Table 5.

Multivariate logistic regression analysis—factors significantly associated with adherence to physical activity recommendations.

Variables	OR	p-Value	95% CI
Model A (all patients included)
${\chi }^2$ 46.9; Nagelkerke $R^2$ 0.64; $p<0.001$
CSHA-CFS (points)	0.38	0.001	0.206–0.688
Motivation (points)	1.13	0.009	1.032–1.245
Duration of hemodialysis (months)	0.99	0.047	0.972–0.999
Awareness of recommendations (1—aware; 0—unaware)	0.08	0.005	0.013–0.464
Vascular access (1—arteriovenous fistulas; 0—other)	0.09	0.007	0.017–0.530
Model B (patients ≥60 yrs included)
${\chi }^2$ 46.9; Nagelkerke $R^2$ 0.67; $p<0.001$
CSHA-CFS (points)	0.25	0.028	0.071–0.861
Motivation (points)	1.19	0.051	1.032–1.245
AIS (points)	0.79	0.036	0.640–0.990
Duration of hemodialysis (months)	0.95	0.025	0.918–0.999
Vascular access (1—arteriovenous fistulas; 0—other)	0.09	0.045	0.008–0.950
Model C (patients <60 yrs included)
${\chi }^2$ 22.7; Nagelkerke $R^2$ 0.69; $p<0.001$
CSHA-CFS (points)	0.13	0.012	0.026–0.633
AIS (points)	1.26	0.058	0.991–1.603
Aware of recommendations (1—aware; 0—unaware)	0.02	0.012	0.001–0.427

Notes: Odds ratio is shown with 95% CI for significant covariates. Variables that yielded p-values of 0.1 or lower in the initial univariate logistic regression analysis of factors were predictive of falls. Logit modeled probability that adherence = yes. Abbreviations: OR, odds ratio; CI, confidence interval.