The prevalence and associated risk factors of pre-diabetes and diabetes among adults in the Batticaloa District, Sri Lanka

Introduction: The prevalence of pre-diabetes and diabetes in Sri Lanka has gradually increased over the last few decades. This study aimed to determine the prevalence and risk factors of pre-diabetes and diabetes among adults in randomly selected Divisional Secretariat divisions in Batticaloa district, Sri Lanka. Methods : A cross-sectional community-based descriptive study was conducted among 582 adults aged ≥ 18 years. A pre-tested and validated interviewer administrated questionnaire was used and a blood sample was obtained to assess the glycosylated hemoglobin level. Statistical analysis was performed using SPSS v 20. Results: Nearly 78% of them were females and mean age of the participants was 50.5 (SD±13.2) years. Pre-diabetes plus diabetes (HbA c ≥ 5.7%) was observed among 63% (95% CI = 33.0 to 41.0). The overall 1 prevalence of pre-diabetes and diabetes was 34.2% (95% CI = 30.3 to 38.2) and 28.9% (95% CI = 25.2 to 32.8) respectively. The mean BMI of the participants was significantly higher among pre-diabetes and diabetes category (dysglycaemia) when compared to the normal category (M = 1.81, 95% CI = 1.01 to 3.01, p = 0.001). Age ≥ 50 years (AOR = 5.88, 95% CI = 2.99 to 11.55), overweight/obesity (AOR = 2.43, 95% CI = 1.18 to 4.97), consumption of white rice (AOR = 2.77, 95% CI = 1.20 to 4.29), and family history of diabetes (AOR = 2.95, 95% CI = 1.51 to 5.77) were found to be the risk factors for dysglycaemia (HbA c of ≥ 5.7%) in the binary logistic 1 regression model. Conclusions


Introduction
The World Health Organization (WHO) has reported that nearly 41 million people die of noncommunicable diseases (NCDs) in the world each year which is equal to 71% of all deaths globally (1). Over 15 million dies prematurely (age 30 -69 years) from NCDs, of whom 85% live in lower-and middleincome countries (1). The majority of these deaths are due to four common categories of noncommunicable diseases such as cardiovascular diseases (heart attack and stroke), diabetes, cancer, and chronic respiratory diseases (2). Diabetes is considered as most common and worst NCDs and associated with significant morbidity and mortality (3).
The incidence of diabetes is increasing rapidly worldwide. Nearly 8.5% of adults aged more than 18 years had diabetes in 2014 and 1.5 million deaths occurred directly due to diabetes globally in 2019 (4). The estimated global prevalence of diabetes was 463 million in 2019 and projected to be 578 million by 2030. Further, 3 in 4 people with diabetes live in low-and middle-income countries (5). Further, the overall prevalence of diabetes for Sri Lankans aged above 20 years was 10.3% in 2006 with a projected prevalence of 13.9% for the year 2030 (6).
The prevalence of pre-diabetes and diabetes globally and in Sri Lanka has been gradually increasing over the last few decades. Further, it has been highlighted that one in five adults in Sri Lanka has pre-diabetes or diabetes, and one-third of them were found to be undiagnosed and the overall prevalence of some form of dysglycaemia was 21.8% (6). A recent Sri Lankan study reported a rise of the urban prevalence of diabetes and pre-diabetes compared to a study conducted before 10 years (6,7). In addition, Sri Lankan study reported that the cumulative prevalence of pre-diabetes and diabetes was 57.9% in urban Colombo in 2019.
In addition, mortality and morbidity from macrovascular and microvascular diseases due to diabetes have shown an alarming increases in Sri Lanka (9). It is important to note that the number of hospital deaths in 2019 due to diabetes was 714 (3.3 deaths per 100,000 population) and the death rate was relatively high in Batticaloa district when compared to other districts (10).
Due to the silent nature of diabetes, a significant percentage of diabetic patients are unaware of their status. Screening and detection of pre-diabetes and diabetes among the public would help to prevent its associated mobility and mortality which is associated with the country's economy. Estimation of glycosylated haemoglobin (HbA c) has been 1 recommended for the diagnosis of abnormalities in glucose tolerance including prediabetes and diabetes (11). Thus, the study was aimed to determine the prevalence and risk factors of prediabetes and diabetes among adults in selected Divisional Secretariat divisions in Batticaloa district, Sri Lanka.

Methods
A community-based cross-sectional descriptive study was conducted among selected Divisional Secretariat (DS) divisions, in the Batticaloa District, Sri Lanka. Batticaloa District has fourteen Divisional Secretariat (DS) divisions and among those, only 4 DS divisions were randomly chosen namely Kattankudy, Manmunai North, Eravurpattu, Koralaipattu based on the feasibility. Male and female adults aged ≥ 18 years and who had lived permanently at least for 2 years in selected DS divisions of Batticaloa district were included in the study. Those who refused to give consent for participating and suffering from severe chronic diseases were excluded.
The sample size was calculated using Lwanga & Lemeshow formula (1991) (12) and the absolute 2 error or precision of the estimate (d ) was 0.05, Z = Standard normal deviation for the selected level of confidence, was 1.96 for a confidence level of 95%. The prevalence of diabetes (39%) from a pilot study was used to calculate the sample size. The calculated sample size was 366. The design effect for the population study was unknown. Thus, the design effect was considered as 1.5. The calculated sample size after applying the design effect was 549. The non-response rate was assumed as 10%. Thus, the sample size after correction for non-responders was 604. Finally, a total of 582 subjects were recruited from all four DS divisions. The number of subjects was decided from each DS division based on the proportions in the population of each DS division. A Grama Niladari division (GND) was considered as a cluster. Voters' registers (updated annually for those aged ≥ 18 years) were used to randomly select the first household in each cluster. The selected household was visited by the study team. The one closest to the right side of the front door of the first house was visited next. This procedure was repeated until the required number of respondents was interviewed in each cluster. In each house visited, all eligible males and females were listed and the person to be interviewed was selected randomly using the lottery method. Written consent was obtained from all potential study participants in each household after providing information before random selection. If any of them refused to participate in the study, that household was not selected.
A validated and pretested interviewer-administered questionnaire (IAQ) was used to collect the data. The IAQ developed in the English language was translated into the Tamil and Sinhala languages using the back-to-back translation method. This IAQ was pretested with a group of subjects who are similar in characteristics in other DS divisions which were not included in the study. Wording problems, the way the questions are asked, length of the questionnaire, language, sequence of the questions were checked during the pretesting, and modifications were done. In addition, anthropometric measurements such as weight and height were also obtained from each participant according to the standard methods (13). In addition, a venous blood sample was collected in heparinized tubes from each participant for HbA c measurement 1 by a registered nurse. Biochemical tests were performed in the certified laboratory using a method that is National Glycohaemoglobin Standardization Programme (NGSP) certified and standardized to the Diabetes Control and Complications Trial (DCCT) assay. Data collection was carried out by a field team of medical and nursing graduates who were trained properly before commencing data collection. Before the beginning of the data collection, trained data collectors were explained about the aim of the research. They were also informed about the anonymity of the information and their privacy of storage.
Data were analyzed using a statistical package for social sciences (SPSS) version 20. Descriptive statistics were used to calculate percentage, mean and standard deviation determined for variables.
Continuous variables were tested for normality and one-way analysis of variance (ANOVA) was used to assess the mean differences between three or 2 more dependent variables. c and unadjusted odds ratio with their 95% confidence interval (95% CI) test were used to check the significance of the differences observed. Univariate and multivariate analyses were used to identify the factors associated with glycaemic status. A probability level of p < 0.05 was set as statistical significance. HbA c values 1 were classified as 5.7% -6.4% -pre-diabetes while a value of ≥ 6.5% was defined as diabetes (11).
Ethical approval for conducting the study was obtained from the Ethics Review Committee of the Faculty of Health Care Sciences, Eastern University, Sri Lanka. Written informed consent for participation was sought prior to recruitment for the study. Permission was obtained from the Divisional Secretary of the selected DS divisions after obtaining permission from Government Agent, Batticaloa District.

Results
Of the 582 participants, nearly 78% were females and 47% were in the age group of 50 -69 years with the mean age of 50.5 (SD±13.2) years range from 18 -83 years. Further, 90% of participants were Sri Lankan Tamil and 79% were married. Nearly 39% of participants had completed primary education and more than half of them were living in rural areas.
A statistically significant difference was observed in the mean differences of BMI across the groups [F (2,579) = 6.64, p = 0.001]. Bonferroni post-hoc comparisons of the three groups indicate that the normal group had a significantly lower BMI (M= 1.81, 95% CI = 1.01 to 3.01, p = 0.001) when compared to diabetes group. Further, there was a statistically significant difference observed in the mean differences of HbA1c level across the groups [F (2,579) = 5.95, p = 0.000]. Bonferroni post-hoc comparisons of the three groups indicated that mean HbA1c level was significantly higher in diabetes group compared to pre-diabetes (M = 2.66, 95% CI = 2.41 to 2.90, p = 0.000) and normal group (M = 3.31, 95% CI = 3.07 to 3.55, p = 0.000) respectively (Table  2).
There was a statistically significant association between glycaemic status and ethnic background, age group, and marital status (p<0.05). In addition, family history of DM, BMI and smoking were also significantly associated with the glycaemic status of the participants (p<0.05) ( Table 3).
There was a statistically significant association between glycaemic status and usage of soft drink and type of rice use among participants (p<0.05) ( Table 4).
Participants' age of ≥ 50 years was 3.5 times more likely to have dysglycaemia than those who were < 50 years (OR 3.59, 95% CI = 2.52 to 5.12). The presence of the family history of diabetes (OR 2.40, 95% CI = 1.62 to 3.55) and overweight/obesity (OR 2.01, 95% CI = 1.40 to 2.89) were nearly 2 times more likely to have dysglycaemia when compared to their counterparts. Usage of soft drinks and using only white rice were nearly 1.5 times (OR 1.45, 95% CI = 1.03 to 2.05) and 3 times (OR 2.97, 95%, CI = 1.84 to 4.77) more likely to have dysglycaemia respectively compared to their counterparts (

Discussion
The results revealed that the estimated prevalence of pre-diabetes and diabetes among the study participants was 34% and 29% respectively. A study conducted in most areas of Sri Lanka except the Northern and Eastern areas, the estimated prevalence of pre-diabetes and diabetes was 11.5% and 10.3% respectively (6). Further, the prevalence of diabetes in the Jaffna district in 2015 was 16.4% (14). In addition, the cumulative prevalence of prediabetes and diabetes was 63% in the present study which is higher when compared to previously reported studies in Sri Lanka, 21.8% by Katulanda et al., 2008 and 57.9% by Somasundaram et al., 2019 (6,8). This increase can be due to many potential factors such as the increased prevalence of obesity, aging, ethnicity, physical inactivity, socioeconomic status, education, and urbanization (15,16). Further, the high prevalence of pre-diabetes and diabetes in this study could be explained by the accuracy of the screening test (HbA1c testing) that was used as American Diabetes Association recommended HbA1c testing is a better diagnostic tool for prediabetes and diabetes screening (11) where other studies use the fasting blood sugar as screen tool. In addition, HbA1c has the potential to reflect the history of mean insulin sensitivity over the preceding period and it serves as a marker of insulin sensitivity in those who have an abnormal glucose tolerance (17).
According to the binary logistic regression, age, BMI, Family history of diabetes and consumption of white rice were found to be the risk factors for development of diabetes. Advanced age was found to be a significant risk factor for the glycaemic status of the participants. Increasing prevalence in advanced age can be due to lack of exercise, increase in psychosocial stress of retirement that has proportionately increased risk of the disease (18). Further, fat deposition and development of insulin resistance are high among the aging population (19). Similarly, the finding is comparable with the Sri Lanka Diabetes and Cardiovascular Study where higher prevalence was observed in participants with advanced age (6).
Further, the BMI of the participants was found to be a risk factor for pre-diabetes and diabetes. Similarly, many studies have shown that increased BMI has been linked to an increased prevalence of diabetes (15). An increased prevalence of obesity has been resulted due to urbanization and many people might be exposed to sedentary lifestyle, higher frequent usage of a vehicle for transportation, watching television frequently and less physical inactivity.
The prevalence of pre-diabetes and diabetes was relatively high among participants who are having a family history of diabetes. Family history of diabetes and history of consanguinity have been identified as risk factors for impaired fasting glucose in Southeast Asians (20). Further, Asian Indians have a strong familial association of diabetes with a high prevalence of diabetes among the firstdegree relatives and vertical transmission through two or more generations (21). People having a family history of diabetes should adopt healthy lifestyles from an early age to prevent or retard the development of diabetes (18).
In addition, white rice consumption was found to be the dietary risk factor for the development of pre-diabetes and diabetes in this study. Usage of white rice was 3 times more likely to develop pre-diabetes and diabetes when compared to brown rice. Similarly, a meta-analysis and systematic review published in the British Medical Journal found that higher consumption of white rice was significantly associated with increased risk of development of type 2 diabetes especially among Asian population (22). Another large cohort study conducted in US found that substituting brown rice for white rice was associated with a 16% lower risk of developing type 2 diabetes (23). In addition, a South Indian study shows a potential benefit on HbA c when substituting brown rice for white rice 1 among overweight adults with metabolic syndrome (24). This difference can be explained that during the processing, majority of the bran and some of the germ are removed from white rice and this grain consists of numerous phytochemicals, minerals, essential fatty acids and fibre which are believed to be the protective factors against type 2 diabetes (25). Also, a meta-analysis found that consumption of brown rice improves the insulin sensitivity due to its insoluble cereal fibre and thereby reduce the risk of type 2 diabetes (26).