Six sets of operational criteria for diagnosing schizophrenia were applied to a systematically ascertained twin series by raters who were blind to zygosity and to the psychiatric status of the co-twin. Assuming a multifactorial/threshold model of transmission, twin correlations in liability and, where possible, approximate broad heritabilities were calculated for each criterion. All definitions resulted in significant monozygotic twin correlations. The highest heritabilities (of approximately 0.8) were given by the Research Diagnostic Criteria and by the categories “probable” plus “definite” schizophrenia according to the criteria of Feighner et al. In contrast, Schneider's first-rank symptoms defined a form of schizophrenia with a heritability of 0 and, together with the criteria of Carpenter et al and Taylor et al, proved to be excessively restrictive, identifying fewer than half of the probands as schizophrenic.
|Original language||English (US)|
|Number of pages||5|
|Journal||Archives of General Psychiatry|
|State||Published - Jun 1984|