Association between secondhand smoke and depression in never-smokers and its moderation by sex. Results from a national sample in Germany

Abstract

Introduction Findings on the association between secondhand smoke (SHS) exposure and depression are ambiguous. While some studies have identified positive associations between SHS exposure and risk of depressive symptoms or major depression (MDD) among former, non- and never-smokers, others did not. Similarly, results concerning sex-specific associations are contradictory. Some studies revealed an increased risk for either men or women only, whereas other studies found no differences between sexes. Our aim was to contribute to this growing body of research. Using a large and representative secondary data set from Germany, we examined whether SHS is associated with depressive symptoms/MDD and whether this association differs between males and females. Methods We pooled data from three waves of the German Health Update telephone surveys from 2009, 2010 and 2012. From the combined sample (n = 62,606) we included only never-smokers, resulting in a subset of 10,808 male and 17,862 female respondents. Exposure to SHS was determined from self-reported information on the number of days per week with SHS exposure. Being exposed on ‘no day per week’ was defined as being not exposed, exposure on at least one day per week was defined as being exposed. The outcome variable of interest was a reported medical diagnosis of a depressive episode or MDD. To determine potential interactions between sex and SHS exposure with respect to likelihood of depression, we applied logistic regression models with an interaction term, adjusting for age, socioeconomic status (low, middle, high), current gainful employment (yes or no), living together with a partner (yes or no), social support (Oslo-3 Social Support Scale), alcohol consumption (AUDIT-C categories), body mass index (BMI confidence intervals (95% > CI) are reported for main effects. Interaction effects were evaluated based on average marginal effects (AME) to avoid bias related to unobserved heterogeneity. Results In our data, 34% of male never-smokers and 21.5% of female never-smokers were exposed to SHS. Female never-smokers were affected by depressive episodes or MDD to a larger extent than male never-smokers. After adjustment for confounders, no significant association between SHS exposure and depressive episodes/MDD could be observed (OR = 0.92; 95% > CI = 0.83, 1.01). There was a significant sex-by-SHS interaction. An inspection of AMEs revealed that male never-smokers exposed to SHS were at a significantly lower likelihood of depression than male never-smokers not exposed to SHS (AME = −0.017; 95% > CI = −0.029, -0.006). No significant differences were observed between exposed and non-exposed females (AME = −0.002; 95% > CI = −0.016, 0.011). Conclusions In our analysis, we found no significant association between SHS exposure and risk of depression in the combined sample of males and females. Among males, the association was counter-intuitive by indicating that exposure to secondhand smoke decreases the likelihood of depression. As sensitivity analyses revealed, the findings were robust to the selection of covariates. Never-smokers exposed to SHS are presumably more frequently involved in social interactions than those not exposed which may be protective for the development of depressive symptoms. This confounder could perhaps not been sufficiently accounted for by adjusting for social support. Our investigation confirms that studies on the association between SHS exposure and depression do not provide a clear picture. Additional research with high-quality primary data accounting for all relevant confounding variables is necessary to allow further insights into the relationship of SHS and depression.