Unmasking Caregiver Heterogeneity: Latent Class Analysis of Abuse Tendency Subtypes for Alzheimer’s Disease Caregivers in China

Abstract Background: Alzheimer's disease patients face higher risks of caregiver abuse and caregiver burden than healthy older adults,. However, the association between abuse tendency by AD caregivers and their burden remains understudied. This study aimed to investigate the association between caregivers' abusive tendencies and burden levels in Alzheimer's disease through latent class analysis. Methods: A total of 541 Alzheimer's disease caregivers from long-term care facilities in Zhejiang Province, China, were included in this cross-sectional study. Demographic characteristics、abusive tendencies and caregiver burden were assessed using scale. Heterogeneous subtypes of abusive tendencies were identified through latent class analysis, and their associations with caregiver burden were examined using multivariate regression models. Rseults: The prevalence of caregiver abuse tendency toward Alzheimer's patients in Zhejiang long-term care facilities was 39.4%, categorized into three latent classes: Class 1 ("high abuse tendency-low coping deficiency"), Class 2 ("overall low abuse tendency"), Class 3 ("high coping deficiency-low abuse tendency"). Significant differences (P<0.05) emerged among classes in marital status, chronic conditions, daily care duration, caregiving training, and patient neuropsychiatric symptoms. Compared with Class 2, caregivers in Class 1 had significantly higher odds of providing more than 12 hours of daily care (OR=0.252, P<0.001), whereas those in Class 3 had lower odds of having received Alzheimer's caregiving training (OR=0.435, P=0.002). Caregiver burden scores also differed significantly across classes (F=12.086, P<0.001). Conclusions: Caregivers for Alzheimer's patients exhibit distinct abuse tendency categories with caregiving burdens. Healthcare providers should implement targeted training programs to alleviate these burdens, reduce the incidence of abuse tendency.