Table 2.

Evaluation of average treatment effects on primary end points: adjusted for patient covariates

Behavior		Effect	Effect estimate	95% CI	p value
Daily calcium intake at 52 weeksa	Model 1	Daily calcium intake at baseline	0.552	(0.532, 0.573)	< 0.001
		Intervention vs. control	0.145	(− 12.508, 12.798)	0.982
		OSE-diet at baseline	16.245	(11.958, 20.531)	< 0.001
		Prior DXA vs. DXA-naïve	− 11.491	(− 29.887, 6.905)	0.214
		High fracture risk at baseline	23.353	(− 0.069, 46.774)	0.051
		Moderate fracture risk at baseline	42.801	(26.211, 59.391)	< 0.001
Proportion with vitamin D supplementation at 52 weeksb	Model 1	Vitamin D supplementation at baseline	20.516	(18.088, 23.27)	< 0.001
		Intervention vs. control	1.007	(0.882, 1.149)	0.923
		OSE-diet at baseline	1.042	(0.990, 1.095)	0.110
		Prior DXA vs. DXA-naïve	0.971	(0.833, 1.131)	0.704
		High fracture risk at baseline	1.023	(0.809, 1.294)	0.846
		Moderate fracture risk at baseline	1.028	(0.846, 1.249)	0.776
Weekly exercise sessions at 52 weeksa	Model 1	Weekly exercise sessions at baseline	0.430	(0.405, 0.455)	< 0.001
		Intervention vs. control	0.092	(− 0.03, 0.214)	0.139
		OSE-exercise at baseline	0.188	(0.153, 0.222)	< 0.001
		Prior DXA vs. DXA-naïve	0.042	(− 0.121, 0.204)	0.606
		High fracture risk at baseline	0.014	(− 0.191, 0.218)	0.894
		Moderate fracture risk at baseline	0.028	(− 0.143, 0.200)	0.741

Italics indicate statistical significance (p < 0.05). High fracture risk is > 20% 10-year fracture risk estimated from the baseline FRAX® score; moderate risk is 10–20% 10-year fracture risk

DXA dual-energy X-ray absorptiometry, OP osteoporosis

^aFor the daily calcium intake and weekly exercise session regressions, the effect estimates represent an adjusted linear effect

^bFor the vitamin D supplementation regression, the effect estimate represents an adjusted odds ratio