Excess-weight Metrics Misrepresent Weight Loss and All Weight Metrics Poorly Predict Early Glycemic Outcome After...

Metabolic and bariatric surgery (MBS) outcomes are reported as weight loss, using percent total weight loss (%TWL), percent excess BMI loss (%EBMIL), or excess body-weight loss (%EBWL). We asked whether these metrics misrepresent weight loss and how well any of them predicts the early glycemic outcome at 6 months.

Using 2023 MBSAQIP data with 6-month follow-up, we studied sleeve gastrectomy and gastric bypass. We examined each metric's dependence on baseline BMI, then tested each as a predictor of HbA1c change and early glycemic response (HbA1c below 7.0% in patients uncontrolled at baseline), against a baseline-severity benchmark.

Of 182,826 patients, 136,406 had a paired 6-month BMI and 27,517 had paired HbA1c, the remainder lost to follow-up. Only %TWL was independent of baseline BMI (rho - 0.016). %EBMIL and %EBWL were strongly baseline-dependent (rho - 0.458 and - 0.337), and %EBMIL exceeded 100%, an uninterpretable value, in 1.46% of patients. For early glycemic response, all metrics discriminated poorly (AUC 0.60 to 0.63). Baseline severity alone reached AUC 0.768. Weight and glycemic responses were discordant in 47.4% of patients (kappa 0.119, balanced accuracy 59.2%).

The excess-weight metrics misrepresent how much weight was lost, most severely at low baseline BMI. No weight metric predicted the early glycemic outcome beyond what baseline severity already explained, and discrimination was weak for all of them. Glycemic and other metabolic outcomes should be measured directly rather than inferred from a weight metric, and prioritized in patients with type 2 diabetes.

• Excess-weight metrics depend on baseline BMI and distort measured weight loss. • %EBMIL exceeded 100%, an uninterpretable value, in 1.46% of a national cohort. • No weight metric, including %TWL, predicted early glycemic response well. • Baseline metabolic severity outperformed every weight metric as a predictor.