Supplements: the few that work and the many that don't

The global dietary supplement market is worth over $150 billion, according to industry estimates. The evidence supporting most of those products could fit on a napkin. Here's an honest, evidence-based look at which supplements are actually worth taking — and which ones are expensive placebo.

The ones that work (with evidence)

Vitamin D — if you're deficient

About 40% of Europeans and 42% of US adults have 25-OH vitamin D below 20 ng/mL; about 13% of Europeans are below 12 ng/mL, clear deficiency. If your level is genuinely low, supplementation is straightforward: 2,000–4,000 IU daily, recheck in 3 months. The evidence is strongest for bone health in deficient individuals. The key word is deficient — in the largely non-deficient population of the VITAL trial (n=25,871), supplementation did not prevent cancer, cardiovascular disease, or depression. More detail is in the article on vitamin D deficiency.

Omega-3 (EPA/DHA) — specific contexts

The evidence for omega-3 supplementation is more nuanced than the supplement industry suggests. The REDUCE-IT trial showed that high-dose EPA (icosapent ethyl, 4g/day) reduced cardiovascular events by 25% in patients with elevated triglycerides already on statins — though critics argue its mineral-oil placebo may have flattered the drug. The STRENGTH trial (4 g/day EPA/DHA versus corn oil) found no benefit; the reason for the difference is debated. General-population supplementation with over-the-counter fish oil capsules (typically 1g with low EPA content) has not shown cardiovascular benefit in most trials.

Magnesium — low intake is common

About half of Americans consume less magnesium than recommended; PPIs and diuretics lower levels further. Deficiency can cause cramps, fatigue, and arrhythmias. If deficient, supplementing is reasonable — within the upper limit for supplemental magnesium of 350 mg/day (US) or 250 mg/day (EU, BfR). Document deficiency first: serum magnesium is simple and cheap, though an imperfect marker of total body stores.

B12 — if levels are low

B12 deficiency is common in older adults (reduced absorption), vegetarians/vegans, and PPI users. In an Oxford cohort of older adults (OPTIMA), B12 in the lowest third (below about 420 pg/mL) was associated with faster brain volume loss; in a related trial, B vitamins slowed brain atrophy, mainly in people with high homocysteine. Supplementation is cheap and effective — 1,000μg oral daily, or intramuscular injection if levels are very low (<200).

The ones that don't work (despite the marketing)

Multivitamins

Large randomized trials — including the Physicians' Health Study II (n=14,641) and the COSMOS trial — have found no benefit of daily multivitamin supplementation for cardiovascular disease in well-nourished adults. For cancer, results were inconsistent (8% fewer cancers in PHS II, no significant effect in COSMOS); COSMOS sub-studies found a small memory benefit of unclear practical relevance. A multivitamin is not harmful in most cases, but it's not doing what people think it's doing.

NAD+ and NMN

Nicotinamide mononucleotide (NMN) and NAD+ precursors are among the hottest supplements in the longevity space. The animal data is genuinely interesting — NAD+ levels decline with age, and restoring them in mice improves various markers of aging. The human data, however, is extremely limited. No large randomized trial has shown clinical benefit. The supplements are expensive ($50–200/month), and the optimal dose, formulation, and long-term safety in humans are unknown.

Collagen supplements

The evidence for oral collagen supplements improving skin, joints, or bones is weak. The theoretical problem is basic biochemistry: collagen is a protein, your digestive system breaks it down into amino acids, and those amino acids don't preferentially reassemble into collagen in your skin. Small studies showing modest improvements have significant methodological limitations.

Testosterone boosters

"Natural testosterone boosters" — typically containing ashwagandha, tribulus, fenugreek, or D-aspartic acid — do not meaningfully raise testosterone levels in men with normal hormonal function. The studies cited by manufacturers are typically small, short, and use surrogate endpoints that don't translate to clinical outcomes.

The evidence hierarchy for supplements

• Strong evidence (supplement if deficient): Vitamin D, B12, iron (if ferritin is low), magnesium, folate (pregnancy)
• Moderate evidence (specific clinical contexts): High-dose EPA (elevated TG on statins), calcium + D (postmenopausal osteoporosis), creatine (exercise performance)
• Weak or no evidence: Multivitamins, NAD+/NMN, collagen, testosterone boosters, most "longevity stacks"
• Evidence of harm: High-dose vitamin E (increased mortality), high-dose beta-carotene (increased lung cancer in smokers), excessive vitamin A

The pattern is clear: supplements work when they correct a documented deficiency. They rarely work as general health enhancers in well-nourished people. And in a few cases, they cause genuine harm. The only way to know which category you're in is to measure first.

This article is general information and not individual medical advice. Decisions about tests and treatment belong in a conversation with your own physician.