The effect of smoking on depressive symptoms

Abstract

Background: People who smoke tend to report higher levels of psychological distress than those who do not smoke. However, studies suggest that smoking reduces many sub-dimensions of distress, including depressive symptoms. A negative (i.e. depression reducing) effect of smoking on depressive symptoms might be observed in survey panel data only after adjustment for the effect of depressive symptoms on smoking. Results from analyses that make that adjustment are presented within groups jointly defined by race and gender, and for all respondents in a representative sample of the U.S.

Method: Reciprocal effects of smoking and depressive symptoms (DSx) are estimated with dynamic panel models. Data are from the Americans' Changing Lives study which assessed smoking and depressive symptoms five times over 25 years. Additional analyses provide estimates of effect of continuing to smoke versus quitting.

Results: The estimated effect of smoking on DSx after adjustment for the simultaneous effect of DSx on smoking and unobserved confounders is negative among women. The estimated effect among men is not statistically significant, and no race difference is observed.

Conclusions: The tendency for smoking to reduce depressive symptoms among American women is masked in analyses that fail to account for the reciprocal effect of depressive symptoms on smoking.

Keywords: Causal Effects; Depression; Gender; Mood; Psychological Distress; Race; Smoking.