Studies of any design were eligible for inclusion, provided they compared vitamin D versus placebo, or vitamin D plus calcium versus calcium alone (with or without concomitant non-osteoporosis medications). Studies of postmenopausal women were eligible; those that also included men aged over 65 years were eligible too, but the data for men were only included in the analysis if the results for women could not be extracted separately. The primary outcomes of the review were spontaneous fall and vertebral or nonvertebral fractures. The secondary outcomes were serum level of the following biochemical markers: serum 25-hydroxyvitamin D, parathyroid hormone, osteocalcin, bone-specific alkaline phosphatase and urinary cross-linked N-telopeptides of type I collagen.
The intervention in the included studies was oral cholecalciferol (vitamin D3) at doses ranging (in most cases) from 300 to 800 international units daily, administered with or without calcium. Some participants were also taking other medications (e.g. corticosteroids or thyroid medications). The participants were aged from 45 to over 80 years and lived in a variety of settings, such as the community, nursing homes and apartments for the elderly (where stated). The inclusion criteria of the primary studies varied with respect to the fracture or osteoporosis history of the participants. In all studies the mean baseline serum 25-hydroxyvitamin D indicated that participants were vitamin D-inadequate. Nearly half of the studies in the review included both men and women, without separate analysis. The duration of follow-up varied from 18 weeks to over 5 years.
The authors did not state how many reviewers selected articles for the review, or how any disagreements were resolved.