Most people are vitamin D deficient

Whether most people are deficient depends almost entirely on where the line is drawn, and the bodies that draw it do not agree. Against the threshold the US Food and Nutrition Board actually uses for deficiency, under 30 nmol/L, roughly one in six people worldwide falls below it, a figure two separate worldwide reviews reach independently, and in the United States the figure is smaller still. Against a 50 nmol/L line, just under half the world falls below. Against 75 nmol/L, about three quarters do. Those are three answers from the same body of data, and the Food and Nutrition Board does not use the highest of them as a deficiency cutoff.

The US numbers are the clearest illustration. Pooling NHANES from 2001 to 2018, 71,685 people weighted to represent the population, 2.6 percent were below 25 nmol/L and another 22.0 percent sat between 25 and 50. That leaves about three quarters of Americans at or above 50 nmol/L, which the Food and Nutrition Board describes as sufficient for most people. The most recent survey cycle, August 2021 to August 2023, moves further the same way: 15.6 percent of Americans sat between 30 and 49 nmol/L, down from 18.8 percent in 2007 to 2008, while the share above 125 nmol/L, the level the Board flags for possible adverse effects, rose from 2.2 to 7.6 percent, and reached 30.6 percent among people taking two or more vitamin D supplements. The NIH Office of Dietary Supplements fact sheet reports that most people in the United States take in less vitamin D than recommended, and that most nonetheless have sufficient serum concentrations, because sun exposure contributes what diet does not.

The thresholds themselves are contested rather than settled. The Food and Nutrition Board places deficiency risk below 30 nmol/L, possible inadequacy from 30 to 50, and sufficiency at 50 and above, while noting above 125 nmol/L can carry adverse effects. The UK currently defines deficiency at under 25 nmol/L, and a collective of UK researchers and clinicians published a position statement in 2026 arguing that it should be raised to 50 to match international practice. ODS states plainly that the serum concentrations associated with deficiency have not been definitively identified. A claim about how many people are deficient inherits all of that unsettledness.

There is a second source of spread that has nothing to do with populations. Different laboratory assays measuring the same blood return different numbers, and the difference is large enough to move national statistics. When NHANES data were recalibrated to reference standards, the share of the US population below 50 nmol/L rose from 22 to 31 percent. When three German national surveys were put through the same standardization, deficiency prevalence fell and sufficiency rose, leaving about 15 percent of German adults below 30 nmol/L. Same procedure, opposite directions, because the original assays were biased differently. The worldwide literature carries the same problem at scale: of 586 publications pooled in a 2025 review, 77 used a certified or standardized assay, and prevalence ran slightly higher in the studies that did not. The ODS fact sheet notes that a result can be falsely low or falsely high depending on the assay and the laboratory.

What every dataset shows is that the burden is not spread evenly, although not always in the same pattern. In standardized NHANES data covering 1988 to 2010, 6 to 10 percent of non-Hispanic white Americans were below 40 nmol/L, against 46 to 60 percent of non-Hispanic Black Americans. Both worldwide reviews find levels lower in winter and spring than in summer and autumn, with deficiency about 1.7 times as common in winter and spring in one of them, and lower in women. Latitude is less tidy: one review found low levels more common at high latitudes, the other found them slightly more common below 40 degrees, and its authors list clothing, skin pigmentation, diet and food fortification among the factors that vary from place to place alongside sunlight. The population-level answer and the individual answer are different questions, and a group average says nothing about any particular person's blood.