Insulin resistance markers improved most at 1,000–1,200 MET-minutes a week
In short
A network meta-analysis of 86 randomized trials involving 4,358 adults with overweight or obesity found lower fasting insulin and HOMA-IR with aerobic, resistance, combined and HIIT training. Estimated doses associated with the largest improvements were 1,265 and 1,091 MET-minutes per week, respectively. HIIT had the largest effects, but low-to-moderate certainty limits claims that it is the best choice for everyone.
Aerobic exercise, lifting, combined training and HIIT all lowered fasting insulin and HOMA-IR versus usual care in a network meta-analysis of 86 randomized trials. The 4,358 participants had overweight or obesity, and 68% were women. HOMA-IR estimates insulin resistance from fasting glucose and insulin.
Should HIIT replace lifting because it ranked first?
HIIT showed mean differences of −2.24 μU/mL for fasting insulin and −0.67 for HOMA-IR. Its SUCRA rankings were first for fasting insulin (85.1%), fasting glucose (97.8%) and HOMA-IR (80.0%). Those percentages describe rankings, not percentage improvements or an individual's chance of success. Resistance training also helped, so the ranking alone is no reason to abandon the big three.
What does 1,000–1,200 MET-minutes per week look like?
The modeled doses associated with the largest improvements were 1,265 MET-minutes per week for fasting insulin and 1,091 for HOMA-IR. The authors suggested approximately 1,000–1,200 as a practical range. Fasting glucose showed no significant dose-response relationship, and the range is not a precise prescription for every person.
These are approximate conversions, not study findings. Multiply activity intensity in METs by minutes. Activities around 3.5–6 METs require different durations: at an assumed average of 5 METs, 200–240 minutes gives 1,000–1,200 MET-minutes. Three 40-minute lifting sessions assumed to average 4 METs plus two 60-minute activities at 5 METs total 1,080. Rest periods and effort change the actual estimate.
How should this inform diet and training logs?
Adding calorie restriction did not produce a significant extra improvement in these markers, but that finding rests on low-confidence indirect evidence. It does not make diet irrelevant. Log lifting sessions in fairlift and review weekly volume; recording duration alongside them helps estimate activity across lifting and cardio. Weight multiplied by repetitions is not itself a MET-minute measure.
Evidence certainty was low to moderate. The network rankings include indirect comparisons rather than head-to-head trials of every exercise mode. This article is not medical advice for diabetes treatment and should not guide medication changes.
Frequently asked questions
Can resistance training alone improve insulin resistance markers?
In a network analysis of 86 randomized trials in adults with overweight or obesity, resistance training lowered fasting insulin and HOMA-IR versus usual care. That does not guarantee an individual's response or establish a diabetes treatment outcome.
How many minutes of exercise equal 1,000–1,200 MET-minutes?
Assuming an average intensity of 5 METs, about 200–240 minutes per week. This is an approximate illustration, not a time prescription directly tested by the review.
Does HIIT's 97.8% SUCRA mean glucose fell by that amount?
No: 97.8% is a ranking measure for fasting glucose, not its percentage reduction. HIIT's mean differences were −2.24 μU/mL for fasting insulin and −0.67 for HOMA-IR.
Does exercising remove the need for calorie restriction?
Adding calorie restriction showed no significant extra benefit for insulin resistance markers, but the comparison was indirect and had low confidence. It does not establish that dietary changes are unnecessary for weight management or other health goals.
Source: PubMed