Another opinion based on research. From this week's New England Journal of Medicine....

Vitamin D Deficiency — Is There Really a Pandemic?

JoAnn E. Manson, M.D., Dr.P.H., Patsy M. Brannon, Ph.D., R.D., Clifford J. Rosen, M.D., and Christine L. Taylor, Ph.D.

N Engl J Med 2016; 375:1817-1820November 10, 2016

In recent years, numerous clinical research articles have concluded that large proportions of North American and global populations are “deficient” in vitamin D.1-3 Most of the evidence cited focuses on one of two observations: that many people have serum concentrations of vitamin D (i.e., 25-hydroxyvitamin D [25(OH)D]) below 20 ng per milliliter (50 nmol per liter), which the Institute of Medicine (IOM) estimated in 2011 was the appropriate level4; or that supplementation with 600 to 800 IU per day — the IOM Recommended Dietary Allowance (RDA) for adults — or more fails to achieve serum concentrations above 20 ng per milliliter in some study participants. Such conclusions, however, are based on misinterpretation and misapplication of the IOM reference values for vitamin D. Because such misunderstandings can have adverse implications for patient care, including unnecessary vitamin D screening and supplementation as well as escalating health care costs due to overscreening and overtreatment, it’s important to clarify the meaning of IOM reference values for vitamin D as they relate to both population health and clinical practice.

To understand the concept of nutrient “deficiency” or “inadequacy,” one needs to know how the IOM nutrient reference values are defined and what they reflect. The IOM develops these reference values, referred to as Dietary Reference Intakes (DRIs), for an array of nutrients. Central to the DRI concept is the biologic reality that the need for any nutrient varies from person to person, generally in a normal distribution across the population. These reference values include an Estimated Average Requirement (EAR) for the nutrient, which is the median of the distribution of human requirements.4 The EAR reflects the most likely requirement for the population, whereas a second DRI reference value, the RDA, reflects the estimated requirement for people at the highest end of the distribution. Practically everyone in the population (at least 97.5%, or within 2 SD of the median) will have a requirement below the RDA.

Because of vitamin D’s established role in bone health (postulated nonskeletal benefits remain under study), the EAR is set at 400 IU per day for persons 1 to 70 years of age and 600 IU per day for persons older than 70 — intakes corresponding to a serum 25(OH)D level of 16 ng per milliliter (40 nmol per liter). The RDAs are 600 IU per day and 800 IU per day, respectively, corresponding to a serum 25(OH)D level of 20 ng per milliliter (50 nmol per liter). Note that the EAR and RDA assume minimal to no sun exposure. Although obesity and overweight are associated with lower circulating concentrations of 25(OH)D, evidence on the relationship with bone health and any implications for modified dietary intake requirements for people with greater adiposity remain inconclusive.4 

A common misconception is that the RDA functions as a “cut point” and that nearly the entire population must have a serum 25(OH)D level above 20 ng per milliliter to achieve good bone health. The reality is that the majority (about 97.5%) of the population has a requirement of 20 ng per milliliter or less. Moreover, by definition of an average requirement, approximately half the population has a requirement of 16 ng per milliliter (the EAR) or less. These concepts are depicted in the population reference-value distribution shown in Panel A, which highlights the relationship between the EAR and the RDA.

In creating its framework for reference values, the IOM anticipated the inherent variability in nutrient requirements and therefore established — and verified by statistical modeling4 — the goal of achieving population levels above the EAR, not the RDA. However, the literature is replete with misapplications of the RDA that treat it as a cut point. Many studies establish “inadequacy” using the RDA, though it is actually at the upper end of the spectrum of human need. Clearly, this approach misclassifies as “deficient” most people whose nutrient requirements are being met — thereby creating the appearance of a pandemic of deficiency.

Applying the correct method to data from the National Health and Nutrition Examination Survey (NHANES) for 2007 through 2010 reveals that 13% of Americans 1 to 70 years of age are “at risk” for vitamin D inadequacy. Less than 6% are deficient in vitamin D [serum 25(OH)D levels <12.5 ng per milliliter4]. The utility of measurement of parathyroid hormone (PTH) concentrations for identifying the optimal level of vitamin D remains controversial; the relationship between serum 25(OH)D and PTH is inconsistent, and no clear threshold defining “sufficiency” has been established.4 Vitamin D is a nutrient of concern, but these levels of deficiency do not constitute a pandemic.

Furthermore, using the RDA-associated serum concentrations of vitamin D to judge whether population groups have inadequate levels or to set intake goals for populations inflates the estimated prevalence of inadequacy and overestimates the needed intake. Indeed, ensuring that 97.5% of the population attains or exceeds vitamin D levels of 20 ng per milliliter would require shifting the entire population to a higher intake (see graph in Panel B). This misapplication of RDA-associated concentrations could cause harm to people whose intake is pushed above the Tolerable Upper Intake Level (UL, the level at which there may be adverse effects), which the IOM has established as 4000 IU daily with a resulting serum 25(OH)D concentration of approximately 50 ng per milliliter (125 nmol per liter). A modeling study by Taylor et al. suggested that shifting the distribution of serum 25(OH)D concentrations in adults 19 to 70 years of age upward so that the RDA-associated concentration of 20 ng per milliliter was achieved in nearly everyone (all but 2.5% of the population) would mean that levels in some people would exceed the UL.5

This problem highlights the concern that universal screening based on inappropriate cut points might lead to routine supplementation in generally healthy populations with adequate vitamin D levels. A preferable option would be to encourage patients and the public to choose foods containing, or fortified with, vitamin D — an approach that will be facilitated by new regulations requiring that vitamin D content be listed on nutrition labels.

Although our focus here is providing clarity about the use of nutrient reference values for estimating the prevalence of inadequacy in population groups, these values are also relevant to clinical settings in which patients are counseled individually. The two key clinical questions are whether to screen for vitamin D deficiency and what vitamin D intake to recommend for individual patients. For optimal decision making, the central issue is whether the patient is generally healthy and free of major risk factors for vitamin D deficiency or whether he or she has a skeletal disorder or significant risk factors for vitamin D deficiency (such as osteoporosis, osteomalacia, malabsorption, use of medications [such as anticonvulsants] that can affect vitamin D metabolism, or institutionalization).4 For healthy patients, routine screening is not recommended by most medical organizations, and the pitfalls would be similar to those described above for population-based studies.

Although the average requirement can be used to estimate the probability that a patient’s 25(OH)D level reflects an inadequate intake, practical counseling on vitamin D intake for healthy patients would use the RDA intake as a guidepost, given that it is impossible to know a given patient’s actual requirement and the RDA will nearly always meet the needs of generally healthy people. For patients who are at high risk or who have a disorder related to calcium metabolism, targeted vitamin D assessment would be appropriate, and vitamin D supplementation at levels above the RDA may be necessary. Although clinical judgment and customized interventions can be used with individual patients, avoidance of overscreening and overprescribing of supplemental vitamin D remains important.