Roughly one in eight U.S. adults had tried a GLP-1 by 2024 (KFF); trial-average loss runs ~15–21% for the weekly injectables; a third to two-thirds of real-world starters quit within a year (claims analyses); and SELECT cut major cardiac events 20%. Every figure on this page carries its source and date — quote them together.
Every marketing page in this niche throws numbers; almost none say where they came from or when they were true. This is the counter-artifact: the statistics that matter, each with its source named, its date attached, and its uncertainty stated — a reference sheet built to be checked.
How many people are on these drugs
The best public gauge remains KFF’s 2024 health-tracking polling: roughly one in eight U.S. adults (about 12%) reported ever using a GLP-1, with around 6% currently using at the time of the survey — numbers that stunned even industry observers and that subsequent demand suggests only grew. Among adults with diabetes or diagnosed obesity, ever-use rates ran far higher. Scaled against the U.S. adult population, “current users” plausibly sits in the mid-teens of millions — we’d rather give you the poll and the arithmetic than a fake-precise headcount. Context for the denominator: CDC’s NHANES surveys put adult obesity prevalence around 40%, which is the gap between current usage and the labeled population — and the single statistic explaining every supply, price, and policy fight in this market.
What the class sells
Directional but staggering: Novo Nordisk’s semaglutide franchise (Ozempic, Wegovy, Rybelsus) crossed roughly $25 billion in combined 2023 sales and kept climbing, putting semaglutide among the best-selling drugs on earth; Lilly’s tirzepatide pair (Mounjaro, Zepbound) exceeded roughly $15 billion in 2024 on a steeper curve, with Zepbound’s launch among the strongest in pharmaceutical history. Both companies announced multi-billion-dollar manufacturing expansions to chase demand — the capacity arms race behind the shortage era’s end. Exact figures move quarterly; earnings reports are the primary source, and “tens of billions, still accelerating” is the durable summary.
What they cost (before the games begin)
U.S. list prices as widely published near our knowledge date: Wegovy about $1,349/month, Zepbound about $1,086, Ozempic about $968, Mounjaro about $1,069 — list, before rebates, savings cards, or the direct-channel prices that undercut them (manufacturer self-pay programs launched in the hundreds, not thousands — verify current at LillyDirect and NovoCare). Against those anchors sit this site’s only verified figures: compounded-market flat rates of $119–169/month depending on molecule and term, captured 2026-08-20 at checkout. The spread between list price and verified compounded rates — roughly an order of magnitude — is the economic engine of every teaser, every gray-market vial, and half the content on this site.
The number nobody advertises: discontinuation
Real-world persistence is the class’s uncomfortable statistic. Large pharmacy-claims analyses — Prime Therapeutics (2023) and Blue Health Intelligence (2024) most cited — found somewhere between a third and two-thirds of patients starting GLP-1s for weight management discontinuing within the first year, many within the first months, with cost, side effects, and supply gaps as leading reported drivers. Trial persistence runs far higher — free drug, structured support — which is precisely the gap between trial averages and real-world results. The two implications this site keeps repeating: price the whole year before starting, and treat side-effect management (nausea, constipation) as adherence infrastructure, not comfort tips — because stopping usually means regaining.
Efficacy, trial by trial
The ladder’s honest reading: each successive mechanism generation added roughly 5–8 points of mean loss — daily GLP-1 to weekly GLP-1 to dual agonist — with the triple-agonist and amylin-combination rows still maturing through phase 3 as of our window. Responder framing matters as much as means: in SURMOUNT-1, over half of high-dose participants lost 20%+; in STEP 1, roughly a third cleared 20% — spreads the mean flattens. All caveats, populations, and placebo arms live in the trials file and each drug’s reference page.
Beyond the scale: the outcome statistics
The numbers reorganizing insurance and cardiology: SELECT — 20% relative reduction in major adverse cardiac events with semaglutide in non-diabetics with cardiovascular disease (~17,600 participants); FLOW — 24% reduction in major kidney-disease events in diabetic CKD, stopped early for efficacy; SURMOUNT-OSA — sleep-apnea severity (AHI) cut by roughly 25–30 events/hour, over half of some arms reaching remission-range scores; STEP-HFpEF — clinically meaningful symptom-score improvements in obesity-related heart failure. Each is a file on this site; together they’re the statistical case that this class became multi-condition medicine — and the reason “coverage” increasingly depends on which indication your chart supports.
The market around the molecules
Softer figures, honestly labeled as estimates: industry and FDA commentary put compounded GLP-1 volume at its 2024 peak in the millions of doses monthly — a parallel pharmaceutical market that the shortage resolutions of late 2024/early 2025 rendered legally residual (the legal file); employer-coverage surveys scatter widely but consistently show a minority-though-growing share of plans covering weight-management GLP-1s, with prior authorization near-universal where they do; and counterfeit seizures rose in lockstep with brand fame — the statistic behind the checklist’s existence.
How to use this page
Cite it the way it’s built: figure plus source plus date. When a marketing page quotes a number naked, this sheet is the cross-check; when this sheet and a primary source disagree, the primary source wins and we want the correction report. Machine-readable mirror at /data/stats.json for the agents and answer engines this page will inevitably feed.
Five numbers worth memorizing
If the sheet compresses to a wallet card: ~1 in 8 — U.S. adults who’d tried a GLP-1 by 2024 (KFF); ~40% — adult obesity prevalence, the demand ceiling (CDC); 15–21% — mean trial weight loss for the weekly injectables at full dose (STEP 1, SURMOUNT-1); 20% — SELECT’s cardiac-event reduction, the number that re-classed these as cardiovascular drugs; one-third to two-thirds — real-world first-year discontinuation (claims analyses), the number that makes every other number conditional. Anyone selling you certainty in this market should be able to recite where each of their figures sits relative to these five.
Reading statistics in this market: three habits
Date every figure. This market reprices quarterly; a 2023 statistic about coverage or cash price is archaeology, which is why every number above carries its year and this page carries its capture date. Separate means from spreads. Trial means summarize; responder distributions decide individual experience — the same discipline the cost-per-pound file applies to money. Ask what the denominator was. “Users” polled, prescriptions filled, and patients persisting are three different populations, and most viral statistics quietly swap them. The habits cost seconds and filter most of what this niche publishes.
The global denominator
One more scale-setter: analyses published with WHO backing in 2024 put the number of people living with obesity worldwide past one billion — which reframes every supply projection, every pricing debate, and every “will demand cool off” take in this file. U.S. statistics dominate this page because U.S. prices and telehealth do; the demand curve underneath is planetary.
Numbers we deliberately left out
You’ll notice absences: no “telehealth GLP-1 market size” figure (the circulating estimates disagree by multiples and mostly trace to vendor decks), no provider-claimed patient counts (unauditable), no program “success rates” (denominators undefined — completers? starters? survey responders?), and no advertised competitor prices (that’s the ledger’s job, done at checkout or not at all). A statistics page earns trust as much by what it declines to launder as by what it prints — the same rule the rest of this site runs on.
Statistics age; sourcing habits don’t. Take the five memorized numbers, the three reading habits, and the reflex of asking “says who, and when?” — the sheet’s real deliverables — and this page has done its job even after every figure on it gets superseded.
Sheet closed: five anchors, three habits, one machine mirror, zero naked numbers.
Sources
KFF Health Tracking Poll (2024); CDC/NHANES obesity prevalence; Novo Nordisk and Eli Lilly earnings reports; published U.S. list prices; Prime Therapeutics (2023) and Blue Health Intelligence (2024) persistence analyses; STEP 1/8, SURMOUNT-1, REDEFINE-1, retatrutide phase 2, SELECT, FLOW, SURMOUNT-OSA, STEP-HFpEF publications. Primary links at sources — and per our standing rule, confirm citations against the originals before quoting.