Formal thought disorder in schizophrenia and bipolar disorder: A systematic review and meta-analysis

Abstract

Historically, formal thought disorder has been considered as one of the distinctive symptoms of schizophrenia. However, research in last few decades suggested that there is a considerable clinical and neurobiological overlap between schizophrenia and bipolar disorder (BP). We conducted a meta-analysis of studies comparing positive (PTD) and negative formal thought disorder (NTD) in schizophrenia and BP. We included 19 studies comparing 715 schizophrenia and 474 BP patients. In the acute inpatient samples, there was no significant difference in the severity of PTD (d = − 0.07, CI = − 0.22–0.09) between schizophrenia and BP. In stable patients, schizophrenia was associated with increased PTD compared to BP (d = 1.02, CI = 0.35–1.70). NTD was significantly more severe (d = 0.80, CI = 0.52–0.1.08) in schizophrenia compared to BP. Our findings suggest that PTD is a shared feature of both schizophrenia and BP but persistent PTD or NTD can distinguish subgroups of schizophrenia from BP and schizophrenia patients with better clinical outcomes.

Introduction

Formal thought disorder (FTD) has been considered a hallmark symptom of schizophrenia. In the early 20th century, schizophrenia was described by the presence of FTD (Solovay et al., 1987, Lake, 2008). In fact, Bleuler indicated the diagnostic importance of FTD in schizophrenia, introducing loosening of association as the pathognomonic symptom of the disorder. FTD has predictive value over prognosis in schizophrenia (Andreasen and Grove, 1986, Wilcox et al., 2012) and is regarded as a marker of illness severity (Roche et al., 2015). The persistent presence of FTD refers to unfavorable prognosis and is accepted as a strong predictor of relapse (Wilcox, 1990, Wilcox et al., 2000). In particular, negative formal thought disorder (NTD) was associated with poor response to treatment (Cuesta et al., 1994) and related to a chronic and persistent course of illness (Andreasen and Grove, 1986, Wilcox et al., 2012). The presence of FTD at the onset of illness was reported to increase relapse rate (Liddle and Barnes, 1990). Furthermore, FTD was found to be the strongest predictor determining conversion from first-episode acute transient psychotic disorder to schizophrenia (Rusaka and Rancans, 2014). Poverty of thought present in the prodromal phase was stated to be predictive of conversion to psychosis and future deterioration (Wilcox et al., 2014). FTD was also associated with occupational and social functioning (Roche et al., 2016) and satisfaction with life (Tan et al., 2014).

Studies on cognitive deficits (Harvey et al., 1990, Bora et al., 2010), neurodevelopmental defects (Bora, 2015, Arango et al., 2014), imaging (Ellison-Wright and Bullmore, 2010, Bora et al., 2012, Hulshoff Pol et al., 2012) and genetics (Lichtenstein et al., 2009, Purcell et al., 2009, Smoller et al., 2013) show considerable similarities between schizophrenia and bipolar disorder (BP). In addition to the overlap of almost all the associated symptoms, psychotic symptoms are also shared features of both disorders during acute episodes (Arango et al., 2014). Similar overlapping findings on FTD in schizophrenia and BP have also been documented. FTD was found to be significantly more frequent in both schizophrenia and mania than in any other psychotic disorders (Marengo and Harrow, 1985). In nonschizophrenic psychoses, FTD was found to be present along a spectrum of severity (Levy et al., 2010, Cuesta and Peralta, 2011). Differences in speech disorganization between patients with schizophrenia and mania were revealed only in certain studies in which the sample consisted of patients with severe FTD (Hoffman et al., 1986, Cuesta and Peralta, 2011). Considering these similarities, Lake (2008) suggested that symptoms of FTD in schizophrenia could be described by manic distractibility and argued that schizophrenia may be the same disease as psychotic mood disorder. Other authors argued that major psychoses including schizophrenia and BP can be best described as a continuum rather than being discrete disorders (Crow, 1990, Cloninger, 1994). It is also important to note that diagnostic criteria used in studies can have influence the level differences in FTD between schizophrenia and BP. For example, many patients with a diagnosis of BP with mood incongruent psychotic features in DSM (Diagnostic and Statistical Manual of Mental Disorders) would receive a RDC (Research diagnostic criteria) diagnosis of schizoaffective disorder.

FTD in schizophrenia and BP have been compared using operationalized clinical scales since Andreasen, 1979b. However, these studies yield inconsistent findings. Some research reveals more global FTD in schizophrenia than in mania (Manoach, 1994, Kravati et al., 2005) and some found no significant difference in FTD between the two diagnostic groups (Harvey et al., 1986, Solovay et al., 1987, Daniels et al., 1988, Khadivi et al., 1997, Wilcox et al., 2012, Jamadar et al., 2013, Mancuso et al., 2015). Manoach (1994) indicated that patients with BP had a greater range of FTD scores compared to patients with schizophrenia. Regarding subtypes of FTD, it has been indicated that patients with mania tend to have more explicit positive formal thought disorder (PTD) in contrast to patients with schizophrenia who manifest significantly more NTD (Harvey et al., 1984, Andreasen and Grove, 1986, Walker and Harvey, 1986, Docherty et al., 1988, Kircher et al., 2014). However, in some studies, patients with mania and schizophrenia were found to manifest equally severe PTD (Andreasen and Grove, 1986, Harvey et al., 1988, Serper, 1993), particularly at acute phases (Harrow et al., 1982, Marengo and Harrow, 1985). Inconsistencies are also seen in a limited number of longitudinal studies. In some follow-up studies, PTD was reported to be more stable in mania, while NTD is more consistent throughout the course of schizophrenia (Harvey et al., 1984, Harvey et al., 1990, Andreasen and Grove, 1986). On the contrary, Earle-Boyer et al. (1986) reported that PTD (incoherence) remained consistent across consecutive admissions in mania whereas both NTD (poverty of speech) and PTD (illogicality) remained stable in schizophrenia.

Some of the inconsistent findings of studies comparing FTD in schizophrenia and BP might be related to low statistical power of individual studies as most of the available studies have small sample sizes. A meta-analysis can be helpful to increase statistical power and provide an estimate of the level of differences in FTD between schizophrenia and BP. The primary goal of this meta-analysis was to synthesize the results of available studies comparing FTD in schizophrenia and BP, and to explore the relationship between the diagnostic specificity of FTD and state of the illness.