Key words: diabetes mellitus, risk factors, rural those in the 20-30 age group. Gender and religion did
population, North Kerala, India not show any statistically significant association with
diabetes. Physical activity was observed as a
SUMMARY protective factor for the development of DM.
Hypertension, especially systolic hypertension,
The aim of the study was to assess the risk factors
emerged as a strong risk factor for T2DM in this study.
associated with type 2 diabetes mellitus (T2DM) in
rural population of North Kerala, India. The study Subjects with systolic hypertension had 4.6-fold
included 100 T2DM cases randomly selected among chance to develop T2DM, making it mandatory to
patients with diabetes admitted to medicine ward and screen all patients with hypertension above 25 years of
200 controls without DM recruited from visitors and age for T2DM irrespective of the presence of other risk
patient attendants at a tertiary care centre in the factors. In conclusion, results of the present study will
northern part of Kerala, India. A questionnaire that be of use in planning primordial, primary and
contained sociodemographic characteristics and risk
secondary measures of prevention at the community
factors was used for data collection. ANOVA was
level.
performed to find the significance of more than two
means. Simple binary logistic regression and multiple
binary logistic regression were performed to find the INTRODUCTION
crude and adjusted odds ratio (OR) and 95%
confidence interval (CI) was calculated to find the Type 2 diabetes mellitus (T2DM) is the commonest
significance of the observed OR. A p value <0.05 was form of diabetes affecting more than 90% of the
considered statistically significant. Study results diabetic population worldwide. There is a rapid
showed those above 50 years of age to have five times upsurge in the number of diabetic patients and this
more chance to get diabetes when compared with
explosive growth is noted in both urban and rural
Corresponding author: Dr. Jayadevan Sreedharan, Assistant Director &
areas. Wild et al. estimated the number of T2DM
Professor, Research Division, Gulf Medical University, Ajman, United Arab patients in the year 2000 at 174 million and predicted
Emirates
E-mail: drjayadevans@gmail.com it to increase to 366 million in 2030 (1).
The majority of the patients with diabetes in One hundred cases with T2DM and 200 unmatched
developed countries are above age 64. It is predicted controls without DM were recruited for the study.
that by 2030, the number of people aged above 64 with Cases were randomly selected among patients with
diabetes will be around 82 million, of which about 48 diabetes admitted to the medicine ward and 200
million in developing countries. India has the largest controls among visitors and patient attendants. A case
diabetic population and it is expected to increase to control ratio of 1:2 was adopted in this study.
174 million in the year 2025 (2). In developing A pretested structured interviewer administered
countries, the majority of people with diabetes are in questionnaire was used for data collection. The
the 45-64 age group (3). questionnaire contained data pertaining to
Seventy percent of the Indian population live in rural sociodemographic characteristics and various risk
areas. It is a well-known fact that urban and rural factors associated with the occurrence of DM.
populations have different lifestyles, work patterns, Collected data were analyzed using SPSS 13 version
and environmental and sociocultural factors. The (IBM, Illinois, Chicago). Frequency, percentage, mean
presentation of T2DM is not uniform throughout the and standard deviation were calculated and
world and there are geographical and ethnic variations associations between variables were assessed using
in the presentation of T2DM across the world. Most chi-square test. Also, t-test was used to find significant
studies from western countries (4-6) and urban studies differences of two means and ANOVA was employed
from India point to lifestyle changes (7), sedentary life to find the significance of more than two means. In
(5), diet and related epidemiological transition (8) as addition to the above statistical tools, simple binary
the major factors in the development of DM. The risk logistic regression analysis and multiple binary
factors hitherto specified in the development of T2DM logistic regression were performed to find crude and
in the western and urban population cannot fully apply adjusted odds ratio (OR). The 95% confidence interval
to a rural setting. So, when there is a rapid upsurge of (CI) was calculated to find the significance of
T2DM in both urban and rural areas, it is an imperative observed OR. In all cases, p value <0.05 was
to identify the factors predisposing to the development considered statistically significant.
of the disease, which affects one out of every five
Indians. There are only limited studies on the subject
RESULTS
from India; of them, the most acclaimed studies are
from Chennai and other metropolitan cities in India.
Gender wise distribution of subjects showed that
The majority of the studies from India are prevalence 71.0% of cases and 59.0% of control subjects were
studies on DM and only very few studies have focused males. A statistically significant association was found
on the profile and risk factors of T2DM. The North between cases and controls with regard to gender
Malabar areas of Kerala state include Kannur, (p<0.05). With regard to age, 83.0% of the case group
Kasargod and Wayanad districts and most of the subjects and more than 56.0% of control group
population in this area have a traditional rural lifestyle. subjects were above age 40. A highly significant
Hence, this study was undertaken to determine the risk association was observed between cases and controls
factors of T2DM among rural population of North with regard to age (p<0.001). The majority of study
Kerala, India. subjects from both case and control groups belonged
to Hindu religious groups. There was no statistically
MATERIALS AND METHODS significant association between diabetes status and
religion. When the participants were compared
This study was conducted at the Academy of Medical according to occupation, it was found that 29% of
Sciences, Pariyaram, which is a Research Centre and control subjects and 25% of cases were homemakers,
Postgraduate Institute situated in the Kannur District whereas 8% of control subjects and 7% of cases were
of North Malabar area in Kerala, India. manual laborers. Other occupational groups included
34
B. Valliyot, J. Sreedharan, J. Muttappallymyalil, S. Balakrishnan Valliyot / RISK FACTORS OF TYPE 2 DIABETES
professionals, mechanics, etc. A highly significant 88.5% of control group subjects and 82% of case
statistical association was observed between diabetes group subjects had a diastolic blood pressure of less
status and occupation (p<0.001). than 90 mm Hg. There was no statistically significant
Dietary pattern showed that among controls 93% association between diastolic blood pressure and
were having mixed diet and 7% were vegetarians, diabetic status. Details are given in Table 1.
whereas among cases 89% were mixed diet consumers Table 2 shows that the mean age was lower in the
and 11% were on vegetarian food. When further case group than in the control group. The mean height
analyzed with chi-square test, dietary pattern did not in both cases and controls was found to be equal.
show any significant association. It was observed that
Difference in the mean weight was not statistically
38% of the cases and 18.5% of the controls were
significant. Age, gender, occupation, physical activity,
tobacco users. When tested with chi-square test,
family history, diet, tobacco use, and systolic blood
tobacco use showed a statistically significant
pressure showed statistically significant correlation
association (p<0.001).
with diabetic status level. The above parameters that
Assessment of self-reported physical activity showed revealed significance were included in the simple
that 42% of cases and 14.5% of control group subjects binary logistic regression model and then in the
were involved in heavy work. Among cases and multiple logistic regression model. The OR and its
controls, 39% and 18% were involved in moderate
confidence interval are shown in Table 3.
physical activity, respectively, whereas 19% of cases
and 67.5% of controls were involved in sedentary Age was found to be a significant factor. In
activity. The association between physical activity comparison with the 20-29 age group, the 40-49 age
with regard to diabetes status was found to be group had a 4.7-fold and 50-55 age group 5.5-fold
statistically significant (p<0.01). likelihood of developing DM. On adjustment for all
With regard to family history of DM, 55.0% of cases other factors, gender was not found to yield
and 37.5% of controls had a family history of DM. The statistically significance. Physical activity was
association observed was high (p <0.001). A detailed measured according to their type of work and was
study of the family history of DM showed that in 24% divided into minimal, moderate and hard physical
of cases and 11.5% of controls mothers were diabetic, activity. For those involved in doing hard activity, the
whereas fathers were diabetic in 10.5% of controls and chance of getting diabetic was by 89% less when
7% of cases. compared to those doing minimal activity, which was
When compared according to body mass index statistically significant. Family history was found to be
(BMI) less than 18.5 kg/m2, both cases and controls an important risk factor with a p value of 0.001.
were equally distributed, whereas 67% of cases and Multiple logistic regression analysis showed an
62% of controls had BMI in the range of 18.5-25 adjusted OR of 3.09. Those with a family history of
kg/m2 .Overweight was recorded in 25% of cases and DM had 3.09-fold greater chance of getting the disease
29% of control subjects. BMI greater than 30 kg/m2 as compared to those without a family history of DM.
was found in only 3% of cases and 4.5% of controls. Tobacco use appeared as a significant risk factor for
There was no statistically significant association the occurrence of DM. The adjusted OR was 2.49,
between BMI and diabetic status. which was statistically significant. Systolic blood
Systolic blood pressure of less than 140 mm Hg was pressure was another risk factor for the development
measured in 93.5% of control subjects and 61% of of DM. After adjusting for all other factors, the OR
cases. When this observation was tested with Pearson observed was 4.69 with a CI of 2.13-10.40. So, age,
chi-square test, systolic blood pressure was found to family history, physical activity, tobacco use and
be a significant factor associated with diabetes status systolic hypertension emerged as significant
(p<0.001). With regard to diastolic blood pressure, independent risk factors for the occurrence of DM.
Table 1. Sociodemographic characteristics and other correlates between cases and controls
Table 2. Distribution of cases and controls according to age, height, weight, body mass index (BMI) and blood
pressure (mean ± standard deviation, SD)
Cases Control
Factor p value
Mean SD Mean SD
Age (yrs) 41.07 9.03 45.63 7.88 p<0.001
Height (cm) 162.71 8.10 162.94 7.6 NS
Weight (kg) 63.56 10.94 62.03 9.28 NS
BMI (kg/m2) 23.98 3.31 23.39 3.09 NS
Systolic blood pressure 125.93 13.51 141.6 19.44 p<0.001
Diastolic blood pressure 81.44 8.94 85.44 10.53 NS
36
B. Valliyot, J. Sreedharan, J. Muttappallymyalil, S. Balakrishnan Valliyot / RISK FACTORS OF TYPE 2 DIABETES
38
B. Valliyot, J. Sreedharan, J. Muttappallymyalil, S. Balakrishnan Valliyot / RISK FACTORS OF TYPE 2 DIABETES
REFERENCES
1. Wild SH, Roglic G, Green A, Sicree R, King H. 11. Ramachandran C, Snehalatha, Kapur A, Vijay V,
Global prevalence of diabetes: estimate for year Mohan V, Das AK, et al. High prevalence of
2000 and projections for 2030. Diabetes Care diabetes and impaired glucose tolerance in India:
2004;27(5):1047-1052. National Urban Diabetes Survey (NUDS).
Diabetologia 2001;44(9):1094-1101.
2. Diabetes atlas, Executive summary, World
Diabetes Foundation, second Edition, 2003. 12. Mohan V, Santhirani CS, Deepa R. Prevalence of
diabetes and IGT in selected south Indian
3. Sicree R, Shaw JE, Zimmet PZ. Global burden of population with special reference to family history,
diabetes. Diabetes atlas, 2nd edn. Brussels: obesity and lifestyle factor – Chennai Urban
International Diabetes Federation, 2003:15-71. Population Study [CUPS-14]. J Assoc Physicians
India 2003;51:771-777.
4. Ziv E, Shaffir E. Psammomys obesus: nutritionally
induced NIDDM-like syndrome on thrifty gene 13. Kokiwar PR, Gupta S, Durge PM. Prevalence of
background. In: Shafrir E, ed. Lessons from diabetes in a rural area of central India. Int J
animal diabetes. London: Smith-Gordon, Diabetes Dev Countries 2007;27(1):8-10.
1995;285-300. 14. Gill JMR, Cooper AR. Physical activity and
prevention of type 2 diabetes mellitus. Curr Opin
5. Heimlich SP, Ragland DR, Lecins RQW. Physical
Sports Med 2008;38(10):807-824.
activity and reduced occurrence of non insulin
dependent diabetes. N Engl J Med 1991;325:147- 15. Karter AJ, Rowell SE, Ackerson LM, Mitchell BD,
152. Ferrara A, Selby JV, et al. Excess maternal
transmission of type 2 diabetes. The Northern
6. Mooy JM, DeVries H, Grootenhuis PA, Bouter California Kaisser Permanent Diabetes Registry.
LM, Hein Major RJ. Stressful life events in Diabetes Care 1999;22:938-943.
relation to prevalence and undetected type 2
16. Shashank RJ, Rakesh MP. Family history and
diabetes. Diabetes Care 2002;15(3):197-201.
pedigree charting – a simple genetic tool for Indian
7. Arunachalam S, Guneshekharan S. Diabetes diabetes. JAPI 2006;54:437-439.
research in India and China today. From literature
17. Kawakami N, Takatwuka N, Shimizu H,
based mapping to health care policy. Curr Sci Ishibarshi H. Effect of smoking on the incidence
2002;82:1086-1097. of NIDDM. Replication and extension in a
8. Ramachandran A, Snehalatha C, Latha F. Rising Japanese cohort of male employee. Am J
prevalence of NIDDM in an urban population in Epidemiol 1997;145:103-109.
India. Diabetologia 1997;40:232-237. 18. Vanderpump MP, Tunbridge WM, French JM,
Appleton D, Bates D, Rodgers H, et al. The
9. Acemoglu H, Ceylan A, Saka G, Plalanci Y, Erten
incidence of diabetes mellitus in an English
M, Turgut S, et al. Risk factors for diabetes in
community – a 20-year follow up of the Wickham
South Eastern Anatolia of Turkey. Int J Epidemiol survey. Diabetes Med 1996;13:741-747.
2001;3(1):1540-2614.
19. Giannini C, Dyck PJ. Basement membrane
10. King H, Aubert RE, Herman WH. Global burden reduplication and pericyte degeneration precede
of diabetes 1995-2025. Prevalence, numerical development of diabetic polyneuropathy and are
estimate and projections. Diabetes Care associated with its severity. Ann Neurol
1998;21:1414-1431. 1995;37:498-504.
20. Rimm EB, Chan J, Stampfer MJ. Prospective 22. National High Blood Pressure Education
study of cigarette smoking, alcohol use and the risk programme working group report on HTN in
of diabetes in men. Br Med J 1995;310:555-559. Diabetes. Hypertension 1994;23:145-158.
21. Anderson EA, Mark AL. The vasodilator actions
of insulin. Implication for the insulin hypothesis of
hypertension revisited. Cardiovasc Risk Factors
1993;3:159-163.
40