Very high levels of diabetes have been reported in urban areas of India (1), but few data are available for rural regions where >70% of the population lives. Data from a new large-scale survey done in 2005 suggest rural India may soon experience the same epidemic of diabetes. A total of 4,535 individuals aged ≥30 years (response rate 81%, mean age 46.8 years) were sampled at random age and sex strata from 20 villages representative of Godavari, a developing rural area of Andhra Pradesh. Data were collected using a structured questionnaire and a brief physical examination with fasting finger-prick blood glucose measured in all participants using B-Braun USV meters (Melsungen, Germany). Fasting venous samples were also done in a random subsample of 1,070 individuals. Estimates of diabetes prevalence for the 20 villages were calculated by applying sampling weights derived from a census done in 2004 with diabetes defined by disease history and/or fasting glucose of 7.0 mmol or over.
On the basis of the finger-prick measurements, the prevalence of diabetes was 13.2% (95% CI 12.1–14.3), of which 6.4% (5.6–7.2) were known and 6.8% (5.9–7.6) were previously undiagnosed. A further 15.5% (14.2–16.8) had impaired fasting glucose (Table 1). Overall estimated mean fasting glucose levels from USV meters was 5.8 mmol/l (5.7–5.9). In the subsample, venous blood measurements gave a lower estimated mean glucose of 5.6 mmol/l (5.4–5.7). The systematically lower levels for the venous samples likely reflect the delay in assay, consequent upon transport of the samples to the local laboratory, and the higher finger-prick estimates are probably a more accurate reflection of the true prevalence of diabetes in this community.
Of those with known diabetes, 67% (61–73) were taking oral hypoglycemic therapy, 3% (1–5) were using insulin, and 46% (40–53) were taking blood pressure–lowering agents. These relatively high levels of treatment suggest that even in fairly poor rural settings, proven preventive therapies are accessible to many and that strategies to improve detection and treatment rates could produce substantial health benefits.
While these data are by no means representative of rural India as a whole, they do provide a reasonably precise and reliable estimate of diabetes and its treatment in the study area. Since much of rural India is likely to develop to a similar or greater extent as the Godavari region of Andhra Pradesh (2), the data provide an early indication of the likely huge burden of diabetes that will occur in rural India in the coming few decades. The generation of new evidence about detection and management strategies suited to resource poor setting is an urgent public health priority for India.
Additional data from this study can be viewed at: http://thegeorgeinstitute.org/