Utility of Current Diagnostic Entities

The utility of current diagnostic entities for genetic studies seems limited, since psychiatric diagnoses incompletely cover the patients’ prodromal phase, age-of-onset, severity of illness, long-term course, and impairment. In particular, psychiatric diagnoses do not offer much information about a patient’s response to treatment and prognosis. Quantitative, syndrome-oriented approaches to psychopathology extend the DSM-IV or ICD-10 phenotype definitions by replacing the dichotomy of the diagnostic schema by dimensional quantities. These quantities allow one to assess intra-familial patterns of psychopathology that do not reach diagnostic significance. Our syndrome-oriented approach to psychopathology includes 16 syndromes and leads to a representation of psychopathology patterns via 16-dimensional quantitative "syndrome vectors" that are extracted from the DIGS and SSCL16 rating instruments.

Quantitative Syndrome Patterns

In a prospective family study of 269 index cases suffering from functional psychoses, 350 affected first- and second-degree relatives, a 20-year follow-up of the index cases, and 105 offspring of the index cases, we applied a multivariate syndrome-oriented approach to psychopathology in order to derive a quantitative measure of the severity of illness and to partition the population into "natural" subgroups. Our main interest was focused on: (1) the question of increasing severity of psychoses across generations (anticipation), (2) the question of differences in the severity of psychoses depending on whether the illness has been transmitted by the maternal or paternal side (genomic imprinting), and (3) the question of age-of-onset shifts toward earlier onset of psychoses in successive birth cohorts (secular trends). Structural analyses revealed clearly distinguishable subgroups of patients characterized by differences in the familial aggregation of syndromes and in the longterm outcome [Stassen et al. 1988; Scharfetter and Stassen 1995; Scharfetter et al. 1999, 2002].

Familial Aggregation Rather than Segregation

We carried out multivariate cluster analysis based on the 16-dimensional, quantitative syndromes derived from the 269 index cases with a diagnosis of functional psychosis, on the one hand, and from the 350 first-degree relatives who exhibited psychopathologic features in our quantitative SSCL-16 syndrome scores, on the other. Our search for "natural" groupings revealed three "core" clusters which were similarly present in the two populations under investigation. There was Striking similarity between mean quantitative syndrome patterns derived by averaging across index cases and mean quantitative syndrome patterns derived by averaging across "affected" first-degree relatives. Despite this striking similarity on the group level, within-family comparisons of syndrome patterns demonstrate that the familial aggregation of syndromes does not occur in a homotypic way, i.e., family data do not provide evidence for genetic segregation. This finding is specifically supported by twin samples ascertained through co-twins suffering from schizophrenic disorders, where highly significant deviations from the mz:dz ratio of 2 indicate the existence of strong non-linearities in the genetic predisposition to functional psychoses.

References