Explore this article's sources with AI
Follow PeptideMethods on Google
Semaglutide Safety by Frailty in SELECT
A SELECT post hoc analysis found no detectable difference in cardiovascular benefit across baseline frailty categories, with important safety limits.
Why we wrote this. A JAMA Medical News item renewed attention to the SELECT frailty analysis, whose reassuring subgroup findings require careful limits and primary source context.
In this article (6 sections)
A post hoc analysis of the SELECT randomized trial found no statistically detectable difference in semaglutide's cardiovascular effect across three baseline frailty categories. The result is reassuring within SELECT, but it does not establish that semaglutide is safe for every person with frailty. The analysis involved adults with established cardiovascular disease and overweight or obesity who did not have diabetes, so its conclusion belongs inside that defined population[1].
The question received renewed attention in a JAMA Medical News article published on October 30, 2026[2]. The underlying evidence is a JAMA Cardiology secondary analysis of SELECT. Readers looking for drug background can visit the semaglutide evidence page, while the United States semaglutide regulation page covers a separate regulatory question. Neither page changes what this trial can establish.
How the frailty analysis was designed
SELECT was a multicenter, double blind, randomized, placebo controlled cardiovascular outcomes trial. It enrolled 17,604 adults aged 45 years or older who had established cardiovascular disease and a body mass index of at least 27, without diabetes. Participants received once weekly semaglutide 2.4 mg or placebo in addition to usual care. The original trial followed participants for a mean of 39.8 months[3]. These are study design details, not instructions for personal treatment.
For the secondary analysis, investigators constructed a 31 item frailty index using the cumulative deficit approach. They classified participants as not frail when the index was 0.210 or lower, more frail from 0.211 through 0.310, and most frail at 0.311 or higher. Of all participants, 5,432 were not frail, 8,349 were more frail, and 3,823 were most frail. The mean age was 61.6 years, and 72.3% of participants were men[1].
That definition matters. A cumulative deficit index summarizes multiple health deficits, but it is not interchangeable with every clinical frailty assessment. SELECT also excluded people with diabetes and required prior cardiovascular disease. The analysis therefore does not represent all older adults, all people living with frailty, or everyone using a GLP-1 receptor agonist.
What happened to cardiovascular outcomes
The primary outcome combined cardiovascular death, nonfatal myocardial infarction, and nonfatal stroke. Compared with placebo, estimated hazard ratios for semaglutide were 0.84 in the not frail category, 0.70 in the more frail category, and 0.92 in the most frail category. Their 95% confidence intervals were 0.65 to 1.07, 0.59 to 0.82, and 0.76 to 1.10, respectively. The interaction P value was .09[1].
An interaction test asks whether the treatment effect differs across subgroups beyond what random variation might explain. Here, the test did not meet the study's statistical threshold for detecting different effects by frailty category. That is not proof that the three hazard ratios are identical, and the individual category estimates should not be ranked as if SELECT directly tested which frailty group benefits most.
The investigators also reported no detectable variation by frailty category for a composite heart failure outcome, all cause hospitalization, or all cause mortality. Interaction P values were .82, .71, and .28, respectively[1]. These secondary findings describe consistency within the analyzed trial. They do not convert a subgroup analysis into a treatment recommendation for an individual patient.
What the study says about safety
The primary analysis reported that adverse events leading to permanent discontinuation showed a different relative pattern across baseline frailty categories, with an interaction P value below .001. The hazard ratio for semaglutide versus placebo appeared lower as baseline frailty increased[1]. This does not mean adverse events disappeared in the most frail group, nor does it show that greater frailty protects someone from side effects.
The broader SELECT result provides essential context. Across the entire randomized population, adverse events that led to permanent discontinuation occurred in 16.6% of participants assigned semaglutide and 8.2% assigned placebo[3]. The subgroup paper and the overall trial answer different questions. One evaluates whether the relative discontinuation pattern varied with baseline frailty, while the other reports total discontinuations by randomized group.
Safety cannot be reduced to one discontinuation endpoint. Frailty may coexist with reduced appetite, mobility limitations, multiple medicines, or other illnesses that affect how a person experiences treatment. SELECT did not test a frailty specific monitoring plan, and this analysis does not provide a rule for starting, continuing, or stopping semaglutide. Clinical decisions require review of an individual's history, goals, symptoms, and current medicines.
Frailty status and quality of life
Between baseline and week 104, participants assigned semaglutide were more likely to move into a better frailty category than those assigned placebo, with an odds ratio of 2.46 and a 95% confidence interval from 1.80 to 3.37. They were also less likely to move into a worse category, with an odds ratio of 0.47 and a 95% confidence interval from 0.34 to 0.65[1].
Health related quality of life, measured with the EQ-5D-5L instrument, appeared to improve more with semaglutide at higher baseline frailty, with an interaction P value of .02[1]. Both findings are secondary outcomes. They are useful signals, but they do not show that semaglutide treats frailty as a syndrome, prevents disability in every patient, or substitutes for a full clinical frailty assessment.
Why the post hoc design limits certainty
A post hoc analysis asks an additional question using data from a completed trial. Randomization still protects the original comparison between semaglutide and placebo, but it does not guarantee that every frailty subgroup comparison has enough statistical power or freedom from chance findings. Multiple outcomes and interactions also increase the need for cautious interpretation.
The original SELECT trial found a primary cardiovascular event in 6.5% of participants assigned semaglutide and 8.0% assigned placebo, corresponding to a hazard ratio of 0.80[3]. The frailty analysis adds evidence that this relative effect was not detectably different across its prespecified index categories. It does not extend the randomized evidence to people who would not have qualified for SELECT.
The most accurate answer to the headline question is therefore conditional. In SELECT, investigators did not detect worse cardiovascular efficacy in participants with higher baseline frailty, and the safety analysis did not reveal a simple signal that discontinuation risk became less favorable as frailty increased. Important limits remain because this was a secondary analysis, the trial population was selective, and adverse event discontinuation is only one dimension of safety.
Bottom line
This analysis supports the consistency of semaglutide's cardiovascular benefit across the frailty categories used in SELECT. It also offers useful quality of life, frailty transition, and discontinuation findings. It does not prove universal safety in frailty, establish identical effects across groups, or provide individualized prescribing guidance. A clinician should interpret the evidence alongside a person's functional status, nutrition, other medicines, cardiovascular history, and treatment priorities.
Medical disclaimer: This article is for educational and journalistic purposes only and does not constitute medical advice. Peptides discussed may be classified as prescription medicines or research chemicals depending on your jurisdiction. Always consult a qualified healthcare professional before using any peptide product. PeptideMethods.com does not sell, distribute, or facilitate the sale of any peptide product.
Frequently asked
Did SELECT prove semaglutide is safe for everyone with frailty?
No. The analysis was limited to SELECT participants with established cardiovascular disease and overweight or obesity who did not have diabetes. It cannot establish safety for every person with frailty.
Did frailty change semaglutide's cardiovascular effect?
Investigators did not detect statistically significant variation in the primary cardiovascular effect across the three baseline frailty categories. That does not prove the effects were identical.
What safety result needs the most context?
The relative pattern of adverse events leading to permanent discontinuation varied by frailty category, but discontinuations were more frequent with semaglutide than placebo in SELECT overall. These comparisons answer different questions.
Sources
- [1]Ostrominski JW et al. Efficacy and Safety of Semaglutide According to Frailty Status: A Post Hoc Analysis of the SELECT Randomized Clinical Trial. JAMA Cardiology. 2026. PMID 42747817.Tier 1 · primary↩
- [2]Bock A. Analysis Asks: Are GLP-1 Drugs Safe in People With Frailty? JAMA. 2026. PMID 42961402.Tier 1 · primary↩
- [3]ClinicalTrials.gov. SELECT cardiovascular outcomes trial, NCT03574597, study record and posted results.Tier 1 · primary↩
No revisions yet. First published .