UK ten-year records show weight loss peaking near 31% at two years, then settling around 25%

Surgeons and diabetes doctors at Salford Care Organisation, part of the Northern Care Alliance NHS Foundation Trust in Greater Manchester, have published a decade of weight records from their bariatric service. The paper appeared online in Obesity Surgery on August 28, 2026.

The headline is a shape rather than a single number. Average weight loss climbed to a peak at two years and then drifted back over the following eight, without returning anywhere near the starting point.

The measure used throughout is percentage total weight loss, written %TWL. It is simply the share of your starting body weight that you have lost. Someone who starts at 300 pounds and loses 90 pounds has a %TWL of 30.

Average loss peaked at 31.3% at two years and sat at 25.1% at ten

Across 2,092 patients, mean total weight loss was greatest at two years, at 31.3% with a standard deviation of 10.7. By ten years the average was 25.1% with a standard deviation of 12.6.

The standard deviation matters more than the average here. A figure of 12.6 at ten years means individual results were spread very widely around that 25.1%. Plenty of people were well above it and plenty were well below. This is a description of a group, not a prediction for a person.

What the numbers do show is that gradual regain after year two was the normal pattern rather than a failure, and that most of the loss was still there a decade later.

Bypass and one-anastomosis bypass beat sleeve by about five percentage points through year five

Up to five years, Roux-en-Y gastric bypass produced 27.9% total weight loss with a standard deviation of 11.5, across 1,080 patients. One-anastomosis gastric bypass produced 28.3% with a standard deviation of 10.1, across 217 patients. Sleeve gastrectomy produced 22.8% with a standard deviation of 12.5, across 795 patients. The comparison was significant at p below 0.005.

This is not a randomized comparison. Surgeons and patients chose the operation together, based on things like reflux, diabetes, starting weight and prior surgery. Those same factors also influence weight loss, so some of the gap between operations reflects who had which operation rather than the operation itself. The authors name this problem directly, calling it confounding by indication.

Deprivation and early weight loss predicted who regained

Six factors were significantly associated with how much weight someone lost in the first two years: age, sex, preoperative body mass index, diabetes, the type of operation, and smoking.

Four factors were significantly associated with regain after year two: age, high blood pressure, the Index of Multiple Deprivation quintile, and how much weight the person had lost in those first two years.

The Index of Multiple Deprivation is an official English measure that ranks small neighbourhoods by income, employment, health, education, crime, housing and local environment. Finding it among the predictors of regain means people from more deprived areas were more likely to regain weight after surgery. That is a statement about circumstances, not about effort, and it is the kind of finding that argues for more follow-up support rather than less.

The other predictor worth noting is early weight loss itself. How the first two years go carries information about the eight that follow.

Only 29% of the group was still being weighed at ten years

This is the limitation that governs how much weight to put on the ten-year figure. Follow-up fell from 86% at six months to 29% at ten years. The 25.1% average at year ten describes the people who were still attending, and people who stop coming to appointments may differ from those who keep coming, in either direction. If people who regained stopped attending, the true ten-year average is worse than 25.1%. If people doing well stopped bothering, it is better.

Other limits the authors state: the study was retrospective and observational; the ten-year sample was small; 75.8% of the cohort was female; there was no comparison of complications, side effects or quality of life; data on diabetes severity such as HbA1c, remission and insulin use was missing; and there was no information on diet or physical activity after surgery.

One service, one region, one health system. The shape of the curve is likely to travel. The exact numbers are less likely to.