The research behind RobustHealth

The answer

~1,600 kcal/day

Typical adult range: 1,200 – 2,200 kcal/day

For a 70 kg, 170 cm, 30-year-old adult, it's around 1,500–1,650 calories at rest depending on sex. The formula: 10 × weight (kg) + 6.25 × height (cm) − 5 × age, then +5 for men or −161 for women. This is your floor — every step, lift, and bike ride adds to it.

The answer

Mifflin-St Jeor

Lands within 10% of measured value in more cases than Harris-Benedict, Owen, or WHO/FAO/UNU

When a fitness app or calorie tracker asks for your weight, height, age, and sex to set a daily target, it is almost certainly running Mifflin-St Jeor underneath. This review is why — it beat the three other commonly-used formulas at predicting how many calories a typical adult burns at rest, and the advantage held in obese adults too, where older equations like Harris-Benedict tend to drift. If a tool you use defaults to Harris-Benedict, switch to Mifflin if the option exists.

The answer

Mifflin-St Jeor is most accurate

Mifflin: 82% within ±10% · Livingston: 79% · Accuracy lower in obese

Across 7 RMR prediction equations tested against indirect calorimetry in 337 adults, the Mifflin-St Jeor equation was the most accurate (82% of estimates within ±10% of measured RMR). Livingston was a close second at 79%. All equations performed worse in obese participants than in non-obese, regardless of which equation was used — meaning prediction error is partly a function of body size, not just the equation choice. The takeaway: Mifflin-St Jeor is the right default for general adults, but expect more error at higher body fat levels.

The answer

Equation accuracy depends on demographics

HB 57.5% · Mifflin 56.4% · WHO/FAO/UNU 55.2% · Owen under-predicts · Subgroup variance is large

Across 362 healthy adults, three RMR equations performed similarly at the population level: Harris-Benedict (57.5%), Mifflin (56.4%), and WHO/FAO/UNU (55.2%) all hit within ±10% of measured RMR for roughly the same fraction of participants. Owen consistently under-predicted. The actually-important finding: accuracy varies a lot by demographic subgroup — sex, BMI, age, and race/ethnicity all modify which equation fits best. The takeaway: the population-level equation choice matters less than recognizing that any equation will under- or over-shoot for specific subgroups.

The answer

Mifflin-St Jeor still best, but limited

Mifflin: 50.4% · HB: 36.8% · BIA: 36.1% within ±10% of IC

In 133 overweight and obese adults, Mifflin-St Jeor was the most accurate of three commonly-used BMR estimation methods, with 50.4% of estimates falling within ±10% of measured RMR (vs Harris-Benedict 36.8%, BIA 36.1%). Notably, all three methods over-predicted BMR by 100–200 kcal/day on average. The takeaway: Mifflin remains the right default even for higher-BMI populations, but expect substantial individual-level variation — only about half of estimates land within ±10% of true RMR, even with the best equation.

The answer

79.5 % within ±5% of DEXA

Lin concordance: 0.89 (men) · 0.86 (women) vs DEXA

For most adults, a tape measure around your neck and waist gets you within five percentage points of what a DEXA scan would show. That's good enough to track change over months — better than calipers, far cheaper than imaging. The errors are roughly balanced (about 9 percent of people get underestimated, 11 percent overestimated), so the method has no systematic bias.

The answer

Comparable

A measuring tape, used correctly, sits in roughly the same accuracy ballpark as calipers and basic bioimpedance for estimating body fat — at a fraction of the cost and with no operator-skill bottleneck. That's the reasoning behind the Navy method this app uses. Source is a 1998 conference abstract; full-text details (exact error bands per method) are not publicly accessible.

The answer

Bad at small changes

ADP: +2.1% body fat · DoD equation: +0.3% body fat

Tape-measure equations are fine for a single snapshot, but they're not sensitive enough to track small body-composition shifts in lean young men over months. If you're using the Navy method to monitor progress, expect it to lag and to under-register fat-mass changes that more direct methods like a Bod Pod or DEXA would catch. Small sample (21 men) — treat as a single signal, not a definitive verdict.

The answer

3-comp + TBW is most accurate

Siri 3-comp: ≤0.4% diff · Simpler methods: ±4.0–10.7% total error

When body-composition methods are tested against a 4-compartment reference (the Heymsfield gold standard), the 3-compartment hydrostatic weighing approach with total body water correction is the most accurate (mean differences ≤0.4% Fat, correlations >0.997). Simpler methods — including DEXA in some cases, skinfolds, and other field tools — carry substantially greater error (±4.0% to ±10.7% Fat). For app users tracking body-fat changes, this reinforces the importance of using a single consistent method over time rather than comparing absolute readings across different measurement approaches.

The answer

Field methods are population-level tools

Equivalence-tested · Lab vs field comparisons · Individual-level error substantial

Field body-fat measurement methods (skinfolds, BIA, circumference) work reasonably well for tracking populations but carry substantial individual-level prediction error compared to DEXA or other gold-standard lab methods. The practical implication for the app: don't cross-reference body-fat numbers from different measurement methods as if they're directly comparable. Track changes within the same method over time, and treat absolute numbers as estimates rather than precise readings.

The answer

1.6 g/kg/day

Most studies land between 1.0 – 2.2 g/kg

For someone at 70 kg, that's roughly 112 g of protein a day. Beginners often benefit from the higher end of the range; trained lifters typically plateau closer to the lower end. Going above 2.2 won't add more muscle.

The answer

1.6 vs 1.2 g/kg/day

Younger (<65): ≥1.6 g/kg · Older (≥65): 1.2–1.59 g/kg

For a 60 kg adult under 65, that's ~96 g protein daily. The same person at 70+ may see the same lean-mass benefit from ~80 g. Reinforces the Morton plateau and clarifies the requirement scales down with age.

The answer

30 g/meal (≥60 yr)

Older (≥60): ≥30 g/meal threshold · Younger (<30): no discrete threshold

For adults over 60, spread protein across 3–4 balanced meals instead of loading dinner. Breakfast is especially impactful after the overnight fast. For under-30s, total daily intake matters more than per-meal distribution. The MPS response lasts ~2–2.5 hours per meal, so spacing meals avoids long anabolic gaps.

The answer

~3 g leucine/meal (older + isolated)

Strongest support: adults >60 with whey/casein · Weakest: under-30s with mixed meals

About half the trials (16 of 29) supported the threshold. The hypothesis holds best when an older adult drinks a whey or casein shake — at that point, the 3 g leucine number matters. For a younger adult eating chicken and rice, total protein quantity outweighs the leucine number. Don't chase the 3 g target in mixed-food meals.

The answer

Yes — a modest drop

Total testosterone fell ~a third of a standard deviation (SMD −0.38); free testosterone similar (−0.37). Larger in Western men (−0.52).

If you cut dietary fat very low — think under about 20% of your calories — expect a small dip in testosterone, not a collapse, and it reverses when fat comes back up. Keep fat above roughly 20–25% of your daily calories as a floor. Response varies between individuals, so this is a population average, not a guarantee for you. Some keto-adapted athletes hold normal testosterone on very-low-carb, higher-fat diets — fat quality and total calories matter as much as the exact percentage, so the app's fat floor is an evidence-based default and its custom macro targets let you set your own if that's you.

The answer

Yes, for hard efforts

Matters most above moderate intensity and in sessions past ~45–60 minutes; low-intensity or short work is largely glycogen-sparing.

Match your carbs to how hard you train, not to a fixed number. On easy or short days you need very little. On long or high-intensity days, going in with low glycogen will cost you output near the end. If your training is hard and frequent, a carbohydrate floor protects performance — this is why the app doesn't let carbs drop to zero for active users. Fat-adapted endurance athletes are a partial exception: they burn fat at intensities where carbohydrate-dependent athletes would deplete glycogen.

The answer

No extra benefit (when calories matched)

Carb+protein vs carb alone: no overall effect · positive only when protein adds energy · zero when energy-matched.

For fast glycogen recovery, total carbohydrate is what does the work — even at higher carb doses (≥0.8 g/kg per hour), adding protein didn't refill glycogen any faster once calories were equal. So don't treat a carb-plus-protein shake as a way to use less carb. Protein after training is still worth eating for muscle repair — just not as a glycogen booster.

The answer

A little

Small effect (SMD −0.16) · roughly 24 mmol/kg of dry muscle spared over ~100 minutes.

Carbs during a long session help mainly by keeping your blood sugar up, with only a small glycogen-sparing bonus on top. For efforts under about 90 minutes it barely matters. For longer or back-to-back sessions the small sparing plus steady blood glucose is worth it — general practice is around 30–60 g of carbs per hour on those days. Don't expect mid-workout carbs to rebuild a depleted tank.

The answer

Yes, with the right setup

Author recommendation: 2.6 – 3.5 g/kg of fat-free mass per day, 3+ resistance sessions/week, prioritised sleep

If you are already trained, recomp is still on the table — but the lever is high protein anchored to your lean mass, not your bodyweight. For a 70 kg lifter at ~20% body fat (about 56 kg FFM), that is roughly 145 – 195 g of protein a day, paired with progressive resistance training at least 3 times a week. Deep, sustained deficits (contest-prep territory) make recomp much harder; moderate intake with strong training is the safer bet.

The answer

~1 kg lean mass gained (lifters)

Men: +0.8 kg FFM, −8.9 kg fat. Women: +0.9 kg FFM, −6.4 kg fat. Non-exercisers lost 1.7 – 2.8 kg of lean mass.

In a 5-month, ~500 kcal/day cut with 1.5 g/kg protein, the people who lifted held onto — and slightly grew — their lean mass while losing fat. The people who only did cardio, or no exercise, lost more lean tissue along with the weight. The signal is consistent across sexes, but the trial was observational, so other lifestyle differences between groups could be doing some of the work.

The answer

Same 3 levers

Higher protein · resistance training (with volume in trained lifters) · moderate caloric restriction

The newer literature reinforces the old recipe. If you want to lose fat without losing muscle, anchor on protein, lift with enough volume for your training age, and don't take the deficit deeper than you need to. The review is a useful pointer to recent primary trials, not a definitive intake number on its own.

The answer

Maybe — sex matters

Women on rising volume: ~+1 kg lean mass. Men averaged ~−2.8 kg lean mass loss across studies.

If you are a trained female athlete cutting on ≥2 g/kg/FFM protein, ramping training volume looks protective and may even nudge you into a small lean-mass gain. If you are a trained male athlete, even high volume and high protein typically won't fully prevent some lean-mass loss in a sustained deficit — expect to lose a bit and aim to minimise it rather than recomp through it.

The answer

2.4 g/kg/day

High protein: +1.2 kg lean, −4.8 kg fat. Low protein (1.2 g/kg): +0.1 kg lean, −3.5 kg fat.

On a deep cut with hard training, doubling protein from 1.2 to 2.4 g/kg/day turned a "barely held on to muscle" outcome into a clear lean-mass gain plus more fat lost. For a 70 kg adult, that's ~170 g of protein a day. Caveat: the regimen was 4 weeks of 6 sessions/week with intervals — not a long-term lifestyle test.

The answer

−3.8% body fat (lifting + deficit)

Lifting alone: −1.6% body fat, +0.7 kg lean. Lifting + deficit: −3.8% fat, −5.3 kg fat mass, lean preserved.

If you are overweight and want to drop fat, the most effective package in the data is lifting plus a moderate caloric deficit. Lifting alone helps. Dieting alone helps too, but typically costs you lean mass. The combination protects what muscle you have while cutting fat about twice as fast.

The answer

Moderate mixed training

Top-ranked overall: moderate mixed > moderate resistance > low-intensity resistance, all combined with a caloric deficit

For pure fat-loss speed, push aerobic intensity. For a recomp goal — keep muscle, lose fat — a moderate-intensity mix of resistance and cardio inside a deficit ranks highest. Resistance-leaning programs preserve lean mass better than aerobic-leaning ones, so build the week around lifting and add cardio as a deficit lever.

The answer

Lifts win for lean mass

AT and CT lost ~1 kg more fat than RT alone. RT alone preserved FFM best; CT did not statistically differ from RT or AT on FFM.

If your priority is keeping muscle, lifting alone protects fat-free mass best. If your priority is losing absolute kilograms of fat, aerobic or combined wins by about 1 kg. The trade-off is real — combining the two gets you most of the way on both, but the lean-mass advantage of pure lifting gets diluted when you stack aerobic on top.

The answer

~16 weeks detraining still retains nuclei

Animal data: ~42% vs 21% cross-sectional area gain in re-loaded vs naive muscle. Human regrowth-acceleration evidence is suggestive but not conclusive.

If you trained seriously before, the cellular machinery you built is probably still there after months off — that's the biology behind "muscle memory." Returning to lifting after a layoff or pregnancy, you should regain muscle faster than the first time around. The animal evidence for that acceleration is strong; the human evidence is consistent but still small.

The answer

Nuclei stay

Fibers shrank to 23% of post-overload size; myonuclear count was statistically unchanged. Recent human work (Pérez-Castillo 2025) extends the finding to ~16 weeks of detraining.

When you train and grow a muscle, you add nuclei to each fiber, and this study showed those nuclei do not leave when the muscle later wastes — at least in mice. The cellular machinery that drives growth is still in place when you come back to training. The result is animal-only by design; human data so far is consistent with it but covers shorter detraining windows.

The answer

Animal scores higher per gram

DIAAS above 1.0: whey, egg, milk, beef · below 0.8: most single plant proteins.

If you already eat plenty of animal protein, you're getting high-quality amino acids and don't need to overthink this. If you eat mostly plants, aim a little higher on total protein and mix sources — for example grains plus legumes — so you cover all the essential amino acids. Quality matters most when total protein is low; at higher intakes the gap between sources shrinks.

The answer

Yes, if it's low-quality

DIAAS ≥1.0 sources meet needs per gram · sources below 0.8 need higher intake or complementary blending.

If most of your protein is animal-based (DIAAS around or above 1.0), the standard target already fits — no adjustment needed. If you eat plant-forward, treat your target as a floor: add roughly 10–20% more total protein, or deliberately combine complementary sources like grains and legumes, to make up for lower per-gram quality. This matters most for older adults and anyone eating near the minimum.

The answer

Only fasted or high-volume

Of 19 acute-feeding trials: 13 showed no benefit, 6 showed a benefit

If you eat a normal meal a few hours before lifting, pre-workout carbs are mostly optional. They start to matter when you train fasted (early morning, no breakfast) or run very high-volume sessions — think 10+ sets per muscle group. The endurance-style "fuel every workout with carbs" rule isn't supported for strength training.

The answer

No benefit

Pooled effect was effectively zero (SMD 0.17, not statistically significant)

For trained endurance athletes, deliberately training with low carb availability did not improve performance over standard high-carb training. The trade-off was real: peak-interval intensity suffered. If you race or train hard, fuel the hard sessions. If you want to experiment with train-low, keep it to easy sessions and don't expect a performance edge.

The answer

Yes — small but real

Roughly twice the effect (SMD 0.38) in sessions over 45 minutes

Carbs before lifting give you a modest edge on total work performed — most of the benefit shows up in longer sessions (over 45 minutes) and when you train fasted. For a quick 30-minute strength session after a normal meal, you can skip the pre-workout carbs. For a 60-minute high-volume day, or an early-morning fasted lift, eating something carby beforehand is the safer bet. As a preprint, treat the exact numbers as provisional.

The answer

−19% CHD events (SFA → PUFA swap)

About 10% lower CHD risk per 5% of energy shifted from SFA to PUFA · CI: 0.70 – 0.95

On RobustHealth, swapping a chunk of your saturated fat — butter, fatty meat, full-fat dairy — for polyunsaturated sources like nuts, seeds, fish, and seed oils is the move with the cleanest randomized-trial backing for heart events. The 19% number is the total package effect over multi-year trials; you won't see anything in your bloods at one week. And the evidence is specifically about the swap, not about cutting saturated fat while replacing those calories with carbs or sugar — that's a separate question.

The answer

−28% CVD mortality (per 5%E SFA → PUFA)

SFA → PUFA: 28% ↓ CVD mortality · MUFA: smaller effect · Carbs: neutral for CVD · Sugar/refined starch: increases events

The headline isn't "eat less saturated fat" — it's "what you eat instead matters more than the cut itself." Swapping a slice of saturated-fat energy for polyunsaturated fat (seed oils, nuts, oily fish) is the substitution with the cleanest mortality signal. Swapping it for white bread, sugar, or refined carbs cancels the gain and may make things worse. That's why the app tracks saturated and polyunsaturated fat as separate macros — and why "low fat" labels can be misleading.

The answer

−21% cardiovascular events

Pooled RR 0.79 (CI 0.66–0.93, 11 RCTs) · No effect on CV mortality, all-cause mortality, MI, or stroke individually · SFA → PUFA/MUFA also improved fasting glucose, HbA1c, and insulin resistance

Cutting saturated fat looks like it lowers your chance of a major cardiovascular event over years of follow-up, but it doesn't clearly extend your life — the trials don't show a mortality benefit. The metabolic upside — better glucose control and insulin sensitivity — comes specifically when you replace the saturated fat with polyunsaturated or monounsaturated sources, not when you just cut total fat. Most of the underlying cohort evidence is methodologically weak; the cleaner randomized signal is the cardiovascular-event one.

The answer

Yes, but often modest and reversible

AT in 27/33 studies · ~30–100 kcal/day in most · Smaller in higher-quality studies · Attenuates during maintenance

The directional finding is robust: most weight-loss studies do find a modest reduction in energy expenditure beyond what lean-mass loss alone explains. But the authors explicitly hedge on magnitude — when higher-quality methods are used, AT values shrink, and in some cases disappear entirely. The effect also doesn't appear to be permanent: it attenuates during weight-maintenance phases and is undetectable 6–24 months after bariatric weight loss. The takeaway: weight loss does have a metabolic cost, but it's smaller and more reversible than the popular "metabolic damage" narrative implies.

The answer

No evidence does not support harm

BMI: 13/18 null · Fat mass: 15/20 null · Lean mass: 0/18 decrease · Metabolic rate: 12/14 null

The evidence reviewed here doesn't support the popular "weight cycling damages your metabolism" claim. Across 23 studies, the dominant finding for body weight, body composition, and metabolic rate was no significant association with cycling. That doesn't mean cycling is risk-free in every population (psychological, eating behavior, and cardiovascular outcomes are studied separately), but the specific worry that repeated dieting will permanently lower your metabolism or destroy lean mass is not supported by the evidence the authors reviewed.

The answer

Multiple adaptations compound

REE↓ beyond lean-mass loss · Thyroid↓ · Leptin↓ · Ghrelin↑ · Effects more pronounced in lean populations

For lean and athletic populations specifically, sustained caloric restriction triggers a coordinated set of adaptations — reduced resting expenditure, falling thyroid hormones, falling leptin (the satiety signal), and rising ghrelin (the hunger signal) — that collectively make further loss harder and regain easier. The authors' practical implication is that periodic maintenance breaks may attenuate these adaptations, though they're honest that individual variability is substantial and the specific timeframes are heuristics rather than physiological thresholds.

The answer

No fixed threshold

Ingested leucine dose predicted the response in older adults only · no blood-leucine variable predicted it in anyone.

Don't chase a magic leucine number per meal — the evidence for a hard "3 g leucine" switch is weak. If you're older, getting enough leucine (from a solid, high-quality protein serving) does help your muscle response, so don't skimp. If you're younger, total protein at the meal matters more than hitting a specific leucine target. Blood-leucine "spikes" from fast proteins aren't the deciding factor.

The answer

25% higher 24-h MPS

EVEN ~30 g/meal vs SKEW 11/16/63 g · FSR 0.075 vs 0.056 %/h (p=0.003) · n=8

When the same ~90 g of daily protein was split evenly across three meals instead of dumped into dinner, the body built measurably more muscle protein over 24 hours. The trial was small (8 people), so treat the exact 25% figure as a directional signal rather than a precise prescription — but it lines up with the broader meal-distribution literature. The practical move: don't leave breakfast and lunch starved of protein and back-load it all at dinner.

The answer

25–30 g/meal

Threshold floor: ~20 g/meal (response blunts in elderly below this) · Same per-meal target applies to young and older adults

About 25–30 g of high-quality protein per meal — three times a day — is the dose this review proposes as the operational target. The threshold matters more in older adults: below ~20 g a meal, the elderly muscle-protein-synthesis response gets blunted; younger adults tolerate smaller meals better. Rather than chasing a higher daily RDA, the practical move is making sure each meal clears the ~25 g bar.

The answer

Sometimes depends on conditions

RT + evidence-based nutrition foundational · Sleep, hormones, measurement limitations also matter

Trained individuals can recomposition under specific conditions: structured progressive resistance training combined with adequate protein and an appropriate calorie balance. Non-training variables — sleep quality and hormonal context — also moderate the outcome, as do real measurement limitations in body composition assessment (DXA, BIA, skinfolds all have noise that can mask small simultaneous changes). The honest framing is that recomposition is possible but harder than pure cutting or pure bulking, and works best for those with higher initial body fat or returning to training.

The answer

Sometimes magnitude unknown

Specific optimal surplus is unknown · Trained-lean populations benefit more · Novices/higher-BF can gain at maintenance or below

The honest answer: it depends on who you are. Trained lean populations likely need a surplus to maximize the RATE of muscle gain. Novices, returning lifters, and people with higher starting body fat can build muscle at maintenance or even slightly below. The specific magnitude of surplus needed isn't validated by current literature — the authors are explicit that "the specific energy surplus required to facilitate muscle hypertrophy is unknown." Practical implication for recomposition eligibility: the exact thresholds the app uses are reasonable defaults, not derived from a single anchor study.

The answer

2.3–3.1 g/kg LBM (contest-prep)

Contest-prep: 2.3–3.1 g/kg LBM · General RT (Morton): 1.6–2.2 g/kg bodyweight · Different populations

For trained, lean populations attempting recomposition — particularly those targeting low body fat — the Helms contest-prep protein floor (2.3–3.1 g/kg of lean body mass per day) is the relevant upper anchor. For more general recomposition scenarios (higher starting body fat, less aggressive deficit), the Morton 2018 general-RT floor (1.6–2.2 g/kg of bodyweight) applies. The denominators differ (LBM vs bodyweight) — comparing the numbers directly is misleading.

The answer

0.5–1 %/week

0.5–1%/week loss · 2.3–3.1 g/kg LBM protein · 15–30% fat · Population: natural bodybuilders

For lean, resistance-trained populations seeking further fat loss, the authors recommend a moderate deficit producing roughly 0.5–1% body weight loss per week. The protein floor they establish — 2.3–3.1 g/kg of lean body mass per day — is notably higher than general-population recommendations and is specifically tuned to muscle preservation under caloric restriction. Important context: this is a contest-prep population, where the goal is extremely low body fat with full muscle preservation. The recommendations may be more aggressive than typical fat-loss users need, but they anchor the upper-bound protein and the lower-bound deficit-rate that the app applies during fat-loss phases.

The answer

~6–8 weeks heuristic, not threshold

REE↓ · Thyroid↓ · Leptin↓ · Ghrelin↑ · Specific timeframes are heuristics, not thresholds

For lean and athletic populations, sustained caloric restriction triggers a coordinated set of adaptations that tilt the system against further loss. The authors' practical guidance is to use periodic maintenance breaks to interrupt these adaptations. The "6–8 week" timeframe cited in app prompts is the paper's discussion-level heuristic — useful as a default, but not a hard physiological threshold.

The answer

For maximum gains yes

Position paper · Bodybuilding/physique focus · Surplus required only to maximize hypertrophy rate

You can build muscle without a surplus — particularly if you're newer to lifting, returning after a break, or carrying enough body fat to fuel new growth from existing stores. But to maximize the rate of muscle gain, the authors argue an energy surplus is required. The paper's practical scope is bodybuilders and physique athletes pursuing peak hypertrophy; the surplus magnitude and macronutrient composition recommendations are tuned to that population.

The answer

0.25–0.5 %/week of bodyweight

0.25–0.5%/week gain · Protein 1.6–2.2 g/kg bodyweight · Per-meal 0.40–0.55 g/kg evenly distributed

For natural off-season muscle gain, the authors recommend a slow, controlled rate of weight gain — 0.25–0.5% of bodyweight per week. That's roughly 0.5–1.0 lb/week for a 200 lb lifter — slower than the popular "bulk" heuristics, deliberately so to limit fat accumulation. Protein at 1.6–2.2 g/kg of bodyweight (note: bodyweight, not lean mass — different from contest-prep recommendations), spread evenly across meals at 0.40–0.55 g/kg per serving. Faster gain rates were associated with disproportionate fat gain in the underlying literature.

The answer

~1.6 g/kg/day

Plateau ~1.6 g/kg/day · Older adults benefit less · Trained individuals benefit more

Protein supplementation drives muscle gain from resistance training, but the dose-response saturates at about 1.6 g/kg of bodyweight per day. Above that, more protein doesn't add more fat-free mass. Two moderators worth surfacing: protein supplementation is less effective with age, and more effective with training experience. The 1.6 g/kg figure is the most-cited quantitative anchor in modern protein-and-RT recommendations.

The answer

More is better with diminishing returns

+0.023 effect size per additional weekly set · No clean plateau identified in abstract

Weekly training volume drives hypertrophy in a graded, dose-response fashion: every additional set per muscle per week added a small but measurable increment to hypertrophy effect size (0.023 per set). The widely-used "10–20 sets per muscle per week" prescription is a practical heuristic that combines this dose-response with real-world recovery considerations — the abstract itself documents continued benefit across the studied range without specifying a clean ceiling.

The answer

150 min/wk moderate (or 75 vigorous)

150–300 min/wk moderate · 75–150 min/wk vigorous · ≥2 days/wk strengthening · Limit sedentary time

The headline number from the WHO 2020 guidelines: 150 minutes of moderate-intensity aerobic activity per week is the floor for "meets minimum recommendations" — or 75 minutes of vigorous activity, or an equivalent mix. Plus muscle-strengthening on at least two days per week. Sedentary time should be limited. The 300-minute upper threshold reflects additional benefit at higher doses without a clearly diminishing return for general-health outcomes. The app's 150-minute weekly aerobic floor is anchored to this guideline.

The answer

150 + 2-3 min cardio + RT days/wk

≥150 min/wk moderate aerobic · 2–3 days/wk resistance · ≥2 days/wk flexibility

The ACSM's combined-modality prescription for general fitness: at least 150 minutes per week of moderate-intensity cardio (or 75 vigorous), resistance training on 2–3 days per week covering each major muscle group, and flexibility work on at least 2 days per week. This is the most-cited general-population prescription in sports medicine and the basis for the foundational-health modality in the app.

The answer

1.0–1.2 g/kg/day minimum (more if active)

General: 1.0–1.2 g/kg · Active: ≥1.2 g/kg · With disease: 1.2–1.5 g/kg · Per-meal: ~25–30 g

For adults over 65, the PROT-AGE Study Group recommends at least 1.0–1.2 g/kg of bodyweight per day — notably higher than the general adult RDA (0.8 g/kg). Active older adults should aim for at least 1.2 g/kg, and those with acute or chronic disease should aim for 1.2–1.5 g/kg. The widely-cited per-meal floor (~25–30 g of high-quality protein) helps operationalize the daily target across meals. The reason for the elevated requirement: older muscle responds less robustly to dietary protein ("anabolic resistance"), so a higher dose is needed to drive the same anabolic response.

The answer

Whole foods limit added sugar/sodium/sat-fat

Added sugar <10% cal · Sodium <2,300 mg/day · Saturated fat <10% cal · Life-stage stratified

The federal DGA recommendations the commentary explains: emphasize nutrient-dense whole foods (vegetables, fruits, whole grains, lean proteins, low-fat dairy); limit added sugars to less than 10% of total calories; limit sodium to less than 2,300 mg/day; limit saturated fat to less than 10% of calories. The 2020–2025 edition was the first to provide explicit guidance from birth through older adulthood. The app's maintenance-calorie / whole-food framing aligns with these consensus recommendations.

The answer

Heavy low-rep wins for strength

Volume-equated · Hypertrophy: similar across protocols · Strength: powerlifting-style superior

When total volume was matched, both heavy/low-rep (7 × 3RM, 3-min rest) and moderate/higher-rep (3 × 10RM, 90s rest) produced similar muscle hypertrophy in trained men. But maximal strength was significantly higher in the heavy/low-rep group. The practical implication: hypertrophy is forgiving of rep-range choice when total volume matches, but maximal strength specifically requires heavy loads, low reps, and longer rest periods. Train for what you're trying to optimize.

The answer

Adequate protein + carbs context-dependent

Strength mode: Morton 1.6–2.2 g/kg bodyweight protein · Contest-prep edge: Helms 2.3–3.1 g/kg LBM · Carbs to fuel sessions

For most strength-training users, the general-RT protein floor (1.6–2.2 g/kg of bodyweight, per Morton 2018) is the relevant anchor — the Helms contest-prep recommendations (2.3–3.1 g/kg of lean body mass) apply specifically to the lean, contest-prep edge case. The cross-cutting principle that applies to both: carbohydrate intake should be sufficient to fuel hard sessions. Strength work is glycogen-dependent — under-fuelling impairs performance and recovery.

The answer

Heavy + multi-set + long rest failure not required

2–5 min rest · Multiple sets > single sets · Heavy + light load combo · Failure not necessary

The Suchomel synthesis identifies several programming principles for maximal strength: rest 2–5 minutes between sets, do multiple sets per exercise, combine heavy and light loads across a training cycle, and — notably — training to failure is NOT necessary for maximum strength gains. For weaker athletes, strength should come before power-specific training. The takeaway: strength prescription is more about consistent progressive overload across heavy-loaded multi-set work than about pushing every set to failure.

The answer

5–10 + 1.2–2.0 g/kg carb + g/kg protein

Carbs: 5–10 g/kg/day scaled to load · Protein: 1.2–2.0 g/kg/day · Energy availability matched to training

The tri-organizational position-stand recommendations: carbohydrate intake of approximately 5–10 g/kg/day, scaled to training volume and intensity — lower on rest/easy days, upper on hard or long sessions. Protein at 1.2–2.0 g/kg/day to support training adaptation. The defining operational principle for endurance: carbohydrate intake tracks training load, rather than sitting at a fixed daily target. This is what distinguishes endurance from other strategy modes in the app.

The answer

Periodize carbs to session demand

Train-low · Train-high · LCHF · Periodized carb diet · Events >30 min to ~24 hr

The contemporary best practice the paper identifies: a "periodized carbohydrate diet" — matching daily and weekly carbohydrate intake to session-specific demands. High carbs for hard sessions and races, lower carbs for easy/recovery days. The "train low" approach (deliberately training with reduced carbs) is a more advanced strategy that may enhance mitochondrial adaptations but should be used selectively, not universally. The chronic LCHF approach is a separate dietary path with its own trade-offs around high-intensity capacity. Strategy choice depends on the event distance and training phase.

The answer

Macro · Meso · Micro aligned to training

Macro: training phase · Meso: weekly structure · Micro: session/meal level

The Stellingwerff framework periodizes nutrition at three timescales: macro (training-phase level — base/build/taper differ in fuelling priorities), meso (weekly hard/easy day patterns), and micro (session-by-session and meal-by-meal carb availability). The central principle the framework operationalizes: nutrition shouldn't sit at a fixed daily target — it should track training prescription. The endurance strategy's training-load-tracked calorie modulation is the application of this framework.

The answer

±4% body fat (adults)

Adult standard error ~4.1% body fat; correlation ~0.79 with underwater weighing.

The formula: BF% = 1.20 × BMI + 0.23 × age − 10.8 × sex − 5.4, with sex = 1 for men and 0 for women. For a 30-year-old man at a BMI of 24 that's roughly 20% body fat; for a woman the same inputs give about 31%. Treat it as a ballpark when you have no measured reading — it's accurate for populations but can be off by 4% or more for you personally, especially if you're very lean or very muscular.

The answer

+508 kcal/day (on UPF)

Net swing: gained 0.9 kg on UPF · lost 0.9 kg on unprocessed

On the ultra-processed week, the same person ate about 500 extra calories a day without trying — roughly a meal's worth. That alone explains a pound of weight change every couple of weeks. The menus matched on paper, so the difference came from the food itself: easier to chew, faster to eat, harder to feel full from. The takeaway isn't "never eat UPF" — it's that calorie counts on a label undercount what UPF actually does to your appetite.

The answer

32 outcomes worsened

Convincing: CVD death +50% · T2D · anxiety +48% · mental disorders +53%. Highly suggestive: all-cause death +21% · obesity +55% · sleep +41%

Across 45 health outcomes the review looked at, 32 showed higher UPF intake tracking with worse health. The four most solid links are cardiovascular death, type 2 diabetes, anxiety, and common mental disorders. Because almost all of the underlying data is observational, you can't read these as exact dose-response — people who eat more UPF also tend to smoke more, exercise less, and earn less. But the direction and consistency across this many outcomes is unusual.

The answer

+15% % mortality (highest vs lowest UPF)

Dose-response: +10% mortality risk for every 10% increase in UPF share of diet

If you go from a diet that's mostly whole foods to one that's mostly ultra-processed, your risk of dying during a typical follow-up window is about 15% higher. The dose-response number is the practical one: every 10 percentage points more of your calories from UPF tracks with roughly 10% higher mortality. Like all UPF mortality data, this is observational — UPF-heavy diets travel with smoking, less activity, and lower income — but the dose-response gradient strengthens the case for treating it as more than coincidence.

The answer

+37% % T2D (highest vs lowest UPF)

Also: hypertension +32% · high triglycerides +47% · low HDL +43% · obesity +32%

If your diet is heavy on ultra-processed food, your risk of developing type 2 diabetes runs about 37% higher than someone whose diet is mostly whole foods. The same direction shows up across blood pressure, cholesterol, and weight gain — your metabolism takes the broad hit, not one specific marker. Only the diabetes link is on firm evidentiary ground here; the others are real but lower-confidence. CVD and death weren't pooled in this review.

The answer

+25% % mortality (vs lowest UPF)

Also: +29% heart disease · +34% stroke · +23% obesity · +20% depression

These numbers compare the highest UPF eaters to the lowest — not "any UPF" vs "none." And because this is observational data, it tells you the association exists, not that UPF directly causes these outcomes. People who eat more UPF also tend to exercise less, smoke more, and eat fewer fruits and vegetables. The signal is consistent enough across studies to warrant cutting back, but don't treat it as a precise dose-response.

The answer

Yes ≈48% fewer events

Major cardiovascular events: OR 0.52 (95% CI 0.32–0.84) across 4 RCTs

This is about as strong as diet evidence gets: in randomized trials, a Mediterranean pattern — olive oil, nuts, vegetables, legumes, fish, whole grains — cut major cardiovascular events by about half versus control diets over 2–7 years. It's the benchmark whole-food pattern the app points to. You don't have to follow it perfectly; even shifting toward it moves you in the protective direction.

The answer

More = less risk dose-dependent

Atrial fibrillation HR 0.65 · coronary heart disease HR 0.95 · stroke / hypertension / CVD mortality HR ~0.97 · each +1 adherence point helps

The more closely you eat Mediterranean-style, the lower your cardiovascular risk trends — and it scales, so partial adherence still counts. The size of the benefit varies by outcome (bigger for atrial fibrillation, smaller for things like blood pressure), and the most reliable evidence comes from the PREDIMED randomized trial rather than the observational studies. Treat it as a direction to move toward, not pass/fail.

The answer

−23% all-cause mortality (60+)

All-cause mortality −23% · cardiovascular mortality −27% · observational, not proof of cause

For adults over 60, sticking closely to a Mediterranean pattern is associated with roughly 23% lower risk of dying from any cause and 27% lower cardiovascular death. Because this is observational, it can't prove the diet itself is the cause — people who eat this way often exercise more and smoke less. Still, it lines up with the stronger randomized evidence, so it's a well-supported way to eat for the long haul.

The answer

−441 kcal/day

Spontaneous intake −441 kcal/day · −4.9 kg over 12 weeks eating freely

Protein is the most filling macronutrient, and this classic study shows why it matters: bumping protein to about 30% of calories led people to eat ~441 fewer calories a day without trying, and to lose about 5 kg over 12 weeks. For you, that means hitting your protein target isn't just for muscle — it quietly makes a calorie deficit easier to sustain because you feel full on less.

The answer

Leptin vs ghrelin

Ghrelin drives hunger before meals · leptin signals fullness · obesity blunts leptin (leptin resistance)

Your appetite isn't pure willpower — it's largely two hormones. Ghrelin spikes when your stomach is empty and makes you seek food; leptin, released from fat tissue, tells your brain you have enough stored energy. The catch: carrying more body fat can make you leptin-resistant, so the fullness signal gets muffled and hunger stays high. That's biology, not a lack of discipline — and it's why steady habits that support these signals matter more than trying to white-knuckle hunger.

The answer

≈1–2 kg (small)

Time-restricted eating −1.37 kg · fewer meals −1.85 kg · earlier calories −1.75 kg · authors: "small, uncertain importance"

When you eat can nudge your weight a little — time-restricted eating, eating fewer times a day, and shifting more food to earlier all led to modest losses of about 1 to 2 kg. But the researchers are upfront that these effects are small and may not matter much clinically. Meal timing is a fine-tuning lever, not a substitute for overall calories and food quality — use it if a pattern helps you eat less, not because the clock itself is magic.

The answer

Convincing for heart + longevity

Convincing-grade: CVD mortality, all-cause mortality, pancreatic cancer · strong: CVD & coronary disease

Fiber is one of the few nutrients where the evidence reaches "convincing" — the top grade — specifically for living longer and lower heart-disease and cardiovascular-death risk. It's observational data, so it can't prove fiber alone causes the benefit, but the consistency across hundreds of studies is why fiber is worth tracking. Most people fall short; getting to ~25–38 g/day from whole foods is a high-leverage habit.

The answer

Yes more good bacteria

Bifidobacterium SMD 0.64 · Lactobacillus SMD 0.22 · fecal butyrate SMD 0.24

Fiber feeds your gut bugs. In randomized trials, more fiber — particularly fermentable fibers found in onions, garlic, legumes, oats, and chicory — grew beneficial Bifidobacterium and boosted butyrate, a compound that nourishes the gut lining and supports metabolism. Because these were controlled trials, the effect is real, not just "healthy people have better guts." Variety of plant fibers matters more than any single supplement.

The answer

+8% diabetes risk per 5 GI units

RR 1.08 per 5 GI units (95% CI 1.02–1.15) · observational, not proof of cause

The kind of carbs you eat seems to matter, not just the amount. Diets with a higher glycemic index — lots of white bread, sugary drinks, refined starches — were linked to modestly higher diabetes risk, about 8% more per 5-unit GI step. This is observational, so it can't prove causation, but it fits what we know about blood-sugar spikes. Favoring lower-GI choices (whole grains, legumes, fruit, and pairing carbs with protein, fat, or fiber) is a sensible hedge.

The answer

+26% type 2 diabetes risk

Type 2 diabetes RR 1.26 (95% CI 1.12–1.41) · metabolic syndrome RR 1.20 (1.02–1.42)

Regularly drinking sugary beverages — soda, sweet tea, energy and fruit drinks — is linked to noticeably higher risk of type 2 diabetes and metabolic syndrome. Because liquid sugar hits fast and doesn't make you feel full, it's an easy thing to over-consume. It's observational data, but the signal is consistent, and swapping even one or two sugary drinks a day for water or unsweetened options is one of the highest-leverage sugar cuts you can make.

The answer

+20% type 2 diabetes risk

Type 2 diabetes +20% · obesity +17% · coronary heart disease +15% · stroke +10%

Across roughly 1.5 million people, those drinking the most sugar-sweetened beverages carried higher risk of obesity (+17%), type 2 diabetes (+20%), coronary heart disease (+15%), and stroke (+10%). It's observational, so other habits play a role, but the pattern is remarkably consistent with the trials on liquid sugar. Cutting back on sugary drinks is one of the clearest, easiest wins in a health-focused diet.

The answer

Likely helps post-exercise inflammation

Reduced key inflammatory markers (e.g. IL-6, CRP) after muscle-damaging exercise · per-marker effects uncertain

Taking omega-3s (EPA/DHA from fish oil) appears to calm the inflammation that follows hard, damaging workouts, which may support recovery. The evidence points that way, but the exact benefit per marker isn't nailed down and some results are mixed. Treat omega-3 as a reasonable recovery-support nutrient — most useful if your diet is low in oily fish — rather than a guaranteed performance booster.

The answer

Lower muscle-damage markers

Reduced creatine kinase, LDH, and myoglobin after exercise

Omega-3s appear to blunt the rise in muscle-damage markers (like creatine kinase) that show up after demanding workouts, which may mean less internal muscle disruption and smoother recovery. It's a modest, supportive effect — helpful alongside the basics of sleep, protein, and sensible training load, not a replacement for them. Oily fish 2–3×/week or a quality EPA/DHA supplement covers it.

The answer

Promising but mixed

Lowered CK/LDH and IL-6 in some trials · TNF-α often unchanged · muscle-mass benefit not yet proven

Omega-3s look promising for recovery — reducing some inflammation and muscle-damage markers — but the evidence is mixed and the authors are openly cautious, since not every marker responds and the best dose and duration aren't established. A fair read: worth including if your diet is low in oily fish, with realistic expectations. It supports recovery at the margins rather than transforming it.

The answer

In mice, yes (no human trial here)

CMC & polysorbate-80 → mucus thinning, inflammation, metabolic syndrome in mice · microbiota-dependent

This landmark study is why emulsifiers are on the radar — but it's a mouse study, so treat it as a red flag to investigate, not proof of human harm. It showed that two detergent-like additives (CMC and polysorbate-80) can erode the gut's mucus barrier and stir up inflammation through the microbiome. Human follow-ups (see the CMC feeding trial in this section) are starting to back parts of it up. Reasonable takeaway: minimizing heavily emulsified ultra-processed food is sensible, without panicking over trace amounts.

The answer

Yes in a controlled trial

15 g/day CMC, 11 days → reduced microbiome diversity, lower SCFAs, mucus encroachment in 2 of 7

This is the first human trial to back up the mouse findings: a common emulsifier, eaten at a realistic dose, measurably disturbed the gut in healthy people over just 11 days. It's a small, short study, and not everyone reacted strongly — but it moves emulsifiers from "worrying in mice" toward "plausibly relevant in people." Whole-food-based eating naturally keeps these additives low.

The answer

CMC & P80 worst; lecithin spared

CMC, polysorbate-80, carrageenans, gums → detrimental · lecithin → minimal impact (ex vivo model)

Not all emulsifiers are equal. In this lab-model screen of 20 additives, the synthetic detergents CMC and polysorbate-80 (plus some carrageenans and gums) most disturbed gut bacteria, while lecithin looked benign. It's an ex-vivo model, so it shows biological potential rather than proven effects in your body — but it's why the app treats CMC and polysorbate-80 more cautiously than lecithin. Check ingredient lists: the additive names, not just "emulsifier," tell the story.

The answer

No overall effect (subgroup signal)

Primary outcome null · overweight subgroup: ↓ insulin sensitivity, ↑ CRP/IL-6 · gut permeability ↑ overall

Contrary to scarier headlines, this trial did not show that carrageenan causes insulin resistance in general — the primary result was null in healthy young men. It did nudge gut permeability up, and hinted at harm specifically in heavier participants, who may be more susceptible. Read it as "possible concern for some, not proven harm for all." If you're metabolically healthy, dietary carrageenan at additive levels has weak evidence of harm; minimizing ultra-processed sources is still reasonable.

The answer

In some, yes (mostly mouse data)

Mouse mechanism confirmed via antibiotics + microbiome transplant · human effect seen only in a subset

This is the landmark study behind the "sweeteners might harm your gut" idea — but read it carefully: the causal proof is in mice, and only some humans reacted. It suggests that in susceptible people, heavy sweetener use could nudge blood-sugar control through gut-bacteria changes. Practically, that means "use with awareness," not panic: occasional diet-soda use in an otherwise whole-food diet has little proven human harm (see the human RCTs in this section).

The answer

Depends on the person

Effect appears in only some individuals · individual microbiome variation likely explains conflicting trials

Whether sweeteners affect you may come down to your personal gut bacteria. This review makes the case that no-calorie sweeteners can shift the microbiome and blood-sugar response — but only in a subset of people, which is why big trials often show nothing on average. That's a "know yourself" situation: if sweeteners seem to bother your digestion or you use a lot daily, cut back; if not, the average human harm signal is weak.

The answer

RCTs: no effect

RCTs: BMI −0.37 kg/m² (not significant) · cohorts: modest ↑ BMI / obesity / hypertension — likely reverse causation

Here's the tension in one study: controlled trials show diet sweeteners don't move your weight either way, while observational data links them to worse metabolic markers. The likely explanation is reverse causation — people already gaining weight switch to diet products, not the sweeteners causing the gain. Bottom line: sweeteners aren't a weight-loss tool, but the trial evidence doesn't support the claim that they cause weight gain or metabolic harm at normal intakes.

The answer

−0.8 kg (vs sugar)

Body weight −0.80 kg · fat mass −1.10 kg · waist −0.83 cm vs caloric comparators · vs water: no difference

Used to replace sugary foods and drinks, low-calorie sweeteners produced small weight, fat, and waist reductions — around 0.8 kg on average. But they were no better than simply drinking water, so they're best seen as a stepping-stone off sugar rather than a magic tool. If diet drinks help you cut sugar, they can assist a little; if water works for you, it works just as well.

The answer

Partly but prep fixes it

Phytate lowers iron & zinc absorption · soaking / sprouting / fermenting / leavening reverse much of it

Plant foods pair minerals with phytate, which cuts how much iron and zinc you actually absorb — one reason plant-based eaters may need more of these minerals. It's not a reason to avoid grains and beans (they're highly protective foods); it's a reason to prepare them well. Soaking, sprouting, sourdough fermentation, and pairing with vitamin-C-rich foods meaningfully boost mineral uptake. The app's bioavailability scoring reflects this gap.

The answer

Phytate + polyphenols

Phytate co-precipitates iron/zinc · polyphenols bind iron · soaking / germination / fermentation restore absorption

The iron and zinc on a bean's nutrition label overstate what you'll absorb, because phytate and polyphenols lock much of it up. The fix is old-fashioned food prep: soaking, sprouting, and fermenting legumes activate enzymes that free the minerals. This is why traditional cuisines soak and ferment beans and grains — and why the app scores minimally processed legumes differently from prepared ones for mineral delivery.

The answer

32.5% → 13% Mg absorption

Fractional Mg absorption: 32.5% (no phytate) → 24% (low dose) → 13% (high dose), dose-dependent

Phytate's reach extends beyond iron and zinc — it also blunts magnesium absorption, here by up to about 60% at whole-grain-equivalent doses. Practically this is a modest trade-off, not a reason to fear whole grains: they still deliver more total magnesium and fiber than refined grains, and preparation (leavening, soaking) recovers much of the mineral. It's why the app models bioavailability, not just the raw milligram counts on labels.

The answer

Absorption varies by food + structure

Most abundant dietary polyphenols aren't always the best absorbed · matrix & processing shift uptake

More polyphenols on paper doesn't always mean more benefit — how a food is built and processed changes what reaches your bloodstream. This is why the app treats polyphenol-rich whole foods (berries, tea, coffee, cocoa, colorful vegetables) as a quality signal, while recognizing that a supplement or extract won't necessarily match whole-food delivery. Variety across polyphenol families beats chasing a single "superfood."

The answer

−7% all-cause mortality

All-cause mortality HR 0.93 (95% CI 0.91–0.95) · observational, not proof of cause

Diets richest in polyphenols — lots of fruit, vegetables, tea, coffee, herbs, and cocoa — track with about 7% lower risk of dying over the study periods. Because this is observational, the polyphenols themselves aren't proven to be the cause; they travel with generally healthier eating. Still, it's a good reason to favor colorful, plant-rich meals. The effect is real but modest — a supporting reason, not a magic bullet.

The answer

Whole fruit lowers risk; juice raises it

Blueberries −26% (HR 0.74/3 servings/wk) · grapes −12% · apples/pears −7% · fruit juice +8%

Eating whole fruit and drinking fruit juice pull in opposite directions for diabetes risk — blueberries, grapes, and apples were protective, juice was harmful, despite similar sugar. Juicing strips the fiber and disrupts the matrix that slows sugar absorption. Practical rule: eat your fruit, don't drink it. Whole fruit is a health food; fruit juice behaves more like a sugary drink.

The answer

The DII a validated score

Scale −8.87 (anti-inflammatory) to +7.98 (pro-inflammatory) · 45 food parameters · 6 biomarkers

"Inflammatory diet" isn't just a buzzword — it's measurable. The Dietary Inflammatory Index turns what you eat into a number based on how each component affects inflammation markers in the blood. Anti-inflammatory foods (fiber, colorful produce, tea, spices, omega-3s) pull the score down; refined carbs, excess saturated fat, and processed foods push it up. This tool underpins the app's inflammatory-quality axis.

The answer

+36% CVD risk (worst vs best diet)

Highest vs lowest DII: RR 1.36 (95% CI 1.19–1.57) · ~8% higher CVD risk per DII point

The way you eat maps onto heart risk through inflammation: the most pro-inflammatory diets carried about a third higher cardiovascular risk than the most anti-inflammatory ones. It's observational, so it reflects association within overall dietary patterns rather than proof — but it's consistent, and the direction is unsurprising. Shifting toward anti-inflammatory foods (produce, fiber, fish, olive oil, spices) is a low-risk, well-supported move.

The answer

Contested ratio hypothesis

Western ~15–17:1 vs ancestral ~1:1 · lower ratios proposed as anti-inflammatory · "ideal ratio" debated

The idea here is that we eat far more omega-6 (from processed seed oils) relative to omega-3 (from fish, walnuts, flax) than our biology evolved for, tilting toward inflammation. It's a plausible and influential framing — but treat specific "target ratios" cautiously, because the evidence is contested and many experts think simply getting enough omega-3 matters more than the ratio. The uncontroversial takeaway: eat more omega-3 sources and lean less on ultra-processed foods heavy in refined oils.

The answer

−0.8 kg (cutting sugar)

Reduce free sugars: −0.80 kg · isocaloric swap for other carbs: −0.04 kg (no change)

Cutting added sugar helps your weight — but mostly because it helps you eat fewer calories, not through some unique fat-storing property. When people cut sugar, they lost about 0.8 kg; when sugar was swapped calorie-for-calorie with other carbs, weight barely moved. The practical takeaway: added sugar (especially in drinks) is easy to over-consume, so trimming it is an effective way to reduce total intake. This is the evidence behind the WHO's "under 10% of calories" free-sugar guideline.

The answer

<25 g/day added sugar

Harmful links across ~45 outcomes · each 250 mL/day sugary drink: +17% coronary heart disease

Across the whole evidence base, higher added-sugar intake is linked to worse outcomes on many fronts, and sugary drinks are the worst offender — each daily serving tracks with meaningfully higher heart-disease risk. The authors suggest keeping free and added sugars under about 25 g/day (roughly 6 teaspoons). Most of this is association rather than proof, but the breadth and consistency make cutting back a safe bet. Whole fruit isn't the concern — it's added sugars and sweet drinks.

The answer

Only in excess (esp. from drinks)

Excess-calorie sugars ↑ liver fat (SMD 1.72) · sugar-sweetened beverages worst · no effect when calories controlled

Sugar drives liver fat mainly when it piles on extra calories — and sugary drinks are the biggest culprit because they're so easy to over-consume. When people ate sugar within their calorie needs or swapped it calorie-for-calorie with other carbs, liver fat didn't rise. So it's less "sugar is uniquely toxic" and more "liquid sugar makes overeating effortless." Keeping added sugar (especially drinks) modest protects your liver as part of overall calorie balance.

The answer

90 g/day whole grain

CVD −22% (RR 0.78) · cancer mortality −15% · all-cause mortality −17% · benefit up to ~210–225 g/day

Aiming for about three servings of whole grains a day — think oats, brown rice, whole-wheat, barley — lines up with substantially lower risk of heart disease, cancer death, and dying early. More helped up to a point (~7 servings). It's observational, so it can't fully prove cause, but refined grains showed none of this benefit, which points to the intact grain (bran, germ, fiber) doing the work. Swapping refined for whole grains is an easy, high-value upgrade.

The answer

−21% colorectal cancer risk

Whole grain (high vs low) RR 0.79 · total fiber: no protection after adjustment (RR 0.99)

In nearly half a million adults, those eating the most whole grains had about 21% lower colorectal cancer risk — but total fiber, after accounting for other habits, wasn't protective on its own. That suggests it's the whole-grain package (fiber plus the germ, bran, and their compounds), not fiber in isolation, that matters. It's observational, so treat it as a reason to favor intact whole grains over refined, not a guarantee.

The answer

≈10% lower fasting insulin

Fasting insulin −15 pmol/L (~10%) on whole grain · improved insulin sensitivity at matched calories/macros

This small but tightly controlled trial isolates grain quality: same calories, same macros, only whole vs refined grains — and the whole-grain diet improved insulin sensitivity and lowered fasting insulin. That means "a carb is a carb" isn't the whole story; the intact grain does something extra for blood-sugar control. It's a small study, so don't over-read the exact numbers, but it supports choosing whole over refined even when the macros look the same.

The answer

Good (from vegetables)

Vegetable nitrate vs lowest intake: CVD mortality HR ~0.51–0.63 · benefit plateaus · observational

Nitrate has a split reputation, and the source is everything. From vegetables — leafy greens, beetroot — higher nitrate intake tracked with substantially lower cardiovascular death in this Australian cohort. That's the good kind: vegetables pair nitrate with vitamin C and polyphenols that steer it toward beneficial nitric oxide rather than harmful compounds. The worry is nitrite added to cured and processed meats (see the IARC card), a completely different context. Don't fear nitrate in vegetables.

The answer

The food matrix decides

Vegetables supply ~80% of dietary nitrate · vitamin C & polyphenols block nitrosamine formation

The nitrate atom is identical whether it comes from spinach or a hot dog — what differs is the company it keeps. In vegetables, vitamin C and polyphenols block the conversion of nitrate/nitrite into cancer-linked nitrosamines, and the nitrate supports healthy blood flow. In cured meats, nitrite is added without those protectors and alongside heme iron and high-heat cooking, tipping toward harmful compounds. Same molecule, opposite outcomes — which is why the app scores them differently.

The answer

Group 1 processed meat

Processed meat: Group 1 (carcinogenic) · red meat: Group 2A (probably) · ~18% higher colorectal cancer per 50 g/day

"Same category as smoking" sounds alarming, but IARC groups reflect how CERTAIN the evidence is, not how dangerous something is. The evidence that processed meat can cause colorectal cancer is strong (Group 1), but the actual risk increase is modest — roughly 18% higher colorectal cancer risk per 50 g/day (about one hot dog). Red meat is a step below (probably carcinogenic). Practical translation: make processed meats an occasional food rather than a daily staple; you don't need to fear an occasional bacon sandwich.

The answer

Selectively yes with ≥6 mmol/day, >3 days

TTE +SMD 0.33 · Distance +SMD 0.42 · Endurance +SMD 0.48 · Peak power +SMD 0.25

Dietary nitrate (typically beetroot juice or nitrate salts) produces ergogenic benefits across some — but not all — exercise performance outcomes. Significant effects: time-to-exhaustion, total distance, muscular endurance, peak power output. Where the benefits don't appear: several other outcomes the umbrella review tested. The dose-response pattern: benefits are stronger with at least 6 mmol/day of nitrate and more than 3 days of supplementation. The takeaway: nitrate is a real but selective ergogenic aid — useful for endurance and peak-power events, less established elsewhere.

The answer

Trivially in clinical populations

Max TTE SMD 0.11 · Submax 0.16 · VO₂peak 0.002 · 6-min walk 0.01 · CVD distance 0.25 (small)

Important context: this paper specifically tests nitrate in patients with cardiopulmonary or metabolic disease — NOT healthy athletes. The pooled effect across most performance outcomes was trivial (SMDs of 0.01-0.16). Even though 46% of individual studies reported some benefit, the meta-analytic synthesis came out essentially null. The authors attribute the weak aggregate effect to study heterogeneity and small samples. The takeaway: the strong nitrate-in-healthy-athletes literature (see Poon 2025) does not extrapolate cleanly to patient populations.

The answer

Possibly a real but unproven signal

Top vs bottom quartile blood erythritol: MACE HR ~1.8–2.2 · promotes clotting in lab/animal models

This study put erythritol — a popular "keto-friendly" sweetener long considered inert — under suspicion. Higher blood levels tracked with more heart attacks and strokes, and it made blood platelets stickier in the lab. But important cautions: the human part is observational (in higher-risk patients), and blood erythritol is partly made by your own body from sugar, so it may be a marker of metabolic trouble as much as a cause. It's enough to warrant a "use with awareness" flag and more research — not a proven danger. If you use a lot of erythritol daily, moderation is prudent.

The answer

Yes made from glucose

Adiposity gainers: ~15× higher blood erythritol · high HbA1c: ~21× higher · endogenous via pentose-phosphate pathway

Here's the twist that complicates the erythritol-and-heart-disease story: your body makes erythritol internally from glucose, especially when blood sugar runs high. Young adults gaining belly fat had dramatically higher blood erythritol even before accounting for diet. That means a high blood level might reflect a body struggling with sugar rather than a spoonful of sweetener — an important caveat when interpreting the cardiovascular findings. It doesn't clear added erythritol, but it argues against panic and for better overall metabolic health.

The answer

+23% CHD per 2% of calories

Each 2% energy from industrial trans fat → ~23% higher CHD risk · raises LDL, lowers HDL

Not all fats are equal, and industrial trans fat (from "partially hydrogenated oil") is the one to eliminate outright — it's the only fat that both raises your bad cholesterol and lowers your good cholesterol at once. Even small amounts matter: a 2%-of-calories bump ties to roughly 23% higher heart-disease risk. There's no safe level, which is why regulators worldwide have banned it. If an ingredient list says "partially hydrogenated," that's a genuine red flag.

The answer

Yes, much worse

Trans fat: all-cause mortality +34%, CHD death +28%, total CHD +21% · saturated fat: no significant association

This large analysis sharpened an important distinction: industrial trans fat is clearly linked to dying earlier and to heart disease, while saturated fat's link to those outcomes was not statistically significant. That doesn't make saturated fat a health food — but it does put trans fat in a category of its own. Practical priority: eliminate trans fat (partially hydrogenated oils) first; treat saturated fat as "moderate," not "avoid at all costs." Observational data, so it maps associations, not proof.

The answer

No dairy trans fat looks neutral

Industrial TFA: RR 1.21 (trend, not significant, p=0.09) · ruminant TFA: RR 0.92 (no association)

The trans fat naturally present in dairy and beef isn't the villain — this analysis found no link between ruminant trans fat and heart disease, while industrial trans fat trended toward harm (it didn't reach statistical significance here, but the broader evidence, including Mozaffarian's review, does support industrial-TFA harm). So you don't need to avoid butter or full-fat dairy over their small natural trans-fat content. The target is manufactured "partially hydrogenated" oils — the distinction the app's scoring makes.

The answer

80–100% absorbed (additives)

Additive phosphate absorbed 80–100% vs 40–60% from whole foods · a concern mainly with reduced kidney function

The phosphorus added to processed foods behaves differently from the phosphorus in whole foods — it's almost completely absorbed, so processed products deliver a bigger phosphate hit than labels suggest. For healthy kidneys this is usually handled fine; the real concern is the large number of people with mild, undiagnosed kidney decline, for whom chronic high additive-phosphate intake can stiffen and calcify blood vessels. Another reason to lean toward whole foods over heavily processed ones.

The answer

44% of top foods

Frozen prepared 72% · dry mixes 70% · packaged meat 65% · baked goods 57% · soup 54% · yogurt 51%

Nearly half of best-selling packaged foods carry added phosphates — and you can't see them on the Nutrition Facts panel, which lumps additive phosphorus in with natural. They cluster in the usual ultra-processed suspects: frozen meals, boxed mixes, packaged meats, and baked goods. If you're watching phosphate (relevant mainly for kidney health), the ingredient list — not the nutrition panel — is where to look for phosphate salts.

The answer

~100% vs ~50% vs ~30% additive/animal/plant

Absorbed: inorganic additives ~100% · animal ~40–60% · plant (phytate-bound) ~20–40%

Not all phosphorus is equal. The inorganic phosphate in additives is almost completely absorbed, animal-source phosphorus about half, and plant-source (locked in phytate) the least. So a whole-food diet naturally delivers less absorbable phosphate than the same milligrams from processed foods — helpful for anyone managing kidney or vascular health. It's another case where the app models bioavailability, not just the number on the label.

The answer

EU: no longer safe (precaution)

Genotoxicity could not be excluded → EU ban (Aug 2022) · still permitted in the US

A European safety review couldn't rule out that titanium dioxide nanoparticles might damage DNA, so — applying the precautionary principle — the EU banned it from food. That's a regulator erring on the side of caution under uncertainty, not proof it causes cancer in people. It has no nutritional purpose (it's purely cosmetic whitening), so avoiding it costs you nothing. The app flags it partly because there's simply no upside to eating it.

The answer

In rats, concerning (not human proof)

E171 → gut immune disruption + preneoplastic lesions in rats · cancer promotion in a carcinogen-primed model

This is a key study behind the titanium-dioxide worry — but it's in rats, and the strongest cancer signal showed up when the animals were also exposed to a chemical carcinogen (E171 nudged the process along rather than starting it). That's a legitimate red flag for a purely cosmetic additive, not proof it causes cancer in people. Since TiO₂ adds nothing but whiteness, there's no reason to seek it out; the app treats it cautiously on exactly that basis.

The answer

Plausibly, if inflamed (animal/in vitro)

TiO₂ nanoparticles worsened colitis via the NLRP3 inflammasome in mice + human cell cultures

If you have an inflammatory gut condition like IBD, this study is worth knowing about: in mice and human cells, titanium-dioxide nanoparticles amplified gut inflammation through a specific immune pathway. It doesn't prove titanium dioxide causes IBD in people, and healthy guts may handle it differently — but for a cosmetic-only additive, "might aggravate inflammation in susceptible people" is reason enough to avoid it. Those managing IBD have the strongest case to skip E171.

The answer

Yes small effect, general population

Significant for some mixes/ages (p 0.001–0.044) · effect small but real · not limited to ADHD kids

This landmark trial found that ordinary artificial food dyes (plus a common preservative) nudged up hyperactivity in typical children, not just those with attention problems. The effect was modest and didn't show up in every subgroup, so it's not a guarantee your child will react — but it was consistent enough that European regulators now require warning labels on foods with these dyes. If a child seems sensitive, cutting artificial colors is a cheap, harmless experiment worth trying.

The answer

~0.4 effect size (ADHD kids)

Effect ~0.4 SD in children with ADHD · smaller effect in general population · independent of parental expectation

Food dyes aren't a myth-level cause of hyperactivity, but they're not trivial either: in kids with ADHD, removing them produced a small-to-moderate improvement (roughly a third to half the size of stimulant medication), with a smaller effect in typical children. Not every child responds. But since eliminating artificial colors carries no downside, it's a reasonable first step for a sensitive or ADHD child — a low-cost experiment, not a cure.

The answer

Concerns, low risk

Red 3 = animal thyroid carcinogen · Red 40 / Yellow 5 / Yellow 6 may carry trace contaminants · dyes are purely cosmetic

This review makes the case that synthetic dyes are inadequately tested and carry theoretical cancer and allergy concerns — most pointedly Red 3, an animal thyroid carcinogen (which the US FDA moved to revoke from food in 2025). Keep perspective: it's authored by a consumer-advocacy group, the strongest data are from animals or contaminant traces, and the estimated real-world cancer risk from approved dyes at normal intakes is low. But because dyes are purely cosmetic, "why eat them at all?" is a fair stance — which is how the app treats them.

The answer

+2.35 mL/kg/min VO₂max

Also: exercise heart rate about 4 bpm lower. Biggest gains in anemic or iron-deficient women.

If you're a woman who trains and runs low on iron — heavy periods, a low-meat diet, or a diagnosed deficiency are the usual causes — correcting it can measurably raise your aerobic capacity and make a given effort feel easier. If your iron is already normal, extra iron won't boost performance and can be harmful, so test before you supplement.

The answer

Maybe

Proposed by the authors: keep 25(OH)D near 100 nmol/L; suggested 4,000–5,000 IU/day with vitamin K. Unproven for performance.

Fixing a real vitamin D deficiency reliably helps how your muscles work; pushing an already-normal level higher to chase extra performance is unproven. If you train indoors, live somewhere with long winters, or have darker skin, it's worth testing your level. Very high doses aren't automatically better and carry their own risks, so supplement to correct a shortfall, not to megadose.

The answer

Only if you're low

Neither magnesium nor vitamin D is a proven performance booster on its own. Where they help is by fixing a shortfall — and both are common shortfalls in people who train hard, diet, or sweat a lot. Aim to get enough through food and correct a measured deficiency; don't expect a supplement stack to add performance if your levels are already fine.

The answer

A modest link

Largest in older or deficient men; strongest for bioavailable (non-SHBG-bound) testosterone.

If your magnesium is low, correcting it may modestly support your testosterone — but magnesium is not a testosterone booster for men who already get enough. The evidence is mostly observational, so treat it as one more reason to hit your magnesium target through food, not as a hormone hack. The men most likely to notice are older, deficient, or training hard.

The answer

Yes — mostly strength

Anaerobic power/strength: improved in 5 of 7 trials. Aerobic capacity: 2 trials improved. Sprint speed: no change in most.

If you compete or train seriously and your vitamin D is low, correcting it is most likely to show up in strength and power work. Don't expect a big endurance or sprint boost, and don't megadose — the benefit comes from reaching adequacy, not from stacking high doses. Test in winter or if you train mostly indoors.

The answer

1–2% bone density (then plateaus)

Same small effect from food or supplements. Not enough, alone, to lower fracture risk.

Getting enough calcium matters, but going above adequate intake buys you only a one-time ~1–2% bump in bone density that won't keep your bones getting stronger. For fracture prevention, calcium works as part of a package — paired with vitamin D and, above all, weight-bearing exercise — not as a standalone lever. Hit your target and don't bother megadosing.

The answer

−30% hip fractures (with vitamin D)

Total fractures −15% · Hip fractures −30%. Effect is for the calcium + vitamin D combination, not calcium alone.

If bone health is your concern — especially as you age — the evidence-backed move is calcium and vitamin D together, not either one solo. Reaching your daily target for both meaningfully lowers fracture risk. This is why the app tracks the two side by side rather than treating calcium as a standalone bone nutrient.

The answer

A weak link

Push-ups vs calcium intake: r = 0.28. Most of the performance difference came from other factors.

Calcium is essential for muscle contraction, but eating extra beyond adequacy won't reliably make you stronger — this snapshot found only a faint correlation, and correlation in a single-timepoint study can't show cause. Cover your calcium target for bone and general health; look to training and protein for strength gains.

The answer

−3.5 mmHg systolic (hypertensives)

Up to −7.2 mmHg at higher intakes · 24% lower stroke risk. No significant effect on heart disease overall.

If your blood pressure runs high, eating more potassium-rich food — beans, potatoes, leafy greens, bananas, yogurt — can shave a few points off, and shifts the odds against stroke. The effect is small per meal but adds up. If you have kidney disease or take certain blood-pressure drugs, check with a doctor before pushing potassium, since your body may not clear it normally.

The answer

Potassium

Systolic drop — potassium: 3.5–9.5 mmHg · sodium reduction: 0.7–8.9 mmHg. Calcium and magnesium: smaller effects.

For blood pressure, the highest-leverage mineral moves are getting enough potassium and not overdoing sodium — the two work as a pair. Calcium and magnesium matter for other reasons but are weaker blood-pressure levers. Tracking sodium and potassium together, as the app does, captures most of the dietary effect on blood pressure.

The answer

Individualize it

Matters most for sessions over ~2 hours. Pair sodium with fluid — over-drinking plain water is a main hyponatremia trigger.

For everyday training under about two hours, normal food covers your sodium. For long, hot, sweaty sessions and ultra-endurance events, you need a deliberate sodium-and-fluid plan matched to your own sweat rate — and the biggest danger is drinking lots of plain water without replacing salt. If you race long, test your sweat losses rather than guessing.

The answer

Enough, not excess

Most benefit reaching adequate intake; effect fades past ~30 mmol/day extra and reverses above ~80. Strongest in hypertensives on high-sodium diets.

Aim to reach an adequate potassium intake from food rather than megadosing supplements — the blood-pressure payoff is front-loaded at getting from low to enough. Piling on high-dose potassium pills isn't just wasteful, it can backfire, and it's risky if your kidneys or medications limit how you clear potassium. Food-first is the safe way to hit the sweet spot.

The answer

Not reliably

Results mixed and inconclusive. High-dose antioxidants may blunt the training adaptations you're working for.

Don't take high-dose vitamin C or E hoping to reduce muscle damage or soreness — the evidence doesn't support it, and routinely megadosing antioxidants around training can dull the signal your muscles use to get fitter and stronger. Get these vitamins from food; save any supplementation for correcting an actual dietary shortfall.

The answer

Blunts stress, not soreness

Lowered oxidative stress and IL-6 · No effect on DOMS or strength recovery.

Taking vitamin C won't make you less sore or get your strength back faster, even though it changes bloodwork markers. Since those same oxidative signals help drive fitness gains, routinely megadosing around workouts is more likely to hurt than help. Eat vitamin-C-rich foods for general health and skip the high-dose recovery pills.

The answer

No clear benefit

IL-6 and MDA: no effect. CRP: reduced but only 2 trials, very-low certainty.

The current best evidence still doesn't give you a reason to take high-dose vitamin C for recovery — the one positive signal (CRP) is too thin to bank on. Cover vitamin C through food, and don't expect a supplement to blunt post-workout inflammation in a way that helps you. Bigger, better trials are needed before any of this changes.

The answer

Only if you're short

Training nudges up the need for B6 and riboflavin — a bigger deal if your diet is already marginal.

If you eat a reasonable, varied diet you're almost certainly getting enough B vitamins, and topping up won't give you an energy boost — that's a marketing promise, not a physiology one. The people who benefit from more are those running low: heavy trainers on restrictive diets, or anyone with a genuine shortfall. Cover them with food first.

The answer

1.26× time to exhaustion

Also lower exercise lactate and ammonia. Small trial (n=32) in untrained adults; industry product.

This hints a B-complex might help endurance, but read it carefully: the people were untrained, the group was small, and the product was commercial. It doesn't prove a B-complex will boost a well-fed, trained athlete. If your diet already covers B vitamins, the likeliest explanation for any benefit is topping up a marginal intake — not a true ergogenic effect.

The answer

Slightly higher

Matters most for active people who diet or eat poorly; effect is small and evidence is mixed.

Training bumps your B-vitamin needs up modestly, not dramatically — and a normal, varied diet easily covers the difference. The people who should pay attention are active dieters or anyone with a restrictive, low-variety diet, who can slip into a shortfall. For everyone else, no special B-vitamin dosing is needed just because you exercise.

The answer

~48% of US adults fall short

Gap is widest in older adults, some minority groups, and people with type 2 diabetes.

Magnesium is one of the nutrients people most commonly under-eat, so it's worth a look at your diet — leafy greens, nuts, seeds, legumes, and whole grains are the richest sources. A normal blood magnesium result doesn't guarantee you're replete, so focus on habitually eating magnesium-rich food rather than relying on a lab number. These links to disease are associations, not proof, but the food advice is low-risk.

The answer

Plausibly, yes

Serum magnesium can look normal despite depleted stores. Processed and refined foods are low in magnesium.

Treat this as a well-reasoned warning rather than proof: the case that hidden magnesium deficiency harms the heart rests on mechanism, not outcome trials. Still, the takeaway is low-risk and worth acting on — lean toward whole, minimally-processed foods rich in magnesium (greens, nuts, seeds, legumes, whole grains) instead of refined ones, and don't assume a normal blood test means you're covered.

The answer

It worsens insulin resistance

Cohort meta-analyses cited by the review: each ~100 mg/day more magnesium ≈ 8–15% lower type-2 diabetes risk (association).

If you're working on blood sugar or body composition, magnesium is a nutrient worth getting right — a shortfall makes your cells less responsive to insulin. The link between higher magnesium intake and lower diabetes risk is consistent but observational, so think "eat enough," not "supplement to prevent diabetes." Prioritize magnesium-rich whole foods, and if you have diabetes, know the disease itself can deplete you further.

The answer

30 g/day for 5 years (safe)

Performance dose: 3–5 g/day · Loading option: 20 g/day × 5–7 days

The position stand reviewed creatine intake up to 30 g/day for 5 years and found no major safety issues across populations spanning infants to seniors. Practical dosing is much lower: 3-5 g of creatine monohydrate daily, with an optional 5-7 day loading phase at 20 g/day to saturate muscle stores faster. No need for cycling or off-periods.

The answer

3× bigger gains (untrained)

Untrained SMD 1.06 · Trained SMD 0.32 · Low dose (3-5g) beats high dose (20g)

An untrained adult adding creatine sees roughly 3× the strength bump a trained lifter does (SMD 1.06 vs 0.32). Counterintuitively, low-dose protocols (3-5 g/day) produced larger pooled effects than high-dose loading (SMD 0.88 vs 0.24). Pairs best with moderate-to-high intensity training (>75% 1RM). If you're already a trained lifter, expect a modest single-digit-percent strength gain over a typical training cycle.

The answer

+4 / +11 kg upper / lower (men)

Upper WMD 4.43 kg · Lower WMD 11.35 kg · Female upper-body: no significant gain

Over 4-12 weeks of resistance training, men adding creatine gained an extra 4.43 kg on upper-body lifts and 11.35 kg on lower-body lifts versus placebo. The female sample was small (40 women across 2 studies) and showed no significant upper-body benefit — could be a real sex difference or a power-limited finding. Most men can expect this to translate to faster strength progression on compound lifts.

The answer

~1 mm extra CSA (very small)

Pooled ES = 0.11 · Individual regions: 0.10–0.16 cm with credible intervals crossing zero

Across 10 imaging-based trials, adding creatine produced muscle-size gains around 0.10-0.16 cm in elbow and knee muscles — a very small effect that's directionally positive but isn't statistically clean for any single region. Translation: creatine is a strength supplement first and a hypertrophy supplement only barely. If you're chasing visible muscle gain, expect strength-mediated progress, not direct mass.

The answer

+5.6 / +1.5 kg squat / cm jump

Bench +1.43 kg · Squat +5.64 kg · Vertical jump +1.48 cm · Wingate peak +48 W

In the largest creatine analysis to date, squat strength rose ~5.64 kg, vertical jump ~1.48 cm, and Wingate peak power ~48 watts versus placebo. Bench press gains were smaller (~1.43 kg). The effects were noticeably bigger in younger males running ≥8-week protocols; older adults and women saw minimal improvements on most measures. For under-30 men chasing power output, this is the most reliable supplement in the kitchen.

The answer

3–6 mg/kg pre-exercise

Endurance: 2–4% gain · Strength: SMD 0.16–0.20 · Endurance: SMD 0.28–0.38

For a 70 kg adult, that's roughly 210-420 mg of caffeine — about 2-4 cups of coffee — taken 45-60 minutes before training. Doses below 2 mg/kg may not help; doses above 9 mg/kg add side effects (sleep disruption, anxiety) without further performance benefit. Individual response varies with genetics; some people see clean gains, others see jitters. Cycling off periodically helps maintain sensitivity.

The answer

0.18 / 0.30 SMD strength / endurance

8 of 9 reviews showed positive effects · Female-specific evidence: limited

Across thousands of participants, caffeine improved strength by SMD 0.18 (small) and endurance by SMD 0.30 (small-to-moderate). Endurance benefits are roughly twice as large as strength benefits. Important caveat: the underlying data is almost all male, so the female response — especially with menstrual-cycle considerations — is still under-studied. For trained women, expect similar but not identical results.

The answer

≥6 mg/kg for strength (men)

Below 6 mg/kg: no significant strength effect · 45 min beats 60 min pre-exercise

For a 70 kg man, the threshold dose for strength benefit is about 420 mg of caffeine taken 45 minutes pre-exercise. Lower doses (e.g., a single cup of coffee, ~80 mg) won't move the strength needle. Endurance is more responsive to lower doses. The 14 RCTs included only 3 with female participants — women showed smaller effects, but the female data is too sparse for a confident female-specific dose.

The answer

Small boost (0.5–10 min efforts)

Overall effect 0.18 (small); in the 0.5–10 min window, exercise capacity climbs to 0.50. Stacking with sodium bicarbonate: 0.43.

Beta-alanine helps most for repeated high-intensity efforts in the half-minute to ten-minute range — rowing intervals, 400–1500 m runs, or high-rep sets pushed near failure. For a single all-out sprint or a long steady endurance session, do not expect much. The commonly studied protocol is about 3.2–6.4 g/day for at least 4 weeks to load muscle carnosine; since total dose did not change the effect, consistency matters more than megadosing.

The answer

Modest edge (best 4–10 min)

Overall effect 0.39; largest for 4–10 min efforts (0.55) and at 5.6–6.4 g/day (0.35).

If you are an experienced male lifter or athlete, beta-alanine can still add a small-to-moderate boost — mostly to hard efforts in the 4–10 minute range, like high-rep sets, HIIT, or middle-distance work. Aim for the higher end of standard dosing, around 5.6–6.4 g/day, taken consistently for several weeks. This analysis was male-only, so the size of the effect in women is still unclear.

The answer

No clear endurance benefit

No significant effect on time-to-exhaustion or time-to-completion; the evidence base is still small.

If you take a citrulline pre-workout hoping to run or ride longer, the pooled evidence does not back that up — endurance was no better than placebo. Citrulline may still help with perceived effort and soreness (see the recovery finding), but do not count on it to extend your endurance. The evidence base is small, especially in women.