A multisite study was initiated by investigators from different regions of the country and funding obtained from Science and Society Division, Department of Science and Technology, Government of India, New Delhi. A central institutional review board approved the study and ethical clearance was obtained from each study site. Sites participating in the study included institutions in northern (Haryana), central (Jaipur), western (Pune), eastern (Kolkata), and southern (Kochi, Gandhigram) India. Study was performed at four urban and five rural locations in the years 2004–2007. Details of the study methodology have been published earlier.24,25
In brief, a proforma was prepared which obtained information regarding social, demographic, diet, physical activity, anthropometric and biochemical variables. This questionnaire has been used in previous epidemiological studies in women in urban slums in Delhi and has been validated for diet, physical activity and other assessments.26
Investigators from each site were centrally trained to ensure uniformity in sampling methodology, questionnaire administration, physical examination and measurements, and biochemical estimations. Written informed consent was obtained from each participant. Sampling involved a systematic stratified strategy at each study site. To ensure uniformity in data collection each site was instructed to identify low and low-middle social status locations in the urban and rural areas. Urban locations were in Jaipur (central), Kolkata (east), Kochi and Pondicherry (south) and rural locations were in Haryana (north), Jaipur (central), Pune (west), Pondicherry (south), and Gandhigram (south). Low middle socioeconomic status locations were identified at each study site. The population at each study location varied from 20 to 30,000 adults and, therefore, the total population at all eight sites was 180,000 to 250,000 adults. At the study site a central point was identified and the study investigators moved house-to-house in a clockwise direction from there till the sampling target (n
= 500 at each location) was completed. We evaluated middle-aged women 35–70 years at all locations. This strategy has been used in previous studies and recommended by the WHO.27
The response rate at each site varied from 50 to 70% and was similar in rural and urban locations.
The baseline questionnaire was designed to collect information on demographic data, family income, educational level, history of chronic illnesses such as coronary heart disease, hypertension, diabetes or high cholesterol, and smoking or tobacco intake. Dietary history was inquired using 2-day 24 hour recalls. A set of standardized cups, glasses and spoons were used to assess the intake of each food item. Food intake data was converted into raw ingredients and its nutritive was calculated using food composition tables in self designed computer software.26
Physical activity was inquired using a previously validated instrument that provides details of all the day long activity.28
In Indian women, the prevalence of leisure time physical activity is low but household chores and work-related physical activity could be substantial. This questionnaire captures all these physical activity domains and is more useful for low-income countries such as India. Physical examination was performed to assess height, weight, waist and hip size, mid-upper arm circumference (MUAC), calf circumference, and blood pressure using techniques recommended by the WHO.29
All the study investigators were centrally trained in measurement techniques for uniformity. Standardised tape-measures and weighing machines that were periodically calibrated were used. Body mass index (BMI) was calculated as weight (kg) divided by squared height (m). Waist-to-hip ratio (WHR) was calculated. Sitting blood pressure was measured using a calibrated digital sphygmomanometer supplied centrally (Omron model SDX, Healthcare Omron Inc, Illinois 60015, USA). Fasting blood sample was obtained in all the study participants. Glucose was determined at the site-based central laboratory using glucose peroxidase method and external quality control and periodical validation from the central laboratory at New Delhi. Blood levels of hemoglobin, fasting glucose and total cholesterol were measured using heme oxidase, glucose oxidase peroxidise and cholesterol oxidase-phenol 4-aminophenazone peroxidase methods, respectively, with quality control.26
The diagnostic criteria have been reported earlier.24,25
Educational status was classified according to number of years of formal education into four categories: illiterate, 1–5, 6–10 and >10 years of formal education. Women who smoked tobacco as bidis or cigarettes or used tobacco in non-smoked form were classified as smokers/tobacco use. Former tobacco users were also identified. Physical activity levels (PALs) were calculated for each woman and PAL <1.4 units taken as sedentary activity as defined in an Indian study.28
Overweight and obesity defined as body mass index (BMI) ≥25 kg/m2
. Truncal obesity was defined by waist:hip ratio of >0.9 for stage II and >0.8 for stage I and waist circumference >90 cm for stage II and >80 cm for stage I.24
Hypertension was diagnosed when the systolic or diastolic BP was ≥140 and/or ≥90 mm Hg on multiple single day measurements or the subject was known hypertensive on medications.25
Hypercholesterolemia was defined by the presence of high total cholesterol (≥200 mg/dl). Diabetes was diagnosed in women with previously diagnosed diabetes or fasting blood glucose ≥126 mg/dl.24
One hundred women at each study location were randomly selected after the initial screening for chronic diseases knowledge, awareness and practices as well as dietary knowledge and behaviors. A total of 900 were evaluated at different study sites before the start of intervention. This questionnaire focussed on health issues among middle-aged women, mainly non-communicable diseases such as cardiovascular disease and their risk factors obesity, hypertension, diabetes and cholesterol; chronic lung disease, osteoarthritis and anemia. Individual study workers were trained by the central team in questionnaire delivery to achieve uniformity.
For multilevel intervention a planned strategy was implemented. We used a population-wide intervention using posters, leaflets and handouts that were designed specifically for the study. These posters were in local languages at every study site and were context specific developed at each of the sites after assessment of abnormal lifestyles and risk factors. For local population we organized street plays focussing on problem of chronic diseases and ways and means to stop tobacco and alcohol use, improve diet and physical activity (Annexure 2, video attachment). Public lectures were organized in the locality of intervention. Population-group based intervention included small group lectures and focussed group discussions. The total duration of intervention was six months.
2.2. Statistical analyses
All the case report forms were transferred to the study management office in Delhi. The data were entered in a customised database using SPSS program (SPSS Inc, Chicago, USA). In a random 10% double entry was made to check for errors. All the analyses have been performed using SPSS version 10.0. All analyses have been performed after age-adjustment within the SPSS statistical package using ANCOVA. The prevalence rates for risk factors are reported as percentage. Age-adjustment was performed using the direct method and 2001 Indian census female population was used as standard. Descriptive statistics are reported for knowledge, attitudes, practices and behaviors questions. Comparison of variables before and after intervention was performed by the non-parametric Mann–Whitney test and U-values were calculated. p values < 0.05 were considered significant.