Table 3.

Multivariable Linear Regression Analyses Showing the Relationships Between Measures of Muscle Mass and Function and Low, Medium, and High Impact Accelerometry Counts

	Low		Medium		High
	β (95% CI)	p	β (95% CI)	p	β (95% CI)	p
Gait speed	1.26 (1.08, 1.48)	<.01	1.68 (1.33, 2.12)	<.01	1.47 (1.14, 1.89)	<.01
Chair stand time	1.00 (0.87, 1.16)	.99	0.95 (0.77, 1.18)	.64	1.03 (0.82, 1.30)	.78
Tandem balance <10 sa	1.05 (0.75, 1.47)	.79	1.01 (0.62, 1.66)	.95	1.05 (0.61, 1.79)	.86
Grip strength	1.10 (0.96, 1.25)	.17	1.08 (0.89, 1.31)	.42	1.14 (0.92, 1.40)	.22
Peak force	1.17 (0.99, 1.39)	.07	1.23 (0.95, 1.57)	.11	1.40 (1.07, 1.84)	.02
Peak power	1.01 (0.84, 1.22)	.90	1.10 (0.84, 1.44)	.50	0.93 (0.69, 1.26)	.64
mCSA	0.95 (0.84, 1.08)	.42	0.96 (0.80, 1.15)	.64	1.01 (0.82, 1.23)	.94

Note: Exponentiated coefficients are presented representing the change in low, medium or high impact counts for an SD increase in muscle function or size exposure (eg, a coefficient of 1.24 represents a 24% increase in low impact counts for a 1 SD increase in gait speed). Adjusted for age, height, weight, smoking, alcohol, IMD, and comorbidities. N = 242.

CI = Confidence interval; mCSA = Muscle cross-sectional area.

^aCoefficient represents the ratio of geometric means with tandem balance 10 s as the reference category.

Low impacts 0.5≤0.5 g<1.0 g, medium impacts 1.0≤g<1.5, high impacts ≥1.5 g