Past 120 minutes, the benefit plateaued. Pairing it with cardio made the effect stronger, according to a BMJ analysis that tracked 147,374 people for up to 30 years.
Two hours a week of strength training, not five, not ten, appears to mark the ceiling for longevity gains, according to three decades of data on more than 147,000 people.
A new analysis of three long-running cohorts, summarized by ScienceDaily from a BMJ Group press release, tracked participants for up to 30 years. The underlying paper was published online in the British Journal of Sports Medicine. The signal is sharp: roughly 90 to 120 minutes per week of strength or resistance training was associated with the largest reduction in all-cause mortality. Past 120 minutes, the benefit stopped climbing.
That ceiling is the part most readers will not expect. The standard fitness message is that more is better, and that the people who train longest live longest. This dataset suggests a different shape. The relationship between weekly strength training and lower mortality risk looked like a plateau rather than a ramp, and the dose at which it flattened was surprisingly modest. Two short sessions a week lands inside it.
The mechanism here is correlation, not prescription. The cohorts were built largely from health professionals, and roughly 78 percent of participants were women (31,540 men and 115,834 women across the Health Professionals Follow-up Study, Nurses' Health Study, and Nurses' Health Study II), so the cleanest reading is that for that population the 90-to-120-minute band carried the strongest signal. A reader outside that demographic can use the range as a target worth aiming for, not a guarantee.
Cause of death mattered. Within that 90-to-120-minute band, the mortality reductions were most pronounced for cardiovascular and neurological disease. That level of exercise was linked to a 19 percent lower risk of death from cardiovascular disease and a 27 percent lower risk of death from neurological disease. The study did not isolate why strength training moved those numbers, but the pattern matches what cardiologists and neurologists have argued for years: muscle mass, insulin sensitivity, blood pressure, and gait stability are all variables these conditions share, and resistance work touches each of them.
Aerobic exercise did not get replaced. Participants who combined strength training with cardio saw larger effects than those who did either alone, which is the second piece of news hidden in the same dataset. The implication is additive, not competitive. A reader who already walks, runs, or cycles has a base to build on. Adding two short strength sessions on top of that is associated with more of the same benefit, not a duplicate of it.
The shape of the dose-response matters. A J-shaped curve, where risk falls as dose rises and then bends back up, is the worry whenever a "more is better" claim is on the table. The BMJ analysis points in the other direction: more is simply not better on this metric. There is a useful read of that for the reader who is already wondering whether to add a third or fourth session. The data says the marginal hour is not where the longevity signal lives.
What counts as strength training, in the study's terms, is also broader than a gym membership. Weights count, but so do press-ups, squats, and lunges, the bodyweight movements most readers can do at home in twenty minutes. That lowers the practical bar for hitting the target, and it widens the group of people for whom the dose is reachable on a normal week.
One thing the ScienceDaily aggregation did not fully pin down: the underlying BMJ paper's full hazard ratios, confidence intervals, and the precise shape of the dose-response curve were not available in the press-release summary. The ScienceDaily article carries the quantitative figures (the 19 percent and 27 percent disease-specific reductions, the 147,374 participant count, the 31,540 men and 115,834 women demographic split) but the primary paper's full statistical appendix remains the definitive source for the most granular dose-band cut-points and hazard ratios. The 90-to-120-minute window and the plateau above it are hypothesis-generating observational findings, not clinical guidelines, and should be read that way.
For the reader, the practical version is short. Two short strength sessions a week, with whatever equipment is on hand, paired with whatever aerobic habit is already in place, lands inside the dose associated with the lowest mortality in a 30-year dataset. Past two hours, the data does not reward the extra effort on this measure. The ceiling is the news, and it is also permission to stop.