A prospective cohort of 1,800 metabolically healthy adults wore consumer continuous glucose monitors (CGMs) for two-week windows every six months over five years. Analysis published this week shows that CGM-derived glycemic variability — specifically the coefficient of variation of post-prandial glucose excursions — flagged future pre-diabetes diagnoses an average of 24 months before standard HbA1c thresholds were tripped.
The result is consistent with a model in which pre-diabetic dysglycaemia manifests first as elevated post-prandial variability — sharp spikes and abnormal recoveries after meals — well before the averaged-over-months HbA1c measure shifts. HbA1c is the diagnostic standard, but it is a lagging indicator by design; the CGM signal is leading.
Translating the finding into clinical practice is non-trivial. CGMs were designed for diabetics, not for population screening; the cost-per-screening is still 5-10× higher than HbA1c, and the variability metric has no established reference range in healthy populations. The authors propose a follow-up trial restricting CGM screening to adults with one or more metabolic risk factors.