In 2015, a blockbuster study came to a surprising conclusion: Middle-aged white Americans are dying younger for the first time in decades, despite positive life expectancy trends in other wealthy countries and other segments of the US population.

The research, by Princeton University’s Anne Case and Angus Deaton, highlighted the links between economic struggles, suicides, and alcohol and drug overdoses.

Since then, Case and Deaton have been working to more fully explain their findings.

They’ve now come to a compelling conclusion: It’s complicated. There’s no single reason for this disturbing increase in the mortality rate, but a toxic cocktail of factors.

In a new 60-page paper, “Mortality and morbidity in the 21st Century,” out in draft form in the Brookings Papers on Economic Activity Thursday, the researchers weave a narrative of “cumulative disadvantage” over a lifetime for white people ages 45 through 54, particularly those with low levels of education.

Along with worsening job prospects over the past several decades, this group has seen their chances of a stable marriage and family decline, along with their overall health. To manage their despair about the gap between their hopes and what’s come of their lives, they’ve often turned to drugs, alcohol, and suicide.

Meanwhile, gains in fighting heart disease have stalled, and rates of obesity and diabetes have ploddingly climbed.

So the rise in mortality for white mid-life people in America since the late 1990s is actually the final stage of a decades-long process. “It’s about the collapse of white middle class,” said Case. Here are the five big takeaways from the researchers’ new opus.

1) Suicides, alcohol, and drug overdose deaths have gone up across the entire country. (Read: It’s not just a rural problem.)

 Brookings

“Deaths of despair” — or suicide, alcoholism, and drug overdoses, particularly from opioid painkillers — are a growing problem for midlife white people.

As you can see on the left-hand map, the epidemic started in the Southwest. Now it’s “country-wide,” the study authors write, and the increase can be “seen at every level of residential urbanization in the US.” So it’s not just a rural problem or an urban problem — it’s both.

The crisis is particularly acute among middle-aged whites. “The deaths of despair come from a long-standing process of cumulative disadvantage for those with less than a college degree,” Case and Deaton write. “The story is rooted in the labor market, but involves many aspects of life, including health in childhood, marriage, child rearing, and religion.”

 Brookings

In an interview, Deaton explained, “The cohort that entered the labor market in the ’70s on down, their jobs earnings and prospects are worse. That affected their marriage prospects. Marriages got screwed up. They had children out of wedlock. Their pain levels [are] going up.” All that contributes to the deaths of despair.

The study authors don’t see the opioid supply as the fundamental factor here, but “prescription of opioids for chronic pain added fuel to the flames, making the epidemic much worse than it otherwise would have been,” they wrote.

The impact of rising deaths of despair on overall mortality was masked until the late 1990s by the decline of heart disease deaths. But recently that has changed too.

2) Deaths from chronic diseases such as diabetes have been rising

County-level mortality from diabetes, urogenital, blood, and endocrine diseases between 1980 and 2014. You can see these trending up all over the country. JAMA

Progress against mortality from heart disease has slowed and stopped, and deaths from cancer, which had been on a steady decline, are also stagnating in this group.

Meanwhile, other chronic diseases have continued to rise in the whole population, particularly among middle-aged white people. Diabetes’ prevalence has exploded in the US over the past 20 years. Nearly 30 million Americans live with the disease today — more than three times the number in the early 1990s. And this may be a major, underappreciated driver of the mortality trend.

3) The least-educated Americans are suffering the most

 Brookings

The rise in mortality among middle-aged whites is largely being driven by those with a high school degree or less. The researchers find that the gap in mortality between more and less educated is increasing, while mortality is also rising for those without a college degree and falling for those with a college degree.

“It looks like there are two Americas,” Case said. “One for people who went to college and one that didn’t.”

The middle-aged whites with less than a bachelor’s degree saw “progress stop in mortality from heart disease and cancer, and saw increases in chronic lower respiratory disease and deaths from drugs, alcohol, and suicide,” the researchers write.

Why education is such an important health indicator is difficult to untangle, Case added. “But when you think about what happens when industries pull out of towns, the tax base implodes, schools [are] not well funded, and the death spiral continues.”

In the past, people with low levels of education could get a job in a factory and work their way up the chain of command. “You could graduate high school, work at Bethlehem Steel, get more money every year as you get more experienced,” Deaton said, “and turn yourself into one of the famed blue-collar aristocrats of the 1970s.” Now, he added, “There’s a feeling that life has gone, and remainders of that life are getting less and less for each generation.”

To be clear, the study authors don’t buy the idea that one’s income relative to what one expected is influencing mortality. Rather, “It’s the life you expected to have relative to your father or grandfather — it’s just not there anymore,” Deaton said.

4) Other nonwhite racial groups aren’t experiencing the same mortality uptick — so it’s not just about income

 Brookings

As you can see here, mortality for middle-aged black people converged with mortality for middle-aged white people with low levels of education in the late 2000s (though the white population overall is still doing better than African Americans). Meanwhile, mortality rates among Hispanics continued to fall.

These other racial groups aren’t necessarily doing any better economically than their white counterparts, which is part of the reason Case and Deaton don’t accept a simple income explanation for the death uptick.

“It is possible that it is not the last 20 years that matters, but rather that the long-run stagnation in wages and in incomes has bred a sense of hopelessness,” they write. “But … even if we go back to the late 1960s, the ethnic and racial patterns of median family incomes are similar for whites, blacks, and Hispanics, and so can provide no basis for their sharply different mortality outcomes after 1998.”

Instead, the researchers think the fact that the overall life prospects for white middle-aged people without a BA have declined over time — they are doing worse than their parents on both a personal and professional level, and probably worse than they expected — is nudging mortality downward. This regression is different from the story of progress in the African American community, for example. Here’s Case and Deaton again.

The historian Carol Anderson argued in an interview in Politico (2016) that for whites “if you’ve always been privileged, equality begins to look like oppression,” and contrasts the pessimism among whites with the “sense of hopefulness, that sense of what America could be, that has been driving black folks for centuries.” That hopefulness is consistent with the much lower suicide rates among blacks, but beyond that, while suggestive, it is hard to confront such accounts with the data.

5) This story is unique to the US

 Brookings

The US, particularly middle-aged white Americans, is an outlier in the developed world when it comes to this mid-life mortality uptick.

“Mortality rates in comparable rich countries have continued their pre-millennial fall at the rates that used to characterize the US,” Case and Deaton write. “In contrast to the US, mortality rates in Europe are falling for those with low levels of educational attainment, and are doing so more rapidly than mortality rates for those with higher levels of education.”

If American wants to turn the trend around, then it has to become a little more like other countries with more generous safety nets and more accessible health care, the researchers said. Introducing a single-payer health system, for example, or value-added or goods and services taxes that support a stronger safety net would be top of their policy wish list. (America right now is, of course, moving in the opposite direction under Trump, and shredding the safety net.)

They also admit, though, that it’s taken decades to reverse the mortality progress in America, and it won’t be turned around quickly or easily. But there is one “no-brainer” change that could help, Case added. “The easy thing would be close the tap on prescription opioids for chronic pain.”

Unlike health care and increasing taxes, opioids are actually a public health issue with bipartisan support. Deaton, for his part, was hopeful. Paraphrasing Milton Friedman, he said, “All policy seems impossible until it suddenly becomes inevitable.”