Rural-Urban Migration and Cardiometabolic Risks

The study was nested within a Cardiovascular Disease Risk Factor Study (CVDRFS) in four Indian cities (Bangalore, Lucknow, Nagpur, and Hyderabad), situated geographically in the north, centre and south of the country, and covering sites where rural-urban migration occurs. For the preliminary work reported here, work was confined to two of the sites, Lucknow and Hyderabad. Participants in the CVDRFS baseline survey, together with their co-resident spouses, were asked about rural-to-urban migration and those responding positively, together with a 25% random sample of non-migrants, were invited to participate in the study. Indian Census 2001 definitions were used to classify areas as urban or rural based on population size and density and non-agricultural employment [13 ]. Migration status was attributed only to intra-generation migrants (i.e. 'first-generation') and of at least one year's duration. Place of origin was identified using a commercial GIS application of the Indian census produced for the study. The software enabled village level unique census identifier codes to be assigned to each participant's place of origin, permitting electronic linkage to relevant census data.
Each participant was asked to identify one non-migrant full sibling of the same sex and closest to them in age. In the case of migrants whose siblings had also migrated, a half-sib, and if not available, then the closest cousin and still resident in the village of origin was recruited instead of the full sibling. For non-migrant workers, siblings who resided in the same city but did not work in the factory were recruited to enable prevalence of obesity and diabetes among factory workers and their sibs to be compared, estimating any healthy worker effect and more generalisable urban prevalence rates than that obtained solely from factory workers. The sampling strategy is shown in Figure 1.