. 2020 Mar 26;2020:3848256. doi: 10.1155/2020/3848256

Table 2.

Differences in mean HOMA values by level of weekly relative physical activity in US men and women, after adjusting for covariates.

Covariate	Weekly relative physical activity level				F	P
	Sedentary-R	Low-R	Moderate-R	High-R
	Mean ± SE	Mean ± SE	Mean ± SE	Mean ± SE
Demographics	2.6^a ± 0.08	2.6^a ± 0.09	2.2^b ± 0.11	2.0^c ± 0.08	11.5	<0.0001
Demographics and lifestyle	2.2^a ± 0.07	2.2^a ± 0.08	1.9^b ± 0.09	1.9^b ± 0.07	6.0	0.0012
Demographics, lifestyle, and waist circ.	2.7 ± 0.09	2.6 ± 0.08	2.5 ± 0.10	2.5 ± 0.08	1.6	0.1937

^a,b,cMeans on the same row with the same superscript letter were not statistically different (P > 0.05). Because of nesting, there were only 59 degrees of freedom in the denominator of each model. The physical activity categories were based on relative MET-minute levels. Participants reporting no regular physical activity were classified as Sedentary, and the remaining adults, each reporting some physical activity in the past 30 days, were divided into sex-specific tertiles. Across the four categories of relative physical activity, weighted percentages were as follows: 34% (n = 2640) reported no regular physical activity (Sedentary), 22.5% (n = 1354) reported Low levels, 22.1% (n = 1309) reported Moderate levels, and 21.4% (n = 1197) reported High levels of physical activity (MET-minutes). Because sample weights were applied to each participant, differences in the size of each category should be interpreted relative to percentages, not N. Means on the same row were adjusted for the covariates in the left column. Moderate and High mean differences in the demographics model were statistically significant at the P=0.0658 level. The demographic covariates were age, sex, race, and year of assessment. The lifestyle covariates were body mass index and cigarette smoking. Waist circ. = waist circumference measured in centimeter.