HIIT edges out other exercise for blood sugar in type 2 diabetes

Exercise is a cornerstone of managing type 2 diabetes, but the practical question people actually face is which kind: steady cardio, weights, a combination, hard intervals, or simply being told to move more. Individual trials rarely compare all of these head to head, so the relative value of each has stayed murky. A network meta-analysis can rank multiple options against each other by pooling every available comparison. Knowing whether the differences between modalities are large or trivial shapes how much anyone should agonise over the choice.

The researchers searched seven databases from inception to May 2024 for randomised clinical trials of at least 12 weeks' duration in people with type 2 diabetes. Eligible trials tested aerobic training, resistance training, combined training, high-intensity interval training or physical activity advice, with HbA1c, the standard marker of long-term blood sugar control, as the outcome. In total, 158 studies covering 17,059 participants were included and analysed in a network meta-analysis, which allows all the modalities to be compared and ranked even when no single trial tested them all.

Every intervention beat the control condition. HIIT was associated with a 0.61% reduction in HbA1c, combined training with 0.58%, aerobic training with 0.58%, resistance training with 0.40% and physical activity advice with 0.35%. In the probability rankings, HIIT came out most likely to be the best option at 82%, followed by combined training at 77% and aerobic training at 76%, with resistance training at 37% and activity advice at 29%. The spread between the best and worst options was real but modest.

This is one of the largest syntheses of exercise and glycaemic control to date, and it confirms the long-standing picture that structured exercise reliably lowers HbA1c in type 2 diabetes. Its contribution is the ranking: intervals and mixed programmes sit at the top, but nothing failed. Notably, the authors highlight that even simple, unsupervised activity advice produced measurable benefit, which matters for people without access to supervised programmes. Open questions include how adherence, intensity and individual preference should weigh against the modest differences between modalities.

If you have type 2 diabetes or are working to avoid it, this is strong evidence that moving in almost any structured way improves long-term blood sugar, and that the differences between exercise styles are smaller than the difference between exercising and not. High-intensity intervals ranked best, so they are a reasonable thing to know about if time is short. Equally, the finding that plain activity advice helped is a reminder that the entry bar is low. The choice you can sustain is the one the data quietly favours.