When an economic model calls a weight-loss drug a good buy, my reflex is to check who set the price, because the answer is usually the drugmaker. Walter Reed’s new numbers came with a price set by the Pentagon, $250 a month, and they left me with a question I hadn’t planned to ask. Why does the Defense Department no longer cover these drugs for weight loss in Medicare-eligible military retirees, when its own physicians now find them worth paying for in people aged 65 to 75?
Taylor Neuman, an internal medicine physician, heads the metabolic health and weight management service at Walter Reed National Military Medical Center. At ObesityWeek in National Harbor, Maryland, she presented a cost-effectiveness analysis that used the Institute for Clinical and Economic Review’s validated Markov model on adults in the Military Health System, which covers more than 9 million people. Among the 17,688 adults aged 65 to 75 with obesity or overweight, at least one related condition and no type 2 diabetes, tirzepatide (Zepbound) cost $15,742 for each quality-adjusted life year it added over lifestyle changes alone. Injectable semaglutide (Wegovy) came in at $25,770 and oral semaglutide at $35,873. The bar the team used was $100,000.
That’s good value, which is different from free money. For tirzepatide to pay for itself outright in the older group, the annual price would have to fall from $3,000 to $1,906.10 or less, and over a lifetime the drug added about $10,998 per person in net cost for 0.7 healthy years. By the team’s own standard, that’s still a comfortable pass.
The younger cohort’s results are where the model gets startling. Among 354,978 adults aged 18 to 64, with an average age of 45 and an average BMI of 35.2, tirzepatide cost $698 per quality-adjusted year, the kind of ratio you expect from a cheap generic, not a brand-name Eli Lilly injectable. It added about $1,046 per person over a lifetime against 1.5 healthy years, and its break-even price came out at $2,942.60 a year, about $57 below what the military pays. Neuman said the drug “would only require a pretty low price reduction to meet that threshold,” and that “our numbers are more significantly cost-effective than what you see in other published studies.” Both semaglutide versions cleared easily too, at $10,838 (injectable) and $14,118 (oral).
Wait, why does diabetes move the math?
I assumed the payoff would come from heart attacks and strokes. The sensitivity analysis points somewhere else: for tirzepatide in the younger group, the inputs that swung the result hardest were the cost of treating type 2 diabetes, how often diabetes develops and baseline BMI. That doesn’t tell you what share of the savings came from diabetes rather than the heart, and the presentation as reported didn’t break that down. It does tell you the verdict leans hardest on the diabetes assumptions.
So why would diabetes carry that much weight? Look at what the drug does in a body. GLP-1 is a gut hormone. Your intestine releases it after a meal, and it nudges the pancreas to put out insulin, slows the stomach so food sits there longer and tells the brain you’ve had enough. Tirzepatide layers a second gut-hormone signal, GIP, on top. In a 45-year-old with a BMI around 35 and no diabetes yet, the drug is pressing on the road to diabetes for decades, and every case it delays or prevents takes years of glucose drugs, monitoring and complication care off the system’s bill. A 70-year-old has fewer of those years left to protect, which fits the older group’s smaller gains. Per person, the model credited tirzepatide with preventing 0.05 heart attacks or strokes in the older bracket against 0.27 in younger adults, roughly a fifth.
What happened on August 31
TRICARE For Life is the plan for Medicare-eligible military retirees and their families, which makes it the coverage most of that 65-to-75 bracket relies on. On August 31, 2025, TRICARE For Life stopped covering GLP-1 drugs approved only for weight loss, such as Wegovy and Zepbound. The Military Officers Association of America attributed the change to new “regulatory controls” described in a TRICARE announcement. Prime and Select members, a mostly younger group, kept weight-loss coverage with prior authorization, and the diabetes versions, Ozempic and Mounjaro, stayed covered across TRICARE with prior authorization. The people who lost access are older beneficiaries without diabetes taking the drugs for their weight, the same no-diabetes profile Walter Reed modeled in its older group.
Medicare’s workaround mostly doesn’t reach them. CMS’s short-term GLP-1 Bridge program, now extended through the end of 2027, gives eligible beneficiaries obesity drugs for a $50 copay, built on a $245 net price the manufacturers agreed to. That’s within $5 a month of the military’s price, so the federal government is already paying roughly what its own model calls good value. But MOAA reports that CMS’s GLP-1 programs require Part D enrollment, which leaves out most TRICARE For Life beneficiaries. More than 800 of them have contacted the group since the change. “I have not felt this good since Airborne School,” one wrote. Another said, “I’ve been on TFL for one month, and immediately my health is in greater jeopardy.” A third put it more bluntly: “That’s not TRICARE For Life … that’s TRICARE For Some.”
The fight in court went badly for the government. Retired Air Force Col. Derence Fivehouse, a former Air Force lawyer, sued to reverse the cut. The Justice Department attorney defending it, Rudy Renfer, was fired March 13, a day after he tried to resign, when the court found that his filings, produced with help from artificial intelligence, were riddled with errors including “fabricated quotations.” Fivehouse himself reported the errors to Judge Robert T. Numbers. The weight-loss coverage still hasn’t come back.
Two prices for the same drug
The usual complaint about GLP-1s is that no payer can afford them at scale, and it isn’t baseless. ICER’s final obesity report, released in December 2025, called the drugs “highly cost-effective” on average at estimated U.S. net prices of $6,829 a year for injectable semaglutide and $7,973 for tirzepatide, while warning that “the US healthcare system will be strained to provide them to most people who need them.” A low cost per healthy year can still mean a huge bill this year, and Walter Reed’s presentation, as reported, included no budget estimate. But ICER’s own suggested fix was for manufacturers to consider steep discounts in exchange for higher volume, and the military’s $3,000 a year shows what that looks like. A payer at ICER’s estimated tirzepatide price spends more than 2.5 times as much for the same drug. I’d stop treating that higher price as a law of nature, since the drugmakers have now agreed to roughly $250 a month with two federal buyers.
Where the model could be too kind
This is a conference presentation, not a peer-reviewed paper. The published account names no funding source, and Neuman reports no relevant financial disclosures. It also doesn’t say how the model handles people who stop the drug, which is where the diabetes savings are most exposed. In a Cleveland Clinic cohort of 7,881 adults, more than 20% quit within three months and another 32% stopped between three and 12 months, citing cost, insurance problems, side effects and shortages. After a year, early quitters had lost 3.6% of their body weight, later quitters 6.8% and people who stayed on 11.9%.
The weight also comes back. In Lilly’s SURMOUNT-4 trial, people switched to placebo after 36 weeks on tirzepatide regained 14% of their body weight over the next year, measured from where they stood at the switch. If the Walter Reed model assumes years of steady use, it is crediting diabetes prevention that many patients won’t stay on the drug long enough to get.
The comparator needs a hard look too. The published account doesn’t define “lifestyle modifications alone,” and what that arm contains changes the answer. The Diabetes Prevention Program, a structured effort aiming for 7% weight loss and 150 minutes of weekly exercise, cut type 2 diabetes risk by 58% across 3,234 high-risk adults and by 71% in those over 60. Fifteen years out, the reduction had narrowed to 27% and was still there. That’s the same diabetes pathway the Walter Reed economics lean on, reached without an injection, and the strongest lifestyle arm is a funded program like that one, not advice at a checkup. The analysis also left out the newer 7.2 mg semaglutide dose.
None of those caveats rescues the coverage decision. They argue for a tighter model and a fairer comparator, while the department’s own physicians found the drug good value for 65-to-75-year-olds at a price Medicare is also paying. The only public explanation MOAA has reported for cutting those patients off is “regulatory controls,” and the Pentagon owes them a longer one.
If I were 70, on TRICARE For Life and watching my blood sugar creep up, I’d enroll in a structured diabetes-prevention program this month instead of waiting on the Pentagon, and I’d send my representative the $15,742 figure.
Sources
- Healio / Endocrine Today: GLP-1s cost-effective in US Military Health System (ObesityWeek 2026, Neuman, Walter Reed)
- MOAA: Weight-loss drug coverage ending for TRICARE For Life users (2025)
- MOAA: Legal case seeking GLP-1 TRICARE For Life reversal makes headlines after lawyer’s firing (March 2026)
- MOAA: Summer AiA 2026, fighting for TRICARE For Life beneficiaries left behind
- ICER: Final evidence report on treatments for obesity (December 2025)
- KFF: CMS extends Medicare’s short-term Bridge program for GLP-1 obesity drug coverage
- JAMA: SURMOUNT-4 randomized withdrawal trial of tirzepatide (Eli Lilly-funded)
- News-Medical: Cleveland Clinic real-world study of GLP-1 discontinuation (Gasoyan et al., Obesity, 2025)
- NIDDK: Diabetes Prevention Program and Outcomes Study results
- JAMA: GLP-1s, from gut hormone discovery to a public health revolution (2026)