Abstract
Many dilemmas confront the clinician's assessment and treatment of children with ADHD. I wish to explore these and also note the models that have attempted to address them. This may lead to an innovative model with potential heuristic value and testable hypotheses.
Core symptom dimensions constitute an initial difficulty. The relationship between hyperactive, impulsive and inattentive symptoms remains unclear [1]. Is the inattentive type distinct from the combined type; a different expression of the same disorder; or both at different stages of development and in different contexts in particular individuals? The age of onset of these core symptoms and their progression also varies considerably.
Comorbid conditions are increasingly recognized, regardless of subtype [2, 3]. Oppositional Defiant Disorder (ODD) is evident in up to half of the children with a combined type ADHD-CT; conduct disorder occurs in 2–3% of children with pre-existing ADHD-CT and comorbid ODD. Children with ADHD-CT and either of these comorbid conditions have worse verbal and visuospatial skills, lower full scale and verbal IQ and poorer academic achievement. In addition, 20–30% of children with ADHD, Inattentive or Combined type, have spelling, reading, writing and/or arithmetic learning difficulties. Anxiety occurs in a similar proportion, while depression, primarily dysthymic disorder, alone or co-occurring with major depressive disorder (so called ‘double depression’), affects 15–30% of children with ADHD-CT. In contrast, a vaguely defined ‘juvenile’ form of bipolar disorder with marked irritability has been described. Four comments can be made about these comorbid patterns: they are common; parent and child reports are discordant and vary within individuals at different stages of development and in different contexts; the greater the number of conditions, the more severe the ADHD-CT; and the response of ADHD-CT symptoms to short-term (up to 6 weeks duration) and/or longer-term (greater than 3 months duration) psychostimulant medication and/or psychotherapy targeting home and/or school environments maybe diminished by comorbidity, particularly anxiety.
Risk factors are either individual, interpersonal, family or social [4]. Individual risk factors include the infant's temperament (for example, decreased rhythmicity, adaptability, span and persistence of attention, and increased threshold of responsiveness and intensity of reaction); reduced verbal and visuospatial working memory and response inhibition executive function abilities; increased sensitivity to noradrenergic activity; and decreased self-esteem. Interpersonal risk factors include impaired relationships with peers, parents and siblings. Family factors over the life cycle are parental psychopathology, decreased flexibility, adaptability, confiding, and nurturing and increased hostile criticism and over-involvement. Social factors include lower socioeconomic status and increased rigidity and inflexibility of attitudes and beliefs about appropriate, responsible and accountable human behaviour, emotional expression and cognition. These factors are not the cause of ADHD. Indeed, they vary in a given child at different points in development.
Psychostimulant and behavioural interventions have been the most studied treatments for ADHD-CT [5]. Psychostimulant medication is a primary treatment modality for children with ADHD-CT. In the short term (up to 4–6 weeks), the core behavioural features of ADHD-CT and executive functions, such as response inhibition and verbal and non-verbal working memory performance, improve in 80%. In the longer-term (greater than 3 months), this improvement may lessen; in contrast, improved executive functions persist. Therefore, constructs of cognition such as executive function may help us understand the reasons behind the poor behavioural response of a subgroup of children who are treated with medication in the longer-term [3]. These reasons may include the complex interplay of the interdependent vulnerabilities of executive functioning deficits, comorbid conditions and psychosocial risk factors that represent a final common pathway of a number of biological and psychosocial disturbances evident at particular periods in early, middle and late childhood, resulting in ADHD-CT. Teacher and parent training programmes that involve reinforcement of positive behaviour and response-cost procedures for undesired behaviour have been associated with short-term improvements in core symptoms, academic performance, social skills, aggression, and oppositional defiant behaviour. In addition, the potential synergistic effect of behavioural interventions with psychostimulants has been reported in the short term; in the longer-term, lower doses of medication were required in the combined medication and behaviour therapy group compared to the medication alone group in the 14 month NIMH Collaborative Study [6]. Future research will need to aid the clinician to tailor validated treatments to the nature and severity of a child's defined impairments and disabilities, which vary in their relative contribution at particular stages and within particular contexts.
Contemporary conceptual models and a potential innovative approach
Four models have been used to address the dilemmas noted above. (i) Pervasiveness of core symptoms has been emphasized through the use of categorical and dimensional measures, to decrease sample heterogeneity and thereby to ensure study of the ‘true’ disorder. High rates of comorbid conditions have limited the success of this approach. (ii) Equifinality and multifinality constructs from the developmental psychopathology discourse [7] and (iii) heterogeneity and pleiotropism models from behavioural genetics [8] have been suggested to deal with the confounding effects of the context- and developmentdependent interplay of core and comorbid symptoms, risk factors and treatment response. (iv) Vulnerability, risk [9] and resilience [10] have also been proffered as a heuristic model that reflects the interplay of core symptoms, comorbid conditions, risk factors and medication and psychological treatment response in different contexts and at different stages of development. While these help cross-sectional description work, they are limited in their power to determine the most appropriate priority of key symptoms, comorbidity, risk factors and treatment response longitudinally. This limitation arises in part from the inability to address the greater degree of interdependence of the four key domains in ADHD compared to other disorders such as schizophrenia. Similarly, a factor of certain severity may be a risk factor in a given patient but be a resilience factor at a lower level of severity. Resilience factors were originally conceptualized as intra-individual in nature, whereas intra-individual and interpersonal aspects may apply.
A logical extension of the former three models addresses the relatively greater degree of interdependence of core symptoms, comorbid symptoms, risk/ resilience factors and treatment response longitudinally in a given child. Although this requires complex correlational statistics and large sample sizes, the advantages for clinician and researcher would be considerable. Just as chaos mathematics has demonstrated that complex orders of relationships can exist within apparent randomness, a more useful pattern of significant statistical effects may exist in the relationships between core symptoms, comorbid conditions, key risk/resilience factors and medication and psychological treatment response than in the four domains themselves.