Percentile-based weight loss benchmarks after sleeve gastrectomy, Roux-en-Y gastric bypass, and duodenal switch: a multi-site cohort study.

BACKGROUND: Weight loss after bariatric surgery varies substantially among individuals, yet outcomes are usually reported as averages. Percentile benchmarks may better support patient counseling. OBJECTIVES: To develop percentile-based weight loss benchmarks after primary sleeve gastrectomy (SG), Roux-en-Y gastric bypass (RYGB), and duodenal switch/single-anastomosis duodeno-ileal bypass (DS/SADI). SETTING: Five-campus academic health system, United States. METHODS: Retrospective cohort study of 2730 primary bariatric patients (1796 SG, 663 RYGB, 271 DS/SADI; 2021 to 2026). Of 2928 eligible patients, 2730 (93.2%) had a postoperative follow-up weight; each contributed one most-recent weight (median follow-up 22.8 months; IQR, 9.7 to 36.1). Percent total body weight loss (%TBWL) benchmarks at the 10th through 90th percentiles were estimated using quantile regression adjusted for age, sex, initial body mass index (BMI), procedure, time, and procedure-by-time interaction, with 95% confidence intervals from bootstrap resampling. RESULTS: Among 2730 primary bariatric patients (1796 SG, 663 RYGB, 271 DS/SADI), the three operations tracked closely through the first 6 months, where DS/SADI was not yet distinguishable from SG (P = .14), then pulled apart after the first year and stayed separated through 36 months. For a reference patient (woman, age 42, BMI 45), predicted median %TBWL at 12 months was 24.5% for SG, 30.0% for RYGB, and 33.4% for DS/SADI; by 24 months it was 23.7%, 30.7%, and 36.2%. In practical terms, RYGB ran about 5.5 percentage points ahead of SG at 1 year and 7.0 at 2 years, and DS/SADI led SG by 8.9 and 12.5. But the range matters as much as the average: at 24 months, %TBWL ran from 10.3% at the 10th percentile to 38.1% at the 90th for SG, 16.3% to 44.3% for RYGB, and 17.8% to 47.1% for DS/SADI, so one in ten SG patients lost about 10% or less while another one in ten lost nearly 40%, and a strong SG result matched a typical RYGB. Internal calibration was good (mean error 1.6 percentage points), and the ordering held across sensitivity analyses. CONCLUSIONS: Percentile benchmarks allow a surgeon to show patients what weight loss really looks like after each operation, side by side and out to 3 years, as a range rather than a single number. This is easier to use and more honest for counseling than the average, and it covers the window beyond the first year, when weight loss is still settling.