# Table 2 The results of multivariable linear and logistic regression models

One year outcome
Multivariable linear regression models Multivariable logistic regression model
TUG KOOS-ADL (range 0–100) VAS-pain (binary)
Predictor Regression coëfficient [95% CI], p-value Predictor Regression coëfficient
[95% CI], p-value
Predictor Odds ratio
[95% CI], p-value
Class: “low gain group” versus “moderate gain group” 3.31 [1.52, 5.09], p < 0.01 Class: “gain group” versus “moderate gain group” 11.97 [8.62, 15.33], p < 0.01 Class: “normal decrease of pain” versus “no/very little pain” 0.92 [0.17,4.84], p = 0.58
Class: “gain group” versus “moderate gain group” −0.56 [−1.75, 0.64], p = 0.36 Class: “sustained pain” versus “no/very little pain” 0.11 [0.03,0.42], p = 0.01
Age (year) 0.00 [−0.03, 0.03], p = 0.96 Age (year) 0.18 [−0.03, 0.39], p = 0.10 Age (year) 0.95 [0.88,1.04], p = 0.07
Gender −0.09 [− 0.58, 0.41], p = 0.73 Gender 2.40 [−1.17, 5.98], p = 0.19 Gender 0.45 [0.13,1.51], p = 0.63
BMI −0.03 [− 0.10, 0.04],p = 0.36 BMI 0.13 [− 0.38, 0.64], p = 0.62 BMI 1.08 [0.9,1.3], p = 0.66
Model R2 0.14 Model R2 0.20 Model R2 0.08
1. BMI = Body Mass Index; CI = confidence interval; R2 = explained variance; *p < 0.05 = statistical significant; Nagelkerkes R2; KOOS-ADL = Knee Osteoarthritis Outcome Score-Activities of Daily Living, TUG = Timed Up and Go; VAS = Visual Analogue Scale 