Content uploaded by Dileepa Ediriweera
Author content
All content in this area was uploaded by Dileepa Ediriweera on Nov 23, 2019
Content may be subject to copyright.
Original Article
High prevalence of Diabetes Mellitus in Sri Lankan urban population –
Data from Colombo Urban Study.
N. P. Somasundaram1, I. Ranathunga1, K. Gunawardana1, D. S. Ediriweera2
1Diabetes and Endocrine Unit, National Hospital of Sri Lanka, Colombo
2Biostatistics and Epidemiology, Faculty of Medicine, University of Kelaniya, Sri Lanka
Copyright: This is an open-access article distributed under the terms of the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original
author and source are credited (CC BY 4.0)
Abstract
Correspondence: e-mail< ish_75@yahoo.com >.
Background:
In recent decades, Sri Lanka has experienced rapid urbanization, with approximately 30% of the population currently residing
in urban areas. We report the age- and sex-specific prevalence of dysglycaemia in an urban population in Colombo, Sri Lanka.
Methods:
Using a stratified random sampling method, 463 subjects (139 men; 324 women) aged 18 years and above were included.
Physical activity was quantified using international physical activity questionnaire (IPAQ). Bio impedence was used to estimate
body fat. Insulin sensitivity was estimated using the HOMA calculations. Prevalence was estimated using weighted age
standardized calculations. Multiple logistic regression analyses were used to study associations to diabetes and prediabetes.
Results:
There were 124 adults in the 18-40 age group (70% female), 209 adults in the 41-60 age group (73% female) and 130 adults
in the > 60 age group (63% female). The overall prevalence of diabetes was 27.6% (95% CI: 23.7-31.4). The prevalence of
diabetes in those aged 18-40 was 12.4% (95% CI: 6.4 -18.4), 36.1% (95% CI: 29.8 – 42.4) in those aged 41 – 60 and 48.3%
(95% CI: 40.7 – 55.8) in those aged >60. Pre-diabetes was detected in 30.3% (95% CI 25.9-34.8) of the population (with either
an HbA1c of 5.7-6.4%, FPG of 110-125 mg/dl or 2 Hr PPG of 140-199 mg/dl). Cumulative prevalence of diabetes and pre-
diabetes in the population was 57.9%.
Conclusions:
This urban study demonstrates that along with the changes in the socio-demographic status, the metabolic profile of the Sri
Lankan adult has transformed, with a high prevalence of dysglycaemia and obesity.
Background:
Sri Lanka, a middle-income country in Asia with a
population of 20 million, has been experiencing rapid and
unplanned urbanization over the recent decades with an
estimated 30% of the population now living in urban and
suburban areas (1). The urban prevalence of Diabetes,
prediabetes and obesity has been rising exponentially
over the past three decades. A study in subjects aged over
20 years indicated a population prevalence of
dysglycaemia (defined as T2DM or IGT or IFG) of
21.8%, which rose to 30% in urban areas (2). Metabolic
syndrome was found in 27.1% of urban adults (3). Physical
inactivity, elevated body mass index (BMI) and central
obesity along with living in an urban area are thought to
be strongly associated with the increased risk of
dysglycaemia.
Methods:
The study population consisted of adult males and
females who were 18 years and above, whose permanent
residence was in the Eastern Kuppiyawaththa local
government (Grama-Niladhari) division of the Colombo
District. The sample represents an urban population
living in Colombo. This local government area was
selected for the community cohort as it is the closest to
the National Hospital of Sri Lanka, which is the main
research center. The study was carried out in 2014/2015.
Sample size
Sample size was calculated using the Lwanga and
Lameshow 1991 formula of n = z2 p (100-p)D/ d2.
Sample size of 600 was arrived at for expected prevalence
of dysglycaemia and obesity of 50%, design effect of 1.2
with a precision of 95% and an anticipated 25% non-
response rate using the EPI 6 sample calculation
software.
Sampling technique
Stratified simple random sampling was used to select a
sample of 463 from the total population of 6473 in the
GN area registered in who belonged in the age categories
of 18-40 years, 40-60 years and above 60 years. People
who are included in the voters list of Colombo district or
lives in Colombo district continuously for at least three
years were included in the sampling process. In order to
ensure the precision of the estimates in the subsample
analysis (according to the age groups) the sample was
divided among the 3 age categories on a weighted basis
that took into account the proportion in the population
and the expected prevalence of dysglycaemia.
Using a random number generator, study subjects were
randomly allocated into the three strata as follows. In the
18-40 age stratum 210 were selected (35% of total
sample) whilst in the 40-60 years stratum 240 were
selected (40% of the sample) and in the above 60 year
stratum 150 were selected (25% of the sample). The
resulting disproportionate sample allocation was
accounted for, by the use of weighted analysis. The
weights were the inversion of the sampling fractions in
the analysis.
Data collection
The participants were recruited at their homes by a team
of researchers who provided an invitation letter and
information documents. On the day of the screening,
informed written consent was taken and data was
collected using an interviewer-administered
questionnaire administered by trained interviewers, to
collect data including socio-demographic data, use of
alcohol, smoking, food frequency and physical activity,
and detailed medical history on previous diagnoses and
treatment. Anthropometric measurements were made
(weight, height, waist circumference, total body fat
estimation and visceral fat percentage using a bio
impedance analyzer- OMRON HBF 516). Blood samples
were taken in a nine to twelve hours fasting stage and in
non-diabetics 75g anhydrous glucose was given and
blood was collected for glucose measurement two hours
later. Plasma Glucose (GOD- PAP5 method, Olympus
AU 480/680/400 analyser), cholesterol (CHOD-PAP
method, Olympus AU 480/680/400 analyser),
triglyceride (GPO-PAP method, Olympus AU
480/680/400 analyser), glycosylated hemoglobin (HPLC
method, Bio-Rad Variant II Turbo analyser), serum
insulin (Chemiluminescent enzyme immunoassay,
Immulite 1000 analyser), corrected calcium (Arzenso III
method, Olympus AU 480/680/400 analyser) and 25-
OH Vitamin D level (Direct Chemiluminescence
method, Advia centaur analyser) were measured in the
blood samples. Once the serum insulin levels were
analyzed, insulin resistance and beta cell function were
calculated using the HOMA calculator (4).Diabetes
mellitus was diagnosed based on the ADA/WHO
criteria. This required either a documented prior
diagnosis of diabetes or a value above the diagnostic
threshold on biochemical testing. The cutoffs included a
FPG above 126mg/dl, 2 Hour PPG above 200 mg/dl, or
HBA1C above 6.5%. Pre-diabetes was diagnosed with
any of the following values: FPG of 110-125 mg/dl or 2
Hour PPG of 140-199 or an A1C of % 5.7-6.4%.
Statistical analysis:
Data analysis was performed in the R programming
language version 3.2.2 (5). Community based prevalence
rates and means with 95% confidence intervals for the
urban study population and for different strata including
age and gender were calculated considering the stratified
sampling methodology using the “Survey” package in the
R programming language (5). Age adjusted prevalence
rates were calculated based on direct standardization
method using the World Health Organization world
standard population. Descriptive data analysis was
carried out to describe study population characteristics.
Exploratory data analysis was done to identify the risk
factors associated with diabetes mellitus. Exposure
variables studied were age, gender, ethnicity, education
level, smoking habits, alcohol consuming habits, family
history of diabetes, hypertension and hyperlipidemia,
past medical history of diabetes, hypertension and
hyperlipidaemia, weight, height, body mass index (BMI),
waist circumference, neck circumference, body fat
percentage, visceral fat percentage, physical activity
which was quantified as metabolic equivalent of tasks
(METS minutes per week) based on International
Physical Activity questionnaire (6), food habits based on
one week food recall in the Food Frequency
Questionnaire and vitamin D levels. Initially, each study
variable was screened with Pearson’s Chi-square test and
simple logistic regression, and the variables significant at
P =0.2 level were subsequently used for multiple variable
analysis. Subsequently, multiple logistic regression was
carried out to investigate the factors associated with
diabetes status and stepwise selection method was
adopted to select significant variables. Ethnicity
consisted of 4 categories (i.e. Sinhalese, Tamils, Moors
and other), the “other” ethnicity had only 4 individuals
and this group was not considered in the analysis. P value
of 0.05 was considered as significant.
Ethical Issues:
Ethical approval was obtained from the Ethical Review
committee of the Faculty of Medicine, University of
Colombo. Documents were encoded to avoid any
identifying character and measurements were taken to
ensure confidentiality.
Results:
A total of 463 subjects gave informed consent and
completed the screening. Most of the respondents were
females (69%). There were 124 subjects in the 18-40 age
group and 70% of these were females. There were 209
subjects in the 41-60 age group and 73% of this stratum
were females. In the over 60-year age stratum there were
130 subjects and 63% were females. The response rate in
each of the above strata was 59%, 87%, and 87% with an
overall response rate of 77.2%. Table 1 summarizes the
basic characteristics of this study population.
This study population’s mean Body mass index was 25.2
kg/m2 (SD 4.8), mean waist circumference was 87.0 cm
(SD 13.0), mean neck circumference was 34.8 cm (SD
3.7), mean total body fat percentage estimated with bio
impedance analysis was 34.3 % (SD 8.3) and the
estimated visceral fat was 9.2% (SD 5.0). Body mass
index was categorized according to global criteria and the
recommended Asian and South Asian criteria (7, 8). 68.2%
of the women and 59.1% of the men were overweight or
obese based on South Asian criteria. The BMI
categorization and distribution are tabulated in Table 2.
Community prevalence for abdominal obesity was 58.1%
based on International Diabetes Federation cut-off
values on waist circumference for determining
abdominal obesity in South Asians (WC – male >=
90cm, female >=80cm) (9).
Prevalence of DM and Pre-diabetes
Estimated community mean fasting plasma glucose level
was 101.3 mg/dL (95% CI: 97.6 – 105.3), HBA1C level
was 6.3% (6.1 – 6.4) and 2 hour PPG in non-diabetic
individuals was 124 (119.7 – 129.2). It is notable that
these are relatively high and the mean HBA1C is in the
prediabetes range. These are tabulated in table 3. Family
history of Diabetes Mellitus in at least one first degree
relative was reported in 43.2 % (95% CI 38.4-48.2) of the
study population and 16.9% (95% CI 13.7-20.1) had
previously been diagnosed to have diabetes mellitus.
Estimated community prevalence of diabetes mellitus
was 27.6% (95% CI 23.7-31.4) and prediabetes was seen
in 30.3% (95% CI 25.9-34.8) of the population. Age
adjusted community prevalence of diabetes was 27.1%
and prediabetes was 30.1%. The community prevalence
of dysglycemia was 57.9%; this was seen in 36.2 % in the
18-40 age stratum, in 70.7% in the 41-60 stratum, and in
83.5% in the over 60 stratum. Thus, population
prevalence with normoglycaemia declined from 63.8% in
the 18-40 age stratum to 29.3% in the 41-60 age stratum
to 16.5% in the over 60 age stratum (Table 4).
Among ethnic groups, Moors had the highest prevalence
of Diabetes Mellitus of 36.1% (95%CI 25.5-46.7)
followed by Sinhalaese (30%, 95% CI 22.4-31.6) and
Tamils (19.4, 95% CI 8.8-29.9). Those with the lowest
educational background had the highest prevalence of
diabetes (39.1% (95% CI 28.2-50). Current tobacco
smokers had higher prevalence of diabetes mellitus
(50.1%, 95% CI 33.4-66.7) compared to those who never
smoked (25.2%, 95% CI 21.4-29.1). Ex consumers of
alcohol had the highest prevalence of diabetes mellitus
(52.9%, 95% CI 31-74.8) compared to non-consumers of
alcohol or current consumers (Table 5).
Hundred and two subjects (16.9%) had prior diagnosis of
diabetes mellitus. We further analyzed this subgroup who
were already diagnosed to have diabetes mellitus and
estimated level of control. Mean HbA1C in those who
had prior diagnosis of diabetes mellitus was 8.3% (95%
CI: 7.9 – 8.8) and mean HOMA ß was 47.6 (95% CI: 33.3
-61.8) indicating declining insulin reserve. HBA1C less
than 7% was found in 29.3% (19.7 – 38.9%), HBA1C
between 7% and 8 % in 29.5% (20.1. % - 38.8%), and
HBA1C above 8% was found in 41.2% (31.0% – 51.5%).
In the population previously diagnosed with diabetes,
blood pressure more than 130/90 mmHg was detected
in 33.4% (95% CI: 24.2 -42.6), LDL cholesterol above
100mg/dl was found in 58.1 (48.9 - 69.0), and triglyceride
above 150mg/dl was found in 41.6% (31.4 – 51.8).
Explorative analysis
Multiple variable analysis showed increasing age, family
history of diabetes, preexisting hypertension, increasing
BMI, increasing neck circumference, higher frequency of
consuming egg yolk and whole grain and less sweet
consumption had significant associations with diabetes.
Lifestyle factors such as level of physical activity or
amount of sitting time recorded with IPAQ did not
demonstrate any significant association with the presence
of diabetes in the analysis (Table 6).
Discussion
The incidence and the prevalence of diabetes mellitus is
a rapidly rising in Sri Lanka as well as globally. Urban
population is at a higher risk due to multiple predisposing
factors. This study was done to ascertain the true urban
prevalence of diabetes mellitus as increasing numbers of
diabetic patients living in urban areas are encountered in
clinical settings. There was an alarmingly high prevalence
of dysglycemia in the urban population studied. This is
of enormous clinical and economic significance as even
the younger population in the 18-40 age stratum had a
prevalence of diabetes mellitus of 12.4% and prediabetes
of 24.8% with potential to conversion to diabetes in the
near future. The high prevalence that has been shown in
our study is higher than urban prevalence of 18%
according to Katulanda et al nine years ago (2). It is
possible that the prevalence has actually increased,
however this study used HBA1C in addition to the FPG
and 2 Hour PPG used in the previous study and this may
explain part of the increase in prevalence. A previous
Colombo suburban study that used all three biochemical
parameters reported a prevalence of 20% (10).
Among the factors explored in this study; increasing age,
increasing BMI, increasing neck circumference and
presence of hypertension as well a family history of
diabetes mellitus in a first degree relative, high frequency
of consuming egg yolk, whole grain and less sweet
consumption are significant in multiple regression
analysis. Even though less whole grain consumption and
more sweet consumption are believed to be associated
with diabetes, our results showed the inverse. This need
to be carefully interpreted as already diagnosed diabetics
tend to eat less sweets and more whole grain as a diabetes
control measure. We also found that Vitamin D level was
not significantly associated with diabetes mellitus and the
results will be discussed in detail in another article.
Several key causative factors have not been explored in
this study; they include genetic and epigenetic factors as
well as foetomaternal environment, childhood feeding
and childhood exercise.
Conclusions
We have detected the highest reported prevalence of
diabetes mellitus in the South Asian region and these
prevalence rates are alarming. The existing pool of
patients with diabetes who are likely to develop
significant morbidity over time is a major policy and
health planning concern. The presence of a large number
of individuals who have prediabetes and can develop
diabetes in the future should prompt urgent nationwide
interventions as well as personalized interventions such
as dietary and exercise counselling. We have previously
reviewed possible public health interventions to prevent
diabetes and other non-communicable diseases in South
Asia (11). In light of the current findings, these
interventions may need to be targeted more towards the
above high risk groups in the urban population.
List of abbreviations
BMI - Body mass index
FPG-Fasting Plasma Glucose,
2 Hour PPG- Post prandial Glucose 2 hours
after 75 g glucose
LDL - Low-density lipoprotein cholesterol,
HOMA ß- Homeostasis Model Assessment
estimate of steady state beta cell function
HDL -High-density lipoprotein cholesterol
TG- Triglycerides
HBA1c-Haemoglobin A1c
TSH-Thyroid stimulating hormone,
WHO-World health organization
WC-Waist circumference
Competing interests:
Authors declare no conflict of interest
Ethics approval and consent to participate
Ethical approval was obtained from the Ethical Review
committee of the Faculty of Medicine, University of
Colombo. All participants who enrolled in the study
signed an informed consent form.
Consent for publication
Not applicable.
Availability of data and materials
The data analyzed in this paper can be made available to
researchers. Requests for access to the dataset used in this
paper should be directed to the corresponding author.
Competing interests
None of the authors have any financial or non-financial
competing interests to disclose.
Funding
Medical Research Institute, Colombo funded the project
Authors’ contributions
NPS and KG designed the study and were involved in
data collection. NPS, DSE, IR and KG were involved in
statistical analysis, interpretation of data and drafting the
manuscript. All authors read and approved the final
manuscript.
Acknowledgements:
None
Authors’ information
NPS is a Senior Consultant Endocrinologist at National
Hospital of Sri Lanka. IR is a Senior Registrar in
Endocrinology at National Hospital of Sri Lanka. KG is
Consultant Endocrinologist at National Hospital of Sri
Lanka. DSE is a Lecturer in Medical Informatics at
Faculty of Medicine, University of Kelaniya, Sri Lanka.
Both sexes
(N=463)
Males
(N=143)
Females
(N=320)
Mean age (SD) years
50.4 (14.8)
50.9 (15.8)
50.2 (14.3)
Mean height (SD) cm
153.3 (9.1)
165.0 (7.5)
152.4 (6.8)
Mean weight (SD) kg
61.5 (12.6)
65.7 (13.0)
59.7 (12.0)
Mean BMI (SD) kg/m2
25.2 (4.8)
24.1 (4.4)
25.7 (4.8)
Mean waist circumference cm
87.0 (13.0)
87.5 (12.9)
86.8 (13.0)
Mean neck circumference cm
34.8 (3.7)
36.3 (3.3)
33.0 (3.4)
Mean body fat %
34.3 (8.3)
26.5 (7.1)
37.8 (6.1)
Mean visceral fat %
9.2 (5.0)
9.8 (5.5)
8.9 (4.7)
Ethnicity
Sinhala
320 (69.1%)
108 (75.5%)
212 (66.2%)
Tamil
56 (12.1%)
12 (8.4%)
44 (13.8%)
Moor
83 (17.9%)
23 (16.1%)
60 (18.8%)
Other
4 (0.8%)
-
4 (1.2%)
Education
Below Grade 5
77 (16.7%)
10 (7.1%)
67 (20.6%)
Up to Ordinary Level
240 (51.9%)
77 (53.8%)
163 (51.1%)
Up to Advanced Level
127 (27.5%)
47 (32.9%)
80 (25.1%)
Above Advanced Level
18 (3.9%)
9 (6.2%)
9 (2.8%)
Tobacco smoking
No
396 (85.6%)
83 (58.0%)
313 (97.9%)
Underweight
Normal
Overweight
Obesity
Global BMI cut offs
7.7(4.9-10.4)
39.6(34.8-44.4)
37.0(32.2-41.8)
15.8(12.3-19.3)
Asian BMI cut offs
7.7(4.9-10.4)
26.8(22.4-31.2)
34.3(29.6-39.0)
31.2(26.7-35.8)
South Asian BMI cut offs
7.6 (4.9 – 10)
26.8 (22.4-31.2)
12.7 (9.6-15.9)
52.8 (47.8 – 57.7)
Table 1. Study group characteristics.
Table 2. Prevalence (%) of underweight, normal weight, over weight and obesity according to Global,
Asian and South Asian BMI categories
Both sexes
Males
Females
FBS
101.3 (97.6 – 105.3)
106.2 (97.1 – 115.4)
99.6 (95.4 – 103.7)
OGTT 2 hr*
124.4 (119.7 – 129.2)
122.9 (113.9 – 132.0)
125.1 (119.4 – 130.7)
HBA1C
6.3 (6.1 – 6.4)
6.4 (6.1 – 6.7)
6.2 (6.0 – 6.4)
Plasma Insulin
5.9 ( 5.2 – 6.7)
5.8 ( 4.9 – 6.7)
6.0 ( 5.0 – 7.1)
HOMA IR
1.6 ( 1.3 – 1.8)
1.5 (1.3 – 1.8)
1.6 (1.2 – 1.9)
HOMA
89.7 ( 77.3 – 102.0)
83.0 (68.9 – 97.2)
92.7 (75.9 – 109.4)
Family history of Diabetes
43.2 (38.4 – 48.2)
45.6 (36.7 – 54.5)
42.2 (36.3 – 48.1)
Past medical history of Diabetes
16.9 (13.7 – 20.1)
18.1 (11.9 – 24.2)
16.4 (12.6 – 20.2)
*In non-diabetic individuals
Normal
Prediabetes
Diabetes
Both sexes
Community
42.1 (37.5 – 46.8)
30.3 (25.9 – 34.8)
27.6 (23.7 – 31.4)
18-40
63.8 (54.1 – 71.6)
24.8 (16.9 – 32.6)
12.4 (6.4 -18.4)
41-60
29.3 (23.3 – 35.3)
34.6 (28.4 – 40.9)
36.1 (29.8 – 42.4)
60 >
16.6 (10.9 – 22.2)
35.2 (27.9 – 42.4)
48.3 (40.7 – 55.8)
Males
Community
40.3 (31.4 – 49.2)
27.2 (19.5 – 34.8)
32.5 (24.6 – 40.4)
18-40
61.1 (45.4 – 76.8)
19.4 (6.7 – 32.2)
19.4 (6.7 – 32.2)
41-60
25.4 (14.7 – 36.1)
33.9 (22.3 – 45.6)
40.7 (28.6 – 52.7)
60 >
16.7 (6.8 – 26.5)
33.3 (20.9 – 45.7)
50.0 (36.8 – 63.2)
Females
Community
43.0 (37.2 – 48.7)
31.8 (26.3 – 37.2)
25.3 (20.8 – 29.7)
18-40
63.6 (53.1 – 74.2)
27.3 (17.5 – 37.1)
9.1 (2.8 – 15.4)
41-60
30.8 (23.6 – 38.0)
34.9 (27.5 – 42.4)
34.2 (26.8 – 41.6)
60 >
16.5 (9.6 – 23.4)
36.1 (27.2 – 45.0)
47.4 (38.2 – 56.7)
Table 3. Estimated glycaemic indices (95% CI) for the study population
Table 4. Crude prevalence of diabetes mellitus and pre-diabetes
Normal
Prediabetes
Diabetes
Ethnicity
Sinhala
41.0 (35.2 – 46.8)
32.0 (26.5 – 37.5)
30.0 (22.4 – 31.6)
Tamil
60.0 (46.7 – 73.2)
20.7 (10.4 – 30.9)
19.4 (8.8 - 29.9)
Moor
32.3 (20.9 – 43.7)
31.6 (21.3 – 41.9)
36.1 (25.5 – 46.7)
Education
Below Grade 5
26.5 (23.6 – 45.2)
34.4 (15.2 – 37.7)
39.1 (28.2 – 50.0)
Upto Ordinary level
41.2 (25.5 – 37.9)
31.7 (34.5 – 47.9)
27.1 (21.5 – 32.6)
Upto advanced level
46.5 (20.4 – 37.4)
28.9 (37.1 – 55.9)
24.6 (17.4 – 31.9)
Above advanced level
69.5 (0.0 – 26.6)
10.8 (47.3 – 91.7)
19.7 (1.6 – 37.8)
Tobacco smoking
Never
44.0 (39.0 – 49.0)
30.7 (25.9 – 35.5)
25.2 (21.4 – 29.1)
Current smokers
33.0 (17.0 – 49.0)
16.9 (6.4 – 27.3)
50.1 (33.4 – 66.7)
Ex-smokers
25.7 (6.3 – 45.1)
45.5 (25.1 – 65.9)
28.8 (10.9 – 46.6)
Alcohol consumers
Never
43.9 (38.6 – 49.1)
30.2 (25.3 – 35.2)
25.9 (21.7 – 30.0)
Current consumers
38.6 (25.9 – 51.3)
31.7 (20.0 - 43.5)
29.7 (19.3 – 40.0)
Ex- consumers
19.5 (3.7 – 35.3)
27.6 (10.0 – 45.2)
52.9 (31.0 – 74.8)
Estimate
Std. Error
z value
Pr(>|z|)
Intercept
-9.04
1.479
-6.11
<0.001
Age
0.04
0.009
3.54
<0.001
Hypertension
0.92
0.260
3.54
<0.001
Family History of Diabetes
mellitus
0.96
0.236
4.08
<0.001
BMI
0.06
0.029
1.98
0.048
Neck circumference
0.10
0.039
2.46
0.014
Egg york
0.28
0.108
2.62
0.008
Whole grain
0.19
0.077
2.43
0.015
Sweets
-0.27
0.111
-2.46
0.014
Table 6. Significant variables at Multiple variable analysis for the presence of diabetes mellitus
Table 5. Prediabetes, diabetes based on categories of education, ethnicity, tobacco, and alcohol
References:
1. Department of Census and Statistics Census Report - 2012, Ministry of Finance and Planning, Sri Lanka; 2012
2. Katulanda P, Constantine GR, Mahesh JG, Sheriff R, Seneviratne RDA, Wijeratne S, Wijesuriya M, McCarthy MI, Adler
AI, Matthews DR: Prevalence and projections of diabetes and pre-diabetes in adults in Sri Lanka--Sri Lanka Diabetes.
Diabetic Medicine 2008;25:1062-9.
3. Katulanda P, Ranasinghe P, Jayawardana R, Sheriff R, Matthews DR. Metabolic syndrome among Sri Lankan adults:
prevalence, patterns and correlates. Diabetol Metab Syndr.2012 May 31;4(1):24.
4. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC (1985). "Homeostasis model assessment:
insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man.". Diabetologia
28 (7): 412–19
5. R Core Team (2016). R: A language and environment for statistical computing. R Foundation for Statistical Computing,
Vienna, Austria. URL http://www.R-project.org/
6. Craig, C. L, Marshall AL, Sjostrom M, Bauman AE, Booth ML, Ainsworth BE et al. (2003). "International physical
activity questionnaire: 12-country reliability and validity." Med Sci Sports Exerc 35: 1381-95
7. World Health Organization. Appropriate body-mass index for Asian populations and its implications for policy and
intervention strategies. Lancet. 2004; 363:157–63
8. Somasundaram N, Rajaratnam H, Wijeyarathne C et al. Clinical guidelines: The Endocrine Society of Sri Lanka:
Management of Obesity. Sri Lanka Journal of Diabetes, Endocrinology and Metabolism 2014; 1: 55-70.
9. Alberti KGM.M, Zimmet PZ, Shaw JE. The metabolic syndrome-a new world-wide definition from the International
Diabetes Federation consensus. Lancet. 2005; 366:1059-62.
10. Pinidiyapathirage MJ, Kasturiratne A, Ranawaka UK, et al. The burden of diabetes mellitus and impaired fasting
glucose in an urban population of Sri Lanka. Diabetic Medicine. 2013;30(3):326-332.
11. Somasundaram NP, Kalupahana NS. Population-based dietary approaches for the prevention of noncommunicable
diseases. WHO South-East Asia J Public Health 2016; 5(1): 22–26
