The Vitamin D Controversy

Few positions in the Root Cause Protocol provoke as much pushback as Morley Robbins’s teaching on vitamin D. While conventional medicine, functional medicine, and most of the “wellness” supplement industry now recommend daily cholecalciferol (D3) capsules — often at 5,000 to 10,000 IU/day — Robbins argues that this practice is biochemically incoherent and clinically harmful for many of the people taking it. The RCP holds that synthetic D3 supplementation depletes magnesium, suppresses ceruloplasmin, antagonizes retinol, and substitutes a steroid hormone for a true nutritional input that the body is designed to make from sunlight and dietary cholesterol. Mainstream endocrinology counters that the harms are overstated, that randomized trials show no excess risk at typical doses, and that confirmed deficiency carries real bone, immune, and metabolic consequences that pills can correct. The evidence is mixed, and a careful reader needs to see both sides before making a decision. This page lays them out.

An important framing note before we begin: this article does not endorse stopping vitamin D supplementation. It presents Robbins’s position, the biochemistry behind it, the trial evidence on both sides, where his arguments have mainstream support, and where they diverge from the consensus. Anyone with osteomalacia, severe deficiency (25-OH-D below 20 ng/mL with rachitic symptoms), pregnancy, or a physician-confirmed indication for D3 should make changes only with their clinician. The audience for this page is the person who has been told to take 5,000 IU/day for life, who feels worse after starting, and who wants to understand what an alternative framework would say.

Table of Contents

  1. The Robbins Position
  2. Vitamin D Is a Steroid Hormone, Not a Vitamin
  3. The Magnesium Burn Rate from Cholecalciferol
  4. D3 and the Retinol / Ceruloplasmin Pathway
  5. The Sunlight Argument
  6. The Endocrine Society Counter-View
  7. VITAL, D2d, and Other Major Trials
  8. Where Robbins’s Position Has Some Mainstream Support
  9. A Balanced Practical Take
  10. How to Monitor Status If You Stop or Reduce D3
  11. Key Research Papers
  12. Connections

1. The Robbins Position

Robbins argues that the rise of routine high-dose D3 supplementation since roughly 2005 has been one of the more damaging interventions in functional medicine. His argument rests on four claims, each of which we examine in detail later in this article.

  1. D3 burns through magnesium reserves. The two hydroxylation steps that convert dietary or skin-derived cholecalciferol into the active hormone calcitriol are magnesium-dependent enzymatic reactions. Pouring exogenous D3 into a magnesium-deficient body forces those enzymes to run at higher flux, accelerating magnesium depletion in a population that is already, on NHANES data, roughly half below the EAR for magnesium intake.
  2. D3 suppresses ceruloplasmin synthesis. Robbins argues that excess D3 reduces retinol availability through cross-receptor inhibition at the retinoid X receptor (RXR), which is the obligate partner for both the vitamin D receptor (VDR) and the retinoic acid receptor (RAR). Because retinol is a required cofactor for ceruloplasmin synthesis, this suppression cascades into reduced ferroxidase activity and the iron dysregulation pattern at the heart of the RCP.
  3. D3 disrupts the calcium / copper / iron axis. Calcitriol up-regulates intestinal calcium absorption. Without K2 and magnesium to direct calcium to bone, excess calcium can deposit in arteries and soft tissue. Robbins links this to the “calcium paradox” literature on vascular calcification.
  4. D3 substitutes a steroid hormone for a nutritional input. Robbins’s most philosophical objection: the body makes its own vitamin D from cholesterol when sunlight strikes skin. Replacing that endogenous synthesis with a daily oral hormone bypasses the natural feedback loops and signaling cascades that come with sun exposure (nitric oxide release, melatonin precursor formation, beta-endorphin response, circadian entrainment).

Robbins’s recommendation is to stop the D3 pill, get sun where geographically and seasonally feasible, eat real fermented cod-liver oil (which provides D in a food matrix with retinol and DHA), and supply the magnesium and retinol cofactors the body needs. He does not recommend stopping D3 in the face of confirmed clinical deficiency — the recommendation applies to “preventive” megadosing.

2. Vitamin D Is a Steroid Hormone, Not a Vitamin

The name “vitamin D” is a historical accident. In 1922, Elmer McCollum and his colleagues identified a fat-soluble factor in cod liver oil that cured rickets and named it the fourth vitamin (after A, B, and C). The naming stuck, but the molecule itself does not behave like a vitamin in any meaningful sense. True vitamins are organic compounds the body cannot synthesize and must obtain from diet. Vitamin D fails this definition: it is synthesized in skin from 7-dehydrocholesterol when ultraviolet B radiation in the 280–315 nm range provides the photolytic energy to open the B ring of the steroid backbone.

Structurally, cholecalciferol is a secosteroid — a steroid in which one of the four rings (the B ring) has been broken open. The active hormonal form, 1,25-dihydroxyvitamin D (calcitriol), binds the vitamin D receptor (VDR), which is a nuclear hormone receptor in the same superfamily as the estrogen receptor, the progesterone receptor, the glucocorticoid receptor, and the thyroid hormone receptor. The VDR partners with the retinoid X receptor (RXR) as a heterodimer, binds vitamin D response elements in the genome, and regulates an estimated 3% of the human protein-coding genome — somewhere between 200 and 1,000 genes depending on the tissue and the analytical methodology.

This is exactly how a hormone behaves. Hormones are exquisitely regulated by feedback loops: parathyroid hormone and FGF-23 modulate the kidney’s 1α-hydroxylase that activates D, calcitriol itself induces CYP24A1 to inactivate excess hormone, and serum calcium provides the upstream feedback signal. Robbins’s argument is that pouring exogenous hormone into this system — especially at doses well above what the skin would synthesize in a reasonable sun exposure — disturbs the regulation in ways that no analogous mainstream practice would tolerate. We do not, for example, prescribe daily cortisol or daily thyroxine to healthy adults to “prevent” deficiency. The asymmetric treatment of vitamin D, Robbins argues, reflects its inaccurate vitamin label, not sound endocrinology.

3. The Magnesium Burn Rate from Cholecalciferol

The biochemistry of vitamin D activation is the most concrete and least disputed element of Robbins’s argument. The conversion pathway runs through three obligate enzymatic steps:

  1. Cholecalciferol (D3) from sun, food, or pill enters circulation bound to vitamin D-binding protein.
  2. 25-hydroxylation in the liver by CYP2R1 (and to a lesser extent CYP27A1) produces 25-hydroxyvitamin D (calcidiol), the storage form measured on standard lab panels. This reaction requires magnesium-dependent enzyme function and Mg-ATP.
  3. 1α-hydroxylation in the kidney by CYP27B1 produces 1,25-dihydroxyvitamin D (calcitriol), the active hormone. This reaction is also magnesium-dependent.
  4. Inactivation by CYP24A1 produces 24,25-dihydroxyvitamin D, removing excess active hormone — and this CYP enzyme is similarly magnesium-dependent.

The 2018 review by Uwitonze and Razzaque in the Journal of the American Osteopathic Association documented the magnesium dependence of all three CYP enzymes in the vitamin D pathway and made the clinical observation that low magnesium status is associated with poor 25-OH-D activation, with people unable to raise their calcitriol meaningfully even on high D3 supplementation, and that magnesium repletion alone (no additional D3) often improves measured vitamin D status. The 2019 Reddy review reached the same conclusion: magnesium is rate-limiting for vitamin D activation, and supplementing D3 in magnesium-deficient subjects forces the activation enzymes to run at higher demand, accelerating a depletion that was already present.

Robbins extends this observation to its logical conclusion. If magnesium is required to activate vitamin D, and if roughly half of the U.S. adult population is below the EAR for magnesium intake, then the cohort being prescribed daily 5,000 IU D3 capsules is largely the same cohort that cannot fully metabolize the D3 they are taking. The result is twofold: storage-form 25-OH-D rises (because the liver can perform 25-hydroxylation reasonably well even at marginal magnesium status), but conversion to active calcitriol may not rise proportionately, and the magnesium reserve is being drained to keep up with the imposed demand. Robbins calls this the magnesium burn rate from D3 supplementation, and he treats it as the single most important harm of the practice.

Mainstream endocrinology does not dispute the magnesium dependence. The 2024 Endocrine Society guideline on vitamin D explicitly recommends adequate magnesium intake alongside D supplementation. Where Robbins and the mainstream diverge is in the clinical implication: the mainstream view is “take both,” while Robbins argues that for many patients the right answer is to repair magnesium first and then revisit whether D3 is needed at all.

4. D3 and the Retinol / Ceruloplasmin Pathway

The second mechanism in Robbins’s argument is more contested. He claims that excess D3 down-regulates retinol-mediated transcription, and because retinol is a required cofactor for ceruloplasmin synthesis, suppressed retinol signaling means suppressed ceruloplasmin and the cascade of iron dysregulation that follows.

The biochemical premise has support. The vitamin D receptor and the retinoic acid receptor (RAR) both heterodimerize with RXR to drive transcription. RXR is a shared, finite resource: when one nuclear receptor partner dominates, it can effectively monopolize available RXR and reduce signaling through the other. Bastien and Rochette-Egly’s 2004 review in Gene on the molecular biology of retinoid signaling laid out the competitive dynamics among VDR, RAR, PPAR, and TR for RXR partnership. So the molecular mechanism for “excess D3 reduces retinol signaling” is real.

What is harder to establish is the clinical magnitude. There are no large RCTs in humans showing that 5,000 IU/day D3 for two years measurably suppresses retinol-driven gene expression or ceruloplasmin levels, and there are no trials directly testing the “D3 lowers ceruloplasmin” hypothesis. Robbins draws the connection from the molecular biochemistry plus clinical observations from the RCP coaching network rather than from controlled trial data. Mainstream nutritionists view this as overreach: the receptor competition is real, but the dose at which it produces clinically meaningful retinol suppression is not established.

Robbins also cites the 2013 Mawson hypothesis (in Medical Hypotheses) that what is often diagnosed as “vitamin D deficiency” might in some cases reflect retinol toxicity or retinol-D imbalance rather than insufficient D intake. This is, by the journal’s own admission, a hypothesis paper rather than evidence, and the broader field has not adopted the framing.

5. The Sunlight Argument

Where Robbins and even his mainstream critics agree is that sunlight is the preferred source of vitamin D when geographically and seasonally feasible. The skin-based synthesis pathway has features that no oral pill can replicate.

UVB radiation between 280 and 315 nanometers strikes 7-dehydrocholesterol stored in the plasma membranes of keratinocytes in the skin’s basal and spinous layers. The B ring of 7-dehydrocholesterol opens, producing previtamin D3, which then undergoes thermal isomerization to cholecalciferol over several hours. Skin can produce a maximum of approximately 10,000 to 25,000 IU of cholecalciferol from a 15-to-20-minute full-body exposure at solar noon in summer at lower latitudes — though this varies enormously with skin pigmentation, age, latitude, and season. Crucially, the skin has a self-limiting feedback mechanism: when previtamin D3 production is excessive, the same UVB radiation converts it to inactive lumisterol and tachysterol, capping the maximum dose. You cannot reach toxic vitamin D levels from sun exposure alone.

Sun exposure also triggers a constellation of effects entirely absent from a D3 capsule:

Robbins’s argument is that the entire signaling cascade triggered by sunlight is missing from a D3 capsule, and that treating the cascade as if it were equivalent to a single hormone is a category error. Even Michael Holick — the most prominent mainstream advocate for vitamin D supplementation — agrees that sunlight is preferred when feasible. Where Holick and Robbins diverge is whether oral D3 is an acceptable substitute when sun is not available. Holick says yes; Robbins says only as a last resort and only with the cofactors.

6. The Endocrine Society Counter-View

The mainstream position deserves a fair hearing. The 2011 Endocrine Society Clinical Practice Guideline, lead-authored by Holick, established the now-standard target of 25-OH-D above 30 ng/mL with daily intake of 1,500 to 2,000 IU/day for adults at risk of deficiency, and 800 to 1,000 IU/day for the general adult population. The Institute of Medicine took a more conservative line at the same time, setting an RDA of 600 to 800 IU/day and a sufficiency threshold of 20 ng/mL. The 2024 Endocrine Society update was notably more cautious than the 2011 version: it dropped the recommendation for population-wide screening, lowered some intake recommendations, and limited the strongest endorsements to high-risk groups (pregnant women, people over 75, children, those with prediabetes).

The Endocrine Society’s case rests on three pillars: (1) confirmed deficiency below 20 ng/mL produces rickets in children and osteomalacia in adults — a real, treatable disease; (2) supplementation reliably raises 25-OH-D, and at the doses recommended (1,000 to 4,000 IU/day) the safety record over decades of clinical use is very good; (3) observational epidemiology has linked low 25-OH-D to higher rates of cardiovascular disease, autoimmune conditions, certain cancers, and all-cause mortality, even if RCT evidence for benefit at typical doses is weaker (see next section).

The Endocrine Society does not endorse the high-dose protocols (10,000 IU/day or weekly 50,000 IU bolus) common in some functional-medicine practices, and explicitly warns about hypercalcemia risk above sustained 25-OH-D levels of 100 ng/mL. So the mainstream position is more conservative than it is sometimes portrayed, and the gap between “Endocrine Society 2024” and “Robbins 2024” is narrower than the rhetoric on either side suggests.

7. VITAL, D2d, and Other Major Trials

The randomized trial evidence on vitamin D supplementation is the single most important corrective to the enthusiasm of the early 2000s. Several large, well-powered trials have now reported, and their findings are sobering for both sides.

The mainstream interpretation: at typical supplementation doses (1,000 to 4,000 IU/day), in already-replete or near-replete populations, D3 produces neither dramatic benefit nor measurable harm on the major hard endpoints. The Robbins-aligned interpretation: the trials do not refute the harms his framework predicts, because they do not measure ceruloplasmin, magnesium status, retinol, or the mineral-balance markers his hypothesis turns on. Both readings are defensible. What is not defensible is the pre-VITAL claim that broad D3 supplementation would meaningfully reduce cancer or cardiovascular mortality — that hypothesis has been tested and has failed.

8. Where Robbins’s Position Has Some Mainstream Support

It is fair to identify the parts of Robbins’s teaching that are not heterodox at all, but rather mainstream physiology that has not yet filtered into routine clinical practice.

So “Robbins is wrong about vitamin D” is too strong. The more accurate framing is that the RCP draws the inferences harder, faster, and more universally than the mainstream literature does, but the underlying biochemical claims are substantially evidence-based.

9. A Balanced Practical Take

Stripping the rhetoric off both sides, a careful reader can arrive at a defensible position that does not require choosing a tribe.

  1. High-dose D3 (above 5,000 IU/day) without magnesium and retinol cofactors is poorly justified outside specific medical indications. The trial evidence does not support routine megadosing for prevention, and the biochemistry argues that magnesium will be the rate-limiting step regardless.
  2. Sunlight, when geographically and seasonally feasible, is superior to a pill. The signaling cascade triggered by UVB exposure is not reproducible orally, and the body’s own feedback prevents toxicity.
  3. Real fermented cod-liver oil provides D in a food matrix with retinol and DHA — the form Robbins prefers. A teaspoon supplies modest D plus retinol and omega-3, in roughly the proportions the body would obtain from traditional Northern-European diets.
  4. Confirmed clinical deficiency (25-OH-D below 20 ng/mL with rachitic or osteomalacia symptoms) is a clinical emergency and should be treated — with D3 if needed. The Robbins position applies most strongly to “preventive” or “wellness” megadosing in asymptomatic adults, not to true deficiency disease.
  5. If you do supplement D3, take magnesium glycinate or malate alongside it. The dose of magnesium should be 200 to 400 mg of elemental magnesium per day, taken with food, to support the activation enzymes and offset the magnesium burn rate.
  6. Reassess at six months. Retest 25-OH-D, ceruloplasmin, serum copper, magnesium RBC, and ferritin. Adjust based on the results, not on a rigid intake target.

10. How to Monitor Status If You Stop or Reduce D3

If a reader decides to reduce or discontinue D3 supplementation, the practical question is: how do I know it is going well? Retest 25-OH-D in three to six months — long enough for the storage form to equilibrate to its new steady state. The interpretation of the resulting number depends on which framework you are using.

Robbins’s informal target is 50 to 65 ng/mL, but achieved through sunlight, fermented cod-liver oil, and adequate magnesium and retinol — not through a daily 5,000 IU capsule. He notes that healthy outdoor workers in Hawaii average around 30 to 40 ng/mL and feel fine, suggesting that the “optimal” band may itself be set higher than the underlying biology requires.

Alongside 25-OH-D, the RCP-aligned reader will also retest ceruloplasmin, serum copper, RBC magnesium, ferritin, and transferrin saturation. If reducing D3 produces an improvement in ceruloplasmin and a fall in ferritin (the hallmark RCP markers), Robbins would interpret that as evidence the prior D3 dose was suppressing copper-iron metabolism. If 25-OH-D drops below 20 ng/mL with onset of bone pain or fatigue, that is a signal to reintroduce D — ideally as fermented cod-liver oil or, if needed, as moderate-dose D3 with magnesium and K2 alongside.

Key Research Papers

Connections

Back to top