Yoga and meditation outdid conventional exercise on diabetes markers, on shaky evidence

Type 2 diabetes rarely travels alone; it comes bundled with cardiometabolic risks spanning glucose control, body composition, blood pressure and lipid metabolism, and exercise is a standard prescription for all of them. Mind-body exercises such as yoga, meditative movement and mindfulness-based training have been proposed as alternatives or additions to conventional exercise, but individual studies have tested many different styles against many different outcomes. This network meta-analysis set out to compare the various mind-body approaches against each other and against conventional exercise across ten cardiometabolic risk factors, to see which style does what best.

The authors searched four databases up to December 15, 2024 for randomised controlled trials of mind-body exercises in people with type 2 diabetes, covering glucose metabolism, body composition, cardiovascular physiology and lipid metabolism. Two reviewers independently extracted data and assessed risk of bias. A network meta-analysis then ranked the interventions, a method that allows comparisons between approaches that were never tested against each other directly. The certainty of the evidence behind each comparison was formally graded, and the review was registered in advance.

Different styles topped different outcomes. Meditative exercise ranked best for fasting plasma glucose. Mindfulness intervention training showed the greatest benefit for glycated haemoglobin, lowering it by 0.78 points, and for blood pressure, with reductions of 13.00 mmHg systolic and 6.00 mmHg diastolic, outperforming conventional exercise. Yoga with meditation was most effective for body mass index, cutting it by 2.90 points. The major qualification is that most comparisons were rated very low certainty, owing to bias within trials and disagreement between them, although consistency checks and publication bias tests were reassuring.

The suggestion that mind-body practice can rival or exceed conventional exercise on some cardiometabolic markers is striking, and it extends a literature that has mostly examined such practices for stress and wellbeing. But the very low certainty ratings mean the rankings are provisional, and some effect sizes are large enough to invite scepticism until stronger trials replicate them. The included trials were also relatively short and their protocols varied, leaving long-term benefits, optimal styles and the right amount of practice unresolved, as the authors themselves acknowledge.

For someone managing type 2 diabetes, this is a reason to take mind-body exercise seriously as more than relaxation: in pooled trials it moved hard numbers like glycated haemoglobin, blood pressure and body mass index. It is not a reason to swap conventional exercise for meditation on the strength of very low certainty comparisons. The fairest reading is that these practices look promising alongside standard activity, with the true size of their benefit still to be pinned down.