The 2.6 Percent: Surgery and Incretins in the Oldest Cohort Yet

A 107,280-patient national comparison finds substantially better weight and cardiometabolic outcomes with metabolic surgery than with incretin therapy in adults 65 and older who have type 2 diabetes — while the surgical arm holds just 2.6 percent of the cohort, and registry data place the age threshold exactly where the question begins.

bariatric surgeryGLP-1diabetes remissionaccess to care

The American population over 65 is projected to double by 2050, and obesity in that group is climbing with it. The two treatments competing for those patients — an operation and an injectable — have never met in a randomised trial. What exists instead is a retrospective cohort in Surgery for Obesity and Related Diseases, drawn from a national electronic-record network, and it is the largest comparison of its kind in this age group.

What 107,280 records show

A University of Virginia and Mass General Brigham team used the Epic Cosmos database to assemble adults 65 and older with type 2 diabetes and a BMI of at least 35 kg/m² who began metabolic surgery or incretin-based therapy between 2017 and 2023: 2,843 operated, 104,437 medicated. Overlap weighting estimated the effect where either choice was clinically plausible.

Mean one-year total weight loss was 23.0 percent after surgery, 6.87 percent after medication. At five years the surgical group had lower cumulative incidence of major adverse cardiovascular events (hazard ratio 0.84, 95 percent CI 0.74–0.95), nephropathy (0.73, 0.64–0.85), retinopathy (0.65, 0.55–0.77) and death (0.66, 0.54–0.80). HbA1c fell 0.80 percentage points against 0.54 — a real difference, modest beside the weight figures. The kidney finding runs with earlier registry comparisons of surgery and GLP-1 therapy, covered here in October.

The number that is not in the abstract

Two thousand eight hundred and forty-three is 2.6 percent of the cohort. That ratio is the finding the paper does not dwell on. Every eligible patient here had diabetes and a BMI of at least 35; roughly one in forty was referred for an operation. Whatever the hazard ratios say about biology, the referral pattern says something about practice.

The endpoints need the same scrutiny. Retinopathy and nephropathy are diagnosed by screening, and patients inside a surgical pathway are examined on a schedule medication-only patients often are not, so a hazard ratio of 0.65 can describe eyes that were looked at as readily as eyes that stayed well. Mortality and cardiovascular findings are less exposed to that objection but inherit the limits of administrative data: the reasons a 68-year-old was cleared for surgery are not fully in the covariates.

Where the risk actually turns

Safety in this population has its own literature, and it converges near 65. A Surgical Endoscopy analysis of 45,775 MBSAQIP gastric bypass patients over 60, matched 1:1 on 22 preoperative characteristics, compared adjacent age bands. Between 60 to 64 and 65 to 69, the older group had higher mortality, more pulmonary complications, unplanned ICU admissions, transfusions, readmissions and non-home discharges. Above that line the gaps narrowed and sometimes vanished.

The same registry, at the far end, reports 2,539 patients aged 75 and older operated between 2015 and 2023 — median age 76, three-quarters treated with a sleeve: complications in 10.4 percent, severe complications in 6.0 percent, mortality 0.3 percent. Sleeve beat bypass on both totals, 8.7 against 15.8 percent and 4.7 against 10.2. Respectable for a septuagenarian operation, and a reminder of how those numbers were earned: a registry describes the patients who were selected and cleared, not the ones never referred.

Remission follows the clock, not the calendar

For diabetes remission, duration of disease matters more than age, which turns the argument toward timing rather than candidacy. A Polish multicentre cohort followed 129 patients aged 60 and older with type 2 diabetes for a median of six years after sleeve gastrectomy. Median total weight loss was 19.4 percent, excess weight loss 47.8 percent, both below the figures quoted for younger patients. Remission came in 62.0 percent, 12.4 percent had no clinical change, and a diabetes history longer than ten years predicted failure.

That is awkward for a field organised around BMI thresholds — the same ones the eligibility criteria on our BMI calculator formalise. A 66-year-old diagnosed at 52 is not the same patient as a 66-year-old diagnosed at 64. The cohort is retrospective and single-country, and remission was read at last contact rather than on a fixed schedule, but the direction matches what surgeons have argued: operate before the beta cell gives up.

The muscle neither arm escapes

For patients over 65 the strongest objection to aggressive weight loss is not perioperative risk. It is lean mass. A review in Current Opinion in Clinical Nutrition and Metabolic Care finds GLP-1 efficacy in older adults comparable to younger ones, with cardiovascular benefit that matters more where disease burden is higher, then lists the costs: gastrointestinal intolerance, muscle loss, ophthalmologic complications, hypoglycaemia with insulin or sulfonylureas, and polypharmacy.

A second review, in Drugs, complicates the measurement. Lean-mass estimates from DXA or bioimpedance do not quantify contractile muscle, so a falling number is not proof of functional loss; strength and physical performance are what count, and pivotal trials rarely prespecified them. Its remedy — functional monitoring, resistance exercise, individualised glycaemic targets — applies equally to the surgical patient, whose loss is faster than any drug’s.

What would settle this is the design that does not yet exist: a prospective cohort, or a trial, of patients over 65 assigned to operation or incretin and followed on function as well as weight. The LABS study, funded by NIDDK, remains this archive’s reference for durability — 4,776 participants with an average BMI of 46.5, operated between 2005 and 2007 — but it was built around a mid-life cohort and never powered to test an age threshold. A field now operating on patients a generation older is still borrowing evidence from a younger one.

Sources

  1. Metabolic and bariatric surgery versus incretin-based therapy in adults aged 65 and older with type 2 diabetes: a nationwide comparative analysis of weight loss and cardiometabolic outcomes (Surgery for Obesity and Related Diseases, 2026)
  2. Determining the age cut-off for increased surgical risk after Roux-en-Y gastric bypass in elderly patients: MBSAQIP 2015–2023 database analysis (Surgical Endoscopy, 2026)
  3. 30-day Outcomes of Bariatric Surgery in Patients Over 75: Insights from the MBSAQIP Database (Obesity Surgery, 2026)
  4. Type 2 Diabetes Mellitus Remission After Sleeve Gastrectomy in Patients 60 Years and Older — A Multicenter Study (Journal of Clinical Medicine, 2026)
  5. The use of glucagon-like peptide-1 agonists as a treatment for obesity in older people: risks and benefits (Current Opinion in Clinical Nutrition and Metabolic Care, 2026)
  6. Glucose-Lowering Therapies in Type 2 Diabetes-Associated Sarcopenia: Mechanisms, Clinical Evidence and a Muscle-Centred Framework (Drugs, 2026)
  7. Bariatric Surgery — health information (NIDDK)
  8. Longitudinal Assessment of Bariatric Surgery (LABS) — study description (NIDDK)
From the blog

New research notes