Table 3.

Linear mixed model-based means at baseline and follow-up, and treatment effects (95% CI).

	Intervention	Control	Effect (95% CI) a	P
Physical performance
Model 1
Baseline	6.8	7.0
2 months	7.1	7.0	0.3 (−0.2; 0.8)	0.27
4 months	7.2	7.2	0.3 (−0.1; 0.7)	0.18
Overall effect: F(2, 107)=1.03, P=0.36
Model 2
Baseline	6.7	7.2
2 months	7.0	7.2	0.3 (−0.2; 0.8)	0.24
4 months	7.2	7.4	0.3 (−0.1; 0.8)	0.13
Overall effect: F(2, 107)=1.30, P=0.28
6-MWT
Model 1
Baseline	525	517
4 months	538	522	7 (−9; 24)	0.38
Model 2
Baseline	523	529
4 months	537	535	8 (−8; 24)	0.33
LAPAQ b
Model 1
Baseline	12	13
2 months	13	13	1 (−0.1; 3)	0.10
4 months	12	13	0 (−1; 1)	0.99
Overall effect: F(2, 108)=1.71, P=0.19
Model 2
Baseline	12	13
2 months	13	12	1 (−0.2; 3)	0.10
4 months	11	12	0 (−1; 1)	0.96
Overall effect: F(2, 110)=1.83, P=0.17
MVPA
Model 1
Baseline	32.1	33.3
4 months	31.7	35.7	−3.9 (−14.2; 6.5)	0.46
Model 2
Baseline	32.9	34.3
4 months	31.6	37.0	−3.9 (−14.3; 6.5)	0.46
Mean activity
Model 1
Baseline	389	394
4 months	375	426	−44 (−123; 35)	0.27
Model 2
Baseline	387	398
4 months	373	430	−46 (−125; 33)	0.25

Model 1: univariable; Model 2: adjusted for sex, age, baseline BMI, and baseline 25(OH)D. Covariates were fixed at the mean values; 6-MWT, 6-min walk test; LAPAQ, LASA Physical Activity Questionnaire, MVPA, moderate-to-vigorous physical activity; c.p.m., counts per minute.

^aThe estimated effect is calculated as (the model-based mean after follow-up−the model-based mean at baseline for the intervention group)−(the model-based mean after follow-up−the model-based mean at baseline for the control group). For example, the estimated effect for LAPAQ, follow-up of 2 months: (13–12)−(13–13)=1.

^bSquare-root-transformed variables.