Roughly 3.27 million people died in the United States in 2022, the most recent year with comprehensive data, translating to an age-adjusted rate of about 833 per 100,000 people.1PubMed Central. Provisional Mortality Data – United States, 2022 That number dropped more than five percent from 2021, when COVID-19 was still pushing the annual count well above three million. But even in a “normal” year, three million-plus deaths is the baseline for a country of 330 million people. The number is large enough to be abstract, so what matters more than the headline figure is what’s behind it: who is dying, from what, at what age, and whether any of it is preventable.
What Kills the Most Americans
Heart disease and cancer have traded places at the top of the mortality charts for decades, and together they still account for close to half of all deaths each year. After those two come unintentional injuries, which have climbed sharply since the mid-1990s. When researchers looked at years of potential life lost before age 70 rather than raw death counts, unintentional injuries actually outranked heart disease, because they tend to kill people at younger ages.2PubMed Central. Leading Causes of Death in the US, 2019-2023 Stroke, chronic lower respiratory diseases, Alzheimer’s disease, and diabetes round out the top tier. The relative importance of each cause shifts depending on whether you count sheer numbers or years of life cut short, a distinction that matters for public health priorities.
Unintentional injuries deserve special attention because of the trajectory. Between 1994 and 2019, years of potential life lost from injuries rose substantially, driven first by motor vehicle crashes and later by drug overdoses. By 2019, unintentional injuries caused over 121,000 deaths among people under 70, and the trend has only worsened since.
Who Dies Younger and Why
The national average life expectancy masks enormous gaps. Between 2000 and 2019, life expectancy for Asian and Pacific Islander Americans rose to about 85.7 years, while for the American Indian and Alaska Native population it stayed flat at around 73.1 years. Black Americans gained nearly four years over that period, reaching 75.3 years, but still trailed White Americans at 78.9 years and Latino Americans at 82.2 years.3PubMed Central. Life expectancy by county, race, and ethnicity in the USA, 2000–2019: a systematic analysis of health disparities Those numbers are national averages; the county-level picture is far more dramatic.
An earlier analysis found that the gap between the longest-lived group (Asian American women) and the shortest-lived group (Black men in high-risk urban areas) was over 20 years. Within the same sex, the spread between best-off and worst-off groups was about 15 years for men and nearly 13 years for women, each several times larger than the simple national Black-White gap.4PLOS Medicine. Eight Americas: Investigating Mortality Disparities across Races, Counties, and Race-Counties in the United States Living in a particular county, belonging to a particular racial group, or both can shift your expected lifespan by a decade or more.
Income and education interact with these patterns in ways that are not purely linear. A large longitudinal study found that the relationship between income and mortality was steepest at lower incomes: an extra thousand dollars of household income mattered far more for survival if you were below roughly $22,500 than above it. At the upper end of the socioeconomic ladder, education was the stronger predictor, with college graduates showing substantially lower mortality than those with only a high school diploma, even after accounting for income.5PubMed. A comparison of the relationships of education and income with mortality: the National Longitudinal Mortality Study In practical terms, poverty shortens lives through material deprivation, while education seems to protect health through pathways like health literacy and occupational safety, even when income is held constant.
The Winter Death Spike
Americans do not die at the same rate year-round. Deaths peak consistently in winter, with roughly 16 percent more people dying in the coldest months than in the warmest ones. That pattern holds across U.S. cities regardless of local climate, whether the average winter temperature is well below freezing or mild enough for short sleeves.6PubMed Central. Winter Season Mortality: Will Climate Warming Bring Benefits? The uniformity of the spike across different temperature zones suggests that cold itself is not the main driver.
A forty-year analysis of U.S. mortality data pointed to influenza as the most likely culprit. The peak timing and autocorrelation patterns of deaths from heart disease, stroke, and diabetes all closely matched the timing of pneumonia and influenza deaths. Seasons dominated by more virulent influenza subtypes produced larger winter spikes, and the age distribution of deaths shifted in ways consistent with influenza epidemiology rather than cold exposure.7American Journal of Epidemiology. Influenza and the Winter Increase in Mortality in the United States, 1959–1999 Weather and shorter daylight hours may play supporting roles, but the primary engine behind the winter rise in American deaths appears to be respiratory viruses circulating through the population.
Where Americans Die
The setting in which people die has shifted considerably. In 2000, nearly half of all deaths occurred in a hospital. By 2018, that figure had fallen to about 35 percent. Meanwhile, the share of deaths happening at home climbed from roughly 23 percent to 31 percent, and deaths in long-term care facilities (including nursing homes and hospice) rose from about 23 percent to 27 percent.8Morbidity and Mortality Weekly Report. QuickStats: Percentage of Deaths, by Place of Death — National Vital Statistics System, United States, 2000–2018 Part of this reflects a genuine shift in preference: more families and patients choose comfort care over aggressive treatment at end of life. Part of it reflects the expansion of hospice services. The proportion of natural deaths occurring in dedicated hospice facilities jumped from 0.2 percent in 2003 to over 8 percent in 2017, accounting for nearly 1.7 million deaths over that period.9PubMed. National Trends in Hospice Facility Deaths in the United States, 2003-2017
These trends have practical implications for how death data are collected. A death in a hospital typically has more detailed clinical documentation and may include an autopsy, while a death at home certified by a private physician may rely on less information about the chain of events. As more deaths move out of hospitals, the quality of cause-of-death reporting could shift, too.
Deaths of Despair
Starting in the late 1990s, midlife mortality in the United States began rising in ways that set it apart from other wealthy countries. Drug overdoses, alcohol-related liver disease, and suicide collectively drove what researchers have labeled “deaths of despair.” A comparative study across the U.S., Canada, and the U.K. found that drug-related mortality was by far the biggest contributor to this trend. England, Wales, Canada, and Northern Ireland showed small but noticeable upticks in drug deaths, while the United States and Scotland experienced very pronounced increases.10PubMed Central. Comparing trends in mid-life ‘deaths of despair’ in the USA, Canada and UK, 2001–2019: is the USA an anomaly?
The American spike is not a natural consequence of being a large or wealthy country. It reflects specific conditions: the widespread prescribing of opioids starting in the 1990s, the subsequent transition to illicit fentanyl, and social and economic dislocations concentrated in certain communities. These deaths show up most starkly in middle-aged White adults without college degrees, though they are not limited to that group. The sheer scale is staggering: drug overdoses alone now kill more than 100,000 Americans per year, a figure that would have been almost inconceivable two decades ago.
How the Numbers Are Counted
Every death in the United States is documented on a standardized death certificate. A funeral director fills in the demographic details, and a certifier, usually a physician, medical examiner, or coroner, records the cause and manner of death. The certifier writes the cause of death in free text, listing the chain of conditions that led to death. When a definitive diagnosis cannot be established, certifiers can note “probable” or “presumed” causes using their best medical judgment.11PubMed Central. Advancements in the National Vital Statistics System to Meet the Real-Time Data Needs of a Pandemic Those certificates feed into the National Vital Statistics System, which produces the official mortality data.
The system works well enough to give reliable total counts, but cause-of-death accuracy is a different story. Death certificates are only as good as the information available to the person filling them out, and certain conditions are systematically undercounted. Dementia is a clear example: one U.K.-based study that compared death certificates against population-based dementia diagnoses found that only about 21 percent of people known to have dementia had it recorded on their death certificate in an earlier cohort. That figure improved to about 45 percent in a later cohort, but it still means more than half of dementia cases go uncaptured.12PubMed Central. Accuracy of death certification of dementia in population-based samples of older people: analysis over time The U.S. system likely faces similar challenges. Conditions that contribute to death without being the immediate cause, like diabetes or obesity, are routinely under-reported on death certificates.
This measurement problem becomes critical during crises. When COVID-19 hit, researchers turned to “excess mortality,” which compares the observed number of deaths in a given period against the number that would have been expected based on historical trends. It sidesteps the question of whether each individual death was correctly attributed by simply asking whether more people are dying than usual.13PubMed Central. Excess Mortality Estimation The approach has its own complexities—estimating the “expected” count requires careful modeling—but it remains one of the most reliable ways to detect when something is killing more people than normal.14PubMed. A Tutorial on Implementing Statistical Methods for Estimating Excess Death With a Case Study and Simulations on Estimating Excess Death in the Post-COVID-19 United States
The Maternal Mortality Puzzle
Official statistics showed a dramatic rise in pregnancy-related deaths in the U.S. over the past two decades, jumping from about 10 deaths per 100,000 live births in the early 2000s to nearly 24 per 100,000 by the late 2010s. That rise made international headlines and fueled urgent policy debates. But the story is more complicated than the headline number suggests. A large share of the apparent increase came from two changes in how deaths are recorded: the adoption of a pregnancy checkbox on death certificates and the introduction of new diagnostic codes for conditions like kidney disease in pregnancy.15PubMed Central. Factors Underlying the Temporal Increase in Maternal Mortality in the United States When researchers statistically adjusted for these surveillance improvements, the steep upward trend largely disappeared.
That does not mean everything is fine. Under an alternative formulation that accounted for changing surveillance, direct obstetrical deaths from conditions like postpartum hemorrhage and eclampsia actually decreased between the early 2000s and recent years, while deaths from conditions like cardiomyopathy and preexisting hypertension increased.16American Journal of Obstetrics & Gynecology. Maternal mortality in the United States: changes in coding, publication, and data release, 2018 And the racial disparity is real regardless of how you measure: Black women face three to four times the pregnancy-related death rate of White women, a gap that persists after accounting for income and education.17PubMed Central. Maternal Mortality in the United States: Updates on Trends, Causes, and Solutions The maternal mortality story is a useful case study in how changes to data systems can create the appearance of an epidemic, even while a genuine and alarming disparity hides within the numbers.
Air Pollution’s Hidden Death Toll
Not all causes of death show up neatly on a death certificate. Fine particulate air pollution (the tiny particles produced by combustion in vehicles, power plants, and industry) is estimated to be responsible for five to ten percent of total annual premature deaths in the contiguous United States.18Nature. Premature mortality related to United States cross-state air pollution A detailed analysis estimated that fine particulate exposure contributed to more than 56,000 excess cardiovascular deaths, about 40,000 excess stroke deaths, roughly 17,500 lung cancer deaths, and nearly 20,000 excess dementia deaths in a single year.19JAMA Network Open. Burden of Cause-Specific Mortality Associated With PM2.5 Air Pollution in the United States
None of these deaths appear in official statistics as “caused by air pollution.” The death certificate will say heart attack, stroke, or lung cancer. The pollution contribution is invisible at the individual level and only becomes apparent through population-level epidemiology. That’s one reason environmental health often gets less policy urgency than infectious disease or road safety: nobody’s death certificate reads “PM2.5.”
How a Century Changed the Causes of Death
A hundred years ago, the American mortality landscape looked nothing like it does now. Until about 1910, infectious diseases far outpaced degenerative diseases as leading causes of death, and death rates were highest in cities, where crowding spread tuberculosis, typhoid, and other pathogens.20American Journal of Epidemiology. The Rise of the Current Mortality Pattern of the United States, 1890–1930 By 1930, that picture had flipped: degenerative diseases like heart disease and cancer had become the dominant killers, and the urban penalty had vanished for infectious disease. The shift is known as the epidemiologic transition, and the U.S. followed what scholars describe as the “classical” pattern, driven by sanitation improvements, better nutrition, and eventually vaccination and antibiotics.21Preventive Medicine. A century of epidemiologic transition in the United States
Understanding that transition puts today’s numbers in context. The fact that most Americans now die from heart disease, cancer, and stroke is not a sign of failure; it’s what happens when a society eliminates the infections that used to kill people in childhood and early adulthood. People live long enough to develop chronic diseases. The trade-off is that the new killers are slower, costlier, and harder to prevent with a single intervention like a vaccine.
How Many Deaths Are Preventable
One of the more sobering questions lurking behind the annual death count is how many of those deaths didn’t have to happen. A meta-analysis pooling data from over 12,500 hospital deaths found that about 3.1 percent were judged preventable by clinical reviewers. When limited to patients who would have been expected to live at least three more months, the figure dropped to roughly 0.5 to 1 percent.22PubMed Central. Rate of Preventable Mortality in Hospitalized Patients: a Systematic Review and Meta-analysis Those percentages sound small, but applied to the millions of hospital deaths that occur each year, even a fraction of a percent represents thousands of people.
Hospital-based preventable deaths are only one slice of the picture. Many more deaths are preventable in a broader public health sense. Smoking, physical inactivity, poor diet, and excessive alcohol consumption contribute to hundreds of thousands of deaths annually from heart disease, cancer, diabetes, and liver disease. The concept of “avoidable mortality” captures deaths that could have been prevented through public health measures or timely medical care, and by that metric the United States consistently performs worse than peer countries. Higher rates of gun violence, drug overdoses, and uninsurance all contribute to a gap that cannot be explained by demographics alone.
What the Aging Population Means for Future Numbers
The annual death count is going to rise regardless of whether Americans get healthier or sicker, simply because the population is aging. By 2060, the number of Americans aged 65 and older is projected to nearly double, with the oldest group, those 85 and up, tripling in size.23PubMed Central. Navigating the Future of Elderly Healthcare: A Comprehensive Analysis of Aging Populations and Mortality Trends Using National Inpatient Sample (NIS) Data (2010-2024) More older people means more deaths per year, even if age-specific death rates hold steady or improve. That distinction between crude death counts and age-adjusted rates is important for anyone trying to interpret year-over-year changes. A rising death count does not necessarily mean the country is getting less healthy; it can simply mean there are more people in the age groups where death is most likely.
The aging wave will also reshape which causes of death dominate the statistics. Dementia, which is already undercounted on death certificates, will become an increasingly significant contributor. Heart failure, falls, and kidney disease are all more common in older populations and will claim a larger share of total deaths. Healthcare systems built around acute interventions will face growing pressure from conditions that require years of management rather than a single fix. The three-million-plus annual death count Americans are accustomed to is not a ceiling. Within a few decades, four million annual deaths may be the norm, even in a country that is technically living longer on average.