Linked Lives, Linked Health: Health Trajectories of Persons with Cognitive Impairment and Their Caregivers' Health

Project Summary More than 20% of people ≥65-years old living in the US are cognitively impaired, with diagnoses ranging from mild cognitive impairment (MCI) to dementia. Because of its progressive nature, as persons with cognitive impairment (hereafter “PCIs”) experience decline in cognitive function and other health outcomes, the health of

their caregivers or care partners (CG) may also be negatively impacted due to factors such as increased caregiving burden and stress. Determining how the health trajectories of all PCIs affect CG health outcomes over the course of cognitive decline will build a scientific foundation to design effective policies to reduce

caregiving cost and improve care quality. The proposed study will investigate the health trajectory patterns of PCIs and their relationships with CG health in the US. Data will be drawn from the annual National Health and Aging Trends Study in 2011–2019 coupled with data from the National Study of Caregiving in 2011, 2015, 2017, and potentially 2021 if available.

We will consider general health status, physical health, and psychological well-being for both PCIs and CGs. Our unique contribution to the field of dementia research is threefold: 1) our proposed study is the first to examine both PCI and CG health using a trajectory approach; 2) we will use high-quality population data; 3) we

will study the full spectrum of cognitive impairment, rather than only the most severe scenarios. We will first describe health trajectories among PCIs and examine how their trajectories predict CG health across time. For each health outcome, we will apply the single-trajectory Bayesian group-based trajectory

model (BGBTM) to identify distinct trajectory groups for PCIs and apply linear regression models to predict CG health. We will also determine how PCI health trajectories are related to CG health trajectories. Applying the dual-trajectory BGBTM, we will visually demonstrate how PCI and CG health trajectories are parallel in time

and estimate the probability of one trajectory pattern among CGs conditional upon one pattern among PCIs. Second, we will determine how the relationships examined above are moderated by caregiving and sociodemographic characteristics of CGs. We hypothesize that a distant relationship with PCIs, high-intensity

caregiving, a heavy caregiving burden, and social disadvantage are associated with adverse health outcomes and trajectories among CGs, and that these characteristics moderate the association between PCI health trajectories and CG health outcomes and trajectories. Finally, we will determine joint trajectories in PCI

cognition and other health outcomes, as well as the impact of PCI cognitive trajectories on CG health. Findings of this study will assist policymakers in understanding the health consequences of caregiving for PCIs, which will build a scientific foundation for the development of effective interventions to improve the

quality of care and reduce long-term care cost. Further, understanding the prognosis for various types of PCI and CG health trajectories may enable better preparation of caregiving and ultimately higher quality care.