Newsletter · 2026-08-17 · Ashutosh Agarwal

Employers Retreat on GLP-1s as Lilly's Pill Advances - The Obesity-Drug Pipeline - Week of August 17, 2026

Obesity-drug podcast intelligence for the week of August 10 to 17, 2026. Starbucks is ending weight-loss coverage in October as GLP-1s reach 11.4% of corporate drug claims, while Lilly's oral orforglipron cleared a Phase 3 efficacy bar in The Lancet and took its first European approval.

The Obesity-Drug Pipeline

Week of August 17, 2026: Employers Retreat on GLP-1s as Lilly's Pill Advances

TL;DR (15 seconds)

• The coverage pendulum swung back toward the bears this week. Starbucks is ending insurance coverage for GLP-1s prescribed for weight loss starting in October (it keeps them for diabetes), and the underlying math is stark: GLP-1s now eat about 11.4% of corporate employer prescription-drug claims, up from 6.9% in 2023, yet only 36% of employers cover them for both diabetes and weight loss. Last week a big PBM (CVS) re-embraced Lilly; this week the employers who actually foot the bill are pulling back. On The Pen GLP-1 News
• Lilly's oral pill got real clinical validation and its first European nod, a genuine advance for the "pill era." New Phase 3 data (the ACHIEVE-2 trial, published in The Lancet) showed once-daily oral orforglipron beat a common diabetes pill on blood sugar and delivered 6–9% weight loss with no empty-stomach restriction; separately, Lilly's oral (branded Foundeo) took its first European approval in the UK, with weekly prescriptions at a new high of about 29,000. DOC Updates · Telltales
• The "200 deaths" headline hit the tape, and the real number, in context, is a bear-scare rather than a bear case. UK regulators logged 216 fatal reports across the GLP-1 class (120 for tirzepatide, 59 for semaglutide, 37 for liraglutide), but against millions of users and 158,000+ total adverse-event reports, all "suspected association, not causation." The genuinely new safety item: UK regulators strengthened acute-pancreatitis warnings. On The Pen GLP-1 News

Everything below comes from a podcast that aired in the last seven days (August 10–17, 2026). The center of gravity this week sat with the clinical and employer-benefits shows rather than the sell-side desks, which pushed the conversation away from "who's winning the P&L" and toward "who's paying for the drugs and how safe are they," the exact axis the bear case runs on.

What's new

1. Employer coverage is backsliding, and Starbucks is the name that makes it concrete. The loudest, most thesis-relevant story this week was payers pulling back, specifically the employers who ultimately pay for most commercial scripts. On On The Pen GLP-1 News (August 12), GLP-1 news commentator Dave Knapp laid out the split screen: Bank of America is keeping coverage despite a "$250 million a year" bill because it "finds that it is worth covering GLP-1s," while Starbucks is ending insurance coverage for GLP-1 medications when they're prescribed specifically for weight loss beginning in October. The company will keep covering them for diabetes and other qualifying conditions, but drops them for obesity.

Why this matters more than one coffee chain's benefits memo: it's a pattern, and the numbers underneath it are moving fast. Per the episode, GLP-1 drugs now account for about 11.4% of corporate employer prescription-drug claims, up from 6.9% in just 2023, and only 36% of employers cover GLP-1s for both diabetes and weight loss. Knapp's framing of the tension is the one every PM should sit with: obesity is a chronic disease, so employers "aren't necessarily looking at the next six months of treatment. They're potentially looking at thousands of workers requiring years of therapy." He also floated the contrarian silver lining: "if employers drop coverage, that forces the drug companies to... look at their prices." Translation for the book: this is the net-price-erosion bear case arriving through the employer door even as the PBM door (CVS, last week) swung open.

"This is just a continuing trend that we're seeing of employers backsliding, dropping coverage." (On The Pen GLP-1 News)

2. There's now a whole business model built on the premise that the current payer plumbing is broken. Two separate podcasts this week featured Jay Bregman, founder and CEO of Andel, a direct-to-employer platform that buys brand-name GLP-1s straight from manufacturers and skips the PBM. On Off the Chart: A Business of Medicine Podcast (August 13), Bregman claimed that less than 2% of people in any given plan actually access GLP-1s because of "massive utilization management and prior authorizations," and said GLP-1s have "broken the traditional model." His pitch: Andel launched in March with Zepbound, acquiring brand drugs directly from manufacturers with no prior authorizations and no formulary.

On GLP-1 Hub (August 13), Bregman put a finer point on the employer's dilemma: coverage becomes unaffordable when "10-20-30% of their population" demand $600–700/month medications, and insurers are dropping coverage when patients hit BMI thresholds despite FDA approval. Treat this as an operator's book-talk, since Bregman is selling a workaround and his "the system is broken" framing is self-interested. But the datapoint that keeps recurring across shows, a $600–700/month list-adjacent cost meeting near-universal demand, is the arithmetic behind every employer that's now flinching.

3. Lilly's oral pill just cleared a real efficacy bar, and got its first European approval. This is the quarter's most under-covered good-news item for Lilly. On DOC Updates (August 10), a monthly diabetes and obesity clinical-education show hosted by two endocrinologists presenting the data, the hosts walked through the ACHIEVE-2 trial, published in The Lancet, testing Lilly's orforglipron (an oral, non-peptide GLP-1) against dapagliflozin (a common oral diabetes drug) in over 960 type-2 diabetics inadequately controlled on metformin. At 40 weeks, all three orforglipron doses (3 mg, 12 mg, 36 mg) were non-inferior on blood-sugar control, cutting A1C by -1.2%, -1.5%, and -1.56% versus -0.81% for dapagliflozin (p<0.0001), and the two higher doses delivered 6–9% weight loss. The clinicians' key commercial tell: unlike Novo's oral semaglutide, orforglipron carries no empty-stomach / no-water restriction, "the ability to harness some of the powers of a GLP-1 without having to do an injection and without forsaking quite as much of the weight loss." One of them called the oral-with-real-weight-loss combination "a little bit of the... holy grail we've been kind of waiting for."

Separately, on Telltales (August 16, a markets recap podcast produced with AI-generated voices), the host noted Lilly's oral pill "took its first European approval in the UK, with weekly prescriptions at a new high of about 29,000." Put the two together and the read is: the oral franchise that missed on launch sales last quarter is now clearing clinical and regulatory hurdles, so the sales ramp may be following the data rather than leading it.

4. Lilly's other headlines this week were a tell about a company with more cash than pipeline. Same Telltales recap put fresh numbers and two strategic moves on Lilly's blockbuster Q2 (reported earlier this month): ~$23 billion of revenue, up 48%, beating consensus by 11%, with adjusted earnings of $8.38 versus $6.58 expected, and full-year EPS guidance raised nearly $3 at the midpoint to $35.50–$36.50. Two moves worth flagging for the competitive map:

• Lilly sued six companies over black-market sales of retatrutide, a drug it hasn't even launched yet. The host's wry read is the right one: "Suing counterfeiters of a product you cannot buy is its own kind of demand data." It also reinforces last month's biologic-classification fight (the Sept 24 Chicago appeal), where the whole game is walling off copycats before launch.
• Lilly agreed to buy three vaccine biotechs, Curevo, Limitech, and one more, for up to $3.8 billion. The host's blunt interpretation: "a company spending obesity money on vaccines, which is what you do when you have more cash than pipeline." For the memo, Lilly is pegged at 55 times trailing free cash flow on $21 billion of trailing FCF, a valuation that only works if the obesity annuity keeps compounding.

5. The "200 GLP-1 deaths" scare made the rounds. Here's the number in context. Because safety headlines move sentiment (and sometimes prescriptions), it's worth getting this one exactly right. On On The Pen, Knapp broke down the UK Medicines and Healthcare products Regulatory Agency (MHRA) "Yellow Card" data behind the viral headline:

• Tirzepatide (Mounjaro): 106,189 suspected adverse reactions, 120 fatal reports
• Semaglutide (Wegovy/Ozempic/Rybelsus): 48,089 suspected adverse reactions, 59 fatal reports
• Liraglutide (Saxenda/Victoza): 4,491 suspected adverse reactions, 37 fatal reports
• Total: 216 fatal reports, the source of the "200+" number

The crucial context Knapp stressed: these are reports of a "suspected association, but not a causation," drawn against a user base "in the millions," in a population (diabetes and obesity) with an already-high baseline mortality. The genuinely new regulatory item is narrower and more real: back in January the MHRA strengthened warnings around acute pancreatitis after roughly 1,300 reports of pancreatitis linked to GLP-1 / GLP-GIP drugs between 2007 and 2025 (including 19 fatal), and is investigating whether genetics make some patients more susceptible. That's textbook pharmacovigilance, a signal being worked rather than a class-wide indictment. Bottom line for the book: no thesis-breaking safety development, but the "deaths" narrative is now ammunition in the payer-pushback debate.

The debate

The bull case (steel-manned): The leader kept executing. Lilly's oral pill just proved it can lower blood sugar better than a standard diabetes drug and drop 6–9% of body weight in a pill, clearing the efficacy doubt that dogged the launch, and it landed its first European approval with scripts at a new high. The demand is so far ahead of supply that Lilly is suing counterfeiters of a drug it hasn't launched. And the safety "bombshell" this week, on inspection, was 216 suspected-association reports against millions of users, the kind of number that looks scary in a headline and benign in a denominator. Penetration is still early: one investor podcast this week floated 60–65% eventual adoption projections. The annuity is intact and broadening.

The bear case (steel-manned): Follow the payer, not the press release. The people who actually write the checks, employers, are backing away: Starbucks is dropping weight-loss coverage in October, GLP-1s have jumped to 11.4% of corporate drug claims in barely two years, and only a third of employers cover both indications. When a category's cost curve goes vertical, coverage gets rationed, and rationed coverage plus $600–700/month economics is exactly the net-price and volume squeeze the Street under-models. Add a safety narrative that, accurate or not, gives benefits managers cover to say no, and a compounding and gray-market channel that still caps brand pricing. The bull's "penetration is early" is the bear's "the payer system will never let penetration get there at these prices."

My read (a framing, not a call): Last week the payer story flipped bullish on the CVS re-embrace; this week it flipped right back on Starbucks and the employer data. That's not noise, it's the actual structure of this market. PBMs and employers are pulling in opposite directions: a PBM that also insures lives (CVS) wants healthy members and will cover the drugs; a self-insured employer just sees a claims line going vertical and rations. The single number that resolves it isn't a trial result, it's net price per script over time, and it's precisely the datapoint the podcasts still aren't giving us cleanly. Until we get it, treat every "coverage expands / coverage retreats" headline as one more oscillation around an unresolved price fight, and note that Lilly's oral, cash-pay, and biologic moves are all attempts to route around that fight rather than win it head-on.

Stocks in play

Read-throughs

• Fast-followers (AMGN, VKTX, Roche): the mechanism debate is still unresolved. The one clinical thread that touches the mechanism question came on Diabetes Care "On Air" (August 10), where clinical researcher Dr. Michael Nauck noted tirzepatide's dual GIP/GLP-1 action reduces A1C and weight more than selective GLP-1s like semaglutide, but that the "paradox," where both GIP agonists (tirzepatide) and GIP antagonists (Amgen's MariTide) seem to aid weight loss, remains mechanistically unresolved. That unresolved paradox is exactly Amgen's binary risk.
• Label expansion: HFpEF got a strong clinical airing. On Parallax by Ankur Kalra (August 10), cardiologist Dr. Barry Borlaug said semaglutide in the STEP-HFpEF trials showed "unprecedented" quality-of-life improvements and substantially fewer worsening-heart-failure events, while tirzepatide in the SUMMIT trial cut worsening heart failure or cardiovascular death by 38–40% as its primary endpoint, and he now recommends prescribing one or the other to appropriate HFpEF patients. This is the cardiologist community operationalizing the CVD and heart-failure label expansions the bulls count on to broaden the payable population beyond "weight loss."
• Pen and auto-injector suppliers (Ypsomed, Gerresheimer, Phillips Medisize): the structural drift keeps working against them. This week's marquee product news was an oral pill clearing efficacy and European approval. Every incremental win for the pill is an incremental headwind for injection hardware.
• Insurers, PBMs and employers (CVS, CI, UNH): active, and net-negative in tone. After last week's CVS re-embrace, the tone flipped: employers dropping weight-loss coverage (Starbucks), the 11.4%-of-claims cost pressure, and Bregman's PBM-bypass pitch all point to friction. Cigna's Express Scripts and UNH's Optum Rx were not named this week; the story is being carried by employers and a start-up middleman rather than the big three PBMs directly.
• Compounding and telehealth gray market: still wide open. On Fat Science (August 10), Dr. Emily Cooper walked through a JAMA "secret shopper" study (July 2026) in which a researcher tested 49 websites selling GLP-1s online and got prescriptions from 45 of them, roughly two-thirds with no clinician contact, most on a simple questionnaire, often with "leading questions" pushing add-ons like B12 or carnitine. This is the same class of study last week's issue flagged; the persistence of an easy-access online channel is exactly what keeps a ceiling on brand pricing and underwrites Lilly's biologic gambit.
• Food, QSR and consumer: one quantified read-through. On the FoodNavigator-USA Podcast (August 10), the discussion pegged GLP-1 users' impact on grocery baskets at a 5.3% reduction in overall spending and a 10.1% drop in savory-snack purchases, a cleaner number than last week's dairy-protein color, and a reminder the second-order consumer trade (packaged food, snacks, impulse buys) is real even when no single stock is named.
• Medtech and bariatric: Stryker as the second-order beneficiary. On Know Your Risk (August 14), a financial commentator framed obesity drugs as a multi-decade investment theme (citing 60–65% eventual adoption) and flagged Stryker (SYK) as a second-order medical-device beneficiary, the "more mobile, more active, more joint-replacement candidates" angle.

What changed vs last week

Issue #3 (August 10) was built around earnings-week fireworks: Lilly's Q2 blowout, the ~61%/39% Lilly-vs-Novo share split, CVS's formulary U-turn re-adding Zepbound, retatrutide's TRIUMPH-2/3 data, and AstraZeneca's CFO announcing an oral GLP-1. Here's the drift into this week:

• Flipped bullish to bearish (again): payer coverage. Last week's headline was a PBM (CVS) re-embracing Lilly. This week it's employers retreating: Starbucks dropping weight-loss coverage in October, GLP-1s hitting 11.4% of corporate claims, only 36% covering both indications. The access axis is oscillating week to week; the underlying fight (cost versus demand) is unresolved.
• Extended and confirmed: Lilly's oral franchise. Last week the story was that Foundeo missed on launch sales and trailed Novo's pill. This week the oral got the two things it needed: strong Phase 3 efficacy data (ACHIEVE-2) and a first European approval with scripts at a new high (~29,000). The narrative moved from "the launch is lagging" toward "the data supports the ramp."
• New this week: Lilly suing six companies over black-market retatrutide sales; Lilly's $3.8B vaccine-biotech acquisitions; the UK MHRA "200+ deaths" safety story and strengthened pancreatitis warning; a clean HFpEF clinical airing (SUMMIT -38 to -40%); a quantified grocery-basket read-through (-5.3% spend, -10.1% savory snacks); and the emergence of an operator (Andel's Bregman) built to bypass the PBM.
• Faded from last week: the CVS re-embrace got no fresh follow-up; retatrutide's TRIUMPH data cycled out (replaced by the counterfeit-lawsuit angle); AstraZeneca's oral program went quiet again; Novo's synthetic-semaglutide and China pivot got no update; the Sept 24 biologic hearing was not re-litigated this week and stays the key carryover catalyst.
• Still open questions: Amgen's MariTide, Viking's VK2735 and Roche's CT-388/CT-996 as standalone stories; the contract manufacturers (CTLT, LNZA, TMO) and pen suppliers; Novo's CagriSema and UBT-251; specific state legislation (e.g., the Tennessee Fair Rx Act); TrumpRx and most-favored-nation pricing; and, still, clean TRx/NRx script numbers with net price per script.