Supplements

Vitamin D3 (the category, not one brand) — and the K2, magnesium and dose question nobody on Instagram gets right

The most over-supplemented and most under-tested nutrient in the wellness aisle — get the blood test before you guess the dose

Vitamin D is the only supplement category where the smartest move is to *measure first, then dose* — and almost nobody does. The 2020–2026 literature has thoroughly debunked the COVID-era influencer claim that 'more is always better' (the VITAL trial, n=25,871, 5-year follow-up, no cardiovascular or cancer mortality benefit at 2,000 IU/day in unselected adults — Manson 2019/2022; the D-Health trial, n=21,315, no all-cause mortality benefit at 60,000 IU/month — Neale 2022; the DO-HEALTH trial, no benefit on falls, fractures or cognition at 2,000 IU/day in healthy 70+ Europeans — Bischoff-Ferrari 2020). But the same literature is *unambiguous* that correcting genuine deficiency (serum 25-OH-D <30 nmol/L / <12 ng/mL) reduces fracture risk, supports immune competence, and is one of the easiest, cheapest endocrine corrections in medicine. The two camps — 'vitamin D fixes everything, take 10,000 IU' vs 'vitamin D does nothing, the trials are null' — are both wrong, and both are derived from misreading the same trials. The honest answer is dose-response: deficient people benefit, replete people don't, and the only way to know which you are is a €25 25-OH-D blood test. The TikTok-driven 'D3 + K2 + magnesium stack' is, for once, broadly defensible — but the specific ratios sold on Instagram are mostly invented.

Vitamin D3 (the category, not one brand) — and the K2, magnesium and dose question nobody on Instagram gets right

What it claims

  • Bone health, fracture prevention, calcium absorption
  • Immune support, respiratory infection prevention, 'autoimmune regulation'
  • Mood support, seasonal affective disorder, depression
  • Cardiovascular, cancer, all-cause mortality (the big COVID-era claims)
  • Hormone support, fertility, testosterone, PCOS

What the label is not telling you

  • You should test before you dose. 25-OH-D, finger-prick or venous, €15–40, available from any private lab and increasingly on the NHS / public-system request list. Sufficiency = 75–125 nmol/L (30–50 ng/mL). Deficiency = <30 nmol/L (<12 ng/mL). Insufficiency = 30–75 nmol/L (12–30 ng/mL). Above 125 nmol/L there is no documented additional benefit and rising risk of hypercalcaemia, kidney stones, and (paradoxically) increased falls in some 2010s trials at very high bolus doses (Sanders 2010; Bischoff-Ferrari 2016 — annual mega-doses increased fall risk). Test once at end-of-winter (Feb–March in the Northern Hemisphere — your lowest point), dose accordingly, retest after 8–12 weeks.
  • The 'mega-dose 10,000 IU/day forever' protocol popularised by certain podcast doctors in 2020–2023 is not supported by trial evidence and carries documented harm signal at the upper end. The Endocrine Society upper tolerable limit is 4,000 IU/day for most adults; clinical correction protocols (50,000 IU/week for 8 weeks under supervision) are time-limited, not lifestyle.
  • D3 (cholecalciferol) > D2 (ergocalciferol) for raising blood levels. D3 is roughly 1.5–2× more effective per IU (Tripkovic 2012, meta-analysis). Most quality supplements are now D3-only; D2 is mostly seen in prescription mega-doses and some vegan formulations. Vegan D3 from lichen (Sports Research, Garden of Life, BetterYou Vegan) is now widely available and works identically to lanolin-sourced D3.
  • The K2 (MK-7) co-supplementation argument is biologically plausible but trial evidence is softer than Instagram claims. Theory: vitamin D increases calcium absorption; K2 helps direct that calcium to bone rather than soft tissue (arteries, kidneys). Some RCTs (Knapen 2013; Brandenburg 2017) support this for vascular calcification. The 'D3 without K2 will calcify your arteries' panic is overstated — the trial signal is real but small, and primarily relevant at high D3 doses (>4,000 IU/day) and in populations with cardiovascular risk. At physiological replacement doses (1,000–2,000 IU/day) the added urgency is modest.
  • Magnesium is genuinely required for vitamin D metabolism — every enzymatic step from 25-OH-D to active 1,25-OH-D requires magnesium as a cofactor. Low magnesium status can blunt the response to supplementation. If your D3 dose isn't moving your blood levels, magnesium deficiency is the most common reason (Uwitonze & Razzaque 2018). The fix is magnesium glycinate at 200–400 mg/day, not more D3.
  • Sun is still the most efficient source for most of the year at most latitudes. 15–30 minutes of arms-and-legs midday sun in summer produces 10,000–20,000 IU of endogenously regulated D3 with no toxicity ceiling. Above ~40° latitude (north of Madrid, north of New York) from October to March, dermal D3 synthesis is functionally zero regardless of how much you stand outside — this is the supplementation window that actually matters.
  • Dose by body weight, not by a flat number. Heath & Mughal 2017 and Ekwaru 2014 — the same 2,000 IU/day raises serum D2.5× more in a 55 kg person than in a 110 kg person. The 'one-size 2,000 IU' default underdoses larger adults and overdoses smaller ones. Reasonable rule of thumb: 25–50 IU/kg/day in winter for adults with confirmed insufficiency, retest at 12 weeks.
  • Spray, oil-based softgel, and liposomal formats meaningfully outperform dry tablets and gummies for absorption. Vitamin D is fat-soluble — take any oral form with the fattiest meal of your day (avocado, eggs, olive oil) and absorption improves 30–50% (Dawson-Hughes 2013). BetterYou's oral sprays, Thorne and Pure Encapsulations softgels in MCT or olive oil, and Carlson liquid drops are functionally the best-absorbed mainstream formats.
  • Gummies are again the worst value SKU. Sugar-loaded, frequently under-dosed against label (2022 ConsumerLab assay found 11 of 28 D3 gummies under-delivered by 10–60%), and the gummy matrix interferes with the fat-soluble absorption advantage.
  • The 'D3 + K2 combo tablet' on Amazon is usually a low-quality compromise. The MK-7 form of K2 is unstable next to certain mineral co-ingredients and degrades in storage. If you're going to combine, separate D3 (softgel in oil) and K2 (separate softgel, MK-7 from natto or fermented source) is the more reliable route.
  • Thorne / Pure Encapsulations / Nordic Naturals / Carlson / Sports Research (€15–35) — Third-party tested, oil-based softgels or liquid drops, accurate label claims, the workhorses. Thorne's D/K2 combo and Pure Encapsulations D3 liquid are the cleanest mainstream options.
  • BetterYou DLux oral spray (€10–18) — Buccal absorption, useful for anyone with gut absorption issues (IBD, post-bariatric, coeliac), and well-tested in independent assays. The format genuinely works.
  • NOW Foods / Solgar / Garden of Life (€10–20) — Decent quality, generally accurate dosing, mainstream availability. Garden of Life Vegan D3 is one of the better lichen-sourced options.
  • Supermarket house-brand D3 tablets (€3–8) — The cheapest tier. Dose accuracy varies; absorption from dry tablet matrix is poorer than oil-based. Fine for correcting frank deficiency on a tight budget; not the optimal vehicle.
  • 'Mega-dose' 50,000 IU weekly OTC SKUs (€20–40) — These are prescription-equivalent doses sold OTC in the US, EU and UK. They have a place in supervised 8-week correction protocols for severe deficiency. They have no place in a lifestyle stack. The single biggest source of preventable D toxicity in the 2020–2025 case-report literature.

Effect on the nervous system

Vitamin D's nervous-system relevance is mostly indirect but consistent. The vitamin D receptor (VDR) is expressed throughout the central nervous system — hippocampus, prefrontal cortex, hypothalamus — and 1,25-OH-D modulates synthesis of serotonin, dopamine and several neurotrophic factors. The clinical signal is clearest in three places. First, mood: meta-analyses (Spedding 2014; updated 2022) find a small but consistent antidepressant effect in supplementation trials limited to participants who started deficient — i.e. correcting deficiency helps mood, supplementing replete people does nothing. Second, seasonal affective patterns: in populations above 40° latitude, the autumn–spring drop in serum D parallels (but does not necessarily cause) the SAD pattern; correcting it is a reasonable, cheap, low-risk move alongside light exposure. Third, sleep: a 2018 systematic review (Gao) found D supplementation modestly improved sleep quality in adults with low baseline D, particularly when combined with magnesium (the cofactor enabling D activation). The harder-to-quantify nervous-system effect is the one our coaching cohort reports most consistently: people whose winter mood, energy, immune resilience and recovery all sag in parallel from November to March — the 'I just hibernate from January' pattern — and whose 25-OH-D tests come back in the 25–50 nmol/L range. Correcting these clients to 75–100 nmol/L over 8–12 weeks, alongside light exposure and magnesium glycinate, produces a reliable subjective shift. The mechanism is probably endocrine-metabolic, not directly neural. The intervention is one of the highest-leverage in the winter stack — but only for the people who are actually deficient. Replete people get nothing from another softgel.

Who it might suit

Adults — all genders — living above ~40° latitude from October to March (most of Northern Europe, Canada, the northern US, anywhere with a real winter); people with darker skin pigmentation at higher latitudes (Fitzpatrick IV–VI, who need 3–5× longer sun exposure to synthesise the same D); people who work indoors during all daylight hours; older adults (skin D3 synthesis drops ~75% from age 20 to 70); pregnant and breastfeeding women (with clinician supervision); anyone with diagnosed osteopenia, osteoporosis, or repeated fractures; anyone with IBD, coeliac, post-bariatric surgery (use the spray or liquid format); anyone with confirmed 25-OH-D <50 nmol/L on testing. Dose to body weight, take with a fatty meal, pair with magnesium glycinate, retest at 12 weeks.

Who should skip it

Anyone with sarcoidosis, primary hyperparathyroidism, certain lymphomas, or other granulomatous diseases — these conditions cause endogenous over-conversion of D and supplementation can precipitate hypercalcaemia. Anyone with a history of kidney stones — discuss with your nephrologist before supplementing above 1,000 IU/day. Anyone whose 25-OH-D already sits above 125 nmol/L — there is no additional benefit and a documented harm signal at higher levels. Anyone supplementing at 10,000+ IU/day without blood-test supervision — you are running an unmonitored mega-dose protocol with no upside over a measured dose. Anyone buying a 'D3 + K2 + magnesium + zinc + boron + boron + boron' kitchen-sink tablet because Instagram said so — you are paying for under-dosed everything; separate products dosed to your actual blood work outperform combo tablets in every category.

Bottom line

The honest hierarchy: (1) Test first — €15–40 25-OH-D blood test, ideally at end-of-winter, before guessing a dose. (2) If insufficient (<75 nmol/L) and above 40° latitude in winter, supplement plain D3 in an oil-based softgel or oral spray, dosed roughly 25–50 IU/kg/day, with the fattiest meal of the day, paired with 200–400 mg magnesium glycinate in the evening, and retest at 12 weeks. (3) Consider K2 (MK-7, 90–180 µg/day) separately if you're dosing D3 above 2,000 IU/day long-term or have cardiovascular risk — useful, not urgent. (4) Ignore the 10,000 IU/day 'just because' protocol — it's not supported by the trial evidence and accumulates real harm signal at the top end. (5) Skip gummies and combo tablets. Brands worth the shelf space: Thorne, Pure Encapsulations, Nordic Naturals, Carlson, Sports Research, BetterYou DLux spray, Garden of Life Vegan D3. The vetted vitamin D options we keep on our shelf — with the dosing protocol matched to your latitude and blood work — live at thecodex.world.