Article — "Resistance Training's Sweet Spot: Converging Cohorts Point to Roughly Two Hours a Week": Article — "Resistance Training's Sweet Spot: Converging Cohorts Point to Roughly Two Hours a Week": Two sources in your evidence pack tell you almost opposite things about strength training and one of the deadliest diseases we track.
One says: do more, live longer — a clean dose that drops all-cause mortality 13 percent, cardiovascular death 19 percent, and neurological death 27 percent, before flattening above two hours a week.
The other, buried in the same study, says: for cancer, only the smallest doses helped. One to twenty-nine minutes a week reportedly cut cancer death 21 percent, per Medical Daily. Push past an hour and the benefit vanishes.
The same knife, apparently, cuts two different directions depending on what it is aimed at. Hold onto that contradiction. It is the crack running through an otherwise irresistibly tidy story.
The same knife, apparently, cuts two different directions depending on what it is aimed at. It is the crack running through an otherwise irresistibly tidy story.
And the story is tidy. A Harvard-led analysis pooled three long-running cohorts — the Health Professionals Follow-Up Study and the two Nurses' Health Studies — following roughly 147,000 adults for up to 30 years, over which reportedly nearly 36,000 died.
Out of that mountain of data came a number precise enough to print on a dumbbell: 90 to 119 minutes a week. Above 120, the curve flattens. More is not better.
Then a second, independent line seems to nod along. In the CHARLS nationwide cohort of 7,258 Chinese adults followed a mean 7.8 years, the weakest grip-strength quartile carried a 37 percent higher risk of new cardiovascular disease (HR 1.37, 95% CI 1.17–1.61), and the slowest chair-rise quartile reportedly a 53 percent higher risk, per the CHARLS study.
Add the mechanism. Skeletal muscle absorbs roughly 80 percent of the glucose your body clears after a meal, and contracting muscle releases myokines that dampen the chronic inflammation underlying heart disease, diabetes, and some cancers.
Three streams — epidemiology, a strength-mortality cohort, muscle biology — converging on the same conclusion. That is exactly the kind of convergence that should make a careful reader more suspicious, not less.
Start with what the headline number is actually made of. Participants reported their exercise on questionnaires every two to four years. That is the raw material behind "90 to 119 minutes."
A person filling out a form in 2003, recalling a typical week, is not a calibrated instrument. Averaging many such reports over decades reduces noise, but it cannot manufacture a precision the underlying measurement never had.
The study's own defenders concede this. As the ScienceDaily writeup puts it plainly: the work was observational, "it can't prove that one directly causes the other," strength training "was self-reported, and the study couldn't capture how hard people trained."
That last clause quietly demolishes the dosing-chart framing. There is no measure of intensity anywhere in the number.
Ninety minutes of drifting through machines at a social pace and ninety minutes of hard, progressive loading are counted identically. A prescription that ignores intensity is not a dose in any pharmacological sense — it is a time budget.
And there is no randomized controlled trial behind any of this. Not one. The 90-to-119 figure has never been tested by assigning people to lift and comparing them against people told not to.
What RCT evidence does exist in this pack points somewhere else entirely. The meta-analysis of 23 randomized trials in adults with diabetes and obesity found resistance training better than aerobic exercise at preserving fat-free mass and improving body fat, while aerobic work won on fasting glucose and fitness — with most outcomes statistically indistinguishable between the two.
That is a real, trial-grade finding: the two modalities are largely complementary. It says nothing about an optimal weekly minute count for longevity, because no trial has asked that question.
Now return to the cancer signal, because it is where the false precision breaks most visibly. If strength training reduced mortality along a smooth biological dose-response, you would expect cancer to behave like cardiovascular disease — more protection with more training, up to a plateau.
Instead it inverts. Protection appears only at 1 to 59 minutes and evaporates above that. The researchers themselves flag a plausible culprit: higher training volumes may elevate IGF-1, a growth factor linked in prior work to certain cancers.
Take that seriously as a methodological warning, not a mechanism-of-aging thread. If a single, coherent dose-response cannot even hold its sign across different causes of death, then "90 to 119 minutes" is not one clean biological optimum. It is a summary statistic that happens to describe cardiovascular and neurological mortality while directly contradicting cancer.
There is also the healthy-user problem, which no amount of statistical adjustment fully dissolves. People who lift weights consistently for thirty years differ from people who do none in ways questionnaires never capture.
The authors adjusted for age, diet, smoking, body weight, alcohol, and aerobic activity. That is diligent. It is also why the finding is worth taking seriously as an association — and exactly why it cannot be read as a law of dosing.
So what actually survives this scrutiny? Quite a lot, in fact — just not the number.
The direction is robust and repeated across independent designs. Strength, and the training that builds it, tracks with lower mortality and lower cardiovascular risk. The CHARLS grip-strength and chair-rise gradients say the same thing in a different population, using measured function rather than recalled minutes.
The combination effect survives too. Across the Harvard cohorts, pairing strength work with regular aerobic activity was associated with up to a 45 percent lower risk of death — with aerobic exercise still doing most of the heavy lifting as the stronger independent predictor.
And the shape of the benefit at the low end survives, which matters more than the peak. People doing just 1 to 29 minutes a week already had lower mortality than those doing nothing. The steepest part of the curve is the climb off zero.
That is the finding with real public-health weight, and it points at a scandal the sweet-spot framing obscures: fewer than 25 percent of American adults meet guidelines for both aerobic and muscle-strengthening activity.
The problem in the population is not people anxiously overshooting 119 minutes. It is the vast majority sitting at zero strength work while cardio gets top billing.
For decades, physical activity guidance treated muscle-strengthening as an afterthought bolted onto an aerobic prescription. The convergence here — imperfect data and all — argues that under-prescription was the real error.
So the honest reading is not "hit 90 to 119 minutes." It is: do meaningfully more strength work than zero, keep doing your cardio, and understand that the largest returns come from starting, not from optimizing.
Treat the two-hour figure as a reasonable ceiling for reassurance — evidence that you do not need to live in a gym — rather than a target calibrated to the minute.
The number is seductive precisely because it feels like precision medicine. It is not. It is a well-measured association wearing a lab coat.
The defensible prescription is older, blunter, and better supported: some strength is dramatically better than none, more than none is where nearly everyone should aim, and pairing it with aerobic exercise is where the biggest numbers live. That message does not fit on a dumbbell. It has the advantage of being true.