Some fairly obvious distinctions between schizophrenia and bipolar disorder

Abstract

In examining the ratio of ‘famous’ individuals with bipolar disorder to those with schizophrenia, it is evident that the ratio greatly favors those with bipolar disorder, suggesting that schizophrenia somehow precludes expert performance or high-level accomplishment. Why might this be so? One possible explanation lies in neurocognitive findings. In this review, I suggest that patients with schizophrenia perform consistently worse than patients with bipolar disorder on a variety of higher level cognitive tasks (though bipolar patients themselves often evince impairment in comparison to healthy controls). Working memory, in particular, appears to discriminate between the groups. However, questions remain, including the effects of state and medications, and the conclusion that I reach does not exclude the possibility that the two disorders share some but not all etiologic or pathophysiologic features.

Introduction

Differences between schizophrenia and bipolar disorder begin with the phenomenological, but they do not end there. With reference to symptoms, mania generally includes activation and increases in goal-directed activity usually not observed in schizophrenia; depression in bipolar disorder can be distinguished from negative symptoms in schizophrenia usually on the basis of subjective feelings of sadness. More significantly still, the course in bipolar illness is usually episodic, with periods of adequate function alternating with periods of impaired function. The course in schizophrenia is marked by relatively more continual impairment in function. Not unexpectedly, outcome also differs between the groups. Two large, methodologically sound studies have concluded that schizophrenia is associated with worse long-term outcome than bipolar disorder (Marneros et al., 1990, McGlashan, 1984).

There are a number of other dimensions in which patients with schizophrenia differ from patients with bipolar disorder. At the anecdotal level, the number of individuals with schizophrenia or bipolar disorder reaching expert performance in a given field (excluding criminality) in the 20th century can be examined. As gleaned from several Internet sites, the ratio of famous individuals who suffered from bipolar disorder to individuals who suffered from schizophrenia is approximately 7:1, which compares to prevalence figures of about 0.8% for schizophrenia and 1.0% for bipolar disorder. This suggests that somehow schizophrenia limits high-level intellectual performance.

Furthermore, the first-line medications used to treat symptoms of these disorders are very different. Mood stabilizers, including lithium, carbamazepine, and valproate, are used to treat bipolar disorder. While their precise mechanism for therapeutic action is unclear, they may have effects on the intracellular G protein second-messenger system or stabilize membrane ion channels. In contrast, the mode of action of neuroleptics in the treatment of schizophrenia is typically attributed to the blockade of D2 receptors. However, it is the case that the medications are used adjunctively: neuroleptics are used to ameliorate psychotic symptoms in manic states; mood stabilizers are used to reduce affective lability or activation in schizophrenia. (Other domains, such as structural and functional neuroimaging and neuropathology, will be covered by other papers in this issue.)

Nevertheless, there are several lines of reasoning that support the conclusion that schizophrenia and bipolar disorder are on a continuum. First, there is overlap in phenomenological symptoms. Thus, patients with bipolar disorder can experience frank psychotic symptoms, including hallucinations, delusions, and thought disorder. There is some overlap in susceptibility markers ascertained in linkage studies, so that both bipolar illness and schizophrenia can be linked to chromosomes 22q. However, there are susceptibility markers found only in schizophrenia, namely on 6p and 8p, while bipolar-only markers are found on 21q, 18p, and 18q (Moldin, 1999)