Cardiovascular diseases (CVD) are among the most common causes of disability and mortality worldwide, encompassing a wide range of disorders including myocardial infarction, coronary heart disease, stroke and heart failure.
Prior meta-analyses have found that depressive symptoms and loneliness are independent risk factors for the incidence of CVD. Most studies examining depression and incident CVD assess depressive symptoms at only a single time point. However, depressive symptoms are not static but rather dynamic over time. Furthermore, the pathways linking depressive symptoms to CVD include both short-term mechanisms and longer-term increases in risk, thus different longitudinal trajectories of depressive symptoms may be associated with differential risk of incident CVD. Preliminary studies have shown that loneliness is also associated with higher risk for CVD, yet evidence is limited. Among the few studies examining loneliness and incident CVD, most have primarily focused on assessing loneliness at one time point. Yet, loneliness is not necessarily stable or constant throughout old age. The effects of changes in loneliness over time on risk of incident CVD are not well understood.
In Chapter 1, I examined the association between depressive symptom patterns measured across 4 time points on the risk of incident stroke over a 10-year follow-up period in the Health and Retirement Study (HRS). Depressive symptom patterns were categorized as consistently low, decreasing, fluctuating, increasing and consistently high. Trajectories of depressive symptoms patterned by high levels of symptoms at multiple time points, were associated with increased stroke risk. However, a trajectory with depressive symptoms that started high but decreased over time was not associated with a higher risk of stroke. In Chapter 2, I examined the association between loneliness and risk of incident stroke in the HRS, through two sets of analyses assessing baseline loneliness scores and changes in loneliness over two time points. Baseline loneliness was associated with increased risk of incident stroke, and chronically lonely respondents over time were at higher risk of stroke compared to those who had chronically low loneliness scores. Increasing or decreasing loneliness patterns were not associated with risk of incident stroke. In Chapter 3, I explored the association between depressive symptom changes across two time points and incident CVD in Japan. We defined four depressive symptoms (consistently low, remitting, recent onset and consistently high) across the two exposure assessments. Elevated depressive symptoms were associated with increased risk of incident CVD in Japan across all depressive symptom change groups, compared to the consistently low group. In summary, findings from each of these chapters contribute new scientific evidence regarding the relationship between depressive symptoms, loneliness and incident CVD. Furthermore, these findings may help inform policies on preventative efforts to reduce risk of incident CVD among middle-aged and older populations.