semaglutide beneficial in frail individuals

 A recent post hoc analysis of a study of semaglutide for weight loss patients found that frail patients did at least as well as patients who were not frail: see semaglutide safe in frail JAMAcardiol2026 in dropbox, or doi:10.1001/jamacardio.2026.3566

 

Details:

-- 17,604 participants were involved in the SELECT trial of participants who had a BMI of at least 27 and established cardiovascular disease but without diabetes, who were randomized to semaglutide 2.4mg weekly vs placebo

    -- 12,732 patients (72.3%) were male and 4872 (27.7%) were female

-- a 31-item frailty index (FI) was assessed using the Rockwood cumulative deficit approach; patients were categorized as not frail (FI up to 0.210), more frail (FI of 0.211 to 0.310), and most frail (FI at least 0.311)

    -- the Rockwood cumulative deficit assesses 30 to 40 diverse variables covering multiple organ systems, focusing on the presence of disease, mobility impairment, and abnormal lab values (ie, not focusing on a single pathology), and has scores ranging from 0 to 1.0

-- the following were largely equivalent between the three FI categories, but differences below are noted

-- mean age 62, race/ethnicity Black 6% (twice as high in the most frail group versus non-frail group) and Latino (13% in non-frail group and 9% in most frail)

-- BMI 33, waist circumference 112 cm

-- blood pressure 130/80, pulse 69

-- smoking status current 17%/ prior 48%/never 34% (never smokers were higher in the non-frail group)

-- medical history (these were more likely to vary by frailty score, as noted below), items below were comparing not frail vs most frail respectively

    -- myocardial infarction: 73% versus 78%

    -- PCI or CABG: 61% versus 72%

    -- symptomatic PAD: 6% versus 17%

    -- stroke 23%: overall; TIA 4.5% overall

    -- heart failure: 10% versus 43%

    -- hypertension: 64% versus 97%

    -- chronic kidney disease: 3% versus 26%

    -- MASLD: 3% versus 17% with MASH 0.1% vs 1.4%

    -- sleep apnea: 5% versus 29%

    -- asthma: 3% versus 14%

    -- COPD: 2% versus 21%

    -- knee osteoarthritis: 6% versus 33%

    -- hip osteoarthritis: 3% versus 17%

    -- gout: 3% versus 19%

-- laboratory data: not much difference in levels between A1c, CRP, eGFR, urine albumin/creatinine ratio, total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides, serum potassium, serum sodium, hemoglobin, ALT, total bilirubin

    -- EQ-5D score: 0.94 in the non-frail and 0.78 in the most frail

        -- scale is from 0 to 1 (0= dead, 1= full life)

    -- EQ-5D VAS score (the visual analog score): 82 in the non-frail and 70 in the most frail

        -- scale is from 100 (best imaginable health) to 0 (worst imaginable health)

-- overall, and not surprisingly, participants with higher frailty scores tended to have higher BMI, more obesity-related complications, and worse health-related quality of life

    --  those with higher frailty scores also tended to be older, be female, and had active or prior tobacco smoking

 

-- primary composite outcome: time to first cardiovascular death, nonfatal MI, or nonfatal stroke (ie, MACE: major adverse cardiovascular events)

-- key secondary outcomes: the individual components of the primary outcome; also, a composite heart failure outcome (cardiovascular death or heart failure events leading to hospitalization or urgent visit), all-cause hospitalization, and all-cause death

-- additional secondary outcomes include kidney composite outcome (kidney death, initiation of chronic kidney replacement therapy, persistent estimate of the eGFR of <15, persistent eGFR decline to <50% of baseline, or onset of urine albumin to creatinine ratio of >300 mg/g), progression to hemoglobin A1c at least 6.5%, changes in health-related quality of life as assessed through the EQ-5D-5L index score and EQ-5D VAS score

-- the post hoc analysis was also designed to assess the efficacy and safety of semaglutide and the effect of semaglutide on frailty trajectory

-- safety events assessed serious adverse events (AEs) leading to permanent discontinuation of the semaglutide irrespective of their seriousness

 

-- mean follow-up was 39 months

Results:

Treatment Effects of Semaglutide on the primary composite outcome by Frailty Status, over mean of 39-month followup:

-- patients on placebo:

    -- the incidence rates (IR) per 100 person years for the primary composite outcome: 

        -- FI ≤0.210 (not frail): 1.5

        -- FI 0.211-0.310 (more frail): 2.5

        -- FI ≥0.31 (most frail): 3.9 

    -- compared with placebo, semaglutide reduced the rate of the primary composite for baseline frailty rating:

        -- F1: IR 1.3, a not statistically significant 16% decrease, HR 0.84 (0.65-1.07)

        -- F2: IR 0.17, a significant 30% decrease, HR 0.70 (0.59-0.82)

        -- F3: IR 3.6, a not statistically significant 8% decrease, HR 0.92 (0.76-1.10)

            -- P=0.09 for the interaction

Treatment Effects of Semaglutide on the heart failure composite outcome by Frailty Status:

-- patients on placebo:

    -- the incidence rates (IR) per 100 person years for the heart failure composite outcome: 

        -- FI ≤0.210 (not frail): 0.5

        -- FI 0.211-0.310 (more frail): 1.2

        -- FI ≥0.31 (most frail): 2.6

    --Compared with placebo, semaglutide reduced the rate of heart failure composite outcome for baseline frailty rating:

        -- F1: IR 0.4, a not statistically significant 15% decrease, HR 0.85 (0.56-1.29)

        -- F2: IR 0.9, a significant 23% decrease, HR 0.77 (0.61-0.97)

        -- F3: IR 2.2, a not statistically significant 15% decrease, HR 0.92 (0.67-1.07)

            -- P=0.82 for the interaction

Treatment Effects of Semaglutide on the all-cause hospitalization outcome by Frailty Status:

-- patients on placebo:

    -- the incidence rates (IR) per 100 person years for the all-cause hospitalization outcome: 

        -- FI ≤0.210 (not frail): 10.1

        -- FI 0.211-0.310 (more frail): 14.1

        -- FI ≥0.31 (most frail): 21.5

    --Compared with placebo, semaglutide reduced the rate of the all-cause hospitalization for baseline frailty rating:

        -- F1: IR 9.0, a significant 11% decrease, HR 0.89 (0.80-0.99)

        -- F2: IR 12.5, a significant 11% decrease, HR 0.89 (0.83-0.96)

        -- F3: IR 18.2, a significant 15% decrease, HR 0.85 (0.73-0.93)

            -- P=0.71 for the interaction

Treatment Effects of Semaglutide on the all-cause death outcome by Frailty Status:

-- patients on placebo:

    -- the incidence rates (IR) per 100 person years for the all-cause death outcome: 

        -- FI ≤0.210 (not frail): 0.7

        -- FI 0.211-0.310 (more frail): 1.6

        -- FI ≥0.31 (most frail): 2.9

    --Compared with placebo, semaglutide reduced the rate of the all-cause death for baseline frailty rating:

        -- F1: IR 0.7, a not statistically significant 4% decrease, HR 0.96 (0.68-1.35)

        -- F2: IR 1.1, a significant 11% decrease, HR 0.72 (0.58-0.88)

        -- F3: IR 2.5, a not statistically significant 15% decrease, HR 0.85 (0.69-1.06)

            -- P=0.28 for the interaction

Treatment Effects of Semaglutide on the progression of HbA1C to a least 6.5% by Frailty Status:

-- patients on placebo:

    -- the incidence rates (IR) per 100 person years for the all-cause death outcome: 

        -- FI ≤0.210 (not frail): 2.5

        -- FI 0.211-0.310 (more frail): 4.2

        -- FI ≥0.31 (most frail): 5.4

    --Compared with placebo, semaglutide reduced the rate of the progression of HbA1C to a least 6.5%

        -- F1: IR 0.9, a significant 63% decrease, HR 0.37 (0.28-0.47)

        -- F2: IR 1.1, a significant 73% decrease, HR 0.27 (0.23-0.32)

        -- F3: IR 1.1, a significant 80% decrease, HR 0.20 (0.15-0.26)

            -- P=0.006 for the interaction

Treatment Effects of Semaglutide on the kidney composite outcome by Frailty Status:

-- patients on placebo:

    -- the incidence rates (IR) per 100 person years for the kidney composite outcome: 

        -- FI ≤0.210 (not frail): 0.3

        -- FI 0.211-0.310 (more frail): 0.5

        -- FI ≥0.31 (most frail): 1.7

    --Compared with placebo, semaglutide reduced the kidney composite outcome for baseline frailty rating:

        -- F1: IR 0.3, a not statistically significant 46% decrease, HR 0.84 (0.49-1.45)

        -- F2: IR 0.5, a not statistically significant 1% decrease, HR 0.99 (0.71-1.40)

        -- F3: IR 1.0, a significant  40% decrease, HR 0.60 (0.44-0.82)

            -- P=0.09 for the interaction

-- Treatment Effects of Semaglutide on the Continuous Outcomes by Baseline Frailty Status Between baseline and week 104, semaglutide decreased:

    --  bodyweight as compared with placebo: 

        -- FI: mean difference –8.16% (–8.58 to –7.73%)

        -- F2: mean difference –8.56% (–8.91 to –8.21%)

        -- F3: mean difference –8.91% (–9.43 to –8.39%)

            -- P = 0.08  for the interaction

    -- waist circumference: P< .001 for interaction

    -- waist to height ratio: P < .001 for interaction

    -- HbA1c value: P < .001 for interaction

    -- high-sensitivity C-reactive protein: P = .03 for interaction

    -- reductions in diastolic blood pressure were attenuated with a higher FI category (P = .03 for interaction).

    -- effects of semaglutide on other continuous end points, including systolic blood pressure, heart rate, and lipid levels were similar across baseline FI categories

    -- participants with a higher baseline FI experienced greater improvement in health-related quality of life as measured by EQ--5D-5L index scores (P = .02 for interaction), driven by the usual activities reported (P = .03 for interaction), self-care (P = .049 for interaction), and mobility (P = .05 for interaction)

        --Benefits of semaglutide on EQ-5D VAS (visual analog scale) scores (P = .45 for interaction) appeared consistent regardless of baseline FI category, per graph below, with some pictorial improvement with increasing stage of fragility:

​​

Treatment Effects of Semaglutide on the Frailty Index:

    -- the FI modestly worsened (increased) in the placebo group (mean [SD] of 0.002 [0.001]) but improved (decreased) in the semaglutide group (mean [SD] of, –0.030 [0.001]) between baseline and week 104 (estimated treatment difference, –0.031 (–0.033 to –0.030); P < .001.

    -- participants randomized to receive semaglutide were more likely to experience an improvement in FI category vs placebo (33.7% vs 18.0%, respectively; odds ratio [OR] for improvement, 2.46 (1.80 to 3.37), P < .001) and less likely to experience progression (10.0% vs 19.3%, respectively; OR for worsening, 0.47 (0.34 to 0.65), P < .001

    -- Among participants with baseline FI up to 0.210 (not frail), treatment with semaglutide additionally reduced transition to higher FI categories vs placebo (19.0% vs 33.1%, respectively; OR 0.47 (0.42 to 0.54), P < .001

Safety Events by Baseline Frailty Status Key safety events:

    -- whether on semaglutide or placebo, incidences of serious AEs increased with a higher baseline FI category.

    -- however, these events occurred less frequently in participants randomized to receive semaglutide vs placebo:

        -- this was true in all FI subgroups: 25.7% vs 27.5% if FI was ≤0.210; 33.6% vs 36.5% if FI was 0.211-0.310; and 43.8% vs 49.0% if FI ≥0.311, respectively

            -- and this seemed to be driven by lower risks of serious cardiac, infectious, and respiratory disorders

        -- incidences of serious gastrointestinal and nutritional disorders were comparable between treatment arms in all FI subgroups

    -- the incidence of pneumonia appeared lower with semaglutide vs placebo with a higher FI. Serious bone and joint events were uncommon among SELECT participants in all FI subgroups and were not increased with semaglutide

    -- AEs leading to permanent discontinuation were more common with higher baseline FI in both treatment groups and occurred more frequently among those randomized to semaglutide vs placebo 

        -- however, relative risks of these events were not enhanced among participants with a higher FI category (15.5% vs 5.5% if FI was ≤0.210; 16.8% vs 8.3% if FI was 0.211-0.310; and 17.6% vs 11.7% if FI was ≥0.311)

        -- participants with a higher baseline FI appeared to exhibit an incrementally lower rate of permanent treatment discontinuation due to AEs with semaglutide vs placebo (P < .001 for interaction)

    -- similar findings were observed for permanent treatment discontinuation for any reason (P = .002 for interaction)

        -- the most common AEs leading to permanent trial product discontinuation were gastrointestinal AEs in all baseline FI categories, on the order of a 5-10 fold increase

        --metabolism and nutrition disorders and cancers were uncommon (<1.5%)

Commentary:

-- frailty is a common health condition, typically associated with decreased physiological reserve and increased vulnerability to stressors in frail individuals, as well as a higher risk of death and further disability

    -- there are some concerns about the risks of adverse health outcomes in frail individuals because of potentially decreased treatment benefits, tolerability, polypharmacy, and self-management capacity; these issues might lead clinicians to be hesitant about starting new medications, and there is some suggestive medical literature that this hesitation may be especially so in obesity management

-- the SELECT trial of overweight patients with pre-existing cardiovascular disease but not diabetes did find impressive benefits of the GLP-1 receptor agonist semaglutide 2.4mg weekly over placebo: https://gmodestmedblogs.blogspot.com/2023/11/semaglutide-decreases-cardiovasc-events.html

    -- this current exploratory analysis of the SELECT trial assessed frail patients also finding that semaglutide was beneficial for many adverse clinical outcomes with no definitive evidence of heterogeneity by frailty status

        -- as noted by figure 1 above, participants' perceptions of the quality of their lives was not only improved with semaglutide but was even more impressively improved in those with higher levels of baseline frailty, and the aspects of the frailty assessment most significantly improved were for mobility, self-care, and conduct of usual activities.

        -- overall, serious adverse events were less common with semaglutide in all categories of frailty; the increase in permanent discontinuation of the semaglutide was higher but less so in those with higher levels of baselin frailty.

            -- in sum, overall adverse effects mostly favored semaglutide, reinforcing that its use had a favorable risk/benefit balance

-- one concern with some medications in frail individuals is that treatment benefits may attenuate and/or adverse effects might counterbalance potential benefits.

    -- a systematic review of 61 trials (DOI: 10.1007/s11606-024-08732-8) in older individuals found that the effects of several medication interventions (eg edoxaban, sacubitril/valsartan, prasugrel, and chemotherapy) varied by the degree of frailty, whereas other treatments (eg, antihypertensives, vaccinations, osteoporosis medications, and androgen medications) demonstrated consistent benefits across different frailty levels. Some non-pharmacological interventions had greater benefits in patients with higher levels of frailty (eg, chair yoga, functional walking, physical rehabilitation, and higher dose exercise program) and some with lower frailty levels (eg, intensive lifestyle intervention, psychosocial intervention), and some benefited all frailty levels (eg, resistance-type exercise training, moderate-intensive physical activity, walking and nutrition or walking). Specific combined interventions (eg, hospital-based disease management programs) demonstrated inconsistent effects across different frailty levels. in sum, there was significant variability in outcomes depending on the med or other intervention in those with frailty, so we should be careful about making global assumptions...

    -- and, as per this current SELECT study in frail individuals, long-term data revealed semaglutide's consistent effects across frailty levels decreasing major adverse cardiovascular events, and all-cause death plus all-cause hospitalization even in the frailest individuals; these benefits reflect improved key cardiovascular-kidney-metabolic risk factors (including body weight) and pretty low levels of treatment discontinuation

        -- patients should be warned that it is common to have some adverse effects (especially GI) on initiating a dose or dose escalation of GLP1s, but this typically gets better with subsequent injections; most (but not all) of my patients on GLP1s who have GI complications are willing to tolerate the initial GI events, continue with the injections, and do well soon thereafter.

-- of note in the above SELECT analysis, there was less heterogeneity of the results about frailty when FI was assessed as a continuous variable than as the 3 discrete buckets used [≤0.210 (not frail),  0.211-0.310 (more frail) and FI ≥0.31 (most frail)]. this is consistent with the obvious:  one patient with an FI of 0.210 vs another with FI of 0.211, for example, is likely not a clinically relevant differnece, yet analyzing their discrete buckets would place these 2 patients into very different categories that is clinically useless

    -- the mechanism of action of semaglutide on the diverse clinical events is not entirely clear. likely many are related to weight loss, though there are "pleiotropic effects", as with statins, leading for example to decreased pneumonia events found in this study, for example as noted in the diagram below:

This circos plot shows an atlas of associations between GLP-1RAs and 175 health outcomes across 12 outcome categories. From the outermost to the innermost rings: (1) 12 outcome categories, (2) outcome names with decreased risks (blue), increased risks (red), and nonsignificant associations (grey), (3) heatmap of risk magnitudes, (4) reduced risk magnitudes, (5) increased risk magnitudes, and (6) statistical significance (negative log-transformed P values; yellow = significant, gray = nonsignificant). The figure highlights the systemic effects of GLP-1s, revealing decreased risks for many health conditions and increased risks for several adverse outcomes, per https://pubmed.ncbi.nlm.nih.gov/39833406/  (see GLP1 myriad effects NatMed2025 in dropbox)

-- i would add a few things:

    -- my experience is that some individuals lose too much weight on GLP1s. this should clearly be monitored closely and may require cutting back on the dose or stopping the med completely

        -- however, cutting back the 2.4mg dose of semaglutide could conceivably not lead to the benefit found in the SELECT study (though this very likely would still help a lot)

        -- there is concern about weight loss by any means in skeletal muscle integrity:

            -- one important approach is to make sure that patients consume a sufficient amount of protein

            -- and there was a pretty impressive trial finding that skeletal muscle was actually improved with GLP1s: https://gmodestmedblogs.blogspot.com/2024/11/glp-1-agonists-improve-skeletal-muscle.html

-- since we are on the topic of GLP-1s, here are a couple of new articles:

    -- Orforglipron is a new oral, nonpeptide glucagon-like peptide 1 receptor agonist, where it was added to insulin glargine for the treatment of type 2 diabetes in this phase 3 multinational study (see diabetes orforglipron JAMA2026 in dropbox, or doi:10.1001/jama.2026.9512), finding:

        -- 546 participants with A1c of 8.5% were randomized to orforglipron daily or placebo, assessed at week 40: the HbA1c decreased 1.8 percentage points, with 5% decrease in weight, with only mild-to-moderate adverse GI events and no significant hypoglycemia (though there was some titration down of glargine dose on the order of 10 units from the baseline of about 36 units)

    -- Mazdutide 9mg is a once-weekly glucagon and GLP-1 receptor dual agonist in a Chinese study of adults with or without type 2 diabetes but with obesity defined as BMI >30 (see obesity wt loss mazdutide JAMA2026 in dropbox, or doi:10.1001/jama.2026.8142)

        -- 461 participants with diabetes, BMI 34. at week 60, the mean change in body weight from baseline was 16.65%; in those without diabetes there was a 19.6% decrease in body weight at week 60

 

Limitations:

-- the SELECT trial had patients with a mean age of 62. This relatively low age does limit generalizability to the patient group with the most frailty: the geriatric population

-- the width of the FI scale was small. the frailist group studied, the "more frail" group, encompassed all with an FI score of at least 0.311 in a scale that runs from 0-1.  the researchers did find that there was a continuous relationship between FI score and clinical results assessed. But, without further explanation, there was a graph buried in the Supplementary materials (their eFigure 3) having a y-axis of 4 of the major outcomes assessed and with the x-axis being "Frailty Index (0-100)", with the maximum being 45.  I assume that this Frailty Index was the same as the one promulgated in the actual article of 0-1 but just a couple of more exponents. if true, then the equivalent FI for the study would be a range of 0 to 0.45. And this suggests that the frailest individuals had less than ½ of the range of the FI. And that would suggest that there are lots of very frail persons who do not fit into that narrow range and are substantially frailer. And perhaps the results of this somewhat constricted study may not be generalizable to individuals with significantly more frailty

-- we do know, as per the diagram above, that GLP-1s have a multitude of benefits throughout the body, not just the weight loss and cardiovascular benefits (eg, as mentioned, decrease in pneumonia). was the benefit simply the decrease in "cardiovascular-kidney-metabolic" risk? we do know that some individuals with high BMIs do not have the "metabolic syndrome". it seems important to disaggregate this in terms of development of targeted future beneficial meds overall and for frail individuals 

   -- and people can be very frail from non-cardiovascular-kidney-metabolic risk factors: neurological disorders (ALS, congenital anomalies....), or accidents (falling from a 10-foot ladder), etc who are very different from those with the array of cardiovascular-kidney-metabolic conditions in the SELECT study that likely apply to the SELCT study individuals. would semaglutide help them?  seems like that would be a useful study to see if the effects of semaglutide on frail individuals is broader than the above study assessed.

so, a pretty clear relationship between semaglutide 2.4mg and many improved clinical outcomes related to cardiovascular-kidney-metabolic risk in patients with frailty

    -- it is unknown based on this study if there is more benefit from higher dosing or using more potent GLP1 type meds

    -- it seems from the above that we do not know anything about individuals who are frailer than the narrow limits of this study, limiting generalization to that less frail group or frail groups who do not have high cardiovascular-kidney-metabolic risk

    -- that all being said, GLP-1 receptor antagonists and the ones with multiple targets as well as GLP-1 receptors (tirzepatide, etc) are amazingly beneficial for oh-so-many things, as pictured in the circos graph above that extend beyond those associated with metabolic havoc

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

to get access to he dropbox: https://www.dropbox.com/scl/fo/pcf4vymtpo77f49wd2ljr/ANyJqB40CuBd-DMEr67XDoI?rlkey=5dsalmuhf8ekd0svsgvn93xm1&dl=0

or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them