Applies to a specific group
This card is written for a sedentary, untrained middle-aged adult with obesity, in a developed country, whose goals are quality of life, physical function, and keeping function into old age. If you are not in it, the recommendations here may not transfer — the population-level picture can differ.
Lift twice a week, do about 150 minutes a week of moderate, low-impact aerobic exercise, and walk more on top of that. Start at about half that for the first eight weeks, then keep the full program going for good. Each piece is at or near the lowest dose that trials found reliably works for its own outcome. Training probably improves physical function and quality of life too. The bet is cheap and reversible, and it targets inactivity, one of the largest risks a sedentary adult can change.
One caveat matters most: no trial has tested this exact package for quality of life, or for keeping function into old age, in middle-aged adults with obesity. Fitness and strength are what you can measure along the way; what you can do in daily life is the goal. -> Resistance Training Prescription - Load Sets and Frequency, Measuring and Raising Cardiorespiratory Fitness, Physical Activity Dose and Mortality, Acting on a Direction Without a Magnitude
Start with strength twice a week, moderate aerobic work, and more steps
Plain numbers trace to the evidence; a † marks a practical choice the evidence leaves open.
| Session | What | How long, how often | How hard |
|---|---|---|---|
| Strength A and B | 5-6 whole-body exercises†, 2-3 sets each | 2x/week on non-consecutive days†, about 30-40 min† | Moderate loads for weeks 1-8†, then build toward heavy; set by reps |
| Moderate aerobic | Brisk walking, stationary cycling, rowing or water exercise: whichever you will keep doing | About 150 min/week, e.g. 5 x 30 min or 3 x 50 min† | About 50% of heart-rate reserve |
| Everyday movement | More steps, less sitting, any intensity | Daily | Step count on a phone or tracker |
| Optional: intervals | Swap one aerobic session for intervals | After 8-12 weeks† | Hard effort; not part of the minimum |
Strength, in detail. 5-6 whole-body exercises†: leg press or sit-to-stand, hip hinge, push, pull, loaded carry. 2-3 sets each (the third set is †). Machines or free weights. Weeks 1-8†: moderate loads. Then build toward heavy. No maximum-lift test is needed†: set the load by reps. Early on, aim for 12-15 reps per set; later, heavier loads for about 5-8 reps per set†. Stop 1-2 reps short of failure†. When you reach the top of the rep range with more than about 2 reps still in hand, add the smallest step next session†. Log the load.
Trials label these loads as below and above 80% of a one-repetition maximum (1RM). A network meta-analysis of 178 strength trials converted rep maxima to %1RM as 100 minus 2.5 per rep, which puts an 8-rep maximum at 80% (Currier et al., 2023). That is a coding rule, not a tested conversion; how much it varies by person or exercise is not known here
Aerobic effort, in detail. About 50% of heart-rate reserve: a brisk effort, breathing harder but still able to talk†. A heart-rate monitor using an age-estimated maximum is a rough check, not a requirement†. On a heart-rate-lowering drug such as a beta-blocker, go by perceived effort instead
How the first six months run:
- Weeks 1-8, build-up. About half the aerobic minutes† and lighter loads, raised in small absolute steps.
- Weeks ~8-24, full program. The whole template, every week.
- Month 6 onward, keep going. Same weekly minutes at the same heart-rate effort. As you get fitter, that effort takes more resistance or pace, which is the progression. No smaller maintenance dose has been tested -> Detraining and Residual Effects in Older Adults, Measuring and Raising Cardiorespiratory Fitness.
Why each piece is in the program, and what you can swap
Even light movement counts, and the least active gain the most from starting. In device-measured cohorts, people in the second-least-active quarter had less than half the mortality hazard of the least active (HR 0.48), and light activity on its own showed a similar gradient (HR 0.60 for its second quarter) (Ekelund et al., 2019). That is why the template counts steps as well as sessions. Steps are the simplest count: risk kept falling up to about 8,000-10,000 steps a day in adults under 60 (Paluch et al., 2022). Both findings are observational associations. -> Physical Activity Dose and Mortality
Do the aerobic work at a moderate effort; intervals are optional. For adults below 10 METs of fitness, about 50% of heart-rate reserve is enough for a gain of at least 1 MET; at 10-14 METs, 65-85% is likely more effective (Ross et al., 2016), a step taken for more fitness, not one this program requires.
The trial closest to this reader enrolled 300 adults with obesity. Its 73-person arm doing about 30 minutes five times a week for 24 weeks raised fitness by roughly 8-9% (Ross’s table gives 7.7%, its text 9.4%) (Ross et al., 2016). A pooled analysis of exercise in adults with obesity saw gains in the same direction but no statistically significant improvement (O’Donoghue et al., 2020), so expect a modest rise. -> Measuring and Raising Cardiorespiratory Fitness
Pick the mode you will keep doing and are least likely to be hurt doing. Aerobic, resistance and combined training each raised fitness in postmenopausal women, in an analysis of 35 training arms from 25 studies (Khalafi et al., 2023); the review did not test the modes against each other.
In 103 trials of knee and hip osteoarthritis, every exercise type beat usual care on pain and function. Aerobic and mind-body exercise had the largest effects, but the single types were not shown to differ from one another. Programs mixing more than one type ranked last, and worse than aerobic or mind-body exercise for pain (Goh et al., 2019). That mixed group also took in trials that did not name a single exercise type, so it does not show that combining modes harms; for knee and hip osteoarthritis it gives no reason to expect a combined program to beat a single mode . The template pairs lifting with aerobic work because each serves a different outcome, not because the pair outranks either alone.
Joint-friendly modes work, and trials in heavier groups showed no smaller benefit. Stationary cycling matched other exercise for knee osteoarthritis (Luan et al., 2020), and water exercise did at least as well as land exercise on balance and walking tests in older adults (Deng et al., 2024). In a companion pooled analysis of knee and hip osteoarthritis from the same research group, the pain benefit was no smaller in trials whose participants averaged a BMI of 30 or more (0.56 versus 0.46 standardized mean difference, P = 0.78) (Goh, Persson, Stocks, Hou, Lin, et al., 2019); a trial-level comparison like this cannot rule out a difference. -> Is the Exercise Modality Doing Any Work, Exercise Therapy for Osteoarthritis Pain and Function
Lift twice a week, two or more sets per exercise. Across 178 strength trials, at least two sets or two sessions a week raised strength, and at least two sets and two sessions raised muscle size, each at 95% probability (Currier et al., 2023). That is the lowest dose the analysis was confident about, not a measured optimum. Which program you pick matters less than starting one: every well-studied prescription beat no training, a few sparsely studied schemes were too imprecise to show it, and most programs could not be told apart from one another (Currier et al., 2023). These trials enrolled healthy adults, not specifically adults with obesity. -> Resistance Training Prescription - Load Sets and Frequency
Move toward heavy loads after the build-up; equipment is your choice. The top-ranked programs for strength used heavy loads, and those for muscle size used multiple sets (Currier et al., 2023). Tendons stiffened under loads above 70% of maximum, but not in the five lower-load programs studied (Bohm et al., 2015), another reason to work up to heavy loads. Free weights and machines built strength equally (Haugen et al., 2023). Bodyweight training has not been compared with loaded training. -> Resistance Training Prescription - Load Sets and Frequency, Tendon Adaptation to Mechanical Loading
If you are dieting, lift as well. Dieting removes muscle along with fat. In six trials of obese older adults cutting calories, adding resistance training three times a week cut lean-mass loss by 93.5% (Sardeli et al., 2018). -> Protein and Resistance Training for Muscle and Strength
Intervals may buy fitness a little faster, and feel worse. In reviews of adults with overweight or obesity, interval training raised fitness as much as or slightly more than moderate continuous work (Poon et al., 2024), but it felt worse during the session, and the hard effort, not the interval format, drove the discomfort (Niven et al., 2020). They suit someone short of time who tolerates hard effort; the minimum does not need them. -> Affective Response to High-Intensity Interval Exercise
Use exercise snacks as a way in, not a substitute. Brief vigorous bouts, lasting from about 20 seconds to 10 minutes in the 14 studies Wan reviewed (Wan et al., 2025), raised VO2max at very low certainty, and the authors judge it unlikely that inactive people will reach recommended levels through snacks alone (Wan et al., 2025). How you split the week matters less than the weekly total: a weekend-warrior pattern carried a similar mortality association to spreading activity out, at very low certainty (Kunutsor et al., 2022). -> Exercise Snacks and Cardiometabolic Health, Physical Activity Dose and Mortality
Watch your joints, plan for lapses, and keep weight expectations modest
Injury and joints
Tendons adapt more slowly than muscle, so raise the load slowly. Tendon can respond to heavy load within about 8 weeks, adapts more clearly over 12 or more, and adapts more slowly than muscle (Bohm et al., 2015). Add load in small absolute steps and judge by how you feel the next day†; the popular 10% weekly rule has no evidence behind it (Damsted et al., 2018). -> Tendon Adaptation to Mechanical Loading, Training Load Progression and Running Injury
Spare the knees first, and leave running out of the minimum†. Higher BMI and a prior knee injury both raise knee osteoarthritis risk (Duong et al., 2025). Whether running is safe for an overweight person with a prior injury and a physically heavy job remains unclear (Alentorn-Geli et al., 2017). -> Knee Osteoarthritis Incidence and Risk Factors
Lifting protects, and exercise rarely harms. Strength training cut sports injuries while stretching did not (Lauersen et al., 2013). Reviews of exercise in chronic pain found no harm beyond passing soreness, though only a quarter of the trials reported harms (Geneen et al., 2017). -> Exercise Interventions and Sports Injury Prevention
Keeping going
Plan for the week that goes wrong. Evidence the reviewers cite suggests advice to exercise, on its own, is insufficient to change much (Geneen et al., 2017). In year-long weight-loss trials, attrition averaged 29% (Franz et al., 2007). A tracker added about 627 steps a day, on low-quality evidence (Brickwood et al., 2019). Supervision helps a little: in adults over 60, supervised programs gave slightly larger gains (robustly only for knee strength) at the same 81% attendance as home programs (Gómez-Redondo et al., 2024), so a home program is a reasonable start. -> Chronic Pain and Physical Activity
A planned break does not reset you; stopping for good does. Older adults who stopped training for 4-36 weeks (mean about 11), while carrying on with normal daily life, still scored ahead of non-exercising controls on function (Buendía‐Romero et al., 2025). A layoff forced by illness, a fall or a hospital stay was not studied. In knee osteoarthritis, though, exercise benefits peaked at 2 months and were no better than usual care by 9-18 months (Goh, Persson, Stocks, Hou, Lin, et al., 2019); the authors suggest the fade may partly reflect poor adherence. -> Detraining and Residual Effects in Older Adults
Exercise alone moves weight less than the calorie count suggests. In a large energy-expenditure dataset, the body offset part of the extra burn by lowering its basal energy use: 28% on average, rising to 49.2% in people at the 90th percentile of BMI (Careau et al., 2021). Eating more can cancel out still more . If weight loss is a goal, change the diet too. -> Exercise Energy Compensation
Keeping score, and special cases
Track the measures, but judge by what you can do. Log the load lifted, time a five-rise chair stand, and check grip (Cruz-Jentoft et al., 2018). If you lose weight, track fitness in litres per minute as well as per kilogram, since the per-kilogram figure rises with weight loss alone; the pooled analysis in adults with obesity used litres per minute (O’Donoghue et al., 2020). These numbers predict outcomes; they are not the outcome. -> Surrogate Outcomes, Sarcopenia Definition and Diagnosis
A fitness test gives a range, not a number. Without a lab, a fast 2-km walk with a heart-rate strap estimates VO2max. Its authors found it relatively valid in medically screened healthy adults with and without obesity, though in their obese men the estimate ran below measured on average, and they advise a practice walk first (Oja et al., 1991). In a later sample of 410 adults aged 18-64, 95% of walk estimates landed within about 12-13 ml·kg-1·min-1 of measured VO2max (about ±3.6 METs), and 95% of repeat walks a week later agreed within about 8 (Marín-Jiménez et al., 2023). An 8-9% gain from a starting VO2max near 25 is about 2 ml·kg-1·min-1, well inside either band, so a before-and-after pair cannot show progress; for tracking, lean on the load and chair-stand log.
In a lab or clinic, a supervised maximal bike test is the other option; read its peak watts through the FRIEND equation rather than the ACSM one, which read 15% high on average. That paper reports no per-person error (Kokkinos et al., 2018). -> Measuring and Raising Cardiorespiratory Fitness
Some people need a different start. With osteoporosis, be wary of moves that bend the spine forward, such as sit-ups and some yoga poses, which were linked to rare vertebral fractures (Kunutsor et al., 2018). If you are frail, the standard program may be too intense; start lower, ideally with professional guidance. -> Frailty
Beyond the template, extra exercise buys less
Past the template, more exercise buys less. For moderate-to-vigorous activity, the mortality curve stops falling steeply after a modest daily amount (World Health Organization, 2020), a broad region of roughly 20-40 minutes a day. The template already reaches that region. More still lowers risk, just less per extra hour: 2000 versus 750 MET-minutes a week carried a mortality HR of 0.82 (World Health Organization, 2020). Fitness and strength also keep improving with more, so doing more is a choice, not a requirement. -> Physical Activity Dose and Mortality, The Underivable Optimum
No study has measured two things that matter here. None tested whether training keeps function intact into old age and then allows a short decline; trials report function over weeks or months. And none tested a smaller dose that keeps the gains, so the plan is to keep going. Most function and quality-of-life trials enrolled older, frail or arthritic people: exercise probably improved mobility and quality of life in frail adults over 65 (Racey et al., 2021). The direction plausibly carries over to this group; the size is not established for it. -> Surrogate Outcomes, Frailty
The template matches published guidance. WHO advises 150-300 minutes of moderate activity a week plus muscle strengthening on at least 2 days (Shailendra et al., 2022). This card treats that range as a sensible target where the mortality benefit has begun to level off, not as the point where benefit begins; the card and WHO differ in purpose, not in evidence. Later in life, add balance training: it carried exercise’s fall prevention in adults with a mean age of 76 (Sherrington et al., 2019). -> Exercise for Preventing Falls in Older Adults
Read this as a starting place, not a measured minimum
- A starting place, not a measured minimum. Each dose is the lowest that worked for its own outcome; the package as a whole is a reasonable bet -> The Underivable Optimum.
- Function and fitness, not weight loss. Judge it by what you can do; for weight, change the diet too -> Exercise Energy Compensation.
- Health only. Time, money and enjoyment are yours to weigh; they shape what you will keep doing.
- No drug substitutes for it on function. A drug can match exercise on blood pressure, but none builds strength -> Blood Pressure Lowering and Cardiovascular Events.
- Not graded against real-world results. This wiki checks that its claims hang together and match their sources, not whether anyone who followed the program ended up better off.
Find the full evidence on the linked pages
The claim pages linked above hold the effect sizes, confidence intervals, dose-response shapes and the trial-versus-observation reasoning behind each directive. Three pages anchor them: Physical Activity for the activity and fitness estimates, Resistance Training Prescription - Load Sets and Frequency for the strength dose, and Big Rocks (Median) for why exercise ranks among the largest levers for this group. The tendon, injury, adherence and detraining detail sits in Tendon Adaptation to Mechanical Loading, Training Load Progression and Running Injury and Detraining and Residual Effects in Older Adults.
Evidence box
Question ’For a median obese, sedentary, untrained, inactive middle-aged adult in a developed country whose objectives are quality of life, functional capacity, and a function-held-then-rapid-decline trajectory (the centenary decathlon): what does the evidence show about aerobic AND resistance training, as co-led questions, for cardiorespiratory fitness, strength, and function together; what is the SIMPLEST weekly program that could plausibly be adequate — the minimum that could work — if a dose threshold is even the right framing, across acquisition and maintenance, accounting for modality, frequency / volume / intensity and measurement, adherence, connective-tissue conditioning, injury, impact, and compensation?‘ Evidence included 36 sources — 28 gold, 5 high, 3 moderate Overall certainty Low (see Rating Certainty of Evidence) Source-selection note 3 source(s) below the gold/high evidence bar feed this page: Marin-Jimenez (method, moderate); Oja (method, moderate); Ross (narrative review, moderate). Their roles and limitations are described in the text. Last updated 2026-10-06 · Independently reviewed: No · Full edit history