Track 3 Eating patterns and behaviour

P63
Effects of energy restriction on small intestinal absorption in normal and
colectomized rats
JM Paoli
1
, N Hwalla Baba
1
, CF Nassar
2
, and N Torbay
1
1
Department of Food Technology and Nutrition, American University of Beirut,
c=o New York Ofce, 850 3rd Ave., 18th Floor, New York, NY 10022-6297,
USA;
2
Department of Physiology, American University of Beirut, c=o New York
Ofce, 850 3rd Ave., 18th Floor, New York, NY 10022-6297, USA
Low- calorie diets for weight reduction are commonly used but their effects on
small intestinal absorption have been inadequately studied. The efciency of
absorption after short periods of energy restriction is unknown. The purpose of
this study was to measure the small intestinal absorption under and after
semistarvartion conditions.
Twenty-two adult male Sprague-Dawley rats were used. Thirteen rats were
colectomized and 9 were sham-operated. After a recovery period of 20 days,
the two groups of rats followed a dietary regimen consisting of three periods.
Period 1, rats received chow ad-libitum for 6 days, period 2 rats received 50%
of their basal intake for 9 days and period 3 rats were given the same amount
of food consumed in period 1 for 9 days. Dietary intake and body weight
change were recorded daily. Stools were collected every three days during the
three periods and analyzed for energy and macronutrient content. Relative
absorption of total energy (RA) was calculated as diet energy minus fecal
energy over diet energy100.
The results showed that body weight gain in the sham rats was
4.2 g=d0.39 during period 1 and 7.6 g=d0.37 during period 3(p<0.05)
while in colectomized rats weight change was similar in periods 1 and 3
(6.0g=d0.62 vs. 6.8g=d0.68). RA increased in the sham-operated rats
from 76.5% in period 1, to 80.4% and 80.2% in periods 2 and 3(p<0.05). RA
in colectomized rats decreased signicantly from 77.8% to 75.4% and 74.3%
respectively (p<0.05). Fecal nutrient analysis showed that RA of carbo-
hydrates in the sham rats was the major contributor to the increase of total
energy RA being 70%, 81% and 77.2% respectively (p<0.05) while in
colectomized rats it decreased signicantly from 67.4% to 67.2% and to
63.4% respectively in the three study periods (p<0.05).
It is concluded that caloric restriction leads to increased efciency of
absorption, mainly of carbohydrates, which may explain the higher weight
gain acquired during the post semi starvation period in the sham rats. The
colectomized rats failed to show a similar adaptation.
P64
Effect of a four versus three meal pattern on mid-nocturnal leptin level,
fat mass and substrate oxidation
C Chapelot
1
, C Marmonier
1
, R Aubert
1
, C Allegre
1
, N Gaussers
1
, L Brondel
1
,
M Fantino
1
, and J Louis-Sylvestre
1
1
Laboratoire de Physiologie du Comportement Alimentaire, UFR Leonard de
Vinci, 74 rue Marcel Cachin, 93017 Bobigny, France
The role of meal frequency in energy balance is still controversial. In France, 4
meal and 3 meal patterns mainly differ by the presence of an afternoon eating
episode called the gouter. We have previously found that in usual gouter
eaters (G), the gouter had the biological and behavioral characteristics of a
meal and could therefore be differenciated from a snack. In the present study,
we investigated the consequences of adding or removing this 4th meal on
body composition, substrate oxidation and plasma parameters involved in
eating behavior (glucose, insulin, leptin, fatty acids (FA) and triglycerides).
In 12 young normal-weight and non restrained G, the gouter was skipped
during 28 days whereas in 12 non gouter eaters (NG) matched for age and
BMI, a 4th meal was created by reducing the usual energy intake at lunch
meals (eaten each day at the laboratory) by 30%. Because of the nocturnal
zenith of plasma leptin levels, body composition was measured at 22:30 and
plasma parameters and substrate oxidation at 00:30 AM on days (D) 0 and 27.
Fat mass (FM) and fat-free mass (FFM) were assessed by Dual Energy X-Ray
Absorptiometry (DEXA) and substrate oxidation (estimated via RQ) by indirect
calorimetry (15min duration). On D1 and D28, subjects were time-blinded at
the laboratory and had to spontaneously request their meals. All meals were
consumed ad libitum.
Results showed that the procedure was effective to create a new sponta-
neous meal pattern (4 to 3 for G and 3 to 4 for NG). Removing the gouter
increased the mean BMI of G (0.220.06kg=m
2
, P =0.018) and specically
FM (428138g, P =0.013). Consistently, leptin levels were increased
(0.70.3ng.mL
71
, P =0.016). RQ was also increased (0.030.01,
P =0.028) and this increase was highly dependant on the individual change
in FFM (r =0.73, P =0.017). Other parameters were not altered. Creating a
4th meal in NG did not change any of these parameters. Across groups, leptin
levels and FM were found to be strongly correlated (r =0.90 and 0.93,
P <10
77
in D0 and D27 respectively). FA levels were weakly but signicantly
correlated with leptin (r =0.30, P =0.009), FM (r =0.36, P =0.01) and insulin
(r = 70.31, P =0.037).
Thus, this intervention study suggests that to switch from a 4 to a 3 meal
pattern leads to an increase in adiposity. Moreover, mid-nocturnal leptin level
is tightly and constantly associated with FM.
P65
Dietary high glycemic index associates with visceral obesity in middle-
aged men
IA Rogozina
1
, IV Dvoryashina
1
, and AA Korobitzyn
1
1
Endocrinology Center, City Clinical Hospital #1, Northern Medical State
University, 1, Suvorova, Arkhangelsk, 163045, Russia
INTRODUCTION: There is growing evidence that development of visceral
obesity is conditioned by neuroendocrine disturbances connected with dietary
disorders. In this study we investigate the relationship between dietary
peculiarities, abdominal fat distribution and hormone regulation in 155 men
from 40 to 60 years.
METHODS: Visceral adipose tissue square (VAT) as determined by CT at L
4 5
level, daily food consumption, revealed by original food questionnaire,
cortisol, testosteron levels, fasting and glucose stimulated immunoreactive
insulin levels, C-peptide and growth hormone levels before and after oral
fat load (130 g fat on 2m
2
body surface) were measured in group of 42 lean
men (ML) without visceral fat abundance (BMI 22.780.24 kg=m
2
; VAT
80.540.24cm
2
), 33 lean men (MLV) with visceral fat abundance (BMI
23.780.17kg=m
2
; VAT 125.3013.18cm
2
) and 83 men (MOV) with total
and visceral obesity (BMI 29.530.29kg=m
2
; VAT 261.6810.60cm
2
).
RESULTS: The food glycemic index was signicantly higher (p<0.001) in
MOV in comparison with MLV and ML. Moreover, in MLV the glycemic index
was higher (p<0.05) than in ML. Daily bread and sugar consumption
prevailed (p<0.01, p<0.001) in MOV in comparison with ML and MLV.
Daily saturated fat and dietary bers consumption didn't differ between
groups. However, the dietary bers index was lower (p<0.001) in MOV
than in MLV. Hormone disturbances in MLV and MOV were characterized by
basal and glucose stimulated hyperinsulinemia; absent of signicantly changes
of C-peptide and growth hormone levels after oral fat load; basal hypercorti-
solemia and lower basal testosteron levels in comparison with ML. In MLV
and MOV C-peptide levels didn't change after fat oral load. There was no
difference between groups in basal C-peptide and growth hormone levels.
Glycemic index positively correlated with VAT (r = 0.31, p<0.05), choles-
terol (r = 0.31, p<0.05), triglycerides (r = 0.38, p<0.05), immunore-
active insulin levels at 1 and 2 hours after standard oral glucose load
(r = 0.41, r = 0.37, p<0.05), testosterone (r = 0.51, p<0.05).
CONCLUSION: High dietary glycemic index associates with hormone and
metabolic disorders accompanied with visceral fat abundance in men inde-
pendently of obesity presence.
P66
Do cariogenic microorganisms predict BMI better than reported intake of
sweets?
B Barkeling
1
, Y Linne
1
, AK Lindroos
2
, D Brikhed
3
, and S Rossner
1
1
Obesity Unit, Huddinge University Hospital, Stockholm Sweden
E-mail: Britta.Barkeling@medhs.ki.se;
2
Institute of Internal Medicine, Goteborg
University, Sweden;
3
Institute of Odontoloty, Goteborg University, Sweden
INTRODUCTION: As a part of the SPAWN (Stockholm Pregnancy and
Women's Nutrition), this study was performed in order to further nd out
which nutritional components are associated with BMI (Body Mass Index),
using subjective as well as more objective measurements. Thus, the aim was to
correlate the BMI to questionnaire data of sweet intake and to salivary counts
of cariogenic microorganisms.
METHODS: 351 women, with a median BMI of 24.2kg=m
2
(range 17.5
47.8) and aged 45 years (range 34 64), quantied in detail their intake of
sweets during the past two months. The dietary history was taken from the
validated SOS (Swedish Obese Subjects) dietary questionnaire. The sweets
included chocolate, candy, cake, pastry, cookies and ice cream. As an objective
test of intake of sugar-containing foods, salivary counts of mutans streptococci
were determined.
RESULTS: The reported intake of sweets did not correlate to BMI or to number
of mutans streptococci (both r
s
=0.01). However, the mutans streptococcus
counts did correlate to BMI (r
s
=0.14, p<0.05).
International Journal of Obesity (2001) 25, Suppl 2, S50S63
2001 Nature Publishing Group All rights reserved 03070565/01 $15.00
www.nature.com/ijo
CONCLUSION: This study indicates that salivary counts of cariogenic micro-
organisms, but not reported intake of sweets, are correlated to BMI. These
results could be due to an unsystematic underreporting of sweets. The
possibility that number of mutans streptococci in saliva could be used as a
future objective marker of a food component, often underreported, warrants
further exploration.
P67
How energy density and diet composition varies across age and sex in 102
UK adults
S Whybrow
1
, S Ferris
1
, and RJ Stubbs
1
1
Rowett Research Institute, Bucksburn, Aberdeen AB21 9SB, UK
We examined how energy density (ED) varied across age and sex groups in 51
men and 51 women, recording 7d weighed intakes. RMR was measured, and
body weight recorded on d 1 and 8, to estimate misreporting. One-way
ANOVA was used to examine differences between groups. Unsurprisingly
energy intake was greater in men. Interestingly, women ate a diet signicantly
lower in ED and %fat. ED declined almost linearly with age. In contrast to the
sex effect, ED decreased as age increased because %protein increased. High
protein foods and lower fat foods tend to be lower in ED.
E MJ ED Wt Kg %Fat %Prot %CHO
Males 9.97 4.13 2.61 31.7 16.5 46.6
Females 7.28 3.63 2.15 28.8 16.5 51.6
SED 0.34 0.22 0.12 1.04 0.58 1.19
F(1,100) 62.14 5.34 14.57 8.59 0.02 17.66
P 0.001 0.023 0.004 0.004 0.897 0.001
Age <30 9.53 4.90 2.10 32.7 14.9 50.1
Age 30 39 8.87 4.01 2.32 30.6 16.1 48.9
Age 40 49 8.47 3.67 2.48 30.0 16.9 48.7
Age >50 7.93 3.25 2.51 28.5 17.5 49.0
SED 0.61 0.28 0.18 1.50 0.79 1.85
F(3,100) 2.23 11.47 1.98 2.03 4.33 0.20
P 0.090 0.001 0.122 0.114 0.007 0.895
These data suggest that ED decreases across age and sex groups, but that the
strategies adopted to attain this outcome differed in these population sub-
groups. Greater attention should be given to the determinants of food and
nutrient selection in different population subgroups.
P68
Taste perception of free fatty acids, i.e. CLA
MMJW Kamphuis
1
, and MS Westerterp-Plantenga
1
1
Department of Human Biology, University of Maastricht PO-box 616,
6200 MD Maastricht, The Netherlands
In rats, free linoleic acid (LA), but not oleic acid has been shown to stimulate
taste receptor cells. Moreover, an inverse relationship between fatty acid taste
perception and fat preferences were shown (Gilbertson et al., 1998)
1
. We
investigated possible LA perception in humans and its relation to satiety and
energy intake.
Subjects were identied as fatty acid tasters (FAT, n=14) when they
distinguished _9 samples and as FA-non-tasters (FANT, n=6) with <9
samples when testing 10 samples of 10 mM linoleic acid against its solvent.
A low concentration free LA as conjugated linoleic acid, oleic acid and no
supplementation was added to a low energy ice cream and a high energy ice
cream. One out of six ice creams was offered ad libitum each week; low energy
without CLA (LE), with CLA (LEC) or with oleic acid (LEO); high energy without
CLA (HE), with CLA (HEC) or with oleic acid (HEO). Food intake, hedonics and
appetite were measured.
FAT and FANT did not differ in bodyweight, BMI and age. No differences in
hedonics, food intake and feelings of hunger and satiety were seen between
FAT and FANT. In the FAT group, but not in the FANT group, there was a
relationship between amount eaten (g) and nsatiety for LEC (r
2
=0.6,
p<0.001). Sensory specic satiety, expressed as npleasantness of taste for
LEC and HEC was higher for FANT than for FAT (p<0.05).
In conclusion, fatty acid tasters could be distinguished by a taste test of
10 mM linoleic acid. CLA induced satiety in fatty acid tasters when it was added
to a low energy food, but did not result in a difference in food intake. CLA
affected sensory specic satiety in fatty acid non-tasters.
1
Gilbertson, Ann N Y Acad Sci, 1998; 855:165
P69
Dietary restraint during weight maintenance with or without exercise
training, in men
MPGM Lejeune
1
, DPC Van Aggel-Leijssen
1
, and MA Van Baak
1
1
Westerterp-Plantenga MS. Dept. of Human Biology, University of Maastricht,
PO-box 616, 6200 MD Maastricht, The Netherlands
Diet plus exercise training programs have been compared with diet programs
with respect to weight maintenance (WM) after an energy restriction (ER)
period. We investigated the effect of dietary restraint during WM after an ER
period, with or without an exercise training program. Subjects were 29 obese,
healthy men (BMI 32.10.4kg=m
2
age 391.4yrs).
Dietary restraint (F1) and general hunger (F3) were determined, using the
Three Factor Eating Questionnaire (TFEQ) at the start and end of the study and
after the ER period. Subjects were divided in a diet (D) or a diet exercise (DE)
group. Both groups participated in a 10-wk ER program, while subjects in the
DE group also participated in an exercise training program. This was followed
by a 40-wk WM phase in which subjects in the DE group continued the
exercise training, while subjects in the D group did not change their activity
pattern. Body weight (BW) and maximal oxygen uptake (VO
2
max) were
measured at the same timepoints as the TFEQ.
BW loss during the ER phase was 15.11.0 kg. During the WM phase BW
regain was 8.00.8kg. The F1 score increased from 50.7 to 100.9
during the ER phase, while the F3 score decreased from 50.6 to 30.5.
BWregain during the WM phase was negatively correlated with the increase in
F1 during the ER phase (r =0.47, p<0.05). The exercise training program did
not induce a difference between groups in changes in BW, VO
2
max and TFEQ.
We conclude that increases in F1 had a clear effect on WM irrespective of
possible exercise effects.
P70
Body weight development 6 years after a dietary intervention with low-
fat products
B Seppelt
1
, and HJF Zunft
1
1
German Institute of Human Nutrition, Aurthur-Scheunert-Allee 114 116,
D-14558 Bergholz-Rehbruecke, Germany
INTRODUCTION: In a 3-months intervention study on the long-term effect of
substituting normal-fat (control group) with fat reduced products (interven-
tion group) a reduction of fat intake by 22 g=d was found. This reduction was
not accompanied by a compensative increase in the intake of other nutrients.
Consequently, in the intervention group a signicant weight loss by 1.5kg was
observed [1]. We were interested whether this effect maintained over a long
period of 6 years without any further intervention.
METHODS: 47 women from the previous study (70 percent of the entire
population) were recruited for the follow-up. Their food intake was measured
with a 3-day weighed food record and the body weight according to the
protocol of the main study.
RESULTS: 6 years after the intervention the food intake was similar to that at
the end of the study. The fat intake was 34.9 energy percent at follow-up vs.
35.7 in the past, the energy intake 1782 kcal=d vs. 1804 kcal=d. However, the
body weight has increased signicantly. The weight gain was 4.2kg on the
average of both groups, among them were six cases with a more than 10 kg
increase. The maximum reached 16.3kg.
CONCLUSION: The results of the 6-year follow-up indicate that a single
intervention without subsequent treatment does not contribute to a long-
term maintenance of the body weight.
REFERENCE: [1] Seppelt B, Weststrate JA, Reinert A, Johnsen D, Lueder W,
Zunft HJF (1996). Long-term effect of fat-reduced foods on energy intake and
body weight. Z Ernaehrungswiss 35:369 377
P71
Neurosteroid plasma levels in obese
R Menozzi
1
, M Bondi
1
, F Petraglia
1
, C Carani
1
, A Velardo
1
, and G Del Rio
1
1
Chair of Endocrinology, Univerisity of Modena, Modena, Italy
Neurosteroids include all steroids synthetized in the brain where they can
modulate central nervous system activity. They have been found also in
peripheral circulation and it has been hypotized that they are able to modulate
reproductive function, development and aging processes. In rats a neuroactive
steroidal modulation of feeding behavior and energy balance has been shown;
allopregnanolone exerted a dose dependent hyperphagic effect while DHEA-S
elicited an anoretic effect. Since no data are available about these compounds
in obesity we found of interest to evaluate the circulating levels of allopregn-
anolone (AP) and DHEA in a group of 39 obese subjects (BMI =36.31.6
W=H=0.950.02, age =41.42.5 ys) and in 57 normal-weight subjects
Track 3 Eating patterns and behaviour
S51
International Journal of Obesity
(BMI =24.11.7; age =35.81.6ys) both at baseline and in response to
functional test (colour word mental stress test -CWT- and CRF infusion,
100mcg iv). Basal AP levels were higher (p<0.01) in obese (1.36
0.23 nmol=L) than in normal subjects (1.060.3 nmol=L); basal DHEA levels
were not signicantly different (p=NS) in obese (4.90.6ng=ml) and in
normal subjects (6.060.7 ng=ml) with the trend to be higher in normal. An
increase in AP (F =3.7, p<0.001) and DHEA (F =6.2 ; p<0.001) plasma
levels was observed in obese subjects after CRF administration beeing the AP
and DHEA peak levels reached at 15' (AP: 2.630.5nmol=L, DHEA:
11.2.00.4 ng=ml); in normal subjects CRF increased signicantly both AP
(F =11.9; p<0.001) and DI IEA levels (F =18, p<0.001) and the AP and
DHEA peak levels were at 60' after CRF infusion (AP: 1.50.3 nmol=L DHEA:
13.4.00.4 ng=ml ). The AP incremental area (i.a.) was signicantly higher in
obese (64.835010.5mmol=L=min) than in controls (33.3267
5.05 nmol=L=min) (p<0.02) while there was no differences in DHEA i.a.
between obese and controls (p=NS). CWT did not change signicantly
(p=NS vs basal) AP and DHEA plasma levels both in obese (AP peak:
1.50.2 nmol=L; DHEA peak: 4.90.5 ng=ml) and in normal (AP peak:
1.20.2 nmol=L; DHEA peak: 5.70.9ng=ml). Our work show higher basal
AP levels and an earlier and higher response of allopregnanolone and DHEA
to CRF infusion in obese subjects respect to normal. This could be related
to the hyper-responsiveness of the hypothalamic-pituitary-adrenal axis to
stimulatory neuropeptides and to stress previously observed in visceral obesity.
Mechanisms responsible for these abnormalities need further elucidations.
A role of neurosteroids in the modulation of feeding behavior it has been
demonstrated in animals the higher basal levels and the hyperresponse of
allopregnanolone in obese may represent one of the mechanisms underlyng
obesity throughout an increase in the food intake.
P72
Personalities and alimentary behaviors in obese patients
A Golay
1
, M Fossati
1
, A-F Allaz
1
, and N de Tonnac
1
1
Division of Therapeutical Education for Chronic Diseases. Department of
Internal Medicine. University Hospital Geneva. 24, rue Micheli-du-Crest, 1211
Geneva 14, Switzerland
INTRODUCTION: The actual tendency in the care of obese patients is the
association of dietetic information with an eating behavior therapy. Studies
attempting to attribute the origin of obesity to psychiatric pathologies are
contradictory. We studied whether certain eating disorders are more specic
to a personality type.
METHODS: We studied eating disorders with the Eating Disorder Inventory
(EDI) test in 281 obese women compared to 252 age- matched non-obese
women. Both obese patients and non-obese volunteers were divided into four
groups depending upon their personality (PERSONA test). This test denes
four types of personality, based on the level of emotion (expansive or reserved)
and the degree of power (dominant or consenting).
RESULTS: According to our study, eating disorders vary between the four
personality groups and were signicantly higher in the facilitating group
(p<0.0l) (consenting and expansive) compared to the three other obese
groups. Neither promoting (expansive and dominant) nor controlling obese
patients (dominant and reserved) present eating disorders. The analyzing
obese patients (reserved and consenting) are reticent when it comes to
consulting (18%) since they distrust others. Analyzing obese patients present
an interpersonal distrust (p<0.05) and an interoceptive awareness (p<0.05).
The group which presents most eating disorders is that of facilitating obese
patients (consenting and expansive). These present eating disorders of the
compulsive types favored by interoceptive awareness (p<0.00l), body dis-
satisfaction (p<0.001), ineffectiveness (p<0.001), and maturity fears
(p<0.05).
CONCLUSION: The diversity, even the absence of eating disorders brought to
evidence by our tests based upon different personality types should allow
better understanding the psychological and behavioral causes of weight gain
and the means for improving compliance in the following of an obese patient.
P73
Detection of facial changes during weight loss: a pilot study
C Kostara
1
, JJ Lara
1
, MEJ Lean
1
, P Vanezis
2
, and M Vanezis
1
1
Department of Human Nutrition and;
2
Department of Forensic Medicine and
Science, University of Glasgow. Glasgow Royal Inrmary, Glasgow G31 2NR,
Scotland
INTRODUCTION: Whether any signicant body change, early during a
weight loss programme, could be detected is uncertain. A positive answer
to this question could be of help in motivating patients to stick longer in a
weight loss programme by improving self-perception. The present study was
carried out as a pilot study in assessing facial changes during weight loss.
METHODS: A novel software technology for facial reconstruction (Vanezis et
al 2000) was used to measure facial changes in ve female participants of an
8-week weight loss programme. Linear measurements of mandible width,
face height, lower third face depth, face width and chin height, were taken at
weeks four and eight. These were expressed as ratios.
RESULTS: Mean weight loss during the 8-weeks was 1.5kg. No signicant
changes were found in BMI or waist circumference during the period of study,
however facial changes were signicant and correlated with both weight and
waist changes. Perceived changes were related to an improved satisfaction with
theprogrammeandattractiveness. Results (meanandSD) areshowninthetable.
Ratio Baseline Week 4 Week 8
Mandible width=face height 1.1410.16 1.1220.17 1.1020.15
Mandible width=lower
third face depth
1.1420.12 *1.1200.14 *1.0860.10
Face height=face width 0.8010.06 0.8130.06 0.8180.06
Chin height=face height 0.3150.03 0.3140.03 0.3160.03
BMI (kg=m
2
) 33.805.40 33.625.46 33.325.85
*P <0.05
CONCLUSIONS: Changes in width of the face, probably mainly reecting loss
of fat over the mandibles, can be quantied at an early stage during weight
loss with this novel technology, which might be useful in evaluating satisfac-
tion, attractiveness and perhaps self-esteem in women losing weight.
REFERENCES: Vanezis P, Vanezis M, McCombe G & Niblett T (2000) Facial
reconstruction using 3-D computer graphics. Forensic Science International
108, 81 95.
JJ Lara was supported by CONACyT in Mexico.
P74
Comparison of vitamins and minerals intake among overweight and non-
overweight schoolchildren
AM Rocandio
1
, L Ansotegui
1
, and M Arroyo
1
1
Nutrition and Food Science. University of Basque Country, Vitoria, Spain
INTRODUCTION: Obesity currently affects 25% of children and the preva-
lence appears to be increasing dramatically.
The objective was to determine differences in vitamins and minerals intake
between overweight and non-overweight schoolchildren and compare with
the recommendations for Spanish population.
METHODS: From a sample of 685 boys and girls (11 years old) 32 were
selected at random to be studied. The schoolchildren with weight-for-height
>90 th percentile was considered overweight. Data were taken during a week
by food weighing method (electronic scale, Soehnle
1
). Family collaborators
were properly trained for this purpose. Data were valorised with Novartis
Nutrition Dietsource (v.1.0. 1997). Statistical analyses were conducted using
the SPSS software.
RESULTS: The mean daily intakes for micronutrients overweight vs non-
overweight children were the next: Vitamins: B1(mg) 1.60.8 vs 1.30.4;
B2(mg) 1.80.6 vs 1.80.5; B6(mg) 1.80.8 vs 1.72.3; B12(mg) 3.5
1.1 vs 6.15.8; Niacin(mg) 30.335.8 vs 22.04.7; Folic acid(mg) 207.2
134.0 vs 249.1309.6; Pantotenic acid(mg) 1.30.4 vs 1.60.4; Biotin(mg)
5.73.8 vs 5.62.3; C (mg) 92.948.4 vs 95.153.7; A(mg) 680.9308.8
vs 901.8602.9; D(mg)2.81.4 vs 3.02.2; E(mg) 10.35.4 vs 10.14.5.
Minerals: Na(mg) 1468.7503.2 vs 1546.4562.8; K(mg) 2217.2571.4
vs 2025.8394.8; Ca(mg) 878.0238.5 vs 827.8331.6; P(mg) 122.7
250.2 vs 1148.1287.5; Fe(mg) 12.82.6 vs 12.12.1; Mg(mg) 276.6
72.1 vs 239.168.8; Mn(mg) 2.11.3 vs 1.80.8; Zn(mg) 8.63.8 vs
10.84.3; I(mg) 132.863.5 vs 128.340.0.
CONCLUSIONS: No signicant overweight and non-overweight differences
were found. Vitamins A and D and minerals calcium, magnesium and zinc
intake was lower than recommendations. The results justify the need for more
nutrition education and better strategies to improve the diets of school-
children.
Track 3 Eating patterns and behaviour
S52
International Journal of Obesity
P75
Food intake patterns associated with overweight in fourth grade children
in Germany
AD Liese
1
, T Hirsch
1
, E von Mutius
1
, and SK Weiland
1
1
Department of Epidemiology and Biostatistics, University of South Carolina,
Columbia, SC 29208, USA
Given the increasing prevalence of childhood overweight, we explored food
intake patterns associated with overweight in two cross-sectional studies of
fourth graders.
All children were participants of the International Study of Asthma and
Allergies in Childhood conducted in 1995=96 in Munich and Dresden,
Germany. Parents completed a questionnaire including a brief, non-quantita-
tive food frequency section on 14 food groups or foods. Height and weight
were measured in a random subsample of children undergoing spirometry.
1738 children age 9 10 years were included with complete data, 846 in
Dresden and 892 in Munich. Overweight was dened as body mass index
greater than or equal to the 90th age- and sex-specic percentile of the
German reference. In Dresden, 95 children were classied as overweight,
in Munich a total of 143.
A higher frequency of cold cut=sausage intake was consistently associated
with overweight in both cities: overweight children were more than twice as
likely to have been consuming 7 or more servings of cold cuts=sausages per
week than their normal weight peers (Dresden: OR 2.12, 95%CI 0.99 4.11;
Munich: OR 2.91, 95%CI 1.34 6.32) adjusting for age and sex in a logistic
regression model. In turn, a pattern for healthful food intake including
vegetables (carrots), fruit juice, vitamin supplements and particularly in
Dresden grains=musli was inversely associated with prevalent overweight.
In summary, the dominant food intake pattern associated positively with
childhood overweight in both cities was characterized by frequent intake of
cold cuts=sausages which represent high calorie and high fat foods. Despite
the limitations of the diet assessment instrument available, our results point
toward plausible and potentially population-specic food intake patterns
associated with prevalent overweight in pre-adolescent children in Germany.
P76
Alteration of metabolic processes and neuropeptide pathways in obese
and non-obese hyperandrogenic women
B Bidzin ska
1
, M Demissie
1
, U Tworowska
1
, and A Milewicz
1
1
Department of Endocrinology and Diabetology, Medical University of Wroclaw,
Pasteur Street No 4, 50-367 Wroclaw, Poland
INTRODUCTION: It is well known that extremes of body weight are asso-
ciated with abnormalities in sex hormones and hypothalamic function. The
role of leptin, galanin, NPY in hyperandrogenic women has not yet been fully
elucidated. Leptin is a potent appetite reducing factor and a possible periph-
eral signal of fat stores. It is very probable that leptin production may be
abnormally regulated in changed ovarian status. The cause of obesity in FOH
remains enigmatic. Insulin resistant PCOS (FOH) patients may have leptin
levels higher than controls. It is also probable that insulin regulates
leptin secretion and adipocytes with greater insulin sensitivity secrete more
leptin. Some FOH women may be resistant with respect to glucose metabo-
lism while other insulin responses such as androgen and leptin production
may be relatively normal. Leptin receptors were identied among others in
hypothalamus, pituitary and gonads. Leptin acting on granulosa cells may
develop environment in the follicle similar to that observed in PCOS. The aim
of our study was to develop the relationship between varied androgenic status
and satiety and appetite stimulating peptides. We investigated patients with
functional ovarian (N=47) and adrenal (N=25) hyperandrogenism, idio-
pathic hirsutism (N=36) and control group (N=42).
METHODS: serum insulin, leptin, NPY and galanin were measured in baseline
condition using commercially available RIA kits.
RESULTS: AUC for insulin observed during OGTT in the group with FAH was
the highest one and signicantly higher than AUC observed in IH group
(p<0.0l) and CG (p<0.0001). Women with functional ovarian (FOH) and
adrenal (FAH) hyperandrogenism had signicantly elevated plasma leptin
levels in comparison to control group (in total groups p<0.05 and p<0.05
respectively) and to women with idiopathic hirsutism (IH)(p<0.05 and
p<0.005 respectively). Similarly, plasma NPY levels were strongly increased
in FOH group in comparison to IH (p<0.05), FAH (p<0.005) and CG
(p<0.005). On the contrary, galanin levels were signicantly decreased in
FOH women in comparison to FAH (p<0.005), IH (p<0.0005) and control
group (p<0.00001).
CONCLUSIONS: It is probable that obesity in women with functional hyper-
androgenism (FOH=FAH) may be due to some shifts in neuropeptides levels
and activity.
P77
Cognitive restraint and misreporting of energy intake in 102 UK adults
RJ Stubbs
1
, S. Whybrow
1
, and S Ferris
1
1
ACERO, Rowett Research Institute, Bucksburn, Aberdeen AB21 9SB, UK
Regression analysis was used to examine the relationship between dietary
restraint and misreporting (MR) in 51 men and 51 women who collected 7 d
weighed dietary intakes. BMR was measured and body weight was recorded
on days 1 and 8 to estimate MR. MR was dened by a decreasing energy
intake (EI)=BMR ratio in subjects who did not lose weight.
The degree of restraint of a subject was negatively related to the ratio of EI
to BMR, although the percentage variance accounted for was not large. More
restrained eaters had a lower EI=RMR and so consequently were classied as
MRs more often. (This was veried by a chi-squared test (w
2
=3.92; P =
0.048)).
EI=RMR = 1:85 0:065
+
Restraint (R
2
= 2:8%)(P < 0:001)
Setting the cut off at 1.2 X BMR revealed that restrained eaters had a lower
reported % E from fat (28% vs. 31% and a lower dietary energy density (ED)
(3.1 vs 4.1MJ=kg) than unrestrained eaters (ANOVA, p<0.006).
The dietary determinants of ED, EI and food intake were generally similar
for restrained and unrestrained eaters. However as the EI=BMR ratio fell the
relationships between predictor variables (diet composition) and the out-
comes (ED, some aspects of El and food intake), became increasingly different.
These data suggest (i) MR increases with increasing cognitive restraint.
(ii) MR is not unspecic w.r.t. dietary ED and macronutrients. (iii) Conse-
quently as MR increases perceived relationships between diet composition ED,
EI and food intake increasingly alter. MR is likely to alter other perceived
relationships between diet and health.
P78
Intentional misreporting, eating restraint, self-esteem, and well being in
women
JJ Lara
1
, JA Scott
1
, and MEJ Lean
1
1
Department of Human Nutrition, University of Glasgow. Glasgow Royal
Inrmary, Glasgow G31 2ER, Scotland
INTRODUCTION: Although widely identied, food intake misreporting
remains a mysterious problem. Given the higher prevalence found in women,
the aim of this study was to explore the prevalence of intentional misreporting
of food intake and its associations with behavioural measures such as eating
restraint, self-esteem, well being, and socioeconomic status in women.
METHODS: Hundred and eighty four female volunteers aged 18 to 65y were
surveyed: 50 were obese patients (BMI _30) attending weight management
groups, 134 were nurses (BMI <25, N=52; BMI 25 29.9, N=45; BMI >30,
N=37). A self-report questionnaire containing three established psychologi-
cal tests (self-esteem, well being and the three eating-factor questionnaire for
eating restraint) (Stunkard & Messick, 1985) and new items to address the
likelihood of misreporting food intake in a diet study was designed.
RESULTS: The proportions of nurses who declared a likelihood to underreport
food intake were 28.8% for BMI range 20 25, 33.3% for BMI 25 30 and
52.6% for BMI >30, while among obese patients the proportion was 46%.
Among the same groups the gures for subjects who declared a likelihood to
overreport food intake were 38.5, 28.9, 10.5, and 32%.
Logistic regression analysis showed that an older age, higher social level,
bigger waist circumference and a higher frequency of previous attempts to
dieting during the past two years were factors characterising misreporters,
declaring intentional underreporting, in the regression model. Psychological
measures were not signicant (P >.10).
CONCLUSIONS: Intentional misreporting is common amongst overweight
and obese women, but this is not the only factor. In this sample social factors
and age shown to be more important than psychological measures in
predicting misreporting. The high prevalence of intentional overreporting is
a surprising new nding, particularly in the obese.
JJ Lara was supported by CONACyT in Mexico.
Track 3 Eating patterns and behaviour
S53
Intentional Journal of Obesity
P79
Acute changes in resting energy expenditure during mild hypocaloric diet
M Bondi
1
, R Menozzi
1
, I Magnoni
1
, M Gola
1
, A Velardo
1
, and G Del Rio
1
1
Chair of Endocrinology, University of Modena, Modena, Italy
It is well known that weight loss is associated with compensatory changes in
energy expenditure, which oppose the long-term maintenance of reduced
body weight. However, no study is available regarding the acute variations of
resting energy expenditure (REE) and body composition in obese patients
during food restriction. We thus monitored REE and body composition in 12
obese subjects (BMI =47.28.5; W=H=0.85). aged 18 59 years, before and
after four days on a mixed hypocaloric diet (600 Kcal=day less than measured
resting energy expenditure ( 72512 KJ); 55% carbohydrate, 20% protein and
25% lipid). The subjects were hospitalized and, after overnight fast, submitted
to resting energy expenditure evaluation by indirect calorimetry and to body
composition assessment by bioimpedentiometry. On the 4th day from hospi-
talization the body weight of patients was signicantly lower than the initial
weight (127.14.7 vs 129.34.7Kg; p<0.000). We observed a signicant
decrease in the REE (2029111 vs 2173108Kcal=24 h, p<0.000) and in
the fat free mass (FFM) (60.53.1 vs 62.13.4 Kg; p< 0.003) but no change
was observed in the fat mass (66.62.4 vs 67.12.3 kg, p=NS). Respiratory
quotient (RQ) did not vary throughout study period (0.7510.013 vs
0.770.014; p=NS). The ratio between FFM and REE was reduced from
36.01.3 to 34.01.3 Kcal per kilogram of FFM per day (p<0.000) with a
reduction of 2Kcal=kgFFM24 h. After four days on food restriction the total
body water (TBW) was reduced from 45.12.2 to 44.02.1 Lt (p<0.002)
while intracellular body water (ICW) did not change throughout study period
(25.31.3 vs 24.91.5Lt; p=NS) suggesting no changes in metabolically
active lean body tissues.
Our results show an acute reduction in REE of about 150Kcal=day during a
low-energy diet; the concomitant acute reduction in FFM of about 1.5 Kg
observed was mainly accounted for by water loss, with minimal, if any,
variations in metabolically active lean body mass. It has been demonstrated
that the reduction in REE after signicant weight loss (i.e. 10% of initial body
weight) exceeds that accounted for by changes in body composition by 15%.
Our data show that a similar degree of metabolic adaptation takes place in a
very early phase of food restriction, suggesting that factors other than body
composition changes may explain the nding. It is well known that food
restriction bluntes Sympathetic Nervous System activity, a main contributor to
the resting energy expenditure. In conclusion the time course of weight loss
deserves ftirther investigations with particular concern to both body composi-
tion and hormonal pattern changes.
P80
Mandatory snacks rich in sugar, starch or fat: effect on energy and
nutrient intake
N Mazlan
1
, G Horgan
2
, and RJ Stubbs
1
1
ACERO, Rowett Research Institute, Bucksburn, Aberdeen AB21 9SB, UK;
2
BioSS, Rowett Research Institute, Bucksburn, Aberdeen AB21 9SB, UK
We examined the effect of mandatory snacks high ( -80% by energy) in fat
(HF), sugar (HS) and starch (HSt) on energy intake (EI) and macronutrient
selection in 8 lean men and 8 lean women. Each subject was studied four
times in 7-d treatment periods, corresponding to control (zero snacks), or
3MJ=d mandatory snacks that were HS, HSt or HF. Treatment order was
randomised across subjects. The snacks were consumed during mid-morning
and mid-afternoon and all had the same energy density (500 KJ=l00g).
Throughout each study day, subjects had ad libitum access to 15 high-protein,
15 high-fat and 15 high-carbohydrate foods, rotated on a 3 d menu. Results
were examined by ANOVA.
Treatment sed P value
Control High
fat
High
Sugar
High
Starch
Inclusive Supplementation
Energy (MJ) 9.8 12.2 11.8 11.8 0.35 <0.001
Fat (MJ) 3.7 5.9 3.7 3.7 0.15 <0.001
CHO (MJ) 4.4 4.4 6.2 6.3 0.18 <0.001
Exclusive Supplementation
Energy (MJ) 9.8 9.1 8.7 8.8 0.35 <0.021
Fat (MJ) 3.7 3.4 3.4 3.4 0.15 <0.172
CHO (MJ) 4.4 4.1 3.8 3.9 0.18 <0.007
Incorporating mandatory snacks into the diet elevated EI, regardless of
composition. The CHO-rich snacks led to signicant, albeit weak compensa-
tion of EI ( -30%) C.F. control. The fat-rich snacks did not. Incorporating
mandatory snacks into the diet between meals elevated EI. Fat was 30% more
efcient than CHO at elevating EI. Nonetheless both HF and high-CHO snacks
markedly elevated EI.
P81
Does energy density drive energy intake because people eat to a constant
weight of food?
N Mazlan
1
, AM Johnstone
1
, G Horgan
2
, and RJ Stubbs
1
1
ACERO, Rowett Research Institute, Bucksburn, Aberdeen AB21 9SB, UK;
2
BioSS, Rowett Research Institute, Bucksburn, Aberdeen AB21 9SB, UK
This study assessed whether lean men eat a constant weight of food daily and
whether energy intake (EI) is a simple function of the energy density (ED) of
foods. 16 men [mean (SD) BMI 22.7 (2.02) kg=m
2
; Age 28.8 (8.6) years] were
each studied four times during a 2-d protocol.
On d 1, subjects were fed a xed diet at 1.6RMR. On d 2, subjects
consumed a mandatory breakfast (08.30) plus a snack (10.30) in one of four
treatments: (i) zero intake, (ii) low energy density (LED, control) [weight (g),
ED (kJ=100g), E (MJ)] 615: 400: 2.46; (iii) high energy density (HED)
1225:400:4.92; (iv) LED2, 615:800:4.91. For the remainder of the day,
subjects had ad lib. Access to 15 high-protein, 15 high-fat and 15 high-
carbohydrate foods. Hunger was tracked hourly using line scales. ANOVA
indicated that subjects were more hungry after the no and LED treatments in
mandatory peroid (F(3,44) =32.35 P <0.001). This persisted throughout the
day on the no breakfast treatment (F(3,44) =3.53 P =0.022). Lunch time Eis
were 5.0, 3.1, 4.2 and 3.2 MJ on the zero, HED, LED and LED2 treatments
(F(23,44) =8.89 P <0.001; sed=0.42). Total ad lib. EI was 11.7, 9.7, 10.2
and 9.6 MJ=d, (F(3,45) =3.17; P =0.033; sed=0.80). Total ad lib. plus
mandatory intakes were 11.7, 14.6, 12.7 and 14.5MJ=d, (F(3,45) =6.08;
P =0.001; sed=0.80). Corresponding food intakes were 2.2, 2.4, 2.5 and
3.1kg=d (F(3,45) =11.78; P <0.001; sed =0.15). EI was not a simple function
of overall ED, since food intake was not constant. Subjects showed a tendency
to compensate for difference in mandatory morning EI (c.f. control). % EI
compensation was stronger for decrements 50% than increments (20 30%)
in prior EI.
P82
No difference in appetite and energy expenditure after intake of different
types of fat
A Flint
1
, BH Jrgensen
1
, A Raben
1
, S Toubro
1
, and A Astrup
1
1
Res. Dept. of Human Nutrition, Center for Advanced Food Studies, RVAU,
Rolighedsvej 30, DK-1958 Frederiksberg C
INTRODUCTION: Dietary fat is an important source of energy. In general, fat
has less satiating power than protein and carbohydrates. However, there is not
a clear picture of the effect of different types of fat on the regulation of
appetite or on energy expenditure (EE). Thus, the aim of this study was to
investigate the short-term effect of intake of 3 different types of fat on appetite
and EE.
METHODS: Three different test meals were given randomly to 19 overweight
(BMI: 26.80.4kg=m
2
), young (25.20.7y) men. The fat-rich test meals
(0.8g fat per kg body weight, 60 E% F), served at 10am, varied only in the
source of fat (polyunsaturated=monounsaturated=trans fatty acids). EE was
measured continuously in a respiration chamber and appetite sensations were
rated by visual analogue scales (VAS) before and every 30 mm for 5 hours after
the meal. At 3pm an ad libitum meal was served, and energy intake (EI) was
registrered. A sensoric evaluation of all meals was given using VAS. Data were
analysed by two-way ANOVA.
RESULTS: There were no differences in basal or postprandial values of appetite
ratings and EE, in the subsequent ad libitum EI or in the sensoric evaluation of
the test meals between the three test days.
CONCLUSION: It can be concluded that on a short-term basis these different
types of fat do not impose any differences in appetite and EE in overweight
humans. However, it cannot be excluded whether other subject groups would
respond differently or whether on a long-term basis the type of dietary fat
makes a difference in the regulation of appetite and body weight.
Track 3 Eating patterns and behaviour
S54
International Journal of Obesity
P83
Short term satiety effects of soup in normal and obese subjects
Rothacker, DQ SDA-NJ, 99 Main Street, Matawan, NJ 07747 USA
INTRODUCTION: In a previous study, we looked for but did not nd BMI
effects on 4-hour satiety with liquid meal replacements (Obesity Res 2000.
8:69S). Using primarily the same subjects, we found that oatmeal of similar
nutritional content was signicantly more satiating (duration and intensity) for
subjects that were normal or underweight, compared with the morbidly
obese. These same subjects are presented here, following the same protocol,
but testing a mixture of solids and liquids (broth-based vegetable soup). All
three food types were of similar nutritional composition: 190 220 calories,
1.5 2g fat, l0 g protein, 4 5g ber, 41 44 g carbohydrates and 23 vitamins
and minerals.
METHODS: Satiety was measured over a 4-hour test period in 93 individuals
(74 F, 19 M; age 52 y = 714; BMI 31 kg=m
2
), of which 30 had BMI _
25 kg=m
2
and 42 had BMI _33 kg=m
2
. Subjects were instructed to consume a
325ml serving following an overnight fast, when they would normally eat
their rst meal of the day then assess their hunger on a 1 (not hungry at al) to
9 (as hungry as I have ever felt) scale before consumption, immediately after
and hourly for 4h. Baseline satiety scores of less than 3 were not used in the
analysis. All satiety tests were run on typical weekdays, separated by a 1 2
day washout period. The 2 separate satiety tests were averaged for each
subject. All tests are 2-tailed.
RESULTS: No baseline differences in satiety between the two BMI dataset
extremes were observed (p=0.867). The duration and intensity of satiety was
greater for the non-overweight subjects; however, only marginal signicance
was attained. No signicant differences were observed between the over-
weight dataset and the two BMI extremes.
Changes from Baseline
Subsets After 1h 2h 3h 4h
All Subjects 74.6 74.3 73.5 72.2 70.6
BMI _25 75.1 75.0 74.2 72.0 71.0
BMI _33 73.9 73.9 73.1 72.6 70.5
Between group difference 0.080 0.086 0.102 0.339 0.420
CONCLUSION: Satiety differences were marginally signicant between obese
and non-obese subjects following ingestion of soup. Obese subjects consis-
tently felt hungrier than normal and underweight subjects following the same
preloads of a broth-based meal replacement soup.
SlimFast Foods, WPB FL supported the study in its entirety.
P84
Short term satiety effects of oatmeal in normal and obese subjects
Rothacker, DQ SDA-NJ, 99 Main Street, Matawan, NJ 07747 USA
INTODUCTION: Satiety differences between obese and non-obese indivi-
duals, if they exist, would likely play a large role in the development of the
disease. This study tested satiety differences 2 4 times per person over a 4-
hour period, following an overnight fast in 93 individuals: 30 were normal
weight (BMI _25 kg=m
2
, and 42 had BMI _33 kg=m
2
.
METHODS: Subjects were prescreened to include only those that normally ate
breakfast and liked instant oatmeal. The 60g premeasured oatmeal packets
contained approximately 200 calories, 2g fat, 10 g protein, 4g ber, and 30g
carbohydrates (minor variations with avor). The oatmeal was reconstituted
with 125ml boiling water. Subjects were instructed consume the entire
serving following an overnight fast when they would normally consume
their rst meal of the day and to assess their hunger on a 1 (not hungry at
all) to 9 (as hungry as I have ever felt) scale before consumption, immediately
after and hourly for 4h. Baseline satiety scores of less than 3 were not used in
the analysis. All satiety tests were run on typical weekdays, separated by a 1 2
day washout period. The satiety score differences from baseline were averaged
for each subject.
RESULTS: Ninty-three (93) subjects completed, resulting in 294 tests of which
35 were considered protocol violations. No gender satiety differences were
found. No baseline differences in satiety by the 2 BMI dataset extremes were
observed (p=0.943). The duration and intensity of satiety was signicantly
greater for the non-overweight subjects.
Changes from Baseline
Subsets After 1h 2h 3h 4h
All Subjects 74.2 73.9 73.1 71.9 70.5
BMI _25 74.9 74.5 73.6 72.3 70.8
BMI _33 73.7 73.3 72.7 71.6 70.0
Between group difference 0.014 0.007 0.046 0.148 0.135
CONCLUSION: Satiety differences with instant oatmeal were signicant
between obese and non-obese subjects. Obese subjects were hungrier earlier
and the intensity of satiation was always less than those subjects with
BMI _25 kg=m
2
. This pattern was not apparent in an earlier study using
liquid meal replacements (Obestity Res 2000 8:69S).
SlimFast Foods, WPB FL supported the study in its entirety.
P85
The relationship of dietary and physical activity pattern with primary
obesity in Indonesian children
D Subardja
1
, SI Sjahid
1
, N Kania
1
, and PS Idjradinata
1
1
Department of Child Health, School of Medicine, Padjadjaran University-
Hasan Sadikin General Hospital Bandung, Indonesia
The purpose of this study is to evaluate the relationship of dietary and physical
activity pattern with the development of primary obesity in children. This
study was conducted to 122 school age children in Bandung municipality
from May to July 2000. Obesity status was determined by body mass index
(BMI) and weight for height (W-H) classications. Interview was done to assess
the dietary pattern i.e. habitual dietary pattern (HD) and 24-hour dietary
record (DR) and daily physical activity pattern i.e. physical activity score (PAS)
and energy expenditure (EE). According to BMI classication, there are 81
obese and 41 non-obese children. While by W-H classication the number of
obese and non-obese children is 47 and 75. The study shows that obese
children consume more calorie than non-obese, and signicant difference
found in calorie intake based on DR between non-obese and obese according
to BMI (p=0.026) and W-H (p=0.001). Macro-nutrient intake based on HD
and RD does not differ signicantly between obese and non-obese children
according to BMI and W-H (p>0.05). Based on HD and BMI degree of
obesity, fat is the most inuencing macronutrient (F =3.392; p=0.006) but
according to W-, calorie has more effects (F =2.557; p=0.042). While by 24-
hour DR in the BMI as well as W-H classication carbohydrate intake was the
most related macro-nutrient. No signicant difference was found in micro-
nutrient intake except for calcium intake by DR according to W-H classication
(p=0.043). Physical activity pattern reveals that obese children have lower
PAS than non-obese (p=0.000) and a signicant negative correlation with
BMI (r =0.615; p<0.001). However, delta calorie in obese was bigger than in
non-obese children and differ signicantly in the BMI classication by HD and
DR (p=0.001, p=0.0068). It is also found that PAS has a stronger effect than
calorie intake (p=0.000, p=0.026, p=0.001) in the development of child-
hood obesity and excessive energy intake and low physical activities have
strong correlation with the development of obesity. Eventhough the most
important factor is calorie intake, the degree of obesity is more related to fat
and carbohydrate intake.
P86
Prevalence differences of overweight and obesity in former East and
West-German fourth grade children
AD Liese
1
, T Hirsch
1
, E von Mutius
1
, and SK Weiland
1
1
Department of Epidemiology and Biostatistics, University of South Carolina,
Columbia, SC 29208, USA
We examined the prevalence of overweight and obesity and potential
correlates in two populations of fourth graders in Germany. All children
were participants of the International Study of Asthma and Allergies in Child-
hood conducted in Munich and Dresden, Germany, in 1995=96, and had
been raised under very different environmental attributes until the German
reunication. Parents completed questionnaires on current dietary habits,
infant nutrition, birth weight and socio-economic factors. Height and
weight were measured in a random subsample of children undergoing
spirometry. Data on 2474 children age 9 10 years were analysed. Over-
weight was dened as BMI > =90th and obesity as a BMI > =97th age- and
sex-specic percentile of the German reference.
In Dresden, 94% of girls and 11.7% of boys were overweight, versus
17.3% and 17.8%, respectively, in Munich (p-value of difference <0.0001
and <0.004). The difference in prevalence of obesity was similarly pro-
nounced: 1.7% of girls and 3.3% of boys in Dresden were obese versus
4.9% of girls and 5.2% of boys in Munich (p-value of difference <0.002 and
Track 3 Eating patterns and behaviour
S55
International Journal of Obesity
<0.10). Children in Dresden were not only lighter but roughly 1cm taller
than their peers in Munich. Subsequently, we explored the contribution of
various correlates of overweight (including current dietary intake and infant
feeding practices, socio-economic status and nationality) using multivariate
logistic regression to explain between-city differences in overweight preva-
lence. Infant feeding practices and socio-economic factors contributed in part
but could not explain the prevalence differences observed. Current dietary
habits in fact exacerbated between-city differences. This suggests that other
factors not assessed by this study such as physical activity are countering these
tendencies at the population level.
In conclusion, we observed marked differences in the prevalence of over-
weight and obesity in fourth grade school children in Dresden and Munich.
These populations share a common cultural and genetic heritage but experi-
enced a very different lifestyle environment in the rst 3 4 years of their life.
Further investigations are needed to unravel what specic components of
western lifestyle play a role in the increasing prevalence and obesity in children.
P87
Validation of body size pictorials and frame perception according obesity
degree in Mexican subjects
JC Lopez-Alvarenga
1
, LY Triana-Carmona
1
, VE Bolado-Garc a
1
, and
J Gonzalez-Barranco
1
1
Obesity Clinic. Instituto Nacional de Ciencias Medicas y Nutricion Salvador
Zubiran. Tlalpan. 14000, Mexico City
INTRODUCTION: Pictorials of body size have been used for European and
USA surveys. Severe obese patients have high prevalence of secondary
hypoxemia and hypercapnia, which can alter the perception. The aim of
this study was to demonstrate if perception of body size, using the scales of
the Swedish Obesity Study (SOS), are reliable and accurate methods for
Mexican obese subjects. Moreover, if obesity quartile shows different percep-
tion of their own body size.
METHOD: Two questionnaires were used: A) Patients were asked to ll a
questionnaire about relatives and their own body size perception using the
SOS pictorials. B) To asses if quartile of BMI had different perception of body
size of other people, we used projections of 19 (10 men and 9 women) slides
from patients who assist to the Obesity Clinic. These photographs were taken
with the same camera, light, background, distance and frontal position.
Subjects lled both questionnaires twice (test-retest) and the order of the
projected photographs was different. Weighted kappa was calculated as
concordance coefcient (CI 95%); value above 0.6, were considered clinically
relevant. Statistical signicance was considered if p<0.05.
RESULTS: There were included 24 males and 78 females, mean age was
50.814.4 years, BMI 35.66.8. Concordance for test-retest for their own
body size was k=0.91 (0.60 1.0), for their fathers k=0.87 (0.8 0.9),
mothers k=0.84 (0.7 1.0), brothers k =0.74 (0.33 1.0), children k =0.86
(0.42 1.0), spouse k=0.86 (0.3 1 1.0). Correlation between patients own
classication of obesity and BMI was r =0.72, p<0.01. No differences
between quartiles of BMI of patients (ANOVA p=0.l7), economic social
status (ANOVA p=0.93) or educational level (ANOVA p=0.31) were found
using accumulative scores of pictorials as dependent variable.
CONCLUSIONS: SOS gures showed high reliability to assess the relatives and
own body size. In spite that pictorial was developed in Europe, it can be used
for Mexican population. We did not nd any problems that severe obese
patients had different perception of slides.
P88
Socio-economic and behavioural determinants of dietary under-reporting
AE Perrin
2
, D Arveiler
1
, N Marecaux
1
, JB Ruidavets
1
, A Bingham
1
, P Amouyel
1
,
J Ferrieres
1
, P Ducimetiere
1
, JL Schlienger
2
, and C Simon
2
1
On behalf of the French MONICA Project;
2
Groupe d'Etude en Nutrition,
Hopital de Hautepierre, 67098 Strasbourg, France
To investigate the socio-economic and behavioural determinants of dietary
under-reporting, we examined the data from the third MONICA survey
conducted in the three French MONICA centres (Strasbourg, Toulouse, Lille)
in 1995 97.
The study population consisted of a representative sample of 974 men
aged 45 to 64. Dietary intake was assessed using a 3-day record method. The
basal metabolic rate was estimated using predictive equations based upon
body weight, age, and sex (according to Schoeld's equations). Subjects who
reported an energy intake below 1.05 times the estimated basal metabolic
rate were dened as under-reporters.
Results indicated under-reporting of energy intake by 15.7% of the
participants. No signicant relationship was observed between under-report-
ing and age group, nor with factors describing socio-economic environment
of the subjects (geographical area, educational and income tax level, eco-
nomic activity, occupation, family size). Underreporting was more frequent in
obese men (p<0.001) and among subjects who met dietary goals (p<0.02).
The proportion of overweight individuals (BMI >30kg=m
2
) who underre-
ported energy intake was 30.3%, compared to a proportion of 6.3%
among normalweight subjects (BMI <25kg=m
2
). Moreover, under-reporting
varied according to physical activity (p<0.001) and to self-reported alcohol
consumption (p<0.001) but not to smoking status. Under-reporters were
over-represented among sedentary people (25.0%) and among so-called non-
drinkers (24.1%). Additionally, the contribution of protein to energy intake
was signicantly higher in under-reporters, and was balanced by a lower
contribution of carbohydrates but not of fat.
Under-reporting in dietary surveys is a potential source of bias and should
be taken into account, particularly when weight-related diseases are the
subject under study. However, interestingly, in middle-aged men, it seems
not to be inuenced by socio-economic factors.
P89
Prevalence of depression in obese and non-obese women: relationship
with eating behaviour and physical activity
G Vansant
1
, M Hulens
1
, W Van der Borght
1
, R Lysens
1
, and E Muls
1
1
University Hospital Gasthuisberg, Catholic University, Leuven, Belgium
INTRODUCTION: The relationship between obesity and psychological health
remains unclear. Nevertheless, it is generally accepted that psychological
distress, if present, may inuence therapy outcome in these patients.
OBJECTIVE: The rst aim of this study was to quantify the prevalence of
depression in a sample of 473 normal weight (n=380; BMI: 22.8
2.8kg=m
2
) and obese women (n=93; BMI: 37.96.1 kg=m
2
). Secondly,
the relation between depression and age, body weight, eating behaviour
patterns and physical activity level was studied.
METHODS: Current levels of depression were measured using the Beck
Depression Inventory (BDI). Eating behaviour patterns were characterised
with the Dutch Eating Behaviour Questionnaire, while the level of physical
activity was quantied using the Baecke Questionnaire.
RESULTS: The prevalence of depression (BDI >12) was signicantly higher in
obese (56%) vs non-obese women (19%). Although age ranged from 16 70
years, no impact of age on the prevalence of depression could be observed
(r =0.05, p=0.297). The degree of depression showed no relationship with
both BMI or level of physical activity. With increasing severity of depression,
both emotional (p<0.0001) and external (p<0.05) eating behaviour pat-
terns were more pronounced.
CONCLUSION: The results demonstrate that depression is a frequent feature
in obese patients. No relationship could be observed with age, body weight or
physical activity indices. However, a clear link between depression and
emotional eating patterns was observed. Further research is needed to
determine the potential causality of the relationships between depression
and the other factors examined in this study.
P90
Dietary and phenotypic determinants of energy intake in 102 UK adults
S Ferris
1
, S Whybrow
1
, and RJ Stubbs
1
1
ACERO, Rowett Research Institute, Bucksburn, Aberdeen AB21 95B, UK
This study examined the extent to which energy intake (EI) was determined by
diet composition and inter-subject variation (phenotypes) using 7d weighed
intakes (food and drink) in 51 men and 51 women.
A multiple linear regression model of EI was obtained using stepwise
regression. Differences between subjects (inter-subject variation) accounted
for 43.4% of the total variation in EI. A further 39.4% of the variation was
explained by including dietary components.
EI(MJ) = 2:83 subject effect 0:10% water 0:66% ED 0:032
%CHO0:044% protein (R
2
= 82:8%)
Replacing the subject effect with the subgroups sex, age, BMI and restraint
gave the following model.
EI(MJ) = 10:09 2:57 sex (female) 0:041 age(30 39) 0:083age
(40 49) 0:84 (age > 50) 0:614 restraint(> 2:65) 0:053%water
0:22% ED 0:071% CHO0:11% protein(R
2
= 55:5%)
Sex, age and restraint do not explain all phenotypic effects since their
substitution for the global subject effect resulted in a reduction of the %
variance explained. These data suggest that the determinants of EI are
multifactorial and that phenotypic factors such as age, sex and psychological
concern about weight are as important as diet composition in determining the
EI of humans consuming their usual diets in their natural environment.
Track 3 Eating patterns and behaviour
S56
International Journal of Obesity
P91
Why lose weight? The reasons reported by overweight men prior to
commencing a weight loss programme
CR Hankey
1
, WS Leslie
1
, and MEJ Lean
1
1
University of Glasgow Department of Human Nutrition, Glasgow Royal
Inrmary, Glasgow, G31 2ER, Scotland
INTRODUCTION: The health benets of moderate weight loss are recognised
by health professionals and feature in the majority of clinical guidelines for
obesity management. However, what factors are important to overweight
men wishing to lose weight?
METHODS: The reasons for attempting weight loss were investigated as part
of a weight management programme carried out in a work-site setting.
Ninety-one male workers in a large oil renery, mean age 41 (SD 8), range
18 55 years, with a body mass index above 25 kg=m
2
(mean 37.0 SD 3.7)
were recruited. Prior to receiving dietary advice subjects were asked as an
open question ``why do you want to lose weight?'' Responses were docu-
mented, categorised by the terms used (health benets, improved tness,
improved well being, improved appearance, other) and ranked in order of
importance. Reasons for weight loss were described for the group as a whole,
according to BMI cut-offs (25.0 29.9, 30.0 39.9 and _40 kg=m
2
) and by
age bands (18 29, 30 39 and 40 55 years).
RESULTS: The message that weight loss is benecial to health was recognised
by all subjects and reported as the principal factor for attempting weight loss.
Physical benets, effects on appearance and well being were reported only
half as often as the primary reason for weight loss. Improved appearance was
considered the primary reason for weight loss by men aged 30 40 years and
ranked second in importance to improving tness in those aged 18 29 years.
These lifestyle factors were less important to the oldest age band who
considered health benets to be of primary importance. Improving health
remained the most important reason for all subjects regardless of BMI.
Subjects with a BMI of 26 30 or greater than 40 kg=m
2
, ranked improving
appearance second in importance to health, while subjects with BMI 30
40 kg=m
2
ranked tness second in importance. In the most overweight group
improving tness was not reported as a reason for weight loss.
CONCLUSION: Overall, obesity is not primarily viewed as a cosmetic issue,
although this is important, especially to those in the younger age group. As
age increases health benets assume greater importance. Improving physical
tness is not perceived as an important goal for the morbidly obese. Over-
weight members of the public appear to have accepted the message that
``weight loss can improve health''.
P92
Attitudes and perceptions of body size of rst year female university
students
M Slabber
1
, M Laubscher
1
, C Van den Heever
1
, and G Joubert
2
1
Department of Human Nutrition, (G24), Faculty of Health Sciences, University
of the Orange Free State, Box 339, Bloemfontein 9300, South Africa;
2
Department of Biostatistics, (G24), Faculty of Health Sciences, University of
the Orange Free State, Box 339, Bloemfontein 9300, South Africa
INTRODUCTION: Body dissatisfaction accompanied by excessive weight loss
behaviors is on the increase amongst adolescent females. Cultural and socio-
environmental factors however play a major role in dieting that may progress
to the development of eating disorders. The aim of this study was to
determine the attitudes of rst year female students, and to compare attitudes
of black (B) and white (W) culture groups to body size.
METHODS: A proportional, stratied sample of 150 rst year female hostel
students (15 25y) was randomly selected to take part in the study. Weights
and heights were measured and body mass index (BMI) calculated. Two series
of photographs of B and W culture groups were respectively standardized
according to ve calculated BMI categories. A structured standardized ques-
tionnaire based on these photographs was used as an attitude scale.
RESULTS: Results showed that although only 11,3%B and 4.l%W were
underweight (BMI of <18.5kg=m
2
), 66%B and 55%W considered under-
weight as ideal. Most subjects (62.3%B and 80,5%W) had normal body
weight (19 24 kg=m
2
) but only 30,2%B and 38%W considered this as
ideal. Most subjects (76.9%B and 77.2%W) considered underweight as
most attractive while none thought that overweight (BMI 24 29,9kg=m
2
)
or obesity (BMI >30 kg=m
2
) is attractive. Most subjects (96.2%B and 93,6W)
considered underweight and normal weight as most healthy.
CONCLUSION: The norm of ``big is beautiful'' might be phasing out amongst
African blacks while a serious problem of increased eating disorders may erupt
with westernization. Primary prevention programs of disordered eating may
need more attention amongst adolescent females in South Africa.
P93
Effect of wine or beer versus soft drink served at a meal on food intake
B Buemann
1
, S Toubro
1
, and A Astrup
1
1
Research Department of Human Nutrition, RVAU, Rolighedsvej 30, 1958
Frederiksberg, Denmark
Results from studies addressing the effect of alcohol on food intake are
ambiguous. If the alcohol is added as a supplement the compensation for
its energy seems to be incomplete. However, the effect of alcohol on energy
intake at a meal should also be tested by using dissolved nutrients with the
same energy density as reference. Ad libitum energy intakes (EI) of a supper
meal were measured with different beverages. Isovolumetric (9mL=kg body
mass) and approximately isocaloric amounts of lager beer, (Carlsberg Hof,
4,6 vol-% alc. 1,66kJ=mL) or a soft drink, (Sprite Regular, 1,73kJ=mL) were
given at two different occations. On a third occation an amount of alcohol
corresponding to that supplied with the beer was given as red wine. Tap water
was available ad libitum with the meals in all the experiments. The test meal
consisted of three different courses: cold pasta salad, goullach with bread and
a rm cake.
Total energy intake did not reach statistical signicance between the
beverages (wine vs beer vs soft drink, 5973343 vs 6181316 vs
5630269kJ, P =0.14). However, the intake of goullach was higher with
wine and beer compared to soft drink (wine vs beer vs soft drink, 1583121
vs 1624125 vs 130877 kJ, P =0.004).
The data imply that alcoholic beverages may stimulate energy intake at a
meal relatively to soft drinks when given in isocaloric quantities with the meal.
However, specic organoleptic beverage=food interactions may also have
played a role in determine the food intake.
P94
Metabolic effects of nicotinamide and diet in obese and diabetic women
from South Africa
M-T Van der Merwe
1
, NJ Crowther
2
, GP Schlaphoff
3
, IP Gray
2
, and
PN L

onnroth
4
1
Carbohydrate and Lipid Metabolism Research Group, University of
Witwatersrand Medical School, 7 York Road, Parktown 2193, Johannesburg,
South Africa;
2
Chemical Pathology, University of Witwatersrand Medical
School, 7 York Road, Parktown 2193, Johannesburg, South Africa;
3
Radiology,
University of Witwatersrand Medical School, 7 York Road, Parktown 2193,
Johannesburg, South Africa;
4
Department of Internal Medicine, University of
G

oteborg, Sweden
INTRODUCTION: 10 Obese (OW) and 10 obese diabetic women (ODW) were
treated for 4 weeks with nicotinamide and 4 weeks of dietary intervention to
investigate the different changes in metabolic indices and body composition.
METHODS: U & E; OGTT with glucose, C-peptide, S-insulin; lactate; FFA
levels. Bio-impedance, 5-level CT scan for bodycomposition.
RESULTS: No weight loss occurred during nicotinamide treatment. During
dietary restriction (l400cal=d) the ODW lost 7.9kg; OW lost 8kg. The weight
loss was associated with a reduction in waist and hip circumference, Kg and %
fat, kg and % lean mass, parasagittal diameter and visceral and SC fat mass.
During nicotinamide treatment there was a small but signicant reduction in
visceral fat mass. Decreases in fasting serum levels were documented in ODW
during nicotinamide treatment for: glucose (9.5 to 7.5mmol=L; p<0.01), C-
peptide (566 to 333pmol=L; p<0.05), insulin (124 to 67pmol=L; p<0.05),
FFA(1113 to 442 mmol=L, p<0.01). The 30 min and 2h OGTT levels showed a
similar trend. In both the ODW and OW the fasting lactate levels increased
(ODW 1397 to 1748 mmol=L; p<0.01 and OW 1545 to 1825mmol=L
p<0.05), but the % in lactate increase post OGTT was higher in ODW than
OW. During nicotinamide treatment no signicant improvement was docu-
mented in OW for glucose, C-peptide, insulin or FFA levels. All the metabolic
improvement in the ODW were maintained during the dietary phase of the
study, although the FFA showed a small rebound (442 to 726mmol=L;
p<0.05).
CONCLUSIONS: 1. ODW showed a dramatic improvement in metabolic
indices and a reduction in insulin resistance during nicotinamide treatment.
2. Nicotinamide treatment followed by dietary intervention leads to an
improved metabolic prole, reduction in visceral fat mass and better body
composition. 3. Nicotinamide treatment did not offer an additional advantage
over dietary intervention in OW.
Track 3 Eating patterns and behaviour
S57
International Journal of Obesity
P95
Age inuences misreport of caloric consumption in a group of morbid
obese women
E Chliamovitch
1
, E Bobbioni-Harsch
2
, T Lehmann
2
, M Volery
1
, C Muggler
1
, Ph
Morel
1
, O Huber
1
, G Chassot
1
, and A Golay
2
1
Department of Digestive Surgery, Geneva University Hospital, Geneva,
Switzerland;
2
Division of Therapeutic Education for Chronic Diseases, Geneva
University Hospital, Geneva, Switzerland
INTRODUCTION: The reliability of alimentary history is made questionable
because of misreport, in particular when dealing with obese patients. The aim
of this study was to investigate the factors that can inuence misreport in a
group of morbidly obese patients.
METHODS: Dietary history was drawn from a group of 57 morbidly obese
patients (age: 372y, bw 119.72kg, BMI 45.01kg=m
2
), before surgical
intervention of gastric bypass, in condition of stable body weight. Caloric
Intake (CI kcal=d) and percent macronutrient composition of the diet were
calculated by computerized program (Prodi 3.0 ); Resting Energy Expendi-
ture (REE, kcal=d) was measured by indirect calorimetry; Energy Requirements
(ER, kcal=d) were calculated as REE1.3 and misreport as ER-CI.
RESULTS: When evaluated in the overall group, misreport corresponded to
71.04% of the calculated ER; however a large individual variability was
observed. The degree of misreport (expressed in % of ER) showed a signicant
(p<0.01), negative relationship with % protein content of the diet, a
signicant, positive relationship with both % carbohydrates (p<0.01) and
% sucrose (p<0.02) and no signicant relationship with % lipid. Misreport
was signicantly (p<0.005), positively related to age whilst no signicant
relationships were observed with body weight or ER. When patients were
divided into two subgroups according to their age (i.e. over 35 y and under
35 y), but with a similar body weight (119.42.9kg and 119.73.4,
respectively) the degree of misreporting corresponded to 713.94.1%
and 12.75.9%, respectively in the older and in the younger group of
patients (p<0.001). The diet of the older group showed a signicantly
higher % protein content (15.70.2 vs 14.00.6, p<0.05) and a signi-
cantly lower % sucrose (6.30.7 vs 10.31.4, p<0.02).
CONCLUSION: In morbidly obese women, underreport mainly occurs in older
subjects, independently of the degree of obesity or energy requirements. It
has been suggested that high fat foods are mainly responsible for obesity
development and maintain; our results agree with this observation, since lipid
content of the diet exceeded 40%. However, underreport concerns mainly
carbohydrates and particularly sweet food. This leads to an articial over-
estimation of protein content of the diet.
P96
Factors inuencing the eating inventory in the Czech population
V Hainer
1
, M Kunes ova
1
, J Lajka
1
, R Mikulova
1
, M Wagenknecht
1
, and
J Par zkova
1,2
1
Obesity Management Centre, 3rd Department of Medicine, 1st Faculty of
Medicine Charles University;
2
Stem=Mark Agency, 128 08 Prague, Czech
Republic
AIM OF THE STUDY: To reveal the factors inuencing items of the Eating
Inventory (EI) in the Czech population. EI evaluates eating behaviour by
restraint (RS), disinhibition (DS) and hunger (HS) scores.
METHODS: A study employed a quota sample of the Czech population which
included 681 men (age: 40.914.7, weight: 83.312.4kg, BMI: 26.22
3.72 kg=m
2
) and 745 women (age: 42.315.1, weight: 67.812.6kg, BMI:
24.764.73kg=m
2
). The following factors which might inuence items of the
EI were evaluated: age, body mass index (BMI), body fat distribution (assessed
as the waist girth), educational level, average monthly per capita income, time
of the onset of obesity and parental obesity. Differences in the items distribu-
tion were assessed by the w
2
test.
RESULTS: Average values of the items of the EI in men were for RS: 4.54.1,
for DS: 5.33.1 and for HS 5.03.5, whereas in women were for RS:
8.25.0, for DS 5.03.5 and for HS 3.63.1. No signicant age-dependent
differences were observed when the upper and lower quartiles of the items of
the EI were compared except for DS which was signicantly lower (p<0.05)
in 45 59 y old men than in 18 29y old men. Obesity was signicantly
associated with higher DS and HS in both genders (p<0.001). A high DS was
more frequently observed (p=0.001) in the individuals with android fat
distribution (waist girth >102cm in males and >88 cm in females) whereas
a higher HS appeared more frequently (p<0.05) only in women with android
body fat distribution. However, no signicant differences in RS related to BMI
and body fat distribution were recorded. Educational level affected only HS in
men: men with secondary or university level of education exhibited lower HS
than those with elementary education (p<0.05). An average monthly per
capita income inuenced RS in women. Women with a higher income scored
more in RS than those with a lower income (p<0.01). Childhood onset
obesity was associated with a signicantly higher scoring in all three items of
the EI. A signicant shift towards higher values of all three items of the EI was
observed in women who reported parental obesity whereas in men signi-
cantly higher values were reported only for DS.
CONCLUSION: In both genders DS is signicantly associated with BMI, body
fat distribution and parental obesity whereas RS is not related to anthropo-
metric measures. HS is related to BMI in both genders. An inuence of parental
obesity, socioeconomic and educational factors on HS and RS is gender
specic. Childhood onset of obesity predisposes for higher scoring in all
three factors of the EI. Supported by the Czech Ministry of Health and by
the grant COST B17 40.
P97
Habitual meal frequency and energy intake regulation in time blinded
men
MS Westerterp-Plantenga
1
, EMR Kovacs
1
, and KJ Melanson
1
1
Dept of Human Biology, University of Maastricht, P0-box 616, 6200 MD
Maastricht, The Netherlands
Daily energy intake is regulated more accurately in nibblers than in gorgers
1
.
Thus, we assessed a possible relationship between habitual as well as intro-
duced meal frequency, blood glucose pattern, macronutrient and energy
intake (El) in men.
A time-blinded within-subject design comparing iso-energetic (1MJ) iso-
volumetric high fat and simple carbohydrate (CHO) preloads, was applied in
twenty healthy men (18 31 yrs; BMI: 22.81.9 kg=m
2
), to assess energy
intake regulation in spite of intervention. Introduced meal frequency, contin-
uous blood glucose levels and patterns, macronutrient and energy intake,
appetite ratings and taste perception from the two test-days were determined.
Habitual meal frequency was determined from controlled 3-day food intake
diaries.
The difference in 24 h El on the two test-days was inversely related to
habitual meal frequency (r
2
=0.56; p<0.001). Habitual and introduced meal
frequency was positively correlated with the number of transient and dynamic
blood glucose declines (r
2
=0.74; p<0.0001). Habitual, but not introduced
meal frequency was positively correlated with average blood glucose level at
baselines (r
2
=0.44; p<0.01), sweetness perception of the preload, percen-
tage energy from CHO and inversely correlated with percentage energy from
fat, hunger suppression during preload consumption, and El during the test-
days (all: 0.76 <r
2
<0.84; p<0.0001).
Habitual meal frequency is of greater signicance in energy intake
regulation in healthy young men than introduced meal frequency.
1. Westerterp-Plantenga MS et al., Appetite 1994; 22:173 182.
P98
Psychological characteristics and subgroups in obesity
K Elfhag
1
, S Rossner
1
, AM Carlsson
2
, and P Rooth
1
1
Obesity Unit, Huddinge University Hospital, SE-141 86 Stockholm,
Sweden;
2
Psychiatric Clinic, S:t Gorans Hospital, SE-112 81 Stockholm, Sweden
INTRODUCTION: Weight reduction is difcult to achieve and sustain. Beside
biological factors, psychological factors are also inuencing weight develop-
ment. We believe a psychological understanding of behaviour in obesity must
partly be sought at a level beyond the immediate and self-reported. Therefore
a projective technique is included in our programs, measuring an unconscious
level.
METHOD: 100 patients awaiting treatment (76 women, 24 men, 20 66
years, BMI 31 60 kg=m
2
) were tested with the Rorschach method (Compre-
hensive System) among other methods.
RESULTS: Rorschach-results suggest that: 1. 50% have a personality function-
ing prone to difculties with emotions, sometimes implying a tendency
towards depression. Such a functioning is shown by an index with a constella-
tion of variables measuring painful affect, various forms of uneasiness, and
defenses toward these feelings off. Cognitive effort, stress and defense of own
view of things is also common in this subgroup. 2.39% of the patients have
difculties in coping or managing complexities in life. These patients have
difculties getting in touch with and making use of psychological resources,
and make little use of affects. Impoverishment and avoiding defence is also
seen in this subgroup. The subgroup with emotional difculties is associated
with relatively higher education, body size experienced as psychological
meaningful, like giving protection or being part of identity, periodic variations
in food-intake, binges and living single. The group with coping difculties has
lower socio-economic level and an irregular or chaotic meal-pattern. Rorschach
results further suggest that patients with very high BMI ( >40) lack access to or
realization of mature affectional needs. Massive obesity (BMI >45) is in
addition associated with less anxiety related to the body. Younger patients
Track 3 Eating patterns and behaviour
S58
International Journal of Obesity
( <30 years) have higher level of anxiety. Among patients >50 years emo-
tional deprivation and feelings of loneliness are more common.
CONCLUSION: Results suggests at least 2 sub-groups with different difculties
affecting weight-regulation. Patients with emotional difculties, often having
higher education, display a more complex psychological pattern. Food might
be used to comfort or soothe affects. For patients with coping liabilities,
mostly from lower socio-economic level, nding strategies to regulate eating
and make changes in life-style might be hard. Also, other sources of enrich-
ment might be lacking.
P99
Reported macronutrient intake in children and adiposity
C Maffeis
1
, A Grezzani
1
, R Gaudino
1
, and L Tato
1
1
Department of Pediatrics, University of Verona, Polyclinic, 37134 Verona, Italy
The purpose of this study was to investigate the potential relationships
between macronutrient intake and energy intake reporting in a group of
30 prepubertal children with different levels of adiposity: 14 overweight
[FM=31.7 (8.6)%] and 16 nonoverweight [FM=16.9 (4.8)%].
Underreporting was calculated as the difference between energy require-
ments (total energy expenditure, measured b, the heart-rate monitoring
method, estimated energy cost of growth) and energy intake, measured
by the diet history method.
Fat mass (kg) was signicantly correlated with energy intake underreport-
ing (r = 70.64, p<0.001). Fat and carbohydrate intake (g=day) were
signicantly correlated with energy intake underreporting (r = 70.62,
p<0.001 and 70.57, p<0.01, respectively). A multiple regression analysis
showed hat fatness, adjusted for age and gender, accounted for ~47% of the
variation in the energy intake underreporting. Fat intake, included with age,
sex and fat mass among the independent variables, explained a further 25% of
interindividual variability of energy intake underreporting.
In conclusion, the inverse association between reported fat intake and
energy intake underreporting suggests caution in drawing conclusions from
data on diet composition of prepubertal children, especially if overweight.
Strategies to increase the awareness of fatty food intake may theoretically
improve the efcacy of prevention and treatment of overweight in children.
P100
Role of eating disorders in weight history and body composition of
women searching weight loss
V Giusti
1
, E Zysset
1
, E Hera ef
1
, RC Gaillard
1
, and P Burckhardt
1
1
University Hospital CHUV, Lausanne, Switzerland
INTRODUCTION: The aim of this study was to evaluate the anthropometric
measurements, eating patterns and weight history in patients attending an
out-patient obesity clinic during 1 year, and to analyse the potential correla-
tion with eating disorders.
METHODS: We have screened 138 consecutive patients, evaluated by the
same physician investigator. The men were only the 14.5% (20) of patients,
and 6.5% of women (8) had specied eating disorder: bulimia nervosa
purging (5) and non-purging type (3). The study was performed in the 110
women without specied eating disorders and searching weight loss. Median
age was 39 years (range 15 68), median body weight 93 kg (range 57 170),
median body mass index (BMI) 35 kg=m
2
(range 24 67.2) and waist circum-
ference 99 (range 73 135). We have distinguished two group of Eating
Disorders: Compulsive Overeating (CO) and Binge Eating Disorder (BED). We
investigated eating behaviour and eating disorders by clinical specic inter-
view and by the Bulimic Investigatory Test of Edimburgh (BITE).
RESULTS: Only 1% of patients had a desired weight loss lower of 10% of initial
weight, corresponding to medical guidelines. The prevalence rate of drop out
was 23% (25=110). The BMI (p<0.004) and waist circumference (p<0.009)
of binge patients were signicantly higher than in patients without eating
disorders. Besides prevalence rate of weight cycling syndrome (p<0.02),
previous utilisation of weight loss programs (p<0.008) and inuence of stress
(p<0.0001) were more important in patients with BED than ED free patients.
Furthermore a positive correlation has been found between weight cycling
syndrome and BMI (p<0.0001), waist (p<0.0001) and body fat
(p<0.0005). Finally the 72% (p<0.0001) of drop out were patients with
eating disorders.
CONCLUSIONS: This study evidences that patients with binge present a
higher obesity with abdominal repartition and a heavy weight history char-
acterised by failure of all treatments at long-term. Besides the recurrently
utilisation of weight loss programs increase the severity of obesity. This high
percentage of ED may explain the failure of conventional diets and the
tendency to weight cycling syndrome. A serious analysis of causes of dis-
crepancy between patients' expectations and professional guidelines concern-
ing weight loss is essential to improve long-term patient compliance and
decrease drop out.
P101
Diet composition and percent body fat in prepubertal children
PE Matz
1
, MS Faith
1
, AM Jorge
1
, and A Pietrobelli
1,2
1
Obesity Research Center, St. Luke's=Roosevelt Hospital Center, Columbia
University College of Physicians & Surgeons, New York, New York, (USA);
2
Pediatric Unit, University of Verona, Verona, (Italy)
INTRODUCTION: The prevalence of overweight and obesity in childhood is a
serious and increasing problem. Although childhood obesity is associated with
increased total energy intake, there is limited information on its association
with specic macronutrient intake. The aim of this study was to determine
whether total energy intake and dietary composition measured precisely
at controlled laboratory meals are related to % body fat (PBF) in 3 7 year
old children.
METHODS: Subjects were 60 prepubertal children, 3 7 years old, participat-
ing in a laboratory feeding study. Specically, the sample included 30 pairs of
twins for a study on the genetics of food intake. Total caloric intake plus
carbohydrate (CHO), fat (FAT), and protein (PRO) (analyzed both as intake in
grams and % total energy intake) were assessed during two ad libitum
laboratory test meals consisting of a broad variety of foods. The average
intake across both visits was analyzed herein. PBF was assessed using Bioim-
pedance Analysis (BIA). Race was assessed by parental report. Correlation
analyses tested the relationships among total energy intake, CHO intake, FAT
intake, PRO intake, and PBF. These relationships were further examined as a
function of gender and race.
RESULTS: PBF was signicantly correlated with total energy intake (r =0.39,
p<0.01), CHO in g (r =0.35, p=0.01), FAT in g (r =0.38, p<0.01), and PRO
in g (r =0.36, p<0.01). PBF was not signicantly correlated with %energy from
CHO, FAT, or PRO. These relationships did not vary based on gender or race.
CONCLUSION: PBF in prepubertal children was associated with total energy
intake but not necessarily percentage intake from particular macronutrients.
These results need replication with larger samples. Future research should
examine the longitudinal effect of macronutrient intakes on adiposity devel-
opment.
P102
Eating behaviour and disorders in diabetic and non-diabetic obese
subjects evaluated on rst access to a diabetes unit
C Taboga
1
, A Angarano
2
, E Biasin
2
, R Ciano
2
, S Mreule
1
, P Mule
2
, L Tonutti
1
,
M Balestrieri
2
, and C Noacco
1
1
Diabetes Unit, General Hospital, 33100 Udine, Italy;
2
Psichyatric Clinic,
Medical Faculty, 33100 Udine, Italy
INTRODUCTION: Correlation between Obesity and Eating Disorders (ED) is
well acknowledged and a high percentage of obese subjects asking for a diet
may suffer from any ED. On the other hand, type-II Diabetes Mellitus (DM II) is
frequently associated with obesity but the prevalence of ED in DM-II is less
investigated and still controversial. Aim of this study is to assess and compare
Eating Behaviours of obese subjects with (DO) or without DM-II (O), both on
rst access to a Diabetes Unit for a therapeutic program.
METHODS: 53 DO and 52 O have been consecutively enrolled in the study.
DO (31M, 22F) had a mean age of 53yrs (range 34 65, n

5<45 yrs.), a
mean Body Mass Index (BMI) =35.7 (range 30.1 64.3); O (7M, 45 F) had a
mean age =48 (30 77, n

21<45 yrs) and a BMI =35.2 (30.1 49.2).

Patients on their rst access were evaluated by the following tests: Eating
Disorder Inventory-2 (EDI-2) and its Symptom Checklist (EDT-SC), Illness
Behaviour Questionnaire (IBQ), Toronto Alexithymia Scale 20 (TAS-20). DO
were either recently diagnosed ( <1yr) or with a longer duration of the
disease (1 20 yrs).
RESULTS: O had higher scores in EDI-2 subscales: drive for thinnes, body
dissatisfaction and asceticism (p<.05), also they used more frequently
laxatives and diuretics to control body weight and had a major use of
psychiatric drugs (p<.05). DO had a higher consumption of drugs and also
they took more physical exercise. DO with a former DM II had higher scores
on asceticism scale in comparison with earlier diagnosed diabetics (p<.05).
The prevalence of ED has been 1.9% in DO and 7.7% in O (p=ns).
CONCLUSION: Prevalence of ED is one of the lowest until now reported, both
in DO and in O. DO showed less tendency to develop ED and less psycho-
logical discomfort in comparison with O, difference probably due to the
younger age and the higher rate of F in the O group. Former diagnosed DM II
subjects did not show a worsening of eating habits over time, on the contrary
they had a higher tendency to develop perfectionistic and rigid behaviour.
This feature could be an adaptation to the disease. From our data ED do not
represent a major concern or a hindrance in the implementation of lifestyle
changes necessary for management of obesity associated with diabetes.
Track 3 Eating patterns and behaviour
S59
International Journal of Obesity
P103
Locus of control and social desirability in relation to weight reduction
B Adolfsson
1
, I Andersson
1
, A Carison
1
, A Hagman
1
, and S Rossner
1
1
The Obesity Unit, Dept of Medicine, Karolinska Institutet, Huddinge University
Hospital, 141 86 Stockholm, Sweden
INTRODUCTION: Locus of Control (LOC) describes to what extent a person
considers herself in charge of her own life. Social Desirability (DS) is dened as
the need to gain approval from others by describing oneself in a socially
favourable light. This study explored if there was any association between
weight reduction, LOC and DS among participants in a weight reduction
programme.
METHODS: Forty-one obese, BMI >30 kg=m
2
, participants in a Day Care
weight reduction programme completed a LOC- and DS-questionnaire before
treatment (0) and after one year of treatment (1). Differences in LOC (0,1) and
DS (0,1) between the participants who had a one year weight reduction of _
5% and the participants who did not was looked for. The result was compared
with one group of healthy and one group of depressed patients. After two
years the participants' weights were registered.
RESULTS: Thirteen of the 24 remaining participants lost _ 5% of original
weight during the rst year of treatment. There was a signicant difference
between the group with a weight reduction _5% and the group with a
weight reduction <5% during the rst year in LOC 0 (p=0.002). LOC 1
(p=0.004) and DS 0 (p=0.026). All participants gained weight during the
second year. A signicant association existed with the control group of
depressed persons.
CONCLUSIONS: An empowering treatment program during the rst year
might have helped to redirect the external LOC of the patients with a weight
reduction <5% to a more internal one, thus probably facilitating a weight
loss. However it was seen that a relatively internal direction of LOC, and a low
need to appear socially desirable, was not enough for a lasting weight loss. As
reported in other studies frequent meetings in a support group may be
needed during a period of maintenance in order to establish new eating
habits and thus leading to a lasting weight reduction.
P104
Differences in nutrient metabolism and energy intake between men and
women
AM Johnstone
1
, SD Poppit
2
, G Horgan
1
, JE Blundell
3
, N King
3
, and RJ Stubbs
1
1
ACERO, Rowett Research Institute, Bucksburn, Aberdeen AB21 9SB, UK;
2
Auckland University, NZ;
3
BioPsychology Group, Leeds University, UK
There is a hierarchy in extent to which cumulative nutrient balance (NB)
relates to subsequent energy intake (EI). Each MJ increase in NB relates
negatively to EI in the order Protein>CHO>Fat.
The present work extends these analyses (Stubbs et al, 1995) to ad libitum
feeding in four residential chamber studies- 2 studies in men (2 n=6) and 2
studies in women. One study in each sex employed a diet manipulation (HF vs
LF) over 7d (M) or 5d (F, n=6). The other used sedentary vs active regimes (F,
n=9), with calorimetry data collected for 7d. For each day, multiple regres-
sion analysis was used to obtain the best predictor of subsequent day's EI
based on cumulative NB.
The above hierarchical relationships were still apparent. Meta-analysis
(MA) of studies indicated that carbohydrate (CHO) and protein (P) balance
had a signicant negative effect on subsequent EI in men (p<0.05), whereas
only protein balance had a signicant negative effect on subsequent EI
(p<0.05) in women. Fat had no suppressive effect on subsequent EI. Further
MA indicated that there was a signicant difference between exercise and diet
manipulation for CHO (t = 73.19; p=0.004) and there was also a signicant
effect between men and women for CHO (t = 73.85; p=0.001). There was
an apparent difference in the physiological determinants of EI across sexes.
Stubbs R.J., Harbron C.G., Murgatroyd P.R. and Prentice A.M. (1995) Am J
Clin Nutr 62, 2, 316 330.
P105
Unhealthy eating behaviours in women
JJ Lara
1
, and MEJ Lean
1
1
Department of Human Nutrition, University of Glasgow, Glasgow Royal
Inrmary, Glasgow G31 2ER, Scotland
INTRODUCTION: Unhealthy behaviours contributing to overeating are com-
monly believed to be more frequent in obese than in lean subjects. Obese
subjects have been documented to spend more time watching TV, compared
to lean subjects, inuencing food choice and decreasing physical activity.
METHODS: The present study reports the frequency from a questionnaire
survey of selected ``unhealthy'' behaviours in 184 women; 50 obese (BMI
_30) patients attending weight management groups, and 134 nurses.
RESULTS: The presence of ``unhealthy'' eating behaviours is shown as
frequencies and percentage (in brackets), in the table.
Nurses
BMI <25
n=52
Nurses
BMI 25 30
n=45
Nurses
BMI >30
n=37
Patients
BMI >30
n=50
Mean age (years),
BMI(kg=m
2
)
32.5, 22.2 38.4, 27.3 39.7, 33.6 46.3, 34.1
Eating too much 7 (13.5) 19 (43.2) 18 (48.6) 25 (50.0)
Watching TV
while eating
23 (44.2) 16 (36.4) 19 (51.4) 18 (36.0)
Snacking while
working
23 (44.2) 16 (36.4) 14 (37.8) 7 (14.0)
Snacking while
watching TV
36 (69.2) 33 (75.0) 29 (78.4) 29 (58.0)
Eating too fast 22 (42.3) 16 (36.4) 9 (24.3) 24 (48.0)
Chewing insufciently 8 (15.4) 5 (11.4) 3 (8.1) 15 (30.0)
Eating in inappropriate
places
7 (13.5) 3 (6.8) 5 (13.5) 5 (10.0)
Bingeing regularly 17 (33.3) 14 (31.1) 10 (27.0) 19 (37.3)
Bingeing intermittently 14 (27.5) 22 (48.9) 16 (43.2) 16 (31.4)
CONCLUSIONS: Apparent differences in behaviours between obese patients
and obese nurses who have not sought help were found. Some patterns e.g.
bingeing, show little difference between obese and lean subjects or may
represent aspects of misreporting in overweight and obese subjects. Obese-
subjects in this study outwith the context of a quantitative dietary assessment
admitted to eat too much. JJ Lara was supported by CONACyT in Mexico.
P106
Eating triggers in female subjects
JJ Lara
1
, and MEJ Lean
1
1
Department of Human Nutrition, University of Glasgow, Glasgow Royal
Inrmary, Glasgow G31 2ER, Scotland
INTRODUCTION: Different factors, or experimental situations in the labora-
tory, have been recognised as triggering or facilitating overeating e.g. comfort
eating.
METHODS: A total of 184 women aged 18 to 65y participated in the study,
50 of them were obese (BMI _ 30) patients attending weight management
groups and 134 nurses (BMI <25, N=52; BMI 25 29.9, N=45; BMI >30,
N=37). As part of a nutrition survey on underreporting, recognition and
experience of situations leading to eating without being hungry (trigger
situations) was sought in an open question.
RESULTS: Ninety seven out of 182 subjects (53%) declared having identied
triggers for eating. The proportion of subjects who identied triggers, the
number of triggers reported and the most common reported situations are
shown in the table.
Nurses
BMI <25
n=52
Nurses
BMI 25 30
n=45
Nurses
BMI >30
n=37
Patients
BMI >30
n=50
Subjects reporting triggers 34 (65%) 22 (49%) 19 (51%) 23 (46%)
One trigger 16 17 0 2
Two triggers 16 4 6 10
Three triggers 1 1 9 11
Four triggers 1 0 4 0
Boredom 21 9 13 5
Stress, Anxiety 5, 2 7, 0 11, 6 12, 8
Other people eating 0 1 11 7
Loneliness 3 1 4 1
CONCLUSIONS: An important proportion of women with different BMI are
aware of situations triggering eating. More lean subjects reported having
identied triggering situations compared to the heavier groups, however
these latter were more likely to report more than one trigger. Whether this
may relate to differences in exposure is unknown, however it could help in the
development of behavioural strategies to tackle weight gain.
JJ Lara was supported by CONACyT in Mexico.
Track 3 Eating patterns and behaviour
S60
International Journal of Obesity
P107
Gastric motility and food palatability in obesity
A Zurakowski
1
, B Zahorska-Markiewicz
1
, J Kolodziejczyk
2
, S Nowak
2
,
M Olszanecka-Glinianowicz
1
, T Sobezyk
1
, and S Staszek
1
1
Dept. of Pathophysiology, Silesian University School of Medicine, 40-752
Katowice, Poland;
2
Dept. of Nuclear Medicine, Silesian University School of
Medicine, 40-752 Katowice, Poland
INTRODUCTION: There are conicting information about differences of
gastric emptying rate between lean and obese subjects and lack of informa-
tion concerning antral motility in both groups. The purpose of this study was
to evaluate the inuence of food palatability on gastric motility in obese
subjects compare to lean controls.
METHODS: Gastric antral motility and gastric emptying was measured after
ingestion of sweet omelette in 2 groups of patients: 9 obese and 9 lean
subjects by a scintigraphic technique. Ratings of fullness, satiety and palat-
ability of the food on seven-point rating scale were recorded during the study.
RESULTS: There were no difference in gastric emptying proles, gastric antral
motility, assessments of satiety, fullness and palatability of the food between
the groups. Correlation between satiation signals and gastric emptying and
antral motility were not found in both groups. Strong correlation were
observed between palatability ratings and gastric emptying rate in lean
group (p<0.01 r = 70.8). This correlation was not found in obese.
CONCLUSIONS: There are no differences in rate of gastric emptying and
gastric antral motility between lean and obese subjects. Gastric content is not
related to the postprandial fullness and satiety in obese and lean objects. We
observed strong negative correlation between palatability ratings and gastric
emptying in lean group. It may indicate better control of food intake in non-
obese subjects.
P108
Frequency of dieting and eating restraint in women, using the Three-
Eating Factor Questionnaire
JJ Lara
1
, and MEJ Lean
1
1
Department of Human Nutrition, University of Glasgow, Glasgow Royal
Inrmary, Glasgow G31 2ER, Scotland
INTRODUCTION: Relationships between dieting and psychological measures
are of interest given the high prevalence of dieters, obese and non-obese, in
most populations. This study assessed the relationship between self-reporting
dieting frequency and psychological measures.
METHODS: Data from 131 female nurses were assessed using the three-eating
factor questionnaire (Stunkard & Messick, 1985).
RESULTS: Mean and SD, are provided in the table for the relevant variables.
CONCLUSIONS: Current dieters showed signicantly greater dietary restraint
than non-dieters, even with similar BMI. Women who are not currently dieting
exhibit a relationship between eating restraint and previous dieting frequency.
Disinhibition and hunger were signicantly related to previous dieting fre-
quency irrespective of current dieting status. These factors might predict poor
weight loss. JJ Lara was supported by CONACyT in Mexico.
Dieting cur-
rently and on
past two years Age (y)
BMI
(kg=m
2
)
Waist circum-
ference (cm)
Dietary
Restraint
Disinhi-
bition Hunger
Current dieters
1 2 previous
diets n=15
36.17.3 27.72.5 84.55.7 13.34.3 5.92.4 2.92.6
_3 previous
diets n=16
39.77.7 29.95.1 87.79.5 13.03.8 9.83.2 6.53.2
Current non=dieters
Zero previous
diets n=53
36.710.6 24.54.9 76.610.9 5.74.2 3.83.0 3.73.0
1 2 previous
diets n=35
34.711.2 27.54.8 82.39.2 7.24.3 7.43.6 5.54.1
_3 previous
diets n=12
36.66.6 31.98.0 91.010.3 9.52.8 9.72.4 6.83.6
P109
Maintenance of reduced weight is associated with active weight loss
efforts
J Westenhoefer
1
, A Stellfeldt
1
, S Fintelmann
1
, R Schoberberger
1
, and B Ludvik
1
1
Fachhochschule Hamburg University of Applied Sciences, Department of
Nutrition and Home Economics, Nutritional and Health Psychology Group,
Lohbruegger Kirchstr 65, 21033 Hamburg, Germany
INTRODUCTION: The Lean Habits Study is an ongoing prospective cohort
study of over 8000 overweight subjects on the association between beha-
vioural patterns and maintenance of reduced body weight.
METHODS: Included in the present analysis were 1500 subjects who began
participation in the BCM diet programme between January and May 1998.
Body weight was measured at one and two year follow-up (1y and 2y EU) for
54.4% resp. 32.9% of the original sample.
RESULTS: After 1y 598 subjects (39,9% of the original sample) had success-
fully reduced their body weight by 5% or 1 BMI unit, after 2y 301 subjects
(20.1%). Weight maintenance or further loss between 1y and 2 y was seen in
111 subjects (7.4%). In the 438 subjects with complete 1y and 2y FU data
there was an average weight gain of 2.6 kg (SD=5.0) between 1y and 2y.
From the 581 subjects who provided 2y FU data, a total of 374 (64.4%) had
meanwhile undertaken some active weight loss efforts in addition to the
original participation in the BCM programme. These subjects maintained their
reduced weights signicantly more often between 1 y and 2y (21.7% vs
14.5%; p<0.05).
Conclusion: Weight stabilisation after weight reduction is a rare event.
Subjects who continue or restart measures to reduce weight during the
follow-up period are more successful in maintaining weight loss.
P110
Eating behaviour and prevalence of overweight and obesity among
schoolchildren
M Pavlovic
1
, S Bijelovic
1
, D Balac
1
, N Majkic-Singh
1
, Z Bolits
1
, and A Kadvan
1
1
Public Health Institute Subotica, Regional Nutrition Centre, Yugoslavia,
http:==www.zzzzsu.org.yu
INTRODUCTION: The eating behaviour are developed from an early age of
child. An unadequate knolige, as well as the past economic situation, has led
us to the incrase of nutritive disorders, such as overweight and obesity. The
aim of this work is to show the quality of familial nutrition and eating habits
among overweight and obese schoolchildren in Yugoslavia.
METHODS: The representative sample of schoolchildren aged 9 10 (3255
boys and 3060 girls) had been chosen for transversal examination in 13
centres. The investigation included the anthropometric and biochemical
parameters (body weight, body height, BMI, TC, HDL-C, LDL-C, TG, LDL-
C=HDL-C), as well as parameters of eating habits and nutrition quality.
Statistical analysis and evaluation of data were done by software in MS-DOS
version, according to NHANES I, National Reference Values, RDA and WHO
recommendation.
RESULTS: The average energy daily value was 2981.96kCal (p<0.01), while
the average participation of proteins was 14.54% (p>0.05), of fats 38.39%
(p>0.05) and of carbohydrates 47.46% (p>0.05). In the structure of family
nutrition cereals were represented by 33.79% (p>0.05), fruit and fruit
products by 4.51% (p<0.01), vegetable and vegetable products by 7.85%
(p<0.01), meat, meat products, eggs, sh and leguminoses by 20.57%
(p<0.01), milk and milk products by 10.19% (p>0.05) and fats, oils and
sugars by 21.70% (p<0.01). The average participation of dietary cholesterol
was 226.69 mg daily (p>0.05), of SFA 11.35% (p>0.05), of PFA 7.76%
(p>0.05), of MFA 9.76% (p>0.05) and of dietary bres 16.46 g daily
(p<0.05). According to BMI (NHANES I), there was found 10.71% over-
weight and 8.78% obese schoolchildren, with the predominance of boys from
the central and south regions of Serbia. The high level of TC, LDL-C and TG
were found in 11.94%, 27.87% and 21.95% of boys and 13.53%, 31.11%
and 27.45% of girls.
CONCLUSION: The nutrition with domination of fats, oils, sugars, meat and
meat products represents the well known risk factors of noncommunicable
diseases, which are, regarding to prevalence of obesity and overweight,
already present at school age. This situation demands adequate preventive
action with the participation of whole society.
Track 3 Eating patterns and behaviour
S61
International Journal of Obesity
P111
Perception of emotions in obese patients
E Frigerio
1
, MG Strepparava
1
, D Perrett
2
, DM Burt
2
, R Bisanti
1
, L Vigna
3
,
G Rezzonico
4
, and GC Secchi
3
1
Psychology Inst., Medical Faculty, Milano University (I);
2
Perception
Laboratory, Schoolof Psychology, St. Andrews University (UK);
3
Dept. Internal
Medicine, Istituti Clinici di Perfezionamento, Milano (I);
4
Experimental,
Environmental Medicine Dept., Medical Faculty, Milano-Bicocca University (I)
INTRODUCTION: One of the main difculties in dealing with obesity is that
patients are often incapable of recognizing inner states and, therefore, also of
following structured diets. Research shows that there is a strong relationship
between the perception of physical states and that of emotions. In this
context, it is then an accepted fact that obese patients have severe difculties
in recognizing their own feelings. This research starts from the hypothesis that
the difculty of recognizing one's own emotions is linked with the incapability
of decoding outside signals. The aim is therefore to evaluate if and to what
extent obese patients have difculties in linking facial expressions to feelings in
general, to particular classes of emotions, and to feelings that specically
involve the subject.
METHODS: 20 female obese patients (mean BMI: 36, mean age: 48yrs), were
studied against a control group. We evaluated their sensitivity to facial
expression of emotions by an experimental tool developed at St. Andrews
University. On an PC screen, 48 male and female faces, front and prole, were
shown to the subjects. Facial expressions ranged from neutral to maximum
intensity of emotional display. Emotions presented were: happiness, anger,
sadness, disgust, surprise and fear. Subjects were rst asked to label the
emotion from the maximum-intensity picture, the asked to point out the
threshold of appearance of each emotional status.
RESULTS: Obese patients show a particularly keen perception of disgust, but
only in front pictures. This is evident in the labelling of full-intensity emotions
as well as in the threshold perception. On the other hand they show a marked
difculty in recognizing fear and sadness. Also, in general their threshold from
neutral to emotional states is higher than that of controls.
P112
Cognitive-behavioral therapy in obese children
P Hosek
1
, S Liskova
1
, and J Plomer
1
1
University of West Bohemia, Department of Biomedicine, Klatovska 51 CZ 320
00 Pilsen, Czech Republic
INTRODUCTION: The aim of this study was to identify psychological distress
associated with obesity in children and modify it during residential weight loss
program.
METHODS: Seventy obese children (41 girls, 29 boys) with mean age 13,7
years (range 12 15) and mean BMI 30,2kg=m
2
(range 23 41) participated
in a 5 weeks weight loss program. A special program with moderate caloric
restriction, elevated physical expenditure and cognitive-behavioral therapy
(both individual- and group-level intervention) has been established. 1.
Anxiety and psychosocial distress 2. self-esteem and 3. social relations were
investigated. In three sessions per week a wide range of methods in therapy
was used: relaxation techniques, autogenic training, psychogymnastic, mod-
ied psychodrama and education strategies.
RESULTS: Ninety percent of subjects admit anxiety experiences, more than
70% of subjects feel dissatisfaction with body shape and discrimination in any
walk of life, only twenty percent disclaim any problem with contemporaries
contact at the time of program beginning.
At the end a signicant weight decrease was obtained (BMI 27kg=m
2
).
The normalization of anxiety prole (Gittelman's Chlidren Manifest Anxiety
Scale) indicates improvement after the treatment.
CONCLUSION: Positive short-term therapeutic effect in weight loss program
for obese children depends mostly on dietary restriction and exercise. Educa-
tion and behavior modication for long-term effect are necessary.
This work was supported by the Czech Ministry of Health (Grant No.
NE=6240-3).
P113
Comparison of energy and nutrients intake among overweight=obese
and non-overweight elderly people
AM Rocandio
1
, MT Garc a-Arias
2
, S Ysart
2
, L Ansotegui
1
, and MC Garc a-
Fernandez
2
1
Department of Nutrition and Food Science, University of Basque Country,
Vitoria, Spain E-mail: knpropaa@vf.ehu.es.;
2
Department of Hygiene and Food
Technology, University of Leon, Spain
INTRODUCTION: Most studies of energy intake among obese and non-obese
people suggest that obese people eat more than the non-obese peers. The
purpose of this study, therefore was to determine differences in dietary intake
between overweight=obese and non-overweight elderly people.
SUBJECTS AND METHODS: Dietary consumption was evaluated in a group of
107 institutionalized elderly people between 65 and 98 (80.546.47). 42
(39.2%) of whom were male and 65 (60.7%) female. Data was compared
with reference values (OMS); the subjects with Body Mass Index >25 was
considered overweight=obese. The method employed involved the weighing
method over a period of seven days. Data were analysed using w
2
tests and
Student's t test (SPSS software v 7.0.).
RESULTS: 67% of examined subjects were considered overweight=obese. The
mean daily energy intake (8375.31161.1kJ) was agree with the recom-
mendations, but there were no signicant differences between both groups
(8333.231129.68kJ=d overweight=obese vs 8349.211255.2 non-over-
weight).
Macronutrient intake
Overweight=
obese
Non-
overweight
Signicant
differences
Protein (g=d) 82.715.0 81.2512.3 NS
Carbohydrate (g=d) 218.5839.14 216.8538.00 NS
Fat (g=d) 86.4819.80 86.1221.25 NS
Saturated fatty acids 27.379.88 27.59.42 NS
Monounsaturated fatty acids 35.248.98 35.639.61 NS
Polyunsaturated fatty acids 10.943.25 11.213.84 NS
NS =No signicant differences (P >0.05).
The mean cholesterol intake was 315.5983.42 mg=d and the differences
were no signicant (304.8586.31 in overweight=obese vs 331.7278.42 in
non-overweight group).
CONCLUSIONS: Although these ndings were not statistically signicant,
data conrm the need for better strategies to improve the diets in institutio-
nalised elderly people.
P114
Analysis of a population of obese patients after a ve-day stay in the
University Hospital of Bordeaux
H Dabadie
1
, C Martinez
1
, M Bellocq
1
, V Adrian
1
, and H Gin
1
1
Department of Nutrition, Hopital Haut-Levoque, Pessac, France
INTRODUCTION: The management of obese subjects often requires a short
hospitalization to initiate nutritional education and weight loss. A unit for ve-
day hospitalization for obesity was created in the Bordeaux University Hospital
4 years ago.
METHODS: Between October 1998 and September 1999, 171 obese subjects
were hospitalized for single 5-day periods in the unit of Department of
Nutrition.
RESULTS: The population included 147 women (86%) and 24 men. Mean
patient age was 50 years (range, 16 to 76 years), mean BMI 39kg=m
2
, and
mean waist-to-hip ratio 0.88. At diagnosis, associated risk factors were noted
in 85 patients (50%): dsylipidemia (60%), hypertension (47%), type 2
diabetes (43.5%), and angina pectoris (12%). Other complications included
depression (55%), hypothyroidism (18%), and obstructive sleep apnea (14%).
At entry, mean intake was 1900kcal=d (range, 630 5400kal=d) furnished
41%, 40%, and 19% by carbohydrates, fats, and proteins, respectively. Mean
proposed intake was 1200 kal=d (range, 800 2000 kal=d). Mean weight loss
during the 5-day period was 1.9 kg (p<0.0001). The distribution of weight
loss was as follows: less than 1kg (18% of subjects), between 1 and 2kg
(68%), between 2 and 3kg (9%), more than 4kg (5%). Among the 95
patients (44%) seen again 1 year later, 84 of them (88%) had lost an average
of 6.2 kg.
CONCLUSION: A 5-day hospitalization for obese patients allows diagnosis of
high-risk subjects (cardiovascular disorders). Weight loss over 5 days was
approximately 2kg for two-thirds of these patients and 88% of those who
were seen at 1-year follow-up had lost around 6kg.
Track 3 Eating patterns and behaviour
S62
International Journal of Obesity
P115
The efcacy of the diet and increased physical activity in the treatment of
overweight and obese patients
Lucyna Ostrowska
1
, and Jan Karczewski
1
1
Department of Hygiene and Epidemiology, Medical Academy of Bialystok
Summary
Obesity is an excessive accumulation of fatty tissue in the organism. Faulty
eating habits and low physical activity are the major cause of overweight and
obesity. The aim of the study was to investigate the efciency of dietetic
treatment and heightened physical activity in patients with excessive body
mass.
The study included 140 patients with overweight or simple obesity (131
women and 9 men), aged 15 65, members of the support group previously
instructed in dietetics. The treatment consisted in the application of a diet,
1000 1500 kcal=day, and the participation in the programme of body-mass
reducing exercise (physical exercise twice a week and exercise in water once a
week for 45 min). Attendance was monitored and those who missed the
classes were excluded from the group. The investigations were terminated
after 2 years.
In group I (exercising for 0 3 months) 39 patients reduced their body
mass by 7.7kg on the average. In group II (exercising for 3 6 months) 59
patients reduced their body mass by the average of 14.5kg. In group III
(exercising for 6 12 months) in 38 patients the average loss of body mass was
20.9kg. In group IV (exercising for over 12 months) 4 patients reduced their
weight by 26 70 kg (mean 50 kg), which gave the 26.5% 47% loss of the
initial body mass.
The patients with overweight and obesity included in the present study
reduced their body mass through the application of an adequately selected
diet and heightened physical activity. Fewer than half of the patients were able
to persevere with over 6 months of the programme.
Track 3 Eating patterns and behaviour
S63
International Journal of Obesity
Reproducedwith permission of thecopyright owner. Further reproductionprohibited without permission.