Do Left-Handed People Die Younger? The Nine-Year Gap That Wasn't
The short version: in 1991, a study of 987 deaths in Southern California reported that left-handers died at 66 on average and right-handers at 75. But left-handedness was suppressed in older generations, so most left-handers who die are relatively young, simply because few old people were allowed to stay left-handed. In our example with identical lifespans, that alone creates an 11-year gap. A 2023 reanalysis concluded that the changing rate of left-handedness explains the original nine-year difference.
Grandpa Frank, 81, writes with his right hand. He was born left-handed, but at his school in the 1950s that wasn't allowed: the teacher tied his left hand to the chair and hit his knuckles with a ruler until he converted. Frank says it built character. The ruler, he says, built most of it.
Mia showed him an old headline: "Left-handers die nine years younger." Grandpa read it twice, looked at his left hand with new suspicion, and asked whether the ruler had added the nine years back. Mia said that was actually a very good question.
The famous study
In 1991, psychologists Diane Halpern and Stanley Coren published a short report in the New England Journal of Medicine. They took 987 recent deaths in Southern California, asked the families whether the person had been left- or right-handed, and compared the average age at death.
- Right-handers: about 75.
- Left-handers: about 66.
A nine-year gap is enormous: about as big as the gap between smokers and non-smokers. That should have been the first warning sign.
The people who weren't left-handed any more
Here's the problem. In Grandpa's generation, many natural left-handers were forced to switch, like Frank, or quietly switched to avoid trouble. So among people born in 1910, very few were recorded as left-handed. Among people born in 1960, many more were. The share of left-handers depends on when you were born.
Now look at who dies in a given year. Most deaths are of old people, who are mostly recorded as right-handed. The left-handers in the death records are disproportionately from younger generations, because that's where most recorded left-handers are. Their average age at death is lower, even if handedness changes nothing about how long anyone lives.
A gap with nothing behind it
Let's build a year of deaths in which left- and right-handers live exactly as long as each other:
- 900 people die at 80. They were born in a generation where only 3% stayed left-handed: 27 left-handers.
- 100 people die at 30. Their generation was 12% left-handed: 12 left-handers.
- Average age at death, left-handers: 64.6. Right-handers: 75.4.
An 11-year gap, created entirely by who was allowed to stay left-handed. Nobody's life was shortened by anything except the arithmetic. This is a sampling bias built into time: the groups being compared came from different generations. A 2023 reanalysis in Archives of Public Health modelled exactly this and concluded that the changing rate of left-handedness explains the original nine-year gap. Other studies following people over time have found small differences or none.
Average age at death in our example year, with identical lifespans
How to spot a cohort trap
1. Ask whether the groups were born at different times. If one group is mostly younger, comparing ages at death will mislead you.
2. Ask whether the label changed over time. Left-handedness, diagnoses, job titles: if the share of people with a label changed between generations, the label carries an age.
3. Follow people, don't count funerals. The fair test follows similar people from birth, or from the same age, and sees who lives longer.
Where it costs you money
"Funds that hold technology stocks have shorter lives" might just mean technology funds were mostly launched recently. "Investors who use apps lose more" might mean app users are younger and newer to markets. Before believing a difference between groups, check whether the groups started at the same time and the same place.
The moral
Mia explained the arithmetic. Grandpa listened, then took a pen in his left hand, which he hasn't done for seventy years, and signed his name. It looked like a spider falling down stairs. "There," he said. "Left-handed again. If the statistics are wrong, I've lost nothing. If they're right, I'm 81 and I've already beaten them by fifteen years." Mia said that wasn't quite how it worked. Grandpa said the ruler would disagree.
Study details are summarised from Halpern and Coren (1991) and later reanalyses. The 900 and 100 death example is illustrative. Educational content, not financial advice.