GLP-1s After 65: What the New Data Actually Shows

Until ECO 2026 opened in Istanbul on May 12, the question 'how well do GLP-1s actually work in adults over 65?' had no clean answer in the trial literature. The original SURMOUNT and STEP programs enrolled adults aged 18+ with no upper age cap, but they enrolled relatively few seniors and almost no one over 75. SELECT, the cardiovascular outcomes trial, was older on average — mean age 61 — and provided the first usable signal that the drugs work in patients with cardiovascular disease, but it wasn't designed to isolate the over-65 cohort.

That changed this week with two simultaneous releases. Prof. Luca Busetto's University of Padova group presented a pooled post-hoc analysis of STEP 1, 3, 4, 5, 8, and 9 covering 358 adults aged 65 and older — 248 received semaglutide 2.4 mg, 110 placebo. Dr. Deborah Horn at UTHealth Houston presented Eli Lilly's pooled ATTAIN-1 + ATTAIN-2 analysis covering 616 adults aged 65 or older on orforglipron (Foundayo). For the first time, prescribers have a real read on whether geriatric patients can expect the trial-headline weight loss when they actually start the drug.

Mean age in the Padova analysis was 69, 72% women, mean baseline weight 102 kg. Semaglutide 2.4 mg produced 15.4% mean body-weight loss vs 5.1% on placebo. 46.8% of the semaglutide arm reached at least 15% loss vs 6.4% on placebo, and 28.6% hit at least 20% loss vs 2.7%. Waist circumference dropped 14.3 cm on drug vs 6.0 cm on placebo. 27% of treated patients reached a healthy BMI under 27 vs 5.5% on placebo.

A couple of numbers stand out. First, 15.4% mean loss is meaningfully lower than the 16-17% reported in the full STEP populations — older patients lose slightly less. That's worth telling a 70-year-old who's read the marketing material. Second, the response distribution skews wider in older patients: more high responders and more partial responders relative to the under-65 cohort. The Padova group attributes this to age-related variability in absorption and lean-mass differences but didn't fully explain it.

The safety picture: serious adverse events occurred in 19.0% of semaglutide patients vs 12.7% on placebo. That gap is larger than the under-65 difference (~8-10 percentage points historically, ~6 points here). Constipation and dizziness were more common with the drug. Sarcopenia and falls — the two outcomes geriatricians care about most — weren't broken out in the public release.

The orforglipron senior analysis pooled 616 randomized participants from ATTAIN-1 (without T2D) and ATTAIN-2 (with T2D): 118 on 6 mg, 135 on 12 mg, 146 on 36 mg, and 214 placebo. The headline finding is more conservative than the Padova number: 'no overall differences in safety or effectiveness' between patients 65+ and younger, per the Foundayo US prescribing label.

13% of the combined ATTAIN-1 + ATTAIN-2 trial population was over 65, and 1% was over 75. Those numbers tell you the trial wasn't powered for a sharp geriatric subgroup readout; they tell you Lilly enrolled the floor of what FDA expected, not the ceiling. The 'no overall differences' framing is a reasonable read of the data Lilly has, but it doesn't mean orforglipron works identically across every decade of life — it means the effect is the same direction and similar magnitude, with confidence intervals that are wider in the older cohort.

Dr. Horn's framing at ECO 2026 emphasized that older adults have higher rates of obesity-related comorbidities (osteoarthritis, cardiovascular disease, sleep apnea, mobility limitation) than younger adults, so the per-patient benefit of substantial weight loss may actually be greater in geriatric populations even if the percentage loss is comparable. That's a defensible argument; whether it actually changes prescribing behavior in primary care is a separate question.

Neither ECO 2026 senior-population analysis reports muscle mass, grip strength, gait speed, or falls. That's the gap the field has been waiting on a year and didn't get on Tuesday.

Sarcopenia and frailty become inflection points around age 65. The cardiometabolic benefits of 15-20% weight loss are unambiguous on paper, but if a quarter of that weight comes off as lean tissue in a patient who started with marginal muscle mass, the calculus gets worse than the BMI number suggests. The 2026 Journal of Cachexia, Sarcopenia and Muscle scoping review covered the broader peptide-and-muscle-wasting literature; Nature Reviews Endocrinology has separately flagged sarcopenia as an under-recognized GLP-1 side effect.

For a 70-year-old patient with sarcopenia risk factors (low protein intake, pre-existing low muscle mass, sedentary), the right framing isn't 'is the drug effective' — it's 'what's the muscle-protection protocol that goes alongside it?' Resistance training two to three times a week, 1.2-1.6 g/kg of protein, and routine grip-strength or short-physical-performance-battery monitoring at six-month intervals. None of that appears in the senior-population analyses presented this week, and it should.

On the same May 12 ECO 2026 schedule, Novo Nordisk presented a SELECT post-hoc analysis stratified by reproductive life stage in women. Perimenopausal women with obesity and heart disease showed a 42% lower risk of major adverse cardiovascular events on semaglutide vs placebo; postmenopausal women showed 13% lower risk. The gap is striking, and Novo's framing emphasized that the perimenopausal benefit may be larger than the population-average SELECT result of 20% MACE reduction.

What that means for the over-65 prescribing question: most patients 65+ are postmenopausal women or men of similar age, and the SELECT benefit is real but smaller in that cohort than the 'GLP-1s cut heart attacks by 20%' headline suggests. The directionally favorable signal is solid; the magnitude is age-dependent and the trial wasn't powered to detect smaller subgroup effects with statistical certainty. A 70-year-old with established cardiovascular disease should expect a real but modest absolute risk reduction from semaglutide — somewhere between the SELECT headline number and zero, depending on baseline risk and concurrent therapy.

Most 65+ patients on a GLP-1 are also on at least three other prescription medications: a statin, an antihypertensive, a diabetes drug (often metformin), and frequently an SSRI or sleep medication. GLP-1s slow gastric emptying and can alter the absorption kinetics of orally administered drugs. The clinically significant interactions are modest but real: levothyroxine absorption may be reduced if taken with food in a slowed-gastric-emptying state, and oral diabetes drugs can stack with the GLP-1 effect to produce hypoglycemia if doses aren't adjusted.

A practical workup for any 65+ patient starting a GLP-1: review all oral medications and consider timing adjustments (most orally absorbed drugs should be taken 30+ minutes before the GLP-1 injection, or moved to a different time of day). Recheck TSH at 6-12 weeks if the patient is on levothyroxine — one tirzepatide analysis found ~29% of hypothyroid patients showed TSH suppression within six weeks, an effect that's pharmacologic, not artifactual. Adjust sulfonylureas and insulin downward before starting, not after the first hypoglycemic event. And run baseline labs: ferritin, B12, vitamin D, and a comprehensive metabolic panel. Senior patients are more likely to start with marginal stores of these nutrients, and GLP-1-related absorption issues will worsen them.

The falls question deserves explicit attention. Rapid weight loss in older adults shifts gait mechanics, blood pressure, and balance. A patient who weighed 95 kg and now weighs 78 kg is a different biomechanical system than the one their physical therapist last saw. PT reassessment at the 10% and 15% weight-loss thresholds is a reasonable habit for any prescriber working with geriatric patients.

The economic case for a 70-year-old patient is harder than the trial data alone implies. Medicare Part D still excludes drugs prescribed solely for weight loss under the 2003 Medicare Modernization Act, which means most over-65 patients in the US pay out of pocket for Wegovy or Foundayo even when their physician thinks the cardiovascular and metabolic benefits justify the prescription.

When the indication is type 2 diabetes (Ozempic, Mounjaro) or cardiovascular risk reduction (Wegovy got that label expansion in 2024, Ozempic in 2025), Part D coverage kicks in. The result is a fragmented coverage map: a 70-year-old with diabetes gets Ozempic at a copay, while a 70-year-old without diabetes but with the same BMI and the same cardiovascular risk pays $1,300+/month for Wegovy. The April 22, 2026 cancellation of CMS's BALANCE pilot — which would have created a coverage pathway for obesity-only Medicare patients — closed off the most promising near-term fix. The Medicare GLP-1 Bridge, scheduled for July 2026, may partially fill the gap at a $50/month patient cost for some beneficiaries, but the rules are restrictive and the rollout uncertain.

For cash-pay seniors, the practical options narrow to the savings-card programs (NovoCare's $499/month self-pay program for Wegovy, Lilly Direct at $499/month for lower Zepbound vials, $399/month for Wegovy HD 7.2 mg). Compounded semaglutide and tirzepatide are mostly off the table after the April 30 FDA proposal to strip them from the 503B bulks list, though the 503A pathway remains open through July 2026 PCAC.

The honest version of the conversation, given what's now in the literature:

Yes, these drugs work in your age group. The Padova STEP pool showed 15.4% mean weight loss on semaglutide 2.4 mg in patients with a mean age of 69. The Lilly ATTAIN pool showed effectiveness consistent with the under-65 cohort on orforglipron. The benefit is real and the side-effect profile is similar to the trial-wide reports.

You'll lose slightly less than the under-65 headlines suggest. Plan for 13-16% on semaglutide 2.4 mg rather than the 16-17% trial mean. Expect waist circumference to drop in proportion. Some patients exceed the trial mean; some come in below it. The drug works on a distribution, not a guarantee.

Muscle protection matters more for you than for a 40-year-old on the same drug. Build resistance training (two to three days a week of compound lifts or progressive bodyweight work) into the plan before starting, not after you've lost 10 kg. Hit 1.2-1.6 g/kg of protein daily. If you can't tolerate that volume of food, prioritize protein over everything else in your remaining appetite. The goal isn't just weight loss — it's losing fat while keeping function.

Keep your existing cardiologist, endocrinologist, and primary-care doctor in the loop. Multiple-prescriber polypharmacy is where the avoidable side effects happen. Recheck labs at 6-12 weeks and at every dose increase.

Finally: if you don't tolerate the dose, hold or step back. The trial-recommended titration was designed for the median patient. Senior patients often do better on a slower schedule and at a lower terminal dose. 14 mg orforglipron is plenty if 36 mg gives you GI issues. 1.7 mg Wegovy is plenty if 2.4 mg makes you stop eating entirely. The drug works on a curve and the right dose is whichever one you can stay on long enough to see results.