Table 3.

Mediating effect of muscle exercise on the association between (A) low household income, (B) low education state, and poor HGS.

(A) Steps (Pathway)	B	SE	OR	95% CI		p	Sobel Test
				Lower	Upper		z (p)
Step 1 (a). I (1Q) → M	0.532	0.151	1.702	1.264	2.292	0.001	2.373 (0.017)
(b) M → G	−0.417	0.169	0.664	0.475	0.928	0.017
Step 2 (c). I (1Q) → G	0.572	0.174	1.772	1.257	2.498	0.001
Step 3 (c’). I (1Q) → G	0.552	0.172	1.736	1.235	2.440	0.002
(B) Steps (Pathway)	B	SE	OR	95% CI		p	Sobel Test
				Lower	Upper		z (p)
Step 1 (a). E (1E) → M	0.494	0.165	1.639	1.184	2.271	0.003	2.489 (0.012)
(b) M → G	−0.407	0.172	0.665	0.474	0.935	0.019
Step 2 (c). E (E) → G	0.862	0.187	2.368	1.635	3.429	<0.001
Step 3 (c’). E (E) → G	0.837	0.187	2.309	1.596	3.342	<0.001

B, estimate; SE, standard error; OR, odds ratio; CI, confidence interval; I (1Q), household income (1Q, reference: 4Q); E (E), education (elementary school, reference: college/university); M, muscle exercise; G, grip strength; HGS, handgrip strength; SES, socioeconomic status. Pathway a, the relationship between the independent and mediating variables; pathway b, the relevance between the mediating and dependent variables; pathway c, the relevance between the independent and dependent variables; pathway c’, the relevance between the independent and dependent variables. Pathways were calculated using multivariable logistic regression analysis.