Liability indicators aggregate many years before transition to illness in offspring descending from kindreds affected by schizophrenia or bipolar disorder.

Objectives: Offspring born to patients with affective and non-affective psychoses display indicators of brain dysfunctions that affected parents carry. Such indicators may help understand the risk trajectory. Methods: We followed up the clinical/developmental trajectories of 84 young offspring born to affected parents descending from the Quebec kindreds affected by schizophrenia or bipolar disorder. We longitudinally characterized childhood trajectories using 5 established risk indicators: cognitive impairments, psychotic-like experiences, non-psychotic DSM diagnosis and episodes of poor functioning, trauma and drug use. Results: Overall, offspring individually presented a high rate of risk indicators with 39% having 3 or more indicators. Thirty-three offspring progressed to an axis 1 DSM-IV disorder, 15 of whom transitioned to a major affective or non-affective disorder. The relative risks for each risk indicator were low in these vulnerable offspring (RR = 1.92 to 2.99). Remarkably, transitioners accumulated more risk indicators in childhood-adolescence than non-transitioners (Wilcoxon rank test; Z = 2.64, p = 0.008). Heterogeneity in the risk trajectories was observed. Outcome was not specific to parent's diagnosis. Conclusion: Young offspring descending from kindreds affected by major psychoses would accumulate risk indicators many years before transition. A clustering of risk factors has also been observed in children at risk of metabolic-cardiovascular disorders and influences practice guidelines in this field. Our findings may be significant for the primary care surveillance of millions of children born to affected parents in the G7 nations. Future longitudinal risk research of children at genetic risk should explore concurrently several intrinsic and environmental risk modalities to increase predictivity.

Replication of linkage with bipolar disorder on chromosome 16p in the Eastern Quebec population.

In a previous study [Maziade et al. (2005); Mol Psychiatry 10:486-499], we provided evidence for linkage (parametric lod score of 4.05) on chromosome 16p for bipolar affective disorder (BP) in 21 kindreds from Eastern Quebec, a population characterized by a founder effect. Using a stringent design, we performed a replication study in a second sample of 27 kindreds (sample 2) collected from the same population and assessed with the same methodologies as in our original sample (sample 1), that is with the same diagnostic procedure and using a common set of 23 markers studied with model-based (parametric) and model-free (nonparametric) linkage analyses. We replicated our initial finding with P values <0.001. Indeed, maximum NPL(all) scores of 3.7 and 3.52 were found at marker D16S3060 in sample 2 for the narrow and broad BP phenotype definition, respectively. For the latter definition, the nonparametric score reached 3.87 in the combined sample, a value that exceeded the maximum NPL score obtained in each individual sample (NPL(all) = 2.32 in sample 1; NPL(all) = 3.52 in sample 2). Moreover, a refined phenotype restricted to BP associated with psychosis yielded significant evidence for linkage in each individual sample (NPL(all) = 2.38 in sample 1; NPL(all) = 2.72) while yielding the best result (NPL(all) score = 3.90) in the combined sample (samples 1 and 2), despite an important reduction in the number of affected individuals. It is also noteworthy that the use of the refined phenotype provided a location of the maximum linkage peak shared by both samples, that is, at marker D16S668 in 16p13.12, suggesting consistency across samples. Our study provided one of the strongest pieces of evidence for linkage with BP in 16p and illustrated the heuristic potential of a replication study in a second sample ascertained from the same population and using homogeneous methodologies.

Chromosome 1q12-q22 linkage results in eastern Québec families affected by schizophrenia.

An impressive LOD score of 6.5 has recently been reported for schizophrenia on chromosome 1q21-22 in large families from eastern Canada [Brzustowicz et al., 2000: Science 288:678-682]. We did not reproduce such a finding in large pedigrees of eastern Québec based on seven markers spanning the 1p13-1q22 region and using both the models and phenotypes of Brzustowicz et al. and those used in our ongoing genome scan in 21 large French Canadian families. There was no significant total LOD scores in that chromosomal region (a maximum of 0.57) either for schizophrenia or bipolar disorders, nor any signal in individual large pedigrees. However, the samples of Brzustowicz et al. and ours differed in terms of their origins, the latter being of French ancestry and the former of Celtic and German descent. Population difference, genetic heterogeneity, and differences in ascertainment might explain the lack of replication. The result reported by Brzustowicz et al. cannot be discarded and should probably be considered as a susceptibility locus for a subset of the schizophrenic population.