Robert P. Heaney Creighton University Omaha, NE rheaney at creighton.edu
Michael F. Holick, Boston University Medical Center, Boston, MA
Initial Date Submitted December 6, 2010; Date Revision Submitted December 21, 2010; Date Final Disposition Set December 23, 2010
Journal of Bone and Mineral Research
Â© 2011 American Society for Bone and Mineral Research, DOI 10.1002/jbmr.328
In the past two years vitamin D supplement sales to consumers have increased by more than 100% per year.1 Now, following the publication of the report 2 on Dietary Reference Intakes (DRIs) for calcium and vitamin D by the Institute of Medicine (IOM), many physicians report that they are decreasing their vitamin D recommendations to patients. This change was explicitly proposed by members of the IOM panel in their various media statements. While a small fraction of consumers may well have all the vitamin D they need, on balance we consider a general downward trend to be harmful for the health of the public.
Both of the authors of this Perspective served as members of the panel that drafted the 1997 report of the IOM on the DRIs for calcium and vitamin D. That report was the first issued by the IOM under the then new evidence-based guidelines for evaluating studies and making recommendations. We are thus familiar with the process and, most importantly, with vitamin D itself. On the basis of this experience, we respectfully dissent from many of the findings and recommendations in the current report and we set forth here a small fraction of the reasons for that dissent.
The IOM report (and its presentation to the media) stressed that its recommendations for vitamin D were based primarily on the intake [and serum 25(OH)D concentration] needed to ensure skeletal health and that, in the panel’s judgement, there was insufficient evidence to make any recommendations in respect to non-skeletal benefits, if any. Second, the report concluded that a serum level for 25(OH)D of 20 ng/mL was sufficient to ensure bone health. And third, they concluded that, since the bulk of the American public had 25(OH)D values that were above 20 ng/mL, most individuals were getting all of the vitamin D they needed and had no reason for further supplementation. These conclusions fail on three grounds: logic, science, and guidance.
Since the panel, in its judgement, concluded that it did not know whether there might be non-skeletal benefits (or at what blood level they could be ensured), then it is patently incorrect to say that they know that people are getting enough. __The most the panel could
logically have said was “Here’s what you need for bone; most people get that much; we do not know whether more would confer possible non-skeletal benefits.”__ That, at least, would have been an honest communication of the state of the issue as the panel apparently understands it. But to state publicly that the general public does not need more goes well beyond what the panel admits it knows.
The statement that skeletal health can be assured at serum 25(OH)D levels of 20 ng/mL is simply incorrect. Without going into an exhaustive recital of all of the evidence pointing to a skeletal need for higher levels, we cite here three illustrative observations which, in our collective judgement, indicate that, instead of 20, a serum level of 30 ng/mL is closer to the bottom end of the acceptable range for skeletal health. First, there is the large randomized controlled trial in the UK which raised serum 25(OH)D level from 21 ng/mL to 29 ng/mL, and produced a 33% reduction in all major osteoporotic fractures combined.3 The fact that other trials, with less good compliance, failed to reproduce that effect does not negate the evidence of a well-conducted trial. Second, there are the many meta-analyses of Bischoff-Ferrari and her colleagues 4 5 demonstrating that, taken overall, fracture reduction with vitamin D does not reproducibly occur below serum 25(OH)D levels of 30, and for some fractures even 40 ng/mL. Finally, there is the demonstration, in a large German autopsy series (strangely misinterpreted by the panel) that osteoid seam width “ the histological hallmark of vitamin D deficiency “ does not reach fully normal values until serum 25(OH)D levels are above 30 ng/mL.6 (N.B.: Of 33 cases with 25(OH)D values between 20 and 30 ng/mL, more than half (18) had elevated osteoid volume. An RDA, by definition, meets the need of 97.5% of the population.) In a closely related finding investigators from South Australia 7 showed seasonal variation in osteoid seam width and mineral appositional rate, reflecting variations in serum 25(OH)D precisely within the 20-30 ng/mL range, i.e., above the IOM panel’s “adequate” level.
Additionally, there is an apparent inconsistency between the recommended intake (600 IU/d for all individuals up through age 70) and the bottom end of the acceptable 25(OH)D serum concentration range (let alone higher values). As virtually universal experience with vitamin D supplementation demonstrates, 600 IU/d, if the body’s sole input of vitamin D, would not be enough to produce a value of even 10 ng/mL, let alone 20 or above. There is a generally recognized “rule of thumb” to the effect that each additional 100 IU of vitamin D/day, raises serum 25(OH)D by approximately 1 ng/mL. That is, in fact, a “rounding up” for convenience of calculation. Several studies indicate that the response increment is closer to 0.7 ng/mL/100 IU. 8,9
Either way, 600 IU/d will not suffice without appreciable solar and dietary input. Furthermore, as is also widely recognized, 600 IU/d produces barely perceptible changes in individuals who are overweight or obese (now better than 50% of the US adult population). Hence the increase from the 1997 DRIs, while welcome, and certainly in the right direction, is simply inconsistent with current professional experience. It not only is inadequate, by itself, to meet even the panel’s recommended serum levels, but this internal inconsistency detracts from the credibility of the whole report inasmuch as it flies in the face of the everyday experience of clinicians who recommend supplements to their patients and measure the resulting responses.
At already noted, the panel indicated that it was uncertain about extra-skeletal benefits - benefits that might accrue at intakes above the new intake recommendations. At the same time, the panel raised the upper level intake “TUIL” to 4,000 IU/d. (The report acknowledges that intakes up to 10,000 IU/d are probably safe for everyone, and applied an uncertainty factor 10 to that 10,000 IU figure to generate the 4,000 IU TUIL. It is important to stress that the TUIL is not a “limit” and, instead, constitutes an assurance of safety for such an intake.) One should have thought that even a very simplistic, game-theory approach would have led to a guidance statement such as the following: “We do not know whether taking more vitamin D than we are currently recommending will help you, but it could, and we can assure you that supplemental intakes up to at least 4,000 IU per day are safe.” Such a statement couched, perhaps, in less straightforward language, would nevertheless provide guidance that both the public and governmental agencies could find useful. Instead, we now have only a confused public.
But beyond these errors and inconsistencies, serious as they are, lies a much deeper flaw in the approach taken by the panel, exemplified by a quote from one of the panel members to the New York Times at the time of the release of the report.11 The statement was simply that the “onus” (i.e., burden of proof) fell on anyone who claimed benefits for intakes higher than the panel’s current recommendations. This is an approach that is correct for drugs, which are foreign chemicals and which do carry an appropriately heavy requirement for proof. For drugs the position of privilege is given to the placebo. And in the current IOM report, the privilege is given to a serum 25(OH)D level that is effectively the status quo. We judge that that is exactly backward for nutrients. The privilege must instead be given to the intake that prevailed during the evolution of human physiology, the intake to which, presumably, that physiology is fine- tuned. So far as can be judged from numerous studies documenting the magnitude of the effect of sun exposure,12,13 the primitive intake would have been at least 4000 IU/d, and probably two to three times that level, with corresponding serum 25(OH)D levels ranging from 40 to 80 ng/mL. The fact that primitive levels would have been higher than current IOM recommendations does not, of course, prove their necessity today. But such intakes should be given the presumption of correctness, and the burden of proof must be placed on those who propose that lower intakes (and lower serum levels) are without risk of preventable dysfunction or disease. The IOM, in its report, has utterly failed to recognize or meet that standard.
Finally, we commend the IOM panel for their concern about safety, certainly an appropriate posture for a body crafting public policy. However, the standards adopted by the panel for taking into evidence papers indicating possible risk were, we note, far lower than those the panel required to indicate benefit. Additionally, many of the purported risks were, on their face, implausible and inconsistent with the experience of population subgroups that routinely have serum levels in the range mentioned by the panel as possibly risky (e.g., ~50 ng/mL). We note that one of the widely accepted Hill1 4 criteria for acceptance of observational data is precisely biological plausibility. Furthermore, we consider it highly implausible that serum levels such as prevailed during hominid evolution could carry more risk than benefit for the populations concerned. Had that been the case, one should have expected that natural selection would have eliminated those prone to such risks.
In this Perspective we have deliberately avoided a mind-numbing laundry list of the vast number of factual inaccuracies and misinterpretations in the report. We are informed that there is a request, through the Freedom of Information Act, to obtain the external review comments submitted to the IOM in response to a prepublication draft. When those materials become available, those interested can review the many problems with the IOM report in detail. For now, our recommendation to the American public is that the IOM report should be taken with a grain of salt (another nutrient the IOM finds risky).
1. Nutrition Business Journal; 2010 Supplement Business Report, Sept. 28, 2010.
2. IOM (Institute of Medicine). 2011 Dietary Reference Intakes for Calcium and Vitamin D. Washington, DC: The National Academies Press.
3. Trivedi DP, Doll R, Khaw KT. Effect of four monthly oral vitamin D3 (cholecalciferol) supplementation on fractures and mortality in men and women living in the community: randomised double blind controlled trial. Br Med J 2003;326:469-474.
4. Bischoff-Ferrari HA, Willett WC, Wong JB, Giovannucci E, Dietrich T, Dawson-Hughes B. Fracture prevention with vitamin D supplementation. JAMA 2005;293:2257-2264.
5. Bischoff-Ferrari HA, Willett WC, Wong JB, et al. Prevention of nonvertebral fractures with oral vitamin D and dose dependency. Arch Intern Med 2009;169:551-561.
6. Priemel M, von Domarus C, Klatte TO, et al. Bone mineralization defects and vitamin D deficiency: histomorphometric analysis of iliac crest biopsies and circulating 25- hydroxyvitamin D in 675 patients. J Bone Miner Res 2010;25:305-312.
7. Need AG, Horowitz M, Morris HA, Moore R, Nordin C. Seasonal change in osteoid thickness and mineralization lag time in ambulant patients. J Bone Miner Res 2007;22:757-761.
8. Heaney RP, Davies KM, Chen TC, Holick MF, Barger-Lux MJ. Human serum 25- hydroxy-cholecalciferol response to extended oral dosing with cholecalciferol. Am J Clin Nutr 2003;77:204-210.
9. Heaney RP, Armas LAG, Recker RR, Grote J, Horst RL. Vitamin D3 is more potent than vitamin D2 in humans. J Clin Endocrinol Metab 2011;(in press).
10. Hathcock JN, Shao A, Vieth R, Heaney RP. Risk assessment for vitamin D. Am J Clin Nutr 2007;85:6-18.
11. Kolata G. Report questions need for 2 diet supplements. New York Times, November 29,2010. www.nytimes.com/2010/11/30/health/30vitamin.html?
12. Holick MF. Environmental factors that influence the cutaneous production of vitamin D. Am J Clin Nutr 1995;62(suppl):638S-645S.
13. Armas LAG, Dowell S, Akhter M, et al. Ultraviolet-B radiation increases serum 25- hydroxyvitamin D levels: the effect of UVB dose and skin color. J Am Acad Dermatol 2007;57:588-893.
14. Hill AB. The environment and disease: association or causation? Proc R Soc Med 1965;58:295-300.