Multivariate Multinomial Regression Analysis for Characteristics of Disability Trajectoriesa
↵a OR=odds ratio; CI=confidence interval; SF-12=12-Item Short-Form Health Survey; 4DSQ=Four-Dimensional Symptom Questionnaire; TSK=Tampa Scale for Kinesiophobia; CSQ-catastrophizing=Coping Strategy Questionnaire catastrophizing scale; SSQ=Social Support Questionnaire. Explained variance (Nagelkerke R2=0.544). Percentage correctly predicted overall: 76.8% (91.5% within fast recovery group, 42.2% within modest recovery group, and 55.9% within continuous high disability group). See Method section of text for measures that were utilized. Model without SF-12 Mental Component Scale because of the high correlation (.67) with distress score (4DSQ). Fast recovery group is reference group for multinomial regression analysis. Cutoff points for dichotomous variables are defined by median score of the total population. The variables of paid work, no sports participation, recurrent complaints, nonmusculoskeletal comorbidity, low health locus of control, and distress (4DSQ) were removed from the model.