Losing Medicaid coverage worsens diabetes control within months, study of 39,000 patients shows

Lose your insurance, and your diabetes gets worse. That is the blunt finding from a case-control study of more than 39,000 adults at community health centers across 20 states, published March 20 in JAMA Health Forum.

The study, led by Nathalie Huguet, an associate professor of family medicine at Oregon Health and Science University, defines insurance churn as two or more consecutive uninsured visits - a marker of sustained coverage loss rather than a brief administrative gap. Patients who experienced this churn had poorer glycemic control, required more intensive medications including insulin, and suffered more acute diabetes complications than matched patients who maintained continuous coverage.

What happened after coverage dropped

The researchers analyzed electronic health records from community health centers - clinics that primarily serve patients living in or near poverty. They identified patients with diabetes who lost insurance coverage and compared them with patients who were nearly identical at baseline but retained their coverage.

The results were consistent across multiple measures. Patients who churned out of coverage showed worsening hemoglobin A1c levels, the standard measure of average blood sugar over roughly three months. They were more likely to be prescribed insulin and other high-intensity diabetes medications. And they experienced more acute diabetes complications during the study period.

The matched design is important. These were not inherently sicker patients who happened to lose insurance. At the start of the study, the churners and the continuously insured looked the same. The divergence in outcomes appeared after coverage was lost.

The medication paradox

One of the more troubling findings involves insulin use. Patients who lost insurance were prescribed insulin at higher rates - but insulin is expensive, often prohibitively so for uninsured patients. The study raises a direct question that Huguet emphasized: how can patients manage complex diabetes treatment when they cannot afford the medications they need?

Community health centers provide discounted medications on sliding-fee scales, which may partially explain why some patients continued receiving prescriptions. But the gap between being prescribed a medication and consistently taking it is wide, particularly for insulin, which requires refrigeration, syringes or pens, blood glucose monitoring supplies, and regular dose adjustments.

The study did not directly measure medication adherence, so it cannot confirm whether patients who were prescribed insulin actually used it consistently. But the combination of worse blood sugar control and more intensive prescriptions suggests that coverage loss disrupted effective disease management even when clinics attempted to maintain care.

Serious complications: an early warning

Acute diabetes complications - including diabetic ketoacidosis and severe hypoglycemia - were more common among patients who lost coverage. But serious long-term complications like amputations and kidney failure did not appear at significantly higher rates during the study window.

Huguet cautioned against reading that as reassurance. Complications such as amputations and end-stage kidney disease develop over years or decades. The study period may simply be too short to capture the full downstream consequences of coverage loss. What the data show instead is an early-stage deterioration: the disease becoming harder to control, the treatment regimen intensifying, and acute episodes increasing. These are warning signs of worse outcomes to come.

Timing and the Medicaid question

The findings arrive at a moment when millions of Americans face potential Medicaid losses. The pandemic-era continuous enrollment requirement, which kept people on Medicaid regardless of eligibility changes, has ended. States have been conducting redeterminations, and millions have been disenrolled - many for procedural reasons rather than actual ineligibility.

This study builds on Huguet's earlier research showing that low-income patients with diabetes are more likely to experience insurance instability in the first place. Together, the two studies form a chain: diabetes patients are disproportionately likely to lose coverage, and losing coverage makes their diabetes worse.

The implications extend beyond individual patients. Community health centers, which serve as the primary care safety net for low-income populations, rely heavily on Medicaid reimbursement. If Medicaid enrollment shrinks, these clinics face a double bind: more uninsured patients needing more intensive care, with less revenue to provide it.

Jennifer DeVoe, a professor of family medicine at OHSU and co-author on the study, pointed to the systemic consequences. Patients without insurance are more likely to use emergency departments for conditions that primary care could have managed. That shift is costlier for the healthcare system and worse for patients, who may arrive with complications that earlier intervention could have prevented.

Limitations of the evidence

The study has several constraints worth noting. It is observational, not experimental. While the matched design strengthens causal inference, unmeasured differences between groups could still influence the results. Patients who lose insurance may also experience other disruptions - job loss, housing instability, stress - that independently worsen diabetes outcomes.

The definition of insurance churn as two or more consecutive uninsured visits captures sustained loss but may miss shorter gaps or patients who lost coverage between visits. The study also cannot distinguish between patients who lost Medicaid, lost employer-sponsored insurance, or never had coverage to begin with, though the community health center population skews heavily toward Medicaid.

The follow-up period was limited, which means the full trajectory of health consequences after coverage loss remains uncertain. Longer studies will be needed to determine whether the deterioration observed here stabilizes, worsens, or reverses if coverage is eventually restored.

Despite these limitations, the core finding is clear and consistent with prior research: insurance coverage is not a bureaucratic abstraction for people with chronic disease. It is a determinant of whether their condition is managed or unmanaged, controlled or spiraling. For diabetes patients living near the poverty line, a gap in coverage is a gap in health.