Prevalence of Diabetes Mellitus in school teachers of Bengaluru

Introduction

Type 2 Diabetes Mellitus (T2DM) is one of the most common non-communicable diseases (NCDs). It is the eighth leading cause of death in most high-income countries according to the WHO factsheet 2012.This epidemic is increasingly affecting people in low and middle-income countries (LMICs) (1-5). Diabetes has emerged as one of the most challenging health problems in the 21st century ( Asian Indians have been consistently found to have a higher prevalence of diabetes in most studies. In India, the first national survey was undertaken in 1971 showing the prevalence of T2DM as 2.3% in the urban areas and 1.2% in the rural areas. Since then, there has been a dramatic increase in the prevalence of T2DM in India. Most recent studies suggest prevalence rates of between 15 to 20% in urban areas and about half of that in rural areas. India has the largest population of T2DM in the world. WHO estimated 19.4 million persons were with DM in India in 1995 and this number is likely to be 57.2 million in 2025 (6). Awareness and knowledge regarding diabetes is still grossly inadequate in India. Massive diabetes education programmes are urgently needed both in urban and rural India(. The presence of various diabetic complications at the time of clinical diagnosis of diabetes is the reason behind the recommendations of screening for type 2 diabetes (7).

Bruhath Bengaluru Mahanagara Palike (BBMP) is the 4th largest municipal corporation in India. There are 3 zones in BBMP with 137 public schools and junior colleges. Job strain results from a combination of high workload and few decision-making opportunities in the workplace. There are studies, which have pointed out associations between job strain and Diabetes Mellitus (8-12). We conducted a cross-sectional study among BBMP teachers and estimated prevalence of diabetes and its association with job stress. Teachers form one of the largest groups of employed people in India. As an essential individual in the process of education, a teacher is supposed to be performing different tasks. It is reported that more than 3% of schools are still without teachers 19% of schools function with single teacher (13). Managing uniform levels of learning in all grades become complex tasks for the teachers (14). To ensure all round development of students, a teacher also has to organize co-curricular activities, sports, and field trips etc. (15). Added to this, teachers are involved in several non-teaching activities by the Government such as census enumeration and elections. In the process of ensuring the welfare of society, there is limited evidence that suggest taking care of the teacher's wellbeing. Several stressors might affect teachers such as inability to cope with teaching problems, managing non co-operative and aggressive children, differential concerns for children learning and due to relationships with other staff members (16,17). They are at risk of developing diabetes due to their nature of job and may be undergoing a high job strain. In the context of limited evidence on the prevalence of diabetes among teachers, we designed this study to screen and identify teachers who are at higher risk of developing diabetes mellitus in Bengaluru.

The study was conducted in public schools in urban Bangalore under the Bruhath Bengaluru Mahanagara Palike (BBMP). The teaching staffs of BBMP run schools are managed by permanent and temporary teachers. The objective of the study was to estimate the prevalence and distribution of diabetes and work place stress factors among school teachers working in BBMP run schools during the year 2013-2014. We also hypothesized that there is an association between stress and T2DM among teachers in BBMP schools. 

Methods

All BBMP school teachers (Nursery, Primary, High School) from the three zones of Bangalore (South, East, West) who were aged 19 to 70 years were selected for the study. The eligibility criteria of a participant were being in teaching profession and willingness to join the study. Convenient sample technique was employed, in which units in the sample are collected with no specific probability structure. Data collection was accomplished from 123 public schools in BBMP area. All three zones of BBMP schools (East, West and South zones) were covered. East zone consisted of 54 schools [high school -14; primary-6; nursery-34 (out of 34, 15 were covered)]. West zone consisted of 38 schools [high school-9; primary-3; nursery-26(out of 26, 12 were covered)]. South zone consisted of 31 schools [high school-10 primary-3; nursery-18 (out of 18, 13 were covered)]. A total number of 559 teachers are employed in high school and primary schools when compared to nursery schools. All the 123 schools were completed and no school was left out. In total, we included 408 school teachers in the sample including primary, high school and nursery schools. We sought permission to conduct the research by doing school to school visit with health check-up camps from educational officer of BBMP in all three zones. 

Inclusion criteria:

● School teachers who are working in Nursery, Primary and High school.

● Teachers who are willing to participate at the time of our survey.

● Teachers who can understand English or Kannada.

 ● Teachers available at the school at time of survey.

Study participants: This study was conducted among all teachers of BBMP schools (Contractual and Permanent, Temporary teachers). We visited school to school and screening was done to all teachers between July 15 to September 15 2014 (Voluntary screening camps). The study protocol is approved by the IEC, Hyderabad and permission to conduct screening camps in all BBMP schools was taken from Educational Officer, BBMP. 

Study design:Our study was cross sectional in nature, in which semi-structured questionnaires were used for data collection. Data was collected by doing school to school survey in BBMP area from each subject through a validated questionnaires regarding individual's age, sex, sedentary occupation, height and weight, dietary pattern, BMI, income status, smoking, ischemic heart diseases. 

Sampling Technique: It is a convenient sampling technique. We chose convenient sample technique in which units in the sample are collected with no specific probability structure. We included all the schools of BBMP in our survey, by preparing the list of nursery, primary and high school and selecting all the teachers who are working in these schools.

A self-administered questionnaire was administered to collect information on job stress and other risk factors. Stressors such as number of travel hours, travel mode, salary satisfaction, travel stress, duration of working time, job control, nature of monitoring, evaluation of work, deciding work schedule, appreciation of work, physical environment like body position, activity, ventilation, lighting, work environment like handling difficult situation, knowledge to perform work, comparison to other workers and effect of power, blaming for someone else mistakes, transparency at work place, abusive communication and discrimination at work.

Outcome Assessment: Saudek et al have provided guidelines for screening and diagnosing diabetes using the random blood sample (. In their article, they recommend that screening standards should include prompt testing and closer follow up with RBS levels of 130 mg/dl or greater (in the place of FBS values of 100 mg/dl or greater or HbA1C values greater than 6.0%). We obtained random blood samples from finger tips ensuring all aseptic precautions in order to asses capillary glucose levels using strips based Accuchek glucometer (Mankind Pharma, India). A value of 130mg/dL was considered as the cut-off. Teachers with capillary blood glucose level of 130mg/dL and above were advised to undergo a confirmatory screening test with intravenous fasting and post prandial glucose levels.

Details on variables that could have possible associations with exposure and outcome were collected. These are physical activity, tobacco use, socioeconomic status, alcohol, working hours after school hours and waist to hip ratio. We measured the height to the nearest centimetre and weight was measured by using adult electronic digital weighing machine. BMI was calculated using the formula, Height in metre sq / weight in Kg. Waist measurement was made at the approximate midpoint between lower margin of the last palpable rib and the top of the iliac crest using a stretch resistant tape and hip girth was measured at the widest portion of the buttocks with the tape parallel to the floor, to the nearest cm.

Data analysis: Binary logistic regression was conducted using SPSS version 20 for studying the relationship between diabetes and multiple stress domains. The outcome variable was recoded to create a dichotomous variable: diabetic and non-diabetic. Nondiabetic category comprised of both those who had normal blood glucose levels and those who were at risk of developing diabetes. Four major contextual domains were considered for analysis:– Work environment, emotional environment, work profile and travel stress. The issues of work environment have been considered as factors influencing stress. All the components of this domain have been found to be associated with health impacts in earlier studies (. This domain elicited responses to questions like 'encountering dilemmas during work', help received thereafter, assessment and its efficiency, knowledge to perform work and comparison to other colleagues at the work place. The major stress factors at work place like number of travel hours, travel stress, mode of transportation, job profile, salary satisfaction, discrimination at work and years of experience were considered in the study (Table-1).

The sub-domains were re coded in order to create three levels of scores – low, middle and high. For work profile, only two levels were constructed. Scores were created for each of the subdomains in a numerically ascending order from negative to positive. Scoring for each of the major domains was done by weighing the best scores of respective sub-domains within, in ascending order from lowest to highest.

Results

Only 401 teachers participated out of 559 with eighty seven teachers not participating due to several reasons. These include being on leave (47), training (33), and disinclination (7) and there were many vacant posts (71). In the study sample of 401 school teachers from BBMP, 74.8 % (n=300) had normal glucose values while 20% (n=83) were at risk of having diabetes and 4.5% (n=18) knew they had T2DM. (Figure.2) 

Based on the employment status, 12 (66.7%)known diabetics were permanent employees in comparison to 5 (27.7%) contractual employees and one (5.5%) temporary employee, because most of the permanent employees fall in the age group of 31 to 60 years. Similarly, number of pre diabetes cases increased from 12 (10.9%) in the age group 31 to 40 years to 32 (48.5%) in the age group 51 to 60 years. Among the teachers who had completed pre-degree, 12 (14.5%) were at risk of developingT2DM (pre diabetes), while the corresponding proportion among degree holders was 33 (39.7%) and 2 (2.4%) in post-graduate degree holders.

The major stress factors like parental/sibling diabetic condition, past history of diabetes, BMI, Waist to Hip ratio were significant in the study in relation with diabetes (Table 1). 

Among the teachers in the study sample, 74% had normal blood glucose levels, 21% had higher glucose levels than normal and were at risk of developing diabetes while 4% were known Diabetics who were on medication. Among the offspring of diabetic parents, 64.9% of teachers had normal glucose levels, whereas 6.6% were diabetic and 28.5% were at risk of developing diabetes. Among the offspring of non-diabetic parents, 80.8%hadnormal blood glucose levels, whereas 16% were at risk of developing diabetes and 3.2%haddiabetes (n=8/250). In our sample of 401 participants, 18were known diabetics. (Table-2) 

About 47.1% of the respondents were in the normal range of BMI (18.5 to 24.99), among whom 15.8% were at risk of developing diabetes. About 33.9% of the teachers had BMI in the overweight category, with 29.4% of them at risk of developing diabetes and 2.9% had diabetes. Among obese teachers (BMI >or=30), 25.5% were at risk of developing T2DM and 8.5% were known diabetics.

Stress Factors at Work Place Associated with Diabetes  

Among teachers who had stress due to travel, 4.6% are known diabetics on medication and 21.4% were at risk of developing diabetes. Job profile is an important stress factor. Amongst those who were at the level of headmaster,13.8% (n=4/29) were known diabetics on medication whereas 27.6% were at risk of developing the same .Amongst those who were at the level of assistant master, 3.8% were known diabetics on medication while 19.2% were at the risk of developing the same. Among teachers at the level of nursery, 3% were known diabetics on medication while 30.3%were at the risk of developing the same. (Table-3)

Around 16.2% of respondents not satisfied with salary were at risk of developing diabetes. We found that there is no significant association between risks of developing diabetes due to discrimination. In respondents with work experience between 5- 10 years, 24.82% of them and 32% with more experience were at risk of getting diabetes. (Table-3)

Crude estimates of the following contextual domains were done without accounting for confounders. None of the stress factors s h owe d relati o n s h i p wit h d ia b etes o n set i n t h e respondents.(Table-4,5) We adjusted for age, gender, waist by hip circumference, family history of diabetes, socio-economic status, marital status, tobacco ever use, moderate physical activity for at least 30 minutes 5 days a week and alcohol use. After adjusting the confounders, none of the contextual domains were found to have association with diabetes (Table-5).

Discussion

We conducted a study of public school teachers and report the prevalence of diabetes in them in Bengaluru. We focussed on the age gradient of diabetes and explored its association with predictors due to stress at the work place. We report that contextual stress domains at work place have no significant association with diabetes. In our study population, the overall prevalence of diabetes was estimated around 4.5%. The highest prevalence was observed in the elder age group. The study also indicates that 20.7% of the population were at risk of developing diabetes, including 26.5% male and 16.6% female teachers. The odds of developing diabetes were found to be higher with advancing age and higher number of years of work but none of the results were significant after adjusting with confounders. 

Earlier studies have shown a wide range of prevalence (1.5 to 19.5%) of DM among different age groups and in different geographical regions of India (. Prevalence estimates of diabetes in urban slums of India were 10.3% ( while in the rest of the urban population have varied from 8.2% to 15.5 %.( . Estimates from China indicate prevalence of 5.2% in men and 5.8% in women (, 5.7% in Indonesia, 8% in Malaysia and Singapore and 11.9% in Thailand (. The world prevalence of diabetes among adults aged 20-79 years was projected to be around 6.4% in 2010 which was further projected to increase exponentially to 7.7% by 2030 (. In this backdrop, our study points out a slightly lower prevalence in comparison to previous urban estimates in India. The underestimation can be due to several reasons including unavailability of results from missed teachers, use of RBS for screening and relative younger age of the teachers. Evidence indicates that participants with more risk factors have greater absenteeism with most significant for individuals with diabetes(. Hence, it is possible that the greater proportion of missed teachers might be attending to treatment of disease or its complications. Further, a major study indicated that variations in blood glucose level undergo progressive shift due to contributions of fasting and postprandial hyperglycemias when the patient's progress from moderate to high hyperglycemias (.Evidence indicates that RBS levels are highest 1–3 hour postprandially and decrease thereafter ( Therefore, non-uniformity in the timing of tests and eating habits may have influenced the variation. Hence, a panel of experts recommended that RBS levels of 130–199 mg/dl can be considered as a positive screening test for diabetes, if persons in this range are subjected to diagnostic testing (. Our study did not have resources for such diagnostic approach and future studies can consider incorporating this suggestion. 

We observed that amongst the teachers who are travelling less than 1 hour to reach their office, 11.2% of them are at the risk of developing diabetes while 2.2% of them are known diabetics on medication. This may probably be a pointer towards insufficient physical activity. Teachers with 5 to 10 years of experience, about 8.7% are at the risk of developing diabetes. Our study could not establish any significant association between stressors at work place or DM. Probable reasons for the same could be skewed distribution of respondents across the age categories, the fact that random blood samples were collected for analysis instead of performing a glycosylated haemoglobin test, the use of a glucometer for analysis, composition and construct of the various domains, the fact that this was a cross-sectional study meant only for screening of diabetes and the want of a larger sample size.

In our study population, we found that there is an increasing risk of developing diabetes with increased age. Our study findings also suggest that risk factors such as obesity, family history of diabetes, waist to hip ratio are associated with an increase incidence of diabetes. Socio economic factors did not show any significant effect on the prevalence of diabetes. The risk also increased with experience and age of the individual and it was observed that males were more prone of developing diabetes than females. It is better that screening is done in early adulthood so that at risk cases could be identified in order to focus on awareness and preventive measures to reduce the risk of diabetes. This has also been recommended by American Diabetes Association (ADA) for screening of asymptomatic adults before 45 years and even earlier (. Testing might also promote the healthy behaviour. For example, teachers who were known diabetics were very keen to check their blood glucose levels.

Many risk factors leading to diabetes are modifiable and therefore provide an opportunity for preventive efforts. It is important to educate the teachers to engage in healthy habits, which reduces the risk of developing diabetes. These include healthy diet, regular physical exercises and weight reduction. These will not only prevent the early onset of diabetes and also will reduce lifethreatening complications. We recommend specific programs for teachers including preventive strategies by modifying lifestyles by increasing physical activity, reducing obesity and overweight as well as adopting healthier lifestyle. We suggest that preventive efforts can be planned to increase awareness in public school teachers through early screening. The key message to them should be to focusing on life style modifications and routine monitoring of blood glucose levels. The education department of the city corporation should liaise with the health centres of the BBMP for regular awareness and screening programs for all NCDs.

Limitations

In our study sample, we cannot reliably infer whether there is under or over estimation of proportion of diabetics as no information was available from many teachers. Few young teachers were not interested in answering the questionnaires as it took long time to answer, and it is possible that they have not answered properly regarding awareness and physical exercise and might have greater chance of having diabetes. Measurement error is an important limitation for variables such as blood glucose, height and weight. We used glucometer for testing blood glucose levels. Future studies should use better measures of glucose such as glycosylated haemoglobin or fasting sugar, which are more reliable indicators of glycaemic control and provide better estimate. As defined in the objectives, the purpose of the study was only screening for diabetes and hence no diagnosis was done in this study, which is an important limitation. Finally, as our study was cross sectional study design, temporality cannot be established. Hence no causal inference could be established.