Written and fact-checked to our published editorial standards.Last updated 2026-08-15. Every clinical claim on this page is sourced to a named primary authority, listed at the foot of the page — see how we research, source, and correct our work.
The short answer
12% of US adults are currently taking a GLP-1 medication, and 18% have taken one at some point — for diabetes, weight loss, or another condition. US sales of the three leading brands topped an estimated $80 billion in 2025, trial weight loss ranges from roughly 13–23% depending on the drug, and real-world persistence is far lower than the trials might suggest: most people who start a GLP-1 for weight loss have stopped within a year.
Every figure below is dated and linked to its source — FDA prescribing labels, KFF polling, CDC survey data, peer-reviewed trials, and named market-research desks. Where sources disagree, we show both rather than pick the more convenient number. See what we deliberately left out for figures we could not verify.
Usage & prevalence
Current GLP-1 use peaks in middle age and skews toward people managing diabetes, heart disease, or their own weight, per a nationally polled KFF survey fielded in late 2025.
Current use by age
18–299%
30–4913%
50–6422%
65+9%
Medicare excludes weight-loss-only prescribing
Current or past use, by condition
Diagnosed with diabetes45%
Heart disease29%
Self-identified overweight/obese23%
Among current and past users, 56% say the drugs are difficult to afford, 27% of insured users still pay full cost out of pocket, 14% stopped because of cost, and 13% stopped because of side effects. Most (76%) got their prescription from a doctor; 17% from an online provider and 9% from a medical spa or aesthetic clinic — the clearest dated figure we found for telehealth’s share of the prescribing channel, though it measures self-reported source, not a claims-based share of all scripts written.
Read those four numbers together and cost stops being a footnote to this subject: it stops more people than side effects do, and it does so among people who are insured. That is the gap between what a drug is priced at and what a given person can reach, which is a different question from which drug works best — we take it separately in what each route to a prescription costs.
Diagnosed-diabetes population, by age (2024)
A separate, nationally representative CDC survey of adults with diagnosed diabetes found 26.5% used a GLP-1 injectable in 2024, ranging as follows:
18–3425.3%
35–4928.9%
50–6433.3%
65+20.8%
Clinical trial efficacy, by drug
Headline mean body-weight reduction from each drug’s pivotal trial. These are not head-to-head comparisons — trial populations, durations, and placebo responses differ — so read this as “what each drug showed in its own trial,” not a ranking. Retatrutide is not FDA-approved as of this writing; it is shown for context only, shaded to mark it as investigational.
ATTAIN-1 · N>3,100 · 72 weeks — first non-peptide oral GLP-1; FDA approved April 2026
Retatrutide 12mg28.3%
TRIUMPH-1 · primary endpoint — vs. 2.2% placebo — NOT FDA-approved; pipeline data only
Every bar above is a mean, which is the balance point of a spread and not a result anyone in the trial actually had. To see the spread instead — how many participants reached 5%, 10%, 15% and 20%, what the placebo arm did at each threshold, and what those thresholds weigh at a given starting weight — use the weight loss calculator.
Side effects: FDA-label incidence rates
Pulled directly from each drug’s current FDA prescribing label on DailyMed — the adverse-reaction rate reported in the pivotal placebo-controlled trial pool, not an estimate. See our side-effects guide for what to do about each one.
One number is almost always quoted without the other. Wegovy’s label reports 44% nausea; the same label reports 16% on placebo, in the same trial over the same weeks. The drug figure on its own cannot separate the medicine from the experience of being in a trial and being asked every week whether you felt sick. The gap can.
What the placebo arm reported
Every rate below is from the product’s own FDA label, next to the placebo arm of the same trial. The gap is the part that belongs to the drug.
These are four separate trials, not a head-to-head comparison, so the products cannot be ranked against each other on these figures — Mounjaro’s placebo arm reported more diarrhea (9%) than Ozempic’s treatment arm did (8.5–8.8%). Ranges are given where a product has several approved doses.
The full cross-product reference
Incidence in the pivotal trial pool for each drug’s approved indication, vs. its own placebo arm. Ranges reflect multiple approved doses.
Drug
Nausea
Diarrhea
Vomiting
Constipation
Wegovysemaglutide 2.4mg, weight management
44%16% placebo
30%16% placebo
24%6% placebo
24%11% placebo
Zepboundtirzepatide 5/10/15mg, weight management
25–29%8% placebo
19–23%8% placebo
8–13%2% placebo
11–17%5% placebo
Ozempicsemaglutide 0.5/1mg, type 2 diabetes
16–20%6.1% placebo
8.5–8.8%1.9% placebo
5–9.2%2.3% placebo
3.1–5%1.5% placebo
Mounjarotirzepatide 5/10/15mg, type 2 diabetes
12–18%4% placebo
12–17%9% placebo
5–9%2% placebo
6–7%1% placebo
Market & spending
2025 US sales, leading brands
Mounjaro39.1B
+67.1% year over year
Zepbound22.9B
+188.7% year over year
Wegovy18.6B
+29.2% year over year
Medicaid GLP-1 spending, gross ($ billions)
20191B
20249B
Medicaid gross spending on GLP-1s rose roughly ninefold from 2019 to 2024, alongside an eightfold increase in prescriptions. Medicare Part D gross spending reached $27.5 billion in 2024, a fivefold increase since 2019. Both figures are gross, before manufacturer rebates — actual net cost to public payers is materially lower.
Global market size: a genuine disagreement between forecasters
Named analyst desks disagree meaningfully on where this market is headed, which is worth showing rather than hiding behind a single confident number:
Morgan Stanley Research: ~$79B (2025) growing to ~$190B by 2035
Goldman Sachs Research: Revised down to ~$95B by 2030 (from a prior $130B estimate)
Persistence: how many people are still on treatment later
This is the least flattering number set on this page, and one most competitor content skips. A JAMA Network Open cohort (patients who started 2018–2023) found 64.8% of non-diabetic weight-loss patients and 46.5% of patients with type 2 diabetes had discontinued within one year.
Separately, Prime Therapeutics’ own book of business shows persistence improving sharply in more recent cohorts — from 33.2% at one year (2021 starters) to 60.9% (H1 2024 starters) — likely reflecting the end of the 2022–2024 supply shortages and better dose-management practice, not a change in the drugs themselves:
2021 cohort33.2%
H1 2024 cohort60.9%
Even so, Prime’s longer-run data shows only 15% still on treatment at two years and just 8% (roughly 1 in 12) at three years for obesity treatment. The JAMA and Prime figures come from different cohorts and time periods and are not directly comparable — we show both, dated, rather than pick one.
Insurance coverage
Employer coverage of GLP-1s for weight loss, by company size
200–999 workers16%
1,000–4,999 workers30%
5,000+ workers43%
up from 28% the prior year
As of KFF’s January 2026 count, 13 state Medicaid programmes cover GLP-1s for obesity specifically under fee-for-service — this count has moved both up and down over the past two years as states add and roll back coverage, so treat it as a snapshot, not a stable figure. Medicare Part D still cannot cover a GLP-1 prescribed solely for weight loss by federal statute; see our cost and coverage guide for the Medicare GLP-1 Bridge programme and how each pathway actually works.
The CDC’s national survey data covers GLP-1 injectable use among adults with diagnosed diabetes specifically — it does not represent the broader weight-loss-only population, where the KFF polling above is the better source.
By race and ethnicity
Hispanic31.3%
Black, non-Hispanic26.5%
White, non-Hispanic26.2%
Asian, non-Hispanic12.1%
By BMI category
Healthy weight16.7%
Obesity32.4%
What we deliberately left out
A few figures came up repeatedly in our research but failed our sourcing bar, so we left them out rather than publish an estimate:
A single “% of GLP-1 prescriptions written via telehealth” figure. We found no primary claims-based study framed exactly this way. The closest verifiable number is the 17% self-reported “online provider” figure cited above.
Compounded-market size and a “% of users who tried a compounded product” figure. Both traced only to secondary aggregation we could not confirm against a named primary report.
SURMOUNT-2’s exact efficacy figures. We could only confirm these via investor-relations reporting of the Lancet publication, not the paper itself, so we left SURMOUNT-2 off the chart above pending direct verification.
Generic market-research-firm CAGR figures (several circulate publicly) that disagree with each other by more than 2x and don’t disclose methodology — we cite only named analyst desks (Morgan Stanley, Goldman Sachs) instead.
Sources
Each source below supports specific statements on this page. We cite the strongest available authority and verify against the current version before publishing.
Global GLP-1 market projectionMorgan Stanley ResearchSupports: Global market-size projection: ~$79B (2025) growing to ~$190B by 2035.
Global GLP-1 market projectionGoldman Sachs ResearchSupports: Global market-size projection: Revised down to ~$95B by 2030 (from a prior $130B estimate).
Wegovy prescribing informationDailyMed, U.S. National Library of MedicineSupports: Wegovy adverse-reaction incidence table (semaglutide 2.4mg, weight management), revised 2026-06-18.
Zepbound prescribing informationDailyMed, U.S. National Library of MedicineSupports: Zepbound adverse-reaction incidence table (tirzepatide 5/10/15mg, weight management), revised 2026-04.
Ozempic prescribing informationDailyMed, U.S. National Library of MedicineSupports: Ozempic adverse-reaction incidence table (semaglutide 0.5/1mg, type 2 diabetes), revised 2026-05.
Mounjaro prescribing informationDailyMed, U.S. National Library of MedicineSupports: Mounjaro adverse-reaction incidence table (tirzepatide 5/10/15mg, type 2 diabetes), revised 2026-04-22.
This page is general education, not medical advice. Talk to your own healthcare provider about your situation before starting, stopping, or changing any medication.