News · Heart & Metabolic
Heart deaths halved since 1990. The causes swapped places.
A JAMA Cardiology analysis of US deaths from 1990 to 2023 found the heart disease death rate down 58.7%, with smoking and air pollution receding as causes while high body weight and blood sugar climbed.
Based on a peer-reviewed Global Burden of Disease analysis in JAMA Cardiology
- 473,000 US deaths from ischemic heart disease in 2023, down 58.7% by age-standardized rate since 1990.
- 88.7% of those deaths were attributable to modifiable risk factors.
- High blood pressure led at 47.2%, then dietary risks at 38.6% and high LDL cholesterol at 28.5%.
- Burden from high body mass index rose 54.5% and from high blood sugar 38.8% since 1990.
- Smoking's contribution fell 33.3% and particulate pollution's fell 74.9% over the same period.
Heart disease is the leading cause of death in the United States, and it is killing far fewer people than it used to. An analysis in JAMA Cardiology puts the fall in the death rate at 58.7% since 1990.
That is one of the great public health achievements of the past three decades, and the same paper shows it has been partly undone from a different direction.
What the numbers look like now
In 2023, there were 473 000 IHD deaths in the United States, where IHD is ischemic heart disease, the narrowing of the arteries that feed the heart muscle.
The decline holds in the recent data too. Between 2010 and 2023, there was a 19.0% decrease in the death rate for males and a 24.5% decrease for females.
Then the number that frames everything else: 88.7% of IHD deaths were attributable to modifiable risk factors. Nearly nine in ten of these deaths trace back to something that could in principle be changed.
The three biggest contributors
High systolic blood pressure, dietary risks, and high LDL cholesterol were the leading risk factors in 2023, at 47.2%, 38.6% and 28.5% of deaths respectively.
Those figures overlap heavily, which is why they add to more than the total. A person with high pressure, poor diet and high cholesterol is counted in all three. Attributable fractions describe how much of the burden each factor touches, not how the burden divides up.
Blood pressure heading the list is unsurprising and worth restating, because it is the risk factor with the best treatments and the widest gap between what medicine can do and what happens in practice.
Diet carries the widest uncertainty of the three, its plausible range running from 17.2% to 56.8%. Measuring what a nation eats is much harder than measuring its blood pressure, and the interval says so plainly.
What went down, and what came up
The interesting part of this paper is the direction of travel for each risk.
Two have collapsed. Smoking and particulate matter pollution had the greatest decrease in attributable IHD mortality since 1990, falling 33.3% and 74.9% respectively. Those are the fruits of tobacco control and air quality regulation, and they are most of the reason the overall rate has halved.
Two have grown. High body mass index increased 54.5% as an attributable cause since 1990, and high fasting plasma glucose, meaning blood sugar, increased 38.8%.
So the composition of the problem has changed underneath the headline. The country stopped smoking and cleaned its air, then got heavier and more diabetic, and the second development has been eating into the gains from the first.
What a modelling study can and cannot say
This is an estimate built from vital records and a broad set of epidemiologic data rather than a count of individually verified causes. Attributing a death to a risk factor is a modelled inference, not an observation.
That is why every figure here carries an uncertainty range, and why some of them are wide. Total deaths are pinned reasonably tightly. Diet is not.
The paper is paywalled beyond its abstract, so the authors’ own account of their limitations was not available for this piece, and these caveats are the conservative ones.
Attribution also cannot settle causation at the individual level. It answers what would change across a population if exposure changed, which is a question about policy rather than about any one person.
Why it is useful anyway
You may be able to lower your risk of certain heart diseases by making heart-healthy lifestyle changes and managing any other medical conditions you have, which is true and easy to say.
What this analysis adds is a sense of proportion about where the remaining ground is. The risks that have been beaten back were beaten back mostly by regulation rather than by individual effort. The two that are rising are the ones currently handled as personal responsibility. Whichever way that reads, the 88.7% figure means the ceiling has not been reached.
People also ask
Why do the risk factor percentages add up to more than 100?
Because a single death can be attributed to several risks at once. Someone with high blood pressure, high cholesterol and a poor diet appears in all three counts. These are attributable fractions rather than slices of a pie, which is why blood pressure at 47.2% and diet at 38.6% can coexist with a combined figure of 88.7% for everything modifiable.
What does ischemic heart disease mean?
It is the condition where the arteries feeding the heart muscle narrow, so the muscle gets less blood than it needs. MedlinePlus describes it as happening slowly over time when a sticky substance called plaque builds up in the arteries that supply your heart muscle with blood. It is the most common form of heart disease and the leading cause of death in the United States.
If deaths are down so much, why does this matter?
Because the decline is not evenly earned. The rate fell 58.7% by age-standardized measure, driven substantially by fewer smokers and cleaner air. Over the same period the burden attributable to high body weight rose 54.5% and to high blood sugar 38.8%. Progress on one front has been partly offset on another.
What is a 95% uncertainty interval?
It is the modelled range the estimate is likely to fall within, and here some of them are wide. Dietary risks are given as 38.6% with a range from 17.2% to 56.8%, which is a large spread. Diet is hard to measure at population scale, and the interval is honest about that. Narrower intervals, like the one on total deaths, reflect better underlying data.
Does this tell me my personal risk?
No. These are population attributable fractions estimated from modelling across whole states, not individual predictions, and this is general information rather than medical advice. Anyone wanting to know their own risk should have blood pressure, cholesterol and blood sugar checked and discuss the results with a clinician.
References
- Benziger, C. P., Stark, B. A., Johnson, C. O., Roth, G. A. Modifiable Risk Factors and Attributable Ischemic Heart Disease Mortality in US States, 1990-2023. JAMA Cardiology, 2026.
- MedlinePlus. Heart Diseases. US National Library of Medicine.
- Centers for Disease Control and Prevention. Heart Disease Facts.