Name: Minimum Effective Dose
Creator: Realised
Published: 2026-06-01

Summary

The first 20-30 minutes of exercise a week buys you most of the survival and fitness benefit; the curve is brutally front-loaded, so the honest minimum — for someone doing nothing — is a fraction of the official 150-minute target, and "I don't have time" is almost never the real barrier.

Why Strong

Tier 1 for the core claim ("a small dose captures most of the all-cause-mortality and fitness/strength benefit; the curve is concave/front-loaded") because: it is convergent across multiple large pooled cohorts (Arem 2015, Paluch 2022, Lancet 2025), strength meta-analyses (Momma 2022, Shailendra 2022), fitness data (Mandsager 2018), and RCT mechanism (low-volume HIIT; single-set strength). The shape of the curve is reproduced everywhere.

NOT a clean Tier 1 on the magnitudes because: the headline mortality percentages are observational and partly confounded (twin/MR deflation), and the specific minimum thresholds (20-30 min/week cardio, single hard set, 5,000-7,000 steps) carry real effect-size uncertainty and population dependence — that part is Tier 2.

NOT Tier 3 because: this is not emerging or thinly-evidenced — it is one of the better-studied questions in exercise epidemiology and physiology; the disagreement is about magnitude, not existence.

Practical takeaway

The honest floor (for someone currently doing nothing):
• Cardio/movement: ~20-30 minutes of brisk walking per week, in any chunks — even 3-4 one-to-two-minute bursts of "get a bit breathless" daily (stairs, hills, carrying shopping) counts and is associated with outsized benefit. Target ~5,000-7,000 steps/day as the high-value band; do not be paralysed by 10,000.
• Strength: one to two short sessions/week, one hard set per major movement (a squat/sit-to-stand pattern, a push, a pull, a hinge/carry) taken close to failure. Total working time can be under 30 minutes/week and still sit in the band where the mortality benefit peaks.
• Fitness, if you have 10 spare minutes: low-volume intervals — e.g. a handful of ~1-minute hard efforts (bike, hill, stairs) with equal rest, inside a ~15-minute session, 1-3×/week — raises VO2max efficiently.

Timing / response windows:
• Insulin-sensitivity and mood/sleep effects: same-day to within ~48 hours of a single session — you feel some return fast.
• Measurable VO2max gains from interval work: ~2-6 weeks.
• Strength: neural gains in 2-4 weeks; visible/strength-meaningful change by 6-12 weeks.
• Mortality/CVD risk reduction: a population-statistics outcome over years — not something an individual will feel, which is exactly why the framing matters.

What "working" looks like:
• Stairs and hills stop making you breathless; recovery between efforts is faster.
• A previously hard load (a heavy bag, getting off the floor, a flight of stairs) becomes easy.
• Resting heart rate drifts down; sleep and mood improve on training days.
• You are consistent — the dose you actually repeat beats the dose you plan and skip.

What to track: weekly step average; sessions completed (binary, not perfection); one strength load and whether it is getting easier; resting heart rate trend. Do not track "did I hit 150 minutes" as a pass/fail gate — that is the framing this entry exists to dismantle.

The progression note: the minimum is the floor, not the destination. Once consistent, adding volume up to roughly the guideline band (and a strength dose past the mortality-minimum, if hypertrophy or performance is the goal) does add benefit. See physical_progressive_overload. The minimum protects the downside; it is not the optimum for every goal.

Evidence detail

Why This Entry Exists

A user reads "150 minutes of moderate activity plus two strength sessions per week" — the WHO/standard guideline — does the math against a life that already feels full, concludes the bar is unreachable, and does nothing. The all-or-nothing framing is the enemy of the sedentary person, and the sedentary person is exactly who has the most to gain. The dose-response curve for nearly every hard outcome (death, cardiovascular disease, fitness, strength) is concave: the steepest gains happen in the move from zero to a little, and it flattens hard after that. The person going from nothing to 20 minutes a week gains far more, per minute, than the person going from 200 to 220.

This entry exists to give that user the floor, not the ceiling — and to do it without lying about how much the floor actually buys.

What bad advice does this protect against?
• The implicit message that sub-guideline activity is "not enough to bother" — false, and the most harmful framing in the space, because it converts "I did a bit" into "I failed."
• The fitness-industry inversion: that meaningful results require high volume, long sessions, expensive programming, or a gym membership. The dose-response data says the opposite.
• The over-correction we have to guard against too: minimum-dose marketing ("just 7 minutes!") that implies the minimum is optimal. It is not. The minimum is the floor that protects most of the downside; more (up to a point) does more. We hold both truths.

Evidence

Strongest first. Funding noted per source. The honest structure here is: the shape of the curve (front-loaded) is Tier 1; the exact numbers are Tier 2 and the observational mortality magnitudes are partly confounded (see Counter-Check).

1. All-cause mortality — the curve is front-loaded (strong, convergent).
A 2019 dose-response systematic review found small amounts of activity (0.1–7.5 MET-hours/week — well below guideline) associated with a ~20% lower all-cause mortality vs. sedentary; the protective effect is already significant at 1–499 MET-minutes/week (government/academic-funded; observational). The flagship analysis in the field (Arem et al. 2015, JAMA Internal Medicine, pooled ~661,000 adults, NIH/government-funded, observational) found that people doing some but below-guideline activity had ~20% lower mortality than the inactive, those at the guideline minimum ~31% lower, and the curve plateaued at ~3-5× the guideline with no added benefit and a hint of attenuation beyond. The first increment is where the steepness lives.

2. Steps — meaningful benefit well below "10,000" (strong, recent).
The 2025 Lancet Public Health systematic review and dose-response meta-analysis found inverse non-linear associations for mortality, CVD, dementia and falls, with inflection points around 5,000-7,000 steps/day; 7,000 vs. 2,000 steps/day was associated with ~47% lower all-cause mortality (academic-funded; observational pooled cohorts). Earlier work (Paluch et al. 2022, 15-cohort meta-analysis, government/academic-funded) put the benefit plateau at 6,000-8,000 steps for adults ≥60 and 8,000-10,000 for under-60s — i.e. benefit accrues steeply long before 10,000, and 10,000 was never an evidence-derived target (it originated as a 1960s Japanese pedometer marketing slogan, manpo-kei = "10,000-step meter"). A 2024 umbrella review identified a protective signal from as low as ~3,100-4,000 steps/day.

3. Strength training — the mortality benefit peaks at ~30-60 min/WEEK (strong, and surprising).
Momma et al. 2022 (British Journal of Sports Medicine, systematic review/meta-analysis, academic-funded, observational): muscle-strengthening activity associated with 10-17% lower all-cause mortality, CVD, total cancer and diabetes, with a J-shaped curve peaking at ~30-60 minutes per week and the benefit attenuating or reversing beyond ~130-140 min/week. Shailendra et al. 2022 (Am J Prev Med, resistance-training-specific meta-analysis, academic-funded) found ~15% lower all-cause mortality, lowest risk at ~40-60 min/week, diminishing beyond. This is one of the most counter-intuitive findings in the literature: for mortality (not maximal hypertrophy), a very small strength dose captures the benefit.

4. Strength/hypertrophy gains — the first set does most of the work (RCT, mechanistic).
Resistance-training dose-response meta-regressions (Schoenfeld, Androulakis-Korakakis et al., and the 2024-25 meta-regressions) show: (a) gains rise with weekly set volume but with steep diminishing returns, especially for strength (Androulakis-Korakakis 2020 found a single hard set near-failure, done 2-3×/week, produced the large majority of strength gains in trained and untrained lifters; multiple sets added more, with a shrinking marginal return); (b) hypertrophy keeps responding to volume longer than strength does, but even there ~4-6 hard sets/muscle/week captures most of the realistic gain for non-competitors; (c) frequency matters less than total volume. Untrained people get near-maximal early gains from a single set per exercise (academic/independent-funded RCTs and meta-analyses).

5. Cardiorespiratory fitness — low-volume interval work works (RCT).
Low-volume HIIT (≤5 min of hard work inside a ≤15-min session) reliably raises VO2max vs. untrained controls and produces aerobic and metabolic adaptations resembling much larger volumes of moderate continuous training (Gibala lab and others; meta-analyses, government/academic-funded RCTs). VO2max — cardiorespiratory fitness — is itself one of the strongest measured predictors of mortality (Mandsager 2018, Cleveland Clinic, ~122,000 patients, treadmill-tested, academic-funded): the gap between low and below-average fitness is larger than the gap between smoking and not, and unlike self-reported activity it is far harder to confound.

6. Exercise "snacks" / VILPA — minutes, not sessions (emerging, striking).
Stamatakis et al. 2022 (Nature Medicine, ~25,000 non-exercisers, wearable-measured, academic-funded, observational): 3-4 minutes/day of vigorous intermittent lifestyle activity (stair-climbing, carrying shopping, brisk uphill walking) associated with ~40% lower cancer mortality and ~50% lower CVD mortality vs. none. Small, accelerometer-measured (so less recall bias than questionnaires), but observational and prone to the confounds below — treat the magnitude as a hypothesis, the direction as well-supported.

Mechanism

Why the curve is front-loaded — why the first bit of movement is worth so much more than the last:
• You are rescuing a physiology that is failing from disuse, not optimising one that is fine. Sedentariness is a distinct pathological state — mitochondrial down-regulation, insulin resistance, endothelial dysfunction, sarcopenia, low heart-rate variability, suppressed lipoprotein lipase from chronic sitting. The first dose of movement switches systems back on: a single bout of exercise improves insulin sensitivity for ~24-48 hours and acutely raises lipoprotein lipase activity. The marginal physiological "switch-on" from zero is enormous; once the switches are on, further volume offers smaller refinements.
• Cardiorespiratory fitness has a steep low end. The untrained have the most headroom; VO2max trainability is largest where fitness is lowest. Going from unfit to merely below-average closes the deadliest part of the fitness gradient.
• Strength: the neural gains come first. Early strength improvement is dominated by neural adaptation (recruitment, rate coding, coordination), which is captured by a small number of hard efforts. Hypertrophy — the slower, volume-hungry process — is what needs more sets, and even it plateaus.
• The recovery framing fits the data exactly. This is not "how much can I push to maximise" — it is "what is the least I can do to keep these systems out of failure." The dose-response curve is the empirical shape of recovery to baseline: most of the protection is in returning to function, not in chasing the far tail.

This is also why the J-shape appears at the top: past a point, training load stops being recovery and becomes a chronic stressor (see Risks and the J-curve).

Risks And Contraindications

• The minimum is a floor, not a prescription to stay there. Misreading "minimum effective dose" as "optimal dose" undertrains people who want strength, body composition, or athletic capacity — for those goals the minimum is genuinely insufficient. Hold the distinction.
• The J-curve at the top end (real, but the top end is far away for almost everyone). Very high chronic endurance volume (commonly cited >10 hours/week of moderate-vigorous, or >45 MET-hours/week) is associated in observational data with higher rates of atrial fibrillation, accelerated coronary artery calcification, and myocardial fibrosis in veteran endurance athletes. For the minimum-dose audience this is irrelevant in practice — but it is the honest counterpart to "more is always better," and it is consistent with the recovery framing: past a point, load stops being recovery and becomes a stressor.
• Deconditioned / cardiac-risk / older novices: sudden vigorous exertion in a previously sedentary person transiently raises acute cardiac event risk; the chronic benefit is large but onboard gradually (start with brisk walking and sub-maximal sets before sprint intervals or near-failure lifting). Anyone with known cardiovascular disease, chest pain on exertion, or significant risk factors should clear vigorous/HIIT work with a clinician.
• Form before load on strength minimums: "one hard set near failure" assumes competent movement. A novice should groove the pattern at low load first; near-failure with poor form on a hinge or squat is an injury route. See beginner_exercise_programming_first_4_weeks.
• Pregnancy, recent surgery, acute illness, uncontrolled conditions: individualised, clinician-guided — the general minimum doesn't override specific medical advice.

Controversy

The controversy: How much benefit does a tiny dose actually buy — and are the headline mortality numbers real or inflated?

Position A — "the minimum buys most of it" (the entry's lean, well-supported as direction): The dose-response curve is concave across essentially every dataset; the steepest segment is zero→a-little; therefore the highest-yield public-health message is "do something, anything, consistently," and the guideline target is a destination, not an entry fee.

Position B — "the minimum-dose framing is over-sold by people selling shortcuts": A real critique. "Just 7 minutes," "one set is all you need," "you don't need cardio" are marketing-adjacent claims that quietly upgrade minimum into sufficient for any goal. For hypertrophy, performance, or maximal fitness, the minimum is demonstrably not enough. Both can be true: minimum = most of the health/survival benefit; minimum ≠ optimum for every outcome.

The Funding-bias dimension (this one is inverted from the usual KB pattern):
• The intervention here is unpatentable and free — walking, a few bodyweight sets, stairs. There is no industry that profits from telling you to do less. That is unusual and worth naming: the commercial bias in fitness runs almost entirely toward more — more sessions (gyms, studios, subscriptions), more equipment, more supplements timed around volume, more coaching. The minimum-effective-dose message is commercially inconvenient to that ecosystem, which is a tell that it is probably under-promoted relative to its evidence.
• The counter-pressure: the "biohacking / time-hack" media economy does profit from minimum-dose content (clicks, books, devices), and that is where over-claim creeps in. So the bias vectors point in opposite directions and roughly cancel — which is part of why we can state the floor confidently.

Realised Position: The front-loaded shape of the curve is as close to settled as this field gets, and it is the single most useful fact for a sedentary person. We coach the floor in a confident voice — "the first twenty minutes a week is the highest-yield thing you will do for your body" — while keeping the evidence store honest about magnitude (see Counter-Check) and about the floor-vs-optimum distinction. This is squarely the recovery, not optimisation, register: the least that returns the body to function.

Counter-Check: What deflates the headline numbers

A truth platform has to say this out loud. The large observational mortality percentages (20%, 31%, 47%) are almost certainly larger than the true causal effect, for three reasons the studies themselves acknowledge:

1. Reverse causation. Sick people move less; some of "inactivity → death" is actually "incipient illness → both inactivity and death." Lagging the analysis 2-4 years shrinks the association.
2. Residual confounding. Active people differ systematically (smoking, diet, income, education, baseline health). Minimally-adjusted models show 16-24% reductions that shrink considerably after proper adjustment.
3. Genetic confounding — the strongest deflator. Finnish twin-cohort and Mendelian-randomization analyses find that within monozygotic twin pairs discordant for activity, the independent mortality benefit of activity is much weaker or absent — much of the observed association tracks shared genetics/familial factors, not the activity itself. RCTs, MR, animal models and twins have collectively failed to prove the causal magnitude that observational headlines imply.

What survives the deflation (and is why we still coach the floor with confidence):
• The effect is widely judged causal but smaller than headline — direction robust, magnitude trimmed.
• Cardiorespiratory fitness and muscular strength as measured outcomes predict mortality far more robustly and are much harder to confound than self-reported activity — and these are exactly what the minimum dose moves. The mechanism (insulin sensitivity, VO2max, strength, endothelial function) is well-characterised and direct, not merely correlational.
• Even a deflated effect on the largest causes of death is enormous in absolute public-health terms, and the cost of the floor is near-zero. The expected value of "do a little" stays overwhelmingly positive even under the pessimistic causal estimate.

So: we do not quote "47% lower mortality" as a personal promise. We say the curve is front-loaded, the mechanism is real, the fitness/strength gains (which are causal for the user) come fast, and the floor is cheap.

Cross-Pillar Connections

• Physical (physical_foundations_for_baseline, physical_progressive_overload, resistance_training_and_body_composition, physical_zone2_cardio, sprinting_highintensity_interval_training, beginner_exercise_programming_first_4_weeks): this entry is the floor; those are the progression and modality detail once the floor is consistent. overtraining_recovery_management is the J-curve's home of record.
• Mental (exercise_mental_health): the mood/anxiety/sleep returns of exercise appear at low doses and fast (often same-day) — for many users the felt mental benefit is the adherence hook that the invisible mortality benefit can never be.
• Sleep (physical_activity_sleep): even minimal regular movement improves sleep onset and depth, and the timing/intensity caveats live there.
• Tier 0.5 (path_to_baseline_protocol): the minimum dose is the physical-pillar expression of "return to baseline function" — the least that keeps the system out of disuse failure.

What would change our mind

Falsifiability: explicit upgrade/downgrade criteria from source

Upgrade the specific numbers if: a large RCT (or a clean MR design) of a genuinely minimal dose vs. nothing reproduces the observational mortality magnitudes; or the strength-mortality J-curve (peak at 30-60 min/week) is confirmed in trials rather than only cohorts.

Downgrade / revise if: twin/MR/RCT evidence continues to shrink the causal mortality effect toward null (we would keep the fitness/strength and quality-of-life claims, which stand on direct measurement and RCTs, and explicitly retire the mortality-percentage framing); or if low-dose work were shown not to produce durable fitness/strength change in real-world adherence conditions.

Already changed our mind once: earlier drafts of this kind of advice leaned on "150 minutes is the bar." The strength-mortality data (peak benefit at 30-60 min/week) and the step data (benefit plateau well below 10,000) both under-shoot the legacy guideline, so the honest floor is lower than the official target — and the 10,000-step number was retired entirely as a marketing artefact.

Industry bias note

Structural incentives the evidence base may reflect

Bias risk on the intervention is LOW but structurally interesting, and it runs opposite to the usual KB pattern. The minimum effective dose is free, unpatentable, equipment-optional. No revenue model rewards telling people to do less. The fitness, supplement, and apparel industries are aligned with more — more sessions, more gear, more programming — so the minimum-dose message is under-promoted relative to its evidence, not suppressed by a single villain but simply unbacked by anyone with a marketing budget (the classic "plausible mechanism + commercially inconvenient" signature of an underfunded-not-disproven truth).

The one genuine over-claim vector is the time-hack / biohacking content economy, which profits from "do almost nothing and thrive" hooks and quietly inflates minimum into optimal. Realised's job is to take the under-promoted true core (the floor is high-yield) without importing the over-promoted false extension (the floor is all anyone ever needs).

Minimum Effective Dose