What percentage of US adults will claim to be currently taking GLP-1 drugs for weight loss in 2030?

TL;DR

My forecast is 20% of U.S. adults reporting current GLP-1 use for weight loss in Gallup’s qualifying 2030 release. The central 80% predictive interval is 12%–30%. The forecast counts current users, not everyone who has tried these drugs, and is conditional on a comparable survey receiving a numerical resolution.

Context

The target is Gallup’s published percentage of all U.S. adults. Gallup first asks about ever taking the listed weight-loss medications, then asks affirmative respondents about current use; the reported population percentage does not use ever-users as its denominator (Gallup questionnaire). I forecast the first qualifying release, not a year-end prevalence or a prescription count.

The information cutoff is October 19, 2026. Historical observations anchor the forecast but do not resolve this future question. The numerical distribution is conditional on publication of a comparable measurement; survey discontinuation and annulment are not encoded as zero use.

Evidence

The full revised Gallup headline history is 3% current use and 6% ever-use in 2024; 8% and 12% in 2025; and 11% and 15% in 2026. These are percentages of all adults, from the article published July 7, 2026, and updated September 8 after revised weighting (Gallup revised series). The latest wave surveyed 5,065 adults online through the probability-based Gallup Panel from May 28 through June 5, 2026; a subsequent analysis reused that wave rather than adding another observation (Gallup follow-up).

This is a short, imperfectly comparable history. The original survey covered March 4–9, 2024, with 5,577 respondents, and asked about injections using semaglutide and liraglutide as examples (original release, May 21, 2024). The newer wording covers medications and also names tirzepatide (updated questionnaire). Three rounded annual observations establish rapid adoption. They do not identify a reliable saturation ceiling or prove that growth will keep slowing.

An independent survey supports the broad scale of adoption without matching Gallup’s definition. KFF’s complete current-use history is 6% in its April 23–May 1, 2024 survey, with 1,479 adults; 12% in October 27–November 2, 2025, with 1,350 adults; and 12% in February 24–March 2, 2026, with 1,343 adults. KFF includes use for diabetes and other conditions, not just weight loss, so I do not average these levels into Gallup’s series (2024 methods, 2025 release, March 2026 topline). The 2025 survey also found that 22% of noncurrent users were interested in taking a GLP-1 for weight loss. That is a recruitment pool, not a conversion rate (KFF demand findings).

Current use depends on both recruitment and continued treatment. A March 2026 claims study followed 33,607 commercially insured adults without diabetes who began Wegovy or Zepbound between January 1, 2021, and June 30, 2024. Its full one-year persistence history was 33.2% for 2021 initiators, N=889; 34.1% for 2022, N=2,480; 40.2% for 2023, N=14,665; and 60.9% for January–June 2024, N=15,573. Switching was allowed, and persistence used a 60-day treatment-gap rule. Follow-up extended through June 2025, so these are not outcomes for people starting in 2026 (Marshall et al., JMCP). This is strong evidence against freezing early, poor persistence into a forecast of future use. It is not evidence of uninterrupted treatment through 2030.

Stopping is not necessarily permanent. A separate study published January 31, 2025, examined 125,474 initiators from January 2018 through December 2023. Among the 48,950 without diabetes, estimated one-year discontinuation was 64.8%; among eligible discontinuers without diabetes, estimated reinitiation within one year was 36.3%. The restart analysis was a selected subset, not all initiators (Rodriguez et al., JAMA Network Open). I therefore model starts, stops and restarts—not cumulative adoption multiplied by a permanent dropout rate.

Oral treatment is already part of the market. Wegovy pills reached U.S. distribution in January 2026, and FDA approved Foundayo/orforglipron on April 1, 2026, both before the latest Gallup fieldwork (Novo launch-period data, FDA approval letter). Further penetration supports growth, but the entire launch effect cannot be credited again after the baseline. Affordability remains a constraint: the manufacturer’s price page lists oral Wegovy at $149 per month for the starter dose, versus $299 for the 9 mg and 25 mg doses. These are dose-specific nominal cash prices, not universal insured copays (current manufacturer offer).

Supply and new products support the upper scenarios. On September 21, 2026, Novo announced an ambition to expand oral-obesity capacity tenfold by 2030. That is a global capacity ambition, not a count of future U.S. users (Novo strategy announcement). Lilly’s August 5 results said its retatrutide clinical package supported a planned U.S. submission in the first quarter of 2027. That is a development milestone, not an approval (Lilly results). I read these developments as making continued expansion feasible, without assuming every new product creates a new user rather than a switcher.

Coverage is expanding unevenly. CMS launched Medicare GLP-1 Bridge on July 1, 2026, with a $50 monthly copay for eligible beneficiaries, but the program currently ends December 31, 2027 (CMS launch release). KFF’s June 29 analysis of 2023 claims estimated that 3.8 million beneficiaries met all eligibility criteria. The program is much narrower than universal Medicare obesity-drug coverage (KFF eligibility analysis). Meanwhile, Mercer’s June 11 release reported that 6% of large employers were dropping weight-loss GLP-1 coverage in 2026, with another 5% planning or considering doing so in 2027. Its April 15–May 8 survey covered 604 organizations, including 481 with at least 500 employees; those percentages describe employers, not adults losing coverage (Mercer survey).

My numerical checks put moderated growth near one adult in five. Continuing the latest Gallup increase of three percentage points annually gives 23% after four years. Reducing each successive annual increment by 10% gives about 20%. Extending ever-use by three points annually produces 27% ever-use; holding the current/ever fraction at its latest value gives 19.8% current use. These are my calculations from the revised Gallup history, not fitted estimates of national transition rates. The current/ever fraction is a cross-sectional ratio, not individual retention.

I turn those checks and the access evidence into the following judgmental mixture. Locations and standard deviations are percentage points of adults; each component is a normal distribution truncated to the physical 0%–100% range.

These weights are forecasting judgments, not historical frequencies. Within-scenario spread includes adoption uncertainty, survey timing, sampling error and measurement drift; sampling error is not added again afterward. The resulting rounded mean and median are both 20%, with a central 90% interval of 9%–33%. The model assigns 9% probability to a published result at or below 11%, and 10% to a result at or above 30%. Today’s current-use level is not a floor.

What's non-obvious

One apparent contradiction in the historical evidence is largely a current-versus-ever problem. Gallup’s older obesity article described 12.4% as “taking” these drugs (October 28, 2025 article). But the chart preserved in Goodbody’s report explicitly asks whether respondents had ever taken them, showing 5.8% in early 2024 and 12.4% in 2025 (preserved chart, printed page 45). I exclude that figure from the current-use trend. It is not evidence that current use subsequently collapsed, or that weighting alone explains a huge revision.

“New to the drug class” also need not mean a new person answering yes to Gallup. Novo’s launch analysis used a twelve-month prescription lookback and excluded compounded products when identifying prior GLP-1 treatment (company data definitions). I infer that its apparent new-user group can include restarters and people switching from compounded treatment. Pills expand access, but prescription growth cannot be translated one-for-one into additional survey users.

Uncertainties

The largest gaps are durable coverage after the temporary Medicare program, long-term persistence on oral products, and nationally representative stop–restart transitions. Existing persistence studies cover selected insured or clinical populations, not a matched sample of all adults (CMS program terms, claims-study population). A longitudinal survey linked to dispensing data—including cash-pay and compounded treatment—would sharpen the forecast. I also could not verify a separate sample size and exact field window for the retrospective 2025 current-use point in the newer Gallup article (Gallup series documentation).

Reporting precision is another modeled uncertainty. I assign 85% probability to whole-percentage reporting and 15% to one-decimal reporting, reflecting the different formats in the newer current-use release and older obesity release. The code integrates over rounding intervals before assigning probability to the requested buckets. It places exactly zero probability below zero, puts a reported zero in the first nonnegative bucket, and retains outcomes above 40% rather than imposing an artificial ceiling.