When bad things happen to children, they can sometimes have long lasting consequences. These children often don’t do as well as their peers at school and at work. Their family lives can be more difficult and their health is likely to be worse. There is no magic wand to stop bad things happening, but how we measure the impact of those bad things is absolutely key to helping these children fare better. That’s what our practitioner review, ‘Twenty years of adverse childhood experience (ACE) scores: advantages, disadvantages and application to practice’ seeks to address.
In the last two decades, considerable weight has been given to Adverse Childhood Experience (ACE) scores – where the number of adversities that a person has experienced are added up to provide a total score. But there are serious concerns about how these ACE scores are used in research, policy and practice.
Being physically or psychologically abused as a child, having a parent who is drug or alcohol-dependent, in prison, or the victim of domestic abuse, are all experiences which are likely to be stressful for children as they grow up. Experiencing more than one of these things is likely to increase that damage. But how do the people seeking to protect, support and help these children best measure those effects in order to give them the help they need to thrive as they get older?
Since its development by Vincent Felliti and colleagues in the late 1990s, the ACE scoring system has become widely used and is highly influential in clinical and policy circles. But our review for the for the Journal of Child Psychology and Psychiatry challenges the heavy reliance on these scores, calls for a more critical view of their conceptualisation, measurement and application and makes a series of recommendations for all those who make use of them.
What is an ACE?
We started by asking ourselves, “What exactly are ACEs?” and came up with the following definition:
‘experiences which require significant adaptation by the developing child in terms of psychological, social and neurodevelopmental systems, and which are outside of the normal expected environment’
There are 10 commonly used ACEs which, broadly speaking, cover abuse, living in a dysfunctional household, parental separation and neglect. But the choice of these has rarely been questioned, and a rationale for including these and excluding others was never given in the original research. Some researchers have ‘branched out’ adding things like poverty, bullying, race discrimination and parenting.
It goes without saying that collecting information on these experiences is challenging. ACE questionnaires differ from organisation to organisation, including different/additional adversities, such as those relating to friends and peers (rejection, bullying, community violence). This makes looking across study findings a bit like comparing apples and oranges.
The fact that people completing these questionnaires only give a yes or no answer to questions about any bad experiences they may have had is in itself problematic. This means that risk is arbitrarily attributed, that the severity, frequency and duration of those experiences are simply not considered. These factors are likely to be important. For example, leaving a young child unattended for a short period might not be considered neglect, whilst repeated and ongoing lack of supervision would.
A 2017 review showed that 77 per cent of studies used unweighted cumulative risk (ACE) scores and only three weighted adversities by their perceived severity. Research dating back as far as the 1970s on adult life events has tried to propose a more sophisticated approach to the weighting of experiences but has been largely ignored in relation to childhood adversities.
How are ACEs measured?
The way in which ACEs have been measured to date is also limited and often problematic. Assigning a ‘cumulative risk score’ based on the number of adversities experienced by an individual and an arbitrarily assigned score between 1 (not stressful) to 100 (very stressful) was first used in 1978 to look at the links between family focused adversities and the risk of bad behaviour in children. The study concluded that children with four or more adversities were most likely to behave poorly.
Fellitti himself used this approach to show strong links between higher ACE scores and poorer health including heart disease, cancer and diabetes. This approach has since been replicated in hundreds of studies to look at other outcomes such as depression, premature death and sleep disorders.
The simplicity of this approach has long been considered a strength – it’s easy to calculate and understand, making it possible to engage non-academics in the long term effects of early life experiences on the population’s lifelong health. It increases the likelihood of clear ‘harder hitting’ findings and makes identifying people at the highest risk a lot easier.
But, or maybe even because of this, the limitations of this approach have been somewhat ignored. Here are some of the limitations we have identified:
- An assumption that each adversity is equally important for outcomes
- Ignores patterns or co-occurrence of adversities. A child who witnesses domestic violence, divorce and has a parent with mental health problems is assigned a score of 3, the same as a child who experiences emotional, physical and sexual abuse implying both children have the same risk of poorer outcomes
- ACE scores tell us nothing about the mechanisms at play – we need to know about the effect of separate adversities, how they interact, co-occur – their individual and combined effects
- The assumption that everyone with the same score will receive the same benefit from an intervention regardless of what those adversities are
- No clear rationale for summing adversities rather than applying a different statistical approach
- Little consideration of the timing, chronicity and discontinuity of adversities
- Considerable focus on adult outcomes compared with children’s –driven by the reliance on retrospective reporting and issues around disclosure
Implications for policy and practice
ACE scores and findings from ACEs research have been used widely in policy, public health and clinical work, for instance in ‘trauma informed’, routine enquiry and ‘ACE awareness’ initiatives. The simplicity of ACE scores has been helpful in highlighting the importance of ACEs amongst broad audiences. However there are concerns about the use of ACEs research and ACE questionnaires in practice. There have also been concerns that evidence from population level studies has been directly translated as individual risk and in a deterministic way, which implies that people who have experienced ACEs will have negative outcomes. There has also been a focus upon individuals and families, rather than on broader structural issues like child poverty, which is strongly related to the likelihood of experiencing ACEs.
Researchers have a responsibility to be clearer about their definitions of adversity – what is an ACE and what is not? Why stick to the usual 10 ACE items when there are other potential ACEs? How do ACEs cluster, and how they predict outcomes individually or together? Longitudinal studies which have prospectively-collected ACEs information are really key to exploring the importance of timing, frequency, duration and severity of ACEs.
Practitioners are urged to communicate messages about risk in a sensitive way which is neither deterministic nor stigmatising and which does not conflate population level and individual level risk. We also urge caution around implementing routine enquiry, particularly on how data are used, the availability of evidence-based interventions and again messages about determinism, stigma and what is an ACE apply here. A focus on the structural ‘causes’ of ACEs is also warranted, particularly focusing on child poverty and inequality.
Heightened policy interest in in this area, including the 2017 select committee inquiry led by Sir Norman Lamb, have provided the opportunity to achieve a step-change in the way we look at adverse childhood experiences. If we seize this opportunity, the research community can play a key role in supporting the desire of policy makers and politicians to be effective in achieving their aim of helping at-risk children, adults affected by earlier bad experiences, the NHS, all those working in this area and UK taxpayers.
Rebecca Lacey is a Senior Research Fellow in life course social epidemiology in the Department of Epidemiology and Public Health at University College London. Her research interests lie in the long-term health effects of early life adversities and effects of social relationships on health using longitudinal population datasets.