Abstract
Despite evidence that Indian immigrants in high-income countries have higher diabetes risk, few studies have directly compared Indian immigrants to both Indians in India and the general population. We compared diabetes prevalence in the Mediators of Atherosclerosis in South Asians Living in America (MASALA) study (Indian immigrants) (N = 686), the Longitudinal Aging Study in India (LASI) (Indians in India) (N = 40,496), and the Health and Retirement Study (HRS) (general US population) (N = 7643), accounting for selective immigration using propensity score matching. We used generalized regression models to assess associations between diabetes and acculturation in MASALA and compare correlates of diabetes across studies. After matching, Indians in India had a higher prevalence of diabetes (37.9 [35.4-40.5]) than Indian immigrants in the US (26.7 % [23.5-30.1]) and the general US population (19.6 % [17.6-21.8]). Higher acculturation was associated with lower diabetes prevalence (prevalence ratio [PR]: 0.68 [0.45-1.04], P = 0.078) and lower HbA1c (difference: 0.205 % [-0.408 to-0.001], P = 0.049). We also identified differences in the magnitude of correlations between diabetes and risk factors, including abdominal obesity (MASALA PR: 1.41 [1.09-1.81], LASI PR: 2.41 [2.29-2.54], HRS PR: 2.52 [2.17-2.93]). Cultural factors, including differences in lifestyle and diet, may play an important role in the high diabetes risk among Indian immigrants and explaining racial disparities in diabetes burden in the US.