Adults who lift weights for roughly 90 to 120 minutes per week face the sharpest reduction in all-cause and cause-specific mortality, according to a study published in the British Journal of Sports Medicine that tracked participants across three large U.S. cohorts for up to 30 years. The finding lands at a time when social-media fitness culture routinely promotes high-volume training programs whose long-term safety has never been tested at population scale, raising a direct question: how much strength training is too much?
Why the 90-to-120-minute resistance training window matters right now
The new BJSM study drew on repeated biennial assessments from the Health Professionals Follow-up Study (HPFS, 1992 to 2022), the Nurses’ Health Study (NHS, 2002 to 2021), and NHSII (2003 to 2021), giving researchers decades of follow-up data on exercise habits and mortality outcomes. That time span is significant because most prior resistance-training research relied on single-point surveys or shorter observation periods, making it difficult to separate a temporary fitness habit from a sustained practice. By capturing self-reported exercise every two years across three prospective cohorts, the authors could model how long-term lifting patterns relate to death from cardiovascular disease, cancer, and other causes after adjusting for aerobic activity.
The dose-response curve that emerged was not a straight line. Mortality risk dropped as weekly resistance training increased toward the 90-to-120-minute range, then flattened or reversed at higher volumes. That nonlinear pattern has practical consequences for anyone designing a weekly workout schedule: adding a fifth or sixth hour of lifting may not deliver additional longevity benefits and could, based on the shape of the curve, erode them.
One reasonable hypothesis is that this mortality benefit will narrow among adults over 65 once researchers add age-stratified hormone and recovery markers to the existing models. Older adults experience slower muscle repair, shifting testosterone-to-cortisol ratios, and higher injury rates, all of which could compress the effective dose window. The current cohort data do not include those biomarkers, so the 90-to-120-minute sweet spot may look different when biology-specific controls enter the equation.
Converging cohort data on resistance training and mortality
The BJSM findings do not stand alone. A separate national cohort study of 416,420 U.S. adults modeled the dose-response relationship between both aerobic and muscle-strengthening activity and all-cause mortality, confirming that the association is nonlinear even after controlling for cardio exercise. That analysis used nationally representative survey data rather than health-professional cohorts, broadening the demographic base and reinforcing the idea that moderate resistance-training volumes carry the strongest signal.
A 2022 systematic review and meta-analysis published in the American Journal of Preventive Medicine synthesized multiple cohort studies on resistance training and mortality from cardiovascular disease and cancer. Its dose-response component identified the same nonlinear relationship, with benefits peaking at moderate volumes before tapering. Taken together, these lines of evidence, spanning different populations, time frames, and analytical methods, point in the same direction: more is not always better when it comes to lifting for longevity.
A CDC-published analysis using the Cancer Prevention Study-II Nutrition Cohort added further texture. That study reported a quadratic trend between muscle-strengthening activity and mortality, meaning the relationship curves rather than climbs steadily. Hazard ratios were adjusted for confounders and aerobic activity, isolating the resistance-training signal from the well-documented benefits of cardio. The emerging consensus is that strength work complements, rather than replaces, moderate-to-vigorous aerobic exercise in a long-term health strategy.
Gaps the current research cannot close
Several blind spots limit how far these results can be applied. All three cohorts in the BJSM analysis relied on self-reported exercise minutes. Participants recalled their weekly training volume on questionnaires every two years, a method known to introduce recall bias. No individual-level adherence logs, gym check-in records, or wearable-device data were used to verify those estimates. The actual minutes spent under load could differ meaningfully from what participants reported, especially among people whose routines fluctuate with work or family demands.
The studies also lack granularity on exercise type. Free-weight barbell training, machine-based circuits, and bodyweight routines all count as “resistance training” in these datasets, yet they impose different mechanical stresses and recovery demands. Without breakdowns by modality, readers cannot determine whether the 90-to-120-minute sweet spot applies equally to heavy squats and deadlifts, light resistance-band work, or mixed gym classes that blend strength and cardio intervals.
Intensity data present a similar gap. None of the cohorts captured load, repetitions, or perceived exertion in a way that allows researchers to distinguish high-intensity strength sessions from moderate ones. Two people logging 100 minutes per week could be doing vastly different things: one might perform low-rep, near-maximal lifts, while another cycles through lighter, higher-rep sets. The mortality curves cannot separate those experiences, leaving open questions about whether the observed plateau at higher volumes reflects overtraining stress, injury risk, or simply the kinds of people who choose to train that much.
Demographic limits also matter. Health professionals and large national survey respondents are not perfect stand-ins for the entire population. People in these cohorts may have higher baseline health literacy, more stable access to medical care, and different occupational activity levels than workers in physically demanding jobs. That context could influence how much additional benefit they derive from formal resistance training and how safely they can sustain it over decades.
What this means for everyday training plans
For adults trying to act on this research today, the clearest takeaway is volume, not variety. Aiming for roughly 90 to 120 minutes of weekly resistance training, spread across two or more sessions to allow recovery, appears to align with the lowest observed mortality risk in current cohort data. For many people, that translates to two or three 30- to 45-minute strength workouts per week, layered on top of regular walking, cycling, or other aerobic activity.
Within that window, consistent practice likely matters more than fine-tuning the exact number of sets or choosing between dumbbells and machines. The cohorts suggest that being reliably in the moderate range over many years is associated with better outcomes than oscillating between inactivity and extreme programs. For those already lifting far beyond two hours a week, the evidence offers a prompt to reassess whether additional volume is serving long-term health goals or mainly chasing short-term performance or aesthetic milestones.
None of this means that lifting more than 120 minutes weekly is inherently dangerous, or that 100 minutes guarantees protection. The findings are observational, not prescriptive. They map patterns in large groups rather than dictating what any one person must do. Individual factors – injury history, joint health, training age, and personal preferences – still shape the safest and most sustainable routine.
What the data do provide is a guardrail in a culture that often celebrates maximal effort. For people new to strength training, they offer reassurance that modest, well-distributed sessions can deliver meaningful longevity benefits without living in the gym. For experienced lifters, they raise a nuanced question: if the goal is to live longer and stay functional, not just to lift heavier, where is the point of enough?
Future research using device-based activity tracking, more detailed logging of exercise type and intensity, and broader demographic samples will be needed to refine that answer. For now, the convergence of multiple cohorts points to a simple, if countercultural, message: in strength training as in many areas of health, a measured dose may be more powerful than an extreme one.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
