HIIT Beats Other Exercise Types for Insulin Resistance in Overweight Adults
A network meta-analysis of 86 RCTs finds HIIT most effective for reducing fasting insulin and HOMA-IR, with an optimal dose of 1000–1200 MET-min/week.
Summary
A comprehensive network meta-analysis pooling 86 randomized controlled trials and over 4,300 adults with overweight or obesity compared four exercise modalities — aerobic training, resistance training, combined training, and high-intensity interval training (HIIT) — on key insulin-resistance markers. All modalities significantly improved fasting insulin and HOMA-IR versus usual care, but HIIT consistently ranked highest across all three biomarkers. Dose-response analyses identified an optimal weekly exercise volume of roughly 1000–1200 MET-minutes — equivalent to about 150–200 minutes of moderate-to-vigorous activity — for maximal insulin-sensitivity gains. Adding caloric restriction to exercise provided no statistically significant additional benefit for insulin-resistance markers, though that finding rests on low-confidence indirect evidence. The overall certainty of evidence was rated low to moderate, underscoring the need for future direct-comparison trials.
Detailed Summary
Insulin resistance is a central driver of type 2 diabetes, cardiovascular disease, and accelerated biological aging — making its modulation through lifestyle intervention a core longevity target. Yet clinicians and patients have lacked clear guidance on which exercise type delivers the greatest metabolic benefit, and at what dose.
This systematic review and network meta-analysis searched four major databases through May 2026, identifying 86 randomized controlled trials enrolling 4,358 adults with overweight or obesity (68% women). Four exercise modalities were compared: aerobic training (AT), resistance training (RT), combined training (CT), and high-intensity interval training (HIIT). Primary outcomes were fasting insulin (FINS), fasting blood glucose (FBG), and the Homeostasis Model Assessment of Insulin Resistance (HOMA-IR). A frequentist network framework was used, with the node-splitting method confirming consistency and restricted cubic spline models estimating dose-response curves.
All four modalities significantly reduced FINS and HOMA-IR relative to usual care. HIIT produced the largest effect sizes (FINS: MD = −2.24 μU/mL; HOMA-IR: MD = −0.67) and ranked first across all three biomarkers by SUCRA probability (FINS 85.1%, FBG 97.8%, HOMA-IR 80.0%). CT and RT emerged as practical alternatives when HIIT is not feasible. Dose-response analyses revealed nonlinear associations, with optimal improvements in fasting insulin at approximately 1,265 MET-min/week and in HOMA-IR at approximately 1,091 MET-min/week — translating to a practical recommendation of 1,000–1,200 MET-min/week. Adding caloric restriction to exercise showed no significant additional benefit for insulin-resistance markers, though this finding is derived from low-confidence indirect comparisons.
Important caveats apply. CINeMA ratings placed overall evidence certainty at low to moderate, driven partly by performance bias from inability to blind participants and personnel, and by missing outcome data. The summary here is based on the abstract only, as the full text was not available. Future head-to-head trials comparing modalities directly, with standardized dose reporting and long-term adherence tracking, are needed to solidify these conclusions.
Key Findings
- HIIT ranked #1 for reducing fasting insulin, fasting glucose, and HOMA-IR across all 86 RCTs.
- Optimal exercise dose for insulin sensitivity is roughly 1,000–1,200 MET-min/week (~150–200 min moderate activity).
- All four exercise modalities — AT, RT, CT, and HIIT — significantly improved fasting insulin and HOMA-IR vs. usual care.
- Adding caloric restriction to exercise showed no significant additional insulin-resistance benefit in this analysis.
- Combined and resistance training are viable HIIT alternatives when interval training is not tolerated or accessible.
Methodology
Frequentist network meta-analysis of 86 RCTs (4,358 participants) identified from PubMed, Embase, Cochrane CENTRAL, and Web of Science through May 2026. Risk of bias was assessed with Cochrane RoB 2.0 and evidence certainty with CINeMA; consistency was verified via node-splitting. Dose-response relationships were modeled using restricted cubic splines with exercise volume expressed in MET-min/week.
Study Limitations
Overall evidence certainty was rated low to moderate by CINeMA, largely due to performance bias from unblinded participants and personnel, and missing outcome data. The dose-response finding for fasting blood glucose did not reach statistical significance, and the caloric-restriction comparison relied on low-confidence indirect evidence. This summary is based on the abstract only, as the full paper was not available for review.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe: