Evidence library

Supplement Evidence Library

Evidence reviews for 37 supplement ingredients, drawing on original research, evidence syntheses, and regulatory guidance. Articles are being standardized to cover relevant forms, interactions, side effects, evidence limitations, and a plain-language verdict.

Evidence reviews and safety notes

Supplement ingredient reviews

37 ingredient articles. Evidence labels are being standardized to the four-level scale below. Open an article to see its current scope, sources, limitations, and safety information.

Grade under editorial review
Magnesium
Sleep, stress, blood pressure — which form matters.
 Grade under editorial review
Omega-3
Heart, mood, inflammation — EPA/DHA dosing.
 Strong
Melatonin
Sleep onset, jet lag, dose-dependent effects.
 Moderate
Vitamin D
Deficiency, mood, immune — who actually needs it.
 Grade under editorial review
St John's Wort
High interaction risk
Mild depression — with serious interaction risks.
 Moderate
CoQ10
Heart, energy, statin-related muscle pain.
 Moderate
L-Tyrosine
Stress resilience, focus under acute pressure.
 Moderate
L-Theanine
Calm focus, caffeine pairing, anxiety support.
 Moderate
Saffron
Mood, depression — what the 34-trial meta-analysis shows.
 Moderate
Garlic
Blood pressure — allicin, aged extract, real effect size.
 Moderate
NMN
NAD+ precursors, aging, energy — human evidence vs hype.
 Grade under editorial review
Probiotics
Strain-specific benefits — which ones actually work.
 Grade under editorial review
Collagen
Skin, joints — hydrolyzed collagen and what trials show.
 Moderate
Berberine
Blood sugar, lipids — metabolic effects and dosing.
 Moderate
Ashwagandha
Stress, cortisol — KSM-66 and the evidence behind it.
 Strong
Creatine
Muscle, cognition — the most-studied supplement.
 Moderate
Stevia
Stevioside vs Reb A vs Reb M — which form actually works.
 Grade under editorial review
Monk Fruit
Forms, mogroside content, safety sources, and the limits of the available human evidence.
 Grade under editorial review
Aspartame
IARC 2B, NutriNet-Santé, JECFA ADI — what the evidence shows.
 Grade under editorial review
Sucralose
Gut microbiome shifts and glycemic effects in human RCTs.
 Grade under editorial review
Saccharin
The bladder-cancer scare that was rat-specific — resolved.
 Grade under editorial review
Acesulfame-K
The unreplicated cancer cohort signal — what it means.
 Strong
Whey Protein
WPC vs WPI vs WPH vs native — MPS dose-response and the leucine trigger.
 Weak
Yeast Protein
Single-cell vegan protein — one manufacturer-funded RCT, weak evidence.
 Grade under editorial review
Plant Protein
Pea/soy/rice/hemp — parity with whey when dose and leucine are matched.
 Moderate
Tongkat Ali
Natural testosterone booster? What human trials actually show.
 Moderate
Psyllium Husk
Nature's Ozempic? LDL and glucose yes — weight loss, weak.
 Grade under editorial review
Bovine Colostrum
Liquid gold for immunity? Athlete URTI data, beauty claims thin.
 Weak
Glutathione
Skin brightening from within? Oral weak, IV unsafe.
 Moderate
Myo-Inositol
PCOS, metabolic outcomes, ovulation and fertility claims reviewed against the available evidence.
 Moderate
Nattokinase
Dissolves microclots? Modest BP effect, bleeding risk.
 Moderate
Astaxanthin
Internal sunscreen? Skin moisture yes, lipid claims weak.
 Weak
Lion's Mane
Brain fog cure? Small pilots, thin human data.
 Weak
Rhodiola
Adaptogen for burnout? Promising but methodologically weak.
 Weak
Shilajit
Mountain mineral pitch — weak data, heavy-metal safety caveat.
 Grade under editorial review
L-Glutamine
Leaky gut repair? One strong RCT in a narrow IBS subgroup; critical-illness harm signal.
 Moderate
Zinc-L-Carnosine
Gastric ulcer healing and H. pylori adjunct — Japan drug, US supplement.
Beyond ingredients

Related health and comparison articles

Health-condition guides and comparison articles are listed separately from the 37 supplement ingredient reviews.

Natural vs Organic
'Natural' has no legal definition; organic is certified — but neither is proven healthier.
Natural vs Chemicals
Why natural doesn't always mean safe, and chemical doesn't always mean dangerous.
 Browse current article
Narcolepsy
Excessive daytime sleepiness & cataplexy — evidence-based treatments are prescription drugs; no supplement has adequate evidence.
 Browse current article
Chronic Rhinosinusitis
Saline irrigation is the strongest non-drug intervention; most supplements lack CRS evidence.
 Browse current article
Rheumatoid Arthritis
DMARDs are standard of care; omega-3 fish oil is the best-supported supplement adjunct.
Evidence-grade legend

How we grade supplement evidence

Every ingredient carries one of four grades. The grade reflects the quality, consistency, and independence of the studies behind a claim — not how often the claim is repeated online.

Strong Strong

Multiple high-quality randomised trials or systematic reviews agree. Effect is consistent and clinically meaningful.

Moderate Moderate

Reasonable trial evidence, but with limits — smaller samples, mixed results, or benefits confined to specific groups.

Weak Weak

Early or low-quality evidence — a few small studies, observational data, or findings that have not been replicated.

Insufficient Insufficient

Not enough reliable human data to judge. Claims rest mainly on mechanism, animal studies, or expert opinion.

A separate flag

Some ingredients also carry a high interaction risk flag. This is independent of the evidence grade — a supplement can be well-studied yet still interact dangerously with common medicines, as with St John's Wort.

Consistent structure

Standard ingredient-article template

Ingredient articles are being standardized around the following six-part structure. Check the visible sections of an individual article for its current coverage.

1
Forms & types
The specific forms that matter for absorption and effect — and the cheaper forms that often disappoint.
2
Interactions
Medicines, other supplements, and conditions that change how the ingredient behaves or raise its risk.
3
Side effects
Reported adverse effects, the populations and doses studied, and important limits or uncertainties.
4
What works & what doesn't
The uses supported by solid evidence, separated clearly from the popular claims that don't hold up.
5
FAQ
Direct answers to the questions people ask most — dosing, timing, safety, and who should avoid it.
6
Citations
Verdict-bearing claims should link to original research, evidence syntheses, or regulators. Contextual sources should be identified and should not determine the grade.
Our methodology

We prioritize original research, evidence syntheses, and regulatory guidance. Study design, relevance, consistency, funding, and disclosed conflicts are considered when interpreting the evidence. Missing funding information is recorded as unclear rather than assumed independent.

Read how we research and grade evidence · See our full source list

Grades reflect the current body of evidence and are revised as new high-quality research is published.