Abstract
OBJECTIVES: To identify the threshold effect of frailty on rehabilitation exercise adherence in patients with ischemic stroke. METHODS: The study included 307 patients diagnosed with ischemic stroke who were given a questionnaire comprising a general information form, the Rehabilitation Adherence Assessment Scale, and the Frailty Assessment Scale. Univariate and multivariable linear regression were employed to determine factors influencing rehabilitation exercise adherence. Subsequently, restricted cubic splines were utilized to fit a smooth curve and detect potential threshold effects. RESULTS: The average score for the rehabilitation exercise adherence index was (54.83 ± 9.32), while the average frailty score was (4.59 ± 2.14). Through univariate analysis, it was found that gender, marital status, living arrangement, household registration type, number of medications, and level of independence are factors influencing rehabilitation exercise adherence. Restricted cubic spline analysis revealed a non-linear relationship between frailty and rehabilitation exercise adherence. The association between frailty score and rehabilitation exercise adherence was found to be insignificant when the score was ≤3.98. Conversely, exceeding this threshold revealed a significant decline in the adherence index, with each additional frailty point correlating with a 2.56-point decrease (p < 0.001). CONCLUSION: Rehabilitation exercise adherence is moderate among patients with ischemic stroke, while the prevalence of frailty is notably high. A frailty score above 3.98 might serve as an early indicator of its impact on adherence. Accordingly, rehabilitation programs need to be adjusted to accommodate disease characteristics and sociodemographic factors.