The ACE study: what 30 years of adverse-childhood-experience research has shown

In 1995, physicians Vincent Felitti and Robert Anda surveyed roughly 17,000 adult patients of Kaiser Permanente's San Diego clinic on their childhood experiences and current health. The instrument they developed — the Adverse Childhood Experiences (ACE) questionnaire — asked about ten categories of childhood adversity: physical, sexual, and emotional abuse; physical and emotional neglect; household substance abuse, mental illness, parental separation, domestic violence, and incarcerated household members.

The findings, published in 1998, were striking. Each additional ACE was associated with elevated risk for a wide range of adult outcomes — depression, suicide attempts, substance abuse, obesity, cardiovascular disease, early mortality (Felitti et al., 1998).

The study became one of the most influential public-health findings of the late twentieth century. The follow-up research has both confirmed and complicated the original.

1. The basic findings replicate

Hundreds of subsequent studies have confirmed the dose-response relationship. People with higher ACE scores show elevated risk for a wide range of adult problems. A 2017 meta-analysis aggregated 96 articles and confirmed the basic pattern across populations and outcomes (Hughes et al., 2017).

The ACE-to-outcome correlations are statistically robust. The effect sizes for severe outcomes (suicide, substance use, severe mental illness) are moderate to large.

2. The causal interpretation is more complicated

The ACE study established correlation. Whether childhood adversity causes adult problems through biological pathways (trauma → HPA-axis dysregulation → disease) or through intermediary mechanisms (childhood adversity → poverty → adult health risks) has been harder to pin down.

Three confounds complicate the simple causal story:

Genetic confounding. Parents who are depressed, alcoholic, or imprisoned share genes with their children that may independently predict adult mental health. Studies controlling for genetic confounding show smaller ACE effects than the original literature, though still real (Baldwin et al., 2021).

Socioeconomic confounding. Adverse childhood experiences cluster with poverty, which independently predicts adult outcomes. Controlling for childhood SES reduces ACE coefficients substantially.

Recall bias. ACE scores are based on adult self-report. People with current mental health problems may recall more childhood adversity. The longitudinal studies that measure ACEs prospectively in childhood show smaller adult-outcome associations than retrospective studies.

3. The interpretation in practice

The current sophisticated reading: ACEs are real risk factors for adult problems, but not deterministic ones, and their causal role is partial. Most people with high ACE scores do not develop the worst outcomes; many people with low ACE scores do.

The original ACE framework can both undersell resilience (most high-ACE adults are functional) and overstate causation (some adult outcomes attributed to ACEs would have occurred anyway).

This isn't a critique of public-health policy that emphasizes early childhood. It is a flag against the popular framing that "childhood determines your adult life" — which is stronger than the data supports.

4. The intervention question

If ACEs predict adult problems, do interventions that reduce ACE exposure improve adult outcomes? The longitudinal evidence here is thinner. Prevention programs (home-visiting, parental support) reduce ACE incidence; their long-term effects on adult outcomes are positive but typically smaller than the correlational ACE literature would predict (Olds et al., 2007).

This pattern — interventions producing smaller effects than the original correlational studies suggest — is common in social science. It suggests substantial non-causal variance in the original associations.

5. The honest summary

ACEs are a useful, validated screening tool for elevated risk. They are not a comprehensive theory of adult outcomes. The original 1998 study established an important correlation that has held up; the strong popular version overstates the causal role.

For an individual reflecting on their own ACE history: high scores indicate elevated population-level risk for various adult problems, not a personal trajectory. Resilience is the rule, not the exception, even at high ACE scores. The framework is useful for clinical screening and policy targeting, not for individual prophecy.