Research That Matters (January 17 - 20, 2008)

PURPOSE: Funded by an RO1 research grant from the National Institute on Aging, the second Family Relations in Later Life project (FRILL II) has completed phase one of the five year study of long-term caregiving to older persons in Athens, GA, Tuscaloosa, AL, and Pittsburgh, PA. Based upon the responses from 419 caregivers (CGs), and using standardized measurements, we were able to explore the relations between components of caregivers' (CG) mental health (depression, anxiety, and resentment), anger, and self-reports of potentially harmful behavior (PHB) to older care recipients (CRs). Previous investigations that examined anger as a mental health component suggested a more complex relationship with other mental health (MH) variables. In the model presented in this paper, anger was examined as a potential mediator or moderator on MH and PHB.

METHODS: Using various sampling methods, including Random Digit Dialing, CG-CR dyads, who met inclusion criteria, were recruited from the three areas of the U.S. Eligible CGs resided in the same household as the CR and provided at least one basic Activity of Daily Living (ADL) or two Instrumental Activities of Daily Living (IADL) to the CR who was at least 60 years of age. Trained project staff independently interviewed each CG and CR dyad in separate rooms within their home. This presentation focuses solely on CGs.

RESULTS: We conducted a Sobel test for mediation and zero-centered interaction variables in a regression model for moderation. Anger was found to be both a mediator and a moderator on the relationship between resentment and PHB in these separate tests. Using a test for moderated-mediation developed by Preacher, Rucker, and Hayes, anger was observed to have a conditional indirect effect such that, at lower levels of resentment, anger did not mediate the relationship between resentment and PHB. But, at moderate and high levels of resentment, anger mediated the relationship between resentment and PHB. Similar findings were observed with depression. Anger was found to be a mediator, but not a moderator, between anxiety and PHB.

IMPLICATIONS: Our findings suggest that anger has a complex relationship with other mental health variables; interventions that focus on reducing anger may consequently reduce the potential of the caregiver to harm the care recipient. In addition, the assessment of anger and resentment may assist in identifying those who are not well-suited for caregiving roles or who may need additional resources and supports for this task. Finally, future research should measure anger as a multidimensional construct.