Two hours a week is a modest slice of time. In a new analysis following 147,374 people across three long-running US cohorts, it was also close to the point where the relationship between strength training and longevity looked strongest.
Participants who reported 90 to 119 minutes of resistance exercise a week had the lowest observed risk of dying during follow-up. Their relative hazard of death from any cause was 13% lower than that of participants who reported none. Beyond roughly two hours, the curve flattened: people training longer still had lower mortality than non-lifters, but the association did not become stronger.
That sounds like a “sweet spot,” and at the population level it may be one. But the phrase needs a quiet footnote. This was an observational study, not an experiment that assigned people to different training schedules. It found an association around 90 to 120 minutes. It did not discover a universal dose that guarantees a longer life.
What the researchers actually tracked
The study, published in the British Journal of Sports Medicine, brought together 31,540 men from the Health Professionals Follow-up Study and 115,834 women from the Nurses’ Health Study and Nurses’ Health Study II. The analytical windows began in 1992 for the men, 2002 for the older nurses and 2003 for the younger nurses. Follow-up continued through 2021 or 2022, depending on the cohort.
That is why “30-year study” is accurate but should not be read as every participant being observed for exactly 30 years. The longest-running cohort contributed up to three decades. Across all three cohorts, researchers documented 35,798 deaths.
The design had an advantage over studies that ask about exercise once and never ask again. Participants reported their resistance training repeatedly, every two to four years, and the researchers calculated cumulative averages. That approach can better reflect a person’s long-term pattern and soften the effect of an unusually active or inactive week. The wider Nurses’ Health Study program has used this repeated-questionnaire model to follow health and lifestyle over decades.
Weekly resistance exercise was grouped into none, 1 to 29, 30 to 59, 60 to 89, 90 to 119 and at least 120 minutes. The analysis then estimated mortality after accounting for measured differences including age, smoking, alcohol, diet, aerobic activity and family history.
The study’s bibliographic record identifies the work as a prospective cohort analysis. That description matters: the investigators watched habits and outcomes unfold rather than controlling who trained and for how long.
What the 13% figure means, and what it does not
The clearest result was a hazard ratio of 0.87 for all-cause mortality in the 90-to-119-minute group, with a 95% confidence interval from 0.81 to 0.95. That translates to a 13% lower relative hazard during follow-up compared with doing no resistance training.
It does not mean that strength training adds 13 years to life. It also does not mean that everyone’s absolute chance of dying falls by 13 percentage points. A hazard ratio compares the rate at which an outcome occurs over time between groups. Its practical meaning for one person depends on age, baseline health, follow-up time and many other risks.
The group reporting at least 120 minutes a week had a hazard ratio of 0.92, or an 8% lower relative hazard than non-lifters. In other words, the study did not find that crossing the two-hour line erased the association or made training harmful. It found no additional mortality benefit beyond 90 to 119 minutes.
That is a narrower conclusion than the headline-friendly idea that “more is worse.” It also places the new article alongside ScienceBlog’s earlier report on the same result, while adding an important distinction: a flatter mortality curve says nothing about whether extra training improves strength, muscle mass, function or performance.
The pattern changed with the cause of death
The lowest points were not identical for every outcome. At 90 to 119 minutes a week, cardiovascular mortality was 19% lower than in the no-training group, with a hazard ratio of 0.81. Mortality from neurological diseases was 27% lower, with a hazard ratio of 0.73.
Cancer mortality looked different. Its statistically significant associations appeared only in the 1-to-29-minute and 30-to-59-minute groups, where relative hazards were 9% and 12% lower. The estimates at higher training levels were not clearly different from the no-training group.
Those uneven curves are a useful restraint on simple stories. “Premature death” is an umbrella outcome. Cardiovascular disease, cancer and neurological disease have different biology, different time courses and different vulnerabilities to measurement error. A single weekly number cannot be assumed to affect every cause in the same way.
Aerobic exercise remained the larger part of the picture
Resistance training did not operate in isolation. When the researchers examined it together with aerobic exercise, people reporting both tended to have lower mortality than those doing either type alone.
The strongest joint association appeared among participants doing 30 to less than 45 MET-hours of aerobic activity plus 60 to 119 minutes of resistance work each week. Compared with people below 7.5 MET-hours and doing no resistance training, their hazard ratio was 0.55, a 45% lower relative hazard.
Aerobic activity on its own was associated with larger reductions than resistance training on its own. At the highest aerobic level, at least 45 MET-hours a week, mortality estimates were similarly low whether participants reported no resistance exercise or up to two hours. This does not make lifting irrelevant. It suggests that the broader movement pattern mattered more than any single gym target.
That fits current US guidance, which recommends aerobic activity and muscle strengthening on at least two days each week. The official Physical Activity Guidelines questions and answers describe frequency, not a mandatory number of lifting minutes. The World Health Organization takes a similar approach, emphasizing both aerobic movement and strengthening of major muscle groups.
Why the study cannot assign the extra years to lifting
People were not randomly assigned to train for zero, 45, 100 or 150 minutes. They chose their own habits, and those choices traveled with other differences. Participants reporting more resistance training were generally younger, had lower body mass indexes, smoked less, ate healthier diets and did more aerobic exercise.
The statistical models adjusted for a substantial list of those factors. Researchers also excluded people with cancer, heart disease or stroke at baseline, delayed the mortality analysis by two years and tested longer delays to reduce the chance that undiagnosed illness caused people to stop training shortly before death.
Those are serious safeguards, but they cannot eliminate healthy-user bias or reverse causation. A person may lift because they are already healthy enough to do so. Better access to healthcare, stronger social support or an unmeasured feature of daily life could influence both training and longevity.
Measurement is another limitation. Minutes were self-reported, and the questionnaires did not capture load, sets, repetitions, effort, rest time or whether respondents counted an entire gym visit rather than active lifting. Calisthenics and Pilates were not included. The sample was about 78% women, roughly 96% white and composed largely of health professionals, which limits confidence that the precise curve applies everywhere.
A useful range is not a magic prescription
Earlier research has not produced one universal number. A 2022 systematic review and meta-analysis found its largest estimated reduction in mortality at roughly 30 to 60 minutes of muscle-strengthening activity a week. Differences in populations, questionnaires and statistical methods can move the apparent bottom of a dose-response curve.
The new analysis is larger and benefits from repeated measurements, but it still cannot tell someone how many sessions to schedule, how heavy to lift or which exercises suit their body. Nor does a mortality plateau imply that time beyond two hours is wasted. People train for reasons the study did not measure, including stronger bones, more muscle, greater independence, rehabilitation, sport and simple enjoyment.
The most defensible reading is humble. Regular resistance exercise was associated with a lower risk of premature death, its lowest point in this dataset was around 90 to 119 minutes a week, and combining strength work with aerobic activity looked better than treating either as the whole answer.
That makes two hours a useful landmark, not a cliff edge. The study offers a clearer contour of population risk. It does not turn a stopwatch into a prescription.