Cognitive Scores among Retired Adults as a mediator on the risk of Depression

Isadora Comens, VP&S Class of 2025

Name: Isadora Coments
School: Vagelos College of Physicians and Surgeons, Class of 2025
Mentor: Ana Jimenez-Bautista, MSW and Antonia Diaz Valdes Iriarte, PhD

View Presentation Poster


Depression is a major global health concern (Vos et al., 2017), especially in the aging population. In this study, we examined one possible mechanism of retirement’s effect on mental health: Do cognition scores mediate the association between retirement and depression Data were obtained from the most recent wave (2018) of the Health and Retirement Study in the United States. We examined the cross-sectional association between retirement status (retired vs. not retired) and depressive symptoms, measured by the CESD-8 scale (coded as risk of depression for CESD-8 scores >3). Then, mediation analyses were conducted for the cognition scores, controlling for age, sex, and race. First, a multiple regression model with cognition scores regressed onto being retired and covariates was conducted. Second, a logistic model – odds ratios, with depression regressed on being retired, cognition, and covariates was conducted.Being retired decreased the cognition score by 0.36 (b=-0.36, p<0.05) points, while one additional point of the cognition score was associated with a decrease of 9%(OR=0.91, p<0.05) on the probability of presenting risk of depression. Additionally, being fully retired increased the probability of risk of depression by 37% (OR=1.37, p<0.05). The indirect effect of being retired on risk of depression was significant and indicated a decrease on risk of depression (IE=-0.03, p<0.05). Given that cognition score does mediate the relationship between retirement and risk of depression, this could serve as one mechanism for the increased risk of depression in retired individuals. Future studies should examine longitudinal data.