Patient's disability and caregiver burden among Chinese family caregivers of individual living with schizophrenia: mediation effects of potentially harmful behavior, affiliate stigma, and social support

The study addresses the growing global and national burden of mental disorders, particularly schizophrenia in China, where limited mental health resources and cultural expectations place caregiving responsibilities largely on families. Caregiver burden encompasses psychological, social, physical, and financial strain that can affect care quality and patient recovery. Prior work has identified disease-related factors (e.g., disability, illness duration, aggressive behaviors) and caregiver-related factors (e.g., low social support, perceived stigma) associated with caregiver burden, but the mechanisms linking these factors are unclear. Guided by Pearlin’s caregiver stress model, the authors hypothesized: (a) patient disability would directly increase caregiver burden; and (b) patient-related (potentially harmful behavior) and caregiver-related stressors (affiliate stigma and social support) would mediate the association between disability and caregiver burden among Chinese family caregivers of adults with schizophrenia.

Prior studies indicate that higher patient disability, longer illness duration, and aggressive behaviors are associated with greater caregiver burden in schizophrenia. Caregiver characteristics, such as low perceived social support and higher perceived or affiliate stigma, also correlate with increased burden. Pearlin’s caregiver stress model posits that patient behavioral problems generate secondary stressors (e.g., stigma), exacerbating caregiver strain. Evidence suggests family caregivers are frequent targets of patient-initiated violence, which is often overlooked but contributes substantially to burden. Social support has been shown to buffer stress and reduce burden, though it may be undermined by illness severity and social isolation. These findings motivated examination of how disability relates to burden via potentially harmful behavior, affiliate stigma, and social support.

Design and setting: Cross-sectional study conducted from April 2021 to March 2022 across four Chinese cities (Wuhan, Changsha, Guangzhou, Shenzhen). Ethics approval was granted by the Human Research Ethics Committee of Central South University (XYGW-2021-41); written informed consent obtained. Sampling and participants: All administrative districts were included as the sampling frame. Two districts per city were randomly selected; 50% of community health centers in each selected district were randomly chosen (63/122 invited; 45 participated). Trained researchers conducted face-to-face interviews at centers using paper questionnaires. Of 1880 eligible patient–caregiver dyads invited, 972 patients enrolled; 104 without family caregivers were excluded; among 868 with caregivers, 375 caregivers refused. Final analytic sample: 493 patient–caregiver dyads with complete data. Participants received RMB 20 as transportation allowance. Inclusion criteria: Patients—aged ≥18, ICD-10 diagnosis of schizophrenia by certified psychiatrists, able to communicate in Chinese, living with family; Caregivers—aged ≥16, primary family caregiver living with the patient, able to communicate in Chinese. Dyads with missing data were excluded. Measures: - Caregiver burden: Zarit Burden Interview (ZBI-22), 22 items (0–4), total 0–88; higher scores indicate greater burden; Chinese version Cronbach’s α=0.96. - Patient disability: 12-item WHO Disability Assessment Schedule 2.0 (WHODAS 2.0), six domains scored 0–4, total 0–48; higher indicates greater disability; α=0.95. - Affiliate stigma: Affiliate Stigma Scale (ASS), 22 items across cognitive, affective, behavioral domains, item score 0–4; higher mean indicates higher stigma; α=0.96. - Social support: Multidimensional Scale of Perceived Social Support (MSPSS), 12 items (1–7), total 12–84; higher indicates greater support; α=0.98. - Patient potentially harmful behavior (PHB): Adapted Potentially Harmful Behavior Scale items (3 psychological: yelling/threatening/verbal abuse; 4 physical: hitting/slapping/shaking/handling roughly/fear of being hit), frequency 0–4; higher sum indicates more frequent PHB; α=0.96. Covariates: Patient gender, age, illness duration, education, marital status, work status; Caregiver gender, age, caregiving duration, education, marital status, chronic disease, work status, kinship. Statistical analysis: SPSS 26.0 and AMOS 23.0 used. Group differences in burden assessed via t-tests/ANOVA. Linear regression estimated associations with caregiver burden (dependent variable); Model 1 unadjusted; Model 2 adjusted for caregiver age, education, work status, kinship, caregiving duration; and patient gender, illness duration, marital status, work status. Pearson correlations examined relationships among disability, PHB, affiliate stigma, social support, and burden. Structural equation modeling (CFA and path analysis) tested the mediation model; fit indices: χ2, RMSEA<0.08, CFI and NNFI>0.85. Mediation tested via bootstrap (5000 resamples) with bias-corrected 95% CIs; indirect effects significant if CI excluded zero. Proportion mediated (PM) calculated for paths.

Sample characteristics: 493 dyads analyzed. Patients: mean age 46.59 (SD 13.99), 53.14% female, 74.04% illness duration >10 years, 46.86% married/cohabiting. Caregivers: mean age 57.26 (SD 12.95), 49.09% female, 77.08% caregiving >10 years, 86.62% married/cohabiting, 39.96% primary education, 31.03% full-time job. Bivariate associations: Higher caregiver burden associated with caregiver older age, lower education, unemployment, being a parent, longer caregiving duration; and patient male sex, longer illness duration, single status, unemployment (p<0.05). Adjusted linear regression (Model 2): - Patient disability: B=0.616 (95% CI 0.479–0.753), p<0.05. - Patient PHB on caregiver: B=0.474 (95% CI 0.232–0.716), p<0.001. - Caregiver social support: B=−0.079 (95% CI −0.158 to −0.002), p<0.05. - Caregiver affiliate stigma: B=13.045 (95% CI 10.227–15.864), p<0.001. Correlations (selected): Disability correlated with PHB (r=0.322), affiliate stigma (r=0.254), lower social support (r=−0.298), and burden (r=0.539); PHB with burden (r=0.391); affiliate stigma with burden (r=0.512); social support with burden (r=−0.348) (all p<0.05). Mediation model fit: χ2(109)=351.747, p<0.001; CFI=0.979; NNFI=0.973; RMSEA=0.06. Direct effect: Disability → caregiver burden β=0.371 (p<0.001). Indirect effects (bootstrap, 5000 resamples): - Disability → PHB → burden: β=0.049 (95% CI 0.024–0.076), p<0.01; PM≈9.12%. - Disability → PHB → affiliate stigma → burden: β=0.026 (95% CI 0.015–0.040), p<0.01; PM≈4.84%. - Disability → affiliate stigma → burden: β=0.036 (95% CI 0.003–0.068), p<0.05; PM≈6.70%. - Disability → social support → burden: β=0.026 (95% CI 0.004–0.051), p<0.05; PM≈4.84%. - Disability → social support → affiliate stigma → burden: β=0.029 (95% CI 0.016–0.046), p<0.01; PM≈5.40%. Non-significant path: Disability → PHB → social support → burden (β=0.003; p=0.06). PHB–social support association was not significant. Overall total effect: β=0.537 (95% CI 0.469–0.610), p<0.01.

Findings confirm that greater patient disability elevates caregiver burden directly and indirectly through increased patient potentially harmful behaviors, higher caregiver affiliate stigma, and reduced social support. The results align with caregiver stress frameworks wherein patient behavioral problems and stigma act as secondary stressors exacerbating burden. The study highlights that family caregivers are frequent targets of patient-perpetrated violence, which contributes substantially to burden yet is often overlooked in outpatient contexts. Social support emerged as a protective factor, mediating the disability–burden relationship, and also linked to reductions in affiliate stigma, suggesting interconnected mechanisms. These insights underscore the need for comprehensive, family-centered interventions in China that target patient behavior management, anti-stigma efforts, and enhancement of caregiver social support to break cycles of isolation and high burden, thereby potentially improving both caregiver well-being and patient outcomes.

Patient disability is linked to higher caregiver burden among Chinese family caregivers of individuals with schizophrenia, with mediation through patient potentially harmful behavior, caregiver affiliate stigma, and social support. Interventions should target these mediators: reduce and manage patient potentially harmful behaviors; implement anti-stigma strategies tailored to families; and strengthen caregiver social support. There is also a need to improve occupational skills and daily functioning for both patients and caregivers and support caregivers in organizing caregiving tasks to promote patient engagement. Future research should employ longitudinal designs to establish causality and examine additional mechanisms such as caregiver coping styles, while health systems should expand community-based mental health services to support families.

The cross-sectional design precludes causal inference despite mediation analyses. Participation bias may be present due to refusals among eligible dyads, potentially affecting generalizability. Although efforts were made to include diverse participants (e.g., transport reimbursement, outreach), some patients and caregivers declined participation, and hospitalized individuals were only partially included. Self-report measures may introduce reporting bias.