Rationale: Alzheimer’s caregivers are at increased risk for cardiovascular disease (CVD) and exhibit higher levels of vascular dysfunction compared to non-caregivers. Sleep has been implicated in the pathogenesis of CVD, however this relationship is not fully understood. Behavioral Activation therapy has shown promise for reducing caregiver distress and decreasing circulating levels of IL-6, and improvements in sleep may mediate these effects. The purpose of the current study is to examine the cross-sectional and longitudinal associations of subjective and objective measures of sleep with markers of CVD risk in Alzheimer’s caregivers and test the efficacy of a brief Behavioral Activation intervention for improving sleep.Design: This dissertation used archival data from two studies. Study 1 was a longitudinal study in which 126 spousal caregivers and 60 non-caregiver controls were assessed annually over 5 years. Study 2 was a randomized controlled trial of Alzheimer’s caregivers testing the efficacy of a brief Behavioral Activation intervention (n = 75) versus an Information Support condition (n = 76) to reduce caregiver distress (i.e., symptoms of depression, negative affect, and sleep disturbance) and markers of CVD risk. Subjective sleep quality and quantity were assessed using the Pittsburgh Sleep Quality Index and Insomnia Severity Index. Objective sleep data including circadian rhythm profiles were obtained from actigraphy records using Actiwatch. Data for markers of CVD risk were obtained through several methods including ultrasounds of the brachial and carotid arteries, blood pressure and baroreflex sensitivity, and blood and urine samples for catecholamine levels and markers of endothelial function. Using a combined sample, bivariate associations and multiple linear regression models were fitted to test associations between sleep variables and markers of CVD risk. Using data from Study 1, longitudinal associations were tested using multilevel models to examine associations between sleep measures and markers of CVD risk over time. Using data from Study 2, multilevel models and ANOVAs were used to test the impact of a brief Behavioral Activation intervention on various measures of sleep. Results: Regression analyses of baseline associations showed several significant associations between circadian activity rhythms and CVD markers. Caregivers with less robust circadian rhythms exhibited impaired baroreflex sensitivity and increased levels of tumor necrosis factor alpha, von Willebrand factor antigen, and plasminogen activator inhibitor 1. Longitudinal analyses revealed that after controlling for caregiver physical activity, stress, and sex; percent sleep was significantly associated with resting heart rate and norepinephrine. Caregivers with higher percent sleep experienced lower resting heart rate at baseline but showed a slower decrease over time. Caregivers with higher percent sleep also exhibited higher levels of norepinephrine at baseline; there was no significant effect on norepinephrine levels over time. There was not a significant effect of the intervention on sleep outcomes. Conclusions: Cross-sectionally, caregivers with less robust circadian activity rhythm profiles exhibit markers suggestive of increased CVD risk. However, this study revealed little evidence of longitudinal associations between measures of sleep and markers of CVD risk over time. Furthermore, the Behavioral Activation intervention was not more effective than the control condition at improving caregivers’ sleep.

Rationale: Alzheimer’s disease caregivers demonstrate significant elevations in depression compared with noncaregivers. Addressing caregiver depression is of high public health importance due to its ties with overall wellbeing, increased risk for cardiovascular diseases (CVD), and ability to sustain caregiving duties. Improving caregiver mental and physical health may not only decrease healthcare costs, but it may also delay institutionalization of Alzheimer’s disease patients. Despite existing behavioral interventions to address caregiver depression and physical health issues, little is known regarding which interventions to recommend based on individual caregiving situations. To improve the impact of caregiver interventions, it is essential to understand the mechanisms through which interventions achieve reductions in depression and CVD risk, as well as the caregiving situation-related factors that may impact response to interventions.

Design: The proposed study employs a data analytic focus on mediators and moderators of a previously completed randomized clinical trial testing the efficacy of the Pleasant Events Program (PEP). A sample of spousal Alzheimer’s Disease caregivers (N = 100; 74% female) was randomized to receive either a Behavioral Activation therapy, named the Pleasant Events Program (PEP), or to a time-equivalent Informational Support control condition (n = 51). The PEP intervention emphasized pleasant events scheduling and reductions in avoidance behaviors (i.e., lack of engagement in activities) in the context of ongoing caregiver responsibilities, whereas the Informational Support intervention emphasized supportive listening and providing information. Participants were assessed at baseline and at 6-weeks (i.e., post-intervention). Co-primary outcomes were CVD risk marker interleukin-6 and depression as measured by the Center for Epidemiologic Studies-Depression Scale. Higher levels of interleukin-6 are implicated in inflammatory processes, constituting higher risk for CVD. The PEP efficacy study reported significant reductions in both depression and interleukin-6 at 6-weeks follow-up. Multiple regression models and the Monte Carlo Method for Assessing Mediation were used to test change in pleasant events, activity restriction, and personal mastery as mediators of depression after PEP. Moreover, depression was tested as a mediator between PEP and reductions in CVD risk markers interleukin-6. Secondly, models were used to investigate constructs measured via standardized assessments (e.g., care recipient disruptive behaviors) as moderators of PEP outcomes.

Results & Conclusions: The current study found that change in personal mastery, activity restriction, and pleasant activities were not significant mediators of change in depression after PEP. Moreover, changes in depression did not mediate changes in cardiovascular risk marker interleukin-6. Caregiver social support, current number of hours working, vulnerability, and disruptive behaviors of care recipients did not significantly moderate treatment response. Higher baseline depression significantly predicted greater reduction in depression in both treatment conditions. The current study was limited in power to detect mediators and moderators of depression and CVD risk change after PEP. Potential caregiver-specific factors that may moderate response to PEP and mediators that account for changes in depression and CVD risk after PEP are still unknown. Thus, future intervention studies utilizing a Behavioral Activation therapy such as PEP should utilize oversampling methods to investigate mediators and moderators of depression and interleukin-6 change.

This study examined an activity restriction/pleasurable activities mismatch model for psychosocial and health-related outcomes. A total of 108 spousal caregivers of patients with Alzheimer's Disease (AD) were assessed for their experience of social and recreational activities over the past month as well as their perception of how restricted they were for engaging in social and recreational activities. Participants were divided into three groups based on their reported activities and activity restriction: HPLR=High Pleasant Events+Low Activity Restriction (i.e., reference group; N=28); HPHR/LPLR=Either High Pleasant Events+High Activity Restriction or Low Pleasant Events+Low Activity Restriction (N=43); LPHR=Low Pleasant Events+High Activity Restriction (N=37). We hypothesized that participants reporting low pleasant events combined with high activity restriction (LPHR) would demonstrate greater disturbance relative to other two groups in multiple outcome domains, including: (a) greater mood disturbance, (b) greater use of negative coping factors, (c) reduced use of positive coping strategies, (d) reduced report of psychological resource factors (e.g., personal mastery, self-efficacy), and (e) increased report of subjective health difficulties (e.g., sleep disturbance). Results generally supported our hypotheses, suggesting that assessment of both constructs is important for best predicting quality of well-being in AD caregivers, and potentially for establishing maximal effect in behavior therapy for caregivers.