ADHD as a multifactorial disorder
Misleading “hype” by scientists and journalists often involves reference to “genes for ADHD”, or for schizophrenia, or for bipolar disorder. The implication is one of a deterministic action by which genes directly cause the psychological condition. This notion is also inappropriately applied to multifactorial disorders such as ADHD. There is no one direct cause—either genetic or nongenetic. Rather, the reality is a complex interacting admixture of multiple genes and multiple environmental risk factors. That is what is meant by multifactorial causation. It is implied by a heritability that is well short of unity. It is possible that within the broad group of ADHD disorders there are a few due to direct gene effects operating in Mendelian fashion but, if so, it is highly probable that these account for a tiny proportion of cases. Rather, the likely scenario is of individual susceptibility genes that increase the risk of ADHD, but whether or not ADHD actually develops depend on the presence or absence of other genes, and of nongenetic risk factors. At the moment, we do not know whether this mixture of genetic and nongenetic risk factors operates additively (i.e.. no particular factor has to be present: rather it is a question of how many out of a large pool are operating) or synergistically (i.e., the emergence of disorder is dependent on the catalytic interplay among particular groups of risk factors).
ADHD as a coinorbid disorder
Along with most other child psychiatric disorders (Angold, Costello, & Erkanli, 1999), it is very common for ADHD to be accompanied by (be comorbid with) supposedly separate and different forms of psychopathology. Thus, it co-occurs with language deficits (see Lewis. Chapter 5, this volume), reading difficulties (see Stevenson, Chapter 6, this volume), and with oppositional/defiant and Conduct Disorders (see Waldman, Rhee, Levy, & Hay. Chapter 7, this volume). There is also less consistent association with other disorders such as Tourette’s syndrome (Pauls, Fredine, Lynch, Hurst. & Alsobrook, Chapter 11, this volume). There are many possible explanations for these comorbid patterns (Caron & Rutter, 1991; Rutter, 1997). each of which has rather different clinical implications. It could be, for example, that they share some key risk factors—such as a genetically influenced sensationseeking trait, or family adversity, or scapegoating—with the syndromic outcomes being different because of other risk factors that are not shared. On the other hand, it could be that the various associated disorders constitute alternative, interchangeable, manifestations of the same genetic liability. In that connection, it is important to recognise that the phenomena might change developmentally, even though the genetic risk factors do not, or vice versa (see Eaves, Maes, Rutter, & Silberg, in press; Hay, McStephen, & Levy. Chapter 4, this volume; Nadder, Silberg, Maes, Rutter & Eaves, 2000). Yet again, it could be that an earlier behaviour, such as inattention or overactivity. provides a risk for a later one, such as antisocial behaviour. This could be either because it serves to put children into risk circumstances, or to render them more susceptible to environmental hazards. The last possibilities involve various forms of gene-environment correlation (rGE) or interaction (G x E). These are discussed in greater detail below, but the point that needs to be made here is that although they crucially implicate the environment in the causal process, standard genetic analyses will identify the effect as purely genetic (see Rutter, Silberg, & Simonoff. 1993; Silberg, Rutter, Neale. & Eaves, in press). In order to differentiate these G x E interplay causal processes from effects solely attributable to a shared genetic liability, it is necessary to have longitudinal data over the relevant age span. It is also necessary to have discriminating measures of the environment, and to use appropriate methods of data analysis. In many ways, the key question is whether a MZ co-twin of a child with ADHD, who does not him- or herself show ADHD, has an increased risk of ODD/CD (or whatever comorbid disorder is being considered). It will be appreciated, however, that such an analysis immediately requires the need to consider measurement error and the arbitrariness of categorical diagnostic thresholds. For obvious reasons, the identification of specific susceptibility genes will help, but only to the extent that, in combination, these genes account for a substantial proportion of the genetic variance.
ADHD as a predominantly male disorder
As Rhee, Waldman, Levy, and Hay (Chapter 8. this volume) emphasise. ADHD is several times more common in males than in females. Their data suggest that a polygenic threshold model may be operating, in which females require a higher genetic “dose” to manifest the disorder. Nevertheless, despite their elegant and well-targeted analyses, they note that there is conflicting evidence, and the issue remains to be fully resolved. Two further points need to be added. First, gender differences in psychopathology are very common (see Moffitt. Caspi, Rutter. & Silva, in press). Thus, autism, ADHD, dyslexia, and developmental language disorders are all far more common in males. They tend to be conceptualised as varieties of neurodevelopmental disorder. Is the explanation of their male preponderance the same for all of these conditions? Conversely, depression and anorexia/bulimia nervosa are more common in females. Does this have a different explanation? Second, even if the threshold liability model is correct, what is its explanation at the molecular level? This remains a hugely important and under explored area.