Bottom line
Longevity conversations rarely dwell on accidental death. For younger people, it is often the thing that ends a life first. World Health Organization figures put injury-related causes in three of the five leading causes of death among people ages 5–29 6.
Road traffic alone kills about 1.16 million people worldwide each year. It is the leading cause of death in that age band 8.
Intuition about risk often fails. People fear the airplane and treat the daily drive as background. One view splits frequency from severity. What sets the death toll is the product of the two 3.
U.S. numbers show this is not a rare tragedy. Traffic deaths were 36,355 in 2019 and rose to 42,915 in 2021, an 18% increase in two years 1.
Official counts line up with that reading: accidental death is common, it gets less attention than it merits, and a sizable share of it can be prevented 16.
What it is
Accidental death, in this article, means death from unintentional injury. That includes road crashes, falls, drowning, poisoning, burns, and also suffocation and electric shock.
The World Health Organization splits "injury" in two. The unintentional half kills about 3.16 million people a year. Intentional injury — violence directed at oneself or at other people — kills about 1.25 million. Together that is about 4.4 million 6.
Why call it underestimated? The first layer is psychological. Most of these deaths happen in an instant. They do not offer the long warning that chronic disease does.
The second layer is arithmetic. One account argues that intuition gets captured by severity. An air disaster is nearly total. A shark bite leads the news 3.
The number of deaths, though, is frequency times severity. A single car crash is fatal well under 1% of the time, yet crashes happen every day. An air disaster is extremely lethal and extremely rare 3.
Multiply the two and the ranking flips. From 2000 through 2018 in the United States, scheduled airline flights killed 778 people. Roads killed 723,530 people over the same span, a gap of nearly a thousandfold 2.
A second comparison lines the miles up and reaches a similar conclusion. The chance of dying in a car crash is about 450 times the chance of dying on a scheduled flight 3.
That is the third layer. The more ordinary a way of dying, the less it looks like news. The less it looks like news, the smaller the place it holds in a person's risk budget.
Why it matters
For people who still have decades ahead, accidents are the leading opponent. In the WHO ranking for ages 5–29, road-traffic injury, homicide, and suicide — three of the top five — are all tied to injury 6.
Road-traffic injury is the first cause of death in that age group. More than half of those who die are pedestrians, cyclists, and motorcyclists: people without a steel shell around them 8.
The global map is sharply uneven. Ninety-two percent of road deaths occur in low- and middle-income countries, and those countries have only about 60% of the world's motor vehicles 8.
The recent U.S. trend has been getting worse, not better. 2021 brought about 6,500 more deaths than 2019. The rise covered every age group, every kind of road, and every month 1.
Early in the pandemic, fewer cars on the road did not make roads safer. That counterintuitive pattern sits inside the same rise. Speeding-related crashes in 2020 were 17% higher than in 2019, and alcohol-related crashes were 14% higher 1.
Drowning is a quiet killer in childhood. About 300,000 people drown worldwide each year, and children under 5 account for nearly a quarter of them. Drowning is the fourth leading cause of death at ages 1–4 and the third at ages 5–14 7.
The drowning death rate among males is more than twice the rate among females. Greater exposure and riskier behavior are the explanations given 7.
Falls are easy to file under "people fall when they are old." WHO figures say more than 680,000 people die from falls each year, and the organization calls this a growing public-health problem that is underestimated 6.
These deaths share a shape. They happen in a moment, there is no second chance, and the room for choices beforehand is large.
Different perspectives
Main positions
One frame fits in a sentence: do not mix up frequency and severity; look at the product 3.
One view puts the most likely fatal accident over the next ten years in a car, one's own or someone else's 2.
The scale cited: in 2018 nearly 37,000 people in the United States died in crashes, produced on average by about 18,000 crashes a day 2.
Where those deaths happen also defies intuition. Among fatal crashes in 2017, only 18% were on freeways. Two-lane country roads alone accounted for 38%, higher than freeways and higher than intersections at 20% 2.
The causes cluster in three things. Among drivers killed on freeways, 31% had alcohol in the body, and nearly 85% of those were over the legal limit. Speeding was involved in about 29% of driver deaths. Distracted driving accounted for about one in ten 2.
The response described has two lines. The first is a not-to-do list: no driving with alcohol in the body, no speeding, no looking at a phone while driving, and no getting on the road short of sleep 2.
The second is a change of frame. Suppose someone on the road today had been given the job of killing you with a car. How would you drive? You would watch every lane change and every hard stop again 2.
On a smaller point, a rule passed on from a truck driver: at an intersection, look left first, then right, because traffic from the left is the first threat 2.
On whether a helmet makes a cyclist take more risks, a 2019 systematic review is cited. Evidence for that kind of risk compensation is weak. The review instead observed that helmeted cyclists behaved more safely 5.
Road-safety expert and former NHTSA head Mark Rosekind has discussed deaths on the road in an interview. The sentence placed most prominently on the episode page says, in substance, that we all move around inside cars and yet tolerate road deaths and injuries that need not be accepted 4.
Rosekind's record is itself a heavy credential on this track. He studied fatigue and jet lag at NASA, served as a member of the U.S. National Transportation Safety Board, and went to the scene of seven major transport accidents. He later became the 15th administrator of the National Highway Traffic Safety Administration 4.
That episode keeps returning to one theme: the role of fatigue in accidental death, and how much of what is being discussed sits under our own control 4.
A separate discussion of alcohol touches the edge of this topic. Driving drunk is a terrible thing, and it can kill you or someone else 9. Getting into a pool or a lake while intoxicated is extremely dangerous. People drown that way 9.
Where views agree and differ
The judgment lines up closely with official data. Road injury is the leading cause of death among young people 68, and crash deaths in the United States have risen rather than fallen 1.
If there is a disagreement, it is with popular intuition. One view ranks the car crash, not the plane crash or the helicopter crash, as the accident of greatest concern. The reason is the arithmetic of frequency times severity 3.
PearHealth's synthesis
We adopt the mental model of frequency times severity. It does not dissolve anxiety. It moves attention from the frightening ways of dying toward the frequent ones.
We adopt two lines of action: keep one's own not-to-do list, and drive defensively on the assumption that other people will break the rules 2.
We also note what the data point to. Alcohol, speeding, distraction, and fatigue are changeable factors. None of them is "luck" 24.
The path we do not take is abandoning travel out of fear. The conclusion of that weighing is not to give up flying and driving, but to do them with the arithmetic and with discipline 3.
For readers with an older adult or a child at home, preventing falls and preventing drowning deserve the same rank of attention as safety on the road 67.
How strong is the evidence
Mortality statistics themselves are hard evidence. They come from death registration that covers the whole population. There is no sampling error in that sense.
The layer that says which behaviors raise risk is also fairly hard. Alcohol, speeding, and distraction appear again and again beside fatal crashes in police crash records 2.
Stories such as "psychological stress caused reckless driving in the pandemic" are still hypotheses. The piece says plainly that the question has not been settled 1.
Helmets and risk compensation are an old difficulty in observational evidence. People who wear helmets may already be more cautious, and that confound is hard to remove. The systematic review itself grants the point 5.
The claim that defensive driving lowers one person's risk of death has almost no randomized trial behind it. No one could really run that trial. It belongs with claims that have a clear mechanism, indirect data, and no experimental test.
Details such as "look left, then right" come from the shape of crashes and from experienced drivers. The direction is reasonable. The precision is unknown 2.
The World Health Organization's direction for action rests on policy and engineering evidence. A United Nations resolution it cites writes "cut road deaths and injuries in half before 2030" as a global goal 8.
What you can do
A not-to-do list from that public discussion, carried into ordinary life:
- Even if you believe you are under the legal limit, stay away from the wheel. One public account states that if there is alcohol in the body, the person does not drive 2.
- No speeding. Keep the phone where it cannot be reached while the car is moving 2.
- When sleep is short, move the appointment or change how you travel. That item sits on the same not-to-do list as driving after drinking 2.
- At an intersection, scan left first and then right before you proceed 2.
Changes to the setting that are worth the effort:
- If an older adult lives in the home, check the floors, the lighting, the handrails, and traction in the bathroom. Falls take more than 680,000 lives a year 6.
- If a young child lives in the home, treat supervision around water as something that cannot share attention. Swimming and water-safety skills are better learned early 7.
- Wear a helmet when cycling. There is no need to worry over the claim that a helmet makes a rider take more risks 5.
Questions worth raising with a doctor:
- Could a medicine I take, or short sleep and daytime drowsiness, affect my driving? 4
- My sleep has been poor for a long time, I am drowsy in the day, and the commute is a long drive. Is it time to rearrange that? 4
- Are my drinking and the way I commute something to talk through together? 2
Common misconceptions
Misconception one: flying is more dangerous than driving. An air disaster is severe enough to frighten people, but it is extremely rare. Multiply the two dimensions and the death toll on the road is higher by orders of magnitude 3.
Misconception two: fatal crashes happen mainly on freeways. In the 2017 figures, freeways were only 18% of fatal crashes. Two-lane country roads were 38% 2.
Misconception three: a helmet makes a rider take more risks and is therefore more dangerous. A 2019 systematic review did not find support for that "risk compensation." It found instead that helmeted riders behaved more safely 5.
Misconception four: an accident is luck, and nothing can be done about it. Alcohol, speeding, distraction, and fatigue — the factors strongly tied to these deaths — are all matters of behavior 2.
Misconception five: drowning and falls are not "big" causes of death, or they are someone else's problem. Drowning takes about 300,000 lives a year. Falls take more than 680,000, and the World Health Organization explicitly lists them as underestimated 67. The death burden of falls sits on older ages. Preventing a fall counts when it happens before a fracture 6.
How it connects to your blood work
No blood test directly measures accident risk. The causes of accidental death sit in behavior and in the surroundings, not in a lab value. The liver enzymes below also cannot be used to judge that risk.
The only overlap is indirect. Alcohol appears again and again in the fatal-crash figures 2, and long-term drinking can leave traces on a blood panel. A drinking history is clarified by asking. It is not read backward from an enzyme.
- Gamma-glutamyl transferase (ggt): an enzyme sensitive to alcohol and to stress on the liver, one trace that long-term drinking may leave.
- Alanine aminotransferase (alt) and aspartate aminotransferase (ast): enzymes that rise when liver cells are injured. They can serve as a cross-check.
On the subject of driving safety, that is as far as a blood report goes. An enzyme reading does not measure accident risk. Drinking is clarified by the history, not by the number.