Curhan and colleagues [1] report that the risk for kidney stone is inversely related to calcium intake in women, a finding consistent with their earlier report in men [2]. They note, however, that supplemental calcium seems to increase the risk for stones; the authors of the accompanying editorial [3] express some puzzlement at that finding.

I suggest that Curhan and colleagues overinterpreted their data and that Coe and colleagues were insufficiently critical in analyzing the data. One test of inferences from such observational data is the finding of at least an ordinal relation between dose and effect. This relation is clearly seen for food calcium in Curhan and colleagues' Table 3, with a highly significant downward trend in risk for stone development as calcium intake increases. However, such a relation is not found with supplemental calcium (the authors' Table 4. Only the group with the lowest supplement intake had a significant increase in risk for stones. The groups with higher intakes had no increase at all. Persons showing an increase consumed less than 500 mg of supplemental calcium per day [that is, the type of intake likely to be associated with multivitamin supplements that contain minerals]. In contrast, supplement intakes greater than 500 mg/d generally reflect ingestion of true calcium supplements. Thus, if the effect described by Curhan and colleagues is anything other than a chance occurrence, it may reflect some other component of the types of supplements likely to be represented in the low-intake group rather than calcium.

Finally, it may be worth recalling that the standard therapy for the renal stone disease of intestinal hyperoxalosis is precisely a high calcium intake (in the form of calcium supplements). The reason, consistent with the explanation given for the efficacy of dietary calcium, is the complexation of oxalate in the intestine by unabsorbed dietary calcium and prevention of its absorption; this process in turn reduces the renal oxalate burden.