Table 4.

Clustered logistic regression models explaining hospitalization in the last year by socio-demographic characteristics, lifestyle and health-related factors, and FIM domains among patients with multimorbidity (n=1173)

Variablea	OR b	95% CI	P value	Nagelkerke R2c	Independent contribution d (%)
Model 1
Gender (male)	1.59	1.16–2.17	0.004
Total				0.012	18.46
Model 2
Gender (male)	1.60	1.16–2.20	0.004
Number of chronic diseases	1.52	1.30–1.78	< 0.001
Total				0.049	56.92
Model 3
Gender (male)	1.63	1.18–2.24	0.003
Number of chronic diseases	1.45	1.24–1.71	< 0.001
Walk	0.80	0.70–0.91	0.001
Total				0.065	24.62

^aOnly variables with P < 0.05 were included in the model

^bFor age, body mass index, number of chronic diseases, and functional independence domains scores, the odd ratios per SD increase are shown

^cNagelkerke R² is the variance of the dependent variable (hospitalization in the last year) explained by all independent variables included in the regression model

^dThe independent contribution of each cluster of predictors to the variation in hospitalization in the last year calculated as individual corresponding R² change/total R² change in the final model × 100%