Ninety minutes a week of strength training was the sweet spot in 147,374 people. More didn't help; none was worse.

The largest long-term study of strength training and death was published today in the British Journal of Sports Medicine: 147,374 health professionals in three American cohorts, followed for up to thirty years, with 35,798 deaths and exercise reported every two years. Compared with people who did no resistance training, those doing 90 to 119 minutes a week had a 13% lower hazard of death from any cause, 19% lower from cardiovascular disease and 27% lower from neurological disease. Above about two hours a week the benefit flattened — more lifting did not buy more years. And the combination mattered most: people doing a solid amount of aerobic activity plus one to two hours of strength work had a hazard of death roughly 45% lower than the least active. Only 46% of participants did any strength training at all.

The caveats, in order of size. It is observational, and the authors say so first: this cannot establish cause and effect. People who lifted were younger, lighter, healthier and more aerobically active to begin with, which is exactly the confounding a study like this cannot fully remove. Exercise was self-reported. The sessions' intensity and content are unknown. And the figures are relative: a 13% lower hazard is 13% of whatever a healthy, well-educated health professional's risk was, which is already low. What survives the caveats is the shape, which matches everything else we know — the first hour of lifting a week is worth far more than the fourth.

Ben’s take

I lift considerably more than two hours a week and compete in powerlifting, and this paper says the hours past the first two are for me, not for my lifespan. I am at peace with that. The reason to keep training is that I love it and the people I train with, which is also, not coincidentally, why I have kept doing it for decades. The paper's lesson is for the 54% who do none: one or two sessions a week is where almost all of the measurable benefit lives, and it is the easiest exercise prescription in this literature to follow.

Loree’s take

Two sessions a week, each shorter than an episode of television, with a resistance band, a pair of dumbbells or your own body weight. That is ninety minutes. The patients who need it most are the ones who have never lifted anything and think the gym is for someone else. This study is theirs: the benefit is front-loaded, the plateau is low, and the thing that pairs best with it is a walk.

Why it matters to a recovering optimizer. The movement fundamental has a dose-response curve like every other one on this site: steep at the start, flat after a modest amount, and best done in combination. Nobody needs a program, a periodization app or a coach to reach ninety minutes a week. They need two appointments with something heavy and the willingness to keep them.