Analysis of latent classes and predictive factors of health behavior among stroke patients
Objective To explore the latent classes of health behavior and explore the predictive factors among stroke patients.Methods A total of 1 250 participants were recruited using cluster random sampling in September 2022.The general information,the modified Rankin scale(mRS),stroke prevention knowledge questionnaire(SPKQ),health behavior scale for stroke patients(HBS-SP),and short form-health belief model scale(SF-HBMS)were administered in the cross-sectional survey.Mplus 8.3 software was used to conduct a latent class analysis(LCA)on the health behavior of stroke patients,and SPSS 27.0 soft-ware was used to carry out multinomial Logistic regression to analyze the predictive factors of different latent classes of health behavior of stroke patients.Results The health behavior of stroke patients obtained three latent classes:low health behaviors-lack of health responsibility group(66.9％,n=794),moderate health behaviors-poor compliance group(11.9％,n=141),and good health behaviors-insufficient exercise group(21.2％,n=251).Compared with good health behaviors-insufficient exercise group,stroke patients with shorter duration education time(B=-0.589,OR=0.555,P=0.036),hemorrhagic stroke(B=0.082,OR=1.086,P＜0.001),fewer comorbidities(B=-0.022,OR=0.978,P=0.026),higher mRS score(B=-0.046,OR=1.047,P=0.004),lower SPKQ score(B=-0.055,OR=0.947,P=0.016),and lower SF-HBMS score(B=-0.085,OR=0.919,P＜0.001)were more likely to be included in moderate health be-haviors-poor compliance group.However,stroke patients with shorter duration education time(B=-0.026,OR=0.974,P=0.003),rural areas dwelling(B=0.800,OR=2.225,P=0.004),fewer comorbidities(B=-0.056,OR=0.945,P＜0.001),lower SPKQ score(B=-0.101,OR=0.904,P＜0.001),and lower SF-HBMS score(B=-0.071,OR=0.931,P＜0.001)were more likely to be included in low health behav-iors-lack of health responsibility group.Conclusion The health behavior of stroke patients has three latent classes.A targeted intervention should be carried out according to the characteristics of different classes to improve their health behavior levels.
StrokeHealth behaviorLatent class analysisPredictive factors
NETL
NSTL
万方数据
维普
