Background: Among Hispanics/Latinos, there is substantial heterogeneity in the prevalence of depressive symptoms and diabetes by background. This study aimed to examine the association between depressive symptoms and incident diabetes among Hispanic/Latino adults of diverse backgrounds.
Methods: The Hispanic Community Health Study/Study of Latinos (HCHS/SOL) is a prospective, community-based study of 16,415 self-identified Hispanic/Latino adults aged 18-74 years at Visit 1 (2008-2011) and 10,914 who have also completed the Visit 2 examination by August 2017. This analysis includes 8,804 participants who were free from diabetes at baseline and attended both Visits. Baseline depressive symptoms were assessed using the CES-D 10-item depression scale and divided into quartiles. Incident diabetes was defined by fasting glucose ≥126mg/dL, 2-hour postload glucose 140-199mg/dL, HbA1c ≥6.5%, or self-report. Accounting for HCHS/SOL complex survey design, we used Poisson regression models to estimate diabetes incidence density ratios (Table).
Results: Baseline BMI was associated with both depressive symptoms and incident diabetes. Overall, 876 participants developed diabetes. In analyses stratified by Hispanic/Latino background, the association between baseline depressive symptoms and incident diabetes was significant for South Americans, Central Americans, and Mexicans. No significant associations between depressive symptoms and diabetes were seen among Hispanics/Latinos of Dominican, Cuban, or Puerto Rican backgrounds.
Conclusions: These findings suggest that the association between depressive symptoms and incident diabetes in Hispanics/Latinos may differ by their background, with South and Central Americans at the highest risk. This difference may be partly explained by variation in participants’ understanding of CES-D questions by Hispanic/Latino background. Future research is needed to understand these novel findings fully and explore their implications for practice and policy.