Meijer J, Simons CJ, Quee PJ, Verweij K,
Acta Psychiatr Scand 2012 Jan;125(1):66-76
OBJECTIVE: The purpose of this study was to examine a range of cognitive measures as candidate phenotypic liability markers for psychosis in a uniquely large sample of patients with psychosis, their unaffected relatives and control subjects.
METHOD: Patients with non-affective psychosis (n = 1093), their unaffected siblings (n = 1044), parents (n = 911), and controls (n = 587) completed a comprehensive cognitive test battery. Cognitive functioning was compared using tests of verbal learning and memory, attention/vigilance, working memory, processing speed, reasoning and problem solving, acquired knowledge, and social cognition. Age- and gender-adjusted z-scores were compared between groups using mixed-model analyses of covariance. Clinically relevant impairment (-1 and -2 SD from control mean) was compared between subject groups.
RESULTS: Patients performed significantly worse than controls in all cognitive domains (z-range -0.26 to -1.34). Siblings and parents showed alterations for immediate verbal learning, processing speed, reasoning and problem solving, acquired knowledge, and working memory (z-range -0.22 to -0.98). Parents showed additional alterations for social cognition. Prevalence of clinically relevant impairment in relatives ranged from 50% (-1 SD criterion) to 10% (-2 SD criterion).
CONCLUSION: Cognitive functioning is a candidate intermediate phenotype given significant small to large alterations in patients and intermediate alterations in first-degree relatives.