Elizabeth C Pino1, Karla Damus2, Brian Jack3, David Henderson4, Snezana Milanovic5, Bindu Kalesan6. 1. Center for Translational Epidemiology and Comparative Effectiveness Research, Boston University School of Medicine, 801 Massachusetts Ave Boston, MA, USA. Electronic address: epino@bu.edu. 2. Department of Family Medicine, Boston University School of Medicine/Boston Medical Center, Dowling 5, 1 Boston Medical Center Place, Boston, MA, USA. Electronic address: damusk@bu.edu. 3. Department of Family Medicine, Boston University School of Medicine/Boston Medical Center, Dowling 5, 1 Boston Medical Center Place, Boston, MA, USA. Electronic address: bjack@bu.edu. 4. Department of Psychiatry, Boston Medical Center, 720 Harrison Avenue, Boston, MA, USA. Electronic address: David.Henderson@bmc.org. 5. Department of Psychiatry, Boston University School of Medicine, 720 Harrison Avenue, Boston, MA, USA. Electronic address: Snezana.Milanovic@bmc.org. 6. Center for Translational Epidemiology and Comparative Effectiveness Research, Boston University School of Medicine, 801 Massachusetts Ave Room 475, Boston, MA, USA. Electronic address: kalesan@bu.edu.
Abstract
BACKGROUND: The complex association between socioeconomic status (SES) and depressive symptoms is not entirely understood and the existing literature does not address the relationship between early-life SES and later-life depression from a life-course perspective, incorporating mediating events. METHODS: Using data from the Wisconsin Longitudinal Study, we employed structural equation modeling to examine how SES measured at age 18 affects depressive symptoms at age 54 directly and through mediating variables college graduation, marriage, and household income level at age 36. RESULTS: The total effect of adolescent SES on later-life depressive symptoms is largely mediated through college graduation. Our final model was driven by the effects of women. The variables contributing most to depressive symptoms in women were the direct effect of being raised in a home with a low SES and the indirect effect of low adolescent SES mediated through non-completion of college. LIMITATIONS: Cohort was exclusively comprised of white, high school graduates born in 1939 (± 2 years). In our analysis we assume that missing values are missing at random (MAR); however, attrition both from death (excluded from our population) and from non-response could be associated with depression, i.e. missing not at random (MNAR). CONCLUSIONS: This study demonstrates the impact of completion of college, particularly among women, and supports the social mobility hypothesis to explain the relationship between adolescent socioeconomic circumstances and late-life health.
BACKGROUND: The complex association between socioeconomic status (SES) and depressive symptoms is not entirely understood and the existing literature does not address the relationship between early-life SES and later-life depression from a life-course perspective, incorporating mediating events. METHODS: Using data from the Wisconsin Longitudinal Study, we employed structural equation modeling to examine how SES measured at age 18 affects depressive symptoms at age 54 directly and through mediating variables college graduation, marriage, and household income level at age 36. RESULTS: The total effect of adolescent SES on later-life depressive symptoms is largely mediated through college graduation. Our final model was driven by the effects of women. The variables contributing most to depressive symptoms in women were the direct effect of being raised in a home with a low SES and the indirect effect of low adolescent SES mediated through non-completion of college. LIMITATIONS: Cohort was exclusively comprised of white, high school graduates born in 1939 (± 2 years). In our analysis we assume that missing values are missing at random (MAR); however, attrition both from death (excluded from our population) and from non-response could be associated with depression, i.e. missing not at random (MNAR). CONCLUSIONS: This study demonstrates the impact of completion of college, particularly among women, and supports the social mobility hypothesis to explain the relationship between adolescent socioeconomic circumstances and late-life health.
Authors: Y A de Vries; M Ten Have; R de Graaf; S van Dorsselaer; N M P de Ruiter; P de Jonge Journal: Epidemiol Psychiatr Sci Date: 2019-12-16 Impact factor: 6.892