Reply to J Hathcock and I Munro

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

Dear Sir:

I agree with the clarification that the tolerable upper intake level (UL) for vitamin D is 17% lower than the no observed adverse effect level (NOAEL). The distinction was omitted from our article (1) for the sake of brevity, but it should have been explained as Hathcock and Munro did in their letters. My comment is that, despite the theoretical principles for deriving the UL for each nutrient, the Food and Nutrition Board (FNB) did not give a reason why the uncertainty factor for vitamin D was chosen other than to state that the value was "conservative" (2). Therefore, it appeared to me that the 17% adjustment was simply adopted to produce a value equal to the vitamin D safety limit of 50 µg/d (2000 IU/d) that was referred to in earlier FNB reports since at least the 1968 edition of Recommended Dietary Allowances (3).

Unlike the ULs for most other nutrients, the UL for vitamin D is not internally consistent across age groups. According to the model for deriving ULs, adjustments rely on body weight ratios (4). On the basis of what is probably a more rigorously established UL for vitamin D for infants and assuming a body weight ratio of 10, the infant data imply that the adult UL should be 250 µg/d; this value is within the adult physiologic production rate for vitamin D (5).

Munro emphasizes the need for published evidence, but we must recognize the reality of a publication bias that ignores or downplays safety and highlights evidence of harm. To illustrate this, evidence for the safety of doses of vitamin D higher than the current lowest observed adverse effect level (LOAEL) was presented in papers published before 1995, when this issue was last reviewed by the FNB. I know of 2 articles that, although they did not focus on the issue, clearly showed that high doses of vitamin D do not cause hypercalcemia in healthy subjects (6, 7). Neither article was mentioned in the FNB review that set upper limits (2). I reviewed at least 3 other studies of healthy subjects in which doses of vitamin D exceeded the NOAEL (5). Although those studies do not mention hypercalcemia, I suggest that it did not occur in those studies either. Any other interpretation implies that the authors failed to consider the effects of vitamin D on calcium or neglected to mention evidence of toxicity; it is hard to imagine either scenario. Instead, authors are inclined to take for granted aspects of nutritional studies showing that no harm was done, and authors do not highlight noneffects (safety) in their publications. Furthermore, because evidence for safety is difficult to support by statistical analysis and because statements about safety are easy prey for critics, such statements are usually eliminated from publications. Thus, the issue of safety may not necessarily require more research designed to provide data for the process of establishing DRIs. Instead, researchers and those involved in the review and publication process should be aware of the need to place more emphasis on the implications of study results for both safety and toxicity.

A report by Barger-Lux et al (8) is particularly relevant because they found no hypercalcemia in 14 men taking 1250 µg vitamin D₃/d for 8 wk; this dosage is 10 times the highest cumulative dosage purportedly given by Narang et al (9). [One of the authors of that study (9) was a member of the appropriate UL subcommittee of the FNB and would have known of this work, which was unpublished at the time the UL was set.] When the data of Barger-Lux et al (8) are taken in the context of Haber's law (toxic efficacy reflects dose times its duration), the evidence against the current UL or NOAEL becomes overwhelming. Haber's law is applied by the Food and Drug Administration to facilitate comparisons among studies that used different dosing protocols, to help establish reference doses (10).

Hathcock states that the change in serum calcium with 95 mg vitamin D/d (the LOAEL) was described by the FNB as "modest," ie, small and statistically difficult to detect. However, on , the final FNB report uses the word modest in the context of the calcium change that Narang et al (9) evoked with 30 mg vitamin D/d (2). Because the article by Narang et al (9) is not readily available but is the only article used by the FNB to define the current UL, NOAEL, and LOAEL, some of its data are reproduced here for comparison (Table 1). Our study had the power to detect an increase in serum calcium as small as 0.06 mmol/L, well within the capability of detecting the 0.19-mmol/L increase (2.62 - 2.43 mmol/L; P < 0.01) that Narang et al (9) reported with 60 mg vitamin D/d. The LOAEL was based on mean serum calcium in the hypercalcemic range, 2.83 mmol/L (11.3 mg/dL), not on a modest increase.

View this table:
TABLE 1 . Doses of vitamin D pertinent to the tolerable upper intake level (UL) and the lowest observed adverse effect level (LOAEL) and their effects on serum calcium¹  
I was pleased that Munro mentioned our comments about the research limitations created by the current UL for vitamin D. Despite the clarification about the UL in the preceding letters, health professionals tend to regard the UL as the toxic dose. They think of vitamin D as a drug, for which the UL is the numerator used to calculate the therapeutic index (ratio of the toxic dose to the effective dose). For example, at the adequate intake (AI) of 15 µg/d for older adults, the perceived therapeutic index for vitamin D is only 3.3. Of the 71 hospital workers who served as subjects in our study, only 1 was a physician. Saying that they did not consider it prudent to take >50 mg vitamin D/d, other physicians declined invitations to take part in the study. The FNB has not made it clear to health professionals that the LOAEL (not the UL) should be used as the numerator in calculations of the therapeutic index for nutrients. As a result, pharmacists almost always warn patients against taking the highest dose of vitamin D available over the counter (25 µg/pill, 1000 IU). They are warning patients against taking the very dose that adults need to ensure that 25-hydroxyvitamin D concentrations exceed the decision point for vitamin D insufficiency (11). To see first-hand the real-life effect of the current UL for the public, I suggest that readers pose a naive question to their local pharmacist about the risk of taking the 1000-IU vitamin D pills.

REFERENCES

1. Vieth R, Chan P-CR, 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. Food and Nutrition Board, Institute of Medicine. Dietary reference intakes for calcium, phosphorus, magnesium, vitamin D, and fluoride. Washington, DC: National Academy Press, 1997.
3. National Research Council. Recommended dietary allowances. 7th rev. ed. Washington, DC: National Academy Press, 1968.
4. Food and Nutrition Board, Institute of Medicine. Dietary reference intakes. A risk assessment model for establishing upper intake levels for nutrients. Washington, DC: National Academy Press, 1998.
5. Vieth R. Vitamin D supplementation, 25-hydroxyvitamin D concentrations, and safety. Am J Clin Nutr 1999;69:842–56.
6. Hartwell D, Hassager C, Christiansen C. Effect of vitamin D₂ and vitamin D₃ on the serum concentrations of 1,25(OH)₂D₂ and 1,25(OH)₂D₃ in normal subjects. Acta Endocrinol (Copenh) 1987; 115:378–84.
7. Stern PH, Taylor AB, Bell NH, Epstein S. Demonstration that circulating 1 alpha, 25-dihydroxyvitamin D is loosely regulated in normal children. J Clin Invest 1981;68:1374–7.
8. Barger-Lux MJ, Heaney RP, Dowell S, Chen TC, Holick MF. Vitamin D and its major metabolites: serum levels after graded oral dosing in healthy men. Osteoporos Int 1998;8:222–30.
9. Narang NK, Gupta RC, Jain MK, Aaronson K. Role of vitamin D in pulmonary tuberculosis. J Assoc Physicians India 1984;32:185–6.
10. Gaylor DW. The use of Haber's law in standard setting and risk assessment. Toxicology 2000;149:17–9.
11. Vieth R. Would prehistoric human 25-hydroxyvitamin D concentrations be beneficial, and how much vitamin D do we need to ensure desirable nutritional targets? In: Burckhardt P, Heaney R, Dawson-Hughes B, eds. Nutritional aspects of osteoporosis. San Diego: Academic Press,2001:173–95.