Familial risk

The epidemiological study by Morgan and colleagues ^{Reference Morgan, Croft, Valuri, Zubrick, Bower and McNeil4} in this issue adds to this burgeoning area of interest in two main ways. First, the findings suggest that there is a three-fold increased risk of intellectual disability in the offspring of mothers with bipolar disorder and major depression as well as those with schizophrenia. These finding are congruent with evidence for genetic overlap between schizophrenia and affective disorders and with suggestions that the three disorders lie on an aetiological and neurodevelopmental continuum. ^{Reference Craddock and Owen5} However, the evidence for an aetiological link between intellectual disability and affective disorders is less strong than for schizophrenia and needs to be a focus for future studies.

One possibility arising from Morgan and colleagues' findings is that the increased rates of intellectual disability seen in the offspring of mothers with affective disorders reflect assortative mating. Indeed, 20% of affected children in the maternal schizophrenia group had a father with a psychiatric illness recorded on the psychiatric case register compared with 40% in the maternal bipolar group, 32% in the maternal unipolar group and 19% in the comparison group. However, the children of women with schizophrenia were more likely to have had no father registered at their birth than children in the other groups (21% compared with 9%, 5% and 4% respectively for maternal bipolar disorder, maternal unipolar disorder and the comparison children). It is also possible that environmental factors are playing a role in the outcomes seen in the offspring of the women with affective disorder. As the authors acknowledge, a significant weakness of the study is the absence of data on psychotropic medication. Although the risk of intellectual disability in offspring was not related to the timing of the onset of maternal psychiatric illness relative to the index birth for the children of mothers with schizophrenia or unipolar major depression, for children of mothers with bipolar disorder, the risk of intellectual disability was significantly increased only if onset of maternal illness pre-dated the index birth. The authors are wisely cautious in drawing conclusions but note that this finding may reflect the role of medication taken during pregnancy and that mood stabilising drugs such as sodium valproate have been associated with an increased risk of congenital anomalies when used in pregnancy.

Obstetric complications

The second important finding of Morgan and colleagues is that obstetric complications, which have long been known to be associated with increased risk of intellectual disability, appear to act independently of familial risk. ^{Reference Morgan, Croft, Valuri, Zubrick, Bower and McNeil4} Obstetric complications have been consistently implicated as environmental risk factors for a range of neurodevelopmental disorders including non-syndromal intellectual disability, autism, ADHD, epilepsy and schizophrenia. The similarity between this range of outcomes and that seen in association with pathogenic CNVs is striking. In the 1950s Pasamanick and colleagues ^{Reference Pasamanick, Rogers and Lilienfeld6} proposed the hypothesis of a ‘continuum of reproductive causality’ consisting of brain damage incurred during pregnancy or during or around birth leading to a gradient of injury extending from fetal and neonatal death through cerebral palsy, epilepsy, intellectual disability and behavioural disorder including schizophrenia. Given recent genetic findings it seems reasonable to modify this concept to encompass a continuum of genetically and environmentally induced neurodevelopmental causality along which lie intellectual disability, epilepsy, autism, ADHD and schizophrenia and possibly the major affective disorders. ^{Reference Owen, O'Donovan, Thapar and Craddock3,Reference Craddock and Owen5} This view recognises the degree of aetiological and symptomatic overlap between diagnostic groups as well as the lack of clear diagnostic boundaries and sees the major clinical syndromes reflecting in part the severity and predominant pattern of abnormal brain development and resulting functional abnormalities as well as the modifying effects of other genetic and environmental factors. ^{Reference Owen, O'Donovan, Thapar and Craddock3}

Implications

As Morgan and colleagues note, the mechanisms by which obstetric complications compromise neurodevelopment and confer risk for psychiatric and other disorders remain to be determined, and this is likely to prove challenging given the variety and complexity of the exposures involved. ^{Reference Morgan, Croft, Valuri, Zubrick, Bower and McNeil4} The nature of the relationship between obstetric and genetic risk factors is also uncertain, but it is parsimonious to assume that they confer risk by acting either additively or multiplicatively on common neurodevelopmental and neurobiological processes. The implication of specific genetic events, including not only CNVs but other classes of rare mutation, offers a means by which such processes can be identified and explored experimentally in animal and cellular models. For example, there is accumulating evidence that the genes disrupted by CNVs that confer risk for schizophrenia converge upon a small set of post-synaptic proteins that have been implicated in synaptic plasticity, which play key roles in neurodevelopment and cognition. ^{Reference Kirov, Pocklington, Holmans, Ivanov, Ikeda and Ruderfer7} It will be possible to study the effects of these mutations in experimental animals and also to model the impact of specific environmental exposures on relevant outcomes. It is also now possible to reprogramme somatic cells from humans into induced pluripotent stem cells and to derive neuronal cells from these stem cells. ^{Reference Han, Williams and Eggan8} These and related techniques offer opportunities to create cellular models of individuals with specific disease phenotypes and to gain insights into developmental and other cellular and molecular phenotypes. Such models will likely have limitations but they offer the possibility of modelling disorders in which multiple genetic risk factors operate in individual cases and for creating platforms by which new therapeutic agents can be tested.

Finally, recent findings implicating shared aetiological factors across a range of disorders remind us how much we still have to learn from studying the interfaces and overlaps between current diagnostic categories. Much research still focuses on cases that satisfy our current largely descriptive DSM and ICD categories and excludes those many ‘comorbid’ cases that do not neatly fit these. Current categories will remain useful for research so long as we expect heterogeneity and overlapping risk factors and mechanisms. However, we must also be prepared to explore novel dimensional and categorical approaches that cut across current categories and that might better capture underlying psychology and biology and to employ longitudinal and developmental designs. Mechanistic insights will also come from studies of endophenotypes. ^{Reference Gottesman and Gould9,Reference Walters and Owen10} Here we will need a combination of top-down approaches, relating specific psychopathological syndromes to phenotypes defined by cognitive psychology and neuroscience rather than diagnosis, and bottom-up approaches relating genetics to fundamental measures of neuronal and synaptic function. These studies will increasingly be informed by close integration with observational and experimental studies of animal and cellular models.