The GLP-1 Revolution: How a Diabetes Drug Became the Fastest-Growing Drug Category in Medicare History

Medicare Part D spending on GLP-1 drugs grew from $2.3 billion in 2019 to roughly $27 billion by 2024. This post traces that explosion through CMS spending data, beneficiary counts, state-level adoption maps, and FDA adverse event reports – showing how a diabetes drug became the fastest-growing drug category in Medicare history.

In 2019, Medicare Part D spent $2.3 billion on GLP-1 diabetes drugs — primarily Trulicity. By 2024, that number had grown to roughly $27 billion, with nearly 4 million Medicare beneficiaries on one of these drugs. Ozempic alone costs Medicare more annually than all opioids combined. The explosion is driven almost entirely by volume — the cost per dose has stayed roughly flat — and it is still accelerating. Mounjaro (tirzepatide) is growing faster than Ozempic ever did, and Wegovy/Zepbound (the obesity-indication formulations) have barely begun to penetrate Medicare coverage. The FDA adverse event database has absorbed more tirzepatide reports in 3 years than semaglutide accumulated in 7, suggesting the pace of adoption is genuinely without precedent in recent pharmaceutical history.

GLP-1 stands for glucagon-like peptide-1, an incretin hormone produced in the gut after meals. Its job is to signal the pancreas to release insulin, slow gastric emptying so nutrients absorb more gradually, and tell the brain that you've eaten enough. In healthy individuals, GLP-1 surges after meals and dissipates quickly — its half-life is roughly two minutes. GLP-1 receptor agonists (GLP-1 RAs) are synthetic molecules designed to bind the same receptors but survive much longer in circulation, producing a sustained hormonal signal that controls blood sugar and suppresses appetite.

The first GLP-1 RA approved in the US was exenatide (Byetta) in 2005 — delivered twice daily by injection and requiring significant patient effort. The category remained a niche diabetes treatment for over a decade: effective but complicated, and competing against established oral options (metformin, sulfonylureas) and insulin. The breakthrough came from a simpler formulation.

Ozempic's arc is the story of the category. Semaglutide (Novo Nordisk) was approved by the FDA in December 2017 as a once-weekly subcutaneous injection for type 2 diabetes. Clinical trials showed not only strong blood sugar control but unexpected weight loss — 10-15% body weight reduction in some patients — and, crucially, reduced cardiovascular events in a high-risk population. Those cardiovascular outcomes data (from the SUSTAIN-6 trial) made Ozempic the first GLP-1 with a proven heart disease benefit, dramatically expanding the patient population that could justify its use.

The next pivotal moment was Ozempic's cultural emergence. By 2022, social media was saturated with accounts of dramatic weight loss from the drug — often from people taking it off-label without a diabetes diagnosis. The #Ozempic hashtag generated billions of views. Celebrities were rumored to be using it. Supply chains strained. The drug had escaped its original clinical context and become a cultural phenomenon.

Then tirzepatide arrived. Mounjaro (tirzepatide, Eli Lilly) was approved in May 2022 for type 2 diabetes. It is a dual agonist — activating both the GLP-1 receptor and the GIP (glucose-dependent insulinotropic polypeptide) receptor. Clinical trial data suggested 20-22% mean body weight reduction, substantially above semaglutide's ceiling. Its SURPASS trial program showed better A1c and weight reduction than Ozempic in head-to-head comparisons. The drug entered Medicare Part D with 54,000 beneficiaries in 2022. By 2024, it had 895,000 — a 16× increase in two years.

Approval timelines for the major drugs in the class:

The T2D and Obesity brands are often the same molecule at different doses and with different regulatory pathways. This distinction matters enormously for insurance coverage.

This chart shows gross Medicare Part D spending on GLP-1 drugs from 2019 through 2024.

The shape is unmistakable. In 2019, the entire GLP-1 category cost Medicare $2.3 billion — dominated by Trulicity (dulaglutide), which was then the market leader. By 2023, spending had reached $22 billion, with Ozempic accounting for $9.2 billion alone. The 2024 full-year figures push the total to approximately $27 billion, with Ozempic ($13.0B), Mounjaro ($6.3B), and Trulicity ($5.5B) as the top three.

To put $27 billion in context:

• It exceeds the total Medicare Part D spending on all cancer drugs in 2019
• It is more than the entire federal budget for the Environmental Protection Agency (~$9B) or the National Science Foundation (~$9B)
• Medicare spent more on Ozempic alone in 2024 than on Humira (adalimumab, long the most expensive drug in the US) at its peak

The growth is not a plateau. The 2025 Q1-Q2 partial year shows Ozempic at $7.6B and Mounjaro at $5.7B in just the first half — both tracking above their 2024 full-year rates. The trajectory has not bent.

What's driving the spending? The answer is almost entirely volume, not price. Average spend per beneficiary on Ozempic has remained roughly $6,000-$7,000 per year (gross, before rebates). The extraordinary growth comes from the number of people on the drug — which increased from roughly 70,000 Medicare beneficiaries in 2019 to nearly 2 million in 2024, a 28× increase.

The beneficiary chart shows just how fast adoption has been. Trulicity had the most Medicare users in 2019-2021 — its patient base was built gradually over five years of diabetes treatment. Ozempic overtook it in beneficiary count by 2022 despite launching years later, driven by higher demand and more aggressive prescribing. Mounjaro's trajectory is steeper than either: launched mid-2022, it reached 895,000 beneficiaries in 2024 with its first full calendar year barely completed.

Wegovy's beneficiary count remains small in Medicare — only 53,000 in 2024 — for a policy reason. Medicare Part D is prohibited by statute from covering drugs "used for weight loss." Because Wegovy is FDA-approved for chronic weight management (not diabetes), Medicare couldn't cover it until the Inflation Reduction Act and subsequent regulatory action clarified that obesity itself is a disease condition. The 2024 PREVENT trial data showing Ozempic reduced major cardiovascular events by 20% in people with obesity (not just diabetes) created a new pathway: Medicare now covers semaglutide for patients with obesity and established cardiovascular disease. But coverage for obesity alone without a comorbidity remains unresolved. Zepbound faces the same situation — only 245 Medicare beneficiaries in all of 2024.

This gap between diabetes-indication and obesity-indication coverage represents the policy frontier. Tens of millions of Americans with obesity could benefit from these drugs. If Medicare were to cover GLP-1s broadly for obesity, estimates of the budget impact range from $35 billion to $145 billion per year.

The market share chart reveals the competitive dynamics within the GLP-1 class. In 2019, Trulicity commanded roughly 98% of GLP-1 spending by virtue of having no real competition. Ozempic began taking share in 2020, and by 2022 had matched Trulicity's spending. Mounjaro's entry in 2022 and rapid growth through 2024 has compressed Trulicity's share from near-total to less than 20%.

This consolidation matters for pricing. Trulicity lost patent exclusivity in 2023 — but generic dulaglutide has been slow to enter the market in meaningful volume, partly because injectable biologics face higher barriers to generic competition than oral drugs. Meanwhile Ozempic and Mounjaro remain fully patent-protected, with estimated expirations not until the late 2020s.

The oral formulation battle adds another dimension. Rybelsus (oral semaglutide, once-daily pill) has carved out a consistent 5-8% of GLP-1 spending. It has lower efficacy than the injectable form but appeals to needle-averse patients and has enabled prescriptions in patient populations that rejected injections. Oral tirzepatide is in development, which would substantially expand access to that class.

One common narrative about GLP-1 growth is that pharma companies have been raising prices aggressively. The per-beneficiary data complicates that picture.

Average spending per beneficiary has remained remarkably stable. Ozempic has held roughly $6,000-$7,000 per beneficiary per year throughout 2020-2024. Mounjaro launched higher (~$8,500 in 2022, reflecting higher doses and higher list price) but has moderated as patient volumes grew. Rybelsus runs substantially cheaper (~$4,500) as an oral daily pill.

This stability in per-patient cost means that list price increases have been modest relative to the volume growth. The Inflation Reduction Act (IRA) capped annual Medicare Part D out-of-pocket costs and restructured plan design to require drug manufacturers to provide rebates; this has kept effective per-beneficiary costs roughly flat even as gross spending has exploded with volume.

The gross figures ($12,000/year for Ozempic is the US list price for cash-paying patients without insurance) are not what Medicare pays. Medicare negotiates through Part D plans and receives manufacturer rebates that can reduce net cost by 40-60%. The actual Medicare expenditure per Ozempic beneficiary is likely closer to $3,000-$4,000 per year — still substantial, but very different from the list price.

The IRA's drug price negotiation provisions specifically target drugs with high Medicare spending, no generic competition, and long patent life. Ozempic and Mounjaro are prime candidates for future negotiation rounds. The first 10 negotiated drugs (from 2023) saved an estimated $6 billion. The potential savings from negotiating GLP-1 prices would dwarf that.

The geographic distribution of GLP-1 beneficiaries is heavily driven by state population — California, Texas, Florida, and New York dominate in raw beneficiary count. But the interesting variation is in the rate of adoption relative to Medicare enrollment.

Several patterns emerge:

• Southern states have high absolute volumes, consistent with higher rates of type 2 diabetes (the South has the nation's highest diabetes prevalence, particularly in the "Diabetes Belt" stretching from the Carolinas through the Gulf states).
• Western states show moderate-to-high GLP-1 use relative to their Medicare populations, partly reflecting the density of endocrinology practices and more aggressive prescribing culture in certain urban centers.
• Rural states often show high rates of diabetes but lower GLP-1 adoption — a gap driven by both provider access (fewer endocrinologists, more reliance on primary care for diabetes management) and coverage barriers (high deductibles, prior authorization burdens that deter prescribing).

The gap between diabetes prevalence and GLP-1 prescribing is one of the less-discussed aspects of the story. Roughly 30 million Americans have type 2 diabetes. About 2 million Medicare beneficiaries are on semaglutide. Even assuming all are diabetic and all are elderly, only a fraction of the eligible population is receiving the drug. The headroom for further volume growth — even before the obesity indication — is substantial.

No drug category grows this fast without accumulating an adverse event record. The FDA's Adverse Event Reporting System (FAERS) captures voluntary reports from patients, physicians, and manufacturers — not clinical trial data, but real-world post-market surveillance.

The heatmap compresses the adoption story into a single view. Each cell shows how many Medicare beneficiaries were on each drug in each year, and what that cost. The pale cells are near-zero adoption (Mounjaro before 2022, Wegovy throughout). The dark cells show the concentration of use — Ozempic in 2024 stands alone as a near-$13B program with nearly 2 million beneficiaries.

Trulicity tells a different story: a large, stable patient base that grew slowly and is now beginning to shrink as Mounjaro and Ozempic have displaced it for new prescriptions. The transition from an older drug (Trulicity) to a newer, more effective one (Ozempic/Mounjaro) is how pharmaceutical markets normally work — but the speed and scale are unusual.

The policy questions around GLP-1s are not abstract budget debates — they determine who can access drugs that could potentially reduce heart attacks, strokes, kidney failure, and premature death at a population scale.

The GLP-1 story is not primarily a story about diabetes drugs. It is a story about what happens when a drug unexpectedly becomes effective for a condition (obesity) that affects roughly 40% of the American population, has limited existing pharmaceutical treatments, and for which payers have historically refused to cover medications.

Several features of this story are genuinely new in pharmaceutical history:

Efficacy that created demand. Most blockbuster drugs achieve high sales through marketing and insurance coverage. GLP-1s created their own demand — social media, celebrity culture, and visible results among early adopters drove patients to seek prescriptions in ways that outpaced pharma marketing budgets. The Ozempic viral moment was not manufactured by Novo Nordisk.

CV outcomes data changing coverage. The expansion of GLP-1 coverage from "diabetes management" to "cardiovascular risk reduction" to (partially) "obesity" follows clinical trial results, not lobbying. The SUSTAIN-6, LEADER, SELECT, and SURPASS-CVOT trials produced mortality reduction data that made it progressively harder for payers to deny coverage on efficacy grounds.

The manufacturing constraint. This class of drugs requires biological manufacturing processes (unlike small-molecule pills) and complex formulation. Both Novo Nordisk and Eli Lilly have each spent billions of dollars on manufacturing capacity expansion. The supply chain is still constrained. This creates an unusual situation where a drug is simultaneously in shortage and generating record revenue — not because of pricing strategy, but because demand outstripped manufacturing capacity at any price.

The obesity treatment paradigm shift. For thirty years, the medical consensus on obesity treatment was behavioral (diet, exercise, lifestyle modification). The failure rate of behavioral intervention alone is high — around 80-90% of patients regain lost weight within 5 years. The GLP-1 clinical data has forced a re-evaluation. If 20% mean weight loss can be sustained pharmacologically, with mortality reduction data to match, the treatment paradigm shifts from "treat the behaviors" to "treat the underlying hormonal dysregulation." This reframing has wide-ranging implications for how obesity is understood, coded, covered, and treated.

Medicare Part D annual spending (2019-2023): CMS Medicare Part D Spending by Drug (dataset ID: 7e0b4365-fd63-4a29-8f5e-e0ac9f66a81b). Annual data with per-year breakdowns for spending, claims, beneficiaries, and average per-unit/per-claim/per-beneficiary costs. "Overall" manufacturer rows used for aggregate drug-level analysis.

Medicare Part D quarterly spending (2024-2025): CMS Medicare Quarterly Part D Spending by Drug (dataset ID: 4ff7c618-4e40-483a-b390-c8a58c94fa15). Provides 2024 Q1-Q4 annual totals and 2025 Q1-Q2 partial year data. "Overall" manufacturer rows used.

State-level prescribing: CMS Medicare Part D Prescribers by Geography and Drug (dataset ID: c8ea3f8e-3a09-4fea-86f2-8902fb4b0920). State-level beneficiary counts aggregated across Ozempic, Mounjaro, Trulicity, and Rybelsus.

FDA adverse events: OpenFDA Drug Adverse Event API (FAERS). Annual report counts by generic drug name (patient.drug.openfda.generic_name). Reaction profiles normalized per 1,000 total reports to enable cross-drug comparison. Administrative/process terms excluded from reaction comparison.

• Spending figures are gross, before manufacturer rebates. Medicare Part D plans negotiate confidential rebates with drug manufacturers; actual Medicare expenditure is lower than gross spending, typically by 40-60% for brand-name drugs. The true net Medicare cost is not publicly reported.
• 2025 data is partial: Quarterly dataset includes 2025 Q1-Q2 (approximately January-June 2025). Full-year 2025 is not yet available.
• FAERS reports are not incidence data: The FAERS database captures voluntary adverse event reports from patients, physicians, and manufacturers. High report counts reflect high usage and high media attention, not necessarily higher risk relative to other drugs. The FDA uses FAERS for signal detection, not causality determination.