Table 4.

Multivariate Analysis of QMT Change during the 2-week Follow-up and Nutritional Risk at Admission Using Three Nutritional Indicators.

Limb	Nutritional risk	B	P value	95% CI		β
				Lower	Upper
Paralytic	GNRI
	Low risk	Reference
	High risk	−0.106	0.070	−0.221	0.009	−0.187
	CONUT
	Low risk	Reference
	High risk	−0.049	0.464	−0.180	0.082	−0.068
	MNA-SF
	Normal nutritional status	Reference
	At risk of malnutrition	0.047	0.386	−0.060	0.153	0.072
	Malnourished	−0.108	0.109	−0.241	0.024	−0.162
Non-paralytic	GNRI
	Low risk	Reference
	High risk	−0.013	0.803	−0.113	0.088	−0.027
	CONUT
	Low risk	Reference
	High risk	−0.022	0.697	−0.137	0.092	−0.038
	MNA-SF
	Normal nutritional status	Reference
	At risk of malnutrition	0.037	0.416	−0.053	0.128	0.070
	Malnourished	−0.143	0.013*	−0.254	−0.031	−0.258

B, partial regression coefficient; β, standardized partial regression coefficient; CI, confidence interval; CONUT, Controlling Nutritional Status; QMT, quadriceps muscle thickness; GNRI, Geriatric Nutritional Risk Index; MNA-SF, Mini Nutritional Assessment - Short Form. *P < 0.01

Multiple linear regression analyses were performed with the QMT change during 2 weeks as the dependent value and the nutritional indicator as the independent value.

All models were adjusted for age, sex, premorbid modified Rankin scale, Charlson Comorbidity Index, National Institutes of Health Stroke Scale, energy intake during two weeks, estimated time of rehabilitation dose, and QMT at admission.

Adjusted R² for each model: 0.305 in the GNRI, 0.292 in the CONUT, and 0.297 in the MNA-SF of the paralytic limb; 0.202 in the GNRI, 0.204 in the CONUT, and 0.260 in the MNA-SF of the non-paralytic limb (all P < 0.01)