Research
The library
Search for what you take or what you are after. Each page says what it was shown to do, the risks reported, and how strong the evidence is. Sources are trials and regulators, never marketing.
Nothing here matches yet. Try a different name, or what you take it for.
Medication
ACE inhibitors
Lower blood pressure and fewer deaths in heart failure and after a heart attack. Cough is frequent. Boxed warning for pregnancy.
Shown to help 6 · Mixed results 1 · Not shown to help 2.
Acetaminophen
Relieves acute pain and tension headache in pooled trials. No effect found on low back pain. Liver failure mostly above 4,000 mg a day.
Shown to help 6 · Mixed results 2 · Not shown to help 2.
ADHD stimulants
Reduced ADHD symptoms in short trials of children and adults. Small lab effects without ADHD. Boxed warning for abuse, misuse and addiction.
Shown to help 8 · Mixed results 1 · Not shown to help 1.
Albuterol
Fast airway opening for about 4 hours and protection before exercise. Heavy use was linked to more asthma attacks and deaths.
Shown to help 5 · Mixed results 1 · Not shown to help 4.
Allopurinol
Lowers uric acid in gout. No kidney or heart benefit in trials. Severe skin reactions in about 5 in 10,000 people.
Shown to help 5 · Mixed results 1 · Not shown to help 4.
Alpha-blockers
Fewer prostate symptoms than placebo and more passage of larger stones. Dizziness and ejaculation problems were common, and small stones passed no faster.
Shown to help 7 · Mixed results 1 · Not shown to help 2.
Amitriptyline
Raised depression response against placebo, with low-quality evidence for pain uses. Dry mouth, drowsiness and weight gain were common.
Shown to help 6 · Mixed results 1 · Not shown to help 2 · Claimed, not studied in people 1.
Antibiotics (common outpatient)
Small gains for sinusitis, sore throat and ear infections, none for colds or bronchitis. Diarrhea, rash, C. difficile and resistance were reported.
Shown to help 7 · Not shown to help 2.
Antihistamines
Relief of hay fever and hives over placebo. Diphenhydramine helped sleep little and impaired driving, and stopping cetirizine can cause severe itching.
Shown to help 2 · Mixed results 3 · Not shown to help 1 · Claimed, not studied in people 1.
Antiplatelets: clopidogrel and low-dose aspirin
About a quarter fewer vascular events after heart attack or stroke. Little net gain for primary prevention. Bleeding is the main harm.
Shown to help 8 · Mixed results 2 · Not shown to help 1.
ARBs (angiotensin receptor blockers)
Lower blood pressure, slower diabetic kidney disease and fewer strokes. Less cough than ACE inhibitors. Boxed warning for pregnancy.
Shown to help 7 · Mixed results 1 · Not shown to help 1.
Atomoxetine and viloxazine
Non-stimulant ADHD drugs with moderate symptom gains, smaller than amphetamines. Nausea, appetite loss and sleepiness are very common, with suicidal-thoughts warnings.
Shown to help 5 · Mixed results 1.
Atypical antipsychotics
Reduced symptoms of schizophrenia, mania and hard-to-treat depression in trials. Weight gain, sleepiness and restlessness were frequent. Boxed warning in dementia.
Shown to help 10 · Mixed results 1.
Benzodiazepines
Short-term relief of panic, anxiety and insomnia in trials. Boxed warnings for opioid use, addiction, dependence and withdrawal. Linked to falls and crashes.
Shown to help 6 · Mixed results 1 · Not shown to help 1.
Benzopyrones
Trials of coumarin for lymphedema disagree. Liver toxicity led Australia to cancel coumarin medicines.
Mixed results 1.
Beta-blockers
About a third fewer deaths in heart failure. Recent heart attack trials disagree. Labels warn against stopping abruptly.
Shown to help 6 · Mixed results 2 · Not shown to help 1.
Bisphosphonates
Fewer spine, hip and other fractures in women with osteoporosis. Rare atypical thigh fractures with long use, and flu-like reactions after infusions.
Shown to help 5 · Mixed results 1 · Not shown to help 2.
Bupropion
Raised depression response and quit rates over placebo, with sexual side effects at placebo level. Insomnia was common, and seizure risk rose with dose.
Shown to help 6 · Mixed results 1.
Buspirone
Reduced generalized anxiety more than placebo. Did not beat placebo when added to an SSRI. Dizziness was the main side effect, without sedation or withdrawal.
Shown to help 2 · Mixed results 1 · Not shown to help 5.
Calcium channel blockers
Lower blood pressure and stroke. More heart failure than diuretics. Ankle swelling is frequent, and diltiazem and verapamil slow the heart.
Shown to help 7 · Mixed results 1.
CGRP migraine drugs
About 1.6 fewer migraine days a month than placebo on the antibodies, and modest attack relief from gepants. Constipation and injection reactions were most frequent.
Shown to help 5 · Mixed results 1.
Clonidine
Lower blood pressure and fewer ADHD symptoms in children. Drowsiness was very common, and stopping suddenly can cause rebound high blood pressure.
Shown to help 5 · Not shown to help 1.
Direct oral anticoagulants (DOACs)
Fewer strokes and about half the brain bleeds of warfarin in atrial fibrillation. More stomach bleeding. Boxed warnings on stopping early.
Shown to help 6 · Mixed results 1 · Not shown to help 1.
Dulaglutide
Lower HbA1c and 12% fewer heart events in type 2 diabetes, with small weight loss. Stomach side effects were very common.
Shown to help 6 · Not shown to help 1.
Ezetimibe and fenofibrate
Ezetimibe adds LDL lowering and a small cut in heart events to statins. Fenofibrate lowers triglycerides but did not cut heart events with statins.
Shown to help 8 · Mixed results 1 · Not shown to help 2.
Finasteride
More scalp hair than placebo over 2 years. Sexual side effects rose. Persistent symptoms after stopping are reported.
Shown to help 4.
Fluoride
Toothpaste, varnish and rinses cut cavities. Supplements lowered cavities in permanent teeth. Mild fluorosis rose in young children, and high water fluoride tracked lower IQ.
Shown to help 4 · Mixed results 1 · Not shown to help 2.
Gabapentinoids
Relieved shingles and diabetic nerve pain in about a third. No benefit for sciatica or low back pain. FDA breathing warning with opioids.
Shown to help 7 · Mixed results 2 · Not shown to help 2.
GnRH drugs for fibroids
About 70% reached normal blood loss against 9% to 19% on placebo. Bone loss is the main limit, smaller with add-back hormones.
Shown to help 8 · Mixed results 1.
H2 blockers
More ulcer healing and heartburn relief than placebo, less than proton pump inhibitors. The effect faded with daily use, and cimetidine did not clear warts.
Shown to help 7 · Mixed results 1 · Not shown to help 1.
Hormonal birth control
Prevents pregnancy, eases acne and period pain. Raises clot and breast cancer risk slightly, lowers ovarian and endometrial cancer, and fertility returns after stopping.
Shown to help 9 · Mixed results 1.
Hydroxyzine
Reduced generalized anxiety more than placebo in short trials. Limited evidence for itch and sleep. Sleepiness was very common. The label warns of QT prolongation.
Shown to help 2 · Mixed results 1 · Not shown to help 2.
Inhaled corticosteroids
Fewer asthma attacks and COPD flare-ups. More thrush, hoarseness and, in COPD, pneumonia. Children grew slightly less.
Shown to help 8.
Insulin
Fewer diabetes complications with tight control in type 1 diabetes. No heart benefit shown in type 2. Low blood sugar and weight gain.
Shown to help 6 · Mixed results 1 · Not shown to help 1 · Claimed, not studied in people 1.
Isotretinoin
The largest acne lesion reductions of any acne drug studied. Dryness and lipid rises are common, and pregnancy exposure causes severe birth defects.
Shown to help 2 · Mixed results 2 · Not shown to help 1.
Ketamine and esketamine
Faster drops in depression scores than placebo and short-lived relief of suicidal thoughts. Dissociation and sedation are very common, and esketamine has a boxed warning.
Shown to help 8 · Mixed results 2.
Ketoanalogues
Added to a low-protein diet, trials report slower kidney decline and lower phosphate. Blood calcium rose. Death rates did not differ.
Shown to help 2.
Lamotrigine
Fewer bipolar relapses and focal seizures than placebo, a modest effect on bipolar depression, and a boxed warning for rare serious rash.
Shown to help 4 · Mixed results 2 · Not shown to help 2.
Letrozole and clomiphene
More live births with letrozole than clomiphene in PCOS, no difference in unexplained infertility. Multiple pregnancy and hot flushes are among recorded harms.
Shown to help 4 · Not shown to help 3.
Levothyroxine
TSH normalized, but symptoms and quality of life did not change in mild hypothyroidism. Over-replacement was linked to fractures.
Shown to help 1 · Mixed results 1 · Not shown to help 3.
Liraglutide
About 5 to 6 kg more weight loss than placebo in a year. Fewer heart events in type 2 diabetes. More nausea and gallbladder disease.
Shown to help 6.
Loop diuretics
Removes fluid and lowers blood pressure modestly. Torsemide matched furosemide on deaths. Low potassium, gout and hearing problems at high doses were reported.
Shown to help 2 · Mixed results 3 · Not shown to help 3.
Low-dose naltrexone
Fibromyalgia pain results disagree across trials. Small, mixed trials in Crohn's and multiple sclerosis. Vivid dreams were more common than on placebo.
Shown to help 1 · Mixed results 3 · Not shown to help 2.
Metformin
Lowers diabetes onset in high-risk adults. Longevity data disagree. Blunted muscle and fitness gains from training in trials.
Shown to help 2 · Mixed results 1.
Methylene blue
A small memory gain in one trial, no benefit in two Alzheimer's trials. Serotonin syndrome reported with SSRIs.
Shown to help 2 · Not shown to help 1 · Claimed, not studied in people 2.
Minoxidil
Topical minoxidil raised hair counts against placebo. Oral and topical did not differ in pooled trials. Oral use brought unwanted body hair.
Shown to help 4 · Mixed results 1.
Mirtazapine
Reduced depression in trials and eased insomnia for about six weeks. Sleepiness, appetite increase and weight gain were very common. Boxed warning on suicidal thoughts.
Shown to help 3 · Mixed results 1 · Not shown to help 4.
Modafinil and armodafinil
Less sleepiness in narcolepsy, sleep apnea and shift work disorder. Only a very small cognitive gain in rested healthy adults. Headache and nausea were common.
Shown to help 6 · Mixed results 2.
Montelukast
Modest gains in asthma and hay fever, less than inhaled steroids. Boxed warning for mood and behavior changes, with mixed study findings.
Shown to help 5 · Mixed results 1 · Not shown to help 2.
Muscle relaxants
About 8 points of 100 less acute low back pain for 2 weeks. Nothing added to an NSAID in two trials. Drowsiness is frequent.
Shown to help 5 · Mixed results 1 · Not shown to help 1.
Naltrexone-bupropion
About 2.5 to 5 kg more weight loss than placebo at a year. Nausea in about 3 in 10. A heart trial ended early.
Shown to help 2 · Not shown to help 1.
NSAIDs
Less soreness after muscle injury. Training results split by dose and age. More heart, stomach and kidney events.
Shown to help 4 · Mixed results 1.
Ondansetron
Less vomiting after surgery, with chemotherapy and in children with stomach bugs. Headache and constipation were common, and the label warns of QT prolongation.
Shown to help 6.
Opioid pain relievers
Relieve pain after surgery. Small effect in chronic pain, none in three acute back and limb pain trials. Boxed warnings for addiction and overdose.
Shown to help 6 · Not shown to help 2.
Oral corticosteroids
Fewer asthma relapses and COPD treatment failures, less joint pain in rheumatoid arthritis. More infections, high blood sugar, fractures and adrenal suppression.
Shown to help 9 · Not shown to help 1.
Orforglipron
About 11% weight loss at the top dose in 72 weeks, and lower blood sugar in type 2 diabetes. Stomach side effects were common.
Shown to help 3.
Orlistat
About 2 to 3 kg more weight loss than placebo at a year. Oily stools and urgency were common.
Shown to help 4.
PDE5 inhibitors
Better erections than placebo. Some gains in thin-air exercise, none at sea level. Nitrates are contraindicated.
Shown to help 3.
Phentermine
Phentermine-topiramate cut about 8 to 9 kg more than placebo at a year. More tingling, dry mouth and a faster heart rate.
Shown to help 3.
Progestins for endometriosis and heavy bleeding
Eased endometriosis pain against placebo and matched leuprolide. Bleeding changes are very common, and the injection carries a boxed bone loss warning.
Shown to help 3 · Mixed results 2.
Proton pump inhibitors
More heartburn relief and esophagitis healing than placebo. Observational links to fractures, kidney disease and dementia were mostly not seen in a large trial.
Shown to help 9 · Mixed results 2 · Not shown to help 2.
Rapamycin
Small aging trials found little change in their main outcomes. No human lifespan data. Transplant use carries infection and lipid risks.
Shown to help 2 · Not shown to help 2 · Claimed, not studied in people 1.
Retatrutide
Grey marketPhase 3 trials found 17 to 25% weight loss over 80 weeks, plus sleep apnea and knee pain changes. Investigational and not approved.
Shown to help 7.
Semaglutide
About 15% weight loss in trials and fewer heart events in people with heart disease. Most weight returned after stopping. Lean mass fell too.
Shown to help 6.
SGLT2 inhibitors
Fewer heart failure admissions and slower kidney disease in large trials. More genital yeast infections, and uncommon ketoacidosis.
Shown to help 8 · Not shown to help 1.
SNRIs
Higher response than placebo in depression and anxiety, and pain relief in diabetic nerve pain. Nausea, raised blood pressure and stopping symptoms reported.
Shown to help 9 · Not shown to help 2.
Spironolactone
Fewer deaths in severe heart failure and lower resistant blood pressure. Improved acne in women. Raises potassium, and breast enlargement in men.
Shown to help 5 · Mixed results 1 · Not shown to help 2.
SSRIs
Modest gains over placebo in depression, anxiety, OCD and PTSD. Sexual side effects and stopping symptoms were common. Boxed warning for people under 25.
Shown to help 11 · Mixed results 2.
Statins
Fewer heart attacks and strokes per unit of LDL lowered. A small excess of muscle pain and new diabetes, and mixed training findings.
Shown to help 3.
Sulfonylureas
Lower HbA1c by about 1.5 points. No heart benefit or harm in recent large trials. Low blood sugar and about 2 kg of weight gain.
Shown to help 3 · Not shown to help 3.
Thiazide diuretics
Lower blood pressure, stroke and deaths. Chlorthalidone and hydrochlorothiazide did not differ on heart events. Low potassium and a small skin cancer signal.
Shown to help 3 · Mixed results 2.
Tirzepatide
15 to 21% weight loss in trials, more than semaglutide head to head. Weight returned after stopping, and a quarter of loss was lean mass.
Shown to help 4 · Not shown to help 2.
Topiramate
About one fewer migraine a month and some weight loss. Tingling and thinking problems were common, and oral clefts rose with pregnancy exposure.
Shown to help 6 · Mixed results 2.
Tranexamic acid
Cut period blood loss about 40% against placebo, with or without fibroids. Clot findings differ between a large cohort and pooled trials.
Shown to help 5 · Mixed results 2.
Trazodone
Reduced depression symptoms like SSRIs, and improved self-rated sleep in short trials. Drowsiness and dizziness were very common. Label warns of priapism.
Shown to help 4 · Not shown to help 1.
Tretinoin and topical retinoids
More acne clearance than vehicle, and modest gains in fine sun-related wrinkles with tretinoin. Peeling, dryness and redness are very common.
Shown to help 4 · Mixed results 3 · Claimed, not studied in people 1.
Triptans
Oral sumatriptan left 28% pain free at 2 hours against 11% on placebo. Tingling and chest tightness are frequent. Contraindicated in coronary disease.
Shown to help 8.
Ulipristal acetate
Controlled fibroid bleeding in about 90% against 19% on placebo. Rare serious liver injury ended its EU fibroid use. US approval is emergency contraception.
Shown to help 6 · Mixed results 1.
Warfarin
About 64% fewer strokes in atrial fibrillation than no treatment. Boxed warning for major bleeding. Many drug, supplement and food interactions.
Shown to help 6 · Mixed results 2 · Not shown to help 1.
Zolpidem
Fell asleep about 22 minutes faster in pooled trials. Boxed warning for sleepwalking and sleep-driving, with next-morning impairment and fracture links.
Shown to help 4 · Mixed results 1.
Hormone therapy
DHEA
Small lean mass and sexual function changes in pooled trials. No strength or quality of life change in older adults. More acne in women.
Shown to help 6 · Not shown to help 2.
Growth hormone
More lean mass and less fat, but no strength gain in healthy people. Swelling and joint pain rose.
Shown to help 4 · Not shown to help 1 · Claimed, not studied in people 1.
hCG
No weight loss effect in controlled diet trials. Sperm production restored in most men with hypogonadotropic hypogonadism in pooled studies.
Shown to help 5 · Not shown to help 2.
Menopausal hormone therapy
Hot flashes and fractures fell in trials. Breast cancer, stroke and clot risks rose with estrogen plus progestin, and muscle did not change.
Shown to help 5 · Not shown to help 3.
Testosterone therapy
Small sexual function and lean mass gains in men with low testosterone. Heart attacks matched placebo, while arrhythmias, erythrocytosis and fractures rose.
Shown to help 5 · Mixed results 1.
Peptides and SARMs
BPC-157
Grey marketNot approved, banned in sport. Healing claims rest on rat studies and one small uncontrolled knee study. FDA found too little safety data.
Shown to help 1 · Claimed, not studied in people 4.
Bremelanotide
Raised a sexual desire score slightly more than placebo in premenopausal women. Nausea reached 40%, and 18% stopped because of side effects.
Shown to help 3.
CJC-1295 and ipamorelin
Grey marketCJC-1295 raised growth hormone in small trials. No muscle or fat loss trials were sourced. Both are unapproved, and FDA cites serious adverse events.
Shown to help 1 · Not shown to help 1 · Not yet researched 5.
Epitalon
Grey marketOne small trial raised a melatonin marker in women. No research on sleep or aging was sourced. FDA found no human safety data.
Shown to help 1 · Not yet researched 2.
GHK-Cu
Grey marketA 1994 trial found more diabetic ulcer closure with GHK-Cu gel. After laser resurfacing, objective skin measures did not differ. No hair trials were sourced.
Shown to help 1 · Mixed results 1 · Not yet researched 2.
KPV
Grey marketNot approved. Gut and skin claims rest on animal and lab studies. FDA found no human exposure data, so its human risks are unknown.
Claimed, not studied in people 3.
MK-677
Grey marketRaised fat-free mass in older adults without changing strength or fat. Blood sugar rose, and one trial stopped early for heart failure.
Shown to help 1 · Not shown to help 1.
MOTS-c
Grey marketNot approved, banned in sport. Metabolic and endurance claims rest on mouse studies. FDA found no human exposure data, so its human risks are unknown.
Claimed, not studied in people 5.
SARMs
Grey marketRaised lean mass in short trials but lowered testosterone and HDL. Online products were often mislabeled, and liver injury cases are reported.
Shown to help 1 · Mixed results 1.
Semax
Grey marketRussian stroke studies reported faster recovery with semax. No focus or memory research was sourced. FDA found too little safety data.
Shown to help 1 · Mixed results 1 · Claimed, not studied in people 2.
Sermorelin
Raised growth rate in children with growth hormone deficiency, less fully than growth hormone. Older men showed no body composition change.
Shown to help 3.
TB-500
Grey marketNot approved, banned in sport. Healing claims rest on animal and lab studies. FDA found no human exposure data, so its human risks are unknown.
Claimed, not studied in people 3 · Not yet researched 2.
Tesamorelin
Cut deep belly fat about 15% more than placebo in adults with HIV. The fat returned after stopping, and blood sugar rose in some.
Shown to help 4.
Supplements
5-HTP
Few quality depression trials. Amplifies SSRI hormone responses. Peak X contaminants found in commercial products.
Shown to help 4.
Activated charcoal
Cuts drug absorption when given soon after. No change in poisoning deaths or gut gas. Also blocks medicines, and aspiration has harmed lungs.
Shown to help 2 · Not shown to help 4.
Aloe vera
Lower blood sugar in small oral trials and faster burn healing on the skin. Whole-leaf extract caused colon tumors in rats.
Shown to help 7 · Mixed results 1.
Alpha-GPC and choline supplements
Dementia trials report cognitive gains, from low-quality studies. Large stroke and brain injury trials found nothing. A cohort linked alpha-GPC to stroke.
Shown to help 2 · Mixed results 2 · Not shown to help 1.
Alpha-ketoglutarate
Biological age claims rest on uncontrolled and cross-sectional data tied to one product. One bone marker trial. Lifespan untested in humans.
Shown to help 3 · Claimed, not studied in people 1.
Alpha-lipoic acid
IV and short oral courses eased diabetic nerve symptoms. Weight fell about 1 kg. Case reports link it to autoimmune low blood sugar.
Shown to help 3 · Mixed results 1 · Not shown to help 1.
Andrographis
Fewer cold and sore throat symptoms in pooled trials. No pooled change in ulcerative colitis. Anaphylaxis and angioedema reported in Thailand.
Shown to help 2 · Mixed results 1 · Not shown to help 1.
Apigenin
We haven't sourced research on this yet.
Apple cider vinegar
Lower fasting glucose in type 2 diabetes. A key weight trial was retracted. Slows stomach emptying. Tablet contents varied.
Shown to help 2 · Mixed results 1.
Arnica
Homeopathic pills did not clearly differ from placebo after surgery. Topical gel matched ibuprofen gel for hand arthritis in one trial.
Shown to help 1 · Mixed results 2 · Not shown to help 2.
Artichoke leaf
Lower LDL and total cholesterol across pooled trials. Lower liver enzymes in small fatty liver trials. No difference from placebo for hangover.
Shown to help 4 · Not shown to help 1.
Ashwagandha
Lower stress, anxiety and cortisol in pooled trials. Liver injury case series include deaths. Germany's risk agency advised against use in some groups.
Shown to help 7 · Mixed results 1 · Not shown to help 1.
Astaxanthin
Mixed muscle damage findings, no performance gain. Small glucose and lipid changes in diabetes. Skin moisture up, wrinkles unchanged.
Shown to help 4 · Mixed results 1 · Claimed, not studied in people 1.
Astragalus
Low-quality trials report kidney markers improving in diabetic kidney disease, and one trial slower eGFR decline. No trials on colds located.
Shown to help 8.
Avocado-soybean unsaponifiables
Small knee osteoarthritis pain reduction against placebo in manufacturer-supported trials. No change in mean joint space loss. Adverse events matched placebo.
Shown to help 1 · Not shown to help 1.
B-complex vitamins
Homocysteine fell, and heart attacks and cognitive aging did not change. Stress fell slightly. Kidney decline sped up in diabetic kidney disease.
Shown to help 2 · Not shown to help 2.
Bacopa
Memory gains in pooled trials, mixed attention findings. No difference in Alzheimer's. Stomach upset in trials.
Shown to help 2 · Mixed results 1 · Not shown to help 1.
BCAAs and EAAs
Less post-exercise soreness and gains in older adults. BCAAs alone stimulate less muscle building than complete protein and did not spare muscle while dieting.
Shown to help 3 · Not shown to help 1.
Beef organ supplements
No benefit research sourced. Liver is rich in preformed vitamin A. Tested glandular products held undeclared thyroid and steroid hormones.
Benefits: no research sourced yet. Risks documented: 3.
Berberine
Lowers HbA1c and LDL modestly. No difference from metformin in trials. Under 1 kg weight loss. Inhibits drug-metabolizing enzymes.
Shown to help 4 · Mixed results 1.
Beta-alanine
Small gains in efforts of 0.5 to 10 minutes after weeks of daily use. No body composition change. Tingling is common.
Shown to help 3 · Not shown to help 2.
Beta-glucan
Oat and barley beta-glucan lowered LDL cholesterol in meta-analyses. Yeast beta-glucan trials reported fewer colds, mostly in small, industry-sponsored studies.
Shown to help 5 · Mixed results 1 · Not shown to help 1.
Beta-sitosterol
Short trials report fewer prostate symptoms than placebo. Plant sterols lower LDL cholesterol 6% to 12%. Heart risk findings conflict.
Shown to help 3.
Bilberry
Rigorous night vision trials were negative. One trial reported less eye muscle tension after screen work. No significant change in blood sugar or cholesterol.
Shown to help 2 · Mixed results 1 · Not shown to help 1.
Biotin
No trials in healthy hair or nails were found. A large MS trial found no benefit. Biotin distorts some lab tests, including troponin.
Shown to help 2 · Mixed results 1 · Not shown to help 1.
Black cohosh
Reviews disagree on hot flashes. Linked to liver injury, including transplants, and carries liver warnings in the UK, EU and Australia.
Mixed results 2 · Not shown to help 2.
Black seed oil
Modest drops in cholesterol, blood sugar and blood pressure in pooled trials. Case reports of kidney injury and bleeding.
Shown to help 6.
Borage oil
Less rheumatoid arthritis pain in small pooled trials. No difference from placebo for eczema. One case report of status epilepticus.
Shown to help 1 · Not shown to help 2.
Boron
A few small trials. Testosterone and strength did not change in bodybuilders. One study without a placebo arm found higher free testosterone.
Shown to help 2 · Mixed results 2 · Not shown to help 2.
Boswellia
Less knee arthritis pain in small trials. No change in Crohn's remission. Side effects matched placebo, with one seizure case.
Shown to help 3 · Mixed results 1 · Not shown to help 1.
Bovine colostrum
Fewer days with cold symptoms in small athlete trials. Lower gut permeability. Mixed performance results. Serum IGF-1 rose in two trials.
Shown to help 2 · Mixed results 1 · Not shown to help 2.
Bromelain
Less pain after wisdom tooth surgery. Swelling results split across reviews. No difference from placebo in a knee arthritis trial.
Shown to help 2 · Mixed results 2 · Claimed, not studied in people 1.
Butcher's broom
Less leg pain and heaviness in venous disease trials. Mixed results on measured swelling. Case reports of colitis and ketoacidosis.
Shown to help 2 · Claimed, not studied in people 1.
Butterbur
Fewer migraine attacks at the higher trial dose. Hay fever relief similar to antihistamines. Liver injury reports, including two transplants.
Shown to help 3.
Butyrate
More ulcerative colitis remission in small add-on trials. Enemas and a children's trial showed no difference. Small, mixed IBS trials.
Shown to help 4 · Not shown to help 2.
Caffeine
Replicated endurance and strength gains at 3 to 6 mg per kg. Costs sleep, raises blood pressure, and pure powders have caused deaths.
Shown to help 5.
Calcium
Bone density rose slightly. Recent reviews found little to no fracture effect. Preeclampsia reviews disagree. Kidney stones rose, and heart findings conflict.
Shown to help 2 · Mixed results 2.
Carbohydrate during exercise
Endurance gains replicate, larger in longer events. Lifting volume rose in longer sessions. Nausea rose with intake. Overdrinking was linked to hyponatremia.
Shown to help 4.
Cat's claw
Fewer painful joints in one small rheumatoid arthritis trial. Mixed osteoarthritis results. Immune trials had no placebo arm.
Shown to help 2 · Mixed results 1.
CBD
Cuts seizures in two rare epilepsies. No pain benefit in 15 of 16 trials. Liver enzyme rises. Most labels inaccurate.
Shown to help 2 · Not shown to help 3.
Centella asiatica
No cognitive gains in a meta-analysis. Small trials on venous swelling and skin. Case reports of liver injury.
Shown to help 3 · Not shown to help 2.
Chanca piedra
Small kidney stone trials disagree. Lower-pole stones cleared more often after lithotripsy. No difference from placebo in stone passage or pain.
Mixed results 1 · Not shown to help 1 · Claimed, not studied in people 1.
Choline
Pregnancy trials mostly found no infant brain change. Alcohol-exposed children showed memory gains. Choline bitartrate raised TMAO and platelet reactivity.
Shown to help 2 · Mixed results 1 · Not shown to help 2.
Chromium
HbA1c fell in diabetes trials of mostly poor quality. Weight fell about 1 kg. Case reports describe kidney failure at high doses.
Shown to help 3 · Not shown to help 2.
Citrulline
Small rep and power gains from a single pre-workout dose. No aerobic effect. Blood pressure fell in older adults. Harms barely studied.
Shown to help 3 · Mixed results 2 · Not shown to help 2.
CLA
Fat and weight change under 1.5 kg in pooled trials. One isomer raised insulin resistance in obese men. CRP rose.
Shown to help 3 · Mixed results 1.
Coenzyme Q10
Statin muscle pain reviews split two and two. Heart failure trials report fewer deaths. Semen quality improved. No exercise performance gain.
Shown to help 5 · Mixed results 5 · Not shown to help 3.
Collagen
Skin results appear only in industry-funded trials. Modest osteoarthritis relief. Tendon gains with training. Few adverse events reported.
Shown to help 5 · Mixed results 3.
Copper
Cholesterol, Alzheimer's progression and arthritis pain did not change. Copper halted decline in deficiency myelopathy. A cohort linked high intake to cognitive decline.
Shown to help 1 · Not shown to help 3.
Cordyceps
Mixed exercise trials, with no change in trained cyclists. Low-quality kidney trials. Wild samples contain arsenic.
Shown to help 2 · Mixed results 1 · Claimed, not studied in people 1.
CPP-ACP
Remineralized white spots more than placebo, not clearly more than fluoride. Gum slowed cavities in one large trial. Made from milk protein.
Shown to help 2 · Mixed results 1.
Cranberry
Fewer UTIs in women with recurrent infections, children and after procedures. No clear benefit in care homes or pregnancy.
Shown to help 1.
Creatine
Lean mass and strength gains during resistance training are well replicated. Small memory gains. No bone effect. Creatinine rises without a measured filtration change.
Shown to help 4 · Not shown to help 1.
D-aspartic acid
Testosterone rose in one short trial of sedentary men. Trained men saw no rise, or a fall at 6 g. Strength unchanged.
Shown to help 1 · Mixed results 1 · Not shown to help 1 · Claimed, not studied in people 1.
D-mannose
No difference from placebo in the largest UTI trial. Earlier small trials reported fewer recurrences. Side effects were mostly diarrhea.
Mixed results 1.
Dandelion
Root powder did not raise urine output beyond water in one small trial. No liver research sourced. Skin allergy appeared in patch tests.
Not shown to help 1 · Claimed, not studied in people 4.
Deer antler velvet
Five of 7 trials found no effect. Hormones and VO2max unchanged. Human IGF-1 was found in 4 of 6 products tested.
Shown to help 1 · Mixed results 1 · Not shown to help 3.
Devil's claw
Short-term back pain relief over placebo in low-quality trials. Mixed osteoarthritis results. Minor stomach upset in about 3%.
Shown to help 1 · Mixed results 1.
Dietary nitrate
Longer time to exhaustion and lower blood pressure in non-athletes. No pooled effect in trained athletes. Product nitrate varies about 50-fold.
Shown to help 5 · Not shown to help 1.
Digestive enzymes
Prescription enzymes raised fat absorption in pancreatic insufficiency. Over-the-counter enzyme trials were small. High pancreatic enzyme doses were linked to colon strictures.
Shown to help 4 · Mixed results 1 · Not shown to help 1.
DIM
Shifted estrogen metabolites in small trials. Lowered tamoxifen's active metabolites. No difference from placebo for cervical dysplasia.
Shown to help 1 · Not shown to help 4.
Diosmin and hesperidin
Less leg heaviness, swelling and hemorrhoid bleeding in trials. No arm volume change in one lymphedema trial. Mild digestive side effects.
Shown to help 2 · Mixed results 2 · Not shown to help 1.
Echinacea
Somewhat fewer colds in prevention trials. Adult colds lasted as long as on placebo. Children had slightly shorter illness and more mild side effects.
Mixed results 2 · Not shown to help 1.
Elastin peptides
Fewer fine wrinkles and more skin moisture than placebo in one 100-person trial. One other trial. No product-related adverse events.
Shown to help 4.
Elderberry
Shorter colds in small, uncertain trials. A larger flu trial found no change. Juice made with leaves and stems caused poisoning.
Shown to help 1 · Mixed results 1 · Not shown to help 1.
Electrolytes
Carbohydrate-electrolyte drinks rehydrate better than water. Cramps tracked prior cramping and pace, not electrolytes. Sodium did not prevent hyponatremia.
Shown to help 2 · Mixed results 2.
Epicatechin
Strength gains in one sarcopenia trial. Blunted aerobic training gains in another. No change in blood pressure or vessel function.
Shown to help 1 · Not shown to help 3.
Evening primrose oil
No difference from placebo for eczema, breast pain, PMS or labor. Side effects were mild and digestive, similar to placebo.
Shown to help 1 · Mixed results 1 · Not shown to help 4.
Fat burners
Thermogenic weight changes cross zero in pooled trials. Ephedra added 0.9 kg a month with more side effects. FDA banned it in 2004.
Shown to help 3 · Not shown to help 1.
Fenugreek
Lower blood sugar in diabetes, from low-quality trials. Small, very uncertain testosterone effect. Warfarin interaction and allergy cases reported.
Shown to help 5.
Fiber supplements
Psyllium lowers LDL cholesterol and helps chronic constipation. Soluble fiber eases IBS, bran did not. More gas. Rare choking and allergy reports.
Shown to help 4 · Mixed results 1.
Fisetin and quercetin (senolytics)
Human senolytic trials paired quercetin with a cancer drug. Small, mixed results. Quercetin alone lowered blood pressure slightly. No lifespan data.
Shown to help 5 · Mixed results 1 · Claimed, not studied in people 1.
Fish oil
Triglycerides fall about 15%. Little change in death or heart events overall. More atrial fibrillation in pooled trials, highest above 1 g a day.
Shown to help 9 · Mixed results 6 · Not shown to help 4 · Claimed, not studied in people 1.
Folate
Folic acid around conception cut neural tube defects sharply. Stroke fell 10% in pooled trials. Cognition did not change.
Shown to help 7 · Not shown to help 1.
GABA
Limited evidence for stress, very limited for sleep. Most trials had industry authors. Some studies found modest blood pressure drops.
Shown to help 1 · Mixed results 3.
Garcinia cambogia
Weight loss about 1 kg more than placebo in pooled trials. One large trial found none. Liver injury cases include a death and transplants.
Shown to help 1 · Mixed results 1 · Not shown to help 1.
Garlic
Lower blood pressure and cholesterol in pooled trials. One trial on colds. Halved an HIV drug's levels in a small study.
Shown to help 5.
Ginger
Less pregnancy and postoperative nausea in pooled trials. Small pain changes. Twice the dropout in arthritis trials.
Shown to help 8 · Mixed results 2.
Ginkgo
Mixed results in dementia. No change in healthy memory, tinnitus or dementia onset. Bleeding case reports, with no clotting changes in trials.
Mixed results 1 · Not shown to help 5.
Ginseng
Small drop in fatigue. Mixed glucose results. Small cognition gains in impairment. Reduced warfarin's effect in a trial. Product content varied widely.
Shown to help 4 · Mixed results 1 · Not shown to help 1.
Glucomannan
LDL about 10% lower. Weight results disagree between meta-analyses. Tablets linked to esophageal blockage. Konjac jellies to choking deaths.
Shown to help 3 · Mixed results 1 · Not shown to help 1.
Glucosamine and chondroitin
Pain differences from placebo fall below what patients notice in large independent trials. Adverse events match placebo. Raised INR reported with warfarin.
Shown to help 2 · Mixed results 2 · Not shown to help 1.
Glutamine
No pooled effect on athletic performance, lean mass, immunity or gut permeability. Higher mortality in one critical-care trial.
Shown to help 1 · Not shown to help 3.
Glutathione
Oral trials disagree on raising body stores. Skin melanin fell in small trials. FDA reported reactions to contaminated IV glutathione.
Shown to help 2 · Mixed results 3 · Not shown to help 1.
Glycine
Three small sleep trials of 3 g report gains. Mixed schizophrenia results. Worsened symptoms when added to clozapine.
Shown to help 1 · Mixed results 2.
GlyNAC
One research group reports wide benefits in small trials. A larger, shorter trial found no glutathione rise. No lifespan data.
Shown to help 3 · Mixed results 1 · Claimed, not studied in people 1.
Grape seed extract
Small blood pressure reductions across meta-analyses. Lower fasting glucose in pooled trials. Small trials on leg swelling and veins.
Shown to help 3 · Mixed results 3.
Green tea extract
Weight change near zero outside Japan. Small LDL and blood pressure drops. High-dose EGCG raised liver enzymes. Liver injury cases include transplants.
Shown to help 5 · Mixed results 1.
Green-lipped mussel
Moderately lower osteoarthritis pain in pooled small trials. Small asthma trials only. Case reports of hepatitis and allergy.
Shown to help 2 · Mixed results 1 · Not shown to help 1.
Greens powders
AG1 raised blood nutrient levels in maker-funded trials. Few metabolic changes. Drug residues and toxins found in algae ingredients.
Shown to help 3 · Mixed results 1 · Not shown to help 1.
HMB
No strength or muscle gain in trained or young lifters. Small gains in older adults and patients. Lower muscle damage markers.
Shown to help 1 · Mixed results 1 · Not shown to help 2.
Horny goat weed
No sexual function trials located. Bone density held steady in postmenopausal women. Case reports of rapid heart rhythm and psychosis.
Shown to help 1 · Claimed, not studied in people 3.
Horse chestnut
Less leg swelling and pain than placebo in venous insufficiency trials. Rare case reports of liver injury and kidney failure.
Shown to help 3.
Huperzine A
Higher Alzheimer's scores in low-quality trials. A US trial found no change at its main dose. Side effects were mostly cholinergic.
Mixed results 2 · Not shown to help 3.
Hyaluronic acid (oral)
Pooled trials found more skin hydration and shallower wrinkles. Small trials found less knee pain. Many were run by one manufacturer.
Shown to help 5.
Inositol
PCOS cycles and ovulation improved, from evidence a guideline review called limited. Gestational diabetes fell in pooled trials. Preterm infant deaths rose.
Shown to help 4 · Mixed results 1 · Not shown to help 1.
Iodine
Iodine improved cognition in deficient children. In mildly deficient pregnancies, child development did not change. Excess intake went with more thyroid disease.
Shown to help 2 · Mixed results 1 · Not shown to help 1 · Claimed, not studied in people 1.
Iron
Less fatigue in iron-deficient adults. VO2 max findings disagree. Gut side effects doubled on ferrous sulfate. Overdose is a child poisoning risk.
Shown to help 5 · Mixed results 2.
Kava
Lower anxiety scores in short trials. A 16-week trial found no difference. Linked to liver failure and transplants, prompting an FDA advisory.
Mixed results 3.
Keratin hydrolysate
Better skin, hair and nail measures than placebo in three small seller-linked trials. No effect sizes reported. Adverse events not counted.
Shown to help 3.
Ketone esters and ketone salts
No change in exercise performance or recovery in pooled trials. Blood glucose falls. Gut discomfort slowed one elite cycling trial.
Shown to help 3 · Not shown to help 3.
Kratom
Grey marketRaised cold-pain tolerance in daily users in one small trial. FDA says it is not a lawful supplement. Linked to overdose deaths and liver injury.
Shown to help 3 · Claimed, not studied in people 1.
L-arginine
Systolic blood pressure about 4 to 6 mmHg lower. Erectile function improved in mild cases. Deaths rose in a heart attack trial.
Shown to help 4 · Mixed results 3.
L-carnitine
About 1.2 kg extra weight loss in pooled trials, less soreness, and a tenfold rise in TMAO, a marker linked to heart disease.
Shown to help 2 · Mixed results 1.
L-Theanine
Faster reaction time after 200 mg. Small gains in self-rated sleep. No anxiety effect in a GAD trial. No serious adverse events reported.
Shown to help 3 · Mixed results 1.
Lactoferrin
Fewer infections in preterm babies in pooled trials, but not in the largest trial. Fewer children with diarrhea. Raised iron levels.
Shown to help 2 · Mixed results 3 · Not shown to help 1.
Licorice root
Glycyrrhizin raised blood pressure and lowered potassium in pooled trials. Lower liver enzymes in liver disease. Small DGL trials reported less indigestion.
Shown to help 3 · Not shown to help 1 · Claimed, not studied in people 1.
Lion's Mane
Higher cognitive scores in small trials of older adults. No overall change in young adults. Stomach upset and rash in one trial.
Shown to help 2 · Not shown to help 1 · Claimed, not studied in people 1.
Lutein and zeaxanthin
Adding it to the AREDS formula did not lower AMD progression in the main analysis. Macular pigment rose. Cataract surgery did not differ.
Shown to help 3 · Mixed results 1 · Not shown to help 3.
Maca
Small trials report better sexual function scores. No change in sperm count or hormones. AST and blood pressure rose in one trial.
Shown to help 3 · Mixed results 1 · Not shown to help 1.
Magnesium
Blood pressure falls about 2 mmHg. Leg cramps in older adults did not change. Sleep trials are small and inconsistent.
Shown to help 2 · Mixed results 2 · Not shown to help 3.
MCT oil
About 0.5 to 0.7 kg less body weight than other fats. Small cognitive gains in Alzheimer's trials. Gut cramping and slower performance during exercise.
Shown to help 4 · Mixed results 1 · Not shown to help 1.
Melatonin
Falls asleep about 7 minutes sooner. Less jet lag after long eastward flights. Labels often wrong. Child ingestions rose 530% in ten years.
Shown to help 5.
Milk thistle
No significant effect on deaths in alcohol-related or viral liver disease. Lower liver enzymes in fatty liver trials, at low certainty.
Shown to help 2 · Mixed results 1 · Not shown to help 1 · Claimed, not studied in people 1.
MSM
Knee pain and function improved in three small trials. Exercise damage markers did not change. Long-term safety is unstudied.
Shown to help 3 · Not shown to help 1.
Mullein
No lung or cough trials located. Mixed herbal ear drops with mullein matched anesthetic drops for ear pain in children.
Mixed results 1 · Claimed, not studied in people 1.
Multivitamin
No change in heart events or death. Small, inconsistent cancer findings. Small memory gains in older adults. Beta carotene raised lung cancer in smokers.
Shown to help 1 · Mixed results 3 · Not shown to help 4.
N-acetylcysteine
No change in sport performance. COPD exacerbations fell 13% to 25% in two meta-analyses. Headache and low blood pressure when combined with nitroglycerin.
Shown to help 4 · Not shown to help 1.
NAD+ infusions
No wellness outcome trials located. Small heart trials used low doses. Infusion symptoms were common in one clinic review.
Shown to help 2 · Not shown to help 1 · Claimed, not studied in people 2.
Nano-hydroxyapatite
Matched fluoride toothpaste for new cavities in maker-funded trials. Independent review called it inconclusive. Less tooth sensitivity in reviews.
Shown to help 2 · Mixed results 1.
Niacin
Niacin raised HDL, and heart events did not fall in two large statin trials. New diabetes and liver toxicity rose.
Shown to help 3 · Not shown to help 1.
Nicotinamide riboside
Raises blood NAD. No gain in insulin sensitivity or muscle. Parkinson symptoms worsened more than placebo in a year-long trial.
Shown to help 2 · Not shown to help 5 · Claimed, not studied in people 1.
Nicotine pouches
Small attention gains, little exercise effect, higher heart rate, and high self-reported addiction. Snus use is linked to higher mortality.
Shown to help 1 · Not shown to help 2.
NMN
Raises blood NAD in trials. Little change in muscle or metabolic markers. No trial measured lifespan. Adverse events matched placebo.
Shown to help 3 · Mixed results 1 · Not shown to help 1 · Claimed, not studied in people 1.
Oral ceramides
Skin hydration rose and water loss fell in small trials. Wrinkle gains faded after stopping. No product-related adverse events reported.
Shown to help 4.
Oral probiotics
Small short-term gum gains with L. reuteri after deep cleaning. No pooled change in gingivitis. Breath odor results disagree.
Shown to help 5 · Mixed results 3 · Not shown to help 1.
Oregano oil
No controlled trial of oregano oil by mouth was located. Small studies of sprays and ointments. Allergic reactions have been reported.
Shown to help 4 · Claimed, not studied in people 1.
Palmitoylethanolamide
Lower chronic pain scores in two meta-analyses. Nerve pain and knee arthritis results split. No major side effects were attributed to it.
Shown to help 2 · Mixed results 2.
Paraxanthine and theacrine
Some cognitive tests improved in small supplier-funded paraxanthine trials. Theacrine raised felt energy, not strength, test scores or time-trial speed.
Shown to help 2 · Mixed results 1 · Not shown to help 2.
Phenibut
Grey marketSmall Russian trials reported less anxiety, none against placebo. FDA says it is not a lawful supplement ingredient. Linked to coma and severe withdrawal.
Shown to help 3 · Claimed, not studied in people 2.
Phosphatidylserine
Memory gains came mostly from cow-brain PS. A soy PS trial found no change. Trials up to 30 weeks found no safety test differences.
Shown to help 1 · Mixed results 2 · Not shown to help 1.
Potassium
Potassium and salt substitutes lowered blood pressure. A large trial found fewer strokes and deaths. Trials excluded people at risk of high potassium.
Shown to help 3 · Mixed results 1 · Not shown to help 1.
PQQ
A few small trials, some run by makers. Memory scores rose in one. No aerobic performance change. No lifespan data.
Shown to help 1 · Mixed results 1 · Not shown to help 2 · Claimed, not studied in people 1.
Pre-workout blends
Small rep and energy gains in single-product trials. Banned and untested stimulants found in tested products. FDA has acted on DMAA and DMHA.
Shown to help 2 · Mixed results 1.
Prebiotics
More bifidobacteria and slightly more frequent stools. No change in IBS symptoms. Inulin worsened flatulence in IBS trials.
Shown to help 4 · Mixed results 1 · Not shown to help 3.
Probiotics
Less diarrhea with antibiotics and fewer H. pylori treatment side effects. Probiotic strains have caused bloodstream infections in ICU patients and preterm infants.
Shown to help 10 · Mixed results 5 · Not shown to help 3.
Protein powder
Modest added muscle and strength during training, a plateau near 1.6 g per kg a day, and measurable heavy metals in tested products.
Shown to help 5 · Not shown to help 1.
Pumpkin seed
Extract capsules matched placebo on prostate symptoms in a large trial. One small trial reports more hair growth than placebo.
Shown to help 2 · Not shown to help 1.
Pycnogenol
Small trials report less leg swelling on flights and in venous insufficiency. Blood pressure reviews disagree. Cochrane drew no firm conclusions.
Shown to help 5 · Mixed results 2 · Not shown to help 1.
Pygeum
Short, older trials report fewer urinary symptoms than placebo in enlarged prostate. Side effects were mild and similar to placebo.
Shown to help 1 · Claimed, not studied in people 1.
Raspberry ketones
No trial of raspberry ketone alone. One company-sponsored blend trial. Rat-only toxicity data. Heart rhythm and vasospasm case reports.
Mixed results 1 · Claimed, not studied in people 2.
Red vine leaf
Smaller leg volume than placebo in three trials of AS 195. Most symptoms changed in only some trials.
Shown to help 2.
Reishi
Cancer and immune data come from small, low-quality trials. No change in diabetes risk factors. Case reports of liver injury.
Shown to help 1 · Mixed results 1 · Not shown to help 2 · Claimed, not studied in people 1.
Resveratrol
Mixed glucose results in diabetes. No overall blood pressure or sirtuin change. Not linked to mortality in one cohort. Blunted training gains in older men.
Mixed results 3 · Not shown to help 2.
Rhodiola
Mixed fatigue and endurance results from small trials. No difference from placebo in depression. Many products adulterated with other species.
Shown to help 1 · Mixed results 3 · Not shown to help 1.
Rose hip
Less osteoarthritis pain in three manufacturer-supported trials, no difference in a later one. Small rheumatoid arthritis trial. Few side effects reported.
Shown to help 2.
SAM-e
No clear difference from placebo for depression alone. Add-on trials disagree. Mania, serotonin-like symptoms and stomach upset were reported.
Mixed results 1 · Not shown to help 3.
Saw palmetto
No difference from placebo on prostate symptoms in a high-certainty review. Hair loss data are thin. Side effects matched placebo.
Shown to help 1 · Not shown to help 3.
Schisandra
Greater leg strength in two small 12-week trials. Lower menopausal symptom scores in one. A related species raised levels of two drugs.
Shown to help 4 · Claimed, not studied in people 1.
Sea moss
Benefit research was not sourced here. US products held arsenic, mercury and bacteria. A thyroid case followed use.
Benefits: no research sourced yet. Risks documented: 4.
Selenium
Cancer and heart events did not change. Thyroid antibodies fell in Hashimoto's thyroiditis. One trial found more type 2 diabetes.
Shown to help 2 · Mixed results 1 · Not shown to help 2.
Serrapeptase
Mixed results for swelling after dental surgery. No pain relief over placebo in dental trials. Case reports of drug-induced pneumonitis.
Shown to help 3 · Mixed results 1.
Shilajit
Small, often maker-funded trials on testosterone, strength and bone. Some samples tested above limits for lead, arsenic and mercury.
Shown to help 4.
Silicon
Less skin roughness and hair breakage than placebo in small maker-linked trials. Spine bone density did not change. No product-related adverse events.
Shown to help 3 · Not shown to help 1.
Sodium bicarbonate
Small gains in short hard efforts and reps. No clear change in running or 200 m swim times. Diarrhea and nausea are common.
Shown to help 3 · Not shown to help 4.
Spermidine
No memory change in a year-long trial. Supplements did not raise blood spermidine. Diet cohorts link intake to mortality, unevenly.
Shown to help 3 · Not shown to help 2.
St. John's Wort
Matched antidepressants in pooled trials. Lowers levels of birth control, cyclosporine, HIV drugs and warfarin. Linked to transplant rejection.
Shown to help 2 · Mixed results 1.
Stinging nettle
Nettle root trials report fewer prostate symptoms, one from a single author with retractions. Leaf trials for allergies, joints and blood sugar are small.
Shown to help 3 · Mixed results 1.
Sulforaphane
Raised excretion of an air pollutant in two trials. Small glucose drop in prediabetes. Lower autism symptom scores in pooled trials.
Shown to help 4 · Not shown to help 1 · Claimed, not studied in people 1.
Tart cherry
Strength recovery and endurance gains in pooled trials. No change in blood pressure. Small, uneven uric acid findings. Sleep findings split.
Shown to help 2 · Mixed results 2 · Not shown to help 2.
Taurine
Small endurance gains and blood pressure about 4 mmHg lower in pooled trials. No cognitive change. Few adverse effects reported.
Shown to help 3 · Not shown to help 2 · Claimed, not studied in people 1.
Thiamine
No change in heart failure. Septic shock reviews split on mortality. Small benfotiamine trials disagree on nerve damage. Rare anaphylaxis after IV doses.
Mixed results 4 · Not shown to help 1.
Thunder god vine
CautionFewer swollen joints in rheumatoid arthritis and less urine protein in kidney disease trials. About one in four users had side effects, including irregular periods.
Shown to help 4.
TMG
Lowers homocysteine. Raises total and LDL cholesterol at higher doses. Body fat findings disagree. Small strength gains.
Shown to help 3 · Mixed results 1 · Claimed, not studied in people 1.
Tongkat ali
Testosterone rose in small, mostly maker-linked trials. Erectile function unchanged overall. EFSA flagged DNA damage in lab tests. One liver injury case.
Shown to help 3 · Not shown to help 1.
Tribulus
No clear testosterone rise from tribulus alone. Pooled erectile trials show higher IIEF scores than placebo. Case reports of kidney and liver injury and priapism.
Shown to help 1 · Mixed results 1 · Not shown to help 2.
TUDCA
Small trials only. Matched ursodiol in one liver trial, raised insulin sensitivity in 20 adults, and slowed ALS decline in a 34-person pilot.
Shown to help 4 · Mixed results 1 · Claimed, not studied in people 1.
Turkesterone and ecdysterone
No turkesterone trial. Ecdysterone trials disagree on muscle gain. Tested turkesterone products held little or none of the label.
Mixed results 1 · Not shown to help 2.
Turmeric
Less knee arthritis pain and modest drops in blood sugar, lipids and CRP in heterogeneous trials. Liver injury case series include a death.
Shown to help 12 · Mixed results 3 · Not shown to help 1 · Claimed, not studied in people 1.
Tyrosine
Working memory held up under cold and multitasking in small trials. Exercise performance mostly unchanged. Older adults did worse at higher doses.
Shown to help 1 · Mixed results 1 · Not shown to help 2.
Urolithin A
Small, mostly maker-funded trials. Walking distance did not differ in pooled data. Some strength and endurance gains. No lifespan data.
Shown to help 1 · Mixed results 2 · Not shown to help 2 · Claimed, not studied in people 1.
Uva ursi
No symptom relief against placebo in acute UTI. More symptoms and kidney infections than an antibiotic. Converted to hydroquinone in the body.
Not shown to help 1.
Valerian
More people report better sleep. Measured sleep onset did not change. Rare adverse events. A small rise in alprazolam levels.
Shown to help 2 · Not shown to help 2.
Vitamin A
Child deaths fell in large trials in Asia and Africa. Beta-carotene raised lung cancer in smokers. High early-pregnancy intake was linked to birth defects.
Shown to help 1 · Mixed results 1 · Not shown to help 3.
Vitamin B12
Oral B12 corrected low blood levels about as well as injections. In people without deficiency, cognition and mood did not change.
Shown to help 4 · Mixed results 1 · Not shown to help 4.
Vitamin B6
Nausea in pregnancy eased in pooled trials. Heart attacks did not change. High intakes cause sensory nerve damage.
Shown to help 4 · Mixed results 1 · Not shown to help 1.
Vitamin C
Colds were slightly shorter, not fewer, except under heavy physical stress. Heart events and cancer did not change. High-dose men had more kidney stones.
Shown to help 4 · Mixed results 3 · Not shown to help 5.
Vitamin D
No change in fractures, falls, cancer or heart events in unselected adults. Respiratory infection effect lost significance in 2025. Annual mega-doses raised falls.
Shown to help 8 · Mixed results 5 · Not shown to help 8.
Vitamin E
Heart events and cancer did not change in large trials. Prostate cancer rose in SELECT. Pooled fatty liver trials found lower liver enzymes.
Shown to help 2 · Not shown to help 3.
Vitamin K2
Fracture findings weakened as trials were updated. Calcification results are mixed. Low doses changed blood thinner control in volunteers.
Mixed results 3 · Not shown to help 1.
Vitex (chasteberry)
Eased premenstrual symptoms and cyclic breast pain against placebo. No change in menstrual blood loss. Fibroid use is untested in people.
Shown to help 3 · Not shown to help 1 · Claimed, not studied in people 1.
White kidney bean extract
Weight loss crossed zero in one independent meta-analysis and reached 1.62 kg in another from 2026. Undercooked bean powder made over 1,000 ill.
Shown to help 3 · Mixed results 1.
White willow bark
Short-term back pain relief over placebo. Arthritis results conflict. Weaker platelet effect than aspirin, and one anaphylaxis case with aspirin allergy.
Shown to help 1 · Mixed results 1 · Not shown to help 1.
Xylitol
Gum lowered children's cavities in reviews. Adult lozenges did not. Large doses loosen stools, and higher blood xylitol tracked heart events.
Shown to help 2 · Mixed results 2 · Not shown to help 1.
Zinc
Colds shortened when zinc was started at onset. No change in catching colds. More nausea and taste effects. Excess has caused copper deficiency.
Shown to help 4 · Mixed results 2 · Not shown to help 2.
Therapies
Acupuncture
Eased chronic pain more than no treatment and slightly more than sham needling. Minor bruising or pain is common, serious harm rare.
Shown to help 12 · Mixed results 4 · Not shown to help 2.
Biofeedback and neurofeedback
Fewer headaches than waiting list. ADHD gains vanished under blinded ratings and sham control. Helped some pelvic floor problems.
Shown to help 8 · Mixed results 2 · Not shown to help 2.
Chiropractic and spinal manipulation
Back pain relief similar to other treatments. No difference from sham for migraine or non-spine conditions. Stroke after neck manipulation is debated.
Shown to help 5 · Mixed results 3 · Not shown to help 3 · Claimed, not studied in people 1.
Cognitive behavioral therapy
Less depression, anxiety and insomnia than with waiting list or placebo, similar to antidepressants, and small effects on chronic pain.
Shown to help 7 · Mixed results 3 · Not shown to help 1.
Colon hydrotherapy
No controlled trials for detox or general health. Perforation, infection outbreaks and coffee enema deaths in case reports.
Shown to help 1 · Mixed results 1 · Not shown to help 1 · Claimed, not studied in people 2.
Cupping
Less chronic pain than no treatment, no difference from sham cupping. Detox claims untested in people. Burns and anemia in case reports.
Shown to help 2 · Mixed results 3 · Not shown to help 2 · Claimed, not studied in people 1.
Dry needling
Eased pain more than mixed comparison groups, not more than sham in well-blinded trials. Bruising common, collapsed lung rare.
Shown to help 6 · Mixed results 3 · Not shown to help 1.
Endometriosis surgery
Eased pain more than sham surgery in small blinded trials and raised pregnancy in mild disease. Endometriosis often returns within 5 years.
Shown to help 5 · Not shown to help 1.
Homeopathy
No condition with reliable evidence beyond placebo in national reviews. FDA actions on belladonna teething tablets and zinc nasal products.
Shown to help 1 · Mixed results 3 · Not shown to help 4 · Claimed, not studied in people 1.
Hypnotherapy and clinical hypnosis
IBS relief beyond education or routine care, similar to CBT. No quit-smoking edge over other support. Adverse events rare.
Shown to help 5 · Mixed results 2 · Not shown to help 2.
IV vitamin therapy
Less fatigue for a day in one trial. No difference from a saline drip in fibromyalgia. Infections and hemolysis in case reports.
Shown to help 1 · Mixed results 1 · Not shown to help 2 · Claimed, not studied in people 2.
Lymphatic drainage massage
No added benefit over placebo drainage for breast cancer lymphedema in the largest trial. Detox and fat loss untested in people.
Shown to help 2 · Mixed results 2 · Not shown to help 2 · Claimed, not studied in people 1.
Massage therapy
Less pain in the short term and less soreness after exercise in pooled trials. Smaller edge over sham. Serious harms appear in case reports.
Shown to help 7 · Mixed results 4.
Osteopathy and osteopathic manipulative treatment
Eased low back pain against mixed comparisons. No difference from sham in a 2024 review. Brief soreness after some visits.
Shown to help 3 · Mixed results 5 · Not shown to help 1.
Ozone therapy
Less back and knee pain in small sham trials of low quality. FDA calls ozone a toxic gas. Stroke and spinal infection in case reports.
Shown to help 3 · Mixed results 3 · Not shown to help 1 · Claimed, not studied in people 1.
Physical therapy and exercise therapy
Less back, knee and hip pain than no exercise. No difference from sham in two trials. Minor flare-ups are more common, serious events are not.
Shown to help 5 · Mixed results 2 · Not shown to help 1.
PRP injections
No difference from saline for tennis elbow or Achilles pain. Mixed for knee arthritis. More hair density in pattern hair loss.
Shown to help 3 · Mixed results 3 · Not shown to help 2.
Reflexology
Less anxiety and better sleep than usual care. No difference from plain foot massage or sham reflexology for pain or mood.
Shown to help 5 · Not shown to help 3.
Reiki and therapeutic touch
Less anxiety than no treatment. Mixed against sham Reiki from untrained people. Practitioners could not detect the energy field.
Shown to help 2 · Mixed results 3 · Not shown to help 1.
Uterine fibroid procedures
Embolization matched surgery on satisfaction but needed more repeat procedures. Hysterectomy cut bleeding most. FDA warns on power morcellation.
Shown to help 6 · Mixed results 3 · Not shown to help 1.
Practices
Cold water immersion
Less muscle soreness after exercise in pooled trials. Smaller strength and muscle gains when used after lifting. Cold shock and hypothermia are documented harms.
Shown to help 6 · Mixed results 5 · Not shown to help 5.
Grounding
Small studies, many funded by a grounding product company, reported better sleep and mood. Inflammation claims rest on blood counts in a few dozen people.
Shown to help 4 · Mixed results 4 · Claimed, not studied in people 1.
Heat therapy
Eased period pain against no treatment and a sham patch, about as much as NSAIDs. Burns and heat rashes are documented harms.
Shown to help 7 · Claimed, not studied in people 1.
Hyperbaric oxygen therapy
Small trials from one research group report cognitive gains. Sham-controlled concussion trials found none. Ear injury is common. FDA warns about unproven uses and fires.
Shown to help 7 · Mixed results 6 · Not shown to help 4.
Infrared sauna
Short-term heart function gains in heart failure trials from Japan. Small recovery studies in athletes. Depression results are mixed. Little data in healthy adults.
Shown to help 11 · Mixed results 5 · Not shown to help 7.
Meditation
Less stress, anxiety and pain than no program in pooled trials, with smaller or no gains over other active programs. Unpleasant experiences are reported.
Shown to help 10 · Mixed results 2 · Not shown to help 3.
Mouth taping
Small uncontrolled studies found less snoring in mild sleep apnea, others found no change. Mouth closure cut airflow in some heavy mouth breathers.
Shown to help 3 · Mixed results 1 · Not shown to help 1 · Claimed, not studied in people 1.
Red light therapy
Higher hair density and less knee and neck pain in pooled sham-controlled trials. Mixed results for performance, acne and mood. Few adverse events in trials.
Shown to help 12 · Mixed results 6 · Not shown to help 1 · Claimed, not studied in people 1.
Sauna
Finnish cohorts link frequent sauna use to fewer heart deaths, observationally. Trials report small blood pressure falls. Harms include burns, fainting and reversible sperm changes.
Shown to help 12 · Mixed results 8 · Not shown to help 4.
Slow breathing and breathwork
Small falls in stress and anxiety in pooled trials. Blood pressure fell in open trials, not against sham devices. Hyperventilating before underwater breath-holds preceded drownings.
Shown to help 8 · Mixed results 3 · Not shown to help 1.
Tai chi and qigong
Fewer falls in older adults, knee arthritis relief similar to physical therapy, and lower blood pressure in trials. Minor aches were the reported harms.
Shown to help 9 · Not shown to help 2.
TENS
Eased period pain and other pain against sham in pooled trials, with mixed labor results. Skin irritation is common, serious harm not reported.
Shown to help 4 · Mixed results 4 · Not shown to help 1.
Whole-body cryotherapy
Insufficient evidence for muscle soreness. No better than placebo in two recovery trials. Small trials report mood and pain gains. FDA has cleared no device.
Shown to help 9 · Mixed results 6 · Not shown to help 1 · Not yet researched 1.
Yoga
Small gains for back pain, mood, anxiety and balance in pooled trials, similar to other exercise. Sprains and strains are the most reported harms.
Shown to help 10 · Mixed results 1 · Not shown to help 1.
Daily life
Daily steps
Fewer deaths and less heart disease with more steps in pooled cohorts, leveling off well below 10,000 in older adults. Programs add about 2,000 steps.
Shown to help 7 · Mixed results 4.
Gardening
Gardening reduced depressive symptoms and raised wellbeing in pooled trials. A garden plot slightly raised fiber and activity in one trial.
Shown to help 5 · Mixed results 2.
Light at night
More light at night linked to poorer sleep, obesity, depression and diabetes. Eye masks improved hospital sleep, blue-blocking glasses gave inconsistent results.
Shown to help 1 · Mixed results 1.
Napping
Naps improved alertness and memory in pooled trials. Naps of an hour or more were linked to more deaths, heart disease and diabetes in cohorts.
Shown to help 4 · Not shown to help 1.
Pet ownership
Dog and pet ownership linked to fewer cardiovascular deaths and more walking. No link with depression, mixed results for childhood asthma.
Shown to help 3 · Mixed results 3 · Not shown to help 1.
Physical activity
About 30% fewer deaths at 150 minutes a week in cohorts, flattening above that. Trials: lower blood pressure, HbA1c and depression scores.
Shown to help 9 · Mixed results 2.
Screen time
Bedtime devices linked to shorter sleep, and screens to myopia, in children. Cutting screens raised children's activity, not weight change. Social media breaks: mixed.
Shown to help 2 · Mixed results 1 · Not shown to help 2.
Shift work
Linked to more heart attacks, diabetes and obesity in pooled studies. A quarter of shift workers had shift work disorder. IARC: probably carcinogenic.
Shown to help 2 · Mixed results 1.
Sitting time
More deaths, heart disease and diabetes with long sitting in cohorts. An hour of daily activity offset sitting, not TV. Breaks lowered blood sugar.
Shown to help 9.
Sleep duration
Lowest death and disease rates near 7 hours in cohorts. In trials, sleep loss raised eating and cut attention, mood and insulin sensitivity.
Shown to help 9 · Mixed results 1.
Sleep timing
Late and irregular sleep linked to depression, weight and cardiometabolic risk in pooled studies. Later school starts added teen sleep.
Shown to help 2.
Smoking
Smoking linked to about triple the death rate. Quitting before 40 avoided most of the excess in cohorts, and gains appeared at every age.
Shown to help 5 · Mixed results 1.
Social connection
Isolation and loneliness linked to more deaths, heart disease, stroke and dementia in pooled cohorts. Loneliness programs had modest effects in trials.
Shown to help 6.
Strength training
Fewer deaths in cohorts, lowest near 30 to 60 minutes a week. Trials: more strength, lower HbA1c, blood pressure and depression scores.
Shown to help 8 · Mixed results 1.
Sunlight
Raised vitamin D in trials and cut new myopia in children. Sunburn is very common and linked to melanoma and other skin cancers.
Shown to help 6.
Time in nature
Lower blood pressure, cortisol and negative mood after time in nature in small pooled studies. Greener surroundings are linked to fewer deaths.
Shown to help 7.
Vaping
More quitters than with nicotine replacement in trials. Linked to asthma, COPD and later smoking in young people. Long-term effects largely unknown.
Shown to help 6.
Volunteering and purpose
Older volunteers and people with more purpose in life had fewer deaths in pooled cohorts. Mood and memory results were smaller in trials.
Shown to help 6 · Mixed results 3.
Food and drink
Added sugar
More sugar raised weight, blood fats and blood pressure in trials. Linked to diabetes, heart disease and tooth decay. No mood lift.
Shown to help 1 · Not shown to help 2.
Alcohol
Light drinking's lower death rate faded after correcting for former drinkers. Linked to several cancers and cirrhosis. Interacts with opioids, sedatives and warfarin.
Shown to help 1 · Mixed results 3.
Coffee
Linked to longer life, less diabetes and less cirrhosis in cohorts. Unfiltered coffee raised LDL in trials. Pregnancy loss and low birth weight links.
Shown to help 7 · Mixed results 1 · Not shown to help 1.
Dairy
Milk not linked to fewer hip fractures overall. No weight change without calorie cuts. Linked to acne and slightly more prostate cancer in cohorts.
Shown to help 3 · Mixed results 2 · Not shown to help 1.
Dark chocolate
Small blood pressure drops in pooled trials. An extract trial missed its main heart target. Lead and cadmium found in many products.
Shown to help 4 · Mixed results 1 · Not shown to help 1.
Eggs
Raised LDL cholesterol in feeding trials. Mostly not linked to heart disease at up to an egg a day. Diabetes links in US cohorts only.
Mixed results 4.
Energy drinks
Strength and endurance improved in pooled studies. Raised blood pressure about 4 to 5 mmHg. Insomnia and jitteriness common. Linked to short sleep.
Shown to help 2.
Fermented foods
Eased bowel symptoms and cut respiratory infections in pooled trials. Mixed results for inflammation and blood pressure. A label warning with MAOI antidepressants.
Shown to help 6 · Mixed results 3 · Not shown to help 1.
Fish and seafood
Linked to less heart disease, stroke and death in cohorts, mainly fatty fish. Mercury and fish or shellfish allergy are the documented risks.
Shown to help 6 · Mixed results 1.
Grapefruit
No significant weight loss in pooled trials. Raises blood levels of many prescription drugs, including some statins and blood pressure drugs, in controlled studies.
Shown to help 2 · Mixed results 1 · Not shown to help 1.
Green tea
Linked to lower death and stroke in Asian cohorts. Small cholesterol drop in drink trials. Inconsistent cancer results. Very hot tea linked to esophageal cancer.
Shown to help 6 · Mixed results 1 · Not shown to help 1.
Non-sugar sweeteners
About 1 kg weight loss when replacing sugary drinks in trials. Linked to diabetes, heart disease and cancer in cohorts. Aspartame classed possibly carcinogenic.
Shown to help 1 · Mixed results 2.
Nuts
Lower LDL cholesterol in pooled trials, with no weight change. Linked to less heart disease in cohorts. Peanut and tree nut allergy are documented risks.
Shown to help 6 · Not shown to help 4.
Olive oil
Linked to less heart disease, diabetes and death in cohorts. No blood pressure change in trials. Many retail extra virgin oils failed purity tests.
Shown to help 5 · Mixed results 3 · Not shown to help 1.
Red meat and processed meat
Processed meat classed as causing bowel cancer. Cohort links to diabetes and heart disease, rated low certainty. One trial found little difference from eating less.
Mixed results 2 · Not shown to help 1.
Salt
Less salt lowered blood pressure in pooled trials. A salt substitute cut strokes and deaths in one large trial. Cohorts disagree on low intakes.
Shown to help 3 · Mixed results 3 · Claimed, not studied in people 1.
Soy foods
Lowered LDL slightly in trials. No change in men's testosterone. Hot flash reviews disagree. Linked to less breast and prostate cancer in cohorts.
Shown to help 4 · Mixed results 1 · Not shown to help 1.
Whole grains
Linked to less heart disease, diabetes and death in cohorts. Small LDL fall in trials, no weight change. Celiac disease and wheat allergy are risks.
Shown to help 5 · Mixed results 2 · Not shown to help 1.
Diets
5:2 diet
Weight loss matches daily calorie restriction. More low blood sugar episodes on diabetes drugs. Some minerals and fiber ran low in diet records.
Shown to help 1 · Mixed results 1 · Not shown to help 1.
Alkaline diet
No bone benefit shown. No cancer trials. Fruits and vegetables slowed kidney decline in small trials from one group.
Shown to help 1 · Not shown to help 2.
Alternate-day fasting
Lowers weight and fat against usual eating. Slightly more loss than daily calorie restriction in short trials. More dropouts in a year-long trial.
Shown to help 5 · Mixed results 1 · Not shown to help 1.
Animal-based diet
We haven't sourced research on this yet.
Anti-inflammatory diet
Observational links to less heart disease, depression and cancer. A rheumatoid arthritis trial found no difference in its main analysis.
Shown to help 5 · Not shown to help 1.
Atkins diet
Weight loss close to other popular diets. Fewer seizures on the modified version. More intake shortfalls in thiamine, folate, vitamin C, iron and magnesium.
Shown to help 2 · Mixed results 1 · Not shown to help 1.
Blood type diet
Some patterns linked to lower weight and heart markers in any blood type. Matching the diet to a person's own type made no difference.
Shown to help 2 · Not shown to help 1 · Claimed, not studied in people 1.
Cabbage soup diet
We haven't sourced research on this yet.
Carnivore diet
No controlled trials. Followers in surveys report weight loss and better blood sugar. LDL cholesterol rose. Nutrient levels were not measured.
Shown to help 5.
DASH diet
Lowers blood pressure in trials, most with less sodium and in hypertension. Small LDL and weight drops. One trial found HDL fell.
Shown to help 5.
Dukan diet
No controlled trials. One interview study reports about 15 kg lost in 8 to 10 weeks, with low vitamin C and folate intake.
Shown to help 1.
Elimination diet
Remission in eosinophilic esophagitis. Slight or no eczema gain. Mixed ADHD results. Nutrient and growth shortfalls in children avoiding several foods.
Shown to help 3 · Mixed results 2.
Fasting-mimicking diet
Monthly 5-day cycles lowered weight, HbA1c and blood pressure in trials. Single trials in Crohn's disease and breast cancer. Few serious side effects.
Shown to help 6 · Not shown to help 1 · Claimed, not studied in people 1.
Flexitarian diet
Little direct research. Weight loss matched an omnivorous diet in one small trial. Cohort data link it to less diabetes.
Shown to help 3 · Mixed results 1 · Not shown to help 3.
Gluten-free diet
Few self-reported gluten-sensitive people react to blinded gluten. No performance change. Higher mercury and arsenic and several nutrient shortfalls.
Mixed results 2 · Not shown to help 3.
High-carbohydrate diet
Race walkers improved more than on keto in small trials. Mixed results elsewhere. More muscle glycogen. Very high intakes linked to mortality.
Shown to help 3 · Mixed results 1 · Not shown to help 1.
High-protein diet
Slightly more lean mass kept and fat lost while dieting. Filtration rose without injury markers in short trials. More digestive complaints.
Shown to help 4 · Mixed results 1.
Intermittent fasting
Weight loss matches daily calorie restriction. Fasting glucose and blood pressure dip slightly. Some lean mass is lost alongside fat.
Shown to help 3 · Not shown to help 2.
Juice cleanse
One small trial found weight loss matching a calorie-matched placebo juice. No toxin research. Case reports of oxalate kidney injury.
Shown to help 2 · Mixed results 2 · Not shown to help 1.
Keto diet
Slightly more weight loss than comparison diets, lower triglycerides, fewer seizures in children with epilepsy, and an LDL rise. Mineral shortfalls reported.
Shown to help 6 · Mixed results 1.
Lectin-free diet
We haven't sourced research on this yet.
Low-carb diet
About 1 kg more weight loss than balanced diets. Short-term HbA1c gains in type 2 diabetes. Lower thiamine status reported.
Shown to help 2 · Mixed results 1.
Low-fat diet
Small drops in weight and LDL. About 1 to 2 kg less loss than low-carb. No change in heart events in one large trial.
Shown to help 2 · Mixed results 2 · Not shown to help 3.
Low-FODMAP diet
IBS symptoms improve more than on habitual diet or standard advice. Bifidobacteria fall. Diet quality scores dropped in one trial.
Shown to help 4.
Macrobiotic diet
Lower blood sugar in small diabetes trials. Deficiencies, rickets and slower growth in children on macrobiotic diets.
Shown to help 1 · Not shown to help 1.
Mediterranean diet
Fewer heart events in large trials of high-risk adults. Weight loss similar to other diets. Mixed cognition results. No weight gain from the added fat.
Shown to help 4 · Mixed results 1.
MIND diet
A 3-year trial found no cognitive difference from a control diet. Cohorts link closer adherence to less dementia.
Shown to help 2 · Not shown to help 1.
Nordic diet
Lower LDL cholesterol and blood pressure and modest weight loss in trials. No change in fasting glucose or inflammation markers.
Shown to help 4 · Mixed results 1 · Not shown to help 1.
One meal a day
Small trials in lean adults only. Fat mass fell at equal calories. Blood pressure, LDL and hunger rose. Glucose results were mixed.
Shown to help 1 · Mixed results 1 · Not shown to help 1.
Ornish diet
Less coronary narrowing in small trials of the full lifestyle program. Weight loss matched other popular diets. High dropout in one trial.
Shown to help 5.
Paleo diet
More weight loss and lower blood pressure and lipids than comparison diets. Mixed blood sugar results. Lower iodine and calcium intake.
Shown to help 5 · Mixed results 1.
Pescatarian diet
Cohorts link it to less heart disease, death and colorectal cancer. Weight loss matched an omnivorous diet in one small trial.
Shown to help 3 · Mixed results 1 · Not shown to help 1.
Pro-metabolic diet
We haven't sourced research on this yet.
Raw food diet
Surveys link it to weight loss and lower cholesterol, and to underweight, missed periods, B12 deficiency, low bone mass and dental erosion.
Shown to help 3 · Mixed results 1.
Sirtfood diet
No trials of the diet. One cohort linked higher sirtfood intake to fewer cardiovascular events. No weight loss research.
Shown to help 1 · Claimed, not studied in people 2.
South Beach diet
Small studies only. Weight and insulin fell in one uncontrolled study. No weight difference from low-fat after gastric bypass.
Shown to help 2 · Not yet researched 1.
Vegan diet
Lower weight and LDL in trials. Lower B12, iodine, calcium, vitamin D and zinc. More hip fractures in a UK cohort.
Shown to help 4 · Not shown to help 1.
Vegetarian diet
Lower weight, LDL and HbA1c in trials. Less heart disease but more stroke and hip fracture in cohorts. Lower B12, iron and zinc.
Shown to help 4 · Mixed results 1.
Very low calorie diet
Diabetes remission in about half of trial participants at one year, fading over five. More weight loss than usual care. Higher gallstone risk.
Shown to help 4.
Whole30
We haven't sourced research on this yet.
Zone diet
Weight loss similar to or below other popular diets. Lower triglycerides than a high-carbohydrate diet in one trial. No micronutrient shortfalls found.
Shown to help 2 · Mixed results 1 · Not shown to help 1.
Devices and wearables
Continuous glucose monitors in people without diabetes
Sensor readings differed from blood glucose by 7 to 19% in people without diabetes. Pooled trials found no change in body mass index.
Matched the reference method 4 · Mixed results 3 · Did not match the reference method 3 · Mixed results 3.
Heart rate variability wearables
Chest straps matched ECG at rest. Wrist and ring sensors differed by 6 to 31% depending on device, and more during movement.
Matched the reference method 5 · Mixed results 8 · Did not match the reference method 1 · Shown to help 1 · Not shown to help 1.
Sleep trackers
Trackers detect most sleep epochs and miss much quiet wakefulness. Sleep stage agreement with polysomnography varies by device. Case reports describe preoccupation with the data.
Matched the reference method 5 · Mixed results 7 · Did not match the reference method 7 · Shown to help 2.
Watch VO2 max estimates
Exercise-based watch estimates matched laboratory VO2 max on average, with individual readings off by up to about 10 mL/kg/min either way.
Matched the reference method 3 · Mixed results 6 · Did not match the reference method 4 · Shown to help 1.
Wrist heart rate and calorie estimates
Wrist heart rate averaged within 1 beat per minute of ECG at rest in pooled studies. Calorie estimates were commonly off by more than 20%.
Matched the reference method 2 · Mixed results 8 · Did not match the reference method 8 · Shown to help 2 · Mixed results 1.
Lab markers
25-hydroxyvitamin D
Cut points for low vitamin D range from 12 to 30 ng/mL. Low levels track with worse outcomes, but supplement trials mostly found no benefit.
Mixed results 3 · Linked in research 4 · Mixed results 1.
Absolute neutrophil count (ANC)
A European guideline calls a count below 500 cells/µL severe neutropenia. Reference intervals start lower in people with the Duffy-null blood type.
Mixed results 1 · Linked in research 2 · Checked, no link found 1.
Alanine aminotransferase (ALT)
Upper limits for ALT run from 19 to 59 U/L depending on sex and source. Higher values are tied to more deaths from liver disease.
Mixed results 3 · Did not match the reference method 1 · Linked in research 2 · Mixed results 1 · Checked, no link found 1.
Albumin
US laboratory sources print adult albumin ranges of 3.5 to 5.0 and 3.7 to 4.7 g/dL. Lower values are tied to more deaths.
Did not match the reference method 2 · Linked in research 3 · Checked, no link found 1.
Alkaline phosphatase (ALP)
US laboratory sources print adult upper limits for ALP of 104 to 129 U/L, and up to 382 for growing children.
Mixed results 1 · Did not match the reference method 1 · Linked in research 2.
Anti-Müllerian hormone (AMH)
Falls with age and predicts egg yield in IVF. Two US bodies state it is not a fertility test for women who are not infertile.
Matched the reference method 1 · Did not match the reference method 1 · Linked in research 2 · Mixed results 2.
Apolipoprotein A1 (ApoA1)
Low ApoA1 tracks heart disease in cohorts, but genetic studies and an infusion trial did not show that raising it lowers risk.
Matched the reference method 1 · Mixed results 2 · Linked in research 2 · Checked, no link found 2.
Apolipoprotein B (ApoB)
Higher ApoB is tied to more heart attacks in cohorts, genetic studies and drug trials. Cut points differ by source, from 60 to 130 mg/dL.
Matched the reference method 1 · Mixed results 1 · Linked in research 4.
Aspartate aminotransferase (AST)
Laboratory upper limits for AST run from 31 to 40 U/L. Muscle is a source too: one weightlifting session raised it for days.
Mixed results 2 · Did not match the reference method 2 · Linked in research 1 · Mixed results 1 · Checked, no link found 2.
Bilirubin (total)
US laboratory sources print adult upper limits for total bilirubin of 1.0 to 1.4 mg/dL. Fasting and Gilbert syndrome raise it.
Mixed results 2 · Did not match the reference method 3 · Linked in research 3 · Mixed results 1.
Blood urea nitrogen (BUN)
US laboratory sources print BUN ranges of 8 to 20 and 6 to 23 mg/dL. Higher values are tied to more deaths in heart failure.
Mixed results 1 · Did not match the reference method 1 · Linked in research 5.
C-peptide
Fasting values ran 0.66 to 2.63 ng/mL in adults without diabetes. In insulin-treated diabetes, below 0.24 is read as absolute insulin deficiency.
Mixed results 2 · Did not match the reference method 2 · Linked in research 2 · Mixed results 1.
Calcium level
US laboratory sources print adult total calcium ranges of 8.4 to 10.2 and 8.5 to 10.5 mg/dL. Albumin correction formulas added no accuracy.
Mixed results 1 · Did not match the reference method 1 · Linked in research 2 · Checked, no link found 1.
Copper level
A US report cites about 64 to 159 µg/dL for adults. Inflammation, oral contraceptives and pregnancy raise the result.
Mixed results 2 · Did not match the reference method 3 · Linked in research 3.
Cortisol
A morning blood level of 5 to 25 µg/dL is the usual range. An Endocrine Society guideline calls below 5 suggestive of adrenal insufficiency.
Mixed results 1 · Did not match the reference method 4 · Mixed results 1 · Checked, no link found 1.
Cystatin C
Cystatin C is used with creatinine to estimate kidney function. Published upper limits run from 0.92 to 1.12 mg/L, and assays have differed.
Matched the reference method 1 · Mixed results 3 · Linked in research 3 · Checked, no link found 1.
DHEA sulfate (DHEA-S)
An adrenal hormone that falls with age. Low levels track mortality, and raising it with DHEA did not change body composition or performance.
Mixed results 2 · Linked in research 1 · Mixed results 1 · Checked, no link found 1.
Erythrocyte sedimentation rate (ESR)
A US government health page gives upper limits of 15 to 30 mm/hr by age and sex. Values rise with age and pregnancy.
Matched the reference method 1 · Mixed results 1 · Did not match the reference method 2 · Linked in research 4.
Estimated GFR (eGFR)
KDIGO grades eGFR from G1 at 90 or above to G5 below 15. Muscle, creatine supplements and a meat meal can lower the estimate.
Matched the reference method 1 · Mixed results 1 · Did not match the reference method 4 · Linked in research 4.
Estradiol
Ranges run from 30 pg/mL or less after menopause to several hundred at ovulation. Routine assays read low levels poorly.
Mixed results 1 · Did not match the reference method 2 · Linked in research 4 · Mixed results 1.
Fasting glucose
US bodies put prediabetes at 100 to 125 mg/dL and diabetes at 126 or above. WHO starts its intermediate band at 110 mg/dL.
Mixed results 2 · Did not match the reference method 1 · Linked in research 4 · Mixed results 1.
Fasting insulin and HOMA-IR
US diabetes bodies have not endorsed a diagnostic range for fasting insulin or HOMA-IR. Published intervals differ, and assays can disagree twofold.
Mixed results 2 · Did not match the reference method 3 · Linked in research 1 · Mixed results 3.
Ferritin
Published cut points for low ferritin run from 11 to 45 ng/mL by source, sex and health state. Inflammation raises the result.
Matched the reference method 1 · Mixed results 2 · Did not match the reference method 1 · Linked in research 3 · Mixed results 1.
Folate level
WHO puts folate deficiency below 3 or 4 ng/mL, depending on the indicator. Serum folate reflects recent intake, red cell folate the longer term.
Matched the reference method 1 · Mixed results 1 · Did not match the reference method 1 · Linked in research 2 · Mixed results 1 · Checked, no link found 2.
Follicle-stimulating hormone (FSH)
Ranges depend on sex, cycle phase and menopause. Two bodies use 25 mIU/mL as a cut point, and results vary from cycle to cycle.
Mixed results 1 · Did not match the reference method 1 · Linked in research 2 · Mixed results 1 · Checked, no link found 1.
Free T3 (free triiodothyronine)
Free T3 ranges sit near 2.1 to 4.4 pg/mL and depend on the assay. A US guideline reports limited use in hypothyroidism.
Mixed results 1 · Did not match the reference method 1 · Linked in research 1 · Mixed results 1 · Checked, no link found 2.
Free T4 (free thyroxine)
Free T4 ranges sit near 0.8 to 1.9 ng/dL and depend on the assay. Higher levels inside the range were linked to more atrial fibrillation.
Matched the reference method 1 · Mixed results 1 · Did not match the reference method 2 · Linked in research 4 · Mixed results 1.
Free testosterone
No harmonized range exists. Published lower limits for men run from 64 to 120 pg/mL, and direct analog immunoassays do not match dialysis.
Mixed results 2 · Did not match the reference method 1 · Linked in research 1 · Mixed results 1 · Checked, no link found 1.
Fructosamine
Fructosamine reflects average glucose over 2 to 4 weeks. A US reference interval is about 195 to 258 µmol/L, and no treatment goals exist.
Mixed results 2 · Did not match the reference method 1 · Linked in research 3.
Gamma-glutamyl transferase (GGT)
Upper limits for GGT run from 33 to 68 U/L by sex and source. Higher values are tied to more deaths and cardiovascular disease.
Matched the reference method 1 · Mixed results 2 · Linked in research 4.
HDL cholesterol (HDL-C)
Low HDL cholesterol tracks heart disease in cohorts. Genetic studies and trials of HDL-raising drugs did not find that raising it lowers risk.
Matched the reference method 1 · Mixed results 2 · Linked in research 2 · Mixed results 1 · Checked, no link found 2.
Hematocrit
US sources put the adult range near 39% to 51% in men and 32% to 46% in women. Cohorts link both ends to more deaths.
Mixed results 1 · Did not match the reference method 2 · Linked in research 4 · Checked, no link found 1.
Hemoglobin
Anemia cut points differ by source, sex, age and pregnancy. Cohorts link both low and high hemoglobin to more deaths.
Mixed results 2 · Did not match the reference method 1 · Linked in research 4.
Hemoglobin A1c (HbA1c)
US bodies put prediabetes at 5.7 to 6.4% and diabetes at 6.5% or above. Other bodies differ, and some conditions distort the result.
Matched the reference method 1 · Mixed results 2 · Did not match the reference method 2 · Linked in research 4 · Mixed results 1.
High-sensitivity C-reactive protein (hs-CRP)
Higher hs-CRP tracks with heart disease and death, but genetic studies suggest it is a marker, not a cause. Repeat results often change band.
Mixed results 2 · Linked in research 4 · Mixed results 1 · Checked, no link found 1.
Holotranscobalamin (active B12)
Adult reference intervals start between 19 and 42 pmol/L depending on the assay. Studies place likely B12 deficiency below 20 to 26.
Mixed results 4 · Linked in research 1.
HOMA-IR
No diabetes body has endorsed a HOMA-IR cut point. Published ones run from 1.85 to 4.65, and the result shifts with the insulin assay.
Mixed results 1 · Did not match the reference method 2 · Linked in research 3 · Mixed results 1.
Homocysteine
Higher homocysteine tracks heart disease and dementia in cohorts. Lowering it with B vitamins did not reduce heart attacks or deaths in trials.
Matched the reference method 1 · Mixed results 2 · Linked in research 1 · Mixed results 3 · Checked, no link found 1.
IGF-1
Reference intervals depend on age, sex and the assay. Both low and high levels go with more deaths in pooled cohorts.
Mixed results 2 · Linked in research 4 · Mixed results 1.
LDL cholesterol (LDL-C)
Higher LDL cholesterol is tied to more heart attacks in cohorts, genetic studies and trials. Cut points differ by source and by a person's risk.
Matched the reference method 2 · Mixed results 3 · Linked in research 2 · Mixed results 2.
LDL particle number (LDL-P)
Higher LDL particle number tracks heart disease, about as well as ApoB. Its cut points come from cohorts and the test's makers, not guidelines.
Mixed results 2 · Linked in research 2.
Lipoprotein(a)
Higher lipoprotein(a) tracks heart attacks in cohort and genetic studies. A 2026 trial that lowered it sharply found no fewer cardiovascular events.
Matched the reference method 2 · Did not match the reference method 1 · Linked in research 3 · Mixed results 1.
Luteinizing hormone (LH)
Ranges depend on sex, cycle phase and menopause. LH is released in pulses, so one sample can differ widely from the next.
Mixed results 1 · Did not match the reference method 2 · Linked in research 3.
Magnesium level
Research groups propose raising the lower limit of serum magnesium from 1.82 to 2.07 mg/dL. Serum holds about 0.3% of body magnesium.
Mixed results 1 · Did not match the reference method 1 · Linked in research 4.
Mean corpuscular volume (MCV)
Low MCV goes with iron deficiency and thalassemia trait. High MCV goes with B12 deficiency, alcohol and liver disease. Ranges differ by age.
Mixed results 1 · Linked in research 4 · Mixed results 1.
Methylmalonic acid (MMA)
WHO puts the cut-off for B12 deficiency at above 271 nmol/L, an expert panel at above 376. Age and reduced kidney function raise it.
Mixed results 4 · Linked in research 2.
Non-HDL cholesterol (non-HDL-C)
Higher non-HDL cholesterol tracks heart disease in cohorts and trials. Cut points run from 85 to 220 mg/dL by source and a person's risk.
Matched the reference method 1 · Mixed results 1 · Linked in research 3 · Mixed results 1.
NT-proBNP
Heart failure is unlikely below 125 pg/mL in European guidelines and less likely below 400 in UK guidance. Obesity lowers the result.
Mixed results 2 · Did not match the reference method 3 · Linked in research 3 · Checked, no link found 1.
Omega-3 index
The developers of the omega-3 index proposed above 8% as its low-risk zone. Cohorts support the link, and supplement trials found small or no benefit.
Matched the reference method 1 · Did not match the reference method 1 · Linked in research 2 · Mixed results 2.
Parathyroid hormone (PTH)
MedlinePlus gives 10 to 65 pg/mL. Upper limits fall to about 46 when people with low vitamin D are left out, and assays differ widely.
Mixed results 1 · Did not match the reference method 2 · Linked in research 3 · Checked, no link found 2.
Platelet count
Ranges differ by sex and age. Low and high-normal counts were linked to more deaths, and tube clumping can give a falsely low count.
Mixed results 1 · Did not match the reference method 1 · Linked in research 4.
Potassium level
US laboratory sources print adult potassium ranges of 3.3 to 5.1 and 3.5 to 5.0 mEq/L. Deaths were lowest at 4 to 4.5.
Did not match the reference method 2 · Linked in research 3.
Progesterone
Under about 1 ng/mL before ovulation and higher after it. Above 3 ng/mL a week before a period is taken as evidence of ovulation.
Mixed results 1 · Did not match the reference method 2 · Linked in research 3 · Checked, no link found 1.
Prolactin
Upper limits run from about 15 to 38 ng/mL by sex and assay. Stress at the draw and macroprolactin can produce a high result.
Mixed results 1 · Did not match the reference method 3 · Linked in research 3 · Mixed results 1.
Prostate-specific antigen (PSA)
No agreed cut point: sources use 4.0 ng/mL, 3.0, or limits that rise with age. Screening trials disagree on deaths prevented.
Mixed results 2 · Did not match the reference method 3 · Linked in research 2 · Mixed results 1.
Red cell distribution width (RDW)
Higher RDW was linked to more deaths and heart disease in large cohorts. Ranges differ by age, and results differ between analyzers.
Mixed results 1 · Did not match the reference method 1 · Linked in research 4.
Red cell folate
WHO puts red cell folate deficiency below 100 or 151 ng/mL, depending on the indicator, and sets 400 for women of reproductive age.
Mixed results 1 · Did not match the reference method 1 · Linked in research 2 · Checked, no link found 1.
Selenium level
Selenoproteins level off between about 70 and 125 µg/L by source. Above that, the result mostly reflects intake.
Mixed results 1 · Did not match the reference method 2 · Linked in research 3 · Mixed results 1.
Serum creatinine
A US survey laboratory lists 0.67 to 1.17 mg/dL for men and 0.51 to 0.95 for women. Muscle, creatine and meat raise it.
Mixed results 2 · Did not match the reference method 3 · Linked in research 4.
Serum iron
US survey manuals print 59 to 158 µg/dL for men and 37 to 145 for women. Results vary about 28% between draws.
Mixed results 3 · Did not match the reference method 2 · Linked in research 3 · Mixed results 1.
Sex hormone binding globulin (SHBG)
Reference intervals differ by sex and rise with age in men. Low SHBG goes with type 2 diabetes, high SHBG with fractures.
Matched the reference method 1 · Mixed results 1 · Linked in research 4 · Mixed results 1.
Sodium level
US laboratory sources print adult sodium ranges of 133 to 145 and 136 to 144 mEq/L. Lower values are tied to more deaths and falls.
Mixed results 1 · Did not match the reference method 1 · Linked in research 4 · Mixed results 1.
Thyroid peroxidase antibodies (TPO antibodies)
TPO antibody cutoffs depend on the assay. Positive results are tied to later hypothyroidism. Levothyroxine trials in antibody-positive women found no change in live births.
Did not match the reference method 1 · Linked in research 3 · Mixed results 1 · Checked, no link found 1.
Thyroid stimulating hormone (TSH)
Sources put the adult upper limit for TSH anywhere from 2.5 to 4.8 mIU/L. Limits rise with age and differ in pregnancy.
Mixed results 4 · Did not match the reference method 1 · Linked in research 5 · Mixed results 1 · Checked, no link found 1.
Total cholesterol
Higher total and non-HDL cholesterol track heart disease. Cut points differ by source, by age and by a person's risk.
Matched the reference method 1 · Mixed results 1 · Linked in research 3 · Mixed results 2.
Total iron binding capacity (TIBC)
US survey manuals print 250 to 400 and 250 to 450 µg/dL for adults. It rises when iron stores are low and falls with inflammation.
Matched the reference method 1 · Mixed results 2 · Did not match the reference method 1 · Linked in research 1 · Mixed results 1.
Total testosterone
Lower limits for men run from 264 to 350 ng/dL by source. Results run lower in the afternoon and after eating.
Mixed results 2 · Did not match the reference method 2 · Linked in research 4 · Mixed results 2 · Checked, no link found 1.
Transferrin saturation
Below 16% to 20% is used to flag iron deficiency, 45% or above to prompt iron overload testing. Results vary between draws.
Mixed results 3 · Did not match the reference method 1 · Linked in research 3.
Triglycerides
Several sources put the fasting cut point at 150 mg/dL. Higher levels track heart disease, but trials that lowered triglycerides gave mixed results.
Matched the reference method 1 · Mixed results 1 · Linked in research 3 · Mixed results 2.
Uric acid (serum urate)
Sex-specific ranges top out near 6 mg/dL for women and 7 for men. Gout guidelines set a treatment target below 6.
Matched the reference method 1 · Mixed results 1 · Did not match the reference method 1 · Linked in research 5 · Mixed results 2.
Urinary iodine
WHO grades the median of a population, with 100 to 199 µg/L adequate. One person's sample varies about 38%.
Matched the reference method 1 · Did not match the reference method 1 · Linked in research 1 · Mixed results 1.
Urine albumin to creatinine ratio
KDIGO puts its albuminuria categories at 30 and 300 mg/g. Risk rose with the ratio, with no threshold.
Mixed results 2 · Did not match the reference method 2 · Linked in research 4.
Vitamin A (retinol) level
WHO counts 20 µg/dL or below as low. The liver holds the level steady, and infection lowers it.
Mixed results 1 · Did not match the reference method 2 · Linked in research 3.
Vitamin B12 level
Published cut points for low B12 run from about 200 to 300 pg/mL. Borderline results are checked with methylmalonic acid or holotranscobalamin.
Matched the reference method 1 · Mixed results 2 · Did not match the reference method 1 · Linked in research 4 · Mixed results 1.
White blood cell count (WBC)
Reference ranges differ by source and start lower in people of African ancestry. Higher counts were linked to more heart disease, diabetes and deaths.
Mixed results 1 · Linked in research 3.
Zinc level
Low cut points run from 57 to 74 µg/dL by age, sex, fasting and time of day. Meals, afternoon draws and infection lower the result.
Mixed results 2 · Did not match the reference method 1 · Linked in research 4.