Aetiology and risk factors - Schizophrenia

Pregnancy and birth complications

Pregnancy and birth complications (PBCs) have been implicated as a risk factor in schizophrenia. In a metaanalysis of 20 case - control studies, subjects who developed schizophrenia were more than twice as likely to have been exposed to PBCs as controls (Geddes & Lawrie 1995). However, the findings from two other large case - control studies suggest that the link between schizophrenia and PBCs may be much weaker than previously assumed (Kendall et al. 2000).

In one case - control study of childhood onset schizophrenia, Matsumoto et al. (1999) reported an odds ratio of 3.5 for PBCs, suggesting a greater risk in very early onset cases. However, in the NIMH study of childhood onset schizophrenia, PBCs were no more common in cases than in sibling controls (Nicholson et al. 1999). Even if there is a significant association, it remains unclear whether there is any causal connection between PBCs and schizophrenia. There is a strong argument that PBCs are consequences rather than causes of abnormal neurodevelopment (Goodman 1988). This view is supported by the finding that schizophrenics have smaller head size at birth than controls (McGrath & Murray 1995), which is likely to be a consequence of either defects in genetic control of neurodevelopment or earlier environmental factors, such as viral exposure.

Puberty

The close temporal association between the onset of puberty and a marked increase in the incidence of schizophrenia suggests that biological (or social) events around puberty may be related to the expression of psychotic symptoms. Galdos et al. (1993) reported an association between the timing of menarche and onset of psychosis in girls. However, this finding has not been supported by subsequent studies. Frazier et al. (1997) found no relationship between the onset of psychosis and indices of puberty in cases recruited to the NIMH study of childhood onset schizophrenia. For both boys and girls the timing of pubertal events were similar for cases and controls.

Psychosocial risks

One possible explanation for an atypically early onset of schizophrenia could be differential exposure to psychosocial adversity, such as higher levels of parental hostility and criticism ('high EE'). High levels of expressed emotion (EE) among relatives of adult schizophrenics has been shown to be a predictor of psychotic relapse and poor outcome (Leff & Vaughn 1985).

Although the role of high EE in precipitating the onset of schizophrenia has not been established, it is theoretically plausible that high EE might act to 'bring forward' the onset of the disorder in a vulnerable individual. Goldstein (1987) reported that parental EE measures of criticism and overinvolvement taken during adolescence were associated with an increased risk of schizophrenia spectrum disorders in young adulthood. However, a causal link is not proven, and the association may reflect either an expression of some common underlying trait or a parental response to premorbid disturbance in the preschizophrenic adolescent. Similarly, high EE in the parents of schizophrenic patients may be a reaction to the illness rather than a cause. There is no evidence that the parents of childhood onset patients show higher levels of EE than do parents of adult onset patients - in fact the reverse may be true. J.R. Asarnow et al. (1994) used the Five Minute Speech Sample to measure parental EE and found that childhood onset schizophrenics were no more likely to have 'high EE' parents than normal controls. Overall, there seems little evidence to suggest that the onset of schizophrenia in childhood can be explained by exposure to higher levels of parental criticism and hostility than that experienced by adult onset patients. Indeed, it appears that, on average, the parents of childhood onset schizophrenics generally express lower levels of criticism and hostility than parents of adult onset patients because of a greater tendency to attribute their childrens' behaviour to an illness which is beyond their control (Hooley 1987).