TY - JOUR T1 - The Contribution of Lifestyle Patterns to Clustering of Elevated Depressive Symptoms within Birth Cohorts JF - International Journal of Epidemiology Y1 - 2021 A1 - Gueltzow, Maria A1 - Maarten J. Bijlsma A1 - van Lenthe, Frank J A1 - Mikko Myrskylä KW - alcohol KW - BMI KW - depressive disorders KW - ethnicity KW - Obesity KW - Physical activity KW - Smoking AB - Some birth cohorts experience a larger burden of depression than others. We hypothesize that lifestyle, i.e. BMI, alcohol consumption, smoking and physical activity, are potential drivers of these generational differences.We analyzed data from US adults aged 50-80 years enrolled in the Health and Retirement Study (N = 163,760 person-years). Birth cohort effects were estimated with the age-period-cohort model approach according to Carstensen. Consequently, we assessed the contribution of lifestyle factors by comparing the predicted probability of elevated depressive symptoms to a counterfactual scenario in which all birth cohorts are assigned the lifestyle factor distribution of the 1945 cohort (counterfactual decomposition analysis). We stratified all analyses by sex and ethnicity.BMI contributes to an increased probability of elevated depressive symptoms of up to 32.7% (95%CI: 190.9-11.23%, 1923 cohort) for cohorts born before 1927 and a decrease of up to 16.7% (95%CI: 0.5-26.8, 1964 cohort) for cohorts born after 1959. Contributions are most pronounced in females and white/Caucasians. Alcohol consumption contributes up to 20% (95%CI: 0.8%;45.3%, 1925 cohort) to cohort effects of elevated depressive symptoms, whereas the magnitude differs by ethnicity. We found no evidence for contributions of smoking or physical activity.Birth cohort effects of elevated depressive symptoms can be partly explained by lifestyle. In particular, mental health of females and the white/Caucasian population may have suffered from the increase in obesity levels in the US.BMI and alcohol consumption, but not smoking or physical activity, contribute to birth cohort differences in depression risk. VL - 50 SN - 0300-5771 IS - Supplement_1 ER -