Prospective associations between multiple lifestyle behaviors and depressive symptoms

Abstract

Background: Our aim was to analyze the associations between multiple lifestyle behaviors and depressive symptoms. Methods: We included 4,725 adults (18-59y), that provided data in routine health evaluations of a hospital in Brazil, followed for a mean period of 3.1 +/- 1.6 years. Physical activity, alcohol consumption (measured using Alcohol Use Disorders Identification Test) and tobacco smoking were categorized as: (1) absence of the behavior (inactivity i.e. not complying with 150 min of moderate-to-vigorous PA/week, not smoking, no risky drinking, i. e. AUDIT<5) during baseline and follow-up; (2) Absence during baseline and presence during follow-up; (3) Presence during baseline and absence during follow-up; (4) Presence during both time points. Depressive symptoms were measured with the Beck Inventory was adopted to analyze patterns of depressive symptoms over time (as exposure). C-reactive protein [HS-CRP]) was assessed and its role in the association was tested. Incidence indicators of behaviors and depressive symptoms were created and used as outcomes. We used crude and adjusted Poisson regression analysis. Results: Fully adjusted models revealed that persistently physical inactive participants (RR:1.71;95% CI:1.33-2.21), those who became physically inactive (1.68;1.19-2.26), with consistently risky drinking (1.62;1.15-2.30), and who became risky drinkers (1.62;1.15-2.30) had higher risk for incidence of elevated depressive symptoms. Vice versa participants with incidence of depressive symptoms over time presented higher risk for physical inactivity (1.44;1.11-1.87) and risky drinking (1.65;1.16-2.34) incidence. HS-CRP did not influence the associations. Limitations: Self-reported physical activity, binary tobacco smoking, and non-probabilistic sampling. Conclusions: There is a prospective relationship between elevated depressive symptoms and adverse lifestyle behaviors.