Although there has been plenty of research linking mortality to various habits and lifestyle factors, scientists at the University of British Columbia (UBC) in Vancouver, Canada — in collaboration with colleagues from other institutions — have identified the 10 factors most likely to lead to death.
The new study appears in the journal PNAS. The research aimed to better understand the stagnation in life expectancy occurring in the United States over the past 3 decades compared with other industrialized countries.
According to the study, the top three factors most closely connected to death are smoking, experiencing divorce, and having engaged in alcohol misuse.
The full list of the top 10 factors that most strongly predict death is as follows:
- smoking (currently)
- history of divorce
- history of alcohol misuse
- recent financial difficulties
- history of unemployment
- smoking (previously)
- lower life satisfaction
- never having been married
- history of using food stamps
- negative affectivity
Analyzing the data
A recent estimate of U.S. life expectancy stood at 78.6 years, while 11 other industrialized nations averaged about 82.3 years. This represented a gain of just 4.9 years for the U.S. between 1980 and 2017, while other nations added an average of 7.8 years.
Biological and medical factors are likely part of the reason, and advances in medicine that aim to address these factors continue. The roles of social, psychological, economic, and behavioral factors in life expectancy are less well understood, and researchers often study them in isolation.
This means that it is unclear which of these nonmedical factors are the strongest predictors of mortality risk — a knowledge gap the new study hoped to fill.
The study, which Eli Puterman of UBC led, included the data of 13,611 adults who took part in the U.S. Health and Retirement Study. This was a nationally representative sample of U.S. adults aged 52–104. The average age was 69.3. Scientists collected these data between 1992 and 2008 and analyzed them in relation to deaths that occurred between 2008 and 2014.
The original data did not capture all potential psychosocial factors that might be linked to mortality. For example, the researchers tracked neither food insecurity nor domestic abuse.
Of the results, Puterman says, “It shows that a lifespan approach is needed to really understand health and mortality.”
We can only understand the long lasting effect of a life event or lifestyle choice by taking the sort of view Puterman and his co-authors embrace.
“For example,” says Puterman, “instead of just asking whether people are unemployed, we looked at their history of unemployment over 16 years. If they were unemployed at any time, was that a predictor of mortality?”
“It’s more than just a one-time snapshot in people’s lives, where something might be missed because it did not occur. Our approach provides a look at potential long-term impacts through a lifespan lens.”
– Eli Puterman