Social Determinants to Depressed Mood Symptoms and All-Cause Mortality in Older Adults

Michelle Lynne Cook, The University of Texas School of Public Health


The purpose of this dissertation was to explore the relationships between specific social determinants of health, depressed mood symptoms, and all-cause mortality in older adults. The Health and Retirement Study (HRS) was utilized for all three papers in this dissertation. The HRS was established in 1992 to provide data on the changing health and economic circumstances associated with aging at both individual and population levels. This nationally representative longitudinal survey is fielded every two years to more than 37,000 individuals aged 51 years and older. The HRS sample is replenished every six years with younger cohorts not previously represented. ^ The first paper of this dissertation reexamined the relationship between depressed mood symptoms and all-cause mortality among 51+ year old Americans utilizing a time-varying predictor (depressed mood symptoms) and potential time-varying confounders, such as marital status and health care utilization. Marginal structural models were utilized to determine the association. Consistent with prior research, depressed mood symptoms were associated with increased all-cause mortality rates; those who had all three symptoms had a 24% increase in odds of death compared to those who did not have any symptoms (p < 0.01). ^ The second manuscript of this dissertation investigated a causal mechanism between perceived discrimination, mastery, and depressed mood symptoms; mastery was tested as the potential mediator. Longitudinal data from the HRS, from 2006 to 2014, were utilized. It was hypothesized and confirmed that increased mastery attenuated the relationship between perceived discrimination and depressed mood symptoms. The percentage of the relationship between perceived discrimination and depressed mood symptoms that was mediated by mastery was 12% (95% CI: 8.4%, 16.8%, p < 0.01). ^ Connecting the first two papers, the final paper of this dissertation determined if mastery in 2010 was a mediator of the relationship between depressed mood symptoms in 2008 and all-cause mortality from 2011 to 2014. A SAS Macro was utilized to determine this relationship, in addition to the difference in coefficients method. Mastery did not mediate the association (2.2%, p > 0.05).^ This dissertation showed that older adults dealing with depressed mood symptoms continue to be a significant public health problem; between 1993 and 2014, 24.2% to 34.4% of HRS adults over the age of 50 reported one or more symptoms of depressed mood. Those with a higher number of depressed mood symptoms had a higher all-cause mortality rate than those who did not have any symptoms. In addition, increased perceived discrimination was associated with an increased number of depressed mood symptoms, both concurrently and temporally; moreover, mastery partially mediated the relationship. In contrast, however, mastery did not mediate the relationship between depressed mood symptoms and all-cause mortality. This suggests that depressed mood symptoms, measured on a week’s worth of symptoms, may fluctuate too frequently to be considered a constant chronic stressor that is associated with an increased all-cause mortality hazard rate. Unless viable interventions or targeted policies are developed to reduce symptoms of depressed mood among older adults, this problem, and the deaths that result because of it, will continue to grow as the proportion of Americans over the age of 50 increases.^

Subject Area

Biostatistics|Behavioral sciences|Epidemiology

Recommended Citation

Cook, Michelle Lynne, "Social Determinants to Depressed Mood Symptoms and All-Cause Mortality in Older Adults" (2017). Texas Medical Center Dissertations (via ProQuest). AAI10272921.