Negative symptoms and neurocognitive deficits in first-episode schizophrenia spectrum disorder

Abstract

Background: Negative symptoms and neurocognitive deficits in schizophrenia are found to be correlated cross-sectionally but their longitudinal relationship remains unclear. Most studies focused on chronic illness and negative symptoms were examined as single construct instead of its subdomains. We aimed to investigate relationship between diminished expression and neurocognition in patients presenting with first-episode schizophrenia utilizing a prospective design. Methods: Ninety-three subjects (42 males, 51 females, mean age: 31.2) with first-episode schizophrenia, schizoaffective disorder or schizophreniform disorder were followed up for 3 years during which regular assessment in neurocognition and symptoms were conducted. Neurocognitive assessments included Wisconsin Card Sorting Test (WCST), digit span, verbal fluency, visual reproduction and logical memory. Cambridge Neurological Inventory (CNI)and PANSS were administered to evaluate neurological soft signs and clinical symptoms respectively. Severity of Diminished Expression (DE) was sum of item scores in Affect, Behaviour and Speech subscales of High Royds Evaluation of Negativity Scale (Mortimer et al. 1989). Results: Neurocognitive deficits were correlated with severity of DE on a cross-sectional basis. When sex, educational level, DUP, chlorpromazine equivalent, positive symptoms, drug-induced Parkinsonism and depression were controlled, regression analysis showed that at year 1, DE was significantly associated with verbal fluency (R square = .12, P < .01); at year 2, DE was predictive of WCST perseverative error (R square = .12, P < .001), WCST categories completed (R square = .10, P < .001), visual reproduction (R square = .16, P = .01) and CNI-motor coordination (R square = .18, P < .05); at year 3, DE was significantly associated with verbal fluency (R square = .09, P < .05) and forward digit span (R square = .05, P <.05). Longitudinally, change ofDE rating (and PANSS negative symptoms score) was uncorrelated with neurocognitive change over 3 years. Baseline DE was not associated with neuropsychological performance at year 3. Conclusion: Neurocognitive deficits and DE were correlated in successive assessments but longitudinal association between changes of these 2 domains in first-episode schizophrenia could not be demonstrated. Our findings indicated that negative symptoms and cognitive deficits are separable illness domains with potentially related underlying neurobiological abnormalities.