Bidirectional Associations Between Alcohol Drinking and Depressive Symptoms Among US Adults Aged 50 to 75: The US Health and Retirement Study

<b>Background/Objectives</b>: Low or moderate alcohol drinking may reduce the risk of depression, but depression may induce alcohol drinking. However, the bidirectional associations between alcohol drinking and depression were inconsistent, and many prior analyses were not properly conducted. This study explored the within-individual bidirectional associations between alcohol drinking and depressive symptoms under a causal analytic framework. <b>Methods</b>: Using data for the baby boomer cohorts (born between 1948 and 1965) from the Health and Retirement Study (HRS), we employed the unit fixed-effect models with lagged measures to examine the within-individual bidirectional associations between the number of alcohol drinks per week and the changes in the eight-item Center for Epidemiological Studies-Depression (CES-D) scores. <b>Results</b>: Among 11,057 participants at baseline, about 48% were drinkers and 19% had a CES-D ≥4, i.e., at a high risk of depression. Among male low/moderate drinkers, increasing alcohol drinking between consecutive visits was significantly associated with a decrease in depression scores after adjusting for prior alcohol drinking (-0.15 points per 7 drinks/week increase, <i>p</i> = 0.009). Conversely, among male drinkers and female heavy drinkers, increasing depression scores between visits increased alcohol drinking after adjusting for prior depression scores (ranging from 0.22 to 0.79 drinks/week per 1 point increase of depression score, all <i>p</i> values < 0.01). <b>Conclusions</b>: The bidirectional associations between alcohol drinking and depressive symptoms were evident only among male drinkers, and alcohol drinking should not be recommended as a solution for preventing or relieving depressive symptoms. <b>Limitations</b>: Measures of alcohol drinking and depression were coarse, and the study cohorts were limited to the US baby boomer generation. Generalizing findings to other populations should be cautious.