点击显示 收起
Osteoporosis Research Center
Creighton University
601 North 30th Street, Suite 4841
Omaha, NE 68131
E-mail: rheaney{at}creighton.edu
Osteoporosis Research Center
Creighton University
601 North 30th Street, Suite 4820
Omaha, NE 68131
Dear Sir:
We are grateful to Sood and Sood for giving us the opportunity both to set the record straight and to clarify a common misconception. They state that our study (1) was "neither designed nor powered" to assess the effect of vitamin D on cancer risk.
First, although cancer was a secondary outcome, it was always a part of the design, and it was the first outcome measure we evaluated after looking at the skeletal endpoint that had been the primary outcome measure. Second, power is an expression of the likelihood of finding a difference of a given size in a population with a given variance. It is an important consideration in planning an investigation and in evaluating a negative outcome. But it is irrelevant with a positive outcome, because the de facto power of such a trial is 1.0. Power relates to a type II error—ie, missing a real effect. When the sought-for difference is found, the only relevant issue is the type I error, which is handled by the observed probability estimate. In the case of our trial, the difference was highly significant by a variety of statistical approaches.
It is important to recognize that the reduction in cancer incidence that we reported was precisely of the same magnitude as was found in a large number of observational studies and meta-analyses of such studies (eg, 2-5). Hence, the observed difference, rather than being surprising, was to have been expected. Moreover, our findings were internally consistent as well, because the serum 25-hydroxyvitamin D [25(OH)D] concentration at year 1 was about as good a predictor of risk as was the fact of treatment.
Sood and Sood express concern about the potential for toxicity, in relation to possible hyperphosphatemia. We call the reader's attention to 2 facts. First, although vitamin D influences phosphate absorption in animal gut sac preparations, it has little demonstrated effect on human phosphate absorption under typical conditions. Net phosphorus absorption in adult humans ranges between 60% and 80% (6), which is already high. Moreover, except in cases of very severe vitamin D deficiency, phosphorus absorption is scarcely influenced by vitamin D status. Second, whereas elevated serum phosphorus is indeed a risk factor for vascular calcification, there is no evidence that vitamin D doses <10 000 IU/d either elevate serum phosphorus or increase the risk of such calcification. The issue of vitamin D toxicity was exhaustively reviewed in this Journal just a few months ago (7), and Sood and Sood may find some reassurance in that report.
We draw attention further to the fact that we used native cholecalciferol (vitamin D3), not calcitriol [1,25(OH)2D3], and we wish to remind Sood and Sood that cholecalciferol does not elevate serum calcitriol concentrations in anyone (8). Calcitriol concentration is a regulated quantity, under the control mainly of parathyroid hormone. Furthermore, given the 3 orders of magnitude difference in biological potency between 25(OH)D and 1,25(OH)2D3, even very low concentrations of 25(OH)D are able to support generous 1,25(OH)2D3 synthesis, which is why vitamin D3 intake is not rate-limiting for 1,25(OH)2D3 concentration. Moreover, as one of us has shown elsewhere (9), calcium absorption efficiency is not related to serum 25(OH)D concentration in the range from 80 to 400–500 nmol/L, irrespective of the vitamin D dose.
Over the 4-y period of our study, 5 women developed kidney stones; 1 of these women was in the vitamin D–treated group. In addition, there were 7 subjects who had myocardial infarctions, which were evenly distributed across the 3 treatment groups. Serum phosphorus was not measured, because there was no way to justify the substantial increase in cost for measurement of a variable that would not have been predicted to change.
Finally, we reject the notion that a serum 25(OH)D concentration such as the one we achieved in our study—which was well below the value typically found in those who work outdoors (10)—would carry with it any risk of pathologic condition or malfunction. If anything, the concentration achieved in our vitamin D–treated participants was well below the best estimates of primitive—ie, physiologic—concentrations. We stress that it is the serum 25(OH)D concentration that is important, not the oral dose of vitamin D3.
ACKNOWLEDGMENTS
Neither author had a personal or financial conflict of interest.
REFERENCES