Shared genetic and familial risk factors

It is implicit from the early onset and marked deficits of the disorders considered for inclusion in this cluster that genetic factors could be salient in their development. The genetic determinants of syndromes that are not included in DSM-IV or ICD-10 but that are similar to the neurodevelopmental disorders, in terms of onset and marked and continuing deficits, support this presumption. For example, Down syndrome and Fragile X syndromes may be recognizable at birth and are associated with significant and continuing mental retardation, and result from trisomy 21 and X-chromosome abnormalities respectively.

Although several candidate chromosomal regions and genes have been proposed, genomic screens have not identified genes for any the neurodevelopmental disorders with the exception of the role of the MECP2 gene in Rett's disorder (Amir et al. Reference Amir, van den Veyver, Wan, Tran, Francke and Zoghbi1999. For reviews of possible loci see: Communication Disorders: Bloodstein & Ratner, Reference Bloodstein and Ratner2008. Learning Disorders: Pennington & Olson, Reference Pennington, Olson, Snowling and Hulme2005. PDDs: Volkmar et al. Reference Volkmar, Paul, Klin and Cohen2005. The importance of genetic factors has nevertheless been confirmed in the Communication disorders (Stuttering: Andrews et al. Reference Andrews, Morris-Yates, Howie and Martin1991; Bloodstein & Ratner, Reference Bloodstein and Ratner2008), Language Impairment (e.g. Expressive Language Disorder, Mixed Receptive-Expressive Language Disorder, Phonological Disorder: Plomin & Dale, Reference Plomin, Dale, Bishop and Leonard2001) ; the Learning disorders (Plomin & Kovas, Reference Plomin and Kovas2005. Reading Disorder: DeFries & Alarcon, Reference DeFries and Alarcon1996; Pennington & Olson, Reference Pennington, Olson, Snowling and Hulme2005; Paracchini et al. Reference Paracchini, Scerri and Monaco2007. Mathematics Disorder: Alarcon et al. Reference Alarcon, DeFries, Light and Pennington1997; Kovas et al. Reference Kovas, Haworth, Harlaar, Petrill, Dale and Plomin2007), and most of the PDDs (Autistic Disorder: Folstein & Rutter, Reference Folstein and Rutter1977; Steffenburg et al. Reference Steffenburg, Gillberg, Hellgren, Andersson, Gillberg, Jakobsson and Bohman1989; Bolton et al. Reference Bolton, MacDonald, Pickles, Rios, Goode, Crowson, Bailey and Rutter1994; Bailey et al. Reference Bailey, Le Couteur, Gottesman, Bolton, Simonoff, Yuzda and Rutter1995; Rutter, Reference Rutter, Volkmar, Paul, Klin and Cohen2005. Asperger's Disorder: Volkmar et al. Reference Volkmar, Klin and Pauls1998; Ghaziuddin, Reference Ghaziuddin2005; Gillberg & Cederlund, Reference Gillberg and Cederlund2005). The strength and scope of this evidence varies for each disorder; for example, most of the genetic research on language has focused on normal linguistic development whereas the genetics of language impairment has been largely unexplored. There is scant genetic information about the written expression, motor and regressive neurodevelopmental disorders such as childhood disintegrative disorder (CDD).

A Neurodevelopmental ‘cluster’ presumes that there may be an overlap between the disorder-specific risks of these disorders that is not shared by disorders in other clusters; that is, the proposed clusters can be delimited by zones of relative rarity. In terms of this Work Group criterion, genomic screens have not identified ‘neurodevelopmental’ genes common to all these disorders. Nevertheless, familial studies which indicate the aggregation of related diseases, and thus it is presumed that disorders within clusters will co-aggregate, do support some common neurodevelopmental genetic risk. In one of the largest investigations into the risk factors of the PDDs, Lauritsen et al. (Reference Lauritsen, Pedersen and Mortensen2005) in a Danish community family sample (total n=943664, autism n=818) show that the risk of Autism, Asperger's and other PDDs are increased in family members (siblings and parents) of autistic probands.

A broad phenotype that spans the neurodevelopmental disorders that increases the risk of language, social communicative deficits and restricted/repetitive behaviours (the ‘autistic symptom triad’) and other forms of cognitive impairment has also been identified. This dimensional model posits that Autism and Asperger's disorders fall at the ‘severe’ end of the genetic spectrum whereas disorders such as mild mental retardation occur at the ‘mild’ end of the spectrum. Subthreshold manifestations of all or some of the symptom triad can be inherited (e.g. Twin studies: Folstein & Rutter, Reference Folstein and Rutter1977; Bailey et al. Reference Bailey, Le Couteur, Gottesman, Bolton, Simonoff, Yuzda and Rutter1995; Le Couteur et al. Reference Le Couteur, Bailey, Goode, Pickles, Robertson, Gottesman, Robertson and Rutter1996. Family studies: Bolton et al. Reference Bolton, MacDonald, Pickles, Rios, Goode, Crowson, Bailey and Rutter1994; Piven et al. Reference Piven, Palmer, Jacohbi, Childress and Arndt1997; Szatmari et al. Reference Szatmari, MacLean, Jones, Bryson, Zwaigenbaum, Bartolucci, Mahoney and Tuff2000). Each symptom domain is moderately to highly heritable but it may be that independent genes are responsible for each symptom domain (Constantino & Todd, Reference Constantino and Todd2003; Ronald et al. Reference Ronald, Happé and Plomin2005, Reference Ronald, Happé, Bolton, Butcher, Price, Wheelwright, Baron-Cohen and Plomin2006a, Reference Ronald, Happé and Plominb; for a review see Happé & Ronald, Reference Happé and Ronald2008). The issue is not resolved.

Genetic data provide less support for the Neurodevelopmental cluster than for the Emotional and Externalizing clusters, where multivariate analyses of genetic-epidemiological data have identified ‘internalizing/emotional’ and ‘externalizing’ genetic liabilities that span most of those disorders (e.g. Kendler et al. Reference Kendler, Prescott, Myers and Neale2003). Similar data are not available to conduct such analyses for all or even most of the neurodevelopmental disorders but there have been comparisons. Kovas et al. (Reference Kovas, Haworth, Harlaar, Petrill, Dale and Plomin2007) in the Twins Early Development study (n=2596) reported a substantial overlap in the genetic risks of reading and mathematical difficulties, as did Markowitz et al. (Reference Markowitz, Willemsen, Trumbetta, van Beijsterveldt and Boomsma2005) in the Netherlands Twin Register (n=1500). Further bivariate and multivariate analyses of twin and family data may serve to identify other neurodevelopmental genetic commonalities and strengthen the justification for grouping these disorders.

In summary, disorder-specific genetic risks have been confirmed in all the neurodevelopmental disorders and there is evidence of a broad neurodevelopmental phenotype. Although these symptoms are heritable, it is unclear whether each symptom domain is caused by independent genes. The limited studies that assess the possible transdiagnostic genetic risk(s) of these disorders at the disorder rather than the symptom level means that the genetic and familial criteria provide weaker support for a Neurodevelopmental cluster than in the case of the Emotional and Externalizing clusters.

Environmental risk factors

Initial conceptualizations posited that deprived early environments caused these disorders. The growing amount of genetic data has meant that formative assumptions such as Kanner's ‘refrigerator’ mothers are now recognized as false (Kanner, Reference Kanner1943).

No environmental factors are necessary or sufficient causes of the neurodevelopmental disorders, even though several perinatal factors have been associated with some of the disorders (e.g. Autism: Larsson et al. Reference Larsson, Eaton, Madsen, Vestergaard, Olesen, Agerbo, Schendel, Thorsen and Mortensen2005). Other environmental risks have been identified in exceptional cases. For instance, toxic exposures to mercury and lead have been associated with isolated cases of autism and intellectual deficits respectively (e.g. Lanphear et al. Reference Lanphear, Hornung, Khoury, Yolton, Baghurst, Bellinger, Canfield, Dietrich, Bornschein, Greene, Rothenberg, Needleman, Schnaas, Wasserman, Graziano and Roberts2005; DeSoto & Hitlan, Reference DeSoto and Hitlan2007).

The utility of this criterion to define a cluster of neurodevelopmental disorders is restricted by the potential importance of perinatal risks that are difficult to define. The limited number of studies that examine environmental risks across disorders or symptom domains also restricts the importance of this criterion when defining a Neurodevelopmental cluster.

Shared neural substrates and biomarkers

Neurodevelopmental disorders are thought to evolve through aberrant processes not typically observed in any phase of normal brain development that alter trajectories in normal brain development (e.g. in Autism: Courchesne et al. Reference Courchesne, Karns, Davis, Ziccardi, Carper, Tigue, Chisum, Moses, Pierce, Lord, Lincoln, Pizzo, Schreibman, Haas, Akshoomoff and Courchesne2001). It is likely that abnormal brain development is a product of gene and perinatal environment. This may explain the persistence of these disorders.

The precise pathophysiology of the neurodevelopmental disorders is unknown although deviations in cortical structures have been identified. The cortical profiles of the learning, communication and motor disorders are the least studied and therefore this review uses the PDDs as an example. In the PDDs, abnormalities in head size and cerebral volume occur. Macrocephaly (head circumference above the 97th percentile) is one of the most consistent physical findings in children with Autism (e.g. Lainhart et al. Reference Lainhart, Piven, Wzorek, Landa, Santangelo, Coon and Folstein1997; Stevenson et al. Reference Stevenson, Schroer, Skinner, Fender and Simensen1997) but at birth these children tend to have smaller heads than normally developing children, with some reports that a marked growth occurs between 6 and 12 months of age (Courchesne et al. Reference Courchesne, Carper and Akshoomoff2003; Hazlett et al. Reference Hazlett, Poe, Gerig, Smith, Provenzale, Ross, Gilmore and Piven2005). Magnetic resonance imaging studies confirm this macrocephaly, showing that Autism is associated with abnormally large brain volumes, although the details differ between studies. A recent meta-analysis of 46 imaging studies found volume increases in total brain, cerebral hemispheres, cerebellum and caudate nucleus but a reduction in the corpus callosum (Stanfield et al. Reference Stanfield, McIntosh, Spencer, Philip, Gaur and Larie2008). There is less evidence that macrocephaly is a defining feature of Asperger's disorder and there has been little investigation of this characteristic in CDD (Cederlund & Gillberg, Reference Cederlund and Gillberg2004). In contrast to Autism, children with Rett's disorder are defined as having a normal head size prior to the onset of regressive symptoms (6–12 months of age; APA, 1994, p. 72, Criteria A and B). The cause(s) of these abnormalities are unclear.

Other neural and biomarker anomalies are evident. Electroencephalography abnormalities occur in some PDDs; for example in autism (Kagan-Kushnir et al. Reference Kagan-Kushnir, Roberts and Snead2005; Coben et al. Reference Coben, Clarke, Hudspeth and Barry2008) and in Rett's disorder (Kaufmann & Moser, Reference Kaufmann and Moser2000). Abnormalities of serotonin concentration and changes in cholinergic and GABAergic indices have also been associated with some of the neurodevelopmental disorders (Chugani et al. Reference Chugani, Muzik, Behen, Rothermel, Janisse, Lee and Chugani1999; Bethea & Sikich, Reference Bethea and Sikich2007) but these findings are far from specific, and none can be considered as either necessary or sufficient to cause the disorder or to provide a ‘neurodevelopmental’ physiological marker.

Although it is presumed that the abnormal developmental trajectory of these disorders are caused by some form of abnormal brain pathology, the identified abnormalities are seldom precisely defined, and to the extent that they have been examined across diagnoses, they seem to be non-specific.

Shared temperamental antecedents

Determining temperamental antecedents in young children affected by developmental disabilities is difficult. In cases where the individual is born with the disorder (MR, PDD) it is not possible to determine the pre-existing temperament. Unlike the Emotional and Externalizing clusters, there is no evidence of a common temperament associated with the disorders considered for inclusion in the Neurodevelopmental cluster.

Symptom similarity and shared cognitive and emotional processing abnormalities

The neurodevelopmental disorders are similar in that they ‘manifest … delay/deviance in maturationally influenced psychological features (i.e. the skills cannot develop unless the necessary neural structure is available) ’ (Rutter et al. Reference Rutter, Kim-Cohen and Maughan2006, p. 278). The disorders in this cluster therefore share the feature of ‘impairment in development’. Moreover, compared to other conditions, such as ADHD, that also involve neurodevelopmental perturbations, the conditions included in the current cluster involve a more stable perturbation in development that persists in the majority of cases. The impairment(s), however, may be in one or several areas, and the pattern of impairment serves to specify the disorder. The developmental areas include deficits in cognition, social interaction, communication and normative behaviour. In the majority of cases, deviance rather than delay in reaching developmental milestones seems a more appropriate characterization, again differentiating the conditions in the current cluster from other so-called ‘childhood’ disorders. Given the integral nature of these domains to everyday life, some degree of functional impairment is also evident. Longitudinal studies show the persistence of the impairment into later life (Maughan & Hagell, Reference Maughan and Hagell1996; Stothard et al. Reference Stothard, Snowling, Bishop, Chipcase and Kaplan1998; Howlin et al. Reference Howlin, Mawhood and Rutter2000; Mawhood et al. Reference Mawhood, Howlin and Rutter2000; Clegg et al. Reference Clegg, Hollis, Mawhood and Rutter2005).

Cognitive impairment is a key feature of the neurodevelopmental disorders. For most of the conditions included in the cluster, standardized tests exist for documenting the severity of the impairment. For example, cognitive impairment characterizes MR, learning, motor, and communication disorders, where standard measures exist for characterizing the impairment. The PDDs also involve cognitive impairment, although less standardized tests exist for identifying associated cognitive deficits in the PDDs than for other disorders included in this cluster. The nature of cognitive impairment varies among the disorders, ranging from deficits confined to specific core processes to more generalized problems. The notion of the presence of relatively ‘preserved’ capacities in the face of specific impairment is well established and even in disorders such as MR, where more pervasive deficit is expected, patterns of relative strengths and weaknesses are still evident (Sattler, Reference Sattler1992). With regard to MR, pervasive and significant deficit (>2 s.d.) in general intellectual functioning and significant limitations in adaptive functioning are expected and form part of the definition of the syndrome. With regard to the disorders that fall under the umbrella of PDD, at least half have intellectual impairment (Chakrabarti & Fombonne, Reference Chakrabarti and Fombonne2001; Yeargin-Allsopp & Boyle, Reference Yeargin-Allsopp and Boyle2002). Autism is associated with more severe MR, whereas in Asperger's disorder, by definition, intellectual impairment is deemed to be not ‘clinically significant’, but deficits in domains other than intellectual functioning, particularly social communication, support the inclusion of Asperger's disorder here. For each of the PDDs, perturbations in social processing are detectable on standardized measures, such as the Autism Direct Observation Scale (ADOS, Lord et al. Reference Lord, Rutter, Goode, Heemsbergen, Jordan, Mawhood and Schopler1989). By contrast, the learning, motor and communication disorders are defined by more isolated deficits in the relevant domain(s), discordant with the individual's chronological age and measured intelligence.

Although the presence of aberrant development sets children with the neurodevelopmental disorders apart from the general population, this factor is not a point of absolute distinction from adults and children with disorders represented in other clusters. In fact, developmental delay and overt cognitive impairment are associated with multiple syndromes. Indeed, the disorders represented in the Neurocognitive cluster also have cognitive deficits as a key feature, as noted by Sachdev et al. (Reference Sachdev, Andrews, Hobbs, Sunderland and Anderson2009). A minor lowering (0.5 s.d.) of intellectual functioning is also associated with some externalizing disorders (Sattler, Reference Sattler1992; Sattler & Hoge, Reference Sattler and Hoge2006; Krueger & South, Reference Krueger and South2009). In the neurodevelopmental disorders the cognitive symptoms arise early and persist as the defining feature of the syndromes. The developmental impairments seem to have neural underpinnings. Cognitive deficits in the neurodevelopmental disorders are persistent, and this is supported by psychometric tests used to reliably track these deficits. In the future, the availability of such tests may allow characterization of the persistence of the core cognitive deficits in other clusters.

High rates of co-morbidity

Grouping the neurodevelopmental disorders presumes that within-cluster co-morbidity is higher than between-cluster co-morbidity. Few studies, however, have examined the differences between these disorders, and unlike analyses of the Emotional and Externalizing clusters there have been no analyses of within- versus between-cluster co-morbidity rates. Thus, for current purposes it is best to examine the co-morbidity of these disorders at the symptom level, and as noted in the last section, there is substantial overlap for all the neurodevelopmental disorders in terms of cognitive, social and communicative deficits, and particularly for intellectual impairment. Stuttering, for example, is associated with delays in the acquisition of language and with poor articulation independent of the stutter itself (Andrews & Harris, Reference Andrews and Harris1964), and MR occurs in most Autism cases (for reviews see Fombonne, Reference Fombonne1999, Reference Fombonne2003).

Co-morbidity is not exclusively limited to the disorders within the Neurodevelopmental cluster. MR (or intellectual impairment) is a key feature in the dementias of the Neurocognitive cluster (Sachdev et al. Reference Sachdev, Andrews, Hobbs, Sunderland and Anderson2009), and emotional and externalizing symptoms may also co-occur with some of the neurodevelopmental disorders (e.g. Kim et al. Reference Kim, Szatmari, Bryson, Striner and Wilson2000). Although the within-cluster co-morbidity between the neurodevelopmental disorders, particularly at the symptom level, supports grouping these disorders together, future analyses comparing co-morbidity rates across disorders and clusters may provide more robust support for this cluster.

Course of illness

The neurodevelopmental disorders have an early age of onset. Given the significant role of genetics in the development of all the neurodevelopmental disorders, it is presumed that these disorders are present from birth. For the PDDs, developmental problems may be recognized in the first year of life (Adrien et al. Reference Adrien, Lenoir, Martineau, Perrot, Hameury, Larmande and Sauvage1993; Osterling & Dawson, Reference Osterling and Dawson1994; Baranek, Reference Baranek1999; Werner et al. Reference Werner, Dawson, Osterling and Dinno2000) although symptom severity and parental concern regarding their child's development can influence when the child receives a diagnosis (Stone et al. Reference Stone, Hoffman, Lewis and Ousley1994; De Giacomo & Fombonne, Reference De Giacomo and Fombonne1998; Twyman et al. Reference Twyman, Maxim, Leet and Ultmann2009). Autism diagnoses typically do not occur until 6 years of age (Howlin & Moore, Reference Howlin and Moore1997; Howlin & Asgharian, Reference Howlin and Asgharian1999) but Autism does tend to be diagnosed earlier than Asperger's disorder (mean age of Asperger's diagnosis is 11 years; Howlin & Asgharian, Reference Howlin and Asgharian1999). The onset of the early regressive symptoms in Rett's and CDD may mean that these conditions are diagnosed earlier than both Autism and Asperger's disorder. Some learning disorders and milder forms of MR may only be detected at a later age when academic testing occurs, typically in elementary school.

The course of the neurodevelopmental disorders is usually continuous, but some speech and motor disorders may be exceptions to this trend (Communication disorders: Stothard et al. Reference Stothard, Snowling, Bishop, Chipcase and Kaplan1998; Johnson et al. Reference Johnson, Beitchman, Young, Escobar, Atkinson, Wilson, Brownlie, Douglas, Taback, Lam and Wang1999; Kloth et al. Reference Kloth, Kraaimaat, Janssen and Brutten1999; Clegg et al. Reference Clegg, Hollis, Mawhood and Rutter2005; Bloodstein & Ratner, Reference Bloodstein and Ratner2008. Learning disorders: Maughan & Hagell, Reference Maughan and Hagell1996; Shaywitz et al. Reference Shaywitz, Fletcher, Holahan, Shneider, Marchione, Stuebing, Francis, Pugh and Shaywitz1999; Mattison et al. Reference Mattison, Hooper and Glassberg2002. Motor Disorder: Cantell et al. Reference Cantell, Smyth and Ahonen1994, Reference Cantell, Smyth and Ahonen2003; Sugden & Chambers, Reference Sugden and Chambers2007. PDDs: Howlin et al. Reference Howlin, Mawhood and Rutter2000, Reference Howlin, Goode, Hutton and Rutter2004; Mawhood et al. Reference Mawhood, Howlin and Rutter2000; Cederlund et al. Reference Cederlund, Hagberg, Billstedt, Gillberg and Gillberg2008). CDD and Rett's disorders by definition have a deteriorating course. Although the clinical profile in each neurodevelopmental disorder does change as the child matures, evidence of the underlying deficit typically persists. A persistent course, more than any other single characteristic, distinguishes the child-onset conditions included in this cluster from those included in other clusters. In some cases where treatment is focused on function, a disorder may seem to remit. However, even here, treatments do not alter course per se; rather, they provide the individual with different skill sets, such as facilitating interpersonal and occupational skills for individuals with MR to compensate for a stable underlying deficit.

Treatment response

The main focus of treatment for the neurodevelopmental disorders is to reduce the maladaptive behaviours and increase functionality in the presence of the enduring deficit. Early and intense educative and behavioural treatments may increase skills and reduce maladaptive behaviours in some cases of PDDs (Sallows & Graupner, Reference Sallows and Graupner2005; Remington et al. Reference Remington, Hastings, Kovshoff, degli Espinosa, Jahr, Brown, Alsford, Lemaic and Ward2007), Developmental Motor (Pless & Carlsson, Reference Pless and Carlsson2000), Communication (Bothe et al. Reference Bothe, Davidow and Bramlett2006) and Learning disorders (Reynolds et al. Reference Reynolds, Nicolson and Hambly2003; Willner, Reference Willner2005). However, few findings have been replicated in randomized controlled trials.

No pharmacotherapies reduce the core social and communicative deficits of autism or Asperger's disorder. These agents are used, however, to control co-morbid emotional and cognitive symptoms. For example, antipsychotics have been used to target co-morbid psychiatric symptoms and aggression in adults with MR (Williams et al. Reference Williams, Clarke, Bouras, Martin and Holt2000). Cholinesterase inhibitors by contrast, may provide direct benefits for some cases of Autism (Chez et al. Reference Chez, Aimonovitch, Buchanan, Mrazek and Tremb2004).

In summary, treatment of the neurodevelopmental disorders is focused on improving functionality and therefore reflects the persisting nature of the disorders. There is a range of treatment options to manage the difficult and disruptive symptoms that accompany the neurodevelopmental disorders.