26/09/2026
The message is still the same - keep moving 🏃♀️💪🍎
There is a dose of resistance training where the mortality curve stops responding. It arrives earlier than most people expect.
Three long-running cohorts of American health professionals were pooled: the Nurses' Health Study, its second wave, and the Health Professionals Follow-Up Study. Together, 147,374 adults followed up to thirty years, with 35,798 deaths. What separates this from most physical activity research is that training time was reported every two years across three decades rather than once at enrolment, so the analysis tracks a habit as it changes rather than a single answer given long ago.
Against people who reported no resistance training at all, those doing 90 to 119 minutes a week had a 13 percent lower rate of death from any cause. Death from cardiovascular disease was 19 percent lower. Death from neurological disease, which covers dementia and Parkinson's, was 27 percent lower and was the largest single drop in the paper. Above roughly 120 minutes a week the curve flattened, and more hours were not associated with any further reduction.
The result the coverage has left out concerns cardio. The authors also examined aerobic activity and resistance training together. Among people already doing a high volume of aerobic activity, mortality ran roughly 42 to 47 percent lower whether they lifted a lot, a little, or not at all. Put the two side by side and the asymmetry is hard to miss: the resistance training association tops out around 13 percent, the aerobic one at more than three times that. Lifting added further reduction at every aerobic level below the threshold, but not at the top of it.
This is observational, and nobody was assigned to lift. People who train twice a week differ from people who never train in income, education, sleep, diet and baseline health, and no statistical adjustment fully removes that. Training was self-reported in minutes, so ninety minutes describes time in a gym rather than a stimulus. The sweet spot is also unstable across the literature: a 2022 pooling of sixteen prospective cohorts found the same J-shaped pattern but placed the maximum benefit at roughly 30 to 60 minutes a week, less than half of what this analysis reports. And this cohort is overwhelmingly white, educated American health professionals.
The flattening is the part most likely to be misread. A plateau in observational data is not proof that additional training does nothing. Very few people in these cohorts trained more than two hours a week, so estimates out at the tail rest on small numbers and become imprecise. What can be said is that the data show no further reduction beyond that point, which is not the same as showing there is none.
If the reason you have not started lifting is a belief that it demands four sessions a week, this data does not support that belief. Two sessions of about 45 minutes sits inside the range where the entire association was observed. If you already train more than that, nothing here suggests cutting back. It suggests those extra hours are buying strength, muscle and capability, all worth having, rather than additional movement on this particular curve. And if you are choosing between adding a third lifting session and adding a walk, the joint analysis points at the walk.
Same finding, two readings. The dose is smaller than expected, and the ceiling arrives sooner.
Zhang et al., Br J Sports Med 2026;60(12):874-883 · PMID 42230125
Momma et al., Br J Sports Med 2022;56(13):755-763 · PMID 35228201
Fukushima et al., J Am Med Dir Assoc 2024;25(3):417-430 · PMID 37925162
