Cognitive functioning in patients with schizophrenia and bipolar disorder: A quantitative review

Abstract

Objective

Evidence suggests that cognitive functioning in bipolar disorder may be impaired even in euthymic states, but it is unclear if the pattern of deficits is similar to the deficits found in schizophrenia. The aim of this study was to review quantitatively the studies on cognitive performance in schizophrenia and bipolar disorder.

Methods

Articles for consideration were identified through a literature search in MEDLINE and PsycLIT in the period between 1985 and October 2004, using the keywords “schizophrenia” combined with “bipolar disorder”, or “manic-depress*” or “manic” combined with “cogniti*” or “neuropsycholog*”. Thirty-one studies were included that: i) evaluated cognitive performance using standardized and reliable neuropsychological testing procedures; ii) compared adult patients with schizophrenia and with bipolar disorder; iii) reported test scores of both patient groups, or exact p-values, t-values, or F-values; and iv) were published as an original article in a peer-reviewed English language journal.

Results

Meta-analyses of all studies indicated that patients with bipolar disorder generally perform better than patients with schizophrenia, but the distribution of effect sizes showed substantial heterogeneity. Results based on a more homogeneous subset of studies that matched patient groups on clinical and demographic characteristics pointed in the same direction, with effect sizes in the moderate range.

Conclusions

Patients with bipolar disorder show better cognitive performance than patients with schizophrenia, even when matched for clinical and demographic characteristics.

Introduction

Accumulating evidence has cast doubt on the Kraepelinian notion (Kraepelin, 1913) that bipolar disorder, in contrast to schizophrenia, is not characterized by cognitive decline (Altshuler et al., 2004, Krabbendam et al., 2000, Martinez-Aran et al., 2004, Van Gorp et al., 1998). A review of 42 studies concluded that cognitive deficits are common in bipolar disorder, predominantly, though not exclusively, during episodes of mania and depression (Quraishi and Frangou, 2002). During recovery, deficits in the areas of verbal memory, sustained attention, and, less consistently, executive functions and visual memory persist (Quraishi and Frangou, 2002). Whether the pattern of cognitive deficits in bipolar disorder is comparable to the deficits observed in schizophrenia is still a matter of debate. Some authors have claimed that remitted patients with bipolar disorder perform notably better than stable patients with schizophrenia (Goldberg, 1999, Quraishi and Frangou, 2002) and that the cognitive profile in bipolar disorder is characterized by selective rather than generalized deficits (e.g., in the area of visual processing) (Goldberg, 1999, Green et al., 1994). Others have emphasized the similarities, stating that in both disorders the cognitive profile is characterized by a relatively generalized pattern of deficits, even though quantitative differences may exist (Hoff et al., 1990, Krabbendam et al., 2000, Martinez-Aran et al., 2004).

Any overlap in cognitive impairment in bipolar disorder and schizophrenia would be in line with many other findings. Although classification systems such as DSM-IV and ICD-10 attempt to distinguish between schizophrenia and bipolar disorder, it is a common observation in clinical practice that many patients do not fit properly into the dichotomy. The frequency of schizoaffective disorder testifies to that (Marneros, 2003, Tien and Eaton, 1992), as does the finding that the positive symptoms of psychosis and the mood symptoms of mania and depression frequently co-occur in the same patient (Brockington et al., 1980, Brockington et al., 1979, Siris, 2000). In addition, risk factors and epidemiological characteristics tend to overlap between schizophrenia and bipolar disorder. Developmental factors such as low educational achievement and delayed motor and language milestones increase the risk for both schizophrenia and affective disorders, although effect sizes tend to be greater for schizophrenia (Jones and Tarrant, 1999, Van Os et al., 1997). Similarly, social factors such as ethnicity and adverse life events affect the risk for affective disorders in particular, but also have their impact on the risk for schizophrenia (Van Os et al., 1998). Familial co-aggregation of schizophrenia and bipolar disorder suggest that the overlap may in part be based on a common genetic aetiology (Cardno et al., 2002, Maier et al., 2002). Recent studies have indeed implied shared susceptibility genes (Bramon and Sham, 2001, Wildenauer et al., 1999).

The apparent overlap between the two disorders in expression and aetiology has for more than a century fuelled spirited discussion as to whether schizophrenia and bipolar disorder are truly distinct diseases (Crow, 1998, Jablensky, 1999, Murray et al., 2004). To further our understanding of the relationship between the two disorders, it is crucial to identify the similarities and dissimilarities. The aim of this study was (i) to review quantitatively the studies on cognitive performance in schizophrenia and bipolar disorder and (ii) to examine whether the magnitude of any difference in performance is influenced by the clinical state of the study population.