2339-P: Spousal Influences on Incident Type 2 Diabetes: A Longitudinal Study of South Asians

Introduction and Objective: Concordance in prevalent diabetes among spouses is well known. It is not clear if having a spouse with diabetes independently predicts incident disease and is relevant for identifying prevention opportunities. Methods: Longitudinal data from 2775 spousal dyads residing in two large metropolises in India, in the CARRS cohort were used to model the relative risk (RR) of incident diabetes by partner’s diabetes status over 10 years, with and without adjustment for health behaviors. Heterogeneity of associations by education, BMI, and chronic conditions was evaluated. Results: At baseline, 30.7% of dyads were discordant in diabetes status. In women without diabetes at baseline, 10-y diabetes risk was 21.5% when their husbands were diabetes-free vs. 24.1% when their husband had diabetes. In men without diabetes at baseline, 10-y diabetes risk was 23.4% when their wives were diabetes-free vs. 31.9% when their wives had diabetes. There was no association between spouse’s diabetes at baseline and incident diabetes in women (Table). In men, the age-adjusted (RR= 1.29; 1.07 - 1.55) and demographic adjusted models (RR=1.26; 1.05 - 1.51) showed higher risk in men when wives had diabetes; this association was mitigated after adjusting for health behaviors and body mass index (RR=1.17; 0.97-1.40). Conclusion: The RR of incident diabetes was higher if the spouse had diabetes for men but not for women. The raised RR in men may be due to shared health behaviors and may be explored in future studies. Disclosure M. Siddiqui: None. M. Deepa: None. G. Rautela: None. R.K. Lata: None. M. Duraivel: None. A.J. Shah: None. S. Argeseanu Cunningham: None. R. Anjana: None. S.A. Patel: None.