Adverse Childhood Experiences

What it is

The instrument is deliberately simple. Ten yes or no items, summed into a single score, designed to make a large population question answerable at scale. Felitti and colleagues published the founding study in the American Journal of Preventive Medicine in 1998, working from a clinic population, and the paper established that the count of childhood adversities tracked adult health outcomes across the population studied.

That is a population-level finding, and as a population-level finding it has held up. The dispute concerns a different use. Once the score is put in front of an individual patient, it is being asked to do something the founding design did not test: to predict, for this person, what will happen next.

This vault's record of the concept comes by way of a popular account, and the note separates three things carefully. The published research is Felitti and colleagues in 1998 and Kennedy and colleagues in 2016. The interviews with Felitti, Robert Anda and Gabor Maté supply the chapter's framing and its confidence, but they are not the source of the instrument. The move from childhood adversity to the contemporary attention crisis is the author's own synthesis.

In effect

The instrument has travelled a long way from a 1998 epidemiological paper. It now appears in clinical screening programmes, the most prominent being the California ACEs Aware initiative launched under Nadine Burke Harris as State Surgeon General, which trains and reimburses clinicians to screen patients for ACE scores.

That programme is where the dispute lives. Paediatric and public health researchers have argued since 2021 that the instrument lacks the predictive precision that individual-level screening requires, and the debate is live scientific disagreement rather than settled practice. The chapter this vault ingested presents the framework without indicating that any of it is disputed, which matters because the contested use is precisely the use the chapter puts it to.

Both halves have to travel together. The population-level association stands. On the individual-screening question, our reading is that the critics have the better of the argument.

What it does not say

It does not say that childhood adversity does not matter. The population-level association is not what is in dispute, and nothing here should be read as minimising it.

It does not say the instrument can tell an individual what will happen to them. A retrospective ten-item count is a blunt measure, and its use as a screening or triage tool for a particular patient is the part under challenge.

It does not escape the problems its method carries. The instrument is a retrospective self-report of childhood events, which places it squarely inside both the self-report problem and the general replication concerns that apply to work of this kind.

It does not support the single most quotable figure in the popular retelling. That number comes from one clinic's own patient population, is cross-sectional and keyed to diagnoses already recorded at that clinic, has no comparison population outside the clinic, and is not a prospective finding. It should not be repeated as a general risk multiplier, and it is not repeated here.

It does not follow from the strongest adjacent evidence either. Kennedy and colleagues found in 2016 that children from Romanian orphanages were four times more likely to have serious attention problems. That is a severe deprivation population, which is what makes it a real finding and also what limits how far it generalises to ordinary childhood adversity.

This is clinical material about a vulnerable population. Nothing on this page is reader-facing without a credentialed reviewer, and nothing on it is guidance.

---