Article — "Beyond Cardio: Combined Aerobic-Plus-Strength Training and the Resistance 'Sweet Spot' for Longevity": Article — "Beyond Cardio: Combined Aerobic-Plus-Strength Training and the Resistance 'Sweet Spot' for Longevity": "In addition, information was not available on the duration of individual workouts or the intensity of strength training sessions."
That's a direct admission from the authors of the June 2026 British Journal of Sports Medicine paper — buried in the limitations section, where inconvenient truths go to be politely ignored.
Yet within a week, ScienceDaily, the Washington Post, and a half-dozen aggregators had crowned a specific number the longevity sweet spot: 90 to 120 minutes of strength training per week, worth a 13% cut in all-cause mortality.
Read that sentence twice. A study that never measured how long or how hard anyone actually trained produced a headline built entirely around training duration.
A study that never measured how long or how hard anyone actually trained produced a headline built entirely around training duration.
Last week in this space, I traced the "resistance sweet spot" through its earlier incarnation — the J-shaped curve that, per News-Medical, put the low point at roughly 40 minutes a week. This week the same beat gets a bigger dataset and a sharper problem: the more you read the actual paper, the less the sweet spot looks like a prescription and the more it looks like an artifact.
Here is what the study is. Three US cohorts — the Health Professionals Follow-up Study, the Nurses' Health Study, and Nurses' Health Study II — pooled into 147,374 people, followed up to 30 years, 35,798 deaths. Serious scale. Over 98% cause-of-death ascertainment. This is not junk. (Evidence level: large prospective observational cohort, self-reported exposure.)
And the headline finding is real as far as it goes: 90–119 minutes of weekly strength training tracked with a 13% lower all-cause mortality risk, a 19% lower cardiovascular death risk, and a 27% lower neurological death risk. Nothing extra above 120 minutes.
But now open the joint-association table — the part almost nobody quoted — and the story inverts.
Strength training alone, 1–119 minutes a week, bought a 7–11% lower risk of death. That's the whole standalone effect. Real, modest, unremarkable.
Aerobic exercise alone, any amount above 7.5 MET-hours a week, was reportedly associated with a 26–43% lower risk of death.
And the combinations dwarf both. Thirty to 44 MET-hours of aerobic activity paired with 60–119 minutes of strength: 45% lower death risk. Push aerobic volume to 45+ MET-hours and the risk drops 53% to 58% — regardless of how much strength training people did.
Read that last clause again. At high aerobic volumes, the strength dose stopped mattering. The curve flattened. You could not titrate your way to more benefit by adding minutes under the bar.
That is not the signal of a magic weekly number. That is the signal of resistance work as a multiplier on aerobic activity — a modifier layered on top of the dominant lever, not a dose to be dialed in on its own.
Which is why "hit 90–120 minutes" is the wrong takeaway and "add strength to your cardio" is the right one. The data supports the second sentence far more cleanly than the first.
The sweet spot problem gets worse once you notice it isn't even one spot.
For cancer death, the benefit peaked at 1–29 minutes a week — a 21% lower risk — and 30–59 minutes gave 18%. That's near the bottom of the dose range.
For cardiovascular and neurological death, the benefit needed 90–119 minutes to show up.
So the "sweet spot" for cancer sits at the opposite end of the dial from the sweet spot for heart and brain. There is no single number. There is a different number for every outcome, and picking 90–120 means quietly privileging some causes of death over others. (Evidence level: same cohort, cause-specific subgroups, some with small case counts the authors flagged as fragile — colorectal, bladder, breast cancer especially.)
The fracturing continues once you leave mortality and ask what you actually want from exercise.
If the goal is mental health, frequency appears to beat intensity. A Norwegian prospective study of 10,460 students found that how often people exercised was the most consistent predictor of who avoided clinical depression and anxiety a year later — more consistent than how hard or how long. (Evidence level: prospective observational, self-reported, self-administered diagnostic instrument.)
If the goal is the metabolic signal, intensity appears to drive it. Rockefeller researchers reportedly compared all-out sprint intervals against 90 minutes of continuous moderate cycling and found the sprints triggered changes in roughly a quarter of detectable plasma proteins — and of 33 proteins linked to lower risk of metabolic disorders, obesity, and type 2 diabetes, all but one moved after sprints while only three moved after the moderate session. (Evidence level: small mechanistic RCT-style comparison in young healthy men, plus UK Biobank association mapping — biomarkers, not outcomes.)
Frequency for mood. Intensity for metabolism. Duration for cardiovascular mortality. Low duration for cancer. The "sweet spot" isn't a coordinate; it's a Rorschach test that returns whatever outcome you were already looking for.
Now the caveat that this entire beat keeps circling, the one that should be printed at the top of every write-up and instead lands in paragraph fourteen of the limitations.
In the BJSM cohort, the people who did more strength training "were generally younger, weighed less, followed healthier lifestyles, and engaged in more aerobic activity." The heavy lifters started out fitter, leaner, and younger.
So when the dose-response curve rises with strength minutes, we genuinely do not know whether we're watching what lifting does or watching who lifts. The curve may be measuring a personality — the kind of 50-something who gets to the gym twice a week — as much as a physiology.
This is the load-bearing weakness of the whole literature. It runs through the racket-sport longevity numbers too — the ones where tennis players reportedly "lived nearly a decade longer" — and even the study's authors there warned the associations "can't prove direct causation," because the people who play tennis differ from the people who don't in a hundred unmeasured ways.
Self-report compounds it. Every exposure here — minutes, MET-hours, frequency — came from questionnaires filled out every two to four years. The paper couldn't distinguish a grueling session from a gentle one, or a 20-minute workout from a 90-minute one. It inferred "90–120 minutes" from weekly totals it admits it couldn't decompose.
Which brings the standing open question into sharper focus than last week. No one has run a randomized controlled trial testing a specific weekly strength dose against mortality. Until someone does, the number floating through the aggregators is a description of a self-selected group, not an instruction.
Here is what survives all of it, stated at the evidence level it earns.
Combining aerobic and resistance training is associated with the lowest mortality in the largest and longest dataset we have — 45% to 58% lower, versus 7–11% for strength alone. That joint pattern is robust, and it directly supports guidelines urging both modalities.
What does not survive is the precision. The "13% at 90–120 minutes" line is a headline artifact — a real association dressed up as a training prescription, in a study that never measured training duration.
So the honest advice fits in six words. Don't titrate the strength dose. Add it.
Do your cardio — that's where the largest, most consistent signal lives. Then put strength on top of it, twice a week, in whatever form you'll actually keep doing.
The exact number of minutes is the part the data doesn't know. Everyone quoting it does.