ANTICIPATION OR ASCERTAINMENT BIAS IN SCHIZOPHRENIA - PENROSES FAMILIAL MENTAL-ILLNESS SAMPLE

Several studies have observed anticipation (earlier age at onset [AAO] in successive generations) in familial schizophrenia. However, whethe r true anticipation or ascertainment bias is the principal originating mechanism remains unclear. Zn 1944 L. S. Penrose collected AAO data o n a large, representative sample of familial mental illness, using a b road ascertainment strategy. These data allowed examination of anticip ation and ascertainment biases in five two-generation samples of affec ted relative pairs. The median intergenerational. difference (MID) in AAO was used to assess anticipation. Results showed significant antici pation in parent-offspring pairs with schizophrenia (sl = 137 pairs; M ID 15 years; P = .0001) and in a positive control sample with Huntingt on disease (n = 11; P = .01). Broadening the diagnosis of the schizoph renia sample suggested anticipation of severity of illness. However, o ther analyses provided evidence for ascertainment bias, especially in later-AAO parents, in parent-offspring pairs. Aunt/uncle-niece/nephew schizophrenia pairs showed anticipation (It = 111; P = .0001), but the MID was 8 years and aunts/uncles had earlier median AAO than parents. Anticipation effects were greatest in pairs with late-AAO parents but remained significant in a subgroup of schizophrenia pairs with early parental AAO (n = 31; P = .03), A small control sample of other diseas es had MID of 5 years but no significant anticipation (n = 9; P = .38) . These results suggest that, although ascertainment-bias effects were observed in parent-offspring pairs, true anticipation appears to be i nherent in the transmission of familial schizophrenia. The findings su pport investigations of unstable mutations and other mechanisms that m ay contribute to true anticipation in schizophrenia.