Introduction
DSM-IV is organized into 16 chapters, and ICD-10 manages a similar number of disorders within 10 chapters. The planning for DSM-V and ICD-11 is in progress. The time is propitious for a fresh look at the organization of disorders into groups or clusters. A new organization could reflect both the risk factors and the clinical manifestations of disorders (Andrews et al. Reference Andrews, Goldberg, Krueger, Carpenter, Hyman, Sachdev and Pine2009a). Five clusters of disorders are proposed: Neurocognitive (Sachdev et al. Reference Sachdev, Andrews, Hobbs, Sunderland and Anderson2009), Neurodevelopmental (Andrews et al. Reference Andrews, Pine, Hobbs, Anderson and Sunderland2009b), Psychoses, Emotional (Goldberg et al. Reference Goldberg, Krueger, Andrews and Hobbs2009b) and Externalizing (Krueger & South, Reference Krueger and South2009).
Here we summarize the evidence for including two boundary disorders in the psychosis cluster, schizotypal personality disorder (SPD) and bipolar disorder (BD), applying 11 spectrum criteria recommended by the DSM-V Study Group (Hyman et al. personal communication, December 2007).
Shared genetic risk factors and familiality
Early family and twin studies show that familial risks are partly shared among schizophrenia, other non-affective psychoses and SPD, and, to a lesser extent, BD and affective psychoses (Kendler & Gardner, Reference Kendler and Gardner1997). These findings are further corroborated by several recent large population-based studies (Mortensen et al. Reference Mortensen, Pedersen, Melbye, Mors and Ewald2003; Laursen et al. Reference Laursen, Labouriau, Licht, Bertelsen, Munk-Olsen and Mortensen2005). Birmaher et al. (Reference Birmaher, Axelson, Monk, Kalas, Goldstein, Hickey, Obreja, Ehmann, Iyengar, Shamseddeen, Kupfer and Brent2009) reported on 388 offspring of 233 parents with BD. Of the 52.1% with axis I disorders, most were mood and anxiety disorders and none received diagnoses of schizophrenia and related disorders. However, some studies do show familial co-aggregation of BD and schizophrenia (Bramon & Sham, Reference Bramon and Sham2001; Craddock & Owen, Reference Craddock and Owen2005). Cardno et al. (Reference Cardno, Rijsdijk, Sham, Murray and McGuffin2002) report common and specific genetic contributions to liability for schizophrenia and BD, and schizo-affective disorder overlapped with schizophrenia and/or BD. In the most definitive study, Lichtenstein et al. (Reference Lichtenstein, Yip, Bjork, Pawitan, Cannon, Sullivan and Hultman2009) analyzed data from over a million nuclear families including more than 75 000 probands with schizophrenia or BD in a national sample from the Sweden registry. About 60% of the variance of each disease is based on genetic effects, and about half of the genetic effect is shared and about half is unique. Owen et al. (Reference Owen, Craddock and Jablensky2007) provide a model for the genetic deconstruction of psychoses.
Until recently, the evidence for shared candidate genes and chromosomal locations in schizophrenia and BD from linkage and gene association research was considered relatively robust. Meta-analyses supported this view (Badner & Gershon, Reference Badner and Gershon2002; Berrettini, Reference Berrettini2003; Lewis et al. Reference Lewis, Levinson, Wise, DeLisi, Straub, Hovatta, Williams, Schwab, Pulver, Faraone, Brzustowicz, Kaufmann, Garver, Gurling, Lindholm, Coon, Moises, Byerley, Shaw, Mesen, Sherrington, O'Neill, Walsh, Kendler, Ekelund, Paunio, Lönnqvist, Peltonen, O'Donovan, Owen, Wildenauer, Maier, Nestadt, Blouin, Antonarakis, Mowry, Silverman, Crowe, Cloninger, Tsuang, Malaspina, Harkavy-Friedman, Svrakic, Bassett, Holcomb, Kalsi, McQuillin, Brynjolfson, Sigmundsson, Petursson, Jazin, Zoëga and Helgason2003; Segurado et al. Reference Segurado, Detera-Wadleigh, Levinson, Lewis, Gill, Nurnberger, Craddock, DePaulo, Baron, Gershon, Ekholm, Cichon, Turecki, Claes, Kelsoe, Schofield, Badenhop, Morissette, Coon, Blackwood, McInnes, Foroud, Edenberg, Reich, Rice, Goate, McInnis, McMahon, Badner, Goldin, Bennett, Willour, Zandi, Liu, Gilliam, Juo, Berrettini, Yoshikawa, Peltonen, Lönnqvist, Nöthen, Schumacher, Windemuth, Rietschel, Propping, Maier, Alda, Grof, Rouleau, Del-Favero, Van Broeckhoven, Mendlewicz, Adolfsson, Spence, Luebbert, Adams, Donald, Mitchell, Barden, Shink, Byerley, Muir, Visscher, Macgregor, Gurling, Kalsi, McQuillin, Escamilla, Reus, Leon, Freimer, Ewald, Kruse, Mors, Radhakrishna, Blouin, Antonarakis and Akarsu2003). However, Sullivan (Reference Sullivan2007) noted that a meta-analysis produces a high number of overlapping genes by chance, and Segurado et al. (Reference Segurado, Detera-Wadleigh, Levinson, Lewis, Gill, Nurnberger, Craddock, DePaulo, Baron, Gershon, Ekholm, Cichon, Turecki, Claes, Kelsoe, Schofield, Badenhop, Morissette, Coon, Blackwood, McInnes, Foroud, Edenberg, Reich, Rice, Goate, McInnis, McMahon, Badner, Goldin, Bennett, Willour, Zandi, Liu, Gilliam, Juo, Berrettini, Yoshikawa, Peltonen, Lönnqvist, Nöthen, Schumacher, Windemuth, Rietschel, Propping, Maier, Alda, Grof, Rouleau, Del-Favero, Van Broeckhoven, Mendlewicz, Adolfsson, Spence, Luebbert, Adams, Donald, Mitchell, Barden, Shink, Byerley, Muir, Visscher, Macgregor, Gurling, Kalsi, McQuillin, Escamilla, Reus, Leon, Freimer, Ewald, Kruse, Mors, Radhakrishna, Blouin, Antonarakis and Akarsu2003) did not find overlap in the highest ranking genes for each disorder. Recent genome-wide association studies (GWAS) cast further doubt on the likelihood that leading candidate genes will be useful in testing validity of diagnostic classes and instead suggest that the contribution of any single gene is very small (Wellcome Trust Case Control Consortium, 2007; Sklar et al. Reference Sklar, Smoller, Fan, Ferreira, Perlis, Chambert, Nimgaonkar, McQueen, Faraone, Kirby, de Bakker, Ogdie, Thase, Sachs, Todd-Brown, Gabriel, Sougnez, Gates, Blumenstiel, Defelice, Ardlie, Franklin, Muir, McGhee, Macintyre, McLean, Vanbeck, McQuillin, Bass, Robinson, Lawrence, Anjorin, Curtis, Scolnick, Daly, Blackwood, Gurling and Purcell2008).
A shared genetic liability for schizophrenia and BD may be associated with psychopathology common to both disorders (e.g. depression, reality distortion) and may not be relevant for differential diagnosis. Investigators have observed an amplified linkage signal in loci shared by the two disorders when they focused on subgroups of families based on psychosis or affective symptoms (Potash et al. Reference Potash, Willour, Chiu, Simpson, MacKinnon, Pearlson, Depaulo and McInnis2001; Hamshere et al. Reference Hamshere, Bennett, Williams, Segurado, Cardno, Norton, Lambert, Williams, Kirov, Corvin, Holmans, Jones, Jones, Gill, O'Donovan, Owen and Craddock2005).
In summary, family, twin and genetic studies provide evidence for both shared and non-shared contributions to schizophrenia and BD. However, specific genes are not yet useful in validating diagnoses. The relevance of genetic effects to the question of grouping BD and schizophrenia together depends on whether the focus is on shared or unique genetic effects.
Environmental risk factors and gene–environment interactions
Several environmental factors have been associated with schizophrenia, some of which are also associated with BD. Substance misuse (Henquet et al. Reference Henquet, Murray, Linszen and van Os2005, Reference Henquet, Krabbendam, de Graaf, Ten Have and van Os2006), prenatal factors, particularly preterm birth (Susser et al. Reference Susser, Neugebauer, Hoek, Brown, Lin, Labovitz and Gorman1996; Brown et al. Reference Brown, van Os, Driessens, Hoek and Susser2000; Clarke et al. Reference Clarke, Harley and Cannon2006; Scott et al. Reference Scott, McNeill, Cavanagh, Cannon and Murray2006; Laursen et al. Reference Laursen, Munk-Olsen, Nordentoft and Mortensen2007), negative life events and childhood trauma (van Os et al. Reference van Os, Jones, Sham, Bebbington and Murray1998; Hammersley et al. Reference Hammersley, Dias, Todd, Bowen-Jones, Reilly and Bentall2003; Read et al. Reference Read, van Os, Morrison and Ross2005) are reported to be linked to both diagnostic classes. Schizophrenia and BD psychotic episodes are also associated with urbanicity, possibly reflecting a precipitant effect of environmental risks in genetically vulnerable individuals (Krabbendam & van Os, Reference Krabbendam, Arts, van Os and Aleman2005; Kaymaz et al. Reference Kaymaz, Krabbendam, de Graaf, Nolen, Ten Have and van Os2006). BD without positive psychotic symptoms, however, is not correlated with urbanicity (Kaymaz et al. Reference Kaymaz, Krabbendam, de Graaf, Nolen, Ten Have and van Os2006). Advanced paternal age and Toxoplasma gondii are associated with schizophrenia and not known to be risk factors in BD (Laursen et al. Reference Laursen, Munk-Olsen, Nordentoft and Mortensen2007; Mortensen et al. Reference Mortensen, Nørgaard-Pedersen, Waltoft, Sørensen, Hougaard and Yolken2007; Perrin et al. Reference Perrin, Brown and Malaspina2007; Torrey & Yolken, Reference Torrey and Yolken2007; Torrey et al. Reference Torrey, Bartko, Lun and Yolken2007; Yolken & Torrey, Reference Yolken and Torrey2008).
These environmental risk factors are common to many disorders and thus are not determinative of cluster membership. Rather, they may contribute to dimensions of psychopathology across diagnostic categories. Researchers are turning their attention to understanding the complex ways in which nature interacts with nurture to produce psychosis. This genotype×environmental interaction (G×E) approach posits a causal role for synergistic co-participation, where the effect of one is conditional on the other (Leboyer et al. Reference Leboyer, Meyer-Lindenberg, Stefanis, Rutten, Arango, Jones, Kapur, Lewis, Murray, Owen, Linszen, Kahn, van Os, Wiersma, Bruggeman, Cahn, Germeys, de Haan and Krabbendam2008). G×E seems a particularly suitable approach for understanding the development of psychosis across the diagnostic categories in DSM. The psychosis phenotype is known to be associated with environmentally mediated risks, yet people display considerable heterogeneity in their response to those environmental exposures (van Os et al. Reference van Os, Rutten and Poulton2008).
Shared neural substrates
Many neuroimaging studies compare schizophrenia to healthy controls (HC) but fewer examine differences between BD and HC. Fewer still directly compare schizophrenia and BD subjects and, of these, only a subset specifies whether psychosis is present in the BD cases. Neural substrate findings are summarized in Table 1.
Table 1. Summary of neural substrate findings in schizophrenia (Sz), bipolar (BP) disorder and healthy normal volunteers (HNV)
MRI, Magnetic resonance imaging; H-MRS, proton magnetic resonance spectroscopy; P-MRS, phosphorus magnetic resonance spectroscopy; DTI, diffusion tensor imaging; PET, positron emission tomography; fMRI, functional MRI; GM, gray matter; WM, white matter; CSF, cerebrospinal fluid; NAA, N-acetylaspartate; Glu, glutamate; PDE, phosphodiesters; FA, fractional anisotropy; GABA, gamma-aminobutyric acid; PME, phosphomonoesters.
↓, Decreased; ↑, increased; +, some evidence; ++, some replications; +++, repeatedly replicated finding.
Magnetic resonance imaging (MRI)
Structural anatomical abnormalities are reported across the psychosis spectrum. Most commonly, gray matter reductions in multiple regions have been identified in schizophrenia and BD patients, when compared to HC. The magnitude of these effects is larger in schizophrenia but similar to BD in abnormal topography. Meta-analyses document a 3–4% whole-brain volume reduction in schizophrenic probands compared to HC (see e.g. Woodruff et al. Reference Woodruff, McManus and David1995; Wright et al. Reference Wright, Rabe-Hesketh, Woodruff, David, Murray and Bullmore2000; Steen et al. Reference Steen, Mull, McClure, Hamer and Lieberman2006; for a review, Harrison, Reference Harrison1999) that is not entirely consistent with the pattern of volume loss in BD (Hoge et al. Reference Hoge, Friedman and Schulz1999). In both schizophrenia and BD, volume reductions are most consistently reported in cortical gray matter, particularly in frontotemporal regions. Relatives of schizophrenic probands and high-risk individuals also have significant, though smaller, volume reductions compared to HC (Seidman et al. Reference Seidman, Faraone, Goldstein, Goodman, Kremen, Tommey, Tourville, Kennedy, Makris, Caviness and Tsuang1999; Lawrie et al. Reference Lawrie, Whalley, Abukmeil, Kestelman, Donnelly, Miller, Best, Owens and Johnstone2001; O'Driscoll et al. Reference O'Driscoll, Florencio, Gagnon, Wolff, Benkelfat, Mikula, Lal and Evans2001; Ho, Reference Ho2007). Of note, there is a concomitant increase in sulcal and ventricular cerebral spinal fluid in schizophrenic probands. Ventricular enlargement also occurs in BD and may be more severe in patients with multiple episodes (Hauser et al. Reference Hauser, Matochik, Altshuler, Denicoff, Conrad, Li and Post2000; Strakowski et al. Reference Strakowski, DelBello, Zimmerman, Getz, Mills, Ret, Shear and Adler2002), but long-term longitudinal studies (comparable to those in early schizophrenia) are lacking. A meta-analysis suggests that right-side ventricular enlargement is the most consistent finding in BD (McDonald et al. Reference McDonald, Zanelli, Rabe-Hesketh, Ellison-Wright, Sham, Kalidindi, Murray and Kennedy2004). The difficulty in relating imaging findings to the diagnostic grouping issue is illustrated by McIntosh et al. (Reference McIntosh, Job, Moorhead, Harrison, Lawrie and Johnstone2005), who report reduced white matter density in the anterior limb of the internal capsule in both syndromes. However, unaffected relatives did not have this reduction and a reduction in frontal subgyral white matter was only observed in cases with a family history of schizophrenia.
Longitudinal changes (mainly gray matter reductions) early in schizophrenia have been repeatedly demonstrated as early as 3 months after treatment (see e.g. DeLisi et al. Reference DeLisi, Sakuma, Tew, Kushner, Hoff and Grimson1997; McCarley et al. Reference McCarley, Wible, Frumin, Hirayasu, Levitt, Fischer and Shenton1999; Rapoport et al. Reference Rapoport, Giedd, Blumenthal, Hamburger, Jeffries, Fernandez, Nicolson, Bedwell, Lenane, Zijdenbos, Paus and Evans1999; Lieberman et al. Reference Lieberman, Chakos, Wu, Alvir, Hoffman, Robinson and Bilder2001; Ho et al. Reference Ho, Andreasen, Nopoulos, Arndt, Magnotta and Flaum2003; van Haren et al. Reference van Haren, Cahn, Hulshoff Pol, Schnack, Caspers, Lemstra, Sitskoom, Wiersma, van den Bosch, Dingemans, Schene and Kahn2003; Theberge et al. Reference Theberge, Williamson, Aoyama, Drost, Manchandra, Malla, Northcott, Menon, Neufeld, Rajakumar, Pavlosky, Densmore, Schaefer and Williamson2007). Lieberman et al. (Reference Lieberman, Tollefson, Charles, Zipursky, Sharma, Kahn, Keefe, Green, Gur, McEvoy, Perkins, Hamer, Gu and Tohen2005) documented global gray matter reductions early in schizophrenia. Early brain volume reduction may stabilize in early adulthood (Woods, Reference Woods1998) or later (Mathalon et al. Reference Mathalon, Sullivan, Lim and Pfefferbaum2001), but most evidence suggests progressive changes during the course of schizophrenia (Arango et al. Reference Arango, Moreno, Martinez, Parellada, Desco, Moreno, Fraguas, Gogtay, James and Rapoport2008; DeLisi, Reference DeLisi2008; Lawrie et al. Reference Lawrie, McIntosh, Hall, Owens and Johnstone2008; Hulsoff Pol & Kahn, Reference Hulsoff Pol and Kahn2008; Wood et al. Reference Wood, Pantelis, Velakoulis, Yucel, Fornito and McGorry2008). It is not known whether progressive changes followed by stabilization is the pattern for BD. Epiphenomena such as substance abuse, therapeutic drugs and intense smoking may contribute to these observations (see e.g. Dorph-Petersen et al. Reference Dorph-Petersen, Pierri, Perel, Sun, Sampson and Lewis2005; Rais et al. Reference Rais, Cahn, Van-Haren, Schnack, Caspers, Hulshoff Pol and Kahn2008).
Subcortical striatal regions may be enlarged in BD compared to HC (Strakowski et al. Reference Strakowski, DelBello and Adler2005) and in schizo-affective disorder (Getz et al. Reference Getz, DelBello, Fleck, Zimmerman, Schwiers and Strakowski2002). The available evidence has found increased amygdala in BD and reduced hippocampi in schizophrenia (Altshuler et al. Reference Altshuler, Bartzokis, Grieder, Curran and Mintz1998; Strakowski et al. Reference Strakowski, DelBello, Sax, Zimmerman, Shear, Hawkins and Larson1999). One study has followed first-episode psychosis subjects with a baseline and repeat MRI (1.5 years later). Both schizophrenia and psychotic BD subjects had smaller left superior temporal volumes than HC, but only schizophrenia subjects had further volume reductions (McCarley et al. Reference McCarley, Wible, Frumin, Hirayasu, Levitt, Fischer and Shenton1999). However, a recent report failed to document gray or white matter differences between BD subjects and non-ill controls (Scherk et al. Reference Scherk, Kemmer, Usher, Reith, Falkai and Gurber2008). Medications may confound relationships between diagnostic cohorts.
Direct comparisons between the first-degree relatives (FDRs) of BD and schizophrenia subjects have revealed prefrontal gray and white matter reductions in schizophrenia relatives but not in BD relatives (McIntosh et al. Reference McIntosh, Job, Moorhead, Harrison, Whalley, Johnstone and Lawrie2006). Hippocampal volumes were reduced in the schizophrenia subjects but not in their relatives. BD and their relatives had normal brain volume indices (McDonald et al. Reference McDonald, Bullmore, Sham, Chitnis, Suckling, MacCabe, Walshe and Murray2006).
Although these studies highlight broad similarity in the brain systems affected in schizophrenia and BD (i.e. frontotemporal), the data arguably provide the most persuasive evidence for neuroanatomical differences between schizophrenia and BD. Important (and potentially opposite) effects of antipsychotic and mood-stabilizing agents cannot be excluded, but the presence of disease-specific abnormalities in the schizophrenia relatives, but not in the BD relatives, strongly suggests that brain volume deficits are not entirely accounted for by a medication effect in schizophrenia. However, given evidence for neurotrophic effects of mood-stabilizing agents (Sassi et al. Reference Sassi, Nicoleeti, Brambilla, Mallinger, Frank, Kupfer, Keshavan and Soares2002) that could increase gray matter volumes in BD subjects, a complex contribution of psychotropic medications (reduced cortical volumes and increased striatal volumes with antipsychotics and increased cortical volumes with mood stabilizers) must be considered in relation to conclusions regarding disease-specific volumetric abnormalities.
Diffusion tensor imaging (DTI)
Many studies have examined white matter tract integrity with DTI in schizophrenia/HC comparisons. There have been several reports of reduced white matter integrity in schizophrenia in a variety of regions (Gur et al. Reference Gur, Calkins, Gur, Horan, Nuechterlein, Seidman and Stone2007). There are fewer reports regarding BD and these are not consistent. Yurgelun-Todd et al. (Reference Yurgelun-Todd, Silveri, Gruber, Rohan and Pimentel2007) report increased white matter integrity, whereas Adler et al. (Reference Adler, Holland, Schmithorst, Wilke, Weiss, Pan and Strakowski2004) report reduced white matter integrity. No direct comparisons between diagnostic groups are currently available.
Magnetic resonance spectroscopy (MRS)
There is abundant evidence of neuronal dysfunction in both schizophrenia and BD. BD is characterized by reduced N-acetylaspartate (NAA), a marker of neuronal viability (Yildiz-Yesiloglu & Ankerst, Reference Yildiz-Yesiloglu and Ankerst2006). However, as in the MRI literature, there are suggestions of neuroprotective effects of lithium, leading to NAA elevations (Brambilla et al. Reference Brambilla, Stanley, Nicoletti, Sassi, Mallinger, Frank, Kupfer, Keshavan and Soares2005). Direct comparisons between the diagnostic groups are lacking.
Glutamatergic indices are increased in medication-naive schizophrenia probands and those at high risk of the disorder (Theberge et al. Reference Theberge, Bartha, Drost, Menon, Malla, Takhar, Neufeld, Rogers, Pavlosky, Schaefer, Densmore, Al-Semaan and Williamson2002; Tibbo et al. Reference Tibbo, Hanstock, Valiakalayil and Allen2004; Chang et al. Reference Chang, Friedman, Ernst, Zhong, Tsopelas and Davis2007). Glutamate indices are also increased in unmedicated BD (Dager et al. Reference Dager, Friedman, Parow, Demopulos, Stoll, Lyoo, Dunner and Renshaw2004). No study has directly compared these metabolites between schizophrenia and BD.
Positron emission tomography (PET) /single photon emission computed tomography (SPECT) neuroreceptor studies
A review of D2 densities in vivo and post-mortem in schizophrenia is suggestive of increased D2 receptors (Zakzanis & Hansen, Reference Zakzanis and Hansen1998). BD with psychosis is also associated with increased striatal D2 receptor density comparable to medication-naive schizophrenia probands. This is not found in non-psychotic BD. These studies suggest that psychosis, irrespective of schizophrenia or mood disorder, is related to increased D2 receptor density. This is consistent with the clinical evidence regarding the efficacy of D2 blockade for psychotic symptoms regardless of diagnosis. The significance of these pharmacological findings has been explicated in a heuristic model linking excessive dopamine to the development of reality distortion symptoms through its role in reward-based behavior, and the assignment of motivational significance to external stimuli (Kapur et al. Reference Kapur, Mizrahi and Li2005; van Os & Kapur, Reference van Os and Kapurin press). This model provides a means of linking dopamine dysregulation with delusion and hallucination through aberrant associations of salience, and suggests that antipsychotic drugs may serve to ‘detach’ an individual from both aberrant and normal motivational salience. Receptor data are particularly vulnerable to drug effects. Reality distortion symptoms occur in many disorders outside the psychosis cluster.
Functional MRI (fMRI)
Functional imaging studies highlight important differences in the activation of frontostriatal networks in schizophrenia and BD patients, depending on task requirements. Working memory performance in schizophrenia has been commonly associated with hypoactivation of the dorsolateral prefrontal cortex (DLPFC), although there is also evidence of hyperactivation in anterior cingulate and frontal pole regions (Glahn et al. Reference Glahn, Ragland, Abramoff, Barrett, Laird, Bearden and Velligan2005). Comparatively fewer studies in BD nevertheless report similar patterns of hypofrontal activation during working memory task performance (see Phillips & Vietta, Reference Phillips and Vieta2007). Other studies using emotional stimuli have revealed a pattern of increased subcortical striatal and limbic activation alongside reduced prefrontal cortex activity in BD (see Green et al. Reference Green, Cahill and Malhi2007). This pattern of activity is also evident in pediatric cases of BD (Dickstein et al. Reference Dickstein, Rich, Roberson-Nay, Berghorst, Vinton, Pine and Leibenluft2007; Pavuluri et al. Reference Pavuluri, O'Connor, Harral and Sweeney2008). In schizophrenia, there is a contrasting pattern of decreased limbic activation and hyperfrontality during emotion processing tasks (e.g. Holt et al. Reference Holt, Kunkel, Weiss, Goff, Wright, Shin, Rauch, Hootnick and Heckers2006). Finally, use of a sentence completion task to study neural correlates of inhibitory control in medicated schizophrenia, BD and HC groups has shown increased activation in the right insula in schizophrenia, whereas BD had reduced activation in this region (McIntosh et al. Reference McIntosh, Whalley, McKirdy, Hall, Sussmann, Shankar, Johnstone and Lawrie2008). Conversely, BD had increased activity in the left DLPFC, whereas schizophrenia subjects had reduced activation in this region. A model including activation of the ventral striatum, middle temporal gyrus, DLPFC and right insula correctly classified 92% of BD and 58% of schizophrenia subjects. Thus, as in other imaging areas, some data suggest similarity and other data suggest important differences between schizophrenia and BD.
Post-mortem
Post-mortem studies show both similarities and differences between schizophrenia and BD. Similarities focus mainly on reduced gamma-aminobutyric acid (GABA) ergic indices in the anterior cingulate and hippocampus. There have been several studies describing evidence of decreased glutamic acid decarboxylase (GAD67) mRNA levels in the limbic lobe of both BD and schizophrenia (e.g. Akbarian & Huang, Reference Akbarian and Huang2006; Benes et al. Reference Benes, Matzilevich, Burke and Walsh2006). Reports from the Stanley Neuropathology Consortium are somewhat consistent with these findings. Evidence of reduced parvalbumin-containing cells in hippocampus layer CA2 (which represent GABAergic interneurons) was found in a direct comparison of both schizophrenia and BD tissue samples compared to HC (Knable et al. Reference Knable, Barci, Webster, Meador-Woodruff and Torrey2004). BD and schizophrenia groups shared about 65% of abnormalities in a variety of mRNA and protein markers related to developmental/synaptic and GABAergic systems (Torrey et al. Reference Torrey, Barci, Webster, Bartko, Meador-Woodruff and Knable2005).
Schizophrenia and BD are associated with conflicting glial profiles. Schizophrenia shows increased neuronal density and glial density in prefrontal regions (Selemon et al. Reference Selemon, Rajkowska, Goldman-Rakic, Watson and Meador-Woodruff1995) and decreased dendritic spines (Glantz & Lewis, Reference Glantz and Lewis2000) without neuronal loss or gliosis (Harrison, Reference Harrison1999). This has been interpreted as reduction of the neuropil, with increased neuronal packing in cortical layers III to VI. In contrast to schizophrenia, BD shows reduced glial and neuronal density coupled with glial hypertrophy in these regions (Ongur et al. Reference Ongur, Drevets and Price1998; Rajkowska et al. Reference Rajkowska, Halaris and Selemon2001). These findings suggest distinct neuropathological substrates in schizophrenia and BD but the samples were small and the potential contribution of medications to these post-mortem findings cannot be dismissed (Dorph-Petersen et al. Reference Dorph-Petersen, Pierri, Perel, Sun, Sampson and Lewis2005).
In summary, the neuroanatomy literature contains compelling reports of differences between BD and schizophrenia, but similarities are also observed. There are few direct comparisons between the two groups using MRI (Altshuler et al. Reference Altshuler, Bartzokis, Grieder, Curran and Mintz1998; McCarley et al. Reference McCarley, Wible, Frumin, Hirayasu, Levitt, Fischer and Shenton1999; Strakowski et al. Reference Strakowski, DelBello, Sax, Zimmerman, Shear, Hawkins and Larson1999; McDonald et al. Reference McDonald, Bullmore, Sham, Chitnis, Suckling, MacCabe, Walshe and Murray2006; McIntosh et al. Reference McIntosh, Job, Moorhead, Harrison, Whalley, Johnstone and Lawrie2006), which all report reduced brain tissue volumes in schizophrenia. Because small effects are expected and the measurement variability with most of these neuroimaging tools is large, concurrent study of both disorders is essential.
Shared biomarkers
Several neurophysiological abnormalities are observed in psychotic disorders before the onset of psychosis, are reported to be stable over the course of the illness, and are only mildly affected by psychotic state and medications. Recent data indicate that, in most part, these neurophysiological deficits are independent of each other, and are observed in non-ill relatives of schizophrenia patients and in subjects with schizotypal traits (Light & Braff, Reference Light and Braff2001; Braff et al. Reference Braff, Freedman, Schork and Gottesman2007; Gur et al. Reference Gur, Calkins, Gur, Horan, Nuechterlein, Seidman and Stone2007; Hong et al. Reference Hong, Summerfelt, Wonodi, Adami, Buchanan and Thaker2007; Turetsky et al. Reference Turetsky, Calkins, Light, Olincy, Radant and Swerdlow2007). Many of the same endophenotypes are now being studied in BD subjects and their FDRs (Hill et al. Reference Hill, Harris, Herbener, Pavuluri and Sweeney2008; Pearlson & Folley Reference Pearlson and Folley2008; Thaker, Reference Thaker2008).
Smooth pursuit and saccadic eye movement abnormalities
Schizophrenia patients with primary and enduring negative symptoms have impairment in smooth pursuit eye initiation (Hong et al. Reference Hong, Avila, Adami, Elliott and Thaker2003). Furthermore, pursuit maintenance and, more specifically, predictive pursuit response are abnormal in FDRs of schizophrenia patients and probands, particularly those with schizotypal symptoms (Holzman et al. Reference Holzman, Proctor, Levy, Yasillo, Meltzer and Hurt1974; Thaker et al. Reference Thaker, Ross, Cassady, Adami, LaPorte, Medoff and Lahti1998, Reference Thaker, Avila, Hong, Medoff, Ross and Adami2003; Avila et al. Reference Avila, Hong, Moates, Turano and Thaker2006; Hong et al. Reference Hong, Turano, O'Neill, Hao, Wonodi, McMahon, Elliott and Thaker2008), and are thought to mark psychosis liability (Hong et al. Reference Hong, Mitchell, Avila, Adami, McMahon and Thaker2006). BD probands and their relatives also show pursuit abnormality similar to the relatives of schizophrenia probands (Rosenberg et al. Reference Rosenberg, Sweeney, Squires-Wheeler, Keshavan, Cornblatt and Erlenmeyer-Kimling1997; Kathmann et al. Reference Kathmann, Hochrein, Uwer and Bondy2003).
In addition to the smooth pursuit eye movement abnormalities, studies have observed abnormality in saccadic inhibition (anti-saccades) and oculomotor delayed responses (which assess spatial working memory) in schizophrenia probands and their relatives, and also in persons with affective disorders (see review by Thaker, Reference Thaker2008).
Sensory gating (P50) deficit
Schizophrenia and BD probands with a lifetime history of psychosis show muted inhibition as measured by P50 responses to a paired click paradigm (Perry et al. Reference Perry, Minassian, Feifel and Braff2001; Sánchez-Morla et al. Reference Sánchez-Morla, García-Jiménez, Barabash, Martínez-Vizcaíno, Mena, Cabranes-Díaz, Baca-Baldomero and Santos2008). This deficit is not consistently affected by medication status or clinical state. BD without a history of psychosis is not associated with abnormal P50 suppression (Olincy & Martin, Reference Olincy and Martin2005). Abnormalities in P50 suppression have also been noted in FDRs of BD-I probands with psychotic features (Schulze et al. Reference Schulze, Hall, McDonald, Marshall, Walshe, Murray and Bramon2007; Hall et al. Reference Hall, Schulze, Sham, Kalidindi, McDonald, Bramon, Levy, Murray and Rijsdijk2008), and in patients with SPD (Cadenhead et al. Reference Cadenhead, Light, Geyer, McDowell and Braff2002).
Prepulse inhibition (PPI)
Patients with schizophrenia show a reduced inhibition of the startle response by a prepulse stimulus (PPI), even when the startle reflex is generally within the normal range (Braff et al. Reference Braff, Geyer and Swerdlow2001). The deficit is also observed in non-psychotic patients, patients not on medications (Swerdlow et al. Reference Swerdlow, Light, Cadenhead, Sprock, Hsieh and Braff2006), and in FDRs who are clinically unaffected (Kumari et al. Reference Kumari, Das, Zachariah, Ettinger and Sharma2005).
Similar to schizophrenic probands, impairment in PPI has been consistently observed in BD during acute episodes (Perry et al. Reference Perry, Minassian, Feifel and Braff2001), and also in remitted patients (Quraishi & Frangou, Reference Quraishi and Frangou2002; Martinez-Aran et al. Reference Martinez-Aran, Vieta, Colom, Torrent, Sanchez-Moreno, Reinares, Benabarre, Goikolea, Brugue, Daban and Salamero2004; Frangou et al. Reference Frangou, Donaldson, Hadjulis, Landau and Goldstein2005; Robinson et al. Reference Robinson, Thompson, Gallagher, Goswami, Young, Ferrier and Moore2006). Three studies have noted PPI deficits in unaffected FDRs of BD probands (Zalla et al. Reference Zalla, Joyce, Szoke, Schurhoff, Pillon, Komano, Perez-Diaz, Bellivier, Alter, Dubois, Rouillon, Houde and Leboyer2004; Frangou et al. Reference Frangou, Donaldson, Hadjulis, Landau and Goldstein2005; Giakoumaki et al. Reference Giakoumaki, Roussos, Rogdaki, Karli, Bitsios and Frangou2007), suggesting that PPI impairment may mark psychosis liability in both schizophrenia and BD, and is associated with shared genetic predisposition to psychotic symptoms. However, these deficits also occur in disorders of other clusters including co-morbid attention deficit hyperactivity disorder, tic disorders, obsessive–compulsive disorder, and Huntington's disease (Braff et al. Reference Braff, Geyer and Swerdlow2001).
P300 evoked potential
BD and schizophrenia have reduced P300 amplitude and increased latency (Souza et al. Reference Souza, Muir, Walker, Glabus, Roxborough, Sharp, Dunan and Blackwood1995; Ford, Reference Ford1999). This deficit is heritable and probably related to the etiology of the two diagnoses as similar impairments are observed in FDRs of both disorder probands (Pierson et al. Reference Pierson, Jouvent, Quintin, Perez-Diaz and Leboyer2000; Turetsky et al. Reference Turetsky, Cannon and Gur2000; van Beijsterveldt et al. Reference van Beijsterveldt, van Baal, Molenaar, Boomsma and de Geus2001; Winterer et al. Reference Winterer, Egan, Raedler, Sanchez, Jones, Coppola and Weinberger2003; Bramon et al. Reference Bramon, McDonald, Croft, Landau, Filbey, Gruzelier, Sham, Frangou and Murray2005).
Early information processing and mismatch negativity
Abnormalities in mismatch negativity are consistently demonstrated in the auditory modality of schizophrenic probands; however, the findings in their unaffected relatives are less consistent (Michie et al. Reference Michie, Innes-Brown, Todd and Jablensky2002; Umbricht & Krljes, Reference Umbricht and Krljes2005; Magno et al. Reference Magno, Yeap, Thakore, Garavan, De Sanctis and Foxe2008). Although not studied extensively, BD patients show unimpaired mismatch negativity (Umbricht et al. Reference Umbricht, Koller, Schmid, Skrabo, Grübel, Huber and Stassen2003).
Neural synchronization deficits
Studies have found abnormality in gamma (30–80 Hz) synchronization in schizophrenia patients (Kwon et al. Reference Kwon, O'Donnell, Wallenstein, Greene, Hirayasu, Nestor, Hasselmo, Potts, Shenton and McCarley1999; Hong et al. Reference Hong, Summerfelt, McMahon, Adami, Francis, Eliott, Buchanan and Thaker2004) and their FDRs. Similar reductions in the gamma band synchronization are also noted in BD (O'Donnell et al. Reference O'Donnell, Hetrick, Vohs, Krishnan, Carroll and Shekhar2004; Maharajh et al. Reference Maharajh, Abrams, Rojas, Teale and Reite2007). The reduced synchrony in the gamma band is correlated to positive symptom ratings such as visual hallucinations, thought disorder, conceptual disorganization, and attention in schizophrenia subjects (O'Donnell et al. Reference O'Donnell, Hetrick, Vohs, Krishnan, Carroll and Shekhar2004; Hermann & Demiralp, Reference Hermann and Demiralp2005).
In summary, although there are extensive data on biomarkers in schizophrenia probands and their relatives, such information is meager in BD. Few studies have directly compared biomarkers in schizophrenia and BD families. The existing data suggest that many of the biomarkers index independent aspects of psychosis risk (Light & Braff, Reference Light and Braff2001; Hong et al. Reference Hong, Summerfelt, Wonodi, Adami, Buchanan and Thaker2007). Measures of early sensory processing such as smooth pursuit initiation, motion perception and mismatch negativity seem to be unique to schizophrenia liability and tend to be associated with negative symptoms (Hong et al. Reference Hong, Avila, Adami, Elliott and Thaker2003; Slaghuis et al. Reference Slaghuis, Bowling and French2005; Urban et al. Reference Urban, Kremlácek, Masopust and Libiger2008), whereas abnormalities in smooth pursuit maintenance, sensory and sensory-motor gating, the P300 component of evoked potential and gamma-band synchronization mark positive psychotic symptoms and are present in both schizophrenia and BD. Table 2 presents a summary of neurophysiological marker findings.
Table 2. Summary of neurophysiological marker findings
SPEM, Smooth pursuit eye movements; PPI, prepulse inhibition; MMN, mismatch negativity; COMT, catecholamine-O-methyltransferase gene; CHRNA7, alpha 7 nicotinic cholinergic receptor gene; NR1, neuregulin 1 gene; DRD2 and DRD3, dopamine receptor D2 and D3 genes; NMDA, N-methyl-d-aspartic acid; GABA, gamma-aminobutyric acid.
+, Some evidence; ++, some replications; +++, repeatedly replicated finding.
a Based on findings in 22q11 deletion syndrome.
b Based on a study in alcoholism.
c Based on findings in (1; 11) translocation carriers. Details and appropriate citations are given in the text.
Shared temperamental antecedents
A strong body of evidence originating in experimental psychology supports the existence of polygenetically inherited variations in temperament and personality factors, which, if inherited in particular combinations, may render an individual vulnerable to the development of psychosis (Green et al. Reference Green, Boyle, Raine, Boyle, Matthews and Saklofske2008). This view is supported by large population-based evidence of the existence of psychotic-like experiences in ostensibly healthy individuals (van Os et al. Reference van Os, Hanssen, Bijl and Ravelli2000, Reference van Os and Kapur2009), and considerable evidence of cognitive, perceptual and psychophysiological characteristics shared by individuals with psychotic disorders, their FDRs and psychosis-prone individuals (Claridge, Reference Claridge1997).
The view that schizotypal characteristics could represent both normal variation in personality and vulnerability to clinical disorders has been difficult to reconcile with the psychiatric notion that these conditions represent distinct classes of ‘illness’, separate from normal function. Although it is beyond the scope of this paper to fully review the evidence in support of this notion, two recent papers have examined whether psychotic-like experiences form a major domain of human variation. Tackett et al. (Reference Tackett, Silberschmidt, Krueger and Sponheim2008) examined the factor structure of a standard set of abnormal personality scales that was augmented by the inclusion of indices of schizotypal personality in a sample designed to be weighted toward psychosis proneness with the inclusion of FDRs of probands with schizophrenia, schizo-affective and BD diagnoses. Elements of the four-factor structure of abnormal personality that has been discussed extensively as a potential dimensional model for abnormal personality in DSM-V (Widiger et al. Reference Widiger, Simonsen, Krueger, Livesley and Verheul2005) emerged, encompassing domains of emotional dysregulation (neuroticism), introversion, compulsivity and antagonism, augmented by an additional fifth dimension of ‘peculiarity’, primarily reflecting the inclusion of the schizotypy scales. Watson et al. (Reference Watson, Clark and Chmielewski2008) reported very similar results in a sample of college students who completed a similarly enriched set of indicators of abnormal personality. Of course, the self-report personality indicators in these studies do not constitute diagnosable psychotic disorder per se, but at least some of these personality indicators have shown predictive validity for psychotic disorder diagnosed through structured interviews (Raine, Reference Raine2006). Studies in this vein generally support a spectrum relationship between schizophrenia and schizotypal personality structure, referring not only to clinical presentation of SPD but also to more broadly defined ‘schizotypal’ personality characteristics within healthy individuals. Some qualities of other cluster A personality disorders may also relate to the peculiarity domain identified in recent factor analytic studies, but the DSM-IV construct that maps best onto this domain is clearly SPD. Endophenotypes, described above as biomarkers, substantiates the relationship between schizophrenia and relatives with schizophrenia spectrum pathology. The relationship between personality and the various other psychotic disorders such as brief reactive psychosis or delusional disorder, has not been a major focus of research.
Shared cognitive and emotional processing abnormalities
Cognitive impairment
Cognitive impairments are consistently noted in schizophrenia, BD, SPD (Green, Reference Green2006), and schizo-affective disorder (Stip et al. Reference Stip, Sepehry, Prouteau, Briand, Nicole, Lalonde and Lesage2005; Torrent et al. Reference Torrent, Martinez-Aran, Amann, Daban, Tabares-Seisdedos, Gonzalez-Pinto, Reinares, Benabarre, Salamero, McKenna and Vieta2007). These deficits are thought to underlie functional disabilities (Bilder et al. Reference Bilder, Lipschutz-Broch, Reiter, Geisler, Mayerhoff and Lieberman1991; Hill et al. Reference Hill, Beers, Kmiec, Keshavan and Sweeney2004; Green, Reference Green2006), and may represent candidate endophenotypes for psychotic conditions that may span current diagnostic groups (Arts et al. Reference Arts, Jabben, Krabbendam and van Os2007; Glahn et al. Reference Glahn, Almasy, Blangero, Burk, Estrada, Peralta, Meyenberg, Castro, Barrett, Nicolini, Raventos and Escamilla2007). Cognitive domains affected in schizophrenia include deficits in sustained attention, visual and verbal episodic memory, working memory, and processing speed (Saykin et al. Reference Saykin, Gur, Gur, Mozley, Mozley, Resnick, Kester and Stafiniak1991; Cannon et al. Reference Cannon, Huttunen, Lonnqvist, Tuulio-Henriksson, Pirkola, Glahn, Finkelstein, Hietanen, Kapiro and Koskenvuo2000; Green et al. Reference Green, Kern, Braff and Mintz2000; Egan et al. Reference Egan, Goldberg, Gscheidle, Weirich, Rawlings, Hyde, Bigelow and Weinberger2001; Dickinson et al. Reference Dickinson, Iannone, Wilk and Gold2004). In general, the pattern of cognitive deficits in BD is similar to the cognitive profile of schizophrenia, although impairment may be somewhat less severe and state dependent in BD (Egan et al. Reference Egan, Goldberg, Gscheidle, Weirich, Rawlings, Hyde, Bigelow and Weinberger2001; Krabbendam et al. Reference Krabbendam, Arts, van Os and Aleman2005; Arts et al. Reference Arts, Jabben, Krabbendam and van Os2007). The FDRs of schizophrenia patients show similar cognitive deficits; although the abnormalities are less salient, this finding supports the heritability of these deficits (Snitz et al. Reference Snitz, MacDonald and Carter2006). Some of the available data, in relatively small samples, suggest that cognitive deficits observed in the BD probands also occur in their FDRs (Antila et al. Reference Antila, Tuulio-Henriksson, Kieseppa, Soronen, Palo, Paunio, Haukka, Partonen and Lonnqvist2007; Trivedi et al. Reference Trivedi, Goel, Dhyani, Sharma, Singh, Sinha and Tandon2008).
In general, cognitive deficits in all domains begin before psychosis and remain stable over the course of schizophrenia (Rund, Reference Rund1998), whereas deficits in attention and executive function seem to be most stable in BD (Burdick et al. Reference Burdick, Goldberg, Harrow, Faull and Malhotra2006; Mur et al. Reference Mur, Portella, Martinez-Aran, Pifarre and Vieta2008). However, fluctuating attentional disturbances have been reported in schizophrenia (Dawson et al. Reference Dawson, Nuechterlein, Schell, Gitlin and Ventura1994). Cognitive impairments in BD tend to vary significantly with clinical state and deteriorate as the illness progresses (Burt et al. Reference Burt, Prudic, Peyser, Clark and Sackeim2000). However, subtle impairments in executive function, verbal fluency, attention and episodic memory are observed in BD relatives and patients even during euthymic phases (Deckersbach et al. Reference Deckersbach, Savage, Reilly-Harrington, Clark, Sachs and Rauch2004; Malhi et al. Reference Malhi, Ivanovski, Szekeres and Olley2004; Pavuluri et al. Reference Pavuluri, Schenkel, Aryal, Harral, Hill, Herbener and Sweeney2006, Reference Pavuluri, O'Connor, Harral and Sweeney2008; Bora et al. Reference Bora, Vahip, Akdeniz, Ilerisoy, Aldemir and Alkan2008). Comparison of cognitive deficits in BD-I and BD-II revealed more severe deficits in BD-I, consistent with a functional distinction between these groups (Simonsen et al. Reference Simonsen, Sundet, Vaskinn, Birkenaes, Engh, Hansen, Jonsdottir, Ringen, Opjordsmoen, Friis and Andreassen2008).
Emotion
Patients with schizophrenia often demonstrate a restricted emotional range and diminished ability to experience pleasure (anhedonia) (see, for example, Bleuler, Reference Bleuler1911; Kraepelin, Reference Kraepelin1917; Carpenter et al. Reference Carpenter, Heinrichs and Wagman1988). Anhedonia correlates significantly with other negative symptoms of schizophrenia such as affective flattening, avolitional pathology including asociality, apathy, alogia, and is independent of positive symptoms and disorganization (Blanchard & Cohen, Reference Blanchard and Cohen2006). Anhedonia has been thought to be a marker of a genetic liability to schizophrenia (Chapman et al. Reference Chapman, Chapman and Raulin1976; Glatt et al. Reference Glatt, Stone, Faraone, Seidman and Tsuang2006) and may be associated with anhedonia in the general population (Tomppo et al. Reference Tomppo, Hennah, Miettunen, Järvelin, Veijola, Ripatti, Lahermo, Lichtermann, Peltonen and Ekelund2009). By contrast, restricted emotional range and anergia are not commonly associated with BD except when secondary to depression. Anhedonia is not a liability marker for BD (Katsanis et al. Reference Katsanis, Iacono, Beiser and Lacey1992; Etain et al. Reference Etain, Roy, Henry, Rousseva, Schürhoff, Leboyer and Bellivier2007).
Symptom similarity
The symptom commonality across the disorders in DSM-IV-TR schizophrenia and related psychoses is based on reality distortion (i.e. hallucinations and delusions). Other criteria A symptoms for schizophrenia are used to differentiate this diagnosis from other psychotic disorders (e.g. disorganization, negative symptoms, psychomotor abnormalities). Reality distortion is associated with many disorders and is not decisive for etiologically based classifications. Bleuler (Reference Bleuler1911) considered hallucinations and delusions as secondary manifestations and Kraepelin (Reference Kraepelin1917) viewed the combination of avolition and dissociative thought, and the difference between chronic course and episodic course as the defining clinical features that separated dementia praecox from manic-depressive disorder. Schneiderian first-rank symptoms are, however, reality distortion phenomena. DSM-IV-TR criteria for schizophrenia can be met by hallucinations and delusions alone, or even just delusions, if bizarre. What is the affinity of SPD and BD in light of a traditional view of schizophrenia or the current DSM emphasis on reality distortion?
The answer for SPD is clear at a definitional level. If magical ideation and perceptual aberrations reach psychotic severity, the case meets criteria for a psychotic diagnosis. Nonetheless, there is a phenomenological similarity where schizotypal pathology can be viewed on a continuum with psychosis, and similarities reviewed above in other spectrum criteria support a close relationship to schizophrenia.
Comparing psychopathology of BD and schizophrenia supports separateness. Much of the manifest pathology of BD is mood disturbance, often with an episodic pattern. Reality distortion may not occur in BD-II, and is often not present during episodes of BD-I. First-rank symptoms occur in both, but more frequently in schizophrenia. Bizarre reality distortion experiences (Jaspers, Reference Jaspers1962) in schizophrenia contrast with the mood-congruent delusions in BD. Thought disturbance is present in both syndromes, but is quite different. Excessive, grandiose and pressured thought and speech is typical in one syndrome whereas disorganized speech with impoverished content is observed in the other. Although depressed affect is present in many cases of schizophrenia, other cases have restricted affect. Excessive affect is typical of BD except in some chronic cases. Avolition defines at least a subtype of schizophrenia and is not associated with BD (Kirkpatrick et al. Reference Kirkpatrick, Buchanan, Ross and Carpenter2001; Fischer & Carpenter, Reference Fischer and Carpenter2009).
In summary, SPD is separated from the psychoses group by definition although manifestations can be viewed as a mild version of schizophrenia. However, BD and schizophrenia are differentiated on a symptomatic basis.
High rates of co-morbidity among disorders as currently defined
Co-morbidity regarding domains of psychopathology is extensive. Depressive, obsessive, anxiety, attention, social, motor, suicidal, sleep disturbance and other psychopathologies are frequently observed in patients with diagnoses in the psychoses cluster. Overlap with many other disorders (van Os et al. Reference van Os, Hanssen, Bijl and Ravelli2000) does not suggest joining the various diagnostic categories in the same grouping. A more meaningful question is whether schizophrenia and other psychotic disorders are co-morbid separate disease entities such as diabetes. Schizophrenia and BD are associated with high prevalence of substance abuse, but this is true of many disorders. Schizophrenia is co-morbid with aspects of the metabolic syndrome including diabetes, but this may relate to lifestyle rather than shared pathology. Co-morbidity at the symptom level can be created by definition as exemplified by the diagnostic category of schizo-affective disorder (Malhi et al. Reference Malhi, Green, Fagiolini, Peselow and Kumari2008).
Course of illness
Schizophrenia does not have a ‘typical’ course (Carpenter & Kirkpatrick Reference Carpenter and Kirkpatrick1988; Harding Reference Harding1988), and the course pattern is variable for BD. Recurrent episodes in schizophrenia tend to be manifestations of symptoms observed previously whereas BD may have distinctive presentations at opposite ends of an affect continuum across episodes. However, more patients with schizophrenia than BD have continuous, rather than phasic, courses. It should be noted that this distinction is based on a proportion of cases rather than unique disorder patterns, and good prognosis cases are often under-represented in long-term course studies.
Both BD affect disturbance and schizophrenia psychosis have typical age–incidence curves with onset in young people (Kennedy et al. Reference Kennedy, Everitt, Boydell, van Os, Jones and Murray2005b) and both display earlier onset in men (Kennedy et al. Reference Kennedy, Boydell, Kalidindi, Fearon, Jones, van Os and Murray2005a). However, core schizophrenia pathologies (e.g. negative symptoms and cognitive impairment) begin much earlier in most cases. Schizophrenia is associated with a decline in cognition during developmental years and trait-like stability of impairments over time (Woodberry et al. Reference Woodberry, Guiliano and Seidman2008). BD is not generally associated with developmental impairment, prompting the suggestion that the major difference between the two conditions is that schizophrenia, but not BD, is developmental in origin (Murray et al. Reference Murray, Sham, van Os, Zanelli, Cannon and McDonald2004). The distinct longitudinal patterns of schizophrenia and BD separated the disorders in the original Kraepelinian concept, but heterogeneity in course is present in each syndrome.
Avolitional pathology begins early and afflicts a minority of cases of schizophrenia. This pathology is rare in the pre-onset and early course of BD. This schizophrenia subgroup is substantially different from BD, but is also distinguished from other schizophrenia subgroups (Kirkpatrick et al. Reference Kirkpatrick, Buchanan, Ross and Carpenter2001).
Treatment response
All antipsychotic drugs reduce the experience of psychosis, regardless of diagnostic category, and may reduce relapse rates. All share a mechanism of action at the dopamine D2 receptor. By contrast, lithium has therapeutic and prophylactic effects for depression and mania in BD patients but is not documented as effective in schizophrenia patients. One study that did examine the effects of lithium in schizophrenia patients found that effects were only evident on mood (Johnstone et al. Reference Johnstone, Crow, Frith and Owens1988). Antidepressant drugs have efficacy in BD, but efficacy has not been established in schizophrenia. Although controversial, there are data suggesting that antidepressant drugs may trigger mania in BD but may decrease psychotic relapse in schizophrenia (Siris et al. Reference Siris, Bermanzohn, Mason and Shuwall1994; Ghaemi et al. Reference Ghaemi, Hsu, Soldani and Goodwin2003). In general, psychopharmacology effects suggest substantial differences in treatment response except with antipsychotic drugs. The latter are not diagnostically specific.
Conclusions
The primary argument for grouping several disorders in a psychosis cluster is shared reality testing pathology and the absence of a compelling case for location elsewhere (e.g. psychosis with Alzheimer's disease). We propose keeping this grouping, incorporating the psychotic disorders already clustered in DSM-IV-TR. There are few data testing the fit between these disorders on the 11 criteria except for similarity in reality distortion symptom criteria. The more challenging issue involves grouping SPD and BD with the psychoses cluster.
The question of adding SPD is cogent because compelling similarities in many of the 11 validating criteria are documented. The main reservation relates to failure of cases to manifest psychotic symptoms and the fact that antipsychotic drugs are not front-line therapy. If psychosis emerges, the diagnosis is changed. A decision on this for DSM-V must also consider how personality disorders are conceptualized.
More complicated is the question of BD. There is substantial overlap in neural substrate, biomarkers, genetic and environmental effects, cognition, and aspects of symptoms, treatment and course. Yet, in each of these areas there are data supporting unique features. There is no consensus on whether any of the data reviewed above are decisive for validation of cluster membership. An additional consideration involves comparing BD with other mood disorders on the 11 proposed criteria (Goldberg et al. Reference Goldberg, Andrews and Hobbs2009a). Similarities between BD and schizophrenia are extensively based on BD-I with psychosis. It is likely that BD-II will overlap extensively with mood disorders, and the question of splitting BD-I and BD-II will have to be addressed. There may be insufficient evidence to de-link these two forms of BD at present. The fact that many BD patients never manifest psychosis argues against BD joining the psychosis chapter, as does the substantial difference in phenomenology.
Grouping BD and SPD with the current psychoses group could facilitate the identification of shared mechanisms of pathophysiology. The cluster could facilitate investigative focus on crucial issues that distinguish between classes within the cluster and that define the porous boundaries across current classes. This process will be facilitated by DSM-V if key dimensions of pathology associated with the grouping are identified and the paradigm for etiological investigations shifts from diagnostic class/syndrome to pathological dimension (Strauss et al. Reference Strauss, Carpenter and Bartko1974; Carpenter & Buchanan, Reference Carpenter, Buchanan, Schulz and Tamminga1989) or endophenotypes (Gottesman & Gould, Reference Gottesman and Gould2003).
There are several limitations in proposing the psychoses cluster based on the 11 spectrum criteria. Five caveats require attention. First, there is insufficient evidence on the etiology and pathophysiology to base group membership on causality. Second, the ultimate determination of group membership for disorders at the border of proposed clusters requires examination of similarities and differences with both clusters. Third, in-depth phenomenology and pattern of illness are not examined in most studies, and state versus trait issues may be crucial. Fourth, psychosis is the main defining feature for grouping these disorders. An alternative is to regard psychosis, especially reality distortion symptoms, as a common manifestation of many disorders and not rare in the general population. Fifth, it is not known whether the 11 spectrum criteria would support the DSM-IV cluster of schizophrenia and related disorders. Groupings would be very different if defined by other trait pathologies of schizophrenia such as avolitional pathology. In summary, the proposed move of BD-I and BD-II from a mood cluster to a psychosis cluster receives only modest support from data relating to the 11 spectrum criteria. This support may be insufficient to overcome tradition and the substantial symptomatic and therapeutic differences with schizophrenia. There is substantial support for SPD, but psychosis as a defining feature may preclude inclusion of a personality disorder in the psychosis cluster.
Declaration of Interest
Dr Carpenter reports European Regional Patent No. 1487998 (6 June 2007) ‘Methods for Diagnosing and Treating Schizophrenia’ with no potential personal financial reward (proceeds pledged to the Maryland Psychiatric Research Center). In the past 12 months Dr Carpenter has been a consultant to Cephalon and Teva. Dr Bustillo is the speaker for CME LCC and reviewed a book chapter for Merck. Dr Thaker has received a research grant from Mitsubishi Tanabe Pharma. Dr van Os is an unrestricted research grant holder with, or has received financial compensation as an independent symposium speaker from, Eli Lilly, BMS, Lundbeck, Organon, Janssen-Cilag, GSK and AstraZeneca. Drs Krueger and Green report no conflicts of interest relating to this paper.
