A quasi-experimental study published in Cureus reports that when five state Medicaid programs withdrew coverage of glucagon-like peptide-1 receptor agonists for obesity, reimbursed dispensing of those drugs collapsed within a single quarter. The withdrawals took effect in late 2025 and January 2026, and other states have narrowed coverage or are reconsidering it.

The drug class at issue is the GLP-1 receptor agonists prescribed for obesity rather than diabetes, a distinction that matters because Medicaid reimburses the two indications separately. The authors note that the benefits of these drugs appear to depend on continued use, which is why an abrupt coverage change is a research question and not just a budget line.

Key results

The five withdrawing states were not marginal players: they accounted for 2,428,709 of the 3,725,732 obesity-indicated GLP-1 prescriptions Medicaid reimbursed in 2025, or 65.2 percent of the national total. Compared with 12 states that continued coverage, withdrawal was associated with an 84.7 percent decline in obesity-indicated dispensing (95% CI: 64.5 to 93.4 percent), equivalent to 29.1 fewer prescriptions per 1,000 enrollees per quarter.

The four states that withdrew on January 1, 2026 reimbursed 661,353 obesity-indicated prescriptions in the preceding quarter and 134,647 in the next, a drop of 79.6 percent. Two states that tightened criteria rather than cutting coverage outright saw declines of 23 to 37 percent.

The diabetes comparison

The study used each state's diabetes-indicated GLP-1 dispensing as a within-state control, and that line barely moved: a change of minus 3.2 percent, with confidence intervals spanning zero. Pre-announcement differences between withdrawing and continuing states were small and not statistically significant, which is the check that makes the headline estimate worth reading.

Nor did older antiobesity medications fill the gap. Dispensing of those drugs rose by at most 2.9 prescriptions per 1,000 enrollees, which the authors describe as little replacement by other agents.

Limitations

This is administrative dispensing data, not patient-level outcomes, so it shows reimbursed prescriptions disappearing rather than what happened to the people who had been taking them. The comparison group was 12 continuing states, and the authors say an approach such as honoring existing authorizations might limit treatment interruption, but add that this has not been evaluated.

Bottom line

Coverage withdrawal was associated with the loss of most reimbursed obesity-indicated dispensing inside one quarter. What remains unmeasured is what those enrollees did next.