Reply to FAJ Muskiet et al

Department of Laboratory Medicine and Pathobiology University of Toronto and Pathology and Laboratory Medicine Mount Sinai Hospital 600 University Avenue Toronto, Ontario M5G 1X5 Canada E-mail: rvieth{at}mtsinai.on.ca

Dear Sir:

I welcome the commentary of Muskiet et al, who present what I see as 4 reasonable questions that probably reflect the thoughts of many readers of our recently published study (1).

1) Is there any evidence that a vitamin D intake of 100 µg/d is more beneficial than one of 20 µg/d? The answer is that no study has ever been attempted to show benefits of vitamin D beyond 20 µg/d. Concerns about vitamin D such as those exemplified by the preceding letter have made it difficult for anyone to be rigorous about the study of vitamin D nutrition in humans.

We cited 6 studies in our article that concluded that if the aim is to keep parathyroid hormone concentrations low, 25-hydroxyvitamin D [25(OH)D] concentrations should exceed 70 nmol/L (1). It remains hypothetical whether that 25(OH)D target level delivers tangible health benefits. Because a mean concentration at a target level implies that one-half of the study group has values below the target, it is important for researchers to know how to ensure that the target is met. Our article gives a realistic sense of how much vitamin D must be consumed to ensure what others consider desirable. In drug development, a phase 1 study like ours would be only the first clinical step. However, Muskeit et al question whether more developmental work is needed for vitamin D because some health benefits were detectable at lower doses.

A wealth of scientific literature, from fields ranging from epidemiology to molecular biology, provides compelling circumstantial evidence that the health benefits of vitamin D extend far beyond effects on just bone. If this kind of preclinical evidence were to exist for any patented molecule, clinical development would be much faster than it is with vitamin D, and it would not suffer from the arbitrary dose restrictions that have constrained nutritional research.

2) Similar 25(OH)D concentrations have been reported by others using 15–20 µg/d, so why use so much more? It is instructive to note that the 2 exceptions seem to be getting all the attention, whereas the majority of publications that present lower 25(OH)D data for the same dose are ignored. At least 25 other studies in which 20 µg/d was used reported average 25(OH)D concentrations <80 nmol/L, as cited previously (2). More recent studies also report this finding. The 2 papers cited by Muskiet et al were the exceptions in terms of the 25(OH)D concentrations attained (3, 4) because they used the "direct" method of measuring 25(OH)D (5). Both laboratories have since stopped using the method, and Meunier (6) now reports lower 25(OH)D concentrations, in line with most of the literature.

3) What would happen if persons in the tropics with abundant sun exposure acquired 100 µg additional vitamin D/d? If serum 25(OH)D concentrations are already >150 nmol/L, then the effective physiologic supply of vitamin D is equivalent to 250 µg/d (2). The 25(OH)D response to a vitamin D dose behaves in a log-dose manner as shown in Figure 2 of my review (2). As a further example, we reported that 25 µg vitamin D/d resulted in average 25(OH)D concentrations of 69 nmol/L, whereas 4 times that amount increased 25(OH)D concentrations by only another 27 nmol/L (1). That increment becomes even smaller as the predose 25(OH)D concentration increases. Thus, an additional 100 µg/d would add marginally to what I regard as the inconsequential risk due to the 250-µ/d vitamin D supply that is physiologic because it is obtainable through sun exposure. Because all available evidence indicates that a long-term vitamin D consumption of 1000 µg/d is needed to cause hypercalcemia, there is a large margin of safety with 100 µg/d. [I welcome any discussion of evidence implicating harm with vitamin D₃ (not D₂) in adults at doses <1000 µg/d. There is simply nothing published about this, except in infants.]

4) Would vitamin D not accumulate in adipose tissue and cause toxicity if adipose tissue were to break down? In our study, we did consider the effect of body mass, but could not detect a correlation between weight and the effect of a vitamin D dose on serum 25(OH)D (1). Muskiet et al describe a study in which the investigators administered enough vitamin D to rats to raise circulating 25(OH)D and vitamin D to 1800 nmol/L, which is in the toxic range, and then measured vitamin D in fat tissue (7). There is a reason so much vitamin D ended up in the rats' adipose tissue. Pharmacologic amounts of vitamin D that are toxic preoccupy circulating vitamin D binding protein; thus, the percentage of vitamin D that is free and unbound increases (2, 8). At toxic doses, the freely circulating vitamin D and its metabolites accumulate in both adipose (7) and muscle (9). The dosage of 100 µg vitamin D/d we used was physiologic and far below the amount that could change the free fraction of circulating metabolites as a result of saturation of vitamin D binding protein. Thus, the deposition of vitamin D in adipose tissue would be no more than what will occur for persons getting a lot of sun exposure. Before doing the human study, my laboratory looked at modest vitamin D supplementation in rats, in which we kept 25(OH)D concentrations well within the normal, human range. There were profound noncalcemic changes in the calcium homeostatic system, including higher tissue vitamin D receptor expression and PTH suppression (10). We used several strategies in the search for vitamin D in the adipose tissue of those rats, but at the doses we used, detected no vitamin D there. Because it was negative, that observation was not published.

I point out that pork and beef products are not meaningful sources of vitamin D nutrition unless the animals have been dosed with enough vitamin D to cause hypercalcemia (7, 9). There is simply no reason to think that the amount of vitamin D in the adipose tissue of animals or humans without vitamin D–induced hypercalcemia should be a concern. In any discussion of vitamin D, we must maintain a context, and note what is physiologic and what reflects true excesses.

REFERENCES

1. Vieth R, Chan PC, MacFarlane GD. Efficacy and safety of vitamin D₃ intake exceeding the lowest observed adverse effect level. Am J Clin Nutr 2001;73:288–94.
2. Vieth R. Vitamin D supplementation, 25-hydroxyvitamin D concentrations, and safety. Am J Clin Nutr 1999;69:842–56.
3. Dawson-Hughes B, Harris SS, Krall EA, Dallal GE. Effect of calcium and vitamin D supplementation on bone density in men and women 65 years of age or older. N Engl J Med 1997;337:670–6.
4. Chapuy MC, Arlot ME, Duboeuf F, et al. Vitamin D₃ and calcium to prevent hip fractures in the elderly women. N Engl J Med 1992; 327:1637–42.
5. Lips P, Dawson-Hughes B, Pols HA, Holick M. An international comparison of serum 25-hydroxyvitamin D measurements. Osteoporos Int 1999;9:394–7.
6. Meunier P. Vitamin D insufficiency: reappraisal of its definition threshold and bone consequences. In: Burckhardt P, Dawson-Hughes B, Heaney R, eds. Nutritional aspects of osteoporosis. New York: Academic Press,2001:167–72.
7. Brouwer DA, van Beek J, Ferwerda H, et al. Rat adipose tissue rapidly accumulates and slowly releases an orally-administered high vitamin D dose. Br J Nutr 1998;79:527–32.
8. Pettifor JM, Bikle DD, Cavaleros M, Zachen D, Kamdar MC, Ross FP. Serum levels of free 1,25-dihydroxyvitamin D in vitamin D toxicity. Ann Intern Med 1995;122:511–3.
9. Montgomery JL, Parrish FC Jr, Beitz DC, Horst RL, Huff-Lonergan EJ, Trenkle AH. The use of vitamin D₃ to improve beef tenderness. J Anim Sci 2000;78:2615–21.
10. Vieth R, Milojevic S, Peltekova V. Improved cholecalciferol nutrition in rats is noncalcemic, suppresses parathyroid hormone and increases responsiveness to 1,25-dihydroxycholecalciferol. J Nutr 2000;130:578–84.