The Resistance Training Dose: 90 to 119 Minutes
Ninety to one hundred and nineteen minutes. Not "two days a week" or "regular strength training"—90 to 119 minutes.
Ninety to one hundred and nineteen minutes. Not "two days a week" or "regular strength training"—90 to 119 minutes. After tracking 147,374 adults for three decades, researchers have finally attached a number to the resistance training prescription for mortality reduction, and the precision changes everything. Physicians counseling patients on cardiovascular risk can cite 150 minutes of weekly cardio with the confidence of decades of dose-response data. Until now, strength training guidance offered only "muscle-strengthening activities on 2 or more days per week"—no duration, no cumulative dose, no clear target for the intervention increasingly recognized as essential to healthy aging.
Last week we established exercise as first-line medicine, with evidence placing physical activity alongside pharmaceutical interventions for chronic disease prevention. This week, the evidence specifies the resistance training dose that delivers maximal longevity benefit—and reveals that precision comes with unexpected complexity.
The dose-response curve extracted from these three decades of prospective cohort data shows 90 to 119 minutes of weekly resistance training associated with a hazard ratio of 0.87 for all-cause mortality—a 13% reduction compared to no resistance training, after adjusting for aerobic activity and confounding factors including age, smoking, diet quality, alcohol intake, and family history. This isn't a linear relationship. Benefits begin at low doses, climb steadily through the first two hours weekly, then plateau completely above 120 minutes. No additional mortality benefit accrues at 130 minutes, 150 minutes, or beyond. The curve flattens.
For cardiovascular mortality, the same 90-119 minute range produces a hazard ratio of 0.81—a 19% reduction. For neurological disease mortality, including dementia, the protection strengthens further: HR 0.73, a 27% reduction. These are population-level effect sizes achieved through approximately 13 to 17 minutes daily of an activity requiring no prescription, no titration.
But cancer mortality tells a different story. The dose-response relationship inverts. Reduced cancer death risk appears only at the lowest resistance training doses: 1 to 29 minutes weekly shows a 21% reduction, 30 to 59 minutes shows HR 0.88 (12% reduction). Above 60 minutes weekly, the cancer mortality benefit disappears—not because higher doses cause harm, but because they deliver no additional protective effect beyond what minimal exposure provides.
This creates an immediate clinical translation problem. If a patient asks "how much strength training should I do to reduce my disease risk?" the honest answer now depends on which disease. Thirty minutes weekly may optimize cancer mortality reduction. Ninety to 119 minutes maximizes cardiovascular and neurological protection. The single "sweet spot" promised in headlines fractures into outcome-specific optima.
The study's design provides unusual longitudinal precision. Rather than measuring resistance training at a single baseline, researchers captured self-reported weekly duration every two years across up to 30 years of follow-up, allowing analysis of long-term training patterns rather than isolated snapshots. The cohorts—Health Professionals Follow-up Study, Nurses' Health Study, and Nurses' Health Study II—skew heavily female (80% women), predominantly white, and toward healthcare professionals with above-average health literacy. Whether these dose-response curves translate precisely to other populations remains unknown, though the biological mechanisms linking muscle strengthening to mortality reduction show no obvious demographic boundaries.
During the monitoring period, 35,798 participants died, providing statistical power to detect cause-specific mortality associations. Those performing 90-119 minutes weekly of resistance training tended to be younger at baseline, weigh less, maintain healthier overall lifestyle patterns, and engage in more aerobic activity than those doing no strength training—confounders the researchers adjusted for statistically, though residual confounding in observational data can never be eliminated entirely.
The plateau above 120 minutes raises immediate questions for the substantial population already training beyond that threshold. Does the curve genuinely flatten, or does measurement error in self-reported data obscure continued benefits at higher doses? The study cannot distinguish. More fundamentally, the researchers captured total weekly minutes but no information on session structure, intensity, load progression, or exercise selection. "Resistance training" encompassed everything from bodyweight squats to heavy barbell work, from single-set circuits to periodized strength programs. A 90-minute weekly total might mean three 30-minute sessions, two 45-minute sessions, or one 90-minute session—the study design cannot detect whether distribution matters.
Intensity data is completely absent. Ninety minutes of resistance training at 40% of one-repetition maximum produces radically different physiological adaptations than 90 minutes at 80% of one-repetition maximum, yet both register identically in this dataset. The mortality benefit may depend as much or more on intensity than duration, but the evidence provides no way to know. Current resistance training guidelines recommend working each major muscle group to near-fatigue at least twice weekly, but this cohort data cannot verify whether participants met that standard or simply accumulated minutes at minimal intensity.
The study explicitly excluded calisthenics and Pilates from its resistance training definition, focusing only on exercises using weights or body weight exercises like push-ups, squats, and lunges. This narrows the actionable population—individuals performing substantial bodyweight training or Pilates-based resistance work remain unstudied, their dose-response curves unknown.
Self-report accuracy presents the study's most fundamental limitation. Participants estimated their weekly resistance training duration from memory every two years. Overestimation is well-documented in exercise self-report data, particularly as intensity increases—people remember working harder and longer than objective measures verify. If systematic overestimation occurred, the true optimal dose might sit considerably below 90 minutes. Conversely, if participants underreported or if the questionnaire captured only formal "gym" resistance training while missing incidental strength activity, the optimal dose might exceed 119 minutes. The dose precision the study appears to provide rests entirely on the accuracy of participants' memory and interpretation of "resistance training."
Joint analyses with aerobic activity reveal the lowest mortality risk among participants combining high aerobic exercise (30 to 45 MET-hours weekly) with 60 to 119 minutes of resistance training: HR 0.55, a 45% reduction in all-cause mortality. For participants achieving very high aerobic activity (45+ MET-hours weekly), adding resistance training produced no additional mortality benefit—cardiovascular fitness alone reportedly drove risk to floor levels regardless of strength training volume.
This finding complicates the clinical prescription. For sedentary patients, resistance training delivers clear mortality benefit. For moderately active patients meeting minimum aerobic guidelines, adding 90-119 minutes of resistance training amplifies protection substantially. But for highly active patients already performing extensive aerobic exercise, the mortality benefit appears to plateau regardless of resistance training volume, even as resistance training provides other benefits—bone density, functional capacity, glucose metabolism—not captured in this mortality-focused analysis.
The cause-specific variations suggest different biological pathways mediating mortality reduction. The strong neurological disease benefit aligns with emerging evidence that resistance training preserves cognitive function and may delay dementia onset, though the researchers note appropriate caution: neurodegenerative diseases begin years before diagnosis, early symptoms may reduce exercise capacity, and neurological causes of death can be difficult to classify accurately. The cardiovascular benefit likely operates through improved glucose metabolism, favorable lipid profiles, and reduced systemic inflammation—adaptations that accumulate with consistent training over years. The cancer mortality pattern remains unexplained. Why would minimal resistance training reduce cancer death risk while higher doses provide no additional benefit?
The study offers no mechanism, leaving the finding as an association requiring replication and investigation.
Translating population-level dose-response curves into individual clinical recommendations requires acknowledging what this evidence cannot tell us. A patient with family history of cardiovascular disease might prioritize the 90-119 minute range targeting the 19% CVD mortality reduction. A patient with cancer history might question whether exceeding 60 minutes weekly offers any mortality benefit, even as their oncologist recommends resistance training for treatment-related muscle loss. A patient already performing three hours weekly of strength training confronts a plateau suggesting their additional effort, at least for mortality outcomes, may be yielding no marginal return.
The precision gained—90 to 119 minutes—transforms vague frequency guidelines into quantified weekly targets. But precision applied to observational data cannot resolve intensity, cannot specify exercise selection, cannot account for individual variation in response, and cannot prove causation. The number is real, derived from three decades tracking nearly 150,000 people. What it means for the individual patient standing in the clinic, asking how much strength training they personally should perform, remains a question requiring clinical judgment layered over population evidence.
Current guidelines recommending muscle-strengthening activities at least twice weekly now have dose support: 90 to 119 minutes weekly, distributed across those sessions, associates with maximal mortality reduction in this cohort. The plateau above 120 minutes suggests diminishing returns, at least for longevity. But the cancer mortality curve peaking at 30-59 minutes, the observational design's inability to prove causation, and the complete absence of intensity data mean this first population-level resistance training prescription arrives with precision on duration and uncertainty on nearly everything else. We have a number. Now we need to understand what it actually measures.