The Evidence Index: 120 Health Myths and Facts, Graded by Science
Search 120 popular health, fitness and nutrition claims, each graded A to F against the research, with a one-line verdict, the key study and a link to the full evidence.
Strong support
Moderate support
Mixed or uncertain
Weak support
False
- B
Creatine makes you gain water weight
Moderate support. The NIH says weight gain is creatine’s only consistent side effect: about 2 to 4 lb (1 to 2 kg) in a month of training, mostly from water retention rather than fat.
- B
Turmeric reduces inflammation
Moderate support. A 2023 meta-analysis of 66 trials found curcumin or turmeric supplements lowered CRP by about 0.58 mg/L, plus IL-6 and TNF-alpha. No study shows cooking with the spice does the same.
- C
Protein shakes build muscle in healthy older adults
Mixed or uncertain. A meta-analysis of 36 trials (1,682 non-frail older adults) found supplements added no muscle or strength, even with weight training, as most already ate enough. A 2022 analysis found small gains when lifters raised total intake.
- C
Rhodiola reduces fatigue and stress
Mixed or uncertain. A systematic review of 11 trials called the evidence contradictory: 3 of 5 mental-fatigue trials were positive, but every study had a high or unclear risk of bias.
- D
Berberine is nature’s Ozempic
Weak support. A 2022 meta-analysis of 37 trials in type 2 diabetes found berberine lowers HbA1c by about 0.63 points, a metformin-like effect. It doesn’t mimic GLP-1 drugs, and no trial shows Ozempic-scale weight loss.
- D
Creatine causes hair loss
Weak support. The fear traces to one 2009 study of 20 rugby players in which DHT rose; hair was never measured. A 2025 randomized trial found no difference from placebo in DHT or hair density over 12 weeks.
- D
Magnesium glycinate is the best-absorbed form of magnesium
Weak support. Its absorption reputation rests mainly on a 12-person 1994 study. In a 2024 crossover trial of 40 adults, blood magnesium rose after citrate and oxide but not after glycinate.
- D
NMN supplements slow aging
Weak support. Small trials lasting weeks show NMN raises NAD+ markers, and one 10-week randomized trial improved muscle insulin sensitivity in prediabetic women. No human study shows it slows aging or extends life.
- D
Zinc supplements boost testosterone
Weak support. A 2023 review of 38 studies, 30 of them in animals, found zinc lifts testosterone mainly when zinc starts low. There’s no reliable evidence extra zinc raises it in men who already get enough.
- F
Creatine damages your kidneys
False. A 2026 meta-analysis of 20 randomized trials found creatine nudges blood creatinine up (0.13 mg/dL) but does not change eGFR or urea. Evidence in people with existing kidney disease is thin.
- F
Magnesium stops night-time leg cramps
False. A 2020 Cochrane review of 11 trials found magnesium no better than placebo for older adults with night cramps (moderate to high certainty); pregnancy trials conflict and are at high risk of bias.
- F
Vitamin D supplements prevent cancer and heart disease
False. The VITAL trial gave 25,871 US adults 2,000 IU a day for about five years and found no drop in cancer or major cardiovascular events. Most were not deficient, so treating real deficiency is a separate question.
- A
Eating more fiber lowers your risk of heart disease
Strong support. WHO-commissioned meta-analyses of 185 cohorts and 58 trials found the highest fiber eaters had 15–30% less heart disease and early death, and trials showed lower weight, blood pressure and cholesterol.
- A
Eating more protein while dieting helps you keep muscle
Strong support. A meta-analysis of 20 randomized trials in adults over 50 found higher-protein diets (about 1.0 g/kg a day or more) kept more lean mass and lost more fat during weight loss.
- B
Adults over 65 need more protein than the RDA
Moderate support. Expert groups (PROT-AGE, ESPEN) advise 1.0 to 1.2 g/kg a day for healthy over-65s, above the 0.8 g/kg RDA, because aging muscle responds less to protein. Long-term outcome trials are still limited.
- B
The Mediterranean diet prevents heart attacks and strokes
Moderate support. A meta-analysis of 6 trials (10,950 people) found about a third fewer major cardiovascular events and strokes, but no drop in overall deaths, and its authors called the evidence limited and variable.
- B
Ultra-processed foods make you eat more
Moderate support. In a tightly controlled NIH trial, 20 adults ate about 500 more calories a day and gained weight on an ultra-processed diet matched for sugar, fat, salt and fiber. The trial was small and short.
- B
Urine color tells you if you’re hydrated
Moderate support. Lab and field studies show urine color tracks urine concentration well enough for everyday checks. A Cochrane review found it, like thirst, unreliable on its own in people over 65.
- C
Fermented foods boost your gut microbiome
Mixed or uncertain. One well-run Stanford trial (36 healthy adults, 17 weeks) found a high-fermented-food diet raised gut microbiome diversity and lowered inflammatory markers. Larger trials haven’t confirmed it yet.
- C
Protein before bed builds more muscle
Mixed or uncertain. In a 2012 study of 16 young men, 40 g of casein before sleep after evening training raised overnight muscle protein synthesis about 22%, a borderline result. Total daily protein matters far more.
- D
Eating an egg a day raises your heart disease risk
Weak support. A 2020 BMJ analysis of 1.7 million people found no link between one egg a day and heart disease (relative risk 0.98). One 2019 US study disagreed, and people with diabetes may differ.
- D
You need 8 glasses of water a day
Weak support. A 2011 BMJ analysis found no good evidence for the rule. Fluid needs vary with body size, heat and activity, and much of your water comes from food and other drinks.
- F
Coffee dehydrates you
False. In its 2005 water report, the National Academies concluded that any diuretic effect of caffeinated drinks seems short-lived, and that coffee and tea still count toward your total daily water intake.
- F
High-protein diets damage healthy kidneys
False. A meta-analysis of 28 trials (1,358 adults without kidney disease) found kidney filtration changed no differently on diets of at least 1.5 g/kg a day than on normal protein. Existing kidney disease is the exception.
- F
Seed oils are bad for your heart
False. A 2014 meta-analysis of 13 cohorts (310,602 people) linked the highest intake of linoleic acid, the main fat in seed oils, to 15% fewer heart disease events and 21% fewer heart disease deaths.
- F
You must eat protein within 30 minutes of a workout
False. A meta-analysis of 23 hypertrophy studies found protein timing made no significant difference once total protein intake was accounted for. Daily protein intake was the strongest predictor of muscle growth.
- F
Your body can only use about 30 g of protein per meal
False. In a 2023 trial, 100 g of protein after exercise drove a bigger and longer (over 12 hours) muscle-building response than 25 g, with little burned for energy. One lab study in young men, but no cap appeared.
- A
Light weights build as much muscle as heavy ones if you train to failure
Strong support. A 21-study meta-analysis found loads of 60% of max or less grew muscle as well as heavy loads when sets went to failure, and a 28-study network analysis agreed. Heavy loads still build more max strength.
- A
You can still build muscle after 50
Strong support. A meta-analysis of 49 studies in 1,328 adults aged 50 and older found strength training added about 1.1 kg (2.4 lb) of lean mass on average, with higher-volume programs adding more.
- B
Blood flow restriction training builds muscle with light weights
Moderate support. A meta-analysis of 14 studies in adults over 60 found light-load BFR training built muscle about as well as heavy lifting, though heavy loads still produced somewhat bigger strength gains.
- B
Doing more sets per week builds more muscle
Moderate support. A 2026 meta-regression of 67 studies found muscle growth rises with weekly sets but with clear diminishing returns, and a 151-trial analysis in adults over 60 ranked low volume best for muscle size.
- B
Lifting fast (power training) beats slow lifting for older adults
Moderate support. A meta-analysis of 20 trials in adults over 60 found fast-lifting power training improved physical function slightly more than slow lifting (SMD 0.30), with no edge for strength or muscle size. Certainty was rated low.
- B
Push-ups build as much chest muscle as the bench press
Moderate support. In an 8-week randomized trial of 18 young men, load-matched push-ups grew the chest and triceps as much as bench pressing. The trials are small and short, but a second study found matching strength gains.
- B
Strength training protects bone density after menopause
Moderate support. A Cochrane review of 43 trials found exercise modestly protected bone, about 1% at the hip with resistance training and 3% at the spine with combined programs. Trial quality was low and fractures were not reduced.
- C
Doing cardio kills your muscle gains
Mixed or uncertain. A 2012 meta-analysis found running alongside lifting cut size and strength gains, but a 2022 analysis of 43 studies found cardio did not reduce muscle growth or max strength, only explosive power.
- C
Training muscles in a stretched position builds more muscle
Mixed or uncertain. A 5-week trial found curls in the stretched range grew the arm 8.9% versus 3.4%, but a 2025 meta-analysis of 12 studies found training at long and short muscle lengths produced similar growth.
- D
Beginners gain 1 to 2 pounds of muscle a month
Weak support. No peer-reviewed trial reports that figure. Studies measure single muscles instead: in 25 untrained men, thigh muscle grew about 9.6% over 8 weeks, with real growth starting around week 3.
- D
Doing 100 push-ups a day will build a bigger chest
Weak support. No study has tested the challenge. In a 4-week push-up trial, strength rose but muscle thickness did not change, and once 100 reps feel easy most sets end too far from failure to drive growth.
- D
Lifting weights boosts your testosterone
Weak support. A 48-study meta-analysis found testosterone rises right after hard exercise but is back to baseline within 30 minutes, and a review of 22 studies in men over 60 found no change in resting levels.
- D
Wearing a weighted vest builds bone density
Weak support. In a 12-month trial of 150 older adults losing weight, wearing a vest about 7 hours a day did not prevent hip bone loss compared with dieting alone. An earlier small trial showed only a nonsignificant trend.
- F
Slow reps (more time under tension) build more muscle
False. A meta-analysis of 8 studies found reps lasting 0.5 to 8 seconds built similar muscle when sets went to failure, and very slow reps of more than 10 seconds appeared to build less.
- F
You have to train to failure to build muscle
False. A 15-study meta-analysis found no significant difference in muscle growth between failure and non-failure training. Sets should still end close to failure, which matters most with light loads.
- A
Losing excess weight lowers inflammation
Strong support. A systematic review of 33 weight-loss studies found CRP, a key inflammation marker, fell about 0.13 mg/L for every kilogram lost, a steady, roughly linear drop across diet, exercise and surgery.
- B
A short walk after meals lowers blood sugar spikes
Moderate support. A 2023 meta-analysis of 8 small crossover trials (116 people) found walking soon after eating blunted post-meal glucose spikes, while the same exercise before the meal did not. Effects were short-term.
- B
Waist-to-height ratio predicts health risk better than BMI
Moderate support. A meta-analysis of 31 studies (300,000+ adults) found waist-to-height ratio flagged diabetes, high blood pressure and heart risk better than BMI, though only by about 4–5%.
- C
Apple cider vinegar helps you lose weight
Mixed or uncertain. A 2025 meta-analysis of 10 small 4–12-week trials found modest weight loss, but half were at high risk of bias, with hints of publication bias. The most-hyped dramatic trial was retracted in 2025.
- C
Dieting permanently slows your metabolism
Mixed or uncertain. Some slowdown beyond weight loss is real: Biggest Loser contestants burned about 500 fewer calories a day than expected six years later. But that was 14 people after extreme, rapid loss.
- D
HIIT’s afterburn effect burns lots of extra calories for hours
Weak support. The afterburn (EPOC) is real but small: a major review found it adds only 6–15% of the workout’s energy cost, roughly 20–45 calories after a 300-calorie session.
- D
Ozempic itself causes ‘Ozempic face’
Weak support. A 2025 systematic review, built mostly on case reports and expert opinion, found no evidence GLP-1 drugs target facial fat. The hollowing most likely reflects rapid, large weight loss itself.
- D
Smart scales accurately measure your body fat percentage
Weak support. They weigh you accurately, but in a 15-device lab comparison single body-fat readings were off by 3.3 to 12.6 percentage points. They track change over weeks far better than the absolute number.
- F
Fasted cardio makes you lose more fat
False. It burns slightly more fat during the session, but in a 4-week randomized trial fat loss was statistically the same: 2.4 lb fasted vs 1.5 lb fed. Daily calorie balance decides fat loss.
- F
HIIT burns more fat than steady-state cardio
False. A 2021 meta-analysis of 54 head-to-head trials found almost identical fat loss, a 0.17 kg gap slightly favoring steady cardio, with high-certainty evidence. HIIT mainly saves time.
- F
Lifting weights burns more fat than cardio
False. In an 8-month randomized trial of 119 adults, aerobic training cut more fat than resistance training, while lifting added more lean mass. Doing both gave the best overall body composition.
- F
Sauna suits help you burn fat
False. The quick scale drop is sweat that returns when you drink. In a controlled trial, a sauna suit added only about 23 calories to an interval workout, roughly 150 sessions per extra pound of fat.
- F
Your metabolism slows down after 40
False. A 2021 Science study of 6,421 people measured with doubly labeled water found size-adjusted metabolism holds steady from age 20 to 60. Midlife weight gain usually reflects muscle loss and less movement.
- A
Any amount of exercise is better than none
Strong support. A 2023 meta-analysis of 94 cohorts (30+ million people) linked even half the 150-minute weekly guideline to 23% lower risk of early death; the biggest gains come from the first increases.
- A
Higher cardio fitness (VO2 max) predicts a longer life
Strong support. A meta-analysis of 33 studies (102,980 people) linked each 1-MET higher fitness level to 13% lower all-cause mortality, and low fitness to 70% higher risk than high fitness. The data are observational.
- B
Enough exercise cancels out a day of sitting
Moderate support. In a meta-analysis of over 1 million adults, 60–75 minutes of moderate activity a day erased the extra death risk of sitting 8+ hours. Heavy TV watching was only partly offset. Data are observational.
- B
Grip strength predicts how long you will live
Moderate support. In the PURE study of 139,691 adults in 17 countries, each 5 kg lower grip strength was linked to 16% higher mortality, a stronger predictor than blood pressure. It mainly signals overall strength.
- B
Japanese interval walking beats regular walking for fitness
Moderate support. In a 5-month randomized trial of 246 adults averaging 63, alternating 3-minute fast and slow walking raised thigh strength and aerobic capacity more than steady walking. Most evidence comes from one Japanese group.
- B
Loneliness raises your risk of dying early
Moderate support. A meta-analysis of studies from 1980 to 2014 linked loneliness to 26% higher odds of death, social isolation to 29% and living alone to 32%. Observational, and poor health can also cause isolation.
- B
Regular sauna use lowers your risk of heart disease
Moderate support. A Finnish study of 2,315 middle-aged men followed for about 21 years linked 4–7 sessions a week to 63% lower risk of sudden cardiac death than one session. Observational, and men only.
- B
Struggling to get up off the floor predicts an earlier death
Moderate support. In 2,002 adults aged 51 to 80, each extra point on the 10-point sitting-rising test was linked to 21% better survival over about six years. It shows an association, not cause and effect.
- C
Being able to do 40 push-ups means a healthier heart
Mixed or uncertain. Firefighters who managed 40-plus push-ups had 96% fewer cardiovascular events over 10 years than those doing under 10, but there were only 37 events, and a study of 8,116 Canadians found no link with death.
- C
Hearing aids prevent dementia
Mixed or uncertain. In the ACHIEVE trial of 977 adults aged 70–84, hearing aids did not slow 3-year cognitive decline overall. A higher-risk subgroup did benefit, which needs confirming, and dementia itself was not measured.
- C
Kettlebell swings are a good cardio workout
Mixed or uncertain. Swings drove heart rate to 87% of max but oxygen use to only 65%. Snatch intervals raised VO2 max about 6% in one 4-week trial, while an 8-week swing trial found no aerobic gain.
- C
Wall sits are the best exercise for lowering blood pressure
Mixed or uncertain. A 270-trial network meta-analysis ranked isometric holds first, cutting systolic pressure 8.2 mmHg, but on few small trials with no head-to-head tests. A 2026 analysis in hypertensive adults ranked combined training and HIIT higher.
- D
Fasting for 16 hours kicks off autophagy
Weak support. Popular autophagy timelines come from yeast and mice. A 2026 review found human data limited and often inconclusive, and no study has pinned down when autophagy rises in people.
- D
Zone 2 is the best type of cardio for longevity
Weak support. A 2025 review found no evidence Zone 2 is the optimal intensity for building mitochondria; in trials, intervals matched or beat it minute for minute. Zone 2 still builds the fitness that predicts longevity.
- F
Sitting is the new smoking
False. Heavy sitting is linked to about 190 extra deaths per 100,000 people a year, versus over 2,000 for the heaviest smokers, roughly a tenfold gap. Sitting matters, but it is not comparable.
- F
Testosterone therapy raises your risk of heart attack
False. The TRAVERSE trial of 5,246 men with low testosterone found heart attacks, strokes and cardiac deaths no more common than on placebo (7.0% vs 7.3%), though atrial fibrillation and lung clots were more frequent.
- A
Most adults need 7 to 9 hours of sleep a night
Strong support. A National Sleep Foundation expert panel set 7 to 9 hours for adults (7 to 8 after 65), and a meta-analysis of 1.38 million people linked short sleep to a 12% higher risk of death.
- A
Therapy works better than sleeping pills for chronic insomnia
Strong support. A 2024 network meta-analysis rated high-certainty evidence that starting with CBT-I beats medication long term: about 41% of patients were in remission months later, versus 28% who started on sleeping pills.
- B
A regular sleep schedule matters as much as hours slept
Moderate support. In 60,977 UK Biobank adults tracked with wrist sensors, more regular sleepers had 20 to 48% lower death risk than the least regular, and regularity predicted mortality better than duration. Observational data only.
- B
Reading on a screen before bed makes it harder to fall asleep
Moderate support. In a controlled sleep-lab crossover trial, reading on a light-emitting e-reader before bed suppressed melatonin, delayed the body clock and slowed sleep onset compared with a printed book. The sample was small.
- B
Sleeping badly makes you more likely to catch a cold
Moderate support. When 153 healthy adults were given cold-virus nose drops, those who had averaged under 7 hours of sleep were 2.9 times more likely to get sick than those sleeping 8 or more. It is a single study.
- C
Blue light glasses help you sleep better
Mixed or uncertain. A 2023 Cochrane review found six small trials split evenly, three helpful and three not (very low certainty), and a 2025 meta-analysis of sensor-measured sleep in 49 people found no significant effect.
- C
Magnesium supplements help you sleep
Mixed or uncertain. A 2025 trial in 155 adults with poor sleep found magnesium bisglycinate improved insomnia scores only slightly more than placebo (effect size 0.2), with hints of bigger gains in people eating little magnesium.
- C
Weighted blankets help you sleep better
Mixed or uncertain. A 120-patient trial in people with psychiatric conditions found large self-rated insomnia improvements, but a trial in 67 autistic children found no gain in measured sleep, and most other studies are tiny.
- D
Taking glycine before bed improves your sleep
Weak support. Support comes from a few small trials, mostly by one Japanese group tied to an amino acid maker; the key 3 g trial had seven participants and found less next-day fatigue after restricted sleep.
- D
Taping your mouth shut at night improves sleep
Weak support. A 2025 systematic review found only 10 small studies (213 patients): two showed sleep apnea improvements, the rest no difference, and reviewers warned of suffocation risk if the nose is blocked.
- D
Tart cherry juice helps you sleep
Weak support. The evidence is two tiny crossover trials: 20 healthy adults slept longer on cherry concentrate, while 15 older adults with insomnia beat placebo only on time awake at night.
- F
A nightcap helps you sleep better
False. A 2025 meta-analysis of 27 studies found that about two drinks already delays and shortens REM sleep, worsening with dose; alcohol only sped up falling asleep at around five drinks.
- F
An afternoon coffee won’t affect your sleep
False. A 2023 meta-analysis of 24 studies found caffeine cut total sleep by about 45 minutes and reduced deep sleep; the authors advise finishing a regular coffee at least 8.8 hours before bed.
- F
Taking more melatonin, like 10 mg, helps you sleep better
False. A 2024 dose-response meta-analysis of 26 trials found melatonin’s effect on falling asleep and total sleep time peaks at about 4 mg a day, so higher doses add no extra benefit.
- F
You can function fine on six hours of sleep
False. In a 14-day sleep-lab trial of 48 adults, six hours a night caused steadily growing cognitive deficits equal to up to two nights without sleep, while people felt only slightly sleepier.
- A
A single workout sharpens your thinking
Strong support. A 2025 meta-review of 30 meta-analyses (383 studies, 18,347 people) found one bout of exercise improves attention, executive function and memory with a small-to-medium effect, across ages and exercise types.
- B
Exercise helps treat depression
Moderate support. A 2024 BMJ network meta-analysis of 218 trials (14,170 people) found moderate symptom reductions from walking, jogging, yoga and strength training, but rated confidence low to very low because only one trial had low risk of bias.
- B
L-theanine plus caffeine sharpens focus more than caffeine alone
Moderate support. In a 27-person placebo-controlled trial, 100 mg of L-theanine with 50 mg of caffeine improved both speed and accuracy on attention switching, while caffeine alone improved only accuracy. Trials are small and effects modest.
- C
Binaural beats boost focus
Mixed or uncertain. A 2019 meta-analysis of 22 studies found a medium average effect on cognition and anxiety, but a 2023 systematic review found only 5 of 14 studies showed the brainwave syncing the theory depends on.
- C
The MIND diet prevents memory decline
Mixed or uncertain. Cohort studies link the MIND diet to slower cognitive decline, but a 3-year NEJM trial in 604 older adults found it no better than a calorie-reduced control diet for thinking scores or brain scans.
- D
NSDR (yoga nidra) boosts your dopamine by 65%
Weak support. The 65% figure comes from a single 2002 PET brain-scan study during yoga nidra meditation; it has not been replicated in beginners using short guided sessions.
- F
Blue light glasses reduce eye strain from screens
False. A 2023 Cochrane review of 17 trials found the lenses may not reduce eye strain, and a double-masked trial of 120 computer users found no difference from clear lenses in symptoms or fatigue tests.
- F
Brain-training games make you smarter
False. In a six-week study of 11,430 people, brain training improved every trained task but produced no transfer to untrained tasks, even closely related ones.
- F
Burnout is a medical diagnosis
False. The World Health Organization lists burnout in ICD-11 as an occupational phenomenon resulting from chronic, unmanaged workplace stress, and states it is not classified as a medical condition.
- F
Multitasking helps you get more done
False. Task-switching experiments show the brain alternates between demanding tasks instead of doing them at once, and every switch costs extra time that grows as the tasks get more complex.
- A
Balance exercises lower fall risk in older adults
Strong support. A 2025 review of 27 trials in adults 60 and over found exercise consistently reduced falls, by roughly 20–45% for strength-and-balance programs and 31–58% for tai chi.
- B
Exercise relieves chronic lower back pain
Moderate support. A 2021 Cochrane review of 249 trials found exercise cut chronic back pain by about 15 points on a 0–100 scale versus no treatment; gains in daily function were smaller, about 7 points.
- B
Exercises can fix forward head posture
Moderate support. A 2018 meta-analysis of 7 randomized trials (627 people) found strong evidence that corrective exercise improves forward head posture and moderate evidence it eases neck pain. It cannot reshape structural curves.
- C
Bad posture causes back pain
Mixed or uncertain. A review of 41 systematic reviews found posture, sitting and bending linked to back pain in some studies but not others, with no consensus that any of them actually cause it.
- D
Barefoot or minimalist shoes prevent running injuries
Weak support. A 2022 Cochrane review of 12 trials found no shoe type reliably cut running injuries, and in one trial runners switching to minimalist shoes showed more bone stress on MRI.
- D
Exercise wears out your knee cartilage
Weak support. A review of 9 randomized trials using MRI found knee-loading exercise did not appear to harm cartilage in people with or at risk of knee osteoarthritis, though the evidence was low quality.
- F
Bed rest is the best treatment for a bad back
False. A Cochrane review of 10 trials found advice to stay active gave small gains in pain and function over bed rest for acute low back pain; for sciatica, the two made little difference.
- F
Surgery is the best fix for a degenerative meniscus tear
False. A 2022 Cochrane review found high-certainty evidence that arthroscopic surgery gives little or no benefit over placebo surgery for degenerative knees, including meniscal tears. Acute injury tears and truly locked knees differ.
- B
Foam rolling reduces post-workout muscle soreness
Moderate support. A 2019 meta-analysis of 21 studies found rolling after exercise eased perceived muscle pain by about 6%, a small-to-moderate effect. The authors judged foam rolling’s effects minor overall.
- B
Ice baths right after lifting blunt muscle growth
Moderate support. A 2019 trial found cold water immersion after every lifting session blunted muscle fiber growth over a training block, though strength gains were unaffected. Save cold plunges for rest days if size matters.
- C
Contrast therapy (hot then cold) speeds muscle recovery
Mixed or uncertain. A meta-analysis of 18 trials found contrast baths eased soreness and strength loss compared with resting, but worked no better than cold water alone, and every trial carried a high risk of bias.
- C
Taking ibuprofen after workouts kills your muscle gains
Mixed or uncertain. Max-dose ibuprofen blocked the post-workout rise in muscle protein synthesis in young men, yet in a 12-week trial of adults in their 60s it was linked to more muscle growth than placebo.
- D
Cold showers boost your immune system
Weak support. In a trial of 3,018 adults, a daily hot-to-cold shower cut self-reported sick-leave days by 29% but did not reduce the number of days people were actually ill.
- F
Your muscles recover more slowly from workouts as you age
False. A meta-analysis of 36 studies found strength loss and recovery after hard exercise were similar in older and younger adults, and older adults actually reported less muscle soreness.
- A
If-then plans help you follow through on goals
Strong support. A meta-analysis of 94 studies found that if-then plans (implementation intentions) had a medium-to-large effect on reaching goals, helping people both get started and avoid getting derailed.
- B
Positive self-talk improves athletic performance
Moderate support. A meta-analysis of 32 sports studies found self-talk had a moderate positive effect on performance, largest when people were trained in it and when the task was new to them.
- C
The marshmallow test predicts a child’s future success
Mixed or uncertain. A larger 2018 replication found the link between waiting time and age-15 achievement was half the original size, and shrank by two-thirds after accounting for family background and early ability.
- D
Grit matters more than talent for success
Weak support. A 2017 meta-analysis of 88 samples (66,807 people) found grit only modestly tied to performance and nearly identical to conscientiousness. At West Point, the admissions score predicted grades better than grit.
- D
Willpower is a limited resource that runs out as you use it
Weak support. A preregistered replication across 23 labs and 2,141 people found the classic ego-depletion effect was close to zero, with a confidence interval that included no effect.
- F
It takes 21 days to form a habit
False. A 2024 review of 20 studies found habits took a median of 59–66 days to form, with averages of 106–154 days and individual times ranging from 4 to 335 days.
- B
Roman legionaries marched 20 miles in five hours
Moderate support. Vegetius set 20 Roman miles in five summer hours as the standard military pace, about 18.4 modern miles in 5 to 6 of our hours. He was describing an ideal training standard, not a measured average.
- F
Knights in full armor could barely move
False. Plate armor weighed 30–50 kg but was spread over the whole body. In a treadmill study, wearers walked and ran, though walking cost 2.1–2.3 times the energy of unloaded walking.
- F
Sumo wrestlers stay healthy despite their size
False. Active wrestlers tend to have low visceral fat, but in an observational study of 95,000+ former athletes, sumo had the most negative link to lifespan of any men’s sport: about 9.8 years shorter.
- F
Vikings were giants
False. Skeletons suggest early medieval northern European men averaged about 173 cm (5 ft 8 in), tall for their era but not giants. The sample was modest, so treat the figure as an estimate.
No claims match that search. Try a broader word, or clear the filters.
Grades reflect the evidence on each claim as stated, checked against the primary sources linked. It is educational, not medical advice. Last updated October 2026.
The Evidence Index is our searchable list of popular health myths and facts, each graded A to F against the research. It covers 120 claims you have probably heard, from creatine and your kidneys to eight glasses of water a day, with a one-line verdict, the key study behind it and a link to the full article. Search a claim, or filter by grade or topic.
How we grade each claim
Every claim is graded as it is usually stated. A grade describes how well the evidence supports that statement. It does not describe how good or bad the thing itself is for you.
- A: Strong support. Consistent evidence from meta-analyses or large randomized trials shows the claim is true.
- B: Moderate support. Good studies support it, but with real caveats, such as a small effect, a narrow group of people or short trials.
- C: Mixed or uncertain. The studies conflict, or there are too few good ones to say either way.
- D: Weak support. There is little human evidence. Most of the support comes from mechanisms, animal studies, anecdotes or very small trials.
- F: False. Good evidence contradicts the claim.
The scale is a simplified version of the way medical guideline groups rate evidence. The GRADE approach asks how confident we can be that an effect is real, and it rates evidence down for risk of bias, inconsistent results, indirect evidence, imprecision and publication bias (1). We use the same questions in plain language, so a claim backed only by cell studies or a single small trial cannot earn more than a D, however often it is repeated online.
“The evidence is mixed” is a real answer, not a cop-out. A C grade means a careful reader should stay open-minded, not that the claim has been proved or disproved.
What grading 120 health claims showed us
| Grade | Meaning | Claims |
|---|---|---|
| A | Strong support | 12 |
| B | Moderate support | 30 |
| C | Mixed or uncertain | 21 |
| D | Weak support | 25 |
| F | False | 32 |
Nearly half the claims in the index, 57 of 120, earn a D or an F. Only 12 earn an A. A few patterns explain most of the low grades:
- A mechanism is not an outcome. A supplement that changes a marker in a lab dish, or an exercise that briefly raises a hormone, has not been shown to change anything you would notice. Several popular claims about testosterone, autophagy and fat burning stop at this step.
- A short-term effect is not a long-term one. Burning slightly more fat during a fasted workout is not the same as losing more fat over months. Trials that measure the outcome you care about often find no difference.
- Animal and small studies get repeated as fact. Many confident claims trace back to mice, to a handful of volunteers or to a single study that was never replicated.
- Who paid for the research matters. Where a key study was funded by a company with a stake in the result, the full article says so.
The high grades are just as useful. Strong evidence backs some of the plainest advice in the index: being fitter predicts a longer life, light weights build muscle when you train close to failure, balance training lowers the risk of falls, and therapy beats sleeping pills for long-term insomnia.
How we check every claim
Every claim in the index links to a full Wonder Drop article that sets out the evidence in detail. For each one we:
- Read the section of our article that answers the claim.
- Identify the key source behind the answer, preferring systematic reviews, meta-analyses, clinical guidelines and large randomized trials.
- Check that the source exists and says what we say it says, using the original abstract or full text on PubMed rather than news coverage.
- Leave out any claim our article cannot support with a verified source, however popular it is.
Building the index also meant re-reading our own articles against their sources. Where we found an article that overstated or misquoted a study, we flagged it for correction rather than grading the claim on a weak foundation.
How to use the Evidence Index
- Search for a word such as creatine, sleep, magnesium or protein. The list narrows as you type.
- Filter by grade to see only the myths (F) or only the strongest findings (A), or by topic.
- Read the full article for the details, the numbers and who each finding applies to. A one-line verdict cannot cover every exception.
- Link to a single claim. Each card has a # link that opens the index at that claim, which is handy if you want to share or cite one answer.
Grades can change. When a large new trial or review changes the picture, we update the article and the grade. Some of our most useful tools sit alongside the index on our free health tools page, including the Zone 2 heart rate calculator and the protein calculator.
Does a low grade mean something is bad for you?
No. A grade is about the claim, not the food, supplement or habit. “Magnesium stops night-time leg cramps” gets an F because good trials found it works no better than a placebo for most adults with night cramps. That says nothing against magnesium in general, which matters for health in other ways. Likewise, a D for “cold showers boost your immune system” does not mean cold showers are harmful. It means the specific promise has little evidence behind it.
The index is educational and is not medical advice. If you have a health condition, take medication or are pregnant, talk to your doctor before you change your diet, supplements or training. See our medical disclaimer.
New to the jargon? Our health and fitness glossary explains research terms such as randomized controlled trial, confounding and relative versus absolute risk in plain English.
Sources
- Guyatt GH, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ. 2008;336(7650):924-926. PubMed 18436948
- The key source for each graded claim is linked on its card in the index above, and every claim links to a full Wonder Drop article with its complete reference list.
This article is for informational purposes only and is not medical advice. See our Medical Disclaimer before changing your exercise, diet, or supplement routine.
See every method in action
New research-backed workouts on YouTube every week — no gym required.
Frequently asked questions
What is the Evidence Index?
It is a searchable list of 120 popular health, fitness and nutrition claims. Each one is graded from A (strong support) to F (false) against the research, with a one-line verdict, the key study behind it and a link to a full Wonder Drop article.
How are the grades decided?
We grade each claim as it is usually stated. A means consistent evidence from meta-analyses or large trials supports it, B means good but limited support, C means the evidence is mixed, D means little human evidence, and F means good evidence contradicts it. Every key source is checked on PubMed before a claim is added.
Does a low grade mean something is bad for me?
No. The grade is about the claim, not the food, supplement or habit. For example, magnesium gets an F for stopping night-time leg cramps, but that says nothing against magnesium for other uses.
Can a grade change?
Yes. When a large new trial or review changes the picture, we update the full article and the grade. The index is reviewed as new evidence is published.
Can I link to or cite a single claim?
Yes. Each claim card has a # link that opens the index at that claim. You are welcome to link to it, and the key source is listed on every card.