Familial liability, obstetric complications and childhood development abnormalities in early onset schizophrenia: a case control study

Early-onset schizophrenia (EOS), manifesting before the age of 18 years, and very early-onset schizophrenia (VEOS), developing before the age of 13 years, are considered more severe and uncommon variants of the adult-onset disorder; these clinical forms may be related to a greater vulnerability by reason of higher risk factors for schizophrenia and may be preceded by more relevant neurodevelopmental abnormalities than the adult onset form of the illness [1‐4].

The study of childhood-onset schizophrenia conducted by the National Institute of Mental Health (NIMH) revealed more severe premorbid neurodevelopmental abnormalities, a higher rate of cytogenetic anomalies, and a seemingly higher rate of familial schizophrenia spectrum disorders than in later onset cases; there was no evidence of increased obstetric complications or environmental stress [5, 6]. Vourdas et al. found that developmental deviance and premorbid abnormalities of social interaction and language-related functions tend to lead to a more precocious onset of schizophrenia [7].

On the other hand, information obtained by parents about risk factors for schizophrenia and early neurodevelopment should be more recent and more reliable for children and adolescents patients than adult onset patients; therefore retrospective studies could lead to more consistent data in case of EOS and VEOS rather than in case of adult onset schizophrenia.

In the present study we examined a sample of 21 patients with EOS and VEOS; the control group was composed of 21 non-psychiatric patients affected by migraine, supposed to have a different pathogenesis for the illness; comparison was made according to age and gender. The aim of the study was the following:

In accordance with neurodevelopmental studies, schizophrenia can be seen as a disorder with age-dependent clinical manifestations that in a minority of individuals start during childhood [23‐28]. Although the prevalence of EOS has not been adequately studied, the American Academy of Child and Adolescent Psychiatry suggest that EOS, and especially VEOS, are predominantly observed in males, with a ratio of approximately 2:1 and that with increasing age, this ratio tends to even out [1]. In our sample there were no significant sex differences as regards either prevalence or mean age of onset. Konnecke et al. have claimed that the difference in the age of onset between men and women is considerably reduced in the presence of a strong genetic vulnerability to schizophrenia and a history of pre and perinatal complications [29]. In 80% of our sample of children with EOS and VEOS we found a familial history of psychiatric disorders with a statistically relevant difference between cases and controls. This confirms the important role of familial loading as a risk factor for schizophrenia, being comparable to the data on adult onset forms [30‐32]. The first family study of childhood-onset schizophrenia was published by Asarnow et al; they found that relatives of probands with childhood-onset disease had an increased lifetime morbid risk for schizophrenia and schizotypal personality disorder as compared to the relatives of children and adolescents with attention deficit hyperactivity disorder and to the relatives of community comparison subjects [33]. Nicolson et al. found a greater morbid risk for schizophrenia spectrum disorders (schizophrenia, schizoaffective disorder, other nonaffective psychotic disorders, schizotypal personality disorder, paranoid personality disorder) in the parents of patients with childhood onset than with adult onset schizophrenia and both these groups were at a higher risk than the parents of community comparison subjects. Nevertheless, in their study, schizophrenia was an uncommon diagnosis in all three groups of parents; no diagnosis of either a schizoaffective disorder or another nonaffective psychotic disorder was present in any of the groups; moreover the parents of patients with childhood onset schizophrenia had a greater morbid risk for a schizotypal personality disorder than the other two groups and for a paranoid personality disorder than the parents of community comparison subjects [34]. In this study we found that the relatives of VEOS/EOS patients differed significantly from the relatives of control subjects in terms of family history of schizophrenia and related disorders and of personality disorders. As reported by Nicolson et al., in this sample too, the parents of VEOS/EOS patients had only diagnosis of personality disorders and never of schizophrenia and related disorders. As reported in other studies on familial liability for schizophrenia, we found that EOS/VEOS is associated with a specific increase in family history for schizophrenia and related disorders as well as for personality disorders, rather than general psychopathology. Furthermore, the different psychopathological expression between first and second degree relatives suggests that we should widen our knowledge about the personality of EOS/VEOS parents. As matter of fact psychopathological traits such as suspiciousness, withdrawal, social avoidance, introversion, diffidence, flattened affectivity are likely to account for the phenotypical expression relevant to familial vulnerability to schizophrenia, characterized by both genetic and environmental factors [35, 36].

While familial risk factors account for a significant rate of predisposition to schizophrenia, there is evidence of an important environmental contribution [17, 32]. Obstetric complications are among the most studied environmental indicators of risk for schizophrenia although discordant data have also been reported about the effective pathogenetic role in schizophrenia disease also because current methods of investigating the relationship between obstetric complication and schizophrenia are reaching the limit of their usefulness [37, 18, 19, 38, 22]. The consequences of obstetric complications lack diagnostic specificity according to the level of hypoxemic stress suffered and to the genetic predisposition of the foetus [14, 39]. There is poor consensus about the pathogenetic mechanisms through which pre and perinatal damage can favour the development of schizophrenia. Rosso et al. proposed a model whereby the neurotoxic effects of fetal hypoxia can trigger early onset of schizophrenia due to premature cortical synaptic pruning [21]. Our data show the presence of environmental risk factors in 55% of the sample, a higher proportion than the literature data on the frequency of obstetric complications in patients with adult onset schizophrenia (7-20% according to Boog G [14], 21.5-31.7% according to Nicolson et al. [6]). Other Authors have also supported this association between obstetric complication and an increased risk for early onset schizophrenia [15, 20‐22]. Moreover, most of the obstetric complications reported in our sample were correlated to fetal hypoxia (bleeding, placental abruption, threatened premature delivery, peripartum fetal hypoxia, emergency cesarean section, forceps delivery). This notwithstanding, as Ordonez et al. in 2005 [40], we didn't find either a relevant association between obstetric complications and VEOS/EOS when compared the patients to control subjects; then we can't support the role of obstetric complications as risk factors of schizophrenia, also when examined more specifically as pre-, peri- and postpartum complications. On the other hand the nature and strength of the association between obstetric complications and schizophrenia has been questioned yet [16] and the hypothesis that exposure to obstetric complications may interact with a genetic liability and increased the vulnerability to schizophrenia remains difficult to assess [39].

Studies of adult onset schizophrenia have demonstrated that developmental delay, early functional impairment and aspecific psychopathologic symptoms are often observed prior to the full emergence of psychotic symptoms [24, 41, 26, 27, 42, 43]. They may be considered as a part of a longitudinal psychotic phenotype in which some aspects are already established in early life; alternatively, the effects may reveal a greater vulnerability depending on exposure to further causes, but which may also be susceptible to protective factors. It is unclear whether neurodevelopmental abnormalities are on the increase in patients with early onset schizophrenia, nor whether do they act to precipitate the earlier onset of the disorder[24, 31, 7]. We found a significant statistical association between childhood developmental abnormalities and the risk for schizophrenia, particularly affecting relational skills and learning. Moreover, about one third of our sample showed a low IQ. The prevalence of mental disturbances is known to be approximately four-fold in mentally retarded subjects as compared to the general population, although the data on prevalence by single diagnostic category are less precise [44, 45]. It is generally agreed that cognitive deficits play an important role in the malfunctioning mechanism underlying the disorder [46]. Cognitive deficits have been documented in practically every domain, being most pronounced in the areas of memory, attention and executive functioning [47]. Low intelligence may be an independent risk factor for schizophrenia rather than the manifestation of a single underlying pathogenetic process.

This study confirms the significant role of familial liability in the pathogenesis of VEOS/EOS, and opens more specific questions about the phenotypic patterns of familial transmission. On the other hand the study doesn't confirm the association between obstetric complications pre, peri and postpartum and VEOS/EOS. Moreover the study shows a statistical association between childhood developmental abnormalities and EOS/VEOS, with particular reference to early relational skills and cognitive ability, in support of the neurodevelopmental model of disease.