Schizophrenia and Bipolar Disorder: Differences and Overlaps

Wolfgang Maier; Astrid Zobel; Michael Wagner


Curr Opin Psychiatry. 2006;19(2):165-170. 

In This Article

Familial and Genetic Risk Factors

Previously, family studies were considered as the crucial test for diagnostic validity and for differentiated disease processes. Recently, linkage and genetic association studies have taken a decisive role in exploring the etiological distinctiveness of diagnostic entities.

Family/Twin Studies

During recent decades, a crucial argument for a fundamental distinction between schizophrenia and bipolar disorder was derived from family studies. 'Breeding true' of diagnoses in families was the ideal; excess morbidity in families of affected ones should only be restricted to the diagnosis of the index case or a similar syndrome. This view was supported by the majority of reports (for example[10]).

More recent family and twin studies are more in favor of an absence of breeding true of diagnoses in families of affected patients. Although the family history of a specific disorder remains the strongest risk factor, significant cosegregation also occurs: an excess of bipolar disorders among relatives of patients with schizophrenia and vice versa.[11,12**]

The first diagnostically unrestricted twin study[13] using blinded diagnostic assessments concluded that the genetic vulnerabilities to schizophrenia and to mania were more overlapping than distinct: strong genetic components were found for schizophrenia (82%) and mania (87%); diagnosis-specific genetic variance accounted for 33% (schizophrenia) and 19% (mania) of variances; common genetic variance shared by both diagnoses accounted for 49% (schizophrenia) and 68% (mania) of variances.

Isolates and Cytogenic Abnormalities

Many large pedigrees living in isolates for centuries reveal a high load of severe mental disorders; in this context, a high degree of diagnostic unspecificity is more the rule than the exception[14,15] with schizophrenia as well as bipolar disorder cosegregating in the same family. This pattern is due to a common underlying susceptibility:

  1. The contribution of identified candidate regions to the manifestation of mental disorders as detected in linkage analysis is not limited to diagnostic boundaries (for example in an eastern Quebec isolate, a region implicated in both disorders was found besides several diagnosis-specific candidate regions[15]).

  2. Cytogenic abnormalities cosegregate with both diagnoses: a balanced translocation involving chromosomes 1 and 11 disrupting a gene DISC1 on chromosome 1q42 was detected to segregate with schizophrenia and bipolar disorder in the same Scottish pedigree.[14] Motivated by this study,[16] more common genetic variants in this gene were investigated in outbred populations and yielded associations to schizophrenia and bipolar disorder.

Much earlier, strong associations were reported between schizophrenia as well as bipolar disorder and the velocardiofacial syndrome, a monogenic disorder resulting from microdeletions in chromosome 22q.[17,18] Interestingly, the underlying gene is located in an often-reported candidate region for both disorders.

Linkage Analysis

Linkage analysis presents a successful approach to map (localize) susceptibility genes into a chromosomal region (candidate region). Limited sample sizes produced inconsistent results for schizophrenia and bipolar disorder but meta-analyses circumvent this deficiency: the most comprehensive meta-analyses for schizophrenia[19] and for bipolar disorder[20] found strong evidence for linkage only to schizophrenia; evidence for linkage to bipolar disorder was only moderate without a cross-diagnostic overlap of candidate regions.

Using another approach, Kohn and Lerer[21**] reviewed evidence of candidate regions on chromosome 6q by only focusing on studies with at least some evidence of linkage in this region. They identified five hotspots on 6q with three of them showing linkage to schizophrenia as well as to bipolar disorder in contrast to another candidate area linked only to schizophrenia, and another one linked to bipolar disorder. Another systematic review[22] of the published genome scans proposed five chromosomal regions along the genome linked to both disorders: 18p11.2, 13q32, 22q11-13, 8p22, 10p14. Indeed, several susceptibility genes shared by both disorders are located in these areas.

Susceptibility Genes

Several positive claims have been there for susceptibility genes for both schizophrenia and bipolar disorder, and some of these overlap and have been reported to be associated with both diseases. During the next few years, studies using more advanced genetic technology and large sample sizes will determine if these in fact are true common genetic risk factors. Strong support for genes implicated in the manifestation of schizophrenia was found: neuregulin-1 gene on 8p22, dysbindin gene on 6p22, and the G72/G30 gene on 13q32, and possibly the COMT gene on 22q11-13, are susceptibility genes.[23,24,25,26,27,28]

The strongest support for specific susceptibility genes common to schizophrenia and bipolar disorder comes from the G72/G30 gene and the neuregulin-1 gene (for G72/G30:[29,30] for neuregulin-1:[31**]).

Among the variants of the gene for dopamine-degrading enzyme COMT, the Val-variant has received up to now most attention; the cumulative evidence, however, is not convincing for schizophrenia;[28] nonetheless, stronger associations emerged from other COMT-gene haplotypes with schizophrenia[24] and with bipolar disorder.[32]

Associations between bipolar disorder and variants of the dysbindin gene have not been reported up to now; instead, maximal strength of association to those gene variants in schizophrenia are mediated by negative symptoms.[33]

Are converging results in favor of a common predisposing susceptibility allele? Association to the same haplotype/marker in schizophrenia and in bipolar disorder in the same population is requested for a positive answer. Indeed, Schumacher et al.[30] were able to support this possibility for the G72/G30 gene, Green et al.[31**] for the neuregulin-1 gene and Shifman et al.[32] for the COMT gene.

Are the two disorders contributing to the observed associations in a global manner or through a specific symptom or symptom pattern? A refined analysis proposed persecutory delusions to explain the association between the G72/G30 haplotype and bipolar disorder. This finding was replicated in an independent sample.[34**] The association to the neuregulin-1 gene is also greatly due to specific symptoms in bipolar cases: mood-incongruent psychotic features proposing a specific effect in this subset of functional psychosis.[31] Thus, the question remains: what is the most appropriate clinical target for an involved susceptibility allele or haplotype, crucial symptoms or diagnoses, which are defined through symptom patterns and additional criteria?

It should be kept in mind that complex behaviors as psychotic and affective disorders are influenced by multiple genes with each of them influencing multiple behavioral components at various physiological functions. On this background it is remarkable that all risk genes identified for schizophrenia and bipolar disorder are involved in the glutamatergic transmission[35] or in the development of neurons and glia cells. These observations point to cross-diagnostic communalities in glutamatergic transmission and neurodevelopment.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.