What insights can we gain from studying early-onset schizophrenia? The neurodevelopmental pathway and beyond

Keywords::

• adolescent
• copy number variation
• IQ
• memory
• schizophrenia

Schizophrenia is a complex and debilitating brain disorder that commonly presents clinically in late adolescents and early adulthood. However, several converging lines of evidence point to its pathogenesis being rooted in early development [1,2]. The current theory posits that schizophrenia is neurodevelopmental in origin [3], with the interaction between genes and environment resulting in the overt expression of clinical symptomatology. Recently, there has been an upsurge of research interest in studying children and adolescents with schizophrenia since they provide an opportunity to understand the underlying neurodevelopment pathway to disease expression.

Early-onset schizophrenia (EOS; manifestation of psychotic symptoms before 18 years of age) occurs in 4% of all schizophrenia cases, representing a rare, usually more severe and familial variant of the adult-onset counterpart [4–6]. EOS lies at one end of the continuum of disease severity and genetic liability [7]. Early-onset cases show clinical, cognitive, genetic and neurobiologic continuity with adult-onset schizophrenia [7–9], albeit with enduring clinical morbidity and psychosocial disability [10]. Thus, the study of this enriched population is likely to provide important pointers to the neurodevelopmental origins and pathophysiological mechanisms of schizophrenia.

Early-onset schizophrenia runs a chronic course, characterized by severe premorbid neurodevelopmental abnormalities, worse long-term outcome and delays in reaching crucial developmental milestones in motor coordination, language, speech and social functioning before psychosis onset [10–16]. Young-onset cases are associated with a greater number of hospitalizations [17], developmental deviance and higher rate of re-admissions [18] compared with adult-onset schizophrenia.

Many studies suggest that cognitive deficits are an extension of the phenotype of schizophrenia, but few studies have investigated cognition in EOS. Overall, EOS shows generalized cognitive deficits across a broad range of ability domains, with the largest effect sizes being reported in general intellectual functioning, verbal learning and memory, and executive function [19–24]. Similar cognitive impairments have been reported in adult-onset cases, although there are some discrepancies in specific domains of cognition. Some studies have reported that visual information processing and sustained attention are selective strengths in EOS [25–27], but this has not been reported in other studies [20,22]. A plausible explanation for these discrepancies may be disease heterogeneity, the effects of medication and varying IQ. In a group of medication-naive EOS patients, Brickman and colleagues reported a large deficit (<3 standard deviations below normative means) in attention function, as measured using the digit span subtest from the Wechsler Intelligence Scale for Children [21]. By contrast, Ueland et al. confirmed the findings documented in earlier reports showing least impairment in sustained attention and early visual information processing [25].

The issue of whether cognitive functioning changes over time in EOS has been a focus of intense interest. Longitudinal studies on adult-onset patients have shown relative stability or minor improvements over a 2- to 20-year period [28–30]. In a review, Kurtz investigated the trajectory of cognitive deficits across the lifespan in community-dwelling outpatients between the ages of 20 and 85 years, and showed that impairments across a majority of cognitive measures remained stable after illness onset [30]. At 13-year follow-up, Gochman et al. prospectively examined a childhood-onset schizophrenia cohort at 2-year intervals on general intellectual functioning and reported that cognitive decline occurred 2 years prior to disease onset and 2 years after the onset of psychotic symptoms, which was subsequently followed by cognitive stability [31]. Therefore, there is no evidence of a progressive loss of cognitive functioning after the initial decline at illness onset.

During normal neurodevelopment, synaptic density increases until 2 years, followed by gradual decline during childhood and a steep fall-off during late childhood and early adolescence [32]. During adolescence, decreases in synaptic density, axon retraction and neuronal fallout in the prefrontal cortex coincide with an increased ability to solve abstract reasoning, response inhibition, processing speed and complex high-order cognitive tasks. The neurodevelopmental hypothesis proposes that patients with schizophrenia may have too many, too few or unnecessary synaptic connections that are eliminated during adolescence; abnormality in pruning would lead to a reduction in synaptic connectivity that goes beyond the liability threshold, ultimately leading to the manifestation of psychotic symptoms [33].

There are numerous studies of brain structural abnormalities in schizophrenia [34]. Neuroimaging studies of adult-onset schizophrenia have consistently reported increased lateral ventricular volume [35], reduced gray and white matter volumes [36–38], particularly in the prefrontal region [39], and a reduction in thalamic size [40]. Similarly, imaging studies on early-onset cases have reported enlarged ventricular size, lower regional gray matter volumes (e.g., prefrontal cortex and cerebellum), white matter abnormalities [41], increased basal ganglia volume [14,17,42] and reduced volume of the thalamus [9,43], cerebellum [42] and anterior cingulated gyrus [44]. Adult-onset studies have also reported hippocampal volume reduction [45]. While four independent studies reported no hippocampal volume reduction in young-onset cases [46–49], this may have been secondary to a lack of statistical power and heterogeneity with respect to clinical characteristics and pharmacotherapy [49]. Indeed, modifications in methodology and the study of a larger sample revealed reduced hippocampal volume in the National Institute of Mental Health (NIMH) EOS cohort [50–52] which, moreover, was shown to be associated with specific abnormalities in memory function [53]. Longitudinal data available from the NIMH EOS cohort (treatment-refractory group of EOS patients) suggest a back-to-front wave of cortical gray matter volume reduction, beginning in the parietal lobes, progressing rostrally to the superior frontal and temporal cortices [54], together with bilateral dorsal-to-ventral frontal gray matter reductions across the medial hemispheric surfaces separated from cingulate and limbic areas [55]. The progressive loss of the cortical gray matter developmental trajectory normalizes in parietal regions and becomes restricted to the prefrontal and superior temporal cortices as EOS subjects move into adulthood [56].

Few studies have used positron emission tomography to investigate neurobiological and neurochemical mechanisms underlying EOS [14], possibly because of ionizing radiation exposure and the need for ready access to a cyclotron (for a review, see [57]).

Schizophrenia has a heritability of 80% [58]. Twin, family and genetic studies support a strong genetic basis [6,59–62]. EOS patients provide a unique opportunity to understand the pathophysiology of schizophrenia at a neurobiological and genetic level as they provide evidence for the development of symptom expression and are associated with higher genetic loading and familiality. The NIMH group have identified four putative schizophrenia-susceptibility genes in childhood-onset schizophrenia implicated in adult-onset schizophrenia, namely DAOA or G72 (previously known as G72/G30) at 13q33.2 [63], GAD1 (2q31.1) [64], NRG1 (8p12) [65] and DTNBP1 (6p22.3) [66]. GAD1 is associated with cortical gray matter loss [64]; NRG1 is associated with reduction in gray and white matter volume (and lateral ventricular enlargement) [65]; while DTNBP1 expression is influenced by cis-acting variation at the gene locus, which may mediate association between DTNBP1 and schizophrenia, and by trans-acting polymorphism on chromosome 8p [67,68].

Compared with adults with schizophrenia, a higher rate of cytogenetic abnormalities has been reported in individuals in the NIMH EOS cohort, implicating a strong association with very early-onset age. These include an inherited balanced 1:7 translocation, mosaic trisomy X (50% XXX), Turner’s syndrome (XO) and 22q11 deletion (velocardiofacial) syndrome [69]. Atypical sex chromosomal anomalies have also been identified at a higher rate in EOS than adult-onset cases [70]. Technological advances in high-density microarrays have provided us with important insights into difficult to detect deleterious mutations that could affect neurodevelopment. Structural genomic variations, or copy number variations (CNVs), are large duplications and deletions involved in disease etiology. Although some CNVs may be benign, and less likely to be fully penetrant owing to the genetic complexity and clinical heterogeneity of schizophrenia, emerging evidence has reported overabundance of rare CNVs on known genes in EOS [71]. These include the genetic loci 22q11.21, 2p25.3 (MYT1L), 16p11.2 (neurexin 1, NRXN1) and 3p25.3 (SRGAP3), some of which have been associated with early neurodevelopmental difficulties (NRXN1) and neuronal development (SRGAP3). Recent studies have shown that rare de novo copy-number mutations exist in familial and nonfamilial (sporadic) forms in childhood disorders [72]. Investigating the association between specific susceptibility genes and cognitive functioning in schizophrenia has been an area of interest. The Maudsley EOS study showed an association of the serotonin receptor 2A gene (5HTR2A) with attention and executive functioning in EOS probands and their first-degree relatives [Vyas NS, Ternouth A, Powell JF, Puri BK. Association of a 5HTR2A gene polymorphism with executive and attentional functions in patients with early-onset schizophrenia and their first-degree relatives. Manuscript Submitted], which supports an earlier study in adult-onset patients [73]. Further similar investigations using a family-based design [74] may help us to better understand the mechanisms underlying the transition to psychosis in apparently healthy siblings of schizophrenia patients who have not passed the period of risk for developing the illness.

In conclusion, EOS represents a rare, chronic and possibly more severe variant of the adult form of schizophrenia, associated with premorbid abnormalities, cognitive deficits, poor psychosocial outcome, higher genetic loading and greater loss of cortical gray matter volume compared with adult-onset cases. Future research could investigate gene–environment interactions to understand the developmental trajectories and disease progression at a young age in schizophrenia. Advances in current methodologies will provide a blueprint for unraveling the impact of rare structural genetic mutations on the neurodevelopmental pathway in EOS. It is anticipated that investigation into the clinical, genetic, cognitive and neuroimaging aspects of EOS will provide a window of opportunity for creating a novel model that systematically explains the neurobiological pathway underlying the occurrence of events leading to the clinical syndrome.

Financial & competing interests disclosure

Sanjiv Kumra is consultant to Bristol-Myers Squibb, AstraZeneca and Schering Plough. He has also conducted studies that were industry sponsored for AstraZeneca and Bristol-Myers Squibb. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.