Recovery & Health

The answer

−18% MPS after one night

Also: cortisol +21% · testosterone −24%

After 30 hours awake, muscle protein synthesis dropped 18% versus a normal night. Cortisol climbed and testosterone fell. The effect is acute — one night doesn't mean permanent loss — but it shows why an under-slept training day is fighting itself.

The answer

Large across all domains

Hardest hit: skill control (−0.87 SMD) · aerobic endurance (−1.0 SMD in non-athletes) · explosive power (−0.63 SMD)

Across 45 studies, every performance measure dropped after sleep loss — and the size of the effect is in the practically meaningful range, not just statistical. Skill-heavy work suffers most. Untrained adults see bigger aerobic hits than athletes do. Early-night sleep loss is worse than late-night.

The answer

+8% belly-fat risk

RR 1.08 (95% CI 1.04–1.12) for habitual <6h sleepers

Compared to people sleeping 7+ hours, those sleeping under 6 hours had an 8% higher chance of developing abdominal obesity over the follow-up window. The signal is consistent across both sexes and across regions. It's not a fast effect — it's a years-long drift that compounds with other risk factors.

The answer

Sleep more

Most effective: sleep extension · naps (20–120 min) · Less reliable: hygiene, light, mindfulness

The reviewers ranked eight different sleep-improvement strategies against performance outcomes. Plain sleep extension — going to bed earlier or napping 20-120 minutes — was the strongest, most reliable lever. Light therapy, sleep hygiene, and removing screens work but less consistently. The cleanest win is the boring one: more time asleep.

The answer

77% of studies show strength drop

Effect varies by muscle group, exercise type, and individual

Across 13 studies and 10,346 people, 77% found that sleep deprivation reduced strength on at least one measure. Effects were stronger in women for some movements (e.g., 24-hour deprivation hit female knee-extension strength). For practical training, the pattern is: sleep poorly, expect underperformance; the size of the hit varies.

The answer

−8% muscle endurance

Strength: −5.5% · Anaerobic power: −5.8% · Vertical jump and anaerobic capacity: no significant change

Showing up dehydrated will quietly cost you ~5–8% on most lifting sets and longer efforts, but won't much affect a single jump or a 10-second sprint. The takeaway isn't to panic-drink before every session — it's to start training already hydrated, because the deficit hits the work that's already hardest to push through.

The answer

Yes, for hot endurance

Continuous endurance gain: moderate (Hedges' g = 0.46) · Strongest above 25°C / 77°F · Cognitive effect: small and inconsistent (5 of 49 tests)

If you're doing long aerobic work in heat, drinking during the session is a real performance lever — not just comfort. For short lifts, intermittent sport, or desk-based mental work, the case is much weaker. The studies looked at fluid intake during the workout, not all-day hydration habits — so "I had my water bottle on my bike" is the move the data actually supports.

The answer

+0.2 RPE points per 1% body-mass loss

Across 0.5–3% dehydration, max observed RPE difference: 0.81 points · Authors' practical-relevance threshold: ~3% body-mass loss

Light dehydration won't make today's session feel meaningfully harder — the math says you'd need to lose around 3% of bodyweight in sweat before you'd reliably notice. For someone at bw_70kg, that's about 2.1 kg / 4.6 lb of fluid. The result is from endurance-trained men in heat — read it as a ceiling, not a license to under-drink on long hot sessions.

The answer

Heat — not fluid loss

Hydration effect on glycogen: not significant (p=0.24, very high study-to-study variability) · Heat effect: significant (p=0.003)

If you're worried about burning through glycogen, the bigger lever in this data is keeping core temperature down — shade, airflow, pacing — not just drinking more. Hydration still matters for sweat replacement and cardiovascular drift, but it's not the primary mechanism behind faster glycogen depletion in hot training. Built on young trained men, so the heat sensitivity may be even higher in less-acclimated people.

The answer

Trivially if at all

ES = −0.06 (95% CrI −0.16 to 0.04) · GRADE: Low certainty · 42% of studies rated low/very low quality

The headline finding from the largest cycle-phase meta-analysis to date: women perform about as well across the menstrual cycle as they do at any other point. The early follicular phase showed a trivial decrement versus other phases, but the credible interval crossed zero and the authors themselves graded the evidence as Low certainty under GRADE. Most underlying studies didn't verify cycle phase with blood samples, which the authors flag as a real limitation. Their practical guidance: personalize rather than prescribe — don't skip a planned session because of where you are in your cycle.

The answer

No clear evidence

Underlying evidence: AMSTAR 1–11, GRADE QoE 2–3 (low-to-moderate) · Primary studies use unreliable phase verification

The authors' position is unambiguous: it is "premature to assume that it is essential to control for the menstrual cycle phase" when designing strength training. Their critique focuses on method — the primary studies generally infer cycle phase from calendar dates or basal body temperature rather than measuring reproductive hormones directly. Until that improves, headlines about phase-based training are running ahead of the data. Their recommendation for future research: within-subject designs with urinary LH or serum hormone confirmation of ovulation.

The answer

Maybe small effects

Isometric SMD 0.60 · Isokinetic SMD 0.39 · Dynamic SMD 0.14 — vs Colenso-Semple 2023 finding no influence

The honest read: this meta-analysis reports a directional pattern — peak isometric and dynamic strength in the late follicular phase, peak isokinetic at ovulation, weakest in the early follicular phase — that another rigorous review of the same literature interprets as no influence. Why both can be true: the underlying studies are small, mostly calendar-based for phase verification, and pooled effects are sensitive to which studies you weight. The takeaway isn't that one camp is right — it's that the underlying evidence isn't strong enough to settle the question.

The answer

Maybe evidence is preliminary

Only 5 of 23 studies tested across multiple phases · Findings labeled "promising," not conclusive

The authors don't make sweeping phase-specific dietary claims, and we shouldn't either. Their two consistent findings: pre-exercise calcium reduces bone resorption regardless of where you are in your cycle, and caffeine works the same throughout the cycle. Beyond that, the iron / omega-3 / hydration story is small studies and preliminary signals. The takeaway: track, observe your own pattern, but don't restructure your nutrition around generic phase advice without strong evidence behind it.

The answer

In this group yes

Correlational within the stressed group · Post-menopausal caregivers only · Mediator: NPY

In this specific population — post-menopausal women providing chronic care — eating more high-sugar/high-fat foods was associated with more abdominal fat, oxidative stress, and insulin resistance. The same association did not appear in matched controls. The candidate mechanism the authors identify is peripheral neuropeptide Y, not cortisol per se. The honest read: stress and food may interact in a way that drives body composition outcomes, but this is one small case-control study in a narrow population — extrapolating to younger people, men, or non-caregivers needs caution.

The answer

Multiple overlapping pathways

Mechanism review · No pooled effect estimates · Integrative framing

The authors don't make a clean single-mechanism claim — they argue chronic stress acts through three overlapping biological systems (autonomic, HPA-axis/cortisol, immune) which interact with behaviour and affect. The result is a "complex and multifaceted" link, not a simple cortisol-causes-fat story. Their practical implication: stress-related obesity probably won't respond to a single biological lever; integrated lifestyle, behavioural, and psychosocial approaches are likely needed together.

The answer

Susceptibility varies

Hair cortisol strongly related to abdominal obesity · But not all obese patients show elevated cortisol · Partly genetic

The honest framing: chronic cortisol exposure is associated with abdominal fat at the population level, but individual responses vary substantially. Some people show elevated hair cortisol and abdominal accumulation; others have similar stress profiles without the same outcome. The authors attribute the variation to differences in glucocorticoid receptor sensitivity (partly genetic) and individual stress-response biology. Practical implication: stress-targeted interventions should be tailored, not assumed to work uniformly.

The answer

Acutely yes transiently

Acute fasting: strong elevation · VLCD/LCD: not significant · Returns to baseline after several weeks

Two clean findings: (1) acute fasting raises cortisol meaningfully; less-severe caloric restriction does not. (2) Even the fasting-induced rise is transient — it appears in the initial period of restriction and returns to baseline after several weeks. The authors' actual mechanistic claim is that elevated cortisol may blunt weight loss during starvation, not that it shifts metabolism toward fat storage. The takeaway for the app: extreme deficits do produce a real but time-limited cortisol response; moderate deficits don't, and the body adapts even to severe restriction within weeks.

The answer

Discretely effects are modest

17 studies · Findings vary by activity amount · Authors flag substantial research gaps

The headline result is more cautious than popular framing. Active people do show some differences in gut microbiota composition compared to sedentary people, but the effects are described as "discrete" and they vary significantly with the amount of activity. The authors flag substantial gaps — most underlying studies don't control well for diet, which is the dominant driver of microbiome composition. The honest takeaway: exercise probably influences your microbiome, but the magnitude and direction are still being characterized.

The answer

Suggestively pilot evidence

B/B ratio modulation p = 0.001 · Pilot meta-regression · Direction/magnitude not in abstract

The pilot meta-regression detected a significant association (p = 0.001) between exercise type and intensity and shifts in the Bacillota/Bacteroidota ratio — one of the more commonly tracked microbiome composition metrics. But the authors are careful with framing: this is pilot/exploratory work, not a definitive dose-response model. The direction and magnitude of the shift, and which exercise patterns drive the largest changes, aren't detailed in the abstract.

The answer

Aerobic, above minimum training duration helps

Aerobic associates with diversity · Resistance does not · WHO-minimum produces no significant change

The findings here are more specific than the popular framing. Aerobic exercise associates with gut microbiota diversity; resistance training does not. Prevotella abundance — one of the more-studied genera — increases with training duration. And exercising only at the WHO minimum recommended dose doesn't produce significant compositional changes — the signal appears at higher doses. The takeaway: modality matters, and dose matters. Strength-only training has weaker microbiome effects than aerobic training does.

The answer

More selectively than commonly framed

No effect on: CK, HR, lactate, glucose, estradiol, SHBG, leukocytes/cytokines, CRP, calcium · Effects on cortisol/T/MPS in body of paper

The abstract emphasizes the null findings: alcohol after resistance exercise doesn't appear to modulate creatine kinase (a muscle-damage marker), heart rate, lactate, blood glucose, several reproductive hormones, inflammatory markers (CRP, leukocytes, cytokines), or calcium. Where alcohol does seem to disrupt — cortisol elevation, testosterone reduction, MPS suppression, amino acid suppression — the abstract leaves to the paper body. The honest takeaway: alcohol's effects on training recovery are real but more selective than "ruins everything" framing.

The answer

No at heavy doses

MPS −24% (ALC+PRO) · −37% (ALC+CHO) · mTOR −54% to −76% · Dose: 1.5 g/kg ≈ 12 drinks

The mechanism is real: protein co-ingestion does not protect muscle protein synthesis from alcohol's suppressive effect. With heavy alcohol (1.5 g/kg, about 12 drinks), MPS dropped 24% even when 25g of protein was consumed afterward. The molecular signal — mTOR phosphorylation — was suppressed 54-76% versus protein alone. Important context that the original framing often misses: this is a heavy-drinking scenario, not moderate consumption. The study establishes the upper-bound disruption pathway; how this scales to 2-3 drinks isn't directly answered by this trial.

The answer

Even 2 drinks disrupt REM

Low (≤0.50 g/kg ~2 drinks): REM onset/duration · High (≥0.85 g/kg ~5 drinks): + sleep & deep-sleep latency

Even small amounts of alcohol — about 2 standard drinks — are enough to delay REM onset and reduce REM duration. At higher doses (around 5 drinks), additional effects appear in sleep-onset latency and deep-sleep latency. The broader sleep-quality measures (total sleep time, efficiency, wake-after-sleep) showed "large uncertainty" across studies — so the cleanest claim is specifically about REM disruption, not "alcohol ruins all sleep quality." The takeaway: if you care about REM-stage sleep (relevant for memory consolidation and emotional processing), even moderate alcohol use has measurable effects.