for girls
USA). Elbow breadth was measured with an anthropometer could not be assessed because there was no data for children
®
(Futrex , MD, USA). of menarche age. In these regression models, the dependent
Body mass index (BMI: body weight/height2 ) was calcu- variable was DXA derived %BF, and the independent variable
lated as a Z-score for age and sex using the World Health was %BF obtained by equations previously developed and
Organization reference.22 Maturity was estimated using the those developed in this study (Table 3). One criterion for
equations of Mirwald et al.23 validity was whether there was a significant difference
The percentage of body fat (%BF) was evaluated with between the intercept and the origin (p > 0.050) because,
®
DXA using equipment Model Discovery Wi (Hologic , MA, if there is difference, the equations would systematically
Table 1 Descriptive statistics of the anthropometric measures in the development and validation samples of Mexican children
and adolescents.
Development sample Validation sample
Boys (n = 191) Girls (n = 207) Boys (n = 100) Girls (n = 103)
% (n) % (n) % (n) % (n)

Sexual maturity
Prepubescent 66.0 (126) 30.9 (64) 75.0 (75) 42.7 (44)
Pubescent 27.2 (52) 38.6 (80) 22.0 (22) 34.0 (35)
Post-pubescent 6.8 (13) 30.4 (63) 3.0 (3) 23.3 (24)
Mean ± SD Mean ± SD Mean ± SD Mean ± SD

Body fat (%) 27.1 ± 9.4 32.6 ± 7.0 25.8 ± 9.0 32.8 ± 7.4
Age (years) 11.3 ± 2.7 11.5 ± 2.9 10.9 ± 2.6 10.9 ± 2.9
Age2 (years2 ) 136.0 ± 59.2a 141.1 ± 67.5a 133.3 ± 56.0 127.9 ± 62.6
Height (cm) 149.2 ± 16.9 146.2 ± 13.1c 144.2 ± 15.3 145.2 ± 13.7
Sitting height (cm) 76.9 ± 8.1 76.5 ± 7.2b 74.3 ± 7.5 75.5 ± 7.9
Sitting height ratio (%) 51.7 ± 1.8a 52.3 ± 1.9 51.6 ± 2.6 52.0 ± 2.7a
Weight (kg) 47.4 ± 15.9 45.0 ± 14.2 42.0 ± 14.3 43.7 ± 14.0
Body mass index (kg/m2 ) 20.7 ± 4.2 20.5 ± 4.1 19.7 ± 3.8 20.3 ± 4.3
Arm circumference (cm) 24.2 ± 4.1 23.9 ± 3.8 23.2 ± 3.8 23.9 ± 3.7
Waist circumference (cm) 73.8 ± 12.9 72.0 ± 11.2 70.1 ± 11.4 72.5 ± 10.8
Triceps skinfold (mm) 14.9 ± 6.6a 16.7 ± 5.5 14.2 ± 6.3 17.2 ± 5.8
Biceps skinfold (mm) 8.3 ± 4.5c 9.0 ± 3.9a 7.7 ± 3.8a 9.8 ± 4.8a
Subscapular skinfold (mm) 14.2 ± 7.8a 15.7 ± 7.2 13.0 ± 7.6b 16.0 ± 9.2
Suprailiac skinfold (mm) 18.9 ± 12.7a 22.0 ± 12.0 16.0 ± 11.3a 21.9 ± 11.9
Calf skinfold (mm) 14.1 ± 6.7a 15.8 ± 6.5 12.4 ± 5.9a 16.0 ± 6.6
Bitrochanteric breadth (cm) 29.7 ± 4.9 29.5 ± 4.2 27.9 ± 4.4 29.3 ± 4.5
Elbow breadth (cm) 5.6 ± 0.6 5.1 ± 0.4 5.4 ± 0.6 5.1 ± 0.4
Significance by the Kolmogorov---Smirnov test.

a p ≤ 0.050.
b p ≤ 0.010.
c p ≤ 0.001.
Table 2 Multiple linear regression models having %BF by DXA as outcome and anthropometric characteristics as predictors in
the development sample of Mexican children and adolescents.
Equations Multiple linear regression models r2 (%) SE (%)

For boys
1 −8.739 − 0.384c (height, cm) + 35.371a (log weight, kg) − 0.892b (BMI, kg/m2 ) − 0.299a 93.4 2.5
(arm circumference, cm) + 0.258c (waist circumference, cm) + 17.732c (log triceps
skinfold) + 6.698b (log subscapular skinfold) + 3.545d (log suprailiac skinfold) + 4.019d
(log calf skinfold)
2 1.511 − 0.152c (height, cm) − 0.431c (arm circumference, cm) + 0.213c (waist 92.7 2.56
circumference, cm) + 21.271c (log triceps skinfold) + 7.386b (log subscapular
skinfold) + 4.981b (log suprailiac skinfold)
3 −1.791 + 0.329 (age, years) − 0.033a (age2 , years2 ) − 0.141c (height, cm) + 0.162c (waist 92.4 2.61
circumference, cm) + 18.516c (log triceps skinfold) + 6.013a (log subscapular
skinfold) + 5.633b (log suprailiac skinfold)
For girls
1 −7.299 − 21.436b (log sitting height) + 17.739b (log BMI) + 20.143a (log waist 85.8 2.65
circumference) + 7.813c (log triceps skinfold) + 6.379b (log subscapular skinfold) + 6.051b
(log calf skinfold) − 16.364a (log elbow breadth)
%BF by DXA, percentage of body fat estimated by dual-energy X-ray absorptiometry; BMI, body mass index; SE, standard error.
p-Values of regression coefficients were:
a p < 0.050.
b p < 0.010.
c p < 0.001.
d p ≤ 0.072.
Table 3 Cross-validation of new and previously published prediction equations in validation sample of Mexican children and
adolescents.
Boys Girls
2 2
Min Max Mean r (%) SE (%) ˛ p Min Max Mean r (%) SE (%) ˛ p
8---18 years n = 85 n = 84
%BF DXA 11.3 47.2 25.1 17.2 52.5 32.9
Equation 1 11.6 44.0 25.6 91.6 2.59 −0.64 0.478 17.2 45.3 33.0 78.1 3.52 0.39 0.841
Equation 2 12.2 44.6 25.7 90.7 2.72 −0.49 0.603
Equation 3 10.6 43.3 25.2 91.5 2.60 0.25 0.776
Slaughtera 8.6 44.7 20.2 78.2 4.17 6.75 0.000 12.4 47.3 25.6 62.8 4.60 11.79 0.000
Slaughterb 10.6 54.1 23.7 80.7 3.92 6.73 0.000 11.2 60.6 27.6 51.0 5.27 14.96 0.000
7---19 years n = 93 n = 92
%BF DXA 11.3 47.2 25.5 17.2 52.5 33.3
Equation 1 11.6 44.0 26.0 91.5 2.66 −0.79 0.370 17.2 45.3 33.3 79.1 3.42 0.45 0.806
Equation 2 12.2 44.6 26.2 90.6 2.78 −0.75 0.423
Equation 3 10.6 43.3 25.6 91.5 2.65 0.03 0.972
Deurenbergc 10.2 31.4 19.5 80.2 4.05 −6.85 0.000 13.8 35.8 23.6 67.6 4.25 3.70 0.097
Deurenbergd 11.3 34.5 20.8 84.8 3.55 −6.54 0.000 13.8 40.1 26.1 67.0 4.29 4.97 0.023
4---10 years n = 29 n = 33
%BF DXA 15.6 44.7 28.9 17.2 45.8 33.1
Equation 1 17.0 39.9 29.5 88.9 2.82 −2.14 0.335 17.2 44.9 32.3 82.1 3.25 2.11 0.430
Equation 2 17.9 40.2 29.8 88.5 2.86 −3.65 0.128
Equation 3 17.6 40.8 29.5 90.2 2.64 −2.59 0.219
Dezenberge 12.7 36.1 25.3 68.1 4.77 −0.57 0.888 12.9 36.4 25.5 62.9 4.65 10.48 0.002
Dezenbergf 15.2 36.2 25.2 77.4 4.02 0.42 0.892 11.3 38.4 26.4 79.8 3.43 7.67 0.002
Dezenbergg 13.1 34.2 23.6 77.7 3.99 2.50 0.385 13.8 38.5 27.5 79.8 3.43 3.95 0.147
Dezenbergh 12.5 33.6 23.0 77.7 4.00 3.32 0.237 12.8 37.8 26.7 79.8 3.43 5.28 0.046
7---13 years n = 66 n = 53
%BF DXA 12.0 44.7 27.3 17.2 52.5 33.0
Equation 1 11.6 41.8 27.6 87.1 2.89 0.17 0.900 17.2 45.3 32.3 83.4 3.36 −0.48 0.822
Equation 2 12.2 42.5 27.8 85.6 3.05 −0.03 0.982
Equation 3 10.7 43.0 27.3 87.8 2.81 0.82 0.522
Huangi 0.9 42.5 24.2 61.9 4.97 13.51 0.000 5.6 46.2 28.2 63.5 4.96 14.25 0.000
Huangj −1.6 40.7 21.9 64.4 4.81 14.85 0.000 9.7 47.9 31.0 63.5 4.96 11.10 0.000
Huangk 4.1 40.8 23.9 58.9 5.16 11.98 0.000 −6.5 44.5 22.7 67.3 4.70 19.47 0.000
8---17 years n = 85
%BF DXA 11.3 47.2 25.1
Equation 1 11.6 44.0 25.6 91.6 2.59 −0.64 0.478
Equation 2 12.2 44.6 25.7 90.7 2.72 −0.49 0.603
Equation 3 10.6 43.3 25.2 91.5 2.59 0.25 0.776
Stevensl 20.6 49.2 31.8 83.4 3.63 −10.89 0.000
%BF DXA, percentage of body fat estimated by dual-energy X-ray absorptiometry; Min, minimum; Max, maximum; SE, standard error;
r2 , coefficient of determination; ˛, intercept; T-SF, triceps skinfold; B-SF, biceps skinfold; SS-SF, subscapular skinfold; SI-SF, supra-iliac
skinfold; C-SF, calf skinfold.
The estimations were made based on:
a T-SF and C-SF.
b T-SF and SS-SF.
c T-SF and B-SF.
d T-SF, B-SF SI-SF, and SS-SF.
e Weight.
f Weight and T-SF.
g Weight, T-SF, and gender.
h Weight, T-SF, gender, and ethnicity.
i Age.
j Age and gender.
k Age, gender, and height.
l Age, race, weight, height, waist, and T-SF.
Equation 1 for boys Equation 3 for boys

A C
Difference = 2080 – 0.060 average 10.00 Difference = 1201 – 0.040 average
Difference in %BF (predicted-measured)

p of β : 0.012 p of β : 0.144
7.50
5.00 5.00
2.50
0.00
0.00
–5.00
–2.50
–5.00
–10.00
10.00 20.00 30.00 40.00 50.00 10.00 20.00 30.00 40.00 50.00
Average %BF by DXA and equation 1 (boys) Average %BF by DXA and equation 3
Equation 2 for boys Equation 1 for girls

B D
Difference = 2198 – 0.057 average 10.00 Difference = 3564 – 0.106 average

p of β : 0.013 p of β : 0.028
10.00
5.00
5.00 0.00
–5.00
0.00
–10.00
–5.00
–15.00
10.00 20.00 30.00 40.00 50.00 10.00 20.00 30.00 40.00 50.00
Average %BF by DXA and equation 2 Average %BF by DXA and equation 1 (girls)
Figure 1 Bland---Altman graphs for the difference in %BF measured by DXA and estimated %BF by new prediction equations based
on anthropometry in the validation sample of Mexican children and adolescents. %BF by DXA, percentage of body fat estimated by
dual-energy X-ray absorptiometry.
over- or under-estimate %BF. Finally, the Bland Altman Results

plots were graphed (Fig. 1 and Supplementary Fig. S1) to
verify whether bias existed in the estimations according A descriptive analysis of the anthropometric variables in the
to adiposity levels. Differences between the measured and development and validation samples is shown in Table 1.
estimated %BF were plotted against the average of mea- The mean age of the development sample was 11.88 years,
sured and estimated %BF. Only equations whose intercept which is higher than that of the validation sample (11.32
was not different from the origin were plotted. The sta- years, p = 0.019). In both samples, half of the participants
tistical analyses were performed using SPSS software (IBM were girls (development sample: 52%, validation sample:
Corp. 2010. IBM SPSS Statistics for Windows, version 20.0. 50.7%). Although most participants were pre-pubescent, the
NY, USA) proportion was higher in the validation sample than in the
Statistical power analysis for different scenarios was development sample (58.6% vs. 47.7%, p < 0.050).
completed using the G*Power software.27 For the devel- For each sex group, different anthropometric character-
opment of equations (Table 2) given a significance level istics were related to %BF (Supplementary Table S1). In boys,
(˛) of 0.050, a sample size of 207, an effect of 85.8% the variables that better predicted %BF were age, stature,
(r2 ), and seven predictors (scenario for the girls’ equa- BMI, circumferences of arm and waist, bitrochanteric and
tion), the statistical power was 1.00. The same result was elbow breadth, sitting stature, weight, sitting stature index,
obtained with the scenario for the three equations for boys. and five skinfolds. Among girls, the best predictors were
For the validation analysis (Table 3), the power obtained age, stature, weight, sitting stature, sitting stature index,
under the different scenarios with respect to the sample BMI, circumference of arm and waist, five skinfolds, and
size (n from 29 to 93), the effect (r2 from 51.0 to 91.6%) bitrochanteric and elbow breadth.
and the number of predictors (1---7) was always satisfactory Using automated procedures, the two models satisfac-
(i.e., ≥0.80). torily predicted %BF in boys (Table 2). With the backward
option, the following variables were included in the model children with obesity,8 and British children.6 The Deurenberg
(called Equation (1)): stature, BMI, arm and waist circum- equations26 were developed in 379 Dutch children and ado-
ferences, body weight, and triceps, subscapular, suprailiac, lescents from 7 to 20 years of age and had low to moderate
and calf skinfolds (r2 = 93%, SE = 2.5%). With the forward predictive capacity (r2 = 41% to 69% in the total sample). In
option, the model (called Equation (2)) included the fol- the Mexican pediatric population, these equations underes-
lowing variables: stature, arm and waist circumferences, timated %BF in boys and overestimated it in girls. However,
and triceps, subscapular, and suprailiac skinfolds (r2 = 92%, the opposite effect occurred in British children.6 Physical
SE = 2.5%). In boys, manual selection produced a model differences of Mexican children and adolescents in com-
(Equation (3)) that included age, stature, waist circum- parison with those who participated in the Slaughter25
ference, and triceps, subscapular, and suprailiac skinfolds and Deurenberg26 studies could explain the results found
(r2 = 92%, SE = 2.0%). In girls, both of the automated pro- in the present study. Participants in the Slaughter25 and
cedures produced the same solution (called Equation (1)) Deurenberg26 samples were taller (155.2 cm and 164.3 cm
with the following variables: sitting stature, BMI, waist cir- vs. 144.2 cm, respectively) and heavier (47.4 kg and 53.9 kg
cumference, triceps, elbow breadth, and subscapular and vs. 42 kg, respectively) than children in the Mexican sam-
calf skinfolds (r2 = 85%, SE = 2.5%). Among girls, the manual ple but had lower %BF (16.7% and 15.1% vs. 25.8%,
selection model did not adequately predict %BF (r2 ≥ 80.0%). respectively).
Table 3 shows the different estimations of %BF obtained In Mexican schoolchildren and adolescents, the estima-
by equations in the validation sample. In all of the tions obtained with the Dezenberg equations did not tend
age groups, the newly developed equations in Mexican to systematically over- or underestimate %BF (i.e., the
schoolchildren and adolescents had high coefficients of intercept did not differ from the origin). However, the pre-
determination (r2 ≥ 85% in boys and ≥78% in girls) and the dictive capacity of the Dezenberg equations (r2 = 68.1% to
lowest SE (≤3.05% in boys and ≤3.52% in girls); in addition, 77.4% in boys and 62.9% to 79.8% in girls) was lower than
their intercepts did not differ from the origin (p > 0.050). those of the equations developed in this study (88.5%---90.2%
Although with the Dezenberg equations the intercept did and 82.1%, respectively). Similar findings were obtained in
not differ from the origin (p > 0.050), their predictive capac- Cuban schoolchildren.28 Nevertheless, in Latino10 and Indian
ity (r2 ≥ 77.7 in boys and <79.8 in girls) was lower than schoolchildren,9 the Dezenberg equations underestimated
that of the equations developed here (r2 ≥ 88 and 82%, %BF; whereas in British schoolchildren theses equations had
respectively). The Stevens equations in boys explained a sig- a bias of 13% in the estimation of %BF measured by deuterium
nificant proportion of variance (r2 = 85%), but the intercept dilution.29
differed from the origin. In both sexes, the Slaughter equa- The Huang equations10 were developed in 96 Latino
tions overestimated %BF (positive intercept, p < 0.050). The schoolchildren from 7 to 13 years of age from the USA and
Deurenberg equations underestimated %BF in boys (negative had an acceptable predictive capacity (r2 = 86%---97% in the
intercept, p < 0.050) and overestimated in girls. The Huang total sample). In the Mexican sample these equations over-
equations overestimated %BF in both sexes. estimated %BF. Although participants of the Huang study
The Bland---Altman plots (Fig. 1A, B and D) revealed that were Latino, they were heavier than the Mexican schoolchil-
developed equations through automatic procedures produce dren in the current sample. These differences could reflect
biased estimations of %BF. These equations overestimated the different environments where children from the USA and
values in children with low adiposity, while in those with Mexico live.
high adiposity, values were underestimated. The equation The Stevens equations19 were developed with a multi-
developed by manual selection in boys yielded estimations ethnic sample of 5374 subjects 8---17 years of age and had a
without systematic bias (Fig. 1C). predictive capacity up to 85% in boys using DXA as the gold
standard. This equation tended to underestimate the %BF in
Mexican schoolchildren and adolescents.
Discussion The Latin American population has certain physical char-
acteristics such as lower height, short lower extremities,
The main aim of this study was to develop equations based higher levels of body fat in the trunk and abdominal regions,
on anthropometric measurements to predict %BF for Mexican and larger body frame.14---16 For this reason, in the present
schoolchildren and adolescents. The four developed equa- study, body characteristics related to linear growth (i.e.,
tions require measures related to adiposity, linear growth, stature or sitting stature) and body frame (i.e., elbow
and body frame. For boys, three equations explained a or bitrochanteric breadth) were measured in addition to
high proportion of the %BF variance (r2 ≥ 92%). Although the those related to adiposity (i.e., skinfolds and waist cir-
equation for girls explained a high proportion of variance cumference). In the equations developed for boys, stature
(85%), this was lower than that for boys. Another aim of the had a negative regression coefficient; whereas in girls, the
study was to analyze the validity of previously developed coefficients for elbow breadth and sitting stature index were
equations. With the exception of Dezenberg’s equations, also negative. This shows that children or adolescents with
none of previously published equations were valid in the higher stature or larger body frames tend to have less adi-
sample of Mexican schoolchildren and adolescents. posity. In addition, in the new equation measurements of
In its original population,5 the Slaughter equations had abdominal and trunk fat were predictors of %BF, includ-
acceptable predictive capacity (r2 from 75% to 78% in ing waist circumference and subscapular and suprailiac
the total sample). However, in the Mexican schoolchildren skinfolds.
and adolescents, these equations overestimated the %BF. One limitation of the study is that a convenient sam-
The same trend was observed Indian children,9 Chinese ple was used. Therefore, generalization of these findings
should be done with caution. Another limitation is that 2. Must A, Anderson SE. Effects of obesity on morbidity in children
although DXA is considered an adequate method to measure and adolescents. Nutr Clin Care. 2003;6:4---12.
adiposity,30 its validity in a Mexican population has not been 3. Rtveladze K, Marsh T, Barquera S, Sanchez Romero LM,
established. In the development of the equations for Mexi- Levy D, Melendez G, et al. Obesity prevalence in Mexico:
impact on health and economic burden. Public Health Nutr.
can children and adolescents, less predictive capacity was
2014;17:233---9.
observed in girls than boys. This difference by sex has been
4. Frignani RR, Passos MA, Ferrari GL, Niskier SR, Fisberg M, Cintra
observed in other samples.19,26 In the future, measurements Ide P. Reference curves of the body fat index in adolescents and
should be identified that increase the predictive capacity their association with anthropometric variables. J Pediatr (Rio
of equations in girls. A main limitation of this research is J). 2015;91:248---55.
that a convenience sample of children and adolescents from 5. Slaughter MH, Lohman TG, Boileau RA. Relationship of anthro-
Mexico City was used, and, therefore, the sample had a pometric dimensions to lean body mass in children. Ann Hum
specific ethnic composition. In this way, although the devel- Biol. 1978;5:469---82.
oped equations provide an improved estimation of %BF in the 6. Reilly JJ, Wilson J, Durnin JV. Determination of body composi-
sample under study, their use in other populations should tion from skinfold thickness: a validation study. Arch Dis Child.
1995;73:305---10.
be extended with caution. In other words, the validity of
7. Dezenberg CV, Nagy TR, Gower BA, Johnson R, Goran MI. Pre-
these equations should be evaluated in other samples of chil-
dicting body composition from anthropometry in pre-adolescent
dren from other regions of Mexico and other Latin American children. Int J Obes Relat Metab Disord. 1999;23:253---9.
countries. This is necessary because the Latin American pop- 8. Chan DF, Li AM, So HK, Yin J, Nelson EA. New skinfold-thickness
ulation is diverse in terms of ethnic ancestry. For example, in equation for predicting percentage body fat in Chinese obese
the southeast region of Mexico, there is a predominance of children. HK J Paediatr (New Series). 2009;14:96---102.
Amerindian ancestry over European, whereas in the North- 9. Kehoe SH, Krishnaveni GV, Lubree HG, Wills AK, Guntupalli AM,
ern region the opposite trend is observed.11 Finally, the Veena SR, et al. Prediction of body-fat percentage from skinfold
predictive capacity of the estimated %BF (with the equa- and bio-impedance measurements in Indian school children. Eur
tions published here) to identify clinically relevant outcomes J Clin Nutr. 2011;65:1263---70.
10. Huang TT, Watkins MP, Goran MI. Predicting total body fat from
(i.e., cardiovascular risk) should be determined.
anthropometry in Latino children. Obes Res. 2003;11:1192---9.
In conclusion, the findings indicate that the equa-
11. Rubi-Castellanos R, Martínez-Cortés G, Muñoz-Valle JF,
tions developed based on anthropometric measurements González-Martín A, Cerda-Flores RM, Anaya-Palafox M, et al.
appropriately predict %BF in a heterogeneous group of Mex- Pre-hispanic mesoamerican demography approximates the
ican schoolchildren and adolescents. In contrast, equations present-day ancestry of Mestizos throughout the territory of
developed in other populations (even with similar ethnic Mexico. Am J Phys Anthropol. 2009;139:284---94.
characteristics) have reduced predictive capacity to esti- 12. Urrejola P, Hernández I, Icaza G, Velandia S, Reyes M, Hodgson
mate %BF. Therefore, the equations developed in this study I. Estimación de masa grasa en niños chilenos: ecuaciones de
can be used to assess the %BF in Mexican schoolchildren and pliegues subcutáneos vs densitometría de doble fotón. Rev Chil
adolescents, as they demonstrate greater validity and lower Pediatr. 2011;82:502---11.
error compared with the previously published equations. 13. Conlisk EA, Haas JD, Martinez EJ, Flores R, Rivera JD, Mar-
torell R. Predicting body composition from anthropometry and
bioimpedance in marginally undernourished adolescents and
Conflicts of interest young adults. Am J Clin Nutr. 1992;55:1051---60.
14. Fernández JR, Redden DT, Pietrobelli A, Allison DB. Waist cir-
cumference percentiles in nationally representative samples of
The authors declare no conflicts of interest.
African-American, European-American, and Mexican-American
children and adolescents. J Pediatr. 2004;145:439---44.
Acknowledgements 15. Martorell R, Mendoza FS, Castillo RO, Pawson IG, Budge CC.
Short and plump physique of Mexican-American children. Am J
Phys Anthropol. 1987;73:475---87.
The authors are grateful for the support of Magdalena
16. Ortiz-Hernández L, López Olmedo NP, Genis Gómez MT, Mel-
Rodriguez-Magallanes --- from the Nutrition, Body Compo-
chor López DP, Valdés Flores J. Application of body mass index
sition, and Energy Expenditure Unit --- who standardized to schoolchildren of Mexico City. Ann Nutr Metab. 2008;53:
the observers who carried out the anthropometric measure- 205---14.
ments. 17. Boutton TW, Trowbridge FL, Nelson MM, Wills CA, Smith EO,
Lopez de Romana G, et al. Body composition of Peruvian chil-
dren with short stature and high weight-for-height, I. Total
Appendix A. Supplementary data
body-water measurements and their prediction from anthropo-
metric values. Am J Clin Nutr. 1987;45:513---25.
Supplementary data associated with this article can 18. Trowbridge FL, Marks JS, Lopez de Romana G, Madrid S, Boutton
be found, in the online version, at doi:10.1016/j.jped. TW, Klein PD. Body composition of Peruvian children with short
2016.08.008. stature and high weight-for-height, II. Implications for the inter-
pretation for weight-for-height as an indicator of nutritional
status. Am J Clin Nutr. 1987;46:411---8.
References 19. Stevens J, Cai J, Truesdale KP, Cuttler L, Robinson TN, Roberts
AL. Percent body fat prediction equations for 8- to 17-year-old
1. Gutiérrez JP, Rivera-Dommarco J, Shamah-Levy T, Villalpando- American children. Pediatr Obes. 2014;9:260---71.
Hernández S, Franco A, Cuevas-Nasu L, et al. Encuesta nacional 20. Lohman TG, Roche AF, Martorell R. Anthropometric standard-
de salud y nutrición 2012, Resultados nacionales. Cuernavaca. ization reference manual. Champaign: Human Kinetics Books;
México: Instituto Nacional de Salud Pública; 2012. 1988.
21. Habicht JP. Standardization of quantitative epidemiological 26. Deurenberg P, Pieters JJ, Hautvast JG. The assessment of
methods in the field. Bol Oficina Sanit Panam. 1974;76:375---84. the body fat percentage by skinfold thickness measurements
22. de Onis M, Onyango AW, Borghi E, Siyam A, Nishida C, Siekmann in childhood and young adolescence. Br J Nutr. 1990;63:
J. Development of a WHO growth reference for school-aged chil- 293---303.
dren and adolescents. Bull World Health Organ. 2007;85:660---7. 27. Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible
23. Mirwald RL, Baxter-Jones AD, Bailey DA, Beunen GP. An assess- statistical power analysis program for the social, behavioral,
ment of maturity from anthropometric measurements. Med Sci and biomedical sciences. Behav Res Methods. 2007;39:175---91.
Sports Exerc. 2002;34:689---94. 28. Fernández-Vieitez JA. Estimación de la composición corporal
24. Hangartner TN, Warner S, Braillon P, Jankowski L, Shepherd J. por dos de las ecuaciones de Dezenberg para niños de 5 a 10
The official positions of the International Society for Clinical años. Rev Cubana Salud Pública. 2003;29:37---41.
Densitometry: acquisition of dual-energy X-ray absorptiome- 29. Wells JC. Predicting fatness in US vs UK children. Int J Obes
try body composition and considerations regarding analysis and Relat Metab Disord. 1999;23:1103.
repeatability of measures. J Clin Densitom. 2013;16:520---36. 30. Toombs RJ, Ducher G, Shepherd JA, De Souza MJ. The impact
25. Slaughter MH, Lohman TG, Boileau RA, Horswill CA, Stillman RJ, of recent technological advances on the trueness and preci-
Van Loan MD, et al. Skinfold equations for estimation of body sion of DXA to assess body composition. Obesity (Silver Spring).
fatness in children and youth. Hum Biol. 1988;60:709---23. 2012;20:30---9.

Validasi Antro Nurhidayah Tiasya Sanas-Compress0

Diunggah oleh

Informasi Dokumen

Judul Asli

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

Validasi Antro Nurhidayah Tiasya Sanas-Compress0

Diunggah oleh

Hak Cipta:

Format Tersedia

Nurhidayah Tiasya Sanas

Nama Peneliti Kesimpulan

Penentuan titik potong skor sindroma metabolik remaja dan penilaian

PENDAHULUAN perhitungan z-score. Penggunaan skor dalam mendeteksi

Jumlah remaja usia 15 – 24 tahun (unit analisis) yang mengikuti survei

Tidak terdapat data berat badan dan

Jumlah remaja, setelah drop missing value pertama

Tidak terdapat data lingkar perut

Jumlah remaja, setelah drop missing value kedua

Tidak terdapat data tekanan

Jumlah remaja, setelah drop missing value ketiga = 3368 orang

Nilai ekstrim (oulier) pada

Gambar 1. Alur pemilihan subjek penelitian

sindroma metabolik jika menderita minimal tiga dari HASIL

Tabel 2. Distribusi ukuran komponen sindroma metabolik berdasarkan jenis

Komponen sindroma Rerata

perempuan memiliki kadar kolesterol HDL lebih tinggi

daripada laki-laki dan bermakna secara statistik (Tabel

Penentuan titik potong skor sindroma metabolik

Penentuan titik potong secara statistik melalui dua

tahap yaitu pertama, dengan melakukan analisis regresi

0.00 0.25 0.50 0.75 1.00

CI:0,72 – 0,88), artinya SSM secara statistik berhubungan

terhadap sindroma metabolik adalah sebesar 0,9119 yang

menunjukkan bahwa SSM memiliki akurasi yang tinggi

secara statistik untuk mendiagnosis sindroma metabolik

pada remaja. Langkah selanjutnya adalah membuat graﬁk

potong. Titik potong probabilitas SSM diketahui dari

Sensitivity Specificity Sensitivity Specificity

Keterangan: A: Kurva ROC untuk laki-laki; B: Kurva ROC untuk perempuan

Keterangan: A: Prevalensi sindroma metabolik berdasarkan kriteria NCEP ATP-III; B: Prevalensi

Tabel 3. Perbandingan titik potong SSM berdasarkan jenis kelamin

n AUC 95% Conf. Interval Titik potong Sensitivitas Spesifisitas

Gambar 6. Perbedaan prevalensi sindroma metabolik berdasarkan jenis kelamin

A: AUC lingkar perut = 0,77

C: AUC IMT = 0,73

0.00 0.25 0.50 0.75 1.00

Prevalensi sindroma metabolik baik berdasarkan Penilaian diagnostik parameter antropometri

Tabel 4. Perbandingan uji diagnosis parameter antropometri terhadap kategori SSM

Parameter Sensitivitas Spesifisitas AUC 95% CI p

potong=71,50; sensitivitas=71%; spesifisitas=67%). remaja perempuan, baik menggunakan kriteria NCEP

Negative Results · January 2020

stunting View project

The user has requested enhancement of the downloaded file.

Model Prediksi Persen Lemak Tubuh Remaja Putri:

Nurul Khairani1, Trini Sudiarti2

ISSN 2528-3170 (printed media); ISSN 2541-5921 (online media) Nutri-Sains

52 Nutri-Sains: Jurnal Gizi, Pangan dan Aplikasinya, Vol 4 No 1

sebaik hasil pengukuran dengan menggunakan hydrostatic weighing jika pengukuran

ISSN 2528-3170 (printed media); ISSN 2541-5921 (online media) Nutri-Sains

Jenis dan Cara Pengambilan Data

54 Nutri-Sains: Jurnal Gizi, Pangan dan Aplikasinya, Vol 4 No 1

menggunakan skinfold caliper Harpenden, London. IMT WHO (Z Score) diukur

Tabel 1. Model prediksi

2. Deurenberg et al. (1991) P LT = 1,20 IMT + 0,23 U – 10,8 JK

ISSN 2528-3170 (printed media); ISSN 2541-5921 (online media) Nutri-Sains

Tabel 2. Status gizi responden berdasarkan IMT WHO (Z Score)

Tabel 3. Hasil pengukuran persen lemak tubuh dan antropometri

Hasil pengukuran persen lemak tubuh dengan menggunakan BIA pada

56 Nutri-Sains: Jurnal Gizi, Pangan dan Aplikasinya, Vol 4 No 1

Tabel 4. Koefisien korelasi antarvariabel penelitian

ISSN 2528-3170 (printed media); ISSN 2541-5921 (online media) Nutri-Sains

Berdasarkan hasil analisis multivariat akhir, maka dapat dibentuk model

tubuh baru (Sitoayu) = 23,28 + 1,56IMT WHO + 0,13ST - 0,62*U. Model