Why it’s time to ACE the way we measure the bad things that happen to children

Avatar photo
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.

Posted on

When bad things happen to children, they can potentially have long lasting consequences throughout their lives – they often don’t do as well as their peers at school and work, in their family lives and their health may well 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. In the last two decades huge weight has been given to Adverse Childhood Experience (ACE) scores – where the number of adversities that a person has experienced have been added up, but there are serious concerns about the uses of ACE scores in research, policy and practice.

Being physically or psychologically abused as a child, having a parent who’s 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 colleagues1 in the late 1990s, the ACE scoring system has become widely used and is highly influential in clinical and policy circles. But in a review just published in the Journal of Child Psychology and Psychiatry2, we challenge the heavy reliance on these scores and call for a more critical view of their conceptualisation, measurement and application.

Defining ACEs

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 organisations to organisation, including different/ additional adversities, such as those relating to friends and peers (rejection, bullying, community violence). This makes comparing 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 review3 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:

  1. It assumes each adversity is equally important for outcomes
  2. The patterns or co-occurrence of adversities are ignored. 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 – an unlikely assumption
  3. 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 and their individual and combined effects
  4. There is an assumption that everyone with the same score will receive the same benefit from an intervention regardless of what those adversities or combinations of adversities are
  5. Lack of rationale for summing adversities rather than applying a different statistical approach
  6. Little consideration of the timing, chronicity and discontinuity of adversities
  7. Considerable focus on adult outcomes compared with children’s – something that’s been 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.

In our JCPP review2 we offer a series of recommendations for researchers and practitioners. 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 not deterministic or stigmatising and 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.

Conflict of interest statement

The author declares no conflicts of interest in relation to this blog


Primary paper: Lacey, R and Minnis, H ‘Practitioner review: Twenty years of adverse childhood experience (ACE) scores: advantages, disadvantages and application to practice. Journal of Child Psychology and Psychiatry, 10.1111/jcpp.13135

Full text available.

Other references

(1) Felitti V, Anda RF, Nordenberg D, Williamson DF, Spitz AM, Edwards V et al. Relationship of Childhood Abuse and Household Dysfunction to Many of the Leading Causes of Death in Adults. Am J Prev Med 1998; 14: 245–258.

(2)  Lacey R, Minnis H. Practitioner review: twenty years of adverse childhood experience (ACE) score research: strengths, limitations and application to practice. J Child Psychol Psychiatry doi:10.1111/jcpp.13135.

(3) Appleton A, Holdsworth E, Ryan M, Tracy M. Measuring childhood adversity in life course cardiovascular research: a systematic review. Psychosom Med 2017; 79: 434–440.


Very interested to know if any research has been done or is being done to understand the impact of ACE’s on ‘Attribution style’ and how that impacts outcomes for individuals – is it a significant factor? Or for that matter any research on how ‘attribution style’ forms and develops in individuals? There is considerable research to suggest the link between optimism and good health for example.

Do you propose using other instruments to measure adversity in conjunction with the ACEs? If so, what questionnaires would you recommend using?

It always seemed to me that we (those who use any social science research) don’t always keep in mind differences between solving a measurement problem (for the social scientist, psychologist epidemiologist, etc) and the problems of lived experience! For me, when I read the first published ACE study many years ago, the ACE study was important because it pointed to something quite simple in a way no other research had done — CHILDHOOD MATTERS. Just as Bruce Perry’s pioneering work on the neurobiology of trauma and neglect in childhood supported this idea and Alice Miller’s writings built on this, ACE research brought this home! Clearly future research amplifying and clarifying the various paths and ameliorating factors related to “aces” (writ small) will continue! But we should not wait until all of the research questions are answered before we put the simple truth that CHILDHOOD MATTERS into public policy and our practices (of all types)!

Most, if not all, of the ten original adverse childhood experiences are associated with unsupportive, harmful parenting. This begs the question…Why aren’t we working to elevate the quality of parenting in communities??? If the primary prevention of aces through a new kind of parenting education…one that reaches everyone, everywhere, all the time isn’t made a priority the ACE Study will go down in history as the greatest lost opportunity ever!

I have just started this aces course. It’s very interesting. Having been through a lot of childhood difficulties myself now I can learn through it.

This article has been written by a researcher. I am a practitioner. The article gives a researcher view of the ACE’s system. They value accuracy and detail.

As a practitioner I seek a sentient story and I look for it in behaviour, posture, vagal tone, medical history and do not expect, at intake, that the person will tell me very much at all. In fact, telling that story is avoided until phase 2 of treatment and may be contraindicated even then.

The ACE’s score is just one of several things we use to help us understand the persons story, strengths and needs, which are gleaned over time. The ACE’s score is designed to be non-traumatising, dignity preserving, and achievable (not too intimidating to complete). A bigger battery of statistically valuable items would be impossible to use – such things scare people away from intake.

The ACE’s score has different usefulness to practitioners v researchers. But even researchers should think about the persons dignity, freedom of choice, ability to process memory etc as i suspect that what the author might do could be re-traumatising. I have often seen young people leave the intake process just because they were given a form to fill out. The ACE’s are a single page and the CYW ones only need the person to write 2 numbers. This protects them from intrusiveness and gives a gross score that helps me understand the depth of the problems I may discover over time.

The researcher describes 7 limitations. Let me address them (for shortness of response I suggest that you read the article as i do not describe here what her criticisms are):

1. Each adversity is roughly equal in outcomes to the others. Not exactly but close enough as different adversities can be remarkably similar in outcome and the same adversity can impact in widely different ways. All the ACE’s do is give me a cumulative idea of the depth of problem and some leads as to causation. More will come out as time progresses.

2. Co-occurrence of adversities is not relevant. ACE’s are about complexity not single incident trauma. So if an adversity is named it probably happened multiple (countless) times and in combination with many different adversities.

3. We do not need to know about the mechanisms at play; “how they interact, co-occur and their individual and combined effects”. The ACE’s score is not a diary of what and when things happened. It is a broad indicator of what sort of life was led prior to treatment. Things will have been forgotten, taken as ‘normal’, or dissociated and not reported in the persons ACE’s score. The more ACE’s there are the less important mechanisms are as people respond to different types of memories as the day plays out and to different types of stimuli. It is messy – deal with it.

4. Her statement is strange. Every person gets a tailored treatment depending on their symptoms. The ACE’s score tells us which people to invest the most resources into. A person with a score of 8 gets intervention even without severe symptoms. The ACE’s score tells us to act before, or in the absence of, symptoms.

5. Adversities are summed rather than applying other statistical methods because it is fast and gives a gross answer. Our empathetic brain makes sense of the score in combination with the story, behaviour, body tension, cognitive skills and emotional display (or not) that we see.

6. The timing, chronicity and discontinuity of adversities would be useful however getting that detail would be exposure therapy and that, if wanted by the person, happens only after phase 1 (safety) has been achieved. That is so far into treatment that the ACE’s score has lost its relevance by then.

7. I cannot comment on her last point. I work with young people and how ACE’s scores play out for adults is not my business.

So, long story short, the ACE’s tests work to give a quick and tolerable glimpse of history. It is a start point for understanding and treatment/intervention. A researcher wants data points that are as reliable as possible. Getting that detail at the start of treatment is likely to hinder therapeutic relationship and to be harmful.

the first time i heard about the ACE was at a talk about its effect on childhood. I live in Australia and their is a life expectancy gap between First Australians and the general population of about 20 years. A concept like ACE, if it contained issues relevant to cultural dislocation, removal from family, placement in government care, colonisation trauma could account for much of this gap. It could also provide an understanding of why many First Australians have a lived experience of alcohol and other drug use (whilst as a whole group they actually have the highest rate of abstinence/non-engagement with alcohol and drugs).

I think it would be worth reviewing the ACE elements and creating a much better set of descripotors that take into account issues such as culture, colonialism, out of home care, gender identity related trauma etc. If it is designed to measure harms and predicate a trajectory based on those harms it probably needs to be much more client centric/focused than it is.

Add a comment

Your email address will not be published. Required fields are marked *