Latent Structure of Negative Symptoms in Schizophrenia Best Conceptualized by 5-Domain Model
Existing 2-dimensional models are likely not sufficient to capture the complexity of negative symptoms in schizophrenia.
The latent structure of negative symptoms in schizophrenia is best described by a 5-factor model, according to study data published in JAMA Psychiatry. This result challenges the existing 2-dimensional conceptualization of negative symptoms.
Prior studies using exploratory factor analyses have characterized negative symptoms using 2 dimensions: motivation and pleasure (MAP) and expressivity (EXP). To evaluate the validity of this model, investigators conducted 3 cross-sectional studies on people receiving outpatient treatment for schizophrenia. Patient symptomatology was captured using 3 separate “contemporary” measures: the Scale for the Assessment of Negative Symptoms (n=268), the Brief Negative Symptom Scale (n=192), and the Clinical Assessment Interview for Negative Symptoms (n=400).
Confirmatory factor analysis was then conducted on each scale to assess fit with 4 proposed models of latent structure. The 4 models were described as followed: (1) a 1-factor model; (2) a 2-factor model with MAP and EXP factors; (3) a 5-factor model with distinct factors for each of the 5 domains from the National Institute of Mental Health consensus development conference; and (4) a hierarchic model with 2 second-order factors reflecting EXP and MAP and 5 first-order factors reflecting the 5 consensus domains. The consensus domains included blunted affect, alogia, anhedonia, avolition, and asociality.
The 1- and 2-factor models provided poor fit for all 3 negative symptom scales for negative symptoms, with comparative fit indexes and Tucker Lewis indexes below the 0.950 threshold, root mean square error of approximations (RMSEAs) above the 0.080 threshold, and weighted root mean square residuals (WRMRs) greater than 1.00. By contrast, the 5-factor and hierarchic models provided “excellent” fit, with comparative fit indexes and Tucker Lewis indexes meeting the 0.950 and 1.00 thresholds for all 3 symptom scales. Notably, the RMSEAs for the 5-factor and hierarchic models fell below the 0.080 threshold for the Brief Negative Symptom Scale and the Clinical Assessment Interview for Negative Symptoms specifically, although not the Scale for the Assessment of Negative Symptoms. Given the success of the hierarchic model, researchers suggested that the 5 consensus domains, as first-order factors, may more properly “account for latent structure” than MAP and EXP dimensions. Given the RMSEAs values for the 5-factor and hierarchic models, researchers also suggested that the Scale for the Assessment of Negative Symptoms may be less effective for measuring symptomatology compared with the other 2 scales.
Future research is necessary to explore the validity of the 5 aforementioned domains in describing schizophrenia symptomatology. These data, however, suggest that existing 2-dimensional models are likely not sufficient in capturing the complexity of negative symptoms in schizophrenia.
Strauss GP, Nuñez A, Ahmed AO, et al. The latent structure of negative symptoms in schizophrenia [published online September 12, 2018]. JAMA Psychiatry. doi:10.1001/jamapsychiatry.2018.2475