It is well-established that adverse childhood experiences (ACEs) such as abuse, neglect or other forms of household dysfunction are associated with poor physical and mental wellbeing later in life. With one in two children in the UK experiencing ACEs, minimising their negative consequences has become a key target of healthcare interventions. As part of this, there have been growing calls to screen for ACEs in primary and secondary paediatric settings, in order to identify vulnerable individuals and provide targeted interventions to prevent or buffer the harmful impact of ACEs on their later health.

This shift towards routine ACE screening is driven by a large body of studies, beginning with the landmark 1998 ACE study, published in the American Journal of Preventive Medicine. In a sample of over 8,000 adults in the US who reported their exposure to 10 categories of ACEs, this study found that as the number of reported ACEs increased, so too did the risk of experiencing a range of physical and mental health problems. Similarly, those who reported four or more types of ACEs (compared to those reporting none) displayed a significantly increased risk for a range of adverse health conditions.

While these findings have been replicated at the population level in multiple studies, the extent to which these ACE ‘scores’ can accurately identify specific individuals who will have physical and mental health problems as a result of experiencing four or more ACEs remains unclear. This is because we still don’t fully understand how additional risk factors that commonly co-occur with ACEs (such as poverty) interact with ACEs to compound their negative impact, nor do we fully understand how sources of individual resilience potentially mitigate these negative consequences. For these reasons, it remains unclear which individuals in the population will go on to develop severe health problems.

This knowledge is referred to as a measure’s ‘sensitivity’ and is essential for any screening activity to be meaningful or effective in identifying the individuals who are most at risk of a poor outcome.

To test the ACE score’s ability to classify individual health risk, we have revisited the data from the original ACE study. We quantified the extent to which this score could accurately differentiate adults with and without various health problems, also known as ‘discrimination’. Based on this measure, we found that the overall ACE score’s ability to predict poor health ranged from poor to fair at best.

We also examined the classification ability of the established screening cut-off point for ‘high-risk’ (4+) and ‘low-risk’ (0–3) ACE groups. Here, we found that only a small proportion of adults with life-threatening health problems had reported experiencing four or more ACEs. We also found that the actual rates of health problems observed among those in the ‘high-risk’ 4+ ACEs group were relatively low, undermining the ability of this threshold to identify those who may be in genuine need of follow-up care or intervention.

Alongside other recent findings from longitudinal cohort studies in the UK and New Zealand, our results further call into question the usefulness of ACE scores for identifying individuals who may be at risk of poor mental health outcomes. In particular, their inability to discriminate individuals who are most vulnerable risks withholding useful services from those who need them the most, as well as referring individuals to services that are not needed. Such activities are wasteful at best, but also risk harming individuals by creating unsubstantiated worry or stigmatisation.

In the face of calls for universal ACE screening in paediatric settings, these findings also offer empirical support to concerns raised in EIF’s recent consultation of experts in the field, which recommended that all frontline ACE screening practices should be halted until validated measures – in other words, those that are sufficiently sensitive for identifying ACEs – become available, with protocols developed for their use.

Taken together, this new evidence sounds a note of caution about the presumed promise of individualised screening under the current ACEs framework, and suggests a need to reconsider our current direction of travel.

Andrea Danese is professor of child and adolescent psychiatry in the Institute of Psychiatry, Psychology and Neuroscience at King's College London. Alan Meehan is a teaching fellow in the Institute.