Experimental Gerontology 48 (2013) 485–491

Contents lists available at SciVerse ScienceDirect

Experimental Gerontology
journal homepage: www.elsevier.com/locate/expgero

Age ≥ 60 years was an independent risk factor for diabetes-related complications

despite good control of cardiovascular risk factors in patients with type 2
diabetes mellitus
Boon How Chew a,⁎, Sazlina Shariff Ghazali a, Mastura Ismail b, Jamaiyah Haniff c, Mohd Adam Bujang c
a
Department of Family Medicine, Faculty of Medicine and Health Sciences, Universiti Putra Malaysia, Malaysia
b
Klinik Kesihatan Seremban 2, Seremban, Negeri Sembilan, Malaysia
c
Clinical Research Centre, Hospital Kuala Lumpur, Malaysia

a r t i c l e i n f o a b s t r a c t

Article history: Providing effective medical care for older patients with type 2 diabetes mellitus (T2D) that may contribute to
Received 17 July 2012 their active aging has always been challenging. We examined the independent effect of age ≥60 years on disease
Received in revised form 7 February 2013 control and its relationship with diabetes-related complications in patients with T2D in Malaysia. This was a
Accepted 8 February 2013
cross-sectional study using secondary data from the electronic diabetes registry database Adult Diabetes Control
Available online 27 February 2013
and Management (ADCM). A total of 303 centers participated and contributed a total of 70,889 patients from
Section Editor: Andrzej Bartke May 2008 to the end of 2009. Demographic data, details on diabetes, hypertension, dyslipidemia and their treat-
ment modalities, various risk factors and complications were updated annually. Independent associated risk
Keywords: factors were identified using multivariate regression analyses. Fifty-nine percent were female. Malay comprised
Aged 61.9%, Chinese 19% and Indian 18%. There were more Chinese, men, longer duration of diabetes and subjects that
Type 2 diabetes mellitus were leaner or had lower BMI in the older age group. Patients aged ≥60 years achieved glycemic and lipid targets
HbA1 but not the desired blood pressure. After adjusting for duration of diabetes, gender, ethnicity, body mass index,
Blood pressure disease control and treatment, a significantly higher proportion of patients ≥60 years suffered from reported
LDL cholesterol
diabetes-related complications. Age ≥60 years was an independent risk factor for diabetes-related complica-
Diabetes complications
tions despite good control of cardiovascular risk factors. Our findings caution against the currently recommended
control of targets in older T2D patients with more longstanding diseases and complications.
© 2013 Elsevier Inc. All rights reserved.

1. Introduction
Age has long been recognized as a significant factor for health (de
Craen et al., 2009; Jaakko, 2004). Life spans and life expectancy have
been used as indicators of socio-economic-political development for
most countries. Many cardiovascular disease risk scoring systems
give larger weightage or marks for the older age groups, indicating
that the older a person becomes, the higher risk the person has for
cardiovascular diseases (Chamnan et al., 2009; Cooney et al., 2011).
The reasons for this association of age and many cardiometabolic
diseases are degenerative processes leading to cellular apoptosis
beyond regeneration or repairs (Navarro and Boveris, 2007). There
were exceptional examples among the children of long-living parents,
suggesting that they had advantageous cardiovascular risk profiles, as
reported by the Leiden research program on aging and Framingham
offspring study (de Craen et al., 2009; Terry et al., 2007). However, the
direction of effect could also be that having cardiometabolic diseases
⁎ Corresponding author at: Department of Family Medicine, Faculty of Medicine and
Health Sciences, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor Darul Ehsan,
Malaysia. Tel.: +60 3 8947 2520x2538; fax: +60 3 8947 2328.
E-mail address: chewboonhow@yahoo.com (B.H. Chew).
cause premature aging and death (Girndt and Seibert, 2010). Many
studies have reported positive relationship of disease control (glycemic,
blood pressure and lipid) in type 2 diabetes mellitus (T2D) and diabetes-
related complications and death. However, the evidence was inconclu-
sive and even against intensive control in the older group of patients
(Huang et al., 2011).
Current research on molecular mechanism of endothelial function
pointed to the chronic excess of oxygen and nitrogen species as the
causes of vascular aging and remodeling (arteriosclerosis and athero-
sclerosis), leading to the increased risk of acute and chronic cardiovas-
cular diseases (Bachschmid et al., in press). This chronic oxidative
stress on top of hyperglycemia in diabetes mellitus was observed to
lead to more oxidative stress, inflammatory reactions and thrombosis
resultant of accelerated interaction between advanced glycation end
products (AGEs) and their receptor for AGEs (Sho-ichi, 2011). AGEs
are the age-related macromolecules that are irreversibly cross-linked
and covalently-bound following many non-enzymatic reaction and bio-
chemical conversion and rearrangement between ketones or aldehydes
and the amino groups of proteins, lipids and nucleic acids (Sho-ichi,
2011). The involvement of defective mitochondrial oxidative phosphor-
ylation leads to apoptosis, deterioration of tissues and organs (such as
the brain, cardio-vascular system, immune system, intestine, macula

0531-5565/$ – see front matter © 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.exger.2013.02.017
486 B.H. Chew et al. / Experimental Gerontology 48 (2013) 485–491
of the eye, Langerhans islets, prostate gland, oocytes of ovaries) structure
and function in aging (Muradian and Schachtschabel, 2001).
Many developing countries in Asia are witnessing an epidemic of
cardiometabolic diseases and fast approaching population aging (Chan
et al., 2009). Malaysia is not exempted from the increasing disease bur-
den of diabetes mellitus and its related complications (Institute for
Public Health, 2008). In 2010, there were about 2.2 million or 8.4% of
total Malaysia population age ≥60 years (Department of Statistics,
2010). At about the same time, we have about one third of the elderly
population (age ≥60 years) confirmed of diabetes mellitus (Institute
for Public Health (IPH), 2011). This group of elderly diabetes patients is
becoming the most challenging group because they frequently presented
with multiple comorbidities, on multiple medications, mainly speaking
in their own mother tongue, forgetful, with diverse health beliefs and
low health literacy (Al-Qazaz et al., 2011; Cigolle et al., 2011). Knowing
their demographic and clinical characteristics would help healthcare
providers to pay appropriate medical attention to their needs, and the
policy-makers in channeling a worthwhile investment.
We examined the older T2D patients' demography, clinical charac-
teristics and the independent effect of age ≥60 years on disease control
and its relationship with diabetes-related complications. The results of
this study could help the country and other countries with similar
health-care systems to better strategize their approaches to support
the older T2D patients to age more actively.
2. Methods
This was a registry-based study using the Adult Diabetes Control and
Management (ADCM) registry. It represents T2D patients from 303
public health centers (289 health clinics, 14 hospitals), which contribut-
ed a total of 70,889 patients (1900 or 2.8% patients were from hospitals)
from inception of the registry in May 2008 until 31st December 2009.
This represents 33.7% (289/858) health clinics and 10.4% (14/135) hospi-
tals from 66.7% (8 states and 2 federal territories) of 15 states and federal
territories in Malaysia. The proportion of T2D patients notified in this
database as compared to the estimated total T2D patients in these states
and federal territories was 5.2% (70,889/1,368,590) (Ismail et al., 2011).
Until 31st December 2009, 3140 (4.4%) were lost to follow-up (defined
as defaulted appointment at the particular clinic for >one year) and
203 (0.3%) patients had passed away.
Adult patients aged ≥18 years were registered. All patients were
informed of the on-going registry and given the opportunity to opt out.
The participation in ADCM was non-mandatory for patients and health
centers. Registrations at local centers were generally performed by
trained physicians, assistant physicians and nurses. All participating
diabetic patients were registered on a paper form or via on-line standard
case record form (CRF) made available in the ADCM website, developed
and maintained by Clinical Research Centre, Ministry of Health, Malaysia.
All online registrations were sent to a central database and analyzed by
using STATA version 9 and PASW 19.0 (SPSS, Chicago, IL). The registra-
tions are per current calendar year, i.e., data from 2009 were registered
in the year of 2009 and latest results in the current year were updated
for patients who were registered in the previous years.
2.1. The Malaysian Health Care system for diabetes patients
This health-care system is generally supported by two main sectors,
namely the public and the private sectors (Awin, 2001). The private
health care is provided by about two-thirds of the country's medical
specialists caring for about one-third of the upper-middle-income
groups of the population. The public health care is organized by Ministry
of Health, structured into public health and hospital care. The health
clinics are generally well-linked to the secondary and tertiary public
hospitals with a seamless referral system. Diabetes care at these public
health clinics is managed by family medicine specialists (FMS), medical
officers (MO), physician assistants and often supported by specialized
nurses and dietitians/nutritionists. In hospitals this is provided by
specialists in internal medicine or endocrinology/diabetology, medical
officers and by specialized nurses. Every patient who is diagnosed with
diabetes mellitus will receive a green booklet (kept by patient) which
comes in pair with a bigger green medical record book (kept at the
health center) that tracks all information pertaining to all medical care
provided. Diabetes patients generally visit their physicians several
times per year. The patients in health clinics generally are managed by
physician assistants and MO (non-specialist primary care physicians),
but once they become uncontrolled or have complications, they will be
transferred under the care of FMS, who may co-manage with the hospital
specialist. Referrals to specialists at secondary or tertiary hospitals are
done for patients with complicated diseases needing treatment and
expertise that is not available in health clinics.
2.2. Definitions of study participants
In this report, we aimed to describe the effect of age (≥ 60 years)
on disease control and its relationship with diabetes-related com-
plications for patients with T2D in Malaysia. The definition of type
2 diabetes was as when their case record fulfilled all these criteria:
(i) either documented diagnosis of diabetes mellitus according to
the WHO criteria or (ii) those whose current treatment consisted
of life-style modification, on oral AHA or insulin. Hyperlipidemia refers
to an increase in concentration of one or more plasma or serum lipids,
usually cholesterol and triglycerides, and the term dyslipidemia is
used for either an increase or decrease in concentration of one or
more plasma or serum lipids. Hypertension was diagnosed if the systol-
ic blood pressure was ≥130 mm Hg or the diastolic blood pressure was
≥80 mm Hg on each of two successive readings obtained by the clinic
physician.
2.3. Clinical investigations
HbA1c ≤ 6.5%, BP b 130/80 mm Hg, LDL ≤ 2.6 mmol/L, triglyceride
(TG) ≤ 1.7 mmol/L and HDL ≥ 1.1 mmol/L were regarded as treatment
targets. Reaching these treatment targets is considered good control
of these cardiovascular risk factors. BMI was calculated as weight
divided by height squared and ≤ 23 kg m 2 was taken as the thera-
peutic target. A standard recommendation for BP recordings has
been published in Malaysia clinical practice guidelines, stating that
a mean of two readings (Korotkoff 1–5) in the rested position with
arm at heart level shall be recorded using a cuff of appropriate size
(Health Technology Assessment, 2008). A BP b 130/80 mm Hg was
the treatment target. The latest results of these clinical characteris-
tics in 2009 were used in analyses. Diabetes complications reported
in ADCM were retinopathy, ischemic heart disease (IHD), cerebro-
vascular diseases (CVD), stroke or transient ischemic attack, ne-
phropathy and diabetic foot problems (DFP). These complications
were retrieved from patients' records. Diagnoses of these complica-
tions were made or confirmed by the attending physician at the clinic
based on the medical symptoms, laboratory results, radiological evi-
dence and treatment history at the clinic and other hospitals. Often
these diagnoses were informed by the relevant hospital specialist in re-
turn referral letter or reported by patient with concordant medication
they were prescribed from hospital. Retinopathy was usually diagnosed
after positive fundus appearance by fundus camera and further con-
firmed by an ophthalmologist. Nephropathy was diagnosed by the
persistent presence, on at least two occasions, separated by at least
three months, of any of the following: microalbuminuria, protein-
uria, serum creatinine > 150 μmol/L or estimated glomerular filtra-
tion rate b60 mL/min (calculated using Cockroft–Gault formula),
diabetic foot problem comprising foot deformity, current ulcer or
amputation, peripheral neuropathy or peripheral vascular disease.
The independent variable of interest was the age ≥60 years. Using
the age at notification, subjects were divided into two groups: those
 B.H. Chew et al. / Experimental Gerontology 48 (2013) 485–491 487

≥60 years (1) and those b 60 years (reference category). The effect of Table 1
Demography and clinical characteristics by age groups.
age ≥ 60 years toward each treatment target and diabetes complica-
tions were identified using multivariate logistic regression with b60 years, n (%) ≥60 years, n χ2 or t p
enter method. The relationship of this variable to the treatment (%) statistic

targets were conducted with adjustment for the patient demogra- Age, mean (SD) 50.22 (7.19) 68.12 (6.54) 343.50 b0.0001
phy: gender, ethnicity, duration of diabetes and BMI. We did not factor Gender Male 15,150 (39.0) 13,789 (43.0) 112.76 b0.0001
Female 23,584 (60.8) 18,257 (56.9)
in pharmacological modalities in the regression analyses because the
Missing 66 (0.2) 43 (0.1)
relationships between anti-hyperglycemic agents and glycemic control, Total 38,800 (100.0) 32,089 (100.0)
anti-hypertensive agents and BP, anti-lipid agents and lipid profiles (54.7)a (45.3)a
were generally linear rather than inverse in nature (Chew et al., 2012a, Ethnicity Malay 25,942 (66.9) 17,960 (56.0) 2397.60 b0.0001
2012b). The relationship of age ≥60 years to diabetes complications Chinese 4823 (12.4) 8628 (26.9)
Indian 7555 (19.5) 5184 (16.2)
was determined with adjustment for the abovementioned patient de- Other 351 (0.9) 237 (0.7)
mography and treatment targets: HbA1c > 6.5%, BP ≥ 130/80 mm Hg, Malaysians
LDL > 2.6 mmol/L, HDL b 1.1 mmol/L, anti-platelet, ACE inhibitors/ Foreigner 65 (0.2) 23 (0.1)
angiotensin receptor blockers (ARB) and statin. These patient demogra- Missing 64 (0.2) 57 (0.2)
Total 38,800 (100.0) 32,089 (100.0)
phy and clinical characteristics were generally accepted to be the poten-
(54.7)a (45.3)a
tial confounders (Chew et al., 2012a, 2012b; Lee et al., in press; Nakao et Diabetes duration in 4.84 (4.53) 7.12 (6.39) 49.81 b0.0001
al., 2012). We looked into the relationship of age ≥60 years and each of year, mean (SD)
the diabetes complications grouped as below: any diabetes complica- b5 18,467 (47.6) 10,717 (33.4) 2077.02 b0.0001
tion (1), macrovascular complications (CVD and/or IHD) (1), micro- 5–9 9618 (24.8) 8896 (27.7)
> = 10 3693 (9.5) 6071 (18.9)
vascular complications (retinopathy and/or nephropathy and/or Missing 7022 (18.1) 6405 (20.0)
DFP) (1); and each of the complication on its own, namely, CVD (1), Total 38,800 (100.0) 32,089 (100.0)
IHD (1), retinopathy (1), nephropathy (1) and DFP (1). There were al- (54.7)a (45.3)a
most equal proportion of unknown status/missing data in each of the BMI, mean (SD) 28.22 (5.69) 26.10 (6.07) −41.74 b0.0001
BMI b 23 12,629 (32.5) 14,072 (43.9) 955.88 b0.0001
complications by age groups of ≥ 60 and b60 years: CVD 39.8% vs.
BMI ≥ 23 26,171 (67.5) 18,017 (56.1)
36.5%, IHD 43.5% vs. 40.8%, retinopathy 53.4% vs. 49.4%, nephropathy Total 38,800 (100.0) 32,089 (100.0)
42.9% vs. 40.2% and DFP 37.8% vs. 34.7%. The normality of each variable (54.7)a (45.3)a
was tested by histogram and box plots and finally confirmed by Systolic BP, mean 134.21 (18.46) 139.86 (20.35) 34.57 b0.0001
Kolmogorov–Smirnov test. Multicolinearity between the variables (SD)
Diastolic BP, 80.36 (10.23) 76.78 (10.81) −40.38 b0.0001
was checked with correlation matrix and inspected for their standard
mean (SD)
error (SE) magnitude. We found no variables correlated with each BP ≥ 130/80 mm Hg 23,544 (75.2) 19,679 (78.2) 70.60 b0.0001
other; r b 0.2 and SEs were all within 0.001 to 5.0. Comparisons of BP b 130/80 mm Hg 7784 (24.8) 5496 (21.8)
mean levels were performed using the Student's t test for unpaired Total 31,328 (100.0) 25,175 (100.0)
(55.4)a (44.6)a
samples and for proportions by use of the Chi square test. A
HbA1c, mean (SD) 8.68 (2.28) 7.94 (2.02) −32.85 b0.0001
p value b 0.05 was considered to be significant at two tails. HbA1c > 6.5% 17,487 (85.5) 13,022 (77.4) 410.04 b0.0001
HbA1c ≤ 6.5% 2956 (14.5) 3798 (22.6)
3. Results Total 20,443 (100.0) 16,820 (100.0)
(54.9)a (45.1)a
LDL-C, mean (SD) 3.24 (1.11) 3.12 (1.08) −11.10 b0.0001
Fifty-nine percent were female. Malay comprised 61.9%, Chinese
LDL-C > 2.6 15,462 (71.5) 11,337 (65.8) 75.54 b0.0001
19% and Indian 18%. Further descriptive reports on demography, clinical LDL-C ≤ 2.6 6164 (28.5) 5885 (34.2)
characteristics, treatment modalities and complication rates had been Total 21,626 (100.0) 17,222 (100.0)
published elsewhere (Ismail et al., 2011). There were relatively more (55.7)a (44.3)a
HDL-C, mean (SD) 1.28 (0.52) 1.32 (0.52) 7.05 b0.0001
Chinese men with longer duration of diabetes mellitus and leaner BMI
HDL-C b 1.1 7527 (34.4) 5258 (30.2) 99.98 b0.0001
in the older age group as compared to the younger age group HDL-C ≥ 1.1 14,348 (65.6) 12,144 (69.8)
(Table 1). Fig. 1 showed the number of treatment targets achieved for Total 21,875 (100.0) 17,402 (100.0)
HbA1c, BP and LDL. After adjustment for duration of diabetes, gender, (55.7)a (44.3)a
ethnicity and BMI, patients aged ≥60 years were related to achieving TG, mean (SD) 2.02 (1.33) 1.83 (1.12) −15.90 b0.0001
TG > 1.7 11,940 (47.0) 8380 (41.2) 151.55 b0.0001
better glycemic and lipid targets but not BP (Table 2).
TG ≤ 1.7 13,461 (53.0) 11,936 (58.8)
About 19.0% of patients had at least one of the diabetes-related com- Total 25,401 (100.0) 20,316 (100.0)
plications. Fig. 2 showed that among those who were ≥60 years signifi- (55.6)a (44.4)a
cantly more suffered from reported diabetes-related complications BMI = body mass index, kg m2, BP = blood pressure, LDL-C = low density
ranging from macrovascular (stroke and IHD) to microvascular (retinop- lipoprotein-cholesterol, HDL-C = high density lipoprotein-cholesterol, TG = triglyceride.
a
athy, nephropathy and DFP). After adjusting for duration of diabetes, Row percentages.
gender, ethnicity, BMI, disease control and treatment, Table 3 showed
that age ≥60 years old was a significant independent risk factor for all
categories of diabetes complications in this study. ethnicity had increased in the elderly group, from 39% to 43% and
12.4% to 26.9% respectively, although the majority remained the
4. Discussion female and Malay. These could be one of the contributing factors for
the better disease control but poor cardiovascular outcomes because
We reported on the patients' demography, treatment targets and men and Chinese tend to have better disease control but higher com-
their relationships toward diabetes-related complications among plication rates (Chew et al., 2012c; Lee et al., in press). The duration of
the elderly (age ≥ 60 years) T2D patients from our national diabetes diabetes was expectedly longer among this elderly group as com-
registry in 2009. People aged ≥60 years comprise 45.3% of the type 2 pared to the shorter in the younger group. T2D patients become pro-
diabetic population in this cohorts. We observed an interesting gressively leaner with age, as was observed in other studies as well as
change in the demography in these elderly T2D patients compared in the present report (Janssen, 2009). We are uncertain whether this was
to those b 60 years old (Table 1). The proportion of men and Chinese due to successful modification of lifestyle, medications (metformin),
488 B.H. Chew et al. / Experimental Gerontology 48 (2013) 485–491

60% 30 *
26.3 ≥ 60 years
50.4 ≥ 60 years *
50% 45.6 25 < 60 years
< 60 years 22
39.6 38.2
40% 20
17.2 *
15.5 14.9
30% 15 *
* 11.7
* 10.6
9.3 *
20% 10 8.4
6.9 6.8
13.1
10.3 3.7
* 4.5
10% 5 3.4
2.2
1.8 1.0 0.8
0% 0
No Treatment One Treatment Two Treatment Three Treatment
Target Achieved Target Targets Targets

Treatment targets: HbA1c ≤ 6.5%, blood pressure < 130/80 mmHg and
low density lipoprotein-cholesterol (LDL-C) ≤ 2.6 mmol/L.
*Fisher Exact’s Test p < 0.0001
TIA= transient ischaemic attack, IHD= ischaemic heart disease, DFP= diabetic foot problems.
Fig. 1. Percentages of treatment targets achieved for HbA1c, blood pressure and LDL-C Any= any diabetes complication (stroke and/or IHD and/or retinopathy and/or nephropathy and/or DFP).
Macrovascular= macroavascular complications (stroke and/or IHD).
by Age groups. Treatment targets: HbA1c ≤ 6.5%, blood pressure b 130/80 mm Hg and
Microvascular= microvascular complications (retinopathy and/or nephropathy and/or DFP).
low density lipoprotein-cholesterol (LDL-C) ≤ 2.6 mmol/L.

co-morbidities, aging or mental disorders. A real-world study of diabetes
therapy (two-thirds of the cohorts aged >60 years) saw significant
weight gain over five years of follow-ups as well as doubling of insulin
therapy (from 14% to 32%) comparing the insulin therapy rate of about
10% in our cohorts (Best et al., 2012; Ismail et al., 2011).
4.1. Disease control
The elderly patients in our study achieved targets of glycemic and
lipid control but not blood pressure target of b130/80 mm Hg. The
proportion of our older patients achieving HbA1c and LDL targets were
comparable to the Asian countries (Hong Kong, India, Korea, Philippines,
Singapore, Taiwan, and Thailand) in the JADE (Joint Asia Diabetes Evalu-
ation) program whose cohorts had a median age of 58 years and slight
difference in the targets setting (35.3% achieved HbA1c b 7% and 34%
achieved LDL b 2.6 mmol/L) (So et al., 2011). However, the propor-
tions that achieved two and three main treatment targets (HbA1c,
BP and LDL-C) were better there. The attenuated association between
Fig. 2. Percentages of diabetes complications in each age group. *Fisher Exact's Test
p b 0.0001. TIA = transient ischaemic attack, IHD = ischaemic heart disease, DFP =
diabetic foot problems. Any = any diabetes complication (stroke and/or IHD and/or
retinopathy and/or nephropathy and/or DFP). Macrovascular = macroavascular com-
plications (stroke and/or IHD). Microvascular = microvascular complications (reti-
nopathy and/or nephropathy and/or DFP).
cardiometabolic risk markers and older people was also reported in the
United States and these Asian countries (Janssen, 2009; So et al.,
2011). Inverse relationship of glycemic control (HbA1c ≤ 7.5%) and age
among the T2D patients was also reported in a primary care setting in
the U.K. (Nagrebetsky et al., 2012). We believe this relationship of older
age and uncontrolled BP would dwindle if higher BP level was accepted
as the target of control as suggested by some recent studies (Mazza et
al., 2012 [In Press]). A Japanese study had reported a normal mortality
in their elderly diabetes patients who achieved HbA1c b 7% and
BP b 145/80 mm Hg as the treatment targets (Katakura et al., 2007).
Table 3
Multiple logistic regression with enter method conducted for diabetes complications
adjusting for patient demographic and treatment targets.

B SE p value Exp(B)a 95% C.I. for

Exp(B)
Table 2 Lower Upper
Multiple logistic regression with enter method conducted for treatment targets
adjusting for patient demographic and clinical characteristic. Any diabetes complications (1), n = 53,393
≥60 years 0.30 0.02 b0.0001 1.35 1.29 1.41
B SE p value Exp(B)a 95% C.I. for
Exp(B) Macrovascular complications (stroke or IHD) (1), n = 53,393
≥60 years 0.66 0.04 b0.0001 1.94 1.78 2.11
Lower Upper
Stroke (1), n = 41,421
HbA1c ≤ 6.5% (1), n = 37,263 ≥60 years 0.58 0.09 b0.0001 1.79 1.49 2.14
≥60 years 0.58 0.03 b0.0001 1.79 1.70 1.90 IHD (1), n = 38,768
≥60 years 0.68 0.05 b0.0001 1.97 1.79 2.16
Blood pressure b 130/80 mm Hg (1), n = 56,503
≥60 years −0.24 0.02 b0.0001 0.80 0.76 0.82 Microvascular complications (retinopathy or nephropathy or foot problem) (1),
n = 53,393
LDL-C ≤ 2.6 mmol/L (1), n = 38,848 ≥60 years 0.20 0.03 b0.0001 1.22 1.16 1.28
≥60 years 0.14 0.02 b0.0001 1.15 1.10 1.21 Retinopathy (1), n = 32,663
≥60 years 0.32 0.04 b0.0001 1.38 1.27 1.49
HDL-C ≥ 1.1 mmol/L (1), n = 39,277 Nephropathy (1), n = 39,161
≥60 years 0.19 0.02 b0.0001 1.21 1.15 1.26 ≥60 years 0.11 0.03 0.001 1.12 1.05 1.19
Diabetes foot problems (1),
TG ≤ 1.7 mmol/L (1), n = 45,717 n = 42,755
≥60 years 0.18 0.02 b0.0001 1.20 1.15 1.25 ≥60 years 0.26 0.04 b0.0001 1.30 1.19 1.41

SE = standard error, df = degree of freedom. SE = standard error, df = degree of freedom, IHD = ischaemic heart disease.
a
LDL-C = low density lipoprotein-cholesterol, HDL-C = high density lipoprotein- Adjusted for variables: diabetes duration, gender, ethnicity, body mass index,
cholesterol, TG = triglyceride. HbA1c > 6.5%, blood pressure ≥ 130/80 mm Hg, low density lipoprotein-cholesterol
a
Adjusted for variables: diabetes duration, gender, ethnicity, body mass (LDL-C) > 2.6 mmol/L, high density lipoprotein-cholesterol (HDL-C) b 1.1 mmol/L,
index ≥ 23 kg m 2. triglyceride > 1.7 mmol/L, anti-platelet, ACEI/ARB and statin.
 B.H. Chew et al. / Experimental Gerontology 48 (2013) 485–491
Poorer disease control for patients with newly diagnosed diabetes
was likely in our cohorts and similar observation had been reported
elsewhere (Edelman et al., 2003). More effort is warranted to achieve
treatment targets for the younger T2D patients who may exhibit
more severe metabolic deterioration and complications if left uncon-
trolled (Hubert et al., 2008; Jaakko, 2004). Early and intensive treat-
ment was recommended in T2D patients with a shorter duration of
disease and without a history of CVD (George, 2011). Past evidence
had indicated that risk of complications could be reduced if diseases
were controlled to targets early after the diagnosis of diabetes (Soe
et al., 2011).
4.2. Diabetes-related complications
Nephropathy was the most common diabetes-related complication
seen in our study. Cardiovascular–renal complications in T2D were
most prevalent in this region (So et al., 2011). Despite good control of
most cardiovascular risk factors in our present registry, the elderly
patients with T2D were still suffering from all types of diabetes
complications even after controlled for cardiovascular-protective
medications. Improved disease control and increased complications
were also observed in French elderly T2D (Pornet et al., 2011). In a
15-year follow-up Chinese multi-provincial study, diabetes and older
age (≥ 45 years) were predictors of coronary heart disease and
ischemic stroke despite well controlled LDL (Liu et al., 2012). Paradoxical
relationship between BMI-mortality, which means lower BMI (18.5 and
24.9 kg/m2) associated with higher mortality (Hazard Ratio 1.70 [95% CI
1.36–2.1]), was reported in African American and Caucasian men (N =
4156; mean age 60 ± 10.3 years) with T2D (Kokkinos et al., 2012).
4.3. Implications on diabetes care
Past evidence has indicated that diabetes mellitus could shorten life
expectancy by about a decade (Franco et al., 2007; Turin et al., 2012)
the elderly cohorts in our study had less than ten years to live based on
the average life expectancy in this country (men 72 years, women
77 years) (Ministry of Health Malaysia, 2011). With age as the
non-modifiable risk factor, health care providers in this country should
emphasize improving quality of life in the elderly patients with T2D rath-
er than aggressive pharmacological therapies to reach the treatment tar-
gets leading to more side effects e.g. hypoglycemia (Sheu et al., 2012).
About two-thirds (61.8%) of the elderly patients with T2D studied at
two rural public health clinics in this country were reported to
experience hypoglycemia in the past year (Ooi et al., 2011). Both mild
and severe symptomatic hypoglycemia have been reported to increase
cardiovascular events, all-cause hospitalization and death up to 2–3
times (Hsu et al., in press; Zoungas et al., 2010). Consequently, a
patient-centered approach with the overriding goal of improving glyce-
mic control while minimizing side effects is now recommended
(Inzucchi et al., 2012). Hence, quality of life as a treatment outcome is de-
sirable with continuing proper choice of pharmacological agents to reach
reasonable treatment targets and having a statin on-board regardless of
baseline cholesterol levels (Fernando, 2011; Jaakko, 2004). Standard of
medical care should be present for these elderly T2D patients with full
support for self-management that is in accordance to their cognitive func-
tion (Punthakee et al., 2012). Better quality of life and self-management is
not impossible in these elderly patients with more chronic complications
(Lu et al., 2006). Doctor–patient communication that leads to better
treatment strategy, patient-centered interventions and active coping be-
havior could prove to be cost-effective, leading to higher quality of life
and improved metabolic control (Huang et al., 2006). More days free of
symptoms and side effects are critical for the elderly with multiple
comorbidities to realize near-normal daily life experience (Wexler et al.,
2006). To achieve these, the community-based and multi-discipline ap-
proach of a primary healthcare clinic is preferred in catering diabetes
care to this group of patients (Browning et al., 2011). However, the facility
489
structure and staff-preparedness at the primary health-care clinics are
crucial for these functions (Watkins et al., 2000). It was reported that
public health-care clinics could be empowered to provide cognitive and
risk assessments, podiatrist care, regular fundus monitoring and early
management of renal replacement therapy and patient education
(Cochran and Conn, 2008; Joy, 2008).
4.4. Limitations
There are some general limitations in a large, population-based
registry study such as this. The proportion of unknown status for
many complications was rather large, especially for retinopathy, ne-
phropathy and IHD. This could well reflect the selective screening
strategies employed within the resource constraint primary health
care or suboptimal inter-professional communication between the
primary care physicians and hospital-based specialists. These screening
rates were rather similar to those in a French study whereby 42% had
missing data on albuminuria (Pornet et al., 2011). However, we have
reason to believe there was no selective screening for the elderly
group ≥60 years based on disease control, and there were about
equal proportion of unknown status/missing data for both age groups.
Owing to the cross-sectional design of this study, we could not ascertain
the sequence of events between onset of the diabetes-related complica-
tions and disease control. Nevertheless, it was possible that good
disease control was only achieved with improved therapy and adher-
ence after impact of complications. Unobservable confounders, beside
those already mentioned, may still exist and bias the results as
evidenced by the modest effect of the models. These other risk factors
not captured in the registry that could contribute to the odds of compli-
cations were socioeconomic status, smoking status, physical activity,
alcohol consumption, health literacy, professional support, mental dis-
order, etc. Further study could examine the possible effect of hypoglyce-
mia in this subgroup of elderly patients as cardiovascular outcomes
were also determined by severe hypoglycemia as a result of intensive
treatment. The U-shaped relationship of HbA1c and risk of diabetes
complications and death as were reported at both the cut-offs of
b6.0% and ≥8.0% (Huang et al., 2011). Therefore, findings should be
interpreted as associations instead of causations. Findings were based
on a single integrated health system and small proportion of the total
T2D patients in Malaysia, and thus may not be generalizable to larger
populations in the whole country. This registry could be improved
with inclusion of more potential risk factors and psychosocial outcomes.
5. Conclusions
We conclude that age ≥60 years was an independent risk factor for
diabetes-related complications despite good control of cardiovascular
risk factors. Thus, T2D in the older patients was a significant cause of
morbidity in this country. It was very likely that good disease control
was only achieved after the onset of complications. The sequence of
onset of complications and disease control, the effect of uncontrolled
hypertension and hypoglycemia, however, requires further study. Our
findings should provide caution on the current recommended clinical
targets control in these patients who were already having more
longstanding diseases and complications. The goals of therapy have to
be re-strategized for safer and more holistic diabetes care including
having quality of life as one of the aims of treatment for elderly T2D.
Author contributions
All authors made (1) substantial contributions to conception and
design, acquisition of data or analysis and interpretation of data,
(2) drafting the article or revising it critically for important intellectual
content and (3) gave final approval of the version to be published.
490 B.H. Chew et al. / Experimental Gerontology 48 (2013) 485–491
Grants
This study received its funding from Ministry of Health, Malaysia.
Conflict of interest
None.
Acknowledgments
We would like to acknowledge the Director General of Health
Malaysia for his support in the registry and his permission to publish
this report. We are grateful to the following steering committee mem-
bers of ADCM who had contributed tremendously to the success of the
registry: Professor Wan Mohamed Wan Bebekar, Dr Zanariah Hussein,
Dr G.R. Letchumanan a/k Ramanathan, Dr Rozita Zakariah, Professor
Fatimah Harun, Dr Cheong Ai Theng, Dr Lee Ping Yein, Dr Sri Wahyu
Taher, Dr Syed Alwi Syed Abd. Rahman, Dr Asmah Zainal Abidin and
Dr Nafiza Mat Nasir. We would like to thank Noor Akma Hassim and
Tee Chin Kim who of Clinical Research Centre, Kuala Lumpur for all
their statistic and logistic support. Special thanks also go to Ms. Lena
Yeap of ClinResearch Sdn Bhd for her team's previous statistical support.
References
Al-Qazaz, H.K., Hassali, M.A., Shafie, A.A., Syed Sulaiman, S.A., Sundram, S., 2011.
Perception and knowledge of patients with type 2 diabetes in Malaysia about
their disease and medication: a qualitative study. Res. Social Adm. Pharm. 7,
180–191.
Awin, N., 2001. A Review of Primary Health Care in Malaysia — A Report for the World
Health Organization Western Pacific Region. Ministry of Health, Malaysia.
Bachschmid, M.M., Schildknecht, S., Matsui, R., Zee, R., Haeussler, D., A. Cohen, R., Pimental,
D., Loo, B.v.d., in press. Vascular aging: Chronic oxidative stress and impairment of
redox signaling—consequences for vascular homeostasis and disease. Ann. Med. 45,
17–36.
Best, J.D., Drury, P.L., Davis, T.M.E., Taskinen, M.-R., Kesäniemi, Y.A., Scott, R., Pardy, C.,
Voysey, M., Keech, A.C., on behalf of the Fenofibrate Intervention and Event
Lowering in Diabetes Study, I, 2012. Glycemic Control Over 5 Years in 4,900 People
with Type 2 Diabetes. Diabetes Care 35, 1165–1170.
Browning, C., Chapman, A., Cowlishaw, S., Li, Z., Thomas, S., Yang, H., Zhang, T., 2011.
The Happy Life ClubTM study protocol: a cluster randomised controlled trial of a
type 2 diabetes health coach intervention. BMC Publ. Health 11, 90.
Chamnan, P., Simmons, R., Sharp, S., Griffin, S., Wareham, N., 2009. Cardiovascular risk
assessment scores for people with diabetes: a systematic review. Diabetologia 52,
2001–2014.
Chan, J.C.N., Malik, V., Jia, W., Kadowaki, T., Yajnik, C.S., Yoon, K.-H., Hu, F.B., 2009. Diabetes
in Asia. JAMA 301, 2129–2140.
Chew, B.H., Ismail, M., Lee, P.Y., Taher, S.W., Haniff, J., Mustapha, F.I., Bujang, M.A., 2012a.
Determinants of uncontrolled dyslipidaemia among adult type 2 diabetes in Malaysia:
the Malaysian Diabetes Registry 2009. Diabetes Res. Clin. Pract. 96, 339–347.
Chew, B.H., Ismail, M., Shariff-Ghazali, S., Lee, P.Y., Cheong, A.T., Ahmad, Z., Taher, S.W.,
Haniff, J., Mustapha, F.I., Bujang, M.A., 2012b. Determinants of uncontrolled hyperten-
sion in adult type 2 diabetes mellitus: an analysis of the Malaysian diabetes registry
2009. Cardiovasc. Diabetol. 11, 54.
Chew, B.H., Cheong, A.T., Ahmad, Z., Ismail, M., 2012c. Men were suffering more
complications of diabetes despite similar glycaemic control and better cardiovas-
cular risk profile: the ADCM study 2008. J. Mens Health 9, 190–197.
Cigolle, C.T., Lee, P.G., Langa, K.M., Lee, Y.Y., Tian, Z., Blaum, C.S., 2011. Geriatric conditions
develop in middle-aged adults with diabetes. J. Gen. Intern. Med. 26, 272–279.
Cochran, J., Conn, V.S., 2008. Meta-analysis of quality of life outcomes following diabetes
self-management training. Diabetes Educ. 34, 815–823.
Cooney, M.T., Cooney, H.C., Dudina, A., Graham, I.M., 2011. Total cardiovascular disease
risk assessment: a review. Curr. Opin. Cardiol. 26, 429–437.
de Craen, A.J., Oleksik, A.M., Maier, A.B., Westendorp, R.G., 2009. Causes of health and
disease in old age: new insights from the Leiden Research Program on Ageing.
Tijdschr. Gerontol. Geriatr. 40, 237–243.
Department of Statistics, 2010. Population Distribution and Basic Demographic Characteris-
tics 2010. Department of Statistics, Malaysia (http://www.statistics.gov.my/
portal/download_Population/files/census2010/ Taburan_ Penduduk_dan_Ciri-
ciri_Asas_Demografi.pdf).
Edelman, D., Olsen, M.K., Dudley, T.K., Harris, A.C., Oddone, E.Z., 2003. Quality of care
for patients diagnosed with diabetes at screening. Diabetes Care 26, 367–371.
Fernando, O., 2011. Cardiovascular implications of antihyperglycemic therapies for
type 2 diabetes. Clin. Ther. 33, 393–407.
Franco, O.H., Steyerberg, E.W., Hu, F.B., Mackenbach, J., Nusselder, W., 2007. Associations of
diabetes mellitus with total life expectancy and life expectancy with and without
cardiovascular disease. Arch. Intern. Med. 167, 1145–1151.
George, D., 2011. Early and intensive therapy for management of hyperglycemia and
cardiovascular risk factors in patients with type 2 diabetes. Clin. Ter. 33, 665–678.
Girndt, M., Seibert, E., 2010. Premature cardiovascular disease in chronic renal failure
(CRF): a model for an advanced ageing process. Exp. Gerontol. 45, 797–800.
Health Technology Assessment, 2008. U. Clinical practice guidelines — management of
hypertension. Expert review of cardiovascular therapy. Ministry of Health, Malaysia,
Putrajaya.
Hsu, P.F., Sung, H.S., Cheng, H.M., Yeh, J.S., Liu,W.L., Chan,W.L., Chen, C.H., Chou, P., Chuang,
S.Y., in press. Association of clinical symptomatic hypoglycemia with cardiovascular
events and total mortality in type 2 diabetes mellitus: a nationwide population-based
study. Diabetes Care (Electronic publication ahead of print).
Huang, E.S., Shook, M., Jin, L., Chin, M.H., Meltzer, D.O., 2006. The impact of patient
preferences on the cost-effectiveness of intensive glucose control in older patients
with new-onset diabetes. Diabetes Care 29, 259–264.
Huang, E.S., Liu, J.Y., Moffet, H.H., John, P.M., Karter, A.J., 2011. Glycemic control, complica-
tions, and death in older diabetic patients. Diabetes Care 34, 1329–1336.
Hubert, K., Berthold, S., Lutz, H., Tim, H., Volker, L., Jacques, K.T.T., Christian, W., Werner, A.S.,
Stephan, M., 2008. Type 2 diabetes phenotype and progression is significantly different
if diagnosed before versus after 65 years of age. J. Diabetes Sci. Technol. 2, 82–90.
Institute for Public Health, 2008. The Third National Health and Morbidity Survey
(NHMS III) 2006, Diabetes. Ministry of Health, Malaysia, Malaysia.
Institute for Public Health (IPH), 2011. National health morbidity survey 2011 (NHMS
2011). Non-Communicable Diseases, vol.II (Putrajaya).
Inzucchi, S.E., Bergenstal, R.M., Buse, J.B., Diamant, M., Ferrannini, E., Nauck, M., Peters,
A.L., Tsapas, A., Wender, R., Matthews, D.R., American Diabetes Association (ADA),
European Association for the Study of Diabetes (EASD), 2012. Management of hyper-
glycemia in type 2 diabetes: a patient-centered approach: position statement of the
American Diabetes Association (ADA) and the European Association for the Study of
Diabetes (EASD). Diabetes Care 35, 1364–1379.
Ismail, M., Chew, B.H., Lee, P.Y., Cheong, A.T., Sazlina, S.G., Jamaiyah, H., Alwi, S.A.R.S., Wahyu,
T.S., Zaiton, A., 2011. Control and treatment profiles of 70,889 adult type 2 diabetes
mellitus patients in Malaysia. Int. J. Collab. Res. Internal Med. Public Health 3, 98–113.
Jaakko, T., 2004. Impact of age on cardiovascular risk: implications for cardiovascular
disease management. Atheroscler. Suppl. 5, 9–17.
Janssen, I., 2009. Influence of age on the relation between waist circumference and
cardiometabolic risk markers. Nutr. Metab. Cardiovasc. Dis. 19, 163–169.
Joy, S.V., 2008. Clinical pearls and strategies to optimize patient outcomes. Diabetes
Educ. 34, 54S–59S.
Katakura, M., Naka, M., Kondo, T., Komatsu, M., Yamauchi, K., Hashizume, K., Aizawa, T.,
2007. Normal mortality in the elderly with diabetes under strict glycemic and
blood pressure control: outcome of 6-year prospective study. Diabetes Res. Clin.
Pract. 78, 108–114.
Kokkinos, P., Myers, J., Faselis, C., Doumas, M., Kheirbek, R., Nylen, E., 2012. BMI–mortality
paradox and fitness in African American and Caucasian men with type 2 diabetes.
Diabetes Care 35, 1021–1027.
Lee, P.Y., Cheong, A.T., Zaiton, A., Mastura, I., Chew, B.H., Sazlina, S.G., Adam, B.M., Syed
Alwi, S.A.R., Jamaiyah, H., SriWahyu, T., in press. Does ethnicity contribute to the
control of cardiovascular risk factors among patients with type 2 diabetes? Asia
Pac. J. Public Health (Electronic publication ahead of print).
Liu, J., Wang, W., Wang, M., Sun, J., Liu, J., Li, Y., Qi, Y., Wu, Z., Zhao, D., 2012. Impact of
diabetes, high triglycerides and low HDL cholesterol on risk for ischemic cardiovas-
cular disease varies by LDL cholesterol level: a 15-year follow-up of the Chinese
multi-provincial cohort study. Diabetes Res. Clin. Pract. 96, 217–224.
Lu, K.-Y., Lin, P.-L., Tzeng, L.-C., Huang, K.-Y., Chang, L.-C., 2006. Effectiveness of case
management for community elderly with hypertension, diabetes mellitus, and
hypercholesterolemia in Taiwan: a record review. Int. J. Nurs. Stud. 43, 1001–1010.
Mazza, A., Ramazzina, E., Cuppini, S., Armigliato, M., Schiavon, L., Rossetti, C., Marzolo,
M., Santoro, G., Ravenni, R., Zuin, M., Zorzan, S., Rubello, D., Casiglia, E., 2012. Anti-
hypertensive treatment in the elderly and very elderly: “always the lower, the
better?”. Int. J. Hypertens. 590683.
Ministry of Health Malaysia, 2011. Health Facts 2010. Ministry of Health Malaysia, Putrajaya.
Muradian, K., Schachtschabel, D.O., 2001. The role of apoptosis in aging and age-related
disease: update. Z. Gerontol. Geriatr. 34, 441–446.
Nagrebetsky, A., Griffin, S., Kinmonth, A.L., Sutton, S., Craven, A., Farmer, A., 2012. Predictors
of suboptimal glycaemic control in type 2 diabetes patients: the role of medication
adherence and body mass index in the relationship between glycaemia and age.
Diabetes Res. Clin. Pract. 96, 119–128.
Nakao, Y.M., Teramukai, S., Tanaka, S., Yasuno, S., Fujimoto, A., Kasahara, M., Ueshima, K.,
Nakao, K., Hinotsu, S., Nakao, K., Kawakami, K., 2012. Effects of renin–angiotensin
system blockades on cardiovascular outcomes in patients with diabetes mellitus: a
systematic review and meta-analysis. Diabetes Res. Clin. Pract. 96, 68–75.
Navarro, A., Boveris, A., 2007. The mitochondrial energy transduction system and the
aging process. Am. J. Cell. Physiol. 292, C670–C686.
Ooi, C.P., Loke, S.C., Zaiton, A., Tengku-Aizan, H., Zaitun, Y., 2011. Cross-sectional study of
older adults with type 2 diabetes mellitus in two rural public primary healthcare facilities
in Malaysia. Med. J. Malaysia 66, 108–112.
Pornet, C., Bourdel-Marchasson, I., Lecomte, P., Eschwege, E., Romon, I., Fosse, S., Assogba, F.,
Roudier, C., Fagot-Campagna, A., ENTRED Scientific Committee, 2011. Trends in the
quality of care for elderly people with type 2 diabetes: the need for improvements in
safety and quality (the 2001 and 2007 ENTRED Surveys). Diabetes Metab. 37, 152–161.
Punthakee, Z., Miller, M.E., Launer, L.J., Williamson, J.D., Lazar, R.M., Cukierman-Yaffee, T.,
Seaquist, E.R., Ismail-Beigi, F., Sullivan, M.D., Lovato, L.C., Bergenstal, R.M., Gerstein,
H.C., for the, A.G.o.I.a.t.A.-M.I., 2012. Poor cognitive function and risk of severe hypo-
glycemia in type 2 diabetes. Diabetes Care 35, 787–793.
Sheu, W.H.H., Ji, L.-N., Nitiyanant, W., Baik, S.H., Yin, D., Mavros, P., Chan, S.-P., 2012.
Hypoglycemia is associated with increased worry and lower quality of life among
 B.H. Chew et al. / Experimental Gerontology 48 (2013) 485–491
patients with type 2 diabetes treated with oral antihyperglycemic agents in the
Asia-Pacific region. Diabetes Res. Clin. Pract. 96, 141–148.
Sho-ichi, Y., 2011. Role of Advanced Glycation End Products (AGEs) and Receptor for
AGEs (RAGE) in Vascular Damage in Diabetes 46, 217–224.
So, W.-Y., Raboca, J., Sobrepena, L., Yoon, K.-H., Deerochanawong, C., Ho, L.-T.,
Himathongkam, T., Tong, P., Lyubomirsky, G., Ko, G., Nan, H., Chan, J., for the, J.P.R.T.,
2011. Comprehensive risk assessments of diabetic patients from seven Asian coun-
tries: the Joint Asia Diabetes Evaluation (JADE) program. J. Diabetes 3, 109–118.
Soe, K., Sacerdote, A., Karam, J., Bahtiyar, G., 2011. Management of type 2 diabetes
mellitus in the elderly. Maturitas 70, 151–159.
Terry, D.F., Evans, J.C., Pencina, M.J., Murabito, J.M., Vasan, R.S., Wolf, P.A., Kelly-Hayes, M.,
Levy, D., D'Agostino Sr., R.B., Benjamin, E.J., 2007. Characteristics of Framingham off-
spring participants with long-lived parents. Arch. Intern. Med. 167, 438–444.
491
Turin, T.C., Murakami, Y., Miura, K., Rumana, N., Kadota, A., Ohkubo, T., Okamura, T.,
Okayama, A., Ueshima, H., 2012. Diabetes and life expectancy among Japanese—
NIPPON DATA80. Diabetes Res. Clin. Pract. 96, 18–22.
Watkins, K.W., Connell, C.M., Fitzgerald, J.T., Klem, L., Hickey, T., Ingersoll-Dayton, B., 2000.
Effect of adults' self-regulation of diabetes on quality-of-life outcomes. Diabetes Care
23, 1511–1515.
Wexler, D., Grant, R., Wittenberg, E., Bosch, J., Cagliero, E., Delahanty, L., Blais, M., Meigs, J.,
2006. Correlates of health-related quality of life in type 2 diabetes. Diabetologia 49,
1489–1497.
Zoungas, S., Patel, A., Chalmers, J., de Galan, B.E., Li, Q., Billot, L., Woodward, M.,
Ninomiya, T., Neal, B., MacMahon, S., Grobbee, D.E., Kengne, A.P., Marre, M.,
Heller, S., ADVANCE Collaborative Group, 2010. Severe hypoglycemia and risks of
vascular events and death. N Engl J Med 363, 1410–1418.