F73. COGNITIVE CLUSTERING IN SCHIZOPHRENIA PATIENTS, THEIR FIRST-DEGREE RELATIVES AND HEALTHY SUBJECTS IS ASSOCIATED WITH ANTERIOR CINGULATE CORTEX VOLUME

Abstract Background Cognitive impairments are a core feature in schizophrenia patients and are also observed in first-degree relatives of the schizophrenia patients. However, substantial variability in the impairments exists within and among schizophrenia patients, first-degree relatives and healthy controls. A cluster-analytic approach can group individuals based on profiles of traits and create more homogeneous groupings than predefined categories. Methods Here, we investigated differences in the Brief Assessment of Cognition in Schizophrenia (BACS) neuropsychological battery (six subscales) among 81 schizophrenia patients, 20 unaffected first-degree relatives and 25 healthy controls. To identify three homogeneous and meaningful cognitive groups regardless of categorical diagnoses (schizophrenia patients, first-degree relatives and healthy controls), cognitive clustering was performed using a k-means clustering analysis approach, and differences in the BACS subscales (verbal memory, digit sequencing, token motor, verbal fluency, symbol coding and Tower of London) among the cognitive cluster groups were investigated. Finally, the effects of diagnosis and cognition on brain volumes were examined. Results As expected, there were significant differences in the five BACS subscales among the diagnostic groups (verbal memory, F2,123=20.6, P=1.90 × 10–8; digit sequencing, F2,123=8.0, P=5.65 × 10–4; token motor, F2,123=16.0, P=6.92 × 10–7; verbal fluency, F2,123=14.8, P=1.79 × 10–6 and symbol coding, F2,123=28.8, P=5.64 × 10–11). The cluster-analytic approach generated three meaningful subgroups: (i) neuropsychologically normal (Cluster 1, N=36), (ii) intermediate impaired (Cluster 2, N=60) and (iii) widespread impaired (Cluster 3, N=30). The cognitive subgroups were mainly affected by the clinical diagnosis (χ2=46.7, P=5.33 × 10-10), and significant differences in all BACS subscales among clusters were found (verbal memory, F2,123=64.1, P=8.49 × 10–20; digit sequencing, F2,123=35.7, P=5.89 × 10–13; token motor, F2,123=71.7, P=2.29 × 10–21; verbal fluency, F2,123=84.2, P=9.05 × 10–24; symbol coding, F2,123=115.6, P=5.70 × 10–29 and Tower of London, F2,123=6.9, P=1.43 × 10–3). The effects of the diagnosis (SCZ