After menopause, 3 to 4 months of exercise lowered insulin; a year did not

How well the body responds to insulin sits at the centre of metabolic health, and it is a live question for women after menopause. Exercise is routinely recommended as a lever, but recommendations are only as strong as the trial evidence beneath them, and individual trials in this group have been modest in size. This review set out to pool the randomised evidence on programmed exercise lasting at least 12 weeks and its effects on insulin sensitivity-related outcomes, including fasting insulin, HOMA-IR, C-peptide, IGF-1 and body measurements. A second question lurked underneath: does a longer programme buy a bigger metabolic return?

The authors searched PubMed, Embase, Scopus, Web of Science and the Cochrane Library from inception through May 2016, in all languages. They included randomised controlled trials of programmed exercise in postmenopausal women and split the interventions into mid-term programmes of 3 to 4 months and long-term programmes of 6 to 12 months. Risk of bias was assessed with the standard Cochrane tool, and results were combined using random-effects models. Seven trials covering 580 women made the cut.

Mid-term programmes of 3 to 4 months lowered fasting insulin by 6.50 pmol/L on average and trimmed HOMA-IR insulin resistance scores by 0.18 compared with controls. They also reduced BMI by 1.48 units, body fat by 2.99 percentage points and waist circumference by 1.87 cm. Long-term programmes of 6 to 12 months showed no significant effect on insulin or HOMA-IR, and their only clear result was a 3.74 cm reduction in waist circumference. Neither programme length shifted circulating IGF-1, glucose or triglycerides, and disagreement between the studies was moderate to low.

The headline direction, that exercise helps insulin handling in postmenopausal women, fits the broad expectation from the wider literature. The twist is the timing pattern: the clearest metabolic improvements came from the 3 to 4 month programmes, while the 6 to 12 month ones had little to show beyond the waistline. That could mean early gains fade, or simply that the longer trials differed in design, participants or adherence; with only three or four trials behind each comparison, the pooled data cannot say which. This is a finding to be extended, not yet a settled rule about how long exercise needs to work.

For a woman past menopause, this is encouraging news about how quickly structured exercise can register on metabolic markers: measurable changes in insulin and insulin resistance within 3 to 4 months. It is not evidence that exercising for longer is pointless; waist circumference still fell in the 6 to 12 month programmes, and these trials were not built to compare durations head to head. The reasonable reading is that metabolic benefits need not take years to appear, and that progress is worth judging by more than one number.