After 2014, U.S. life expectancy fell for 3 straight years. This striking trend is not associated with other wealthy countries in the world and has given rise to a cottage industry of speculation on causes, with varied social, cultural, and political actors making use of the findings for preferred narratives. Some of this speculation arose after Princeton economists Anne Case and Angus Deaton coined the term “deaths of despair,” an easily misunderstood phrase. Case and Deaton used the term to refer to fatal drug overdoses, alcohol-related diseases, and suicides. “We think of all these deaths as suicides, by a very broad definition,” these economists have written,“and we attribute them to a broad deterioration in the lives of Americans without a college degree who entered adulthood after 1970.”
In late 2019, a National Institute on Aging-supported review offered a comprehensive examination of falling U.S. life expectancy. This study used data from the CDC, National Center for Health Statistics, and U.S. Mortality Database to trace life expectancy trends over a longer time frame and analyze mortality rates for particular age cohorts. It paints a complicated picture of poor U.S. mortality trends, which are not driven just by our polysubstance epidemics, and a decidedly unhappy one. The authors write, “According to one estimate, if the slow rate of increase in US life expectancy persists, it will take the United States more than a century to reach the average life expectancy that other high-income countries had achieved by 2016.”
What did this study find about U.S. life expectancy?
This study found that U.S. life expectancy, between 1959 and 2016, rose to 78.9 years from 69.9 years. But, following 2014, life expectancy began declining. “A major contributor,” this study’s authors write, “has been an increase in mortality from specific causes (eg, drug overdoses, suicides, organ system diseases) among young and middle-aged adults of all racial groups, with an onset as early as the 1990s and with the largest relative increases occurring in the Ohio Valley and New England.”
This study also finds that:
- Mortality increases are concentrated among Americans at “midlife”, or those between 25-64 years of age. The all-cause mortality rate rose by 6% between 2010-2017 for Americans between these ages.
- Rising mortality between 2010-2017 led to 33,307 excess deaths.
- More than 32 percent of these deaths happened in Kentucky, Indiana, Ohio, and West Virginia—or the Ohio Valley states. The upper New England states also had some of the largest mortality increases, though they account for a lower share of the overall total.
- These rising mortality figures are found among all racial groups in the U.S.
These trends do not reflect significant changes in violent crime. The U.S. is much less violent than it used to be, and although it still has higher violent crime rates than other rich countries, things have gotten much better in the last 30 years. Other medical conditions, such as infectious diseases and cancers, were also not behind these changes, as outcomes for certain medical problems actually improved.
These researchers offer some answers. They first examine life expectancy from 1959 to trace when it started to change and, recognizing that fatal overdoses started to rise in the nineties, next look at cause-specific mortality between 1997-2017. They note that U.S. life expectancy first diverged from other wealthy countries in the eighties. Our rate of increase in life expectancy was not as fast as in other wealthy countries. In the nineties, cause-specific mortality rose for Americans in midlife. By 1998, the U.S. fell below the life expectancy average of other rich countries. Then U.S. life expectancy increases ended in 2010 and started falling in 2014. Fatal overdoses account for an important share of this increase — they rose by almost 387% between 1999-2017. Deaths from alcoholic liver disease rose by around 41 percent, and suicides by around 38 percent. But other causes were, or still are, at play, too. Hypertensive disease rates, for example, also rose — by about 79 percent. And obesity death rates were higher, rising by 114 percent. The authors observe that one study found that, for women, cardiovascular and respiratory conditions accounted for nearly as many deaths as fatal overdoses.
What’s going on?
It’s a tough question. The authors write, “The largest relative increases in midlife mortality occurred among adults with less education and in rural areas or other settings with evidence of economic distress or diminished social capital.” But these observations aren’t necessarily explanatory factors in our life expectancy decline. This review considers different explanations and evidence in their favor. Isn’t it really just our drug epidemics — the rising death toll from heroin in the sixties and seventies, cocaine in the eighties, and then the three-stage opioid epidemic of prescriptions, heroin, and synthetics? Well, the authors say, this is a significant part of the story, but far from complete. Suicides and alcohol-related liver diseases have also contributed substantially and the timing is off because our life expectancy divergence started in the eighties, “and involved multiple diseases and nondrug injuries.” They also point to 2 studies suggesting that only 15 percent of our life expectancy divergence can be explained by fatal overdoses.
What about smoking? Americans smoke less than we used to, but smoking more decades ago could still kill more people today. And Americans are often more obese than their peer country counterparts, so could that account for the differences? This review points to research on other countries, like Australia, that resemble the U.S. in smoking and obesity but haven’t followed our marked life expectancy divergence. Health care? The U.S. famously spends more to cover fewer people than its peers, and Americans also face higher costs of care, but the authors note that this wouldn’t account for why we have more deaths from some diseases and not others, and from suicides or obesity-related deaths “which originate outside the clinic.” Might the problem be “deaths of despair” after all, then, or a large increase in psychological distress? There’s “inconclusive evidence” that depression and anxiety, which can also harm physical health, rose over the relevant time period, and, the authors say, it’s also hard to figure out the link between conditions like depression and all the rising specific causes of death.
This review continues like this, carefully weighing the relative evidence about which factors make sense and under which conditions. For socioeconomic status, the authors say that the timing works, because the U.S. experienced pronounced economic churn in the eighties and nineties, with the most economically impacted areas and people also experiencing the largest mortality rises. But then localized data on income and employment don’t always align with the geographic and demographic trackers. What’s the answer? It’s likely that there isn’t a particular cause, or at least one we can decisively identify right now. The authors say these various possible causes “are not independent and collectively shape mortality patterns.” They call for the accumulation of more rigorous evidence, gathered from machine learning, migration research, and cohort studies, and interdisciplinary research, given how many different areas may contribute, and attempts to answer comparative questions about why some states and regions have worse life expectancy than others, and why other rich countries do better. These are sensibly modest conclusions but gravely important — they affect our view of the most important challenge we face.
These numbers do not consist of entirely dreary findings. Life expectancy has risen in some U.S. states, deaths from some causes, like violence, have declined, and black men’s life expectancy has improved. Deaths from screenable prostate, breast, cervical, and colon cancer are down, as are cardiovascular diseases. Reports say that the cancer death rate has dropped 29 percent, leading to about 3 million fewer deaths relative to a steady mortality rate. But supporting data has been stubbornly consistent. University of Pennsylvania researchers have found that outside of metropolitan areas following 1990, mortality increased even though it declined for many other demographic groups.1
Many death certificates tell a different story. Around 75,000 people died in the U.S. in 2017 from liver disease and alcohol-related conditions, a steep rise from 1999, when 36,000 died from those causes. Women used to die at lower rates from these conditions, but that gap has closed.2 As reported first by Case and Deaton and reiterated in their upcoming book, rising morbidity and mortality among whites due to accidents, drug overdoses, alcoholism, liver disease, and suicides means reduced overall longevity. These deaths are alarming businesses, too, as there are more suicides in the workplace than in the past.3 I expect that suicides, as well as overdose deaths, are undercounted. The most recent Florida data suggests that 1 out of 3 opioid overdose deaths, and an even greater share of cocaine deaths, are not reported.
- Elo, I.T., Hendi, A.S., Ho, J.Y., Vierboom, Y.C., Preston, S.H. (2019). Trends in non-Hispanic white mortality in the United States by metropolitan-nonmetropolitan status and region, 1990–2016. Pop Dev Rev. 45(3), 549-583. doi: 10.1111/padr.12249. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/31588154
- National Institute on Alcohol Abuse and Alcoholism (NIAAA). (2020). Alcohol-related deaths increasing in the United States [news release]. Retrieved from https://www.niaaa.nih.gov/news-events/news-releases/alcohol-related-deaths-increasing-united-states
- Bureau of Labor Statistics. (2019). National Census Of Fatal Occupational Injuries in 2018 [news release]. Retrieved from https://www.bls.gov/news.release/pdf/cfoi.pdf
1. Woolf, S.H., Schoomaker, H. (2019). Life Expectancy and Mortality Rates in the United States, 1959-2017. JAMA 322(20), 1996-2016. doi: 10.1001/jama.2019.16932. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/31769830