Table 2.

Odds Ratios of Being in the Bottom Sex-specific 20% of the Distribution of ALMI, ALM Residuals, Grip Strength, and Muscle Quality By Conditional BMI Change (in the intervals 15–36 and 36 to 60–64 years; adjusted for sex; N = 1,081)

Interval of BMI change	OR (95% CI) of being in the bottom sex-specific 20% of the distribution of the specified measure of muscle per 1 SD change in BMI in each interval
	ALMI	ALM residuals	Grip strength	Muscle quality
15 to 36 y	0.33 (0.27, 0.42)	0.65 (0.55, 0.78)	1.00 (0.85, 1.18)	1.46 (1.23, 1.72)
36 to 60–64 y	0.35 (0.28, 0.43)	0.94 (0.80, 1.10)	1.03 (0.88, 1.21)	1.47 (1.25, 1.74)
p-value*	.81	.003	.76	.93

Notes: ALMI = appendicular lean mass index; ALM residuals = appendicular lean mass residuals; BMI = body mass index; OR = odds ratio.

There was little evidence of sex interaction in any of these models: p = .06 for ALMI, .12 for ALM residuals, .23 for grip strength, and .76 for muscle quality; sex-stratified OR of low ALMI associated with change in BMI between 15–36 and 36 to 60–64 y, respectively, were in men: 0.27 (0.19, 0.38), 0.39 (0.29, 0.53); in women: 0.41 (0.30, 0.56), 0.31 (0.23, 0.41).

ORs represent the odds ratio of being in the bottom 20% of the distribution of the specified measure of muscle (compared with the odds for being in the top 80%) per 1 SD change in BMI in the specified interval conditional on earlier BMI (ie, 1 SD change in the residuals from sex-specific models in which each BMI measure is regressed on the earlier measure(s) (ie, 36 on 15 and 60–64 on 36 and 15).

OR < 1: greater gain in BMI in the specified age interval associated with reduced odds of low levels of the specified measure of muscle.

OR > 1: greater gain in BMI in the specified age interval associated with increased odds of low levels of the specified measure of muscle.

*From Wald test of the difference between the two coefficients.