DAY 9

Health & Longevity: Supplement Science
What Works, What's Hype, How to Decide

2026-05-31 · BigCat's Vitality Protocol
Evidence base: ISSN position stands, Cochrane reviews, large RCTs (VITAL/ASCEND). Three evidence tiers: human hard endpoints, mechanism, animal models.
SUB · Evidence tiers / decision framework
The Supplement Pyramid: Food First, Then Test, Then Supplement
Food → Test → Supplement
One-line takeaway
A supplement's value scales with how deficient you are. Big gains in the deficient zone, near-zero in the sufficient zone, net harm in the excess zone. Test, then dose by deficit — don't shop by podcast.
Evidence level
RCT: VITAL (Manson 2019, NEJM, n=25,871, 5.3-yr follow-up) found that routine vitamin D (2000 IU/d) and Omega-3 (1 g/d) did not reduce the composite cancer/CV endpoint — but prespecified subgroups with baseline 25(OH)D <20 ng/mL and low fish intake did benefit. Lesson: "no effect on average" ≠ "no effect on you." Stratification is everything.
Visualization: dose-response curve
Supplement dose-benefit curve (typical U-shape) Deficient Sufficient (optimal window) Excess Big gain ≈ 0 Net harm Benefit ↑ Toxicity ↑
Fat-soluble vitamins (A, D, E), iron, selenium, zinc all show a U-shaped dose-response. "More is better" is a linear-thinking mistake.
Actionable protocol
Three-tier decision:
① Food tier: 1.6–2.2 g/kg protein, 30 g fiber, 2 servings fatty fish/week — most "deficiencies" are actually diet gaps.
② Test tier: annual 25(OH)D, ferritin, B12, Omega-3 Index, magnesium (RBC Mg > serum), HbA1c — target above the "normal" floor.
③ Supplement tier: only for measured deficits or hard-endpoint evidence. For each addition ask: hard-endpoint RCT? Am I deficient? Is the ROI worth it?
Women's note + common myths
Premenopausal women have high iron-deficiency rates (menstrual loss). Pre-conception and pregnancy folate 400–800 μg/d is one of the few RCT-backed "musts" (neural tube defect prevention). Do not blindly supplement iron in perimenopause — postmenopausal iron needs drop sharply, and excess iron is linked to CV and neurodegenerative risk (Mascitelli 2014).
Myth 1: "Natural = safe" — herbal–heavy metal and hepatotoxicity events appear in USP MedWatch every year.
Myth 2: "Stack more, win more" — fat-soluble vitamins and trace minerals antagonize each other (calcium blocks iron, excess zinc depletes copper).
Myth 3: "US/Japan brands are purer" — FDA does not pre-approve supplements. Look for USP / NSF / Informed Sport third-party certification.
This week + reflection
THIS WEEK
Lay out every supplement bottle in your home and photograph it. For each ask: do I have personalized evidence for this (a deficiency test or a hard-endpoint RCT)? Retire the ones you bought on a hunch.

Reflection: if the supplement industry had to pass Phase III RCTs like drugs, how many bottles would still be on the shelf?
SUB · Well-evidenced / muscle & cognition
Creatine: 5 g/day, the Most Under-Used Evidence-Backed Supplement
Creatine Monohydrate
One-line takeaway
Creatine monohydrate has the strongest evidence base of any sports supplement after protein: 3–5 g/day, +5–10% strength, +1–2 kg lean mass, and a growing dataset on cognition and mood. 30-year clean safety record.
Evidence level
Meta: Kreider 2017 (ISSN position) synthesizing 500+ studies — creatine has the largest, most reproducible effect size on high-intensity performance. Lanhers 2017 (Sports Med meta): creatine + resistance > resistance alone (+1.4 kg lean mass). Cognition: Forbes 2023 (Eur J Nutr systematic review): 5–20 g/d improves memory, especially in vegetarians, older adults, sleep-deprived states. Mood: Smith-Ryan 2021 review — 5 g/d as SSRI adjunct accelerates remission.
Actionable protocol
GoalDoseTiming / form
Strength training3–5 g/dayAnytime; ideally with a meal + carbs
Cognition / aging5 g/dayDaily; saturation 2–4 weeks
Vegetarian5 g/dayLow baseline → largest gain
Not recommended"20 g loading phase"GI upset, no extra benefit
Form: "Creatine Monohydrate" only (skip HCl / ethyl ester / buffered — no advantage, higher cost). Creapure is the German reference raw material. Water: +500 ml/day. Contraindication: known severe kidney disease. In healthy adults, 5 g/d long-term safety is well established (Kreider 2017).
Women's note + common myths
Women's baseline creatine stores are 70–80% lower than men's, so the response is larger. Stacy Sims recommends 5 g/d for perimenopausal women — muscle, cognition, mood, all three. Energy metabolism tightens in the luteal phase; creatine helps maintain training performance. Pregnancy: small studies show no harm signal but no large RCT — default to avoiding.
Myth 1: "Creatine is a hormone" — it's an amino-acid derivative; no hormonal action.
Myth 2: "Stop creatine and the muscle drops" — only a slight water shift (<1 kg). Lean tissue is retained.
Myth 3: "Bad for kidneys" — multiple RCTs in healthy kidneys disprove this; creatine raises serum creatinine (the marker) but not creatinine clearance. Tell your doctor if a kidney panel is ordered.
This week + reflection
THIS WEEK
Buy a tub of Creapure monohydrate (~$15 for 100 g, ~20 days). Take 5 g every morning in warm water or coffee. After 28 days, retest your 5RM squat / deadlift against baseline.

Reflection: why does a supplement with 30 years of safety data and a clear effect size still live in public perception as "for gym bros only"?
SUB · Test-first / micronutrients
Vitamin D + Magnesium: Test, Then Dose — Don't Guess
Vitamin D + Magnesium
One-line takeaway
Vitamin D and magnesium have the highest real-world deficiency rates. But "everyone takes 2000 IU" is a crude starting point. Test 25(OH)D and RBC magnesium, dose to the gap, retest at 3 months.
Evidence level
Vitamin D: VITAL main result null, but prespecified subgroups with low baseline showed cancer-death HR 0.83 and a 22% drop in autoimmune disease incidence (Hahn 2022, BMJ). Fracture RCT (Bischoff-Ferrari 2017): 800 IU/d + calcium cut hip fracture 30%. Magnesium: Veronese 2016 meta — low Mg strongly linked to T2D, hypertension, CV events; RCTs show supplementation improves insulin sensitivity, sleep quality, migraine frequency.
Actionable protocol
NutrientTestTarget / dose
Vitamin DSerum 25(OH)D, annual30–50 ng/mL; typically 1000–2000 IU/d, deficient → 4000 IU/d, retest at 8 weeks
Vitamin K2Not routinely testedMK-7 90–180 μg/d; synergy with D to direct calcium to bone (Knapen 2013 RCT)
MagnesiumRBC Mg (not serum — far more sensitive)200–400 mg elemental/day; glycinate / threonate / citrate (avoid oxide — <5% absorption)
CalciumDiet estimate (food diary)1000–1200 mg/d, food first; supplemental >500 mg/d carries a CV-risk signal
Stack timing: D + K2 at breakfast with fat (fat-soluble); magnesium at bedtime (sleep + avoids absorption competition with Ca/Zn). Avoid: calcium with iron or zinc — mutual antagonism.
Women's note + common myths
Low magnesium is linked to PMS and menstrual migraines (glycinate 300 mg at bedtime is a cheap intervention). Perimenopausal bone protection: D + K2 + Mg + heavy lifting is the foundational stack (see Day 8). Pregnancy: D supplementation under physician guidance, target ≥30 ng/mL (ACOG).
Myth 1: "Serum Mg normal = not deficient" — 99% of body Mg is intracellular; serum is insensitive. RBC Mg is the reliable test.
Myth 2: "Higher is better" — vitamin D >100 ng/mL raises risk of hypercalcemia and kidney stones; 4000 IU/d is a conservative untested ceiling.
Myth 3: "Magnesium oxide works because it makes me poop" — laxation ≠ systemic absorption. Oxide is near-useless for repletion.
This week + reflection
THIS WEEK
Book a 25(OH)D + RBC Mg test (most labs offer it as an add-on, ~$20–40). Do not start or change a dose before the results land. Decide after the numbers arrive.

Reflection: "test first, then dose" sounds obvious — why is it so rare in supplement decisions? Is it that marketing is fast and lab work is slow, or do we prefer the illusion of action over delayed gratification?
SUB · Hype detection / anti-aging
NMN / NR / Resveratrol: Great Story, Where's the Human Data?
NMN / NR / Resveratrol
One-line takeaway
NMN, NR, and resveratrol have seductive mechanisms (NAD+ pathway, sirtuin activation) and animal lifespan data — but almost no human hard-endpoint RCTs. They're "science worth tracking," not "mature interventions worth hundreds of dollars a month."
Evidence level
NR/NMN: human RCTs consistently show blood NAD+ does rise (Martens 2018, Yoshino 2021), but downstream clinical endpoints (performance, insulin sensitivity, aging biomarkers) move little or not at all. Resveratrol: early Howitz/Sinclair animal data failed to replicate; Pollack 2017 meta — CV markers mixed, no clear hard endpoint. The marketing-to-clinical-reality gap is enormous. Attia (2023 podcast): personally stopped resveratrol; on NMN "watchful, not validated."
Visualization: evidence vs marketing heat
Hard-endpoint evidence (green) vs marketing heat (red) Creatine 500+ RCTs Omega-3 REDUCE-IT etc. Vitamin D Effective if stratified NMN/NR Mechanism Resveratrol Evidence Marketing
Green = hard-endpoint RCT strength; red = marketing/search heat. NMN/NR/resveratrol are severely inverted.
Actionable protocol
Stance: keep "NAD+ class" in the watch pool, not the core protocol.
Substitute: want NAD+ up? Lift weights + Zone 2 + adequate sleep — free and hard-endpoint-backed.
If you insist on trying: NR or NMN 500 mg/d for 12 weeks; measure pre/post performance (VO2max, grip), HOMA-IR, optionally GrimAge/PhenoAge clocks. No change → stop.
Don't bother with resveratrol — Sinclair's follow-on data retracted; isolated hepatotoxicity cases exist.
Women's note + common myths
Perimenopausal women are the prime target of "anti-aging stack" marketing. Mary Claire Haver's stance: the same monthly spend put toward an MHT consultation, strength-training coaching, and DEXA scans returns vastly more than an "NAD+ bundle." Prioritization, not all-or-nothing.
Myth 1: "Blood NAD+ up = anti-aging" — surrogate endpoint ≠ clinical benefit (recall the early errors of LDL ≠ MI, BMD ≠ fracture).
Myth 2: "David Sinclair takes it, so I should" — single celebrity protocols ≠ population evidence; he himself revises it constantly.
Myth 3: "It's cheap, so evidence doesn't matter" — $1000/year × 30 years deserves RCT-grade scrutiny.
This week + reflection
THIS WEEK
List every "anti-aging" supplement you take and annualize the spend. The same dollars could buy: ① one full blood panel; ② 8 sessions of strength coaching; ③ a DEXA + VO2max test. Which helps the 5-years-from-now you most?

Reflection: when there's a 10–20-year gap between mechanism studies, animal data, and human hard endpoints, paying to be a guinea pig now vs waiting for Phase III RCTs — which posture is more rational, and what does it depend on?

📚 Resources

🧠 Reflections

① Why does the chain "mechanism → animal → human hard endpoint" so often break for NAD+ compounds?
Mouse lifespan is a single inbred strain in standardized cages — a reproducible phenotype. Humans are heterogeneous, long-lived, and complex in exposure. Historically, fewer than ~5% of nematode/mouse lifespan-extending interventions translate into clinical benefit in humans (rapamycin is a rare exception). For a mechanism to graduate into the "mature intervention" list, it must cross this translational gulf — typically a 10–20-year wait.
② "Test before you supplement" — why is it easy to say, hard to do?
Testing has friction: book, draw, wait, read, decide. Buying a supplement is one click. Behavioral economics: low-friction wrong beats high-friction right, repeatedly. The fix isn't moral willpower — it's process design: bundle an annual blood panel with your physical, and defer all supplement decisions until results land.
③ Public celebrity/expert protocols — popular science or noise amplifier?
Double-edged. Good: makes N=1 experiments transparent and sparks discussion. Bad: viewers misread "they do it" as "the evidence is in." Healthy posture: treat celebrity stacks as research direction hints, not prescriptions. Downgrade each line item by one evidence tier when you evaluate.
④ An intervention is "harmless but possibly useless" — pay to try, or skip?
Three variables: cost (annual × discounted lifespan), opportunity cost (the marginal return of putting that money into a higher-certainty intervention), psychic tax (cognitive load of swallowing pills daily). <$70/yr, plausible mechanism, clean safety → fine to try. Thousands of dollars annually, mechanism only animal-level, eight capsules a day — usually not worth it.
⑤ When supplements become identity ("I'm in the longevity tribe"), how do you avoid fooling yourself?
Identity-laden decisions are the hardest to reverse — stopping = admitting you were wrong earlier. The best guardrail is a pre-committed stop rule: at the outset, write down "if marker X has not moved by week 12, I will stop." It frees the future you from sunk cost.
← Back to Index