Bariatric surgery was linked to fewer diabetes complications than obesity drugs

Design

Nationwide overlap-weighted retrospective cohort

Population

379,858 US adults with obesity and type 2 diabetes

Primary outcome

Five-year nephropathy, retinopathy, MACE and mortality

Effect

Death 2.1% vs 3.7% (HR 0.61, 0.55–0.69); MACE HR 0.78 (0.73–0.84)

This US nationwide cohort compared 32,401 adults with obesity and type 2 diabetes who had metabolic bariatric surgery with 347,457 who started obesity management medications, mostly incretin-based. Overlap weighting balanced the groups on measured characteristics. Median follow-up was about four to five years.

Surgery was associated with lower five-year nephropathy (4.5% vs 8.3%; HR 0.55), retinopathy (3.6% vs 7.4%; HR 0.60), major adverse cardiovascular events (6.1% vs 8.1%; HR 0.78) and all-cause death (2.1% vs 3.7%; HR 0.61). At one year, total weight loss was 25.4% after surgery and 6.4% on medication, and HbA1c below 6.5% was reached by 76.7% versus 47.2% of those starting above it. Results held across sensitivity analyses.

The weight loss on drugs here is far below trial figures, which almost certainly reflects real-world discontinuation and under-dosing. That is part of the point: surgery delivers its effect without depending on years of adherence and supply. For surgeons, this supports keeping surgery on the table for patients with diabetes who are drifting on medication, rather than treating drugs as the default end-point.

• Discuss metabolic surgery with eligible patients with type 2 diabetes, including those already on GLP-1 drugs
• Ask about drug persistence and actual weight loss when assessing patients referred from medical clinics
• Record baseline renal function and retinal screening before surgery to track microvascular benefit
• Expect real-world weight loss on medication to be well below trial figures

Why it matters

Challenges the assumption that incretin drugs have made surgery a second-line option for diabetes.

Don't overread it

This is observational and compares surgery with real-world drug use, not with patients who stayed on full-dose therapy.

The statistics, in plain English

Hazard ratios of 0.55 to 0.78 describe how much lower the event rates were after surgery. Because patients were not randomised, unmeasured differences — such as health motivation or access to care — could explain part of the gap, even after careful weighting.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.