An Ecologically-Informed Approach to Caregiver Burden (WITHDRAWN)

Objective: Caregiver burden has been associated with a range of poor health outcomes, including increased mortality. Although multiple risk factors for burden have been identified, risks are often attributed dichotomously to the caregiver or the care recipient. This study is informed by an ecological approach, including environmental/contextual factors to better understand which aspects of the caregiving dyad risk for burden can be attributed to, providing a starting point for targeted interventions.

Methods: This study uses a subset of data from the Los Angeles Caregiver Resource Center (LACRC), one of 11 such centers distributed across California since 1989. Data for this pilot study was entered from caregiver intake, assessment, and reassessment forms from 2014-2016 (n= 51). Information was originally collected by clinical social workers at the LACRC, entered by student interns, and provided to the research team. To be eligible for LACRC services, caregivers either cared for someone aged 60 and older with dementia and/or multiple chronic conditions or a younger person with a memory impairment. Burden was assessed using the Zarit Burden Index, and risk factors included in the analysis were based on those identified in previous literature as well as demographic variables for the caregiver and recipient. Following data reduction using an ordinary least squares regression, significant variables were organized by the characteristics of the caregiver, care recipient, and contextual factors surrounding the dyad (e.g. hours of paid/unpaid care received, hours spent caregiving) in a sequential model. The first model included caregiver characteristics, the second added care recipient characteristics, and, finally, contextual factors were included.

Results: Results indicate that characteristics of the caregiver were the most influential in explaining variance in caregiver burden. Specifically, in the final model (R²=.757), higher caregiver depression scores were associated with greater burden (b=1.104, t=4.20, p=.007). Employment status, too, was associated with burden, such that caregivers who were not working were more likely to experience higher levels of burden (b=-.257, t=-2.22, p=.039). Social support, considered a contextual factor in this model, was also significant; caregivers reporting they received “somewhat less” help than they needed were more likely to experience burden than those who did not need additional help (b=1.103, t=3.66, p=.002). Interestingly, care recipient factors were not significant in explaining variance in burden.

Discussion: An ecologically-informed model provided a novel way to understand caregiver burden. These findings have implications for practice at CRCs and similar services. First, these findings suggest that burden may not be the result of objective care circumstances (e.g. care recipient functional status), but rather the caregiver’s characteristics. Based on these findings, future research using longitudinal data is needed to determine whether targeting caregiver depression would reduce burden. Furthermore, given the minimal relationship with contextual factors in this model, it is possible that providing additional services to the dyad is not likely to decrease burden unless these impact caregiver psychosocial wellbeing. Following this pilot, next steps will include adding additional cases (n=200), as well as variables for specific LACRC services utilized by caregivers as contextual factors that may impact burden.