Cardio plus weights topped the list for HbA1c

Exercise is a standard part of managing type 2 diabetes, yet the advice given to older adults rarely specifies which kind of training to do or how much of it is needed. Most trials compare a single modality against usual care rather than against each other, so the relative merits of aerobic work, weights, interval training and traditional mind-body practices have stayed unclear. The dose question is just as open: at what weekly volume do glycaemic markers actually start to shift? This review set out to rank modalities and to model dose against HbA1c, fasting blood glucose and 2-hour postprandial glucose in people aged 60 and over.

The authors searched seven databases from inception to 18 August 2025 for randomised controlled trials in adults with type 2 diabetes where the mean participant age was at least 60 years. Thirty-four RCTs involving 2461 participants met the criteria. Using a Bayesian random-effects network meta-analysis, they compared continuous aerobic exercise, combined aerobic and resistance exercise, traditional Chinese sports, resistance exercise, high-intensity interval training and control. Exercise dose was standardised as MET-min per week and entered into dose-response models. Modalities were ranked using SUCRA probabilities across HbA1c, fasting glucose and 2-hour postprandial glucose.

For HbA1c, all five active modalities beat control: combined training reduced it by 1.05%, traditional Chinese sports by 0.80%, HIIT by 0.63%, continuous aerobic exercise by 0.48% and resistance exercise by 0.44%. Combined training ranked highest with a SUCRA of 94.86%. For fasting glucose, only combined training (-1.44 mmol/L), continuous aerobic exercise (-0.98 mmol/L) and traditional Chinese sports (-0.77 mmol/L) separated from control, with combined training again ranked first at 88.03%. Evidence on 2-hour postprandial glucose was limited and only combined training showed a significant reduction (-3.15 mmol/L). Dose-response modelling was non-linear, with significant improvements beginning at about 520 MET-min per week for HbA1c and 500 MET-min per week for fasting glucose.

This reinforces the long-standing message that exercise improves glycaemic control, and adds a comparative layer: mixing aerobic and resistance work looked better than either alone, which fits the physiological logic that the two stimuli act through partly different routes. It also puts numbers on volume, suggesting a threshold below which effects are hard to detect and a curve that is not simply linear. The rankings come from a network of indirect comparisons rather than head-to-head trials, so they indicate probability rather than proof. Open questions include the optimal dose for each individual modality, how intensity interacts with volume, and why postprandial glucose evidence remains so thin.

If you are older and managing type 2 diabetes, this is a reason to think that a programme containing both cardio and resistance work has the broadest glycaemic payoff, rather than picking one and ignoring the other. The traditional Chinese sports result is a useful signal that lower-impact practices are not merely token activity. The dose findings suggest that total weekly volume matters and that very small amounts of activity may be too little to register on HbA1c. None of this is a prescription, and how any individual responds will depend on medication, diet and starting fitness.