In the 14-year ARMMS-T2D follow-up of 262 adults with type 2 diabetes (mean BMI 36.4, HbA1c 8.5%), 58% randomized to metabolic-bariatric surgery reached remission (HbA1c <6.5% off glucose-lowering drugs) versus 3% with intensive medical-lifestyle care. Shorter diabetes duration, no insulin use, fewer drug classes, higher fasting C-peptide and greater first-year weight loss predicted remission. Longer duration and more medication classes predicted relapse after initial remission.
The duration and insulin findings fit the beta-cell reserve model: surgery restores glycemic control best when the pancreas still produces enough insulin, and the higher C-peptide signal points the same way. Earlier surgery, rather than waiting for medication escalation, may therefore give the best odds of durable remission, a point worth raising with a physician for anyone with obesity and recently diagnosed diabetes. The 58% figure after a median of 11 years is a notably long horizon, since most remission data stop at five years, and the relapse analysis shows remission is often a window rather than a permanent cure. Caveats matter: the surgical arm combined three procedures, the sample is modest, the predictor analyses are observational even though the original randomization was not, and fewer participants remain at the latest time points. Weight change at one year is an outcome of surgery, not a baseline trait, so it cannot guide patient selection beforehand. Editorially, this is confirmatory but valuable, sharpening who benefits most rather than overturning the case for surgery.