The phrase "insurance coverage collapse" implies something national and sudden — a switch thrown, a class of drugs cut loose. That is not what happened in 2025. What happened was a map forming: a patchwork of public plans, state Medicaid programs, and self-insured employers quietly deciding, one budget cycle at a time, that GLP-1s for weight loss cost more than they were willing to carry.
My thesis is simple and, I think, underappreciated: the retreat is real, but it is not happening to the people most headlines imply. It is concentrated almost entirely among payers who bear the drug cost directly — states insuring their own employees, Medicaid programs, and large companies that self-fund their health plans. Where a commercial insurer can raise premiums to absorb the hit, coverage has mostly held. Where the payer eats the cost with no way to pass it on, coverage is being cut. Your access now depends less on your BMI than on who signs your paycheck and which state you live in.
Understand that split and the "collapse" stops looking like a wave and starts looking like what it is: an actuarial map.
The arithmetic that broke the public plans first
Start with the number that made every benefits director flinch. Wegovy carries a list price around $1,349 a month; Zepbound lands near $1,059; Ozempic and Mounjaro sit in the $970 to $1,070 range. Net prices after rebates are lower, but for a self-funded plan the exposure is roughly a thousand dollars per member per month, indefinitely, for a chronic condition affecting a large share of the covered population.
North Carolina was the tell. Its State Health Plan, covering some 740,000 people, projected that GLP-1 weight-loss coverage would cost more than $1 billion over six years — and in early 2024 it voted to stop covering the drugs for obesity entirely, cutting off roughly 25,000 members already on them. That decision, more than any FDA action, is the real origin of the "collapse" story, and it came from a treasurer with a spreadsheet, not a regulator with a safety concern.
West Virginia's public employees' insurance agency ended its GLP-1 weight-loss pilot on similar math. Connecticut's state plan watched GLP-1 spending climb into the tens of millions and spent 2024 and 2025 debating restrictions. The common thread is not ideology. It is that a state cannot deficit-spend its health plan, and a drug that a quarter of your workforce may qualify for at $1,000 a month is a line item that eats everything around it.
Medicaid is the sharpest line on the map
Nowhere is the geography starker than Medicaid, where coverage for obesity is genuinely a coin flip by state. As of the last full count, only about a dozen state Medicaid programs covered GLP-1s for weight loss at all; the rest cover them only for diabetes, or not at all for this purpose. That is the single fact that most cleanly proves my thesis: same drug, same clinical profile, same federal program — and whether it is covered depends entirely on which side of a state line you happen to live on.
The states that do cover it are feeling the bill. Programs that opened the door have reported obesity-drug spending running into the tens of millions annually, and several have layered on prior-authorization gates, BMI thresholds, and step-therapy requirements to slow the bleed. Coverage on paper and coverage in practice have drifted apart — a plan can technically "cover" a drug while making it hard enough to obtain that utilization stays low. That gap, I'd argue, is where a lot of the real 2025 story lives, and it never shows up as a clean "dropped coverage" headline.
Then Washington declined to be the backstop
The one move that could have overridden the map was federal. Medicare has been barred by statute since 2003 from covering drugs "used for weight loss" — a line written into the Part D law long before anyone imagined a drug like semaglutide. In late 2024 the outgoing administration proposed reinterpreting that language to let Medicare and Medicaid cover anti-obesity medication, a change the government's own estimates pegged at roughly $35 billion for Medicare and $11 billion for Medicaid over a decade.
In April 2025 the incoming administration declined to finalize it. That single non-decision is the most consequential event in this entire story, and it got a fraction of the coverage the individual employer cuts did. It meant the federal government would not be the payer of last resort — that the map would be allowed to stand, with tens of millions of Medicare beneficiaries on the wrong side of a twenty-two-year-old sentence. There is a narrow exception: after the SELECT cardiovascular-outcomes trial, Medicare Part D plans were permitted to cover Wegovy for heart-attack and stroke risk reduction in people with established cardiovascular disease. But that is coverage for a heart indication that happens to use a weight-loss drug — the obesity door stayed shut.
Same drug, same clinical profile, same federal program — and whether it's covered comes down to which side of a state line you live on. That is not a collapse. It is a map.
The commercial side that didn't collapse
Here is the counter-argument I take most seriously, because it complicates my own framing. If coverage were truly collapsing, you would expect the large commercial insurers to be leading the retreat. Mostly, they aren't. A fully insured commercial plan prices the GLP-1 risk into next year's premiums and moves on; the cost is real but it is distributed and recoverable. The high-profile 2025 cuts — Blue Cross Blue Shield of Michigan ending weight-loss coverage for certain large-group commercial members at the start of the year, various employer carve-outs — are notable precisely because they are exceptions worth reporting, not the baseline.
And there is a genuine escape valve the "collapse" narrative tends to ignore: manufacturer direct-pay. Lilly and Novo now sell vials directly to cash patients in the roughly $499-a-month range, as this magazine has covered. That is not insurance, it is not cheap, and it is not available to everyone — but it means "my plan dropped it" no longer equals "I cannot obtain it" the way it did in 2022. The channel shifted more than the access did, for those who can pay.
So is "collapse" the wrong word? Partly. I'd defend it only in a specific sense: for the public-plan and Medicaid populations, in the states that pulled back, coverage did not erode gradually — it was cut on a date, by a vote, for a budget reason. For those people, "collapse" is exactly right. For the commercially insured in a supportive state, almost nothing changed. Both things are true at once, which is why the national headline is always wrong and the map is always right.
Why the map is now the real eligibility test
Step back and the pattern is clean. Clinical guidelines say who should get these drugs. The map decides who can. And the map is drawn not by medicine but by who absorbs the cost: self-insured public payers cut first and hardest, Medicaid splits by state, commercial insurance mostly holds, and the federal government declined to redraw any of it.
I think that is the honest 2025 story, and it is more uncomfortable than "insurers dropped Ozempic." It is that the United States built, without deciding to, a system where the same person with the same diagnosis is covered in one state and not the next, covered as a teacher here and not there, covered for a heart indication but not for the obesity that drives it. That is not a policy. It is the absence of one, rendered as geography. When you read "coverage collapse," don't picture a wave. Picture a map — and check which side of the line you're standing on.
Ozemback — July 2026