PROJECT SUMMARY/ABSTRACT Dementia and other neurodegenerative diseases lead to profound cognitive, emotional, and functional impairments. As these diseases progress, the person with dementia (PWD) becomes increasingly dependent on a caregiver for functional, psychological, and economic assistance. It is well-documented that caring for a PWD is associated with considerable declines in health and well-being. Evidence suggests that many caregivers continue to experience these adverse effects for years after caregiving has ended (after the death of the PWD), though this area has received less attention. There is also striking variability in the extent to which caregivers experience these consequences – both during caregiving and after it has ended. Research on the sources of these individual differences has largely focused on factors related to the external environment, the PWD, or the caregiver that contribute to increased vulnerability in caregivers. Although research has linked PWD-caregiver relationship quality with declines in caregiver well-being, little is known about the specific interpersonal mechanisms that contribute to caregiver outcomes. For example, the positive emotional qualities of the PWD-caregiver social connection may buffer against the negative effects of caregiving stress, whereas lower quality social connection may drive increases in caregivers’ loneliness. In the proposed research, I will measure PWD-caregiver social connection through observational measures derived from basic affective science, including dyadic coding of emotional behavior in a laboratory-based interaction between PWDs and caregivers (Study 1; N = 186) and text analysis of emotional language caregivers used when describing a recent time they felt connected to the PWD (Study 2; N = 533). I will then determine the impact of PWD-caregiver social connection on caregivers’ health and well-being (Aim 1), both concurrently (in current caregivers) and longitudinally (in former caregivers, after caregiving has ended). Additionally, I will compare the relative strength of these associations between current and former caregivers, and I will investigate possible mechanisms (e.g., loneliness) through which PWD-caregiver social connection is linked to caregivers’ health and well-being. I will then evaluate the predictive value of PWD-caregiver social connection on caregivers’ health and well-being, above and beyond an optimal set of predictors identified through machine learning (Aim 2). The proposed dissertation research will prepare me to conduct future studies investigating the socioemotional mechanisms that may influence caregivers’ health trajectories. This F31 will provide the necessary support to accomplish the following research goals: (1) to further expertise in basic affective science methodologies (e.g., dyadic behavioral coding, text analysis); (2) to achieve advanced proficiency in statistical techniques for longitudinal and machine learning analyses; and (3)...