Vitamin D3 + K2
TL;DR
- Do not supplement without a baseline 25(OH)D test first — Phuket sun exposure likely maintains sufficiency
- D3+K2’s most evidence-solid benefit is bone health via osteocalcin carboxylation (9-RCT meta-analysis, n=2,570; PMID 41268154)
- K2 MK-7 improves arterial stiffness in healthy postmenopausal women (PMID 25694037, n=244, 3-year RCT) but NOT in CKD (PMID 32817311, null)
- D3 does NOT lower blood pressure (46-trial meta-analysis, PMID 25775274, null), does NOT improve sleep (PMID 34881361, n=189 RCT null), does NOT improve HRV (PMID 28361788, BEST-D null)
- VITAL RCT (PMID 30415629) is the anchor for all cardiovascular nulls: 2,000 IU/day D3, n=25,871, ~5 years — no CVD or cancer prevention
- The “44% sleep onset latency improvement” claim circulating in trade press has no verifiable peer-reviewed primary source — treat as folklore
- Co-supplementation with K2 is pharmacologically justified when D3 >2,000 IU/day (D3 upregulates VKDPs without K2 → more inactive ucOC and MGP)
Why It Matters for Vitals
Ben: 65 kg, Phuket (~7°N), daily training, 3 days post wrist injury, Retatrutide week 4 (0.5 mg), GHK-Cu 1 mg/day, Creatine starting.
| Factor | Relevance to D3+K2 |
|---|---|
| Phuket sun exposure | Chronic UVB at 7°N likely sustains 25(OH)D >30 ng/mL without supplementation. Baseline 25(OH)D is mandatory before supplementing. Blind dosing risks hypervitaminosis D and hypercalcemia. |
| 65 kg bodyweight | Standard adult dosing (2,000–4,000 IU/day) is body-weight-appropriate |
| Retatrutide (week 4, 0.5 mg) | GLP-1 agonist reduces appetite → potential reduced dietary fat intake → may mildly impair fat-soluble D3 absorption. Take D3+K2 with the largest meal of the day. |
| GHK-Cu 1 mg/day | No known PK or PD interaction with D3/K2; independent pathways |
| Creatine (starting) | No interaction with D3/K2; D3+K2 bone support is relevant for contact sport context but no acute performance benefit |
| Wrist injury (3 days post) | D3 role in bone healing is theoretically supportive but not clinically significant acutely; not a standalone justification for high-dose D3 |
| Daily training | Athletic context does not independently justify D3 above IOM thresholds absent confirmed deficiency |
Mechanism Summary
D3 (cholecalciferol)
D3 requires two-step activation:
- Skin/liver → 25-hydroxyvitamin D [25(OH)D] via 25-hydroxylase. 25(OH)D is the major circulating metabolite and the clinical status marker. Half-life: ~15 days.
- Kidney → 1,25-dihydroxyvitamin D [calcitriol, 1,25(OH)₂D] via 1α-hydroxylase (CYP27B1). Calcitriol is the active hormone.
D3 upregulates synthesis of vitamin K-dependent proteins (VKDPs): osteocalcin (bone) and matrix Gla protein (MGP) (vascular). Without adequate K2, these proteins remain undercarboxylated (inactive), potentially diverting calcium away from bone and toward soft tissue.
Key PK: 1 IU = 0.025 mcg cholecalciferol; 1 mcg = 40 IU. 25(OH)D steady state: ~5 half-lives × 15 days ≈ 75 days (~2.5 months) on daily dosing.
K2 (MK-7, menaquinone-7)
MK-7 serves as an essential cofactor for γ-glutamyl carboxylase, which carboxylates specific glutamate residues in VKDPs to their active form. It carboxylates osteocalcin (promotes bone mineralization) and MGP (inhibits vascular calcification).
Key PK (MK-7): Half-life ~68–72 hours; Tmax 2–6 hours; detected in serum up to 48+ hours after a single dose. MK-7 redistributes in LDL-C/VLDL-C particles, extending extrahepatic tissue exposure.
Critical distinction: MK-7 is the only K homolog that meaningfully supports extrahepatic VKDP carboxylation at nutritional doses (~60 µg/day). MK-4 is rapidly cleared (not detectable after a single oral dose). All positive vascular/bone RCTs used MK-7.
Evidence Grades
✅ Confirmed
| Claim | Evidence | Population |
|---|---|---|
| K2 MK-7 (180 µg/day × 3 years) improves arterial stiffness and vascular elasticity; dp-ucMGP decreased 50% vs placebo | RCT, n=244, 3-year; cfPWV improved; PMID 25694037 | Healthy postmenopausal women |
| D3 + K2 improves bone turnover markers; reduces undercarboxylated osteocalcin (ucOC) | Meta-analysis 9 RCTs, n=2,570, median 12 months; PMID 41268154 | Postmenopausal women / mixed adults |
| High-dose D3 WITHOUT K2 increases circulating ucOC and MGP | Meta-analysis/synthesis; J Nutr Sci Vitaminology 2024 | Mixed adult populations |
| Combined exercise + D3 + calcium improves femoral neck BMD vs exercise alone | Meta-analysis RCTs; DOI 10.3390/nu17243866 | Postmenopausal women |
| MK-7 pharmacologically superior to MK-4 for extrahepatic VKDP carboxylation at nutritional doses | Head-to-head PK study; PMID 3502319, PMID 7230802 | Healthy adults |
| Co-supplementation justified when D3 >2,000 IU/day | Mechanistic + ucOC/MGP RCT evidence | Adults on high-dose D3 |
| MK-7 2–3×/week is PK-supported | Half-life ~68–72 h; steady state within ~7–14 days | Healthy adults |
| D2d trial (4,000 IU/day × 3 years): safe in overweight/obese adults, no hypercalcemia/nephrolithiasis increase | RCT, n=2,423; PMID 9352576 | Overweight/obese adults with prediabetes |
| MK-7 safe at 360 µg/day × 3 months | Human trial; PMID 8202544 | Healthy adults |
❌ Confirmed Null
| Claim | Evidence | Source |
|---|---|---|
| D3 2,000 IU/day does NOT prevent CVD or cancer in generally healthy adults | VITAL RCT, n=25,871, ~5 years: HR 0.96 (95% CI 0.85–1.09) for major CVD | PMID 30415629 |
| D3 does NOT lower blood pressure | Individual patient data meta-analysis, 46 trials, n=4,541 | PMID 25775274 |
| D3 does NOT improve sleep quality, duration, or insomnia severity | RCT, n=189, 20,000 IU/week × 4 months: no PSQI, sleep duration, or insomnia severity improvement | PMID 34881361 |
| K2 MK-7 does NOT improve arterial stiffness in CKD patients | K4Kidneys RCT, n=159, CKD stages 3b–4, MK-7 400 µg/day × 12 months: null for cfPWV, augmentation index | PMID 32817311 |
| D3 does NOT improve HRV or lower resting heart rate | BEST-D trial, n=305, 800–2,000 IU/day × 12 months | PMID 28361788 |
| D3 alone does NOT prevent fractures; requires calcium co-administration | VITAL + prior meta-analyses | PMID 30415629 |
| D-Health trial (60,000 IU/month D3, n=21,315): null for primary MACE endpoint | RCT; PMID 37380191 | PMID 37380191 |
| MK-4 is not detectable in serum after single oral dose; does NOT accumulate with repeated dosing at nutritional levels | Head-to-head PK study, healthy women | PMID 3502319 |
⚠️ VITAL (PMID 30415629) is the primary anchor for all cardiovascular null claims. No subsequent trial has overturned this result.
⚠️ K2 vascular benefit is population-specific. Confirmed in healthy postmenopausal women (PMID 25694037) but NULL in CKD patients (PMID 32817311).
⚠️ Insufficient / Contested
| Claim | Issue |
|---|---|
| ”44% sleep onset latency improvement” from D3 | No verifiable PubMed-indexed primary source. Trade-press folklore. Do not cite as fact. |
| D3 → HRV improvement in healthy replete adults | Small n signals in disease populations; BEST-D null in healthy adults |
| D3 → athletic performance in 25(OH)D-replete athletes | No credible evidence |
| K2 → reversal of existing vascular plaque | May slow calcification progression; does NOT reverse established CVD |
| D3 → acute bone healing acceleration in wrist injury | Theoretical plausibility; no acute human RCT |
| D3 → mood/mental health in non-deficient adults | Mechanistic hypothesis; weak-to-null RCT evidence |
| K2 → longevity / all-cause mortality | Observational Asian associations not confirmed in Western RCTs |
| D3 + K2 and BP | Contested — one meta-analysis of 30 trials found modest diastolic reduction only; 46-trial meta-analysis found no effect |
Wearable / VitalsSync Implications
D3+K2 supplementation — if deficiency is confirmed — may influence wearable signals as follows:
| Signal | Direction | Evidence | Wearable Confidence |
|---|---|---|---|
| Bone turnover markers (CTX, osteocalcin) | ↓ ucOC, ↑ carboxylated osteocalcin | 9-RCT meta-analysis; PMID 41268154 | Lab measure only; not wearable |
| Arterial stiffness (cfPWV) | ↓ in healthy postmenopausal women on MK-7; NULL in CKD | PMID 25694037 vs. PMID 32817311 | Not Apple Watch-accessible directly |
| 25(OH)D levels | ↑ if deficient and supplementing | Well-established PK | Lab test required |
| HRV / resting heart rate | No change expected in healthy adults | BEST-D null; PMID 28361788 | Do not expect improvement |
| Sleep quality metrics | No change expected in deficient adults | PMID 34881361 null | Do not expect improvement |
| Blood pressure | No change expected | 46-trial meta-analysis null; PMID 25775274 | Do not expect improvement |
Practical wearable implication: If Ben is deficient in 25(OH)D and corrects that deficiency, there may be indirect benefits to subjective energy and recovery. No evidence that D3 supplementation in a replete person will improve any wearable-tracked metric.
Protocol
Blood Work / Baseline
Mandatory before supplementing:
| Test | Target | Why |
|---|---|---|
| Serum 25(OH)D | Confirm deficiency/insufficiency before supplementing; retest at ~3 months if supplementing | 25(OH)D is the functional status marker; Phuket sun may already produce sufficiency |
| Calcium (corrected), eGFR, PTH | Baseline kidney and calcium status | Safety monitoring if supplementing at 4,000 IU/day |
Dose Decision Logic
IF 25(OH)D < 20 ng/mL:
→ D3 2,000–4,000 IU/day + MK-7 100–180 µg/day; retest at 3 months
IF 25(OH)D 20–29 ng/mL:
→ D3 1,000–2,000 IU/day + MK-7 100–180 µg/day; retest at 3 months
IF 25(OH)D ≥ 30 ng/mL:
→ No D3 supplementation indicated; reassess every 6 months
IF CKD, sarcoidosis, on warfarin:
→ Consult physician before supplementing K2
For Ben: Given Phuket sun exposure, 25(OH)D ≥30 ng/mL is plausible. Do not supplement without the baseline test. If supplementing, stay ≤4,000 IU/day (pragmatic upper boundary per D2d trial, PMID 9352576).
Timing
- D3 + K2 together — always co-supplement; D3 upregulates VKDPs and increases K2 demand
- With largest meal — standard practice; retatrutide may reduce fat intake; compensate by taking with the largest available meal
- Consistency — daily or 2–3×/week both acceptable for D3; MK-7 2–3×/week is PK-supported given 68–72 h half-life
- Steady state — do not expect 25(OH)D changes before ~75 days of consistent dosing
Product Quality
| Criterion | Recommendation |
|---|---|
| K2 form | MK-7 only (menaquinone-7). MK-4 is rapidly cleared and not effective at nutritional doses. Verify label. |
| K2 source | Natto fermentation; look for “MenaQ7” or equivalent standardized extract |
| D3 form | Cholecalciferol (D3); no advantage to D2 in supplementation |
| MK-7 dose | 100–180 µg/day (used in positive RCTs); 360 µg/day used safely for 3 months |
| D3 dose verification | Trusted brands with third-party testing (NSF, USP, ConsumerLab) |
Safety Profile
Vitamin D3
| Threshold | Value | Source |
|---|---|---|
| IOM UL (adults) | 4,000 IU/day | IOM/FNB |
| Serum 25(OH)D toxicity threshold | >60 ng/mL (>150 nmol/L) associated with hypercalcemia risk | IOM 2011 |
| Clinical toxicity threshold | >25,000 IU/day sustained; 25(OH)D ~500 nmol/L | PMC10195747 |
| Hypercalciuria risk | Increases at >800 IU/day (RR 1.64 vs placebo); 37-RCT meta-analysis | PMC7897606 |
Dose-dependent hypercalcemia rates (Calgary RCT, 3-year, healthy adults):
| Dose | Hypercalcemia | Hypercalciuria |
|---|---|---|
| 400 IU/day | 0% | 17% |
| 4,000 IU/day | 3% | 22% |
| 10,000 IU/day | 9% | 31% |
Vitamin K2 (MK-7)
| Parameter | Value | Source |
|---|---|---|
| CRN Highest Observed Intake (HOI) | 375 µg/day | CRN safety analysis |
| MK-7 NOAEL (rat) | 10 mg/kg/day | PMC3172146 |
| Acute toxicity | No deaths at 5,000 mg/kg | PMC3172146 |
| Human trial (360 µg/day × 3 months) | No GI, muscle, or other AEs | PMC8202544 |
No established UL for MK-7 (no adverse effects reported from high food/supplement intakes in humans).
Key Contraindications / Interactions
| Drug/Condition | Effect | Severity |
|---|---|---|
| Warfarin / VKA anticoagulants | K2 antagonizes anticoagulant effect; requires INR monitoring | High — contraindication |
| Granulomatous disease (sarcoidosis, TB) | Ectopic CYP27B1 in macrophages → excess 1,25(OH)₂D → hypercalcemia | High — relative contraindication |
| CKD (eGFR <30) | Impaired 1,25(OH)₂D degradation; elevated hypercalcemia risk | High |
| Thiazide diuretics | Reduce renal calcium excretion → hypercalcemia risk with D3 | Moderate |
| CYP24A1 mutations | Found in ~10–15% of calcium stone formers; vulnerability to D-induced hypercalcemia | Moderate |
| Chronic corticosteroids | Reduce intestinal calcium absorption; antagonize D-mediated calcium homeostasis | Moderate |
| Anticonvulsants (phenytoin, phenobarbital) | Induce CYP450 → increased D catabolism → potential deficiency | Moderate |
| Orlistat | May reduce fat-soluble D absorption; monitoring advised | Mild |
Safety Notes for Ben
- Ben is not on warfarin, does not have CKD, sarcoidosis, or documented CYP24A1 mutations — no high-risk contraindications present.
- Phuket sun exposure means risk of accumulating 25(OH)D above 60 ng/mL on 4,000 IU/day is real. Baseline 25(OH)D first.
- Retatrutide-related reduced food intake may mildly impair D3 absorption; take D3+K2 with the largest available meal.
Evidence Summary — Additional RCTs (Batch 37 source update)
The following trials were identified in the canonical k2-menaquinone.md source (2026-04-18) and incorporated into this note:
| Trial | Population | Result | Key PMIDs |
|---|---|---|---|
| Trevasc-HDK (Kidney Int Rep 2023) | 178 hemodialysis patients, MK-7 360 µg 3×/week × 18 months | CAC relative mean difference 0.85 (95% CI 0.55–1.31); null; MACE HR=0.98 | PMID 37705910 |
| Valkyrie (JASN 2020) | 132 hemodialysis patients, MK-7 + D3 × 18 months | No effect on coronary/aortic/valve calcification, death, stroke, CV events — null | PMID 31704740 |
| RenaKvit (Clin Kidney J 2021) | 48 dialysis patients, 2 years | Improved dp-ucMGP but no effect on CAC, AAC, or cfPWV — null | PMID 34476095 |
| AVADEC (Circulation 2022) | 365 elderly men with AVC score >300, MK-7 720 µg/day + D3 25 µg × 24 months | AVC increased 275 AU vs 292 AU placebo (P=0.64); all secondary endpoints null — null | PMID 35465686 |
| Dahl et al. JACC Adv 2023 | Mixed population | No significant reduction in mean CAC progression; hypothesis-generating signal only in CAC≥400 subgroup | PMID 38938724 |
Pattern across all RCTs: dp-ucMGP improves consistently (confirming biological activity), yet vascular calcification and clinical events do not change. This is a classic surrogate endpoint fallacy — improving a biomarker does not reliably translate to clinical benefit.
MK-4 at pharmacological doses (45 mg/day)
Japanese osteoporosis trials used MK-4 at 45 mg/day (45,000 µg — 50–500× typical supplemental doses). The primary meta-analysis supporting fracture reduction claims (Iwamoto & Sato, Expert Opin Pharmacother 2013) was formally retracted (PMID 23346882). This dose is irrelevant to typical supplement use.
Claims Registry
| Claim | Grade | Primary Source | Notes |
|---|---|---|---|
| K2 MK-7 (180 µg/day × 3 yr) improves arterial stiffness in healthy postmenopausal women | Confirmed (population-specific) | PMID 25694037 | Null in CKD/HF patients |
| D3 + K2 reduces ucOC, improves osteocalcin carboxylation | Confirmed | PMID 41268154 | 9 RCTs, n=2,570 |
| High-dose D3 without K2 increases ucOC and MGP | Confirmed | J Nutr Sci Vitaminology 2024 | Pharmacological rationale for co-supplementation |
| D3 + exercise improves femoral neck BMD in postmenopausal women | Confirmed | DOI 10.3390/nu17243866 | Exercise + D3 + Ca synergy |
| MK-7 pharmacologically superior to MK-4 at nutritional doses | Confirmed | PMID 3502319, PMID 7230802 | Key formulation distinction |
| AVADEC, Trevasc-HDK, Valkyrie, K4Kidneys, RenaKvit: MK-7 does NOT reduce vascular calcification or CV events | Confirmed Null | PMID 35465686, 37705910, 31704740, 32817311, 34476095 | Consistent null across 5 RCTs |
| Dahl 2023: no significant CAC reduction; only hypothesis-generating signal in CAC≥400 subgroup | Null with post-hoc signal | PMID 38938724 | Requires prospective confirmation |
| MK-4 45 mg/day fracture reduction meta-analysis | Retracted | PMID 23346882 (retraction) | Irrelevant to typical supplemental doses |
| D3 2,000 IU/day does NOT prevent CVD (VITAL) | Confirmed Null | PMID 30415629 | PRIMARY anchor |
| D3 does NOT lower blood pressure | Confirmed Null | PMID 25775274 | 46-trial meta-analysis; Cochrane 2015 |
| D3 does NOT improve sleep quality/duration/insomnia | Confirmed Null | PMID 34881361 | n=189 RCT |
| K2 MK-7 does NOT improve arterial stiffness in CKD | Confirmed Null | PMID 32817311 | K4Kidneys RCT |
| D3 does NOT improve HRV or lower resting HR | Confirmed Null | PMID 28361788 | BEST-D trial |
| D3 does NOT prevent fractures alone; requires calcium co-administration | Confirmed Null | VITAL + meta-analyses | |
| D3 does NOT prevent cancer incidence | Confirmed Null | PMID 30415629, PMID 30809758 | VITAL + ViDA |
| MK-4 does not accumulate with repeated dosing at nutritional levels | Confirmed Null | PMID 3502319 | |
| ”44% sleep onset latency improvement” | Insufficient | No verifiable primary source | Trade-press folklore only |
| D3 → HRV improvement in healthy adults | Insufficient | BEST-D null | Observational/small-n RCT signals insufficient |
| D3 → athletic performance in replete athletes | Insufficient | No credible RCT | |
| K2 → reversal of existing vascular plaque | Insufficient | No RCT evidence | May slow progression only |
| D3 → acute bone healing acceleration | Insufficient | No acute human RCT | Theoretical plausibility only |
| K2 → all-cause mortality / longevity | Insufficient | Observational Asian data, no Western RCT confirmation | |
| MK-7 2–3×/week is PK-supported | Probable | PMID 3502319, PMID 9237441 | Half-life-based; not endpoint-tested |
| Co-supplementation justified at D3 >2,000 IU/day | Probable | Mechanistic + ucOC/MGP data | |
| MK-7 safe at 360 µg/day × 3 months | Confirmed | PMID 8202544 | Human trial |
| MK-7 HOI 375 µg/day (CRN) | Probable | CRN safety analysis | No established UL |
Related Notes
- Retatrutide — primary stack partner; take D3/K2 with largest meal
- GHK-Cu — tissue repair stack partner; no known interaction
- Creatine — ergogenic stack partner; bone support context for contact sport
- HRV — wearable signal context; no D3-attributable HRV signal expected in replete adults
- MK-7 vs MK-4 — PK distinction; MK-7 only, MK-4 ineffective at nutritional doses
Built from batch-17 Phase 1 (W1 Clinical, W2 PK, W3 Safety, W4 Skeptic) | Canonical: vitamin-d3-k2-canonical.md | Last updated: 2026-03-31