Adverse childhood experiences and mental health in young adultsa longitudinal survey

Think back to the stories that shape us. Not the bright, postcard moments, but the hard ones — the nights you didn't feel safe, the upheavals you didn't choose, the harms that arrived before you had a say. Psychologists call them adverse childhood experiences, or ACEs. We've known for years that they are tied to worse mental health. What's been less clear is what happens right as teens step over the threshold into adulthood. Is that bridge a place where early adversity shows up in a new way? Schilling, Aseltine, and Gore set out to capture that moment in motion. In the late 1990s, they drew a probability sample of high school seniors from nine public schools in seven Boston-area communities chosen for their socioeconomic and ethnic diversity. They interviewed students in 1998, then returned in 2000 — a clean two-year window that spans graduation, first jobs, college, and all the churn of launching. Out of an original frame of one thousand five hundred seventy-eight seniors, one thousand one hundred forty-three were interviewed in person at baseline, with another sixty-six by phone. Two years later, one thousand ninety-three of those original respondents were re-interviewed. Not bad for a community study tracking busy eighteen-year-olds into their early twenties. The team asked about ACEs at the first interview — parental separation, time spent removed from home, a parent's drinking or drug problems, physical abuse or assault, sexual abuse or assault, serious neglect, witnessing a serious injury or murder, being threatened with a weapon, or held captive. Each ACE counted as a yes or no. They also built a cumulative ACE score by summing the ten items and, to avoid a few extreme values warping the analysis, they truncated that sum at five. Two years later, they measured three outcomes and put them on the same scale by standardizing to a mean of zero and a standard deviation of one: depressive symptoms using a modified twelve-item Center for Epidemiologic Studies Depression scale, frequency of drug use in the past year, and frequency of antisocial behavior like fighting or theft. They were careful about the guts of the design. Attrition wasn't random — roughly thirty percent of the baseline sample didn't make it to the follow-up, and it hit groups differently. Whites had about twelve percent attrition, while Hispanics were closer to thirty-one percent, with Blacks in between. The team ran logistic regressions to probe who was lost and controlled for socioeconomic status using parents' education, alongside gender and race or ethnicity, in all outcome models. And because running lots of tests can make chance look like signal, they used a simple, sensible rule: when they tested, for example, ten ACEs against three outcomes across racial groups, the pattern of significant findings had to exceed what you'd expect from noise before they interpreted individual coefficients. Before we get to the punchline, it helps to know who experienced what. In this urban, socioeconomically disadvantaged cohort, parental separation topped the list at twenty-seven point five percent. The gender splits were sharp in places. Girls were far more likely to report sexual abuse or assault — about eleven point seven percent versus one point four percent in boys. Boys, in contrast, more often reported witnessing a serious injury or murder, nineteen point seven percent compared with thirteen point eight percent in girls, and were more likely to say they had been physically assaulted, ten point eight percent versus four point five percent. Being threatened with a weapon or held captive also skewed male in this sample, fifteen point five percent for boys and four point three percent for girls. Race and ethnicity told a different story about exposure. About one in six White students reported a parent with a drinking or drug problem — higher than both Black and Hispanic peers. Witnessing a serious injury or murder cut the other way: that was reported by about a quarter of Black and Hispanic youths, roughly double the White rate of twelve point nine percent. Hispanics also had relatively high rates of being threatened with a weapon, held captive, or kidnapped, around sixteen percent compared with eleven point two percent for Whites. Already, you can hear the complexity: different groups face different kinds of adversity, and any single label like "high risk" hides a lot of texture. Now, the two-year outcomes. The short version is simple: ACEs cast a wide shadow. Having any ACE at all predicted more depressive symptoms, more drug use, and more antisocial behavior two years later, even after controlling for gender, race or ethnicity, and socioeconomic status. The size of those "any ACE" effects was remarkably similar across domains — roughly a quarter of a standard deviation for depression, about zero point twenty-six, for drug use, zero point twenty-seven, and for antisocial behavior, zero point twenty-two. That tells you there's a general strain of vulnerability running through these very different outcomes. It's not just sadness or just acting out; it's a broad shift in how young people are doing. The pattern got stronger when you considered accumulation. Each additional ACE moved the needle in a nearly linear way. The slope of that line was close across the three outcomes — about zero point sixteen standard deviations per ACE for depression, zero point seventeen for drug use, and zero point seventeen for antisocial behavior. Put that in everyday terms: stack up four ACEs, and you're right around a one standard deviation increase on each of these problems. That's not a footnote; that's meaningful change in everyday life — in mood, in risk-taking, in behavior that can draw the attention of schools, police, or employers. Zooming in on specific experiences, a few stood out as particularly potent. Sexual abuse or assault was linked to a large jump in depressive symptoms two years later — on the order of zero point seventy of a standard deviation. Being "sent away from home," which can capture removal due to danger or instability, packed an even bigger punch for behavior: about zero point eighty-one for drug use and zero point ninety-eight for antisocial behavior. Physical assault also showed strong associations, especially for drug use, and witnessing serious violence registered as well. The point isn't to rank traumas but to see a consistent picture — a range of adversities, from chronic family strain to acute violence, forecast difficulties across emotional and behavioral domains as teens become young adults. And that's the average picture. The study got more interesting when the researchers asked if the link between early adversity and later outcomes looks the same for girls and boys. Not always. One place where the gap was striking was sexual victimization. For drug use at follow-up, the coefficient for boys who reported sexual abuse or assault was about one point eighty-five, compared to zero point twenty-nine for girls. That doesn't mean girls weren't affected — they were — but the signal was far stronger among boys in this sample. On antisocial behavior, the cumulative burden of ACEs was also more tightly chained to outcomes for boys: the slope of the sum-of-ACEs line was around zero point twenty-five for boys, versus zero point zero nine for girls. In other words, stacking ACEs seemed to push boys' antisocial behavior up more steeply over that two-year span. Race and ethnicity layered on another pattern. When Schilling and colleagues split the sample into White, Black, and Hispanic groups, the mental health impact of ACEs often looked larger among White youths. Take depressive symptoms: the cumulative ACE effect for Whites was about zero point twenty-two, while it was closer to zero point zero three for Blacks and zero point zero five for Hispanics. Or think about that "four ACEs equals roughly one standard deviation" rule of thumb for drug use. Among Whites, that slope got you there. For Blacks and Hispanics, the authors estimated you'd need more than twenty ACEs to reach the same increase — an extrapolation that dramatizes how much flatter the line was for these groups in this dataset. No one is saying twenty ACEs happen; the point is that comparative vulnerability per ACE looked higher for Whites here. That finding will raise eyebrows, and it should. The authors offer a few possibilities — ideas like differential coping, socialization, or what some call "steeling" effects, where kids exposed to frequent stressors develop some resilience to their psychological impact. It could also reflect measurement issues, differences in which adversities were most common, or the fact that risk piles up in different ways across communities. The pattern wasn't uniform across every ACE and every outcome — there were exceptions — but the overall tilt was clear enough to warrant attention. Let's pause for a moment on what this study can and cannot tell us. It's prospective, which is a real strength — ACEs were recorded before the outcomes, not remembered after the fact — and it covers a crucial developmental window. The outcomes were standardized and thoughtfully constructed, and the models consistently controlled for gender, race or ethnicity, and socioeconomic status. The team took attrition seriously and tried to keep multiple comparisons from fooling them. That's solid epidemiological homework. At the same time, it's not an experiment. We can't randomly assign adversity. Some unmeasured factor — say, early temperament or neighborhood changes — could shape both exposure to ACEs and later behavior. The sample is urban and socioeconomically disadvantaged by design, which is the point for understanding risk in those communities, but it limits generalizability. Attrition was higher for Hispanics and Blacks than for Whites, which could bias subgroup estimates despite careful controls. And the two-year window is a strength for timing, but it can't tell us whether an ACE at age six versus age sixteen matters in different ways. So, what do we take away? First, the link between early adversity and near-term mental health in the launch to adulthood is robust and broad. It shows up across depression, drug use, and antisocial behavior. Second, accumulation matters. The more ACEs, the steeper the climb in risk, with roughly four ACEs mapping onto a one standard deviation uptick in these outcomes. Third, context matters. Sexual victimization in particular carried heavier consequences for boys in this sample, and White youths often showed stronger per-ACE associations than Black or Hispanic peers, even as the kinds of adversity they faced differed. Those are not just statistical curiosities. They flag places where prevention might bite hardest — family and community supports that reduce the pileup of adversities, trauma-informed approaches that recognize how sexual violence can reverberate differently for boys, and interventions tuned to the specific patterns of exposure and vulnerability in different communities. As Schilling and colleagues argue, these are public health problems with public health solutions. If you're listening and thinking, but isn't resilience real? Yes. Many young people with ACEs will not develop depression, will not use drugs, and will not offend. This study doesn't negate that. What it does is shift the odds, and it shows that those odds are already shifting by the time someone is twenty, not just decades later. That's early enough to matter. Early enough to act. And maybe that's the deepest point here. The transition to adulthood is not a reset button. It's a hinge. What came before still bends the door. But hinges can be reinforced. With the right supports — fewer early adversities, more stable homes and schools, better mental health care — you change the angle of that swing. The signal in this study is clear. The earlier we shore up the hinge, the quieter the squeak two years down the line.