And Then You're Dead

The Autopsy / Case File · Chronic Disease Progression

Four Times the Count

Diabetes was the seventh leading cause of death in America in 2023, by the number of death certificates that named it as the underlying cause. A separate, larger number — more than four times as large — counts every certificate where diabetes shows up at all. The distance between those two counts is where most of diabetes actually kills people.

CDC's National Diabetes Statistics Report, published in January 2026 using 2023 data, found 95,190 U.S. death certificates listing diabetes as the underlying cause of death. On a total of 399,342 certificates, diabetes appeared as either the underlying or a contributing cause — more than four times the narrower count. Research on death-certificate accuracy explains most of that gap: diabetes is recorded on only about 40 percent of the death certificates of people who actually had it, and among those certificates, only 10 to 15 percent list diabetes itself as the underlying cause. Roughly 65 percent of deaths among people with diabetes get attributed instead to heart disease and stroke — the complications diabetes causes, not the disease that caused them. Diabetes is separately the leading cause of kidney failure in the country, responsible for 59 percent of new end-stage renal disease cases as of 2022, up from 46 percent in 2016.

And Then You're Dead · September 2026 · 13 min read

Join the discussion →
A finger with a small drop of blood held next to a blood glucose meter displaying a reading, the test strip glowing blue where it's inserted into the device
A blood glucose meter reads a fingerstick test. CDC found diabetes named as the underlying cause of death on 95,190 U.S. death certificates in 2023 — and named as any cause, underlying or contributing, on 399,342. Photo: Roy Zuo (RoyZuoMedia), CC BY-SA 4.0, via Wikimedia Commons.

Every death certificate asks a physician to answer a deceptively hard question: not just what killed this person, but what chain of causation led there, ending in a single line for "underlying cause of death" that vital statistics systems use to build the leading-causes-of-death rankings the public actually sees. Diabetes has a documented, specific problem with that line. It rarely kills anyone through diabetes itself, in the way a heart attack or a car crash has an obvious, immediate cause a physician can write down. It kills through years of downstream damage to the heart, the kidneys, the nerves, and the vascular system — damage that eventually produces a heart attack, a stroke, kidney failure, or a fatal infection following an amputation, and it's that final event, not the disease that spent years causing it, that most often ends up on the certificate.

What "Underlying Cause" Actually Captures

CDC's 2023 data makes the scale of that gap concrete. Diabetes ranked as the seventh leading cause of death nationally based on the 95,190 certificates naming it as the underlying cause — the single condition vital statistics systems trace the fatal chain of events back to. But diabetes was mentioned somewhere on 399,342 certificates that year, whether as the underlying cause or as a contributing condition physicians noted alongside whatever they did list as underlying. That second, larger number is closer to counting every death where diabetes was medically present and relevant; the first, smaller number is what actually produces diabetes's official ranking and the public health attention that follows a ranking. A condition mentioned on over four times as many certificates as the ones crediting it as the actual cause is a condition whose true mortality footprint the standard public ranking substantially understates.

0
2023 death certificates mentioning diabetes as an underlying or contributing cause, versus 95,190 naming it as underlying alone
0
of deaths among people with diabetes attributed instead to heart disease and stroke

Why Physicians Write Down the Complication, Not the Disease

Research specifically examining death-certificate accuracy for diabetes has found that the condition is documented on only about 40 percent of the certificates of decedents who are independently known to have had it — meaning six in ten certificates for people who actually had diabetes don't mention it at all. Among the certificates that do mention it, the share listing diabetes as the underlying cause rather than a contributing one is small: 10 to 15 percent, by the research cited in physician-certification studies. The larger pattern researchers describe is straightforward: diabetes is recorded more often when the underlying cause listed is cardiovascular disease, and recorded less often when the underlying cause is cancer or another condition less obviously connected to it in a certifying physician's mind — a bias toward crediting diabetes only when the complication looks like a textbook diabetic outcome, and toward omitting it when the death, on the surface, looks like something else. Roughly 65 percent of deaths among people with diabetes are attributed to heart disease and stroke specifically, the two complications diabetes accelerates most directly through vascular damage — meaning most of diabetes's actual lethality shows up in the mortality statistics as cardiovascular disease, filed under a different name.

Diabetes rarely kills anyone through diabetes itself. It kills through years of damage that eventually produces the heart attack, the stroke, or the kidney failure that gets written down instead.

The Kidneys and the Limbs, Named Separately Too

The same attribution pattern plays out in two of diabetes's other major complication pathways, each tracked by its own federal surveillance system rather than folded back into a diabetes-specific count. The National Institute of Diabetes and Digestive and Kidney Diseases' United States Renal Data System found that diabetes was the underlying cause behind 59 percent of new end-stage renal disease cases in 2022, up from 46 percent in 2016 — meaning most Americans newly starting dialysis or receiving a kidney transplant got there because of diabetes, in a dataset that counts them as kidney-failure patients, not as diabetes deaths, even when the ESRD eventually proves fatal. Diabetes-linked lower-extremity amputation follows a similar path: nontraumatic lower-extremity amputation rates among adults with diabetes fell 43.8 percent between 2000 and 2013, from 7.5 to 4.2 per 100 person-years, before stabilizing at that lower level — real, measurable progress — and research has separately established that people living with both end-stage renal disease and diabetes together face a substantially higher amputation risk than people with either condition alone. An amputation that leads to a fatal post-surgical infection or cardiovascular event gets coded to that infection or event. The diabetes that made the amputation necessary in the first place is, once again, the upstream cause a death certificate's single "underlying cause" line was never built to preserve.

What Gets Funded Follows What Gets Counted

The distinction between the 95,190-certificate figure and the 399,342-certificate figure isn't just an accounting curiosity for epidemiologists — it shapes where federal and state chronic-disease attention and funding actually flow, because leading-causes-of-death rankings are one of the primary inputs public health agencies use to prioritize research dollars, screening programs, and public awareness campaigns. Heart disease and stroke, the two conditions absorbing most of diabetes's real mortality burden under the current certification pattern, already sit at or near the top of that ranking and receive funding and attention commensurate with that position. Diabetes, ranked seventh on the strength of the narrower undercount, receives a correspondingly smaller share of that same prioritization system — even though the research on certificate sensitivity suggests its actual contribution to American mortality sits far closer to, or inside, the causes ranked above it. A chronic disease whose true toll is filed mostly under other diseases' names is, by construction, a chronic disease whose public health response is sized to a fraction of the problem.

The 2024 provisional numbers add a wrinkle to that picture rather than resolving it. Provisional CDC data for 2024 tallied 94,294 diabetes deaths at a rate of 26.4 per 100,000 — a continued decline from the pandemic-era peak of 103,294 deaths and 31.1 per 100,000 recorded in 2021, when COVID-19's direct physiological interaction with diabetes drove both diseases' mortality figures upward together. That decline is real and worth noting on its own terms. But it's a decline measured entirely within the narrower, underlying-cause-only count — the one already shown to capture a minority of diabetes's actual mortality footprint. A falling number built on top of a four-times undercount is still, at its core, a number describing a fraction of the disease's real reach.

A Ranking Built to Undercount Its Own Subject

None of this is evidence of fraud or carelessness on the part of individual physicians completing death certificates under time pressure with an ambiguous, multi-decade causal chain to summarize in one line. It's evidence that the entire system for producing a single "underlying cause of death" per certificate is structurally mismatched to a disease like diabetes, which doesn't kill in one identifiable moment but through a slow, branching cascade of organ damage that a certifying physician has to collapse into whichever downstream event happened to be the final one. The 399,342-certificate figure — diabetes mentioned anywhere, underlying or contributing — is the closer approximation of the disease's real mortality footprint. The 95,190-certificate figure, the one that actually produces diabetes's official rank among the nation's leading causes of death, is the one the public sees, cites, and builds health policy attention around.

And Then You're Dead

A death certificate has room for one underlying cause, and diabetes usually loses that slot to whatever organ finally failed first. The heart attack gets the credit. The stroke gets the credit. The kidney failure gets its own federal tracking system entirely. Diabetes, the condition that spent years making all three more likely, shows up in the data mostly as a footnote to causes that took its name off the certificate on their way out.

That's it. That's the whole thing.

Comments

Loading…

Comments are anonymous and moderated.

Sources