Coming off a GLP-1 after 65: sarcopenia, falls, and why the plan changes

Most of what gets written about stopping a GLP-1, including our own article on the subject, is written for a general adult population. The regain data, the muscle-preservation advice, and the "plan your exit" framing all apply after 65 too. But something else is layered on top once someone is 65 or older, and it rarely gets its own treatment: a body that was already losing muscle and bone before the medication started, at a pace research has quantified independently of any drug.

That is not a reason to avoid a GLP-1 later in life, and it is not a reason to panic about one already in progress. It is a reason to know what the evidence specifically says about this age group, where that evidence is reassuring, where it is genuinely unsettled, and what changes in practice.

The baseline math is already working against muscle after 65

Sarcopenia, the age-related loss of muscle mass and function, is not a rare condition confined to frail nursing-home populations. A systematic review and meta-analysis pooling 151 studies and more than 690,000 participants found prevalence estimates for adults 60 and older ranging from about 10 percent using the strict EWGSOP2 criteria to about 27 percent using broader muscle-mass-only definitions, with severe sarcopenia affecting roughly 2 to 9 percent.² A separate, more specific concern, sarcopenic obesity, the coexistence of reduced muscle with excess body fat, is estimated to affect 10 to 20 percent of older adults and about 20 percent of older adults specifically seeking obesity treatment.¹ Sarcopenic obesity carries elevated cardiometabolic and frailty risk beyond what either low muscle or high fat carries alone.⁹

None of that requires a GLP-1 to be true. It is the starting line a meaningful share of older adults are already standing on before a prescription is written. Reviews of GLP-1 trial data generally, across all ages, place lean mass at roughly 15 to 40 percent of total weight lost.¹⁰ Losing that same percentage from a smaller muscle reserve to begin with is a different proposition than losing it from a larger one, even though no trial has directly quantified how much different.

What the GLP-1 trials actually show about people 65 and older

Here the picture is more reassuring than the framing above might suggest. A 2026 post hoc analysis of the SURMOUNT and SUMMIT tirzepatide obesity trials, isolating adults 65 and older, found clinically meaningful weight reduction with favorable cardiometabolic and quality-of-life differences comparable to the trial's younger participants. On safety, the analysis found no clinically meaningful differences in adverse event rates between tirzepatide and placebo in older adults for gastrointestinal tolerability, falls, fractures, depression outcomes, pancreatitis, or renal, hepatic, and gallbladder-related events. Fractures and falls occurred more often in the older subgroup overall, as expected at that age, but at comparable rates on tirzepatide and placebo.⁵

A separate, differently designed study complicates that reassurance. A 2026 real-world analysis in the Journal of Clinical Endocrinology & Metabolism followed 46,177 adults 65 and older with type 2 diabetes, comparing those newly starting a GLP-1 against those starting an SGLT2 or DPP-4 inhibitor. Over a median follow-up of 34.7 months, GLP-1 users showed an 11 percent higher risk of fragility fracture (hazard ratio 1.11, 95% CI 1.01 to 1.21).⁴ That is a real signal from a large, comparator-controlled dataset.

These two findings have not been reconciled, and this article will not pretend to. They come from different populations, different designs, and different comparators: obesity-trial participants versus a type 2 diabetes cohort, a randomized trial versus an observational comparison, placebo versus other diabetes medications. The honest summary is that randomized evidence in people 65 and older does not currently show an elevated fracture or fall signal, while a large real-world study in an older diabetes population does. Anyone weighing a GLP-1 in this age group deserves to know both results exist, not just the more comfortable one.

What actually predicts falls, and what does not

It is tempting to assume that losing muscle mass directly raises fall risk, but the largest analysis on this question complicates that assumption. A 2025 meta-analysis of 34 cohort studies covering 57,449 older adults, paired with a Mendelian randomization analysis using UK Biobank and FinnGen data, tested which sarcopenia-related traits actually predict falls. Walking speed showed a strong, consistent relationship: each unit increase was linked to a 33 percent reduction in fall risk in the cohort data (odds ratio 0.67), a relationship the genetic analysis supported as likely causal (odds ratio 0.64, p = 0.001). Grip strength showed only a small protective association that did not hold up as causal. Appendicular lean mass, the muscle-mass number closest to what a DXA scan reports, showed no significant association with falls in either analysis.³

That lines up with something this site says about muscle loss generally: track function, not the mass number. For fall risk specifically, gait speed is the function that matters in the evidence, not how much lean tissue shows up on a scan, and it is something anyone can informally track over time without special equipment.

Why the standard off-ramp advice needs an update after 65

The exit itself is the same event for everyone: appetite returns, and weight tends to follow it back. In the STEP 1 trial extension, participants who stopped semaglutide regained about two-thirds of their lost weight within a year, with mean net weight loss falling from 17.3 percent to 5.6 percent.⁸ Discontinuation itself is common at any age; roughly 65 percent of GLP-1 users in a large real-world cohort had stopped within 12 months.¹¹ None of that changes after 65.

What changes is the cost of the same event. A younger adult who regains weight as a mix of fat and muscle is drawing down a larger reserve, with more physiological capacity to rebuild lean tissue afterward through training. An older adult, especially one who started with sarcopenia or was already trending toward sarcopenic obesity, is drawing down a smaller reserve and faces a harder, slower rebuild.¹ No trial has directly measured lean-mass change through discontinuation and regain in an older-adult-specific GLP-1 population, so this is a reasoned extrapolation, not a demonstrated finding. But the direction is consistent across every source cited here: less reserve at the start, the same regain pattern at the end, and a real-world fracture signal specific to this age group that a younger user does not carry. The practical shift is less a different protocol than a shorter margin for error; a missed month of training matters more against a smaller starting reserve.

What protects muscle and bone after 65, specifically

The two levers with the strongest evidence are not age-specific inventions, and one was tested directly in this population. A meta-analysis of randomized trials in older, obese adults losing weight through caloric restriction found that adding resistance training offset about 93.5 percent of the lean-mass loss that dieting alone produced, roughly 0.82 kg of preserved lean tissue.⁷ That trial population is the closest match in this article to its readers: not young athletes, older adults under a real caloric deficit.

Protein is the second lever. A dose-response meta-analysis of 49 studies found protein supplementation continued improving resistance-training outcomes up to about 1.6 grams per kilogram of body weight per day.⁶ That target does not shift for older adults in the pooled data, though hitting it against a suppressed appetite and, for some, reduced kidney reserve, is a conversation for a prescribing clinician rather than a number to chase blindly. Our guide to the best exercises to preserve muscle on a GLP-1 covers the movement patterns that matter most regardless of age or equipment.

What the evidence does not say

Several honest limits apply here, more than in most articles on this site, because this population is underrepresented in the trials that would answer the question directly. No randomized trial has isolated a structured resistance-training and protein program specifically in older adults coming off a GLP-1; that evidence is strong but drawn from broader older-adult weight-loss populations, not from people discontinuing these medications. The fracture and falls data point in different directions depending on study design, and that tension is unresolved rather than one study being wrong. The falls meta-analysis found gait speed predictive of falls but did not test whether GLP-1-driven weight loss changes gait speed itself, so its relevance here is inferential. And the sarcopenic obesity prevalence figures describe the general older population, not GLP-1 users specifically, since these trials do not typically report body composition broken out by decade of age.

The bottom line

Nothing about turning 65 makes a GLP-1 unsafe, and the largest randomized evidence available in this age group, the tirzepatide post hoc analysis, is genuinely reassuring on falls and fractures specifically. But the same age group carries a documented, independent risk of sarcopenia and sarcopenic obesity before any medication is involved, a large real-world study found an elevated fracture signal in an older diabetes population, and the broader falls literature says the muscle-mass number is less important than how fast someone walks. None of that argues against the medication. It argues for treating resistance training and protein as closer to non-negotiable after 65 than at 35, for tracking function rather than a scan, and for planning the exit, if and when it comes, with a clinician who knows the full picture rather than with the same generic timeline written for a 40-year-old.

Frequently asked