Article Text

Download PDFPDF

Should we aspire to screen preschool children for conduct disorder?
  1. P Wilson1,
  2. H Minnis2,
  3. C Puckering2,
  4. C Gillberg2
  1. 1
    General Practice and Primary Care, Faculty of Medicine, University of Glasgow, Glasgow, UK
  2. 2
    Child and Adolescent Psychiatry, Faculty of Medicine, University of Glasgow, Glasgow, UK
  1. Correspondence to Dr P Wilson, General Practice and Primary Care, Faculty of Medicine, University of Glasgow, 1 Horselethill Road, Glasgow G12 9JT, UK; p.wilson{at}


Life-course persistent conduct disorder is a major public health problem with implications for individual health and for society at large. Recent evidence has shown that treatment can be cost-effective, and early symptom detection is improving. We weigh the evidence for and against the introduction of population screening for conduct disorder in early childhood using Wilson and Jungner’s criteria.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Historically, the mental health of infants and young children has received scant attention from health services, but numerous studies in humans and animals have demonstrated the crucial nature of the early weeks and months of life in shaping physical and mental health lifecourse trajectories (see the WAVE report for a summary; This has encouraged development in the field of infant mental health. There is now good evidence that mental health problems in early childhood can be reliably recognised and have longitudinal stability. We argue here that at least one of these problems—life-course persistent conduct disorder (LP-CD—see box 1)—could reasonably be addressed as a component of routine child health surveillance.

Box 1 Conduct disorder and related conditions: a summary (ICD-10)

Conduct disorder

A repetitive and persistent pattern of dissocial, aggressive, or defiant conduct. Such behaviour, when at its most extreme for the individual, should amount to major violations of age-appropriate social expectations, and is therefore more severe than ordinary childish mischief or adolescent rebelliousness. Isolated dissocial or criminal acts are not in themselves grounds for the diagnosis, which implies an enduring pattern of behaviour.

Moffitt1 describes a dual taxonomy to reconcile two apparently incongruous facts about antisocial behaviour: it shows continuity over time, but its prevalence also increases almost tenfold during adolescence. She suggests that there are two distinct categories, each with its own natural history: a small life-course-persistent group engages in antisocial behaviour of one sort or another at every life stage, whereas a larger adolescence-limited group is antisocial only during adolescence. In the life-course-persistent group, children’s neuropsychological problems interact cumulatively with their “criminogenic” environments during development. In the adolescence-limited group, a “maturity gap” encourages teens to mimic the antisocial behaviour of their peers.

Antisocial (dissocial) personality disorder

Personality disorder, usually coming to attention because of a gross disparity between behaviour and the prevailing social norms, and characterised by:

  1. callous unconcern for the feelings of others;

  2. gross and persistent attitude of irresponsibility and disregard for social norms, rules and obligations;

  3. incapacity to maintain enduring relationships, though having no difficulty in establishing them;

  4. very low tolerance to frustration and a low threshold for discharge of aggression, including violence;

  5. incapacity to experience guilt or to profit from experience, particularly punishment;

  6. marked proneness to blame others, or to offer plausible rationalisations, for the behaviour that has brought the patient into conflict with society.

An increasing emphasis on evidence-based practice has led to rationalisation of child health surveillance policies. For example, in the UK, universal screening involving face-to-face contact between health professionals and children2 may no longer extend beyond the first weeks of life—and will be replaced with a more “targeted approach.” There is little agreement on the best mechanism for identifying the most vulnerable children, and some might argue that “targeting” is what health visitors and community nurses are already doing in countries lucky enough to have this service. The accuracy of current identification practice is, of course, not known. We think that a population screening model might offer a useful conceptual framework for thinking about what effective targeting might involve.

In this article, we consider the case for population screening for LP-CD in the preschool years. We refer throughout to whether such a programme could meet Wilson and Jungner’s criteria for population screening for non-infectious disease:3

  • The condition should be an important health problem.

  • The natural history of the condition should be understood.

  • There should be a recognisable latent or early symptomatic stage.

  • There should be a test that is easy to perform and interpret, acceptable, accurate, reliable, sensitive and specific.

  • There should be an accepted treatment recognised for the disease.

  • Treatment should be more effective if started early.

  • There should be a policy on who should be treated.

  • Diagnosis and treatment should be cost-effective.

  • Case-finding should be a continuous process.

More recently published criteria4 follow similar principles.

Condition should be an important health problem

Conduct disorder (CD) is present in 4.9% of 5–10-year-olds,5 and antisocial (dissocial) personality disorder (APD) in 4.1% of young adults.6 There is now good evidence that antisocial behaviour beginning before age 10, known as Lifecourse Persistent Conduct Disorder (LP-CD), is less common than antisocial behaviour beginning in adolescence, but that symptoms are likely to persist.7 This has profound implications for physical and mental health of individuals, and for society: detained young offenders (almost all of whom have CD8) are at nine times greater risk of all cause mortality.9 Scott et al10 demonstrated that the attributable costs of conduct disorder to age 28 were about 10 times those of people without conduct problems. When crime is fully taken into account, lifetime costs to society may be more than 10 times higher than Scott’s data suggest.11

Natural history of the condition should be understood

There are now thought to be two typical life-course trajectories for CD, at least in boys: early-onset life-course persistent conduct disorder (LP-CD) in which there is likely to be psychiatric comorbidity, significant family dysfunction and which also has a high genetic loading, and adolescent-limited CD, which is likely to be more related to the social environment.1 12 13 Genetic and environmental factors interact in the development of antisocial behaviour.14 Twin studies show that some of the genetic influences on antisocial behaviour and attention-deficit/hyperactivity disorder (ADHD) are common to both.15 16

CD is a marker for adult psychopathology. Among 26-year-olds with APD, 62.5% developed CD or related problems at ages 11–15.6 Strikingly, CD also seemed to be a phase in the development of a wide range of adult mental health problems, from eating disorders to schizophrenia, though it predicts these other conditions with a lower degree of specificity.6

There should be a recognisable latent or early symptomatic stage

Use of a simple child behaviour checklist and assignment of high-/low-risk status among kindergarten attenders strongly predicted involvement of criminal justice, mental health and supported education services 7 years later, with sensitivities of 36–61% and specificities of 78–84%.17 The specificity issue would need to be carefully addressed in design of a screening programme, however, if false-positive children are to avoid possible negative effects of the “high risk” label.

Although there is considerable debate on the relationship between oppositional-defiant disorder (ODD) and CD, there is good evidence in some children that ODD represents an early symptomatic stage of CD,18 although ODD is also a precursor of various other child psychiatric disorders.19

There should be a test that is easy to perform and interpret, acceptable, accurate, reliable, sensitive and specific

Whereas Jungner and Wilson were primarily interested in biochemical disorders which might lend themselves to simple screening procedures, LP-CD is a multifactorial condition with both environmental and genetic determinants. Once the condition is entrenched, the diagnosis itself can be made fairly straightforwardly using ICD-10 or DSM-IV criteria. The key issue, however, is whether a reliable diagnostic test can be performed before these diagnostic criteria are met. Reviewing the literature on the early antecedents of disruptive behaviour disorders suggests that certain following factors are associated with conduct disorder:

  • male sex;5

  • ADHD;20

  • ODD;19

  • social deprivation;5

  • parental smoking;21 22 23

  • maternal and family history of problem alcohol use;24

  • maternal history of conduct disorder;22

  • poor family cohesion;25

  • parental separation and early adoption;26 27 28

  • coercive29 and harsh30 parenting;

  • low scores on the HOME31 assessment;32

  • emotional/behavioural dysregulation at 9 months;29

  • antenatal and postnatal depression;25

  • interaction between abuse and low activity monoamine oxidase A genotypes.33

The determinants of CD are becoming clearer and more measurable. For example, the strong association between harsh parenting and aggression is well known,34 and there are now simple and acceptable measures of harsh parenting.35

There is also an increasing knowledge of genetic risk factors. For example, boys who have been maltreated and who have a common functional polymorphism at the neurotransmitter metabolising enzyme gene MAOA are twice as likely to develop CD (80% vs 40%) as those without this polymorphism.33 It would be fairly simple to set up screening for this genotype, but ethical considerations would need to be addressed before this could be implemented.

More work clearly needs to be done before we have a robust early test for the development of LP-CD. A large birth cohort incorporating data on parent–child interactions and genetics would probably provide most of the required data.

There should be an accepted treatment recognised for the disease

The effectiveness of treatments for CD has been demonstrated. To give just one example, Webster–Stratton parenting groups have been shown to be highly effective in the treatment of 3–8-year-olds with the condition (effect size >1).36 The treatment is also acceptable to parents,37 though parents with multiple social and psychiatric problems are the least likely to engage with behavioural family interventions.38

Treatment should be more effective if started early

There is little direct evidence concerning the very early treatment of LP-CD, partly because early diagnosis is difficult. Nevertheless, Olds’ relatively loosely targeted primary prevention programme involving nurse home visitation and parenting support offered to high-risk mothers in the first 2 years of life markedly reduced antisocial behaviour at age 15.39 It is possible that effect sizes would be greater with more accurate targeting of cases.

The treatment of CD in early childhood with parenting groups is relatively effective,36 while effective intervention in adolescence requires more intensive input and is less likely to be successful.40

There is a good deal of evidence that treatment of APD is extremely difficult and resource-intensive—and benefits are at best modest.41 42

There should be a policy on who should be treated

The question of which families are most likely to benefit from intervention is the key issue, and accurate early identification of LP-CD would be the cornerstone of an intervention policy. Uptake of child health surveillance services is generally higher among those who least need it,43 but there is some evidence that parents of children at highest risk for conduct disorder are actually those most keen for intervention.37 More vigorous outreach and support with practical problems such as transport and childcare may be important in ensuring this wish for help is translated into effective engagement with services. These supports would need to be enshrined in public policy documents, and financial plans of councils and health boards or watered-down versions of interventions will be destined for use only by the affluent and well.

Diagnosis and treatment should be cost-effective

Studies by Olds et al,44 Scott et al36 and Edwards et al45 have demonstrated cost-effectiveness. For example, in the Edwards et al study of the Incredible Years parenting intervention with preschool children at risk of conduct disorder (over the clinical cut-off on a well-validated screening instrument), £1344 was all that was required to bring the symptoms of the “average child” in the intervention group down to subclinical levels. When balanced against the lifetime costs of conduct disorder (individuals with conduct disorder engender lifetime costs 10 times greater than those without10) this clearly demonstrates financial benefit.

Case-finding should be a continuous process

The identification of vulnerability can clearly not be a once-for-all exercise. A continuous process of screening in early childhood should be feasible in countries with universal child public health nursing services at various ages, although this may not fit easily with all the proposals of the Hall 4 report.2 46

Why a screening programme rather than primary prevention and opportunistic case finding?

Introduction of a screening programme could be seen as running the risk of “medicalising” ubiquitous social problems, but it can also be defended as good preventive medicine: people who have not received timely help and who become burdened with LP-CD become massive users of healthcare as well as other resources.

Primary care workers and families find it difficult to argue for scarce resources based on the opportunistic finding of putative risk factors for disruptive behaviour problems. Within the UK there are very few services offering interventions to families with young children with developing mental health problems,47 and the perception of many primary care workers is that it is much easier to obtain services when problems are well established and much damage has been done than it is to obtain interventions to minimise problems at an early stage.48 A robust tool for predicting risk, used universally, would make early provision of appropriate help to families more feasible. Furthermore, a screening programme could allow resource allocation to be led more by need than by demand.

In contrast, the evidence for the effectiveness of community-based programmes to prevent disruptive behaviour problems is fairly weak. Outcomes from large-scale UK community-based interventions such as Sure Start49 and Starting Well50 have been disappointing, although long-term follow-up studies39 may be required to demonstrate the value of these interventions loosely targeted on the basis of deprivation in defined geographical areas. Nevertheless, the evaluation of Sure Start49 has suggested that the most vulnerable children may actually be damaged by untargeted interventions, and we would certainly expect specifically targeted interventions to be more effective. Effective treatment needs to include intervention in parenting, not just support for parents,51 52 which has been shown to be ineffective in modifying poor parenting.

There are two possible drawbacks to a screening programme. First, there is the problem of potential inequity. Generally, uptake of screening is higher among affluent families43 where the risk of CD is lower.5 While acknowledging this may be the case, we do not consider it an insuperable challenge. Furthermore, UK policy in relation to child health surveillance2 explicitly promotes the targeting of high-risk families we describe here in terms of screen positivity.

Second, there is the risk of stigmatisation of families where children have positive screening results. The introduction of a screening programme would certainly require considerable publicity, probably easily obtained in the context of current media fascination with parenting issues. We think that a well-argued case for screening and its potential benefits for vulnerable families could be effective in promoting its acceptability—though this clearly requires confirmation in clinical trials.

What might a screening programme involve?

Drawing together the evidence for and against screening, there are clear gaps in our knowledge: the accuracy of current identification practice remains unknown, the predictive utility of screening questionnaires to identify CD remains uncertain, and we do not yet have the data to establish the sensitivity and specificity of an appropriate test battery at different developmental ages. For this reason, it would be premature to suggest details of a screening programme. Conversely, there are well-validated measures—both questionnaire-based and observational—for harsh and insensitive parenting, postnatal depression, parental life history and lifestyle. DNA collection and analysis are now fairly non-invasive and inexpensive using cheek swabs. Identification of “screen-positive” children based on such sensitive data would nevertheless be inconceivable without an ongoing trusting relationship with a professional such as a health visitor or public health nurse.53 Both the evaluation of Sure Start49 and Olds’ work54 suggest that health-led services are those most likely to succeed. Health visitors in the UK already carry out screening for family vulnerability, albeit without the support of a rigorous framework:55 we consider that, along with laboratory services, appropriately trained nurses could effectively screen and treat conduct disorder, calling in help from specialist services only in the most problematic cases. Secondary prevention strategies, such as treatment of comorbid ADHD, may help prevent some of the most adverse outcomes such as substance misuse in later life.56

We do not know if a screening programme and targeted intervention would improve outcomes for children and their families. There are few data on current involvement of community nurses in the field of mental health promotion and early intervention, and it is possible that much good work is already being done, either explicitly or implicitly. We certainly do not know the current sensitivity and specificity of public health nurses’ judgements in the early detection of LP-CD. For this reason, we would hope to see pragmatic trials of screening compared with current provision, with intervention and comparison groups having equal access to resources. A good start would be a complex interventions approach57 to the question of screening utility in conduct disorder beginning with small-scale pilot studies. We would hope that this could lead, in years to come, to potential reductions in mental and physical health burden for some of the most vulnerable members of society and a potentially massive reduction in economic burden for society as a whole.

To summarise:

  • LP-CD is a major public health problem.

  • The natural history of the condition and its risk factors are becoming increasingly well understood, although there are gaps in knowledge.

  • Early intervention is cost-effective.

  • There is a lack of clear policy on how best to improve the mental well-being of very young children and prevent serious psychopathology.

  • There may be a case for a formal screening programme for conduct disorder, but research on infant mental health is required to strengthen this case.



  • Funding PW received support for protected time for research in infant mental health from the Chief Scientist Office of the Scottish Executive Health Department during the preparation of this paper.

  • Competing interests CP is the joint author and trainer for the Mellow Parenting programme. Mellow Parenting is now a charity, but she is not a trustee. She is employed from time to time by this charity to run training courses.

  • The article was conceived and drafted initially by PW, who acts as guarantor. Responsibility for redrafting the article was shared between all the authors.

  • Provenance and peer review Not commissioned; externally peer reviewed.