Structural brain abnormalities have been reported in schizophrenia. We tested the hypothesis that these abnormalities represented a marker for the genetic liability to schizophrenia in a sample of people with schizophrenia and their relatives from families multiply affected with the disorder.

We compared 31 people with schizophrenia, 57 relatives and 39 unrelated control subjects. Volumetric measurement of brain structures was carried out using stereological principles from three-dimensional reconstructed magnetic resonance images.

Subjects with schizophrenia had larger lateral ventricles than their relatives and the normal control subjects. Relatives who were ‘presumed obligate carriers' had larger left lateral ventricles than other relatives and the control subjects. Subjects with schizophrenia showed smaller whole brain and cerebellar volumes and larger lateral ventricles than their age–and gender-matched unaffected siblings.

In families multiply affected with schizophrenia lateral ventricular enlargement distinguishes people with schizophrenia and presumed obligate carriers from other relatives and unrelated control subjects. These changes may be a marker for a genetic liability to schizophrenia.

There have been several reports of linkage between genetic markers on the X chromosome at Xq26.3-28 and bipolar affective disorder in family samples obtained from distinct ethnic and geographical origins. As part of a genome search in a series of 23 UK and Icelandic families, specifically selected for their large size and power to resolve the issue of linkage heterogeneity, we have tested the hypothesis that there is a locus for a genetic subtype of bipolar affective disorder which is linked to this region.

In families selected on the basis of absent male to male transmission for affective disorder, we performed two-point and FASTMAP multipoint linkage analyses with markers spanning the region between the genetic loci DXS102 and F8

We found negative lod scores for several models of affection status in families selected under stringent and relaxed criteria for the absence of male to male transmission.

In the family sample we have obtained, our study provides no support for the presence of a locus increasing genetic susceptibility to bipolar affective disorder in this region of the X chromosome. It is likely that our finding reflects heterogeneity of linkage for bipolar and genetically related unipolar disorder that exists in specific ethnic populations. Alternatively the X-linked subtype of the disorder may have been present only in a few of our small families resulting in loss of power to detect the Xq26.3-28 linked subtype.

A susceptibility locus for schizophrenia in the pseudoautosomal region has been proposed on the basis of a possible excess of sex chromosome aneuploidies among patients with schizophrenia and an increased sex concordance in affected sib pairs. Several studies investigating this hypothesis have produced conflicting evidence.

In a series of Icelandic and British families, we used lod score and sib pair linkage analyses with markers for the MIC2 and DXYS14 loci on the pseudoautosomal XY region.

Lod and sib pair linkage analysis with these markers produced strongly negative scores. Heterogeneity testing also produced negative results.

We conclude that the present study provides no support for the involvement of either the pseudoautosomal region or the nearby region of the sex chromosomes in the aetiology of schizophrenia.