Symptom structure of acute mania: A factor study of the 24-item Brief Psychiatric Rating Scale in a national sample of patients hospitalized for a manic episode

Abstract

Background

Despite increasing interest in dimensional psychopathology and the use of symptom clusters in clinical research, factor analytic studies of mania are rare. Most studies included not only manic patients, but also patients with a mixed episode or other severe mental disorders. We aimed at further elucidating the symptom structure of manic states.

Methods

As part of a national survey of acute psychiatric inpatient care, all patients admitted to a random sample of Italian public and private facilities during an index period underwent a standardized assessment, including the 24-item Brief Psychiatric Rating Scale (BPRS-24). Eighty-eight patients (90% of all manic patients admitted) with an ICD-10 diagnosis of Bipolar Affective Disorder, Current Episode Manic with complete data were included in this study. Principal axis factor analysis with Varimax rotation was performed on BPRS-24 items.

Results

Four factors were extracted, explaining 51% of total variance. They were interpreted as Mania, Disorganization, Positive Symptoms, and Dysphoria. The distribution of the Disorganization factor was positively skewed, with most patients relatively free from disorganization symptoms and some patients showing varying degrees of severity.

Limitations

The sample size was relatively small; also, patients were not administered a structured diagnostic interview. However, reasonably large samples are usually sufficient when communalities are high. Also, the manic episode is a clear-cut diagnostic entity easily identified by experienced clinicians, and the independent BPRS-24 ratings corroborated the diagnosis.

Conclusions

The identification of a Mania, Positive Symptoms, and Dysphoria factor is consistent with most previous studies. The identification of a Disorganization factor in a sample including only manic patients is a new finding that may have clinical implications, as its distribution suggests the possibility of distinguishing two patient groups, which may require different interventions to achieve optimal therapeutic response. The factorially derived BPRS-24 subscales may be useful for evaluation of treatment effects in clinical trials of antimanic agents.

Introduction

Recently, dimensional psychopathology has attracted increasing interest (Van Praag et al., 1990, Goldberg, 2000). This approach emphasizes the variety of clinical pictures in patients sharing the same categorical diagnosis, which is ascribed to the different weight in each patient of a small number of symptom clusters. While factor analysis has often been used to investigate the symptom structure of schizophrenia (Blanchard and Cohen, 2006, Peralta and Cuesta, 2001) and depressive disorders (Biondi et al., 2005), factor analytic studies of mania are relatively rare.

Several investigations were performed on heterogeneous samples including manic patients and patients with other severe mental disorders (Lindenmayer et al., 2004, Maziade et al., 1995, McIntosh et al., 2001, Peralta et al., 1997, Serretti et al., 2001, Serretti and Olgiati, 2004, Toomey et al., 1998, Ventura et al., 2000). However, they mainly aimed at identifying common factors shared by psychotic disorders, rather than elucidating the specific symptom structure of mania.

Few studies included only patients with bipolar disorder (Beigel and Murphy, 1971, Cassidy et al., 1998, Daneluzzo et al., 2002, Dilsaver et al., 1999, Double, 1990, Gonzalez-Pinto et al., 2003, Perugi et al., 2001, Rossi et al., 2001, Sato et al., 2002, Swann et al., 2001). While they contributed to the understanding of the psychopathology of mania, most of them have limitations such as small sample size (Beigel and Murphy, 1971, Double, 1990), inclusion of patients with mixed episode (Cassidy et al., 1998, Daneluzzo et al., 2002, Dilsaver et al., 1999, Gonzalez-Pinto et al., 2003, Sato et al., 2002), and use of assessment instruments specifically designed for mania (Beigel and Murphy, 1971, Cassidy et al., 1998, Double, 1990) or affective disorders (Gonzalez-Pinto et al., 2003, Rossi et al., 2001). Given that disorder-specific instruments are relatively narrow in scope and cannot cover the full spectrum of psychopathology, factorial solutions derived from such instruments may not depict the whole psychopathological picture.

This study was performed within the PROGRES-Acute project, a national survey of psychiatric inpatient care in Italy (de Girolamo et al., 2007, Gigantesco et al., 2007). The availability of psychometric data from a national sample of patients hospitalized for a manic episode provided an opportunity to further investigate the symptom structure of mania.