Research paperFruit and vegetable intake, physical activity, and depressive symptoms in the African American Health (AAH) study

•Intake of vegetables show some protective effect on clinically relevant levels of depressive symptoms (CRLDS).•Physical activity, even in the form of leisurely walking, appears to protect against CRLDS.•Dietary patterns and physical activity may have greater benefits when used in combination than when used separately.•Promotion of leisurely walking appears to be warranted because of positive effects on depressive symptoms.

BackgroundHealthy diet and physical activity (PA) have been associated with reduced depressive symptoms, but few studies have examined them simultaneously in African Americans.AimsTo investigate fruit and vegetable intake (FVI) and PA as predictors of clinically-relevant levels of depressive symptoms (CRLDS) in African Americans.MethodsAfrican American Health (AAH) is a population-based longitudinal study of African Americans in St. Louis, MO, who were born in 1936-1950 (inclusive) and empaneled in 2000-01 (wave 1). At wave 8, participants self-reported fruit and vegetable intake (FVI) and completed the Yale Physical Activity Scale. At both waves 8 and 10, the CES-D 11-item scale was used to identify those who met criteria for CRLDS. Sequential logistic regression modeling was used to examine the associations of components of FVI/PA with CRLDS, both cross-sectionally (n = 680, including imputed values) and longitudinally (n = 582, including imputed values). Modeling employed gender, age, perceived income adequacy, and education as potential confounders.ResultsCross-sectionally, vigorous PA, and leisurely walking PA, were independently associated with lower odds of CRLDS in all but the fifth model and green vegetables in all models. Longitudinally, green vegetables and interactions between the FVI summary score, the PA summary score, and other factors at wave 8 were most consistently associated with CRLDS at wave 10. In both cross-sectional and longitudinal models, the socioeconomic variables showed the strongest association as risk factors for CRLDS.LimitationsBoth FVI and PA were self-reported rather than observed, our cohort had limited geographic- and age-ranges, and confidence intervals for some results were broad.ConclusionsGreen vegetables, total FVI, and various aspects of PA showed protective effects regarding CRLDS. Therefore, the promotion of such lifestyles is likely to help prevent CRLDS in this population.