Dr. Lesser hypothesizes that calcium or vitamin D deficiency might have biased our main study finding that alendronate prevents bone loss over 4 years (1). We elected to assess dietary calcium intake and offer nutritional advice rather than provide calcium and vitamin D supplementation. We believed that this naturalistic approach was more appropriate because the prevalence of calcium and vitamin D deficiency is likely to be low in healthy postmenopausal women. We also considered the issue of whether calcium supplementation can prevent bone loss in younger postmenopausal women to be controversial (2). The controversy did, however, prompt us to analyze the effects of calcium intake on bone loss in the placebo group (3).

The average daily calcium intake in the placebo group was 900 mg at baseline, and the range was wide enough to allow meaningful analyses of bone loss in women with high and those with low calcium intake. No association was found between change in BMD and calcium intake over a range of daily calcium intake (from <100 mg to 2700 mg) (3). In addition, the participants' baseline characteristics, including calcium intake, were similar across study groups (4). Finally, bone loss in the placebo group was similar across study centers; this finding excluded a major influence of potential vitamin D deficiency resulting from scarce sunlight exposure (4).

The effect of antiresorptive therapy on BMD is closely associated with protection against fracture (5). Earlier studies have shown that alendronate reduces the incidence of fractures by about 50% in postmenopausal women with osteoporosis. Because the expected prevalence of osteoporotic fractures in early postmenopausal women with normal bone mass is low, fractures were not an end point in our study (1, 4). Fractures were, however, captured as adverse events, and we recorded 153 nonvertebral and 23 vertebral fractures over 4 years (1). As expected, most fractures were traumatic. The incidence of fractures was similar across study groups, with one exception: The incidence tended to be lower in the hormone replacement group (1). The small number of nontraumatic fractures did not allow any definitive conclusion on effect on fracture incidence (1, 4). Still, the effect of alendronate on BMD implies that over the longer term, alendronate will prevent not only bone loss but also fractures (5). Additional studies should be done to investigate this further.