Distress disclosure and health behaviours in patients with stroke: a theory-informed serial mediation model of social support, self-efficacy, and self-management

Abstract

Aim To examine the association between distress disclosure and health behaviours among patients with stroke and to test whether social support, self-efficacy and self-management statistically mediate this association independently and in sequence. Methods A convenience sample of 473 patients with stroke was recruited from the Department of Neurology, Baotou Central Hospital, Inner Mongolia, China, between August 2025 and March 2026. Participants completed a sociodemographic and clinical characteristics form, the Distress Disclosure Index, Social Support Rating Scale, Stroke Self-Efficacy Questionnaire, Stroke Self-Management Scale and Health Behaviour Scale for Stroke Patients. Descriptive statistics, Pearson correlations, hierarchical multiple linear regression and serial mediation analyses were conducted using IBM SPSS Statistics 29.0 and PROCESS macro version 5.0, Model 6. Age, stroke type, modified National Institutes of Health Stroke Scale (modified NIHSS) score and Barthel Index (BI) score were included as covariates. Results The mean health-behaviour score was 60.65 (SD 12.32), indicating a moderate level. Distress disclosure was positively correlated with health behaviours (= 0.828,< 0.001) and remained a significant correlate after adjustment for covariates (B = 1.333, beta = 0.819,< 0.001). The total effect of distress disclosure on health behaviours was 1.3327 (95% CI: 1.2494 to 1.4160), the direct effect was 0.2032 (95% CI: 0.0635 to 0.3429), and the total indirect effect was 1.1295 (95% CI: 0.9875 to 1.2884). Social support, self-efficacy and self-management each showed significant indirect statistical effects. The largest specific indirect effect was through social support alone (effect = 0.4749; 35.63% of the total effect). The full serial indirect association through social support, self-efficacy and self-management was also significant (effect = 0.0276; 95% CI: 0.0121 to 0.0479). Conclusion Distress disclosure was strongly associated with health behaviours among patients with stroke, and this association was accompanied by significant indirect statistical associations via social support, self-efficacy and self-management. The findings support an expression-support-efficacy-management framework that may inform the design and prospective evaluation of communication-centred and self-management-oriented nursing interventions.