A study that followed elderly people in Japan for 28 years found that they were remarkably accurate in their life expectancy estimates. Those who expected to live longer tended to do so.
However, their predictions revealed one striking flaw: many underestimated their remaining life. Among the participants whose results were available, the researchers report that approximately 59 percent lived beyond their life expectancy.
A team of researchers led by sociologist Shohei Okamoto of the University of Tsukuba analyzed data from the Japan Study of Aging and Health Dynamics—a nationally representative survey of adults aged 60 and older across Japan.
In 1996, 2,447 participants were asked what age they expected to live. Of these, 393 responded that they did not know. The remaining 2,054 people reported ages ranging from 63 to 120 years.
Over the next 28 years, 1,514 of those who provided life expectancy estimates died. Approximately 98 percent of death dates were obtained from official residence records or family reports, allowing the team to compare predicted and actual life expectancies.
People who expected to live longer tended to live longer. This relationship remained even after the researchers controlled for age, gender, socioeconomic characteristics, and health-related factors.
So, can simply expecting a long life help you live longer?
This study, available as a preprint on the medRxiv platform and not yet peer-reviewed, cannot answer that question. Because it was observational, it does not prove that expecting a longer life extends a person’s lifespan or that positive thinking helps people stay alive.
Instead, people may know things about their health, habits, or family history that influence both their prognosis and life expectancy. For example, a person who feels healthy may expect to live longer, and therefore have a higher likelihood of doing so.
Although the researchers accounted for a wide range of health-related factors, unaccounted characteristics could still influence both expected and actual life expectancy.
The predictions were also far from personalized countdowns. In the team’s statistical model, each additional year of life expectancy was associated with only approximately 0.12 additional years of expected survival.
Errors varied across groups. Women and people with higher levels of education were more likely to live longer than expected, while those with poorer health were less likely to do so.
One possible explanation lies in a well-known number: 80.
Many survey participants chose 80 as the age they expected to reach. This was close to the average life expectancy for both sexes in Japan in 1996, suggesting that people may have based their predictions on a generally accepted population average.
However, life expectancy at birth does not correspond to the life expectancy of a person who has already reached old age. A person who lives to age 70 has overcome the early life risks that affected the average life expectancy calculated at birth.
Therefore, an estimate based on a person’s current age and gender may be more relevant than a single figure for the entire population. A broad average may also conceal a trend toward longer life expectancy for women and people with higher levels of education. However, the study does not establish that the use of the value 80 was the cause of the error.
The study’s results regarding education provided another unexpected twist. Participants with more years of education were less likely to choose ages divisible by 5, such as 75, 80, or 85. But less rounded answers did not make their predictions more accurate: they also often underestimated their life expectancy.
Miscalculating life expectancy can have consequences beyond simple curiosity about the future.
“Underestimating life expectancy can lead to insufficient savings for retirement, while overestimating it can lead to excessive frugality or even poverty,” the researchers write.
The research team suggests that more accurate information about remaining life expectancy by age and gender could help people make more prudent financial decisions.
This study has important limitations. Participants estimated their life expectancy only once, in 1996, so the researchers were unable to track how a diagnosis, medical advances, or the death of a loved one might have changed their expectations.
The data also lacked detailed information on their parents’ life expectancy. And because the study only included older adults in Japan, the findings cannot be automatically applied to younger people or populations in other countries.
These results do not indicate that people can extend their lives simply by expecting to live longer. But our predictions may contain clues gleaned from what we know, consciously or not, about our health and circumstances.
However, most participants overlooked one important possibility: their future may extend longer than they imagined.
About the Author
