The Vitamin D Paradox — Why It's So Important and So Misunderstood

Vitamin D is essential. Without it, children develop rickets and adults develop soft, painful bones, and the molecule acts on tissues throughout the body. Yet the largest randomized trials of vitamin D supplements in generally healthy adults, including a 25,871-person trial, found no reduction in cancer or cardiovascular events. The two facts are not in conflict. They describe the difference between fixing a deficiency and adding more to a body that already has enough.

Educational content, not medical advice. Talk to your doctor before changing any medication, treatment, or routine. Information here reflects current research but every person's situation differs.

Key takeaways

• Vitamin D works more like a hormone than a typical vitamin: skin makes it from sunlight, then the liver and kidneys convert it to its active form.
• The evidence is strongest for preventing rickets and osteomalacia (adult bone softening) and weakest for the many conditions it was once expected to prevent.
• Blood-level targets are disputed. The U.S. Institute of Medicine linked 20 ng/mL (50 nmol/L) to meeting the needs of most people, while some experts argued for higher.
• Randomized trials in people who mostly started with adequate levels found no benefit for cancer, cardiovascular events, or depression.
• A 2024 Endocrine Society guideline advises against routine vitamin D testing and against doses above the standard recommended intake for healthy adults under 75.

What vitamin D actually is

Despite the name, vitamin D does not fit the usual definition of a vitamin, which is a nutrient your body cannot make. Your skin can make it. Once made, it behaves like a hormone: it travels through the blood, is chemically modified, and then binds receptors inside cells to change which genes are active. The general logic of that kind of signaling is explained in how hormones work.

An analogy helps. Think of vitamin D as a raw ingredient shipped to a factory. The skin produces a precursor, the liver does the first processing, and the kidney finishes the job. Only the finished product, called calcitriol, is the active hormone. A blood test usually measures the intermediate stored form, so a result is more like checking the warehouse than the shop floor.

Two forms

Vitamin D2 (ergocalciferol) comes from fungi, including mushrooms exposed to ultraviolet light, and from yeast. Vitamin D3 (cholecalciferol) is the form made in human skin and found in animal foods such as fatty fish and egg yolks. Supplements and fortified foods use one or the other.

The three steps of activation

1. Skin. Ultraviolet-B light converts 7-dehydrocholesterol, a molecule closely related to cholesterol, into previtamin D3, which then becomes vitamin D3 in the skin. The biology of the organ that does this is covered in how the skin works, and the cholesterol connection is explored in cholesterol explained.
2. Liver. The liver adds a hydroxyl group (an oxygen-hydrogen unit) to make 25-hydroxyvitamin D, written 25(OH)D. This is the form measured in blood tests. See how the liver works.
3. Kidney. The kidney adds a second hydroxyl group to make 1,25-dihydroxyvitamin D, the active hormone. The kidney's wider role is described in how the kidneys work.

The practical consequence is that sun exposure, diet, and supplements all feed the same pool. Your body does not know where a given molecule came from, a point that matters when comparing studies of sunlight with studies of pills.

A short history: how a disease led to a vitamin

Vitamin D was identified through a childhood disease. Rickets, in which growing bones soften and bend, was common in industrial cities. Observers noticed in several European countries between the 1820s and 1860s that children given cod liver oil did not develop it.

In the early twentieth century, experiments made the link explicit. Edward Mellanby produced rickets in dogs kept indoors on an oatmeal diet and cured it with cod liver oil, around 1918 to 1919. Elmer McCollum then found that cod liver oil treated with oxidation lost its vitamin A activity yet still cured rickets, which led him to conclude in 1922 that a distinct factor existed, which he named vitamin D. In 1923, Harry Steenbock and Archie Nelson confirmed this and showed that ultraviolet light could also relieve the disease. Adolf Windaus received the 1928 Nobel Prize in Chemistry for work on sterols and their connection with vitamins. Sources differ by a year on some of these dates, so the sequence matters more than any single year.

In the United States, milk fortification with vitamin D at 400 IU per quart began in the 1930s as a public health measure against rickets. The story fits a broader pattern in nutrition science, where a deficiency disease points to a missing factor, as in the work described in Eijkman's chickens and beriberi and the Lind scurvy trial.

Why levels fall: sun, season, and skin

The skin makes vitamin D only when ultraviolet-B light reaches it, and the angle of the sun controls how much arrives. A 1988 study by Anthony Webb, Lisa Kline, and Michael Holick in the Journal of Clinical Endocrinology and Metabolism tested this directly. In Boston (42.2 degrees north), sunlight on cloudless days from November through February produced no previtamin D3 in skin or in a laboratory solution of 7-dehydrocholesterol. In Edmonton (52 degrees north), the ineffective period ran from October through March. At 34 degrees and 18 degrees north, sunlight still produced it in midwinter.

The authors called this a "vitamin D winter" that lengthens with latitude. Readers should treat the exact months as a measurement from one study with one method, not a universal calendar.

Skin pigmentation also matters. Melanin, the pigment that protects against sun damage, absorbs some of the ultraviolet light that would otherwise drive production. Sunscreen reduces ultraviolet reaching the skin in the same way, and laboratory studies show reduced synthesis under it. Dermatology groups have disputed the claim that everyday sunscreen use causes widespread deficiency, because people rarely apply it perfectly or for long.

Sun exposure also carries a trade-off, since ultraviolet radiation is linked to skin cancer. This is one reason there is no single recommendation of how much sun to seek, and why advice about "safe" exposure varies between countries and skin types.

How blood levels are read, and why the cutoffs are disputed

Clinicians judge vitamin D status by the blood concentration of 25(OH)D. It is reported in two unit systems, which causes constant confusion. Nanomoles per liter (nmol/L) is common internationally, and nanograms per milliliter (ng/mL) is common in the United States. Multiply ng/mL by 2.5 to get nmol/L.

Threshold	ng/mL	nmol/L	Where it comes from
Commonly used deficiency cutoff	12	30	Used in the 2016 European analysis and in post hoc analysis of the D2d trial
Average requirement (Institute of Medicine)	16	40	Corresponds to the estimated average requirement for bone health
Meets needs of most people (Institute of Medicine)	20	50	Matches the 600 IU recommended intake for most adults
Higher target argued by some experts	30 or more	75 or more	Robert Heaney argued that bone outcomes needed more than 30 ng/mL
Toxicity (Endocrine Society definition)	about 150	375	Hypercalcemia becomes a risk

The Institute of Medicine (now the National Academy of Medicine) published its dietary reference intakes in 2011. Its committee concluded that 20 ng/mL met the needs of 97.5 percent of the population and described a U-shaped pattern for some outcomes, with risks at both low and high levels. Critics argued the cutoffs were too low. The Endocrine Society's 2024 panel took a different approach and could not set blood-level thresholds at all, because the available trials did not identify outcome-specific benefits tied to particular levels.

A further complication is the lab method. Different assays can give different numbers for the same blood sample, and the European analysis described below used an international standardization program to correct for that. A single reading from one lab is a rough guide rather than a precise verdict.

How common is low vitamin D?

Headline numbers vary because the answer depends on the cutoff. In 2016, Kevin Cashman and colleagues pooled standardized data from 55,844 people in Europe and published the result in the American Journal of Clinical Nutrition. They found that 13.0 percent had 25(OH)D below 30 nmol/L, with a higher rate in winter samples (17.7 percent) than in summer samples (8.3 percent). Using a 50 nmol/L cutoff, the prevalence was 40.4 percent. Dark-skinned ethnic subgroups had a prevalence of the lower cutoff that was 3 to 71 times higher than white populations.

Cutoff	Share below it in the pooled European data
30 nmol/L (12 ng/mL), all samples	13.0%
30 nmol/L, winter samples	17.7%
30 nmol/L, summer samples	8.3%
50 nmol/L (20 ng/mL), all samples	40.4%

Source: Cashman et al., 2016, American Journal of Clinical Nutrition, as summarized in the abstract.

The size of that gap between the two cutoffs is a lesson in itself. If 13 percent have a level most clinicians would call deficient and 40 percent fall below the Institute of Medicine's 20 ng/mL, then the debate over cutoffs changes the apparent scale of the problem by a factor of three. Claims that a fixed number of people worldwide are "deficient" depend entirely on which line someone draws.

What deficiency actually does

Severe, prolonged deficiency causes rickets in children and osteomalacia in adults. Both involve poorly mineralized bone, meaning the bone matrix does not take up enough calcium and phosphate. Children with rickets can develop bowed legs and growth problems, and adults with osteomalacia can have bone pain and muscle weakness. The mechanism is that active vitamin D increases calcium absorption from the gut, so with too little, the body cannot keep bone and blood calcium in balance. Ordinary bone biology, including how bone is continuously rebuilt, is covered in how bones remodel throughout life.

Rickets and osteomalacia are well established, treatable, and rare in places with fortification and adequate nutrition. Everything beyond this point on the evidence ladder is weaker, which is where the paradox begins.

What the big trials found

For decades, observational studies reported that people with low vitamin D had more heart disease, cancer, infections, and depression. That is a strong hint, but a hint is not proof, and the proof requires randomized trials. Their logic is explained in why randomized trials are powerful.

Outcome	Trial or review	Design and size	Result
Cancer and cardiovascular events	VITAL (Manson et al., NEJM)	25,871 U.S. adults, 2,000 IU/day of D3, median 5.3 years	Cancer HR 0.96 (0.88 to 1.06); major cardiovascular events HR 0.97 (0.85 to 1.12)
Type 2 diabetes in prediabetes	D2d (Pittas et al., NEJM 2019)	2,423 adults, 4,000 IU/day, about 2.5 years	HR 0.88 (0.75 to 1.04); not statistically significant
Depression	VITAL-DEP (Okereke et al., JAMA 2020)	18,353 adults, 2,000 IU/day, median 5.3 years	No significant difference from placebo
Respiratory infections	Martineau et al., BMJ 2017	25 trials, 11,321 participants	Overall adjusted odds ratio 0.88 (0.81 to 0.96)
Respiratory infection and COVID-19	CORONAVIT (Jolliffe et al., BMJ 2022)	6,200 UK adults, test-and-treat	No reduction in infection or COVID-19
Fractures and falls	Bolland et al., Lancet Diabetes Endocrinol 2018	81 trials, 53,537 participants	No reduction in total fracture, hip fracture, or falls

VITAL

VITAL is the largest trial. It randomized 25,871 U.S. men aged 50 or older and women aged 55 or older to 2,000 IU of vitamin D3 daily or placebo, in a design that also tested omega-3 fatty acids. After a median 5.3 years, 793 people in the vitamin D group and 824 in the placebo group had been diagnosed with cancer, a hazard ratio of 0.96. Major cardiovascular events numbered 396 and 409, a hazard ratio of 0.97. The authors reported no excess risk of high calcium or other adverse events (Manson et al., NEJM, via PMC).

A hazard ratio of 1.00 means no difference, and a confidence interval that spans 1.00 means the data are compatible with no effect. Those intervals do span it.

D2d and diabetes

The D2d trial gave 4,000 IU daily to adults with prediabetes and followed them for about 2.5 years. New diabetes arose in 293 people on vitamin D and 323 on placebo, a hazard ratio of 0.88 with a confidence interval from 0.75 to 1.04. A post hoc analysis of the 103 participants with deficiency (below 12 ng/mL) reported a hazard ratio of 0.38. Because post hoc subgroup results are exploratory and the group was small, this is a hypothesis for further trials, not a conclusion. For background on the disease itself, see type 2 diabetes biology.

Depression

VITAL-DEP, an ancillary study within VITAL, followed 18,353 adults and found no significant difference in depression risk or mood scores between vitamin D and placebo. The authors concluded the results do not support vitamin D for depression prevention in the general population, and they added that benefit in particular subgroups cannot be ruled out. For the neuroscience of the condition, see depression biology.

Infections and COVID-19

The picture for infections is more nuanced. Adrian Martineau and colleagues pooled individual participant data from 25 trials (11,321 participants, 10,933 in the primary analysis) and found an overall odds ratio of 0.88 for acute respiratory infection, a small protective effect. The CORONAVIT trial in 2022 then tested a test-and-treat approach, offering vitamin D (800 or 3,200 IU/day for six months) to people whose blood levels were below 75 nmol/L, against no offer. Among 6,200 adults, 299 had the primary outcome, and the approach did not reduce infections or COVID-19. The authors noted a high baseline of suboptimal levels in the population. How the immune response works in general is covered in how the immune system recognizes threats.

Bones, fractures, and falls

One of the most surprising reversals concerns the original headline use of vitamin D. In a 2018 meta-analysis, Mark Bolland, Andrew Grey, and Alison Avenell pooled 81 randomized trials involving 53,537 participants. Supplementation did not reduce total fractures, hip fractures, or falls, even in trials of doses above 800 IU per day, and effects on bone density were small or inconsistent. The authors made an exception for preventing rare conditions such as rickets and osteomalacia in high-risk groups. They also noted that only 4 trials, involving 831 participants, reported mean baseline levels below 25 nmol/L, so the question for severe deficiency was less well tested.

The U.S. Preventive Services Task Force took a similar line in 2018 and recommended against vitamin D supplementation to prevent falls in community-dwelling older adults, reversing a 2012 position. A falls benefit, once widely described as well supported, is no longer an accurate reading of the trial evidence.

Why the paradox arises

How can a molecule be essential and yet add nothing when given to healthy people? Four explanations fit the evidence, and several probably contribute.

Trials mostly enrolled people who were not deficient. An essential nutrient follows a saturation curve. Below a threshold, each increment matters a great deal. Above it, more produces little or nothing. Think of a car's oil level: running a quart low causes real harm, but adding a second quart on top of a full engine does nothing useful. A trial that gives 2,000 IU to people whose blood levels are already adequate tests the flat part of the curve. The Endocrine Society panel made this point, noting that many large trials were not designed for the outcomes they reported and that participants often began with levels most would consider adequate.

Low vitamin D can be a marker, not a cause. People who are ill, inactive, or housebound spend less time outdoors, so illness can lower vitamin D rather than the reverse. In that case, a low 25(OH)D is like a smoke detector going off. It tells you something is wrong without being the fire. The distinction between association and causation is spelled out in how correlation differs from causation.

Genetic evidence points to a threshold. A 2022 Mendelian randomization study by Sutherland, Zhou, and Hyppönen in Annals of Internal Medicine used genetic variants as a natural experiment in 307,601 UK Biobank participants of European ancestry. The association with all-cause mortality was L-shaped. Genetically predicted odds of death were 25 percent higher at 25 nmol/L than at 50 nmol/L, rose more steeply at lower concentrations, and plateaued by about 50 nmol/L. This suggests that harm is concentrated at severely low levels, which matches the flat curve above. The study was limited to people of European ancestry.

The measurement may not mean the same thing in everyone. This is the least intuitive piece, and it has a racial dimension.

The black-white paradox

In a 2013 study in the New England Journal of Medicine, Camille Powe and colleagues analyzed about 2,085 black and white adults from a Baltimore cohort. Black participants had lower total 25(OH)D (an average of 15.6 versus 25.8 ng/mL) and lower levels of vitamin D-binding protein, the carrier that transports vitamin D in blood (168 versus 337 micrograms per milliliter). Yet black participants had higher bone mineral density. Genetic variation explained most of the difference in binding protein. The authors estimated that bioavailable vitamin D, meaning the portion not tightly bound to its carrier, was similar between groups (2.9 versus 3.1 ng/mL in the subgroup analyzed), and they suggested that the standard test might overdiagnose deficiency in black adults (ScienceDaily summary).

The finding has been contested. A 2016 letter in the same journal by Carrie Nielson and colleagues reported that when free 25(OH)D was measured directly, or calculated using a polyclonal binding-protein assay, levels were lower in black men than white men in the United States, and that the earlier result depended on the monoclonal assay used (NEJM letter, via PMC). The practical lesson is cautious: a single total 25(OH)D threshold may not apply equally across populations, and the question is unresolved. That is also a reason the 2024 guideline did not recommend routine testing on the basis of complexion alone.

How to read a vitamin D trial

Because so many headlines rest on single studies, it helps to know what questions to ask of one. The same five questions apply to most nutrient research.

Who was enrolled? A trial in people who already have adequate blood levels cannot say much about people with low ones. The Endocrine Society panel noted that study populations often began with levels most would consider adequate. The few trials in people with very low baselines are small, which is why Bolland and colleagues flagged that just 4 trials in their pooled analysis, with 831 participants, started from a mean level below 25 nmol/L.

What dose and for how long? A dose of 2,000 IU for five years tests a different question than a single large bolus or a six-month course. Tripkovic's meta-analysis of D2 versus D3 is a reminder that even the form and the dosing schedule can change the blood level achieved.

What was the outcome? Many trials were designed around one outcome and then reported others as secondary or exploratory. The D2d trial's finding in the 103 deficient participants is a good example. It looks striking, but because it was a post hoc subgroup analysis, it is a lead to test, not a result to rely on.

How big was the trial? Bolland and colleagues observed that smaller, shorter trials tended to find stronger effects of vitamin D than larger, longer ones, a pattern familiar across medicine, in which small studies are noisier and more likely to overstate. When a large trial such as VITAL disagrees with earlier small ones, the larger result is usually the more reliable.

Did anyone replicate it? An effect seen once is a hypothesis. An effect seen across independent trials and meta-analyses is a finding. The infection data show how this plays out: pooled analyses found a small benefit, and a later large trial did not.

None of these checks needs a statistics degree. They explain why the same molecule can look miraculous in observational headlines and modest in trials, and why a careful reader waits for the randomized evidence before treating a supplement as proven.

What guidelines say

Official recommendations come in two types: how much to eat (intake) and what to do about blood tests. They are easy to blur together.

Source	Position
National Institutes of Health / Institute of Medicine	Recommended dietary allowance of 15 mcg (600 IU) per day for ages 1 to 70 and 20 mcg (800 IU) per day from age 71; infants 10 mcg (400 IU)
Institute of Medicine tolerable upper intake	4,000 IU per day for most adults, with lower limits for children (1,000 IU for infants up to 6 months, 1,500 IU for 6 to 12 months, 2,500 IU for ages 1 to 3, 3,000 IU for ages 4 to 8)
Endocrine Society, 2024	Suggests against supplement doses above the standard intake for healthy adults under 75, and against routine 25(OH)D testing in all studied populations

The Endocrine Society's 2024 guideline, "Vitamin D for the Prevention of Disease," did identify groups where supplementation beyond the standard intake may help. These are children and adolescents aged 18 and younger (to prevent rickets and reduce respiratory infections), adults 75 and older (potential mortality benefit), pregnant people (to potentially lower risks such as pre-eclampsia and preterm birth), and people with prediabetes (to potentially slow progression to diabetes). For adults 50 and older with an indication for supplementation, it suggests daily lower-dose vitamin D over non-daily, higher-dose regimens (Endocrine Society news release).

The upper intake level is a safety ceiling, not a target. The committee that set it cautioned against treating it as a recommended dose. Whether a particular person should supplement is a conversation for a clinician who knows their medical history, diet, and medications. Nothing here is a dosing recommendation.

Does the form matter: D2 versus D3?

Both forms raise blood levels, but not equally. A 2012 meta-analysis by Tripkovic and colleagues in the American Journal of Clinical Nutrition found that D3 raised serum 25(OH)D more effectively than D2. The advantage appeared mainly with larger single doses given at intervals, and it was lost with daily dosing, so the difference is smaller than some marketing implies. The authors called for more work on metabolic pathways and differences by age, sex, and ethnicity.

Too much: what toxicity actually looks like

Vitamin D toxicity is real but uncommon, and it comes almost entirely from extremely high doses. Because active vitamin D increases calcium absorption, the danger is hypercalcemia, meaning too much calcium in the blood, which can cause nausea, vomiting, weakness, frequent urination, and eventually kidney damage. The Endocrine Society defines toxicity at 375 nmol/L, which is about 150 ng/mL.

A review of case reports of toxicity from overcorrection of deficiency found patients with levels from 150 to 1,220 ng/mL and serum calcium of 11.1 to 23.1 mg/dL. Manufacturing and dosing errors feature prominently in published cases. One report described a supplement labeled at 400 IU per drop that delivered about 12,000 IU daily to an infant, producing a level of 422 ng/mL. Another described a product containing 2,000 IU per drop given at 50,000 IU per day, with a level of 294 ng/mL and calcium of 18.7 mg/dL.

Trial data show how far ordinary doses sit from this territory. VITAL, at 2,000 IU per day, reported no excess hypercalcemia, kidney stones, or gastrointestinal symptoms, and D2d, at 4,000 IU per day, found no significant excess of kidney stones or high calcium. The point is not that large doses are risk-free but that the hazard lies in mega-doses and errors, not in the range typically discussed for supplements. Anyone using high-dose products or taking medicines that affect calcium should discuss it with a clinician.

Sources: sun, food, and supplements

Because the body treats all sources the same, the question is which is practical for a given person. Sunlight is limited by season and latitude, as the Webb study showed, and involves a skin cancer trade-off. An often-cited review by Michael Holick in 2007 suggested that exposing arms and legs for 5 to 30 minutes between 10 a.m. and 3 p.m. twice a week could be sufficient for many people, but that estimate depends on latitude, season, and skin tone, and it comes from a clinician's review rather than a controlled trial.

Food sources are limited. Fatty fish and egg yolks contain some, and fortified products such as milk and some cereals add more, a legacy of the rickets campaigns. Mushrooms exposed to ultraviolet light produce D2. Supplements are inexpensive and standardized, which makes them a common route when other sources fall short. The point of the paradox is that a person who already has adequate levels gains little from adding more, so the question depends on whether a real shortfall exists.

For readers comparing this to other nutrition questions, how much protein you actually need and metabolic syndrome and obesity work through similar gaps between popular belief and data.

Common misconceptions

"If a little is good, more is better." Above adequacy, the curve is flat for most outcomes tested, and very high levels carry risk of hypercalcemia.

"Vitamin D is a general immune booster." The evidence is narrower. A pooled analysis found a small reduction in respiratory infections, while a large 2022 trial of a test-and-treat approach found none for infections or COVID-19.

"A blood test is the only way to know you are fine." The 2024 guideline suggests against routine testing in all studied groups, partly because no outcome-specific benefit tied to levels has been shown. Testing may still be reasonable in specific clinical situations, which is a matter for a clinician.

"Observational studies proved vitamin D prevents disease." They showed associations, and randomized trials mostly failed to confirm causation in healthy adults. See why headlines overstate breakthrough science for how such gaps arise.

"One cutoff applies to everyone." The cutoff itself is contested, and the Powe study suggests the total 25(OH)D measurement may carry different meaning across populations.

Frequently asked questions

Q: Is vitamin D really a vitamin? A: Technically it behaves like a hormone. The skin makes it from sunlight, the liver and kidneys convert it, and the active form binds receptors in many tissues. It still counts as a vitamin in nutrition because people who lack sun exposure must get it from food or supplements.

Q: What blood level counts as deficient? A: There is no single agreed number. Many studies use below 30 nmol/L (12 ng/mL) for deficiency. The Institute of Medicine tied 50 nmol/L (20 ng/mL) to meeting the needs of most people, and some experts argued for 30 ng/mL or more. The 2024 Endocrine Society panel did not set thresholds.

Q: Does vitamin D prevent cancer or heart disease? A: In the VITAL trial of 25,871 adults taking 2,000 IU per day for a median of 5.3 years, vitamin D did not significantly reduce cancer or major cardiovascular events. The trial mostly enrolled people without severe deficiency, so it does not test what happens when a true shortfall is corrected.

Q: Does it help with COVID-19 or colds? A: Evidence is mixed. A pooled analysis of 25 trials found a small reduction in acute respiratory infections overall, but the 2022 CORONAVIT trial of 6,200 UK adults found no reduction in infections or COVID-19 with a test-and-treat approach.

Q: Can you get too much vitamin D? A: Yes, though it is uncommon and usually involves very high doses or dosing errors. The result is hypercalcemia. The Endocrine Society defines toxicity at about 150 ng/mL (375 nmol/L). The standard upper intake level for most adults is 4,000 IU per day.

Q: Is D3 better than D2? A: A 2012 meta-analysis found D3 raised blood levels more than D2, mainly with larger intermittent doses, and the difference disappeared with daily dosing. Which one suits a person depends on diet, preference, and a clinician's advice.

Q: Why do some people with low readings have healthy bones? A: A 2013 study found black American adults had lower total 25(OH)D but higher bone density, and estimated similar bioavailable levels. The finding is debated because later direct measurements gave different results. It suggests that total 25(OH)D alone may not tell the whole story.

The takeaway

Vitamin D is a case study in how an essential nutrient can be both vital and overhyped. Deficiency causes bone disease, and low levels are common in winter and at higher latitudes, so correcting a genuine shortfall matters. Beyond that, large randomized trials in mostly adequate populations repeatedly found little or no benefit, and observational links likely reflect a mix of confounding and thresholds. The best current reading is that the benefits cluster at the low end of the range, which is why current guidance focuses on specific groups instead of everyone. When a headline claims that vitamin D prevents a disease, the right question is who was studied, what their levels were, and what the randomized trials showed.

Related Reading

• The Science of Sleep — Why You Need It and What Happens Without It
• Why Vanilla Is So Expensive — The Biology and Economics of the World's Second-Priciest Spice
• The Science of Exercise — Why It Works and How Much You Actually Need
• Why Wagyu Beef Is So Expensive — The Science of the World's Most Marbled Meat
• How Much Caffeine Is Too Much? The Science of Coffee and Energy Drinks