Japan Study Finds People Often Underestimate Their Life Expectancy

How old do you believe you will live to be?

The response might reveal more than you think.

A study following older people in Japan for as long as 28 years indicates that individuals can judge their own longevity with surprising accuracy. In general, those who anticipated living longer went on to do so.

Their estimates did, however, reveal a notable tendency: many people underestimated the time they still had. Of the participants for whom outcomes were available, the researchers found that roughly 59 percent lived beyond the age they had expected to reach.

Japan study links expected lifespan with longevity

The research team, led by University of Tsukuba social scientist Shohei Okamoto, examined information from the Japanese Aging and Health Dynamics Study. This nationally representative survey covered people aged 60 and over throughout Japan.

In 1996, researchers asked 2,447 participants the age they thought they would live to. Of this group, 393 did not know. The other 2,054 respondents supplied an age between 63 and 120.

During the following 28 years, 1,514 people who had provided an estimate died. Official resident records or reports from relatives provided about 98 percent of the dates of death, enabling the researchers to compare expected lifespans with actual ones.

Participants who believed they had more years remaining typically lived longer. This relationship persisted when the team adjusted for age, sex, socioeconomic characteristics and health-related factors.

Can expecting to live a long life, then, actually help a person live longer?

The study, which is available as a preprint and has not yet undergone peer review, cannot resolve that question. As an observational study, it cannot demonstrate that anticipating a longer life increases lifespan, or that positive thinking keeps someone alive.

Rather, people may possess knowledge about their health, routines or family history that influences both their expectation and their longevity. For instance, someone who feels well may predict a longer life and may also be more likely to have one.

While the researchers adjusted for several health factors, other unmeasured traits may still have influenced both anticipated and actual lifespan.

Life expectancy estimates were not personalised forecasts

The predictions were not precise personal countdowns either. In the team’s statistical model, every extra year a person expected to live corresponded with only around 0.12 extra years of estimated survival.

The degree of error also differed between groups. Women and those with more years in education were more likely to live past their expected age, whereas participants in worse health were less likely to do so.

One potential reason centres on a familiar figure: 80.

A large number of participants selected 80 as the age they expected to attain. Since this was close to Japan’s combined-sex life expectancy at birth in 1996, it suggests that people may have based their forecasts on a well-known population average.

Yet life expectancy at birth differs from the remaining life expectancy of a person who has already reached older age. Someone who has made it to 70 has already avoided the early-life risks included in an average calculated from birth.

A calculation that takes account of a person’s present age and sex may consequently be more useful than one figure for the entire population. A general average may also hide the pattern that women and more highly educated people tend to live longer. The study does not, however, show that using 80 as an anchor caused these inaccuracies.

The education result included a further detail. People with more years of education were less inclined to give ages divisible by 5, including 75, 80 and 85. Providing less rounded figures did not improve the accuracy of their forecasts, though: they too frequently underestimated how long they would live.

Why underestimating life expectancy matters

Getting longevity wrong may have implications that extend beyond mere curiosity about the future.

"Underestimating life expectancy may lead to insufficient retirement funds, while overestimating it could result in unnecessary frugality or even poverty," the researchers write.

The team proposes that more clearly presented information on remaining life expectancy, based on a person’s current age and sex, might assist people in making better financial choices.

There are significant limitations to the study. Participants gave their lifespan estimates just once, in 1996, meaning the researchers could not observe how a diagnosis, medical progress or the loss of a close person could alter their expectations.

The dataset also did not contain detailed information on how long participants’ parents had lived. In addition, because the research involved only older adults in Japan, its findings cannot automatically be extended to younger people or to populations in other places.

The findings do not indicate that people can lengthen their lives simply by expecting to live longer. Yet their predictions may hold clues based on what they know, either consciously or unconsciously, about their health and circumstances.

Even so, most participants overlooked one key possibility: their future might continue for longer than they had imagined.

The study is available on medRxiv.

This article was fact-checked by Michael Irving and edited by Michael Irving. Although we take pride in our process, we are only human. If you notice an error, please tell us.