Caregiving Intensity and Caregiver Burden Among Dementia Caregivers: The Moderating Roles of Social Support

Abstract Objective: Using the stress-coping theory, the aims of the present study were to test what levels of caregiving intensity (hours actually spent on caregiving every day) posed the most negative influence on caregiver burden as well as how social support moderated such associations among dementia caregivers. Methods: Data from the baseline assessment of the Resources for Enhancing Alzheimer’s Caregiver Health (REACH II) (N = 637) were used. Caregiver burden (12-item Zarit caregiver burden scale), caregiving intensity (caregiving hours), and social support (Lubben social network, received support, satisfaction with support, and negative interactions) were the main measurements. Separate multivariate regression models were conducted with Stata 16. Results: The results showed that the relationships between caregiving hours and caregiver burden were nonlinear after controlling all of the socio-demographic variables. Further analyses showed that when caregiving hours reached 13.50 hours per day, the levels of burden were the highest. In addition, received social support, satisfaction with social support, and social network significantly moderated the relationship between caregiving hours and caregiver burden among dementia caregivers when they were examined separately. However, only social network played a significant moderator role when examining the four social support indicators simultaneously. Discussion and conclusion: These findings suggest the need for programs and practices on educating caregivers regarding how to identify, approach, and gain social support/s, especially in how to broaden the caregivers’ social network while caring for a family member with dementia.

organizations and perform a wide range of professional and lay functions. Participant data indicate that the program has appealed to a wide variety of learners including participants who serve caregivers generally (60.8%) and grandparents raising grandchildren specifically (81%). A small majority (55.6%) of the agency-based learners reported serving, on average, more than 40 grandfamilies annually. Self-reported learning levels were notable ranging from a mean low of 3.46 out of 4 points (N = 157, SD = 0.59) for the volunteer recruitment and mentorship programming module to a mean high of 3.79 (N = 167 ,SD = 0.45) for the caregiver self-care module. Evaluation results from the first seven learner cohorts underscore the efficacy of program content as well as the utility of performing an initial program needs assessment to guide curriculum development. Practice implications for future continuing education efforts targeting grandfamilies professionals and lay leaders include: the need for easily accessible online education in combination with supplemental training opportunities addressing topics such as the long-term impact of substance use disorder and trauma combined with locally relevant content on grandfamilies and legal resources. People living with dementia (PwD) can often reliably selfreport their health; yet, there are limited data on their and their primary caregiver's self-reported overall health (excellent, very good, good, fair, and poor). We used data from the Aging, Demographics, andMemory Study (2001-2009) to quantify the association between PwD's cognitive impairment (Minimental State Exam), physical limitations (scale [0-10] of activities of daily living), and behaviors (scale [0-12] of behavioral symptoms on the Neuropsychiatric Inventory Questionnaire) and self-reported health. We estimated two ordered logistic regressions estimating: 1) PwD's self-reported health (analyzed n=308); 2) primary caregiver's self-reported health (analyzed n=135; 173 PwD did not have primary caregiver in the survey). We controlled for the PwD demographics, chronic conditions, and if they lived in the community. The regression estimating caregiver's self-reported health also controlled for the caregiver's relationship to the PwD, and whether the caregiver lived with the PwD. PwD's self-reported health was lower (4% excellent; 16% very good; 22% good; 30% fair; 30% poor) than caregivers (14% excellent; 27% very good; 32% good; 23% fair; 4% poor). For PwD, one-additional physical limitation, but not cognition or behavior, was associated with 1.15 (95%CI: 1.01,1.30) times greater odds of self-reporting poor health compared to all other categories. For caregivers, one-additional behavior, but not cognition or physical limitations, was associated with 1.17 (95%CI: 1.01,1.37) times greater odds of self-reporting poor health. For PwD, interventions targeting physical limitations may increase self-reported health, but for caregivers, interventions targeting behavioral symptoms may increase self-reported health.

CAREGIVER-PROVIDER COMMUNICATION ABOUT PAIN IN PERSONS WITH DEMENTIA
Catherine Riffin, 1 Sylvia Lee, 2 M. Cary Reid, 3 Keela Herr, 4 and Karl Pillemer, 5 1. Weill Cornell Medical College,London,England,United Kingdom,2. Weill Cornell Medicine,New York,New York,United States,3. Weill Cornell Medical College,New York City,New York,United States,4. University of Iowa,Iowa City,Iowa,United States,5. Cornell University,Ithaca,New York,United States Pain in older persons with dementia (PWD) is severely under-detected and under-managed. Family caregivers can play an important role in addressing these disparities by acquiring the requisite skills to communicate PWD's pain symptoms and behaviors to health care providers, but little is known about how caregivers of dementia patients and their providers approach such pain-related discussions. We employed qualitative methods to explore the perspectives of family caregivers of PWD (n=18) and health care providers (geriatricians, general internists, neurologists, emergency room physicians) involved in PWD's treatment (n=16) regarding pain communication. We specifically focused on participants': 1) priorities and expectations for communicating about pain and dementia, 2) challenges to communicating about pain and dementia, and 3) strategies and recommendations for optimizing communication about pain and dementia. Analyses revealed that caregivers and health care providers expected to receive accurate, detailed information from one another, but uncertainty in both groups around differentiating pain behaviors from dementia symptoms acted as a barrier to effective information exchange. Additional challenges to productive pain-related discussions were identified by caregivers, including provider fatalism and lack of interpersonal skills, and by providers, including patient-caregiver disagreement about pain symptoms and unreliable caregiver reporting. Participants endorsed using practical approaches, such as pain scales and logs, as well as rapport-building strategies, such as affirmation of caregivers' input, to facilitate collaborative discussions. Objective: Using the stress-coping theory, the aims of the present study were to test what levels of caregiving intensity (hours actually spent on caregiving every day) posed the most negative influence on caregiver burden as well as how social support moderated such associations among dementia caregivers. Methods: Data from the baseline assessment of the Resources for Enhancing Alzheimer's Caregiver Health (REACH II) (N = 637) were used. Caregiver burden (12-item Zarit caregiver burden scale), caregiving intensity (caregiving hours), and social support (Lubben social network, received support, satisfaction with support, and negative interactions) were the main measurements. Separate multivariate regression models were conducted with Stata 16. Results: The results showed that the relationships between caregiving hours and caregiver burden were nonlinear after controlling all of the socio-demographic variables. Further analyses showed that when caregiving hours reached 13.50 hours per day, the levels of burden were the highest. In addition, received social support, satisfaction with social support, and social network significantly moderated the relationship between caregiving hours and caregiver burden among dementia caregivers when they were examined separately. However, only social network played a significant moderator role when examining the four social support indicators simultaneously. Discussion and conclusion: These findings suggest the need for programs and practices on educating caregivers regarding how to identify, approach, and gain social support/s, especially in how to broaden the caregivers' social network while caring for a family member with dementia.

CHILDREN RAISED BY GRANDPARENTS OR NONGRANDPARENTS: WHICH HAVE GREATER ODDS OF HIGH ACADEMIC PERFORMANCE?
Tamar Shovali, 1 Melissa Bright, 2 and Kerstin Emerson, 3 1. Eckerd College,St Petersburg,Florida,United States,2. University of Florida,Gainesville,Florida,United States,3

. University of Georgia, Athens, Georgia, United States
There is a 19-fold greater likelihood that children removed from parental care will be raised by a grandparent than any other caregiver. Theorists and practitioners highlight the importance of monitoring academic progress to understand the benefits of at-risk youth living in grandfamilies. Using a nationally representative dataset we examined academic performance for children in three caregiver (N = 814) categories: Grandparent (73.1%), Foster parent (12.7%), Other (nonfoster, nonkin/nonfoster; 14.2%), with significance testing across groups. Children were between 6-17 years with grandfamilies and foster families caring for significantly younger children compared to the "other" group. Overall, 76% of children were reported to have high academic performance in math and 79.6% had high academic performance in reading/writing. Grandparents were caring for a significantly higher proportion of non-Hispanic White children with statistically significantly higher reported academic performance in math and reading/ writing compared to nongrandparents. Logistic regression model A showed for both foster parent (AOR 0.57, CI: .35-.91) and other (AOR 0.55, CI: .35-.86) caregiver groups were significantly negatively correlated with high math performance compared to grandparents. Model B showed the same statistically significant and negative relationship to reading/writing performance outcomes for foster parent (AOR 0.56, CI: .02-.35) and other (AOR 0.51, CI: .01-.32) caregivers compared to grandparents. Controlling for relevant caregiver and child variables both models suggest that children living with grandparents have 55% greater odds of high academic performance compared to children raised by nongrandparents. Findings support placement of children with grandparents. Supporting grandfamilies with appropriate social services will be reviewed. Research indicates that family caregivers of individuals living with dementia are at risk for high levels of stress, depression, physical health declines, and illness. The health and well-being of family caregivers is critically important to a long-term care system that is dependent on them to continue their caregiving role. In-depth individual and focus group interviews of 16 dementia caregivers were conducted to explore the emotional experiences of caregiving stress during transitions of individuals living with dementia to a higher level of care. Data were transcribed verbatim, checked for accuracy, and analyzed by at least two members of the research team. Line-by-line coding, memo writing, and constant comparative analyses were conducted until redundancy, when no new themes were discovered. Caregivers described various levels of feeling overwhelmed and symptom progression leading to the move to a nursing facility. Social isolation featured prominently, with caregivers describing a gradual erosion of their social network and socializing opportunities because of their caregiving responsibilities and the care recipient's deteriorating symptoms. Caregivers described feeling isolated and stigmatized. One caregiver said, "…you're being less invited, you're being less involved. People don't know how to deal with you…I don't know if they become the pariah or I become the pariah." At the same time, maintaining social connections and having help with caregiving featured prominently in the coping mechanisms described. The health of caregivers is equally as important as the person living with dementia, and programs, interventions, and resources should be a priority for supporting families through transitions.

EXPLORING THE STRESS AND STRENGTHS ARISING FROM THE COMPLEXITIES EXISTING WITHIN GRANDFAMILIES
Catherine Tompkins, 1 Alice Zic, 1 Ellie Carlson, 1 Ali Purvis, 1 Loriena Yancura, 2 and Danielle Nadorff, 3 1. George Mason University,Fairfax,Virginia,United States,2. University of Hawai'i,Honolulu,Hawaii,United States,3. Mississippi State University,Mississippi State,Mississippi,United States The opioid crisis and other social problems continue to increase the number of grandparent-headed households in the U.S. There are challenges and benefits that result from grandparents parenting their grandchildren. Grandparents often report the joy in watching their grandchildren grow but also report on the complexities that may lead to stress. Two-hundred forty-one grandparents were recruited using Qualtrics' Online Panel Service. In addition to a standardized perceived stress scale and demographic questions, participants responded to open-ended questions related to the benefits and challenges of residing within a grandparentheaded household. This presentation focuses on comparing demographics, perceived stress, benefits and challenges of 108 current grandparent caregivers to 133 grandparents who were no longer the head of household at the time of the survey. Grandparents who currently were raising their grandchildren had a higher perceived stress score (p=.03) compared to grandparents who had raised their grandchildren in the past. An exploration of the demographic variances and responses to the open-ended questions, will begin to explain this statistically significant difference in reported stress. An additional complexity arising for a subsample of 10 current