Relationship of premorbid functioning with negative symptoms and cognition in first-episode schizophrenia: a three-year prospective follow-up study

Abstract

Background: Premorbid adjustment is an important prognostic factor of schizophrenia. Numerous previous studies examined premorbid functioning as a unitary construct without further delineation into its sub-domains. Relatively few studies focused on first-episode population and the follow up periods were mostly within 1 to 2 years. We aimed to investigate the differential relationships of premorbid social and academic domains with symptoms and cognitions in first-episode schizophrenia-spectrum disorder over 3 years utilizing a prospective design. Methods: Ninety-three Hong Kong Chinese aged 18 to 55 years presenting with first-episode DSM-IV schizophrenia, schizophreniform disorder or schizoaffective disorder were examined using Premorbid Adjustment Scale (PAS), IRAOS, PANSS, HEN1 and a battery of cognitive assessments. Psychopathological and cognitive evaluations were conducted at baseline, clinical stabilization after first-episode, 12, 24 and 36 months. Subjects were categorized into three groups based on the classification method of Haas & Sweeny (1992). Following the same procedure, yet due to small size of deteriorating group, subjects were divided into two groups, namely good and poor premorbid groups (combining stable-poor & deteriorating groups) with respect to academic and social premorbid functioning. Results: There was no significant difference among PAS groups in sociodemographics, duration of untreated psychosis (DUP), baseline symptomatology and cognitive functions, and treatment characteristics. Repeatedmeasures ANOVAs (5 time points × 3 PAS groups) revealed significant between-subject main effect of premorbid adjustment typology on total negative symptom score (HEN total score; F=4.1, p<0.05) and HEN Diminished-Expression2 (DE, F=4.4, p<0.05), but not HEN Avolition-Asociality (AA)2 scores and cognitions over 3 years. Concerning social functional dimension, repeated-measures ANOVAs (5×2 groups) showed that poor-premorbid group had significantly higher HEN total (F=5.2, p<0.05), DE (F=4.5, p<0.05) and AA (F=5.5, p<0.05) scores than goodpremorbid group, and the results remained significant even when DUP was added as covariate for further analysis. No significant differences between two groups in cognitions were observed. Regarding academic dimension, good-premorbid group was associated with female sex (chi-sqaure=5.1, p<0.05), higher educational attainment (t=1.7, p<0.01), and superior performnace on verbal fluency (F=3.9, p=0.05), logical memory (F=4.2, p<0.05) and WCST perseverative error (F=5.8, p<0.05) over 3 years as revealed by repeated-measures ANOVAs (5×2 groups). Results on WCST perseverative error remained significant when sex, educational level and DUP were controlled for. No significant difference on negative symptom level between two groups was noted. Discussion: In a Chinese cohort of first-episode schizophrenia-spectrum disorder, we found that premorbid social functioning was associated with severity of negative symptoms while poorer academic adjustment was related to male sex, lower scholatsic achievement and worse cognitive functioning. The current findings thus lend support to the differential patterns between social and academic dimensions of premorbid functioning in relation to demographics, symptom manifestations and cognitive impairment in the initial 3 years of first-episode schizophrenia.