Weightlifting beats running for blood sugar control, researchers find (2025)

Olympic weightlifting beats running blood sugar control measures in a rigorous head-to-head comparison — a finding that could reshape how clinicians and fitness professionals approach exercise prescriptions for the nearly 40 million Americans living with or at risk of Type 2 diabetes. Published on October 30, 2025 in the Journal of Sport and Health Science, the Virginia Tech study is the first controlled preclinical trial to pit endurance exercise and resistance exercise directly against each other in a diet-induced obesity and hyperglycemia model, and resistance training came out decisively ahead on every primary metabolic measure.

The Study That Changed the Cardio-First Conversation

For decades, aerobic exercise — running, cycling, swimming — dominated medical guidelines for metabolic disease management. The intuitive logic was hard to argue with: steady-state cardio burns calories, elevates heart rate, and commands a large evidence base confirming reductions in HbA1c and fasting glucose. Current American Diabetes Association (ADA) guidelines acknowledge resistance training as a useful complement to aerobic work, but aerobic exercise has historically occupied the top tier of most clinical exercise prescriptions for Type 2 diabetes prevention and management. Resistance training was rarely elevated to equal status, let alone superior status, in blood sugar discussions.

That framing is now being challenged. A team of seven researchers led by Zhen Yan, Professor and Director of the Center for Exercise Medicine Research at the Fralin Biomedical Research Institute at Virginia Tech, designed an eight-week experiment to do what no prior preclinical study had systematically done: put endurance exercise and resistance exercise in the same controlled environment, under identical dietary and metabolic stress conditions, and measure every meaningful outcome side by side.

"There is plenty of evidence in humans that both endurance exercise, such as running, and resistance exercise, such as weightlifting, are effective in promoting insulin sensitivity. Our data showed that both running and weightlifting reduce fat in the abdomen and under the skin and improve blood glucose maintenance with better insulin signaling in skeletal muscle. Importantly, weightlifting outperforms running in these health benefits." — Zhen Yan, Virginia Tech

The paper was co-authored by Robert J. Shute, Ryan N. Montalvo, Wenqing Shen, Yuntian Guan, Qing Yu, Mei Zhang, and Yan, and was funded by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (a division of the NIH) and the Red Gates Foundation — institutional backing that lends the work substantial credibility and suggests a clear pipeline toward human translation.

Engineering a Mouse Gym: How Researchers Built the Resistance Exercise Model

One of the most technically remarkable aspects of this research is the apparatus itself. A standard mouse running wheel — the endurance exercise (EEX) model — is a well-established and validated tool in metabolic research. A mouse resistance exercise model that reproduces the progressive overload principle of human weightlifting, however, did not previously exist in any standardized, reproducible form. The Virginia Tech team built one from scratch.

Mice in the resistance exercise (REX) group lived in specially designed cages where food could only be accessed through a hinged, weighted lid. To eat, each mouse had to lift the lid while wearing a small shoulder collar, producing a biomechanical movement analogous to a squat or a clean-and-press. Crucially, the load placed on the lid was gradually increased across the eight weeks — a direct replication of the progressive overload principle that underlies all effective human strength training programs, from beginner dumbbell weightlifting routines to advanced Olympic weightlifting protocols. This is not incidental: progressive overload, not static load, is what drives the neuromuscular and metabolic adaptations that make resistance training effective. The model's novelty is that it reproduces this mechanism in a controlled, measurable way for the first time in preclinical obesity research.

The four experimental groups were:

• NC-SED: Normal chow, sedentary (healthy baseline control)
• HFD-SED: High-fat diet, sedentary (obese and hyperglycemic control)
• EEX: High-fat diet plus voluntary wheel running (endurance exercise)
• REX: High-fat diet plus progressive weighted-lid lifting (resistance exercise)

All exercising mice stayed on the same high-fat diet for the full eight weeks. That design ensured any metabolic differences between the EEX and REX groups came down to the type of exercise — not differences in caloric intake, baseline body weight, or diet composition. Body composition was assessed using EchoMRI, a non-invasive quantitative magnetic resonance technique that measures fat mass and lean mass with high precision — far more granular and reliable than simple scale weight, and capable of distinguishing visceral fat from subcutaneous fat depots.

The Results: Where Resistance Weightlifting Beats Running

Both exercise modalities delivered meaningful benefits compared to sedentary obese mice — a baseline finding that aligns with decades of human clinical data confirming that any structured exercise improves metabolic health versus no exercise at all. But the degree and profile of benefit diverged significantly between the two modalities. The full picture, drawn from the published abstract on PubMed, looks like this:

Outcome Measure	Endurance Exercise (Running)	Resistance Exercise (Weightlifting)	Advantage
Overall weight gain attenuation	Significant improvement	Significant improvement	Tie
Total fat mass reduction (EchoMRI)	Yes (fat mass reduced, lean mass preserved)	Yes (fat mass reduced, lean mass preserved)	Tie
Visceral fat suppression	Improved	Significantly greater suppression	Weightlifting
Subcutaneous fat suppression	Improved	Significantly greater suppression	Weightlifting
Brown adipose tissue (brown fat) mass	Significant increase	No significant change	Running
Aerobic exercise capacity (run-to-fatigue)	Significantly improved	No significant change	Running
Skeletal muscle weight (on HFD)	Increased relative to NC-SED baseline	No significant change vs. HFD-SED	Mixed — see note
HOMA-IR (insulin resistance index)	Improved vs. HFD-SED	Significantly greater improvement	Weightlifting
Glucose tolerance (GTT)	Improved vs. HFD-SED	Significantly greater improvement	Weightlifting
Insulin tolerance (ITT)	Improved vs. HFD-SED	Significantly greater improvement	Weightlifting
Cardiac function changes	No significant change vs. HFD-SED	No significant change vs. HFD-SED	Tie (neither worsened nor improved)
Skeletal muscle contractile properties	No significant change vs. HFD-SED	No significant change vs. HFD-SED	Tie

Note on skeletal muscle weight: EEX mice showed an increase in skeletal muscle weight compared to normal-chow sedentary (NC-SED) controls — meaning the high-fat diet plus endurance exercise combination produced a muscle-weight outcome that exceeded even the healthy baseline group, likely reflecting an adaptive response to sustained aerobic loading under caloric surplus conditions. This should not be interpreted as a straightforward muscle-building advantage for running over weightlifting; REX mice did not show a parallel significant increase versus the sedentary obese group, yet outperformed EEX mice on all three insulin-sensitivity metrics.

The three headline metabolic wins for resistance training — HOMA-IR, glucose tolerance (GTT), and insulin tolerance (ITT) — are precisely the metrics clinicians managing Type 2 diabetes risk care about most. HOMA-IR (Homeostatic Model Assessment of Insulin Resistance) is a validated formula derived from fasting glucose multiplied by fasting insulin, divided by a constant; in practice, a HOMA-IR score below 1.0 is considered optimal insulin sensitivity, while scores above 2.5–3.0 suggest clinically significant insulin resistance. A lower HOMA-IR means the body is managing blood sugar efficiently without requiring large insulin spikes. The fact that weightlifting beats cardio on all three of these measures simultaneously — and does so independently of muscle mass gains, as discussed below — is the core finding driving the paper's clinical significance.

One particularly notable nuance: running produced one exclusive benefit that weightlifting did not — a significant increase in brown adipose tissue (brown fat) mass. Unlike white fat, which stores energy passively, brown fat is metabolically active: it generates heat through a process called non-shivering thermogenesis, burning stored lipids and glucose in the process. Higher brown fat volume is independently associated with improved insulin sensitivity and reduced obesity risk in humans. This divergence strongly suggests that the two exercise types activate distinct but complementary physiological pathways, and that combining them may produce synergistic metabolic benefits greater than either modality alone.

Why Weight Lifting Beats Cardio for Blood Sugar: The Molecular Story

The most intellectually provocative finding in the study was this: the superior blood sugar benefits of resistance exercise were not explained by changes in muscle mass or overall exercise performance. The REX mice did not develop significantly heavier skeletal muscles, nor did they perform better on treadmill run-to-fatigue tests. Yet their insulin signaling markers were superior across every tested measure. This decoupling of metabolic benefit from muscle hypertrophy suggests a fundamentally different — and still partially uncharted — mechanism is at work.

Exercise physiology research offers several well-established candidate mechanisms for why resistance training may produce superior insulin sensitization independent of muscle mass gains. Resistance training — particularly when performed progressively and at sufficient intensity — selectively recruits large, fast-twitch (Type II) muscle fibers that sustained endurance activities like running and cycling do not heavily engage. These fibers, when subjected to progressive mechanical loading, are known to upregulate glucose uptake capacity — effectively expanding the body's metabolic capacity for blood sugar clearance beyond what aerobic training alone achieves. Two key molecular cascades are central to this process: the mTOR pathway (mechanistic target of rapamycin — a master regulator of cell growth, protein synthesis, and metabolic homeostasis) and calcium signaling pathways within muscle cells, both of which are activated by resistance-type mechanical loading and independently enhance insulin sensitivity according to a well-established body of exercise physiology literature.

A third mechanism central to blood glucose regulation is the GLUT4 glucose transporter. GLUT4 is the primary protein responsible for moving glucose from the bloodstream into muscle cells in response to insulin. A well-established body of exercise physiology research demonstrates that resistance training — particularly when it engages fast-twitch fibers under progressive load — increases GLUT4 expression and improves its translocation to the cell membrane. More GLUT4 at the cell surface means more glucose can be cleared from the blood per unit of insulin secreted, which is precisely what a lower HOMA-IR score reflects. The study authors note improved insulin signaling in skeletal muscle in both the abstract and the accompanying Virginia Tech press release; the GLUT4-mediated glucose uptake mechanism described here represents the established exercise physiology framework that most parsimoniously explains those observed signaling improvements, though the precise molecular mapping in this model's specific skeletal muscle tissue remains a subject for follow-up mechanistic work.

Yan's group believes these molecular-level findings could inform entirely new pharmacological targets for Type 2 diabetes — compounds designed to activate the mTOR and calcium signaling cascades specifically triggered by progressive resistance loading, independent of the exercise itself. In an era when GLP-1 receptor agonists like semaglutide (Ozempic) are dominating the metabolic disease conversation, Yan was careful to emphasize that no existing drug replicates the comprehensive, multi-system, multi-pathway benefits of a balanced exercise program — and that exercise remains the most cost-effective intervention available.

Does Lifting Weights Get Your Blood Flowing? The Circulatory Connection

A persistent question surrounding resistance training is whether it meaningfully improves circulation — the kind of cardiovascular benefit traditionally associated with aerobic work. The short answer: yes, lifting weights does get your blood flowing, and the mechanism matters for glucose control. Established exercise physiology research demonstrates that resistance exercise increases limb blood flow during and after sessions, in part through nitric oxide-mediated vasodilation — a well-documented acute response to mechanical and metabolic stress in working muscle. With consistent training over weeks, resistance exercise is associated with capillary expansion within skeletal muscle tissue, a structural adaptation that improves the delivery of oxygen, insulin, and glucose directly to muscle cells. These findings are well supported in the exercise physiology literature and are consistent with the structural adaptations the Virginia Tech model was designed to induce.

This vascular dimension of resistance training is directly relevant to blood sugar control. Insulin and glucose must be delivered to muscle tissue via the bloodstream before GLUT4 transporters can take them up. Improved vascular conductance — more capillaries, greater blood flow per unit of muscle — reduces the diffusion distance between capillary and muscle fiber. The result is faster, more complete glucose clearance after meals. So when people ask whether lifting weights gets your blood flowing, the question is really about more than cardiovascular effort during a session — it is asking whether resistance training builds the vascular infrastructure needed for long-term metabolic efficiency. The evidence says it does.

How Long Does Blood Sugar Stay Elevated After Exercise?

For patients and clinicians managing blood glucose day-to-day, the temporal dynamics of exercise's effects matter enormously. The acute impact of a single bout of aerobic exercise typically lowers blood sugar during and immediately after the session — but the downstream improvements in insulin sensitivity can last 24 to 48 hours post-workout. This is precisely why the ADA recommends avoiding more than two consecutive days without exercise: the insulin-sensitizing window closes, and metabolic gains begin to erode.

Resistance training presents a more complex short-term picture. Intense weightlifting can transiently raise blood glucose during the session itself, due to the acute release of cortisol and adrenaline, which signal the liver to release stored glucose (glycogenolysis) to fuel the working muscles. This transient elevation is normal and not clinically concerning in most people. The Mayo Clinic notes that vigorous exercise — including resistance training — can affect blood sugar levels for up to 24 hours after the workout as muscles replenish glycogen stores and insulin sensitivity remains elevated. The downstream improvements in insulin sensitivity are what the Virginia Tech study captures and quantifies across eight weeks, and it is that cumulative, chronic adaptation — not a single session's acute response — where resistance training appears to most decisively outperform aerobic exercise. People managing blood glucose should monitor levels before, during, and after beginning any new resistance training regimen, as adjustments to medication or carbohydrate intake may be warranted during the adaptation period.

Caveats and the Limits of a Mouse Model

Any intellectually honest reading of this study must grapple carefully with what a mouse model can and cannot tell us. The findings are promising, and several features of the study design support meaningful translation to humans. The key molecular pathways identified — mTOR signaling, GLUT4 transporter regulation, calcium-mediated insulin sensitization — are highly conserved between rodents and humans, lending mechanistic credibility to the results. Mouse studies have successfully predicted human responses to exercise interventions in prior metabolic research. However, several important limitations deserve direct acknowledgment.

The psychosocial, behavioral, and environmental factors that drive obesity in real human populations — ultra-processed food environments, chronic psychological stress, sleep disruption, sedentary occupations, socioeconomic constraints on healthy eating and exercise access — cannot be replicated in a laboratory cage. The high-fat diet model captures one dimension of metabolic disease (dietary lipid overload) without reproducing the full complexity of human Type 2 diabetes pathogenesis, which also involves genetic predisposition, gut microbiome variation, and decades of cumulative lifestyle exposure.

A population-level consideration that deserves serious attention is aging. As humans age, they progressively lose skeletal muscle mass through sarcopenia — a process that directly worsens glucose regulation, reduces resting metabolic rate, and increases metabolic disease risk. Resistance training's capacity to build and preserve lean mass becomes increasingly valuable with each decade of life. The Virginia Tech study's finding that weightlifting's blood sugar benefits occurred independently of significant muscle mass gains raises a compelling follow-up question: would those benefits be even larger in an older model that did produce meaningful hypertrophy? Human trials in aging populations are a critical next step.

The study used exclusively male mice throughout — a limitation the authors explicitly acknowledge. Well-documented sex differences exist in metabolic disease progression, fat distribution patterns, and hormonal responses to both aerobic and resistance exercise (particularly the roles of estrogen and testosterone in insulin signaling). Whether female subjects — rodent or human — would show the same resistance-exercise advantage requires separate investigation. Human trials replicating this design across both sexes, a range of ages, and varying baseline fitness levels will be necessary before these findings can be translated into revised clinical guidelines.

What This Means for Exercise Prescriptions and Diabetes Prevention

Yan's translational message is deliberately inclusive rather than prescriptive. Rather than declaring resistance training the outright winner and telling patients to abandon the treadmill, he frames the finding as an expansion of legitimate, evidence-backed options — particularly for the large population segment that cannot safely engage in sustained endurance activity due to joint conditions (osteoarthritis, prior injury), cardiovascular contraindications, mobility limitations, or simple preference. For this group, the study provides strong preclinical evidence that progressive resistance training — from structured dumbbell weightlifting programs to machine-based gym routines — delivers blood sugar control benefits that are at least equal to, and likely superior to, running.

"The findings also bring good news for people who, for any number of reasons, cannot engage in endurance-type exercise. Weight training has equal, if not better, anti-diabetes benefits." — Zhen Yan, Virginia Tech

For those who can perform both modalities, the data argue clearly for a combined training approach. Running's unique capacity to expand brown fat mass and improve aerobic exercise capacity complements weightlifting's superior performance on HOMA-IR, glucose tolerance, and insulin tolerance. The two modalities appear to activate distinct biological pathways that do not substantially overlap — which means their benefits are likely additive rather than redundant. Yan's own synthesis recommendation, shared via the Virginia Tech press release, is unambiguous: "The take-home message is that you should do both endurance and resistance exercise, if possible, to get the most health benefit."

From a public health standpoint, the implications extend well beyond individual exercise prescriptions. Current ADA exercise guidelines recommend a minimum of 150 minutes per week of moderate-intensity aerobic activity, plus resistance training on two or more days per week. This study's findings suggest that the resistance training component may be doing more metabolic work than the guidelines' secondary billing implies — and that for patients unable to meet the aerobic minimum, resistance training alone may be a more effective primary intervention than previously credited. Revised guidelines based on human trial replication of these findings could meaningfully shift clinical practice.

If the molecular pathways identified — particularly the mTOR and calcium signaling cascades activated by progressive mechanical loading — can be mapped in granular detail through follow-up mechanistic work, they may yield pharmacological targets that are additive to existing diabetes drugs. At a moment when GLP-1 agonists are already restructuring the obesity drug market, exercise-mimicking compounds that selectively activate the specific insulin-sensitizing pathways triggered by resistance training could represent a meaningful next frontier in metabolic medicine — one that could benefit patients who cannot exercise due to disability or advanced disease.

Key Takeaways

• Resistance training outperformed endurance exercise on the three most critical metabolic metrics — HOMA-IR, glucose tolerance (GTT), and insulin tolerance (ITT) — in the first controlled, direct preclinical comparison of its kind, published October 30, 2025 in the Journal of Sport and Health Science.
• Both exercise types significantly reduced fat mass and improved blood glucose maintenance versus sedentary obese controls, confirming that any structured exercise is better than none for people at risk of Type 2 diabetes.
• The weightlifting advantage was independent of muscle mass or aerobic performance gains, pointing to distinct molecular mechanisms — particularly mTOR signaling, GLUT4 upregulation, and calcium signaling pathways in fast-twitch muscle fibers — as the likely primary drivers, consistent with established resistance-training exercise physiology.
• Running retained two exclusive benefits: increased brown adipose tissue mass and improved aerobic exercise capacity — differences that strengthen the case for combining both modalities rather than choosing one.
• Lifting weights does get your blood flowing: resistance training promotes capillary expansion in skeletal muscle and improves vascular conductance, enhancing delivery of insulin and glucose to muscle cells and supporting long-term metabolic efficiency.
• Blood sugar improvements from exercise persist for 24–48 hours post-workout; consistent, progressive training — not isolated sessions — is what produces the durable metabolic adaptations measured in this eight-week protocol.
• The findings are especially significant for people who cannot perform sustained cardio — due to joint issues, age-related mobility limits, or other contraindications — confirming that progressive weightlifting delivers "equal, if not better, anti-diabetes benefits."
• Molecular insights from this work may inform new drug targets designed to pharmacologically replicate or amplify the specific insulin-sensitizing pathways activated by resistance exercise.
• Key limitation: the study used exclusively male mice; human trials across sexes, ages, and fitness levels are required before findings can update clinical guidelines.

What Comes Next

The Virginia Tech team's most immediate next steps involve human validation at scale. The logical trial design would be a randomized controlled study comparing matched volumes of resistance and endurance training in people with prediabetes or early-stage Type 2 diabetes, using HOMA-IR and continuous glucose monitoring (CGM) as co-primary endpoints, with secondary endpoints including visceral fat volume (measured by MRI or DEXA), brown fat activity (measured by PET-CT), and inflammatory biomarkers. Stratifying participants by age, sex, and baseline fitness level would directly address the key limitations of the preclinical model.

Parallel mechanistic work will almost certainly focus on the specific insulin-signaling cascade changes — particularly GLUT4 expression, mTOR activation, and calcium-mediated signaling — observed in skeletal muscle tissue from the REX group. Mapping these pathways in human muscle biopsies taken before and after a structured resistance training intervention would provide the molecular bridge between mouse and human needed to justify pharmaceutical investment in exercise-mimicking compounds.

For now, the message to clinicians, patients, and the fitness industry is both evidence-backed and immediately actionable: the weight room belongs alongside — and for many patients, ahead of — the running track in discussions of metabolic disease management. The preclinical evidence is compelling, the mechanistic rationale is sound, and the safety profile of supervised resistance exercise is well established. The research now points clearly in one direction: pick up the weight.