Deep Dive: Why You No Longer Need to Test for Vitamin D

Clinical Deep DivesGenerated July 2026 · Clinical Deep Dive0 CommentsA 25-hydroxyvitamin D [25(OH)D] comes back at 22 ng/mL. The patient is asymptomatic. Ten years ago you'd have called that insufficient and started a supplement. Today, the same number might not even exist as a category — because in 2024, the Endocrine Society dropped the 30 ng/mL sufficiency target it wrote in 2011, and told physicians to stop routinely testing healthy adults altogether (JCEM)Nobody updated the annual physical order set. Vitamin D remains one of the most-ordered labs in American primary care, and testing volume hasn't meaningfully budged since the reversal. That gap between what the guideline says and what the requisition form does is where the breakdown occurs.What actually changedThe 2024 guideline (JCEM) is narrower than its 2011 predecessor in almost every direction. For healthy adults under 75, it suggests against supplementing beyond the standard Recommended Dietary Allowance (RDA) (600–800 IU/day) and against routine 25(OH)D testing — including in patients with obesity or darker skin pigmentation, two groups the 2011 version flagged for testing. It also declines to name any target blood level for disease prevention, because, as guideline chair Marie Demay, MD, of Harvard Medical School and Massachusetts General Hospital put it when the guideline dropped: "we do not recommend routine testing for vitamin D levels in any of these groups" — referring to children, pregnant patients, adults 75 and older, and people with prediabetes, the four groups the guideline does carve out for empiric, no-testing supplementation (Endocrine Society).The evidence base behind the reversal is four large, concordant, null trials: VITAL in the US (25,871 adults, no reduction in cancer or cardiovascular events) (NEJM), D-Health in Australia, FIND in Finland, and — the one that kills the "the null trials were all American" objection — DO-HEALTH, a five-country European trial of 2,000 IU/day in adults 70 and older that found no benefit on fractures, infections, blood pressure, physical performance, or cognition, even though 40% of participants started out deficient (JAMA, UEF Connect). USPSTF has landed in the same place for over a decade, calling the evidence on screening asymptomatic adults insufficient to weigh benefits against harms — first in 2014, again in 2021 (USPSTF).Two camps, one number, no resolutionThis isn't a new fight — it's a rematch. In 2011, the Institute of Medicine (IOM) set adequacy at ≥20 ng/mL, covering 97.5% of the population, and explicitly warned that higher cut-points were overstating how much deficiency actually existed (NAM/NCBI). The Endocrine Society's own 2011 guideline disagreed, setting sufficiency at ≥30 ng/mL based on calcium-absorption and parathyroid hormone (PTH) suppression data (JCEM). The 2024 guideline effectively sides with the IOM — and Michael Holick, MD, PhD, the Boston University endocrinologist who chaired the 2011 guideline, has pushed back hard, arguing the newer document leaned too exclusively on RCTs and discounted decades of association data linking low vitamin D to "poor pregnancy outcomes, early childhood dental caries, increased risk for autoimmune disorders, increased risk for upper respiratory tract infections including COVID-19, advancement of prediabetes to type 2 diabetes, cardiovascular disease, neurocognitive dysfunction, mortality and accelerating mortality from deadly cancers" (Endocrine Practice).Neither side has a trial that randomized patients to different serum targets — the entire dispute rests on how to weight the same PTH and absorption data. That's genuinely unresolved, and it's worth knowing which camp your specialist colleagues are in before you argue with them about a 24 ng/mL result.Four groups still get vitamin D, no blood draw requiredThe 2024 guideline carves out empiric, higher-than-RDA supplementation, without testing, for:Children and adolescents (1–18): to prevent rickets and possibly reduce respiratory infections.Pregnant patients: potential reduction in pre-eclampsia, preterm birth, and neonatal mortality.Adults 75 and older: a modest mortality signal.Adults with high-risk prediabetes: to slow progression to diabetes.The prediabetes call is the one worth pressure-testing. It rests on a pooled analysis of three trials showing a 15% relative reduction in progression (about 3 percentage points absolute over three years — a number needed to treat (NNT) near 30) (Endocrine Society JCEM). But every one of those trials was null on its own primary endpoint, and one used an active vitamin D analogue rather than the cholecalciferol (vitamin D3, the supplement form) you'd actually prescribe. The American Diabetes Association has seen the same pooled data and still won't issue a positive recommendation, citing uncertain risk-benefit at the doses studied (ADA). Two major US bodies, same trials, different conclusions — that's a live disagreement, not a rounding error.Where testing still earns its keepThe "don't test the healthy" message has a hard boundary, and it's the same boundary the 2011 guideline drew: chronic kidney disease (CKD), malabsorption (celiac, inflammatory bowel disease (IBD), bariatric surgery), granulomatous disease, hepatic failure, and patients on enzyme-inducing medications. None of that changed in 2024 — the new guideline narrows testing in the general population, not in patients with an actual reason for impaired vitamin D metabolism.What your peers are saying"The goal of this guideline was to address the vitamin D requirements for disease prevention in a generally healthy population with no underlying conditions that would put them at risk of impaired vitamin D absorption or action... we do not recommend routine testing for vitamin D levels in any of these groups." — Marie B. Demay, MD, Harvard Medical School and Massachusetts General Hospital, chair of the Endocrine Society 2024 guideline panel (Endocrine Society, 2024)The bottom line of these guidelines is that there is a very limited role for vitamin D supplementation and for screening for vitamin D deficiency in the general population. - JoAnn E. Manson, MD, DrPH (Medscape)I have little truck with the wildly popular supplement. In all of clinical research, I believe that there is no molecule with stronger data for correlation and weaker data for causation. - F. Perry Wilson, MD, MSCE (Medscape)The wildcard: multiple sclerosisThe newest data point complicates the tidy "avoid high-dose bolus dosing" rule. The D-Lay MS trial, published in JAMA in 2025, gave patients with clinically isolated syndrome (CIS) either 100,000 IU of cholecalciferol (vitamin D3) every two weeks or placebo, for 24 months. Disease activity — relapse or new MRI lesions — occurred in 60.3% of the vitamin D group versus 74.1% on placebo (HR 0.66 [95% CI, 0.50–0.87]; P=0.004; NNT 7.2) (JAMA).Two caveats keep it from being a mandate. The result was carried by MRI lesion counts, not clinical relapse rate, which didn't reach significance on its own (17.9% vs. 21.8%; HR 0.69 [95% CI, 0.42–1.16]; P=0.16). And it's a single trial in a narrow population — recent clinically isolated syndrome, not established or progressive relapsing-remitting multiple sclerosis (RRMS). It's a reason to check and correct vitamin D in a patient with a new demyelinating event. It is not evidence for population-wide high-dose dosing.Bottom line Don't screen asymptomatic, otherwise healthy adults for vitamin D — the evidence against benefit is now large, international, and consistent.For patients already taking 1000–4000 IU/day of vitamin D on their own, there is no trial evidence of harm at these doses, and advising a return to the standard RDA is reasonable but not mandatory — the guideline addresses testing and new empiric prescribing, not existing supplementation. Patients with obesity or darker skin pigmentation who have been tested do not require routine retesting; absent a new clinical indication, no further workup is needed regardless of prior results.Do supplement, without testing, in kids, pregnancy, patients 75 and up, and prediabetes (while knowing the ADA doesn't fully endorse that last one). Keep testing where a real disease process makes deficiency plausible: CKD, malabsorption, granulomatous disease. And if a patient presents with a first demyelinating event, checking and correcting vitamin D now has trial support behind it — nowhere else does an intermittent high-dose bolus.DEEP DIVE:Vitamin D: A Low Number Is Not a DiseaseMedscape Medical News © 2026 WebMD, LLCSend comments and news tips to news@medscape.net. Cite this: Stop Testing Vitamin D in Healthy Patients, Says the Society That Told You to Start - Medscape- July 23, 2026.