Lorenzo Mazzarini is with the Bipolar Disorder Program, Institute of Neuroscience, Hospital Clinic, University of Barcelona, IDIBAPS, CIBER-SAM, Barcelona, Spain; and with the NESMOS Department, Sapienza University, and Sant’Andrea Hospital, Rome, Italy. Eduard Vieta is with the NESMOS Department, Sapienza University, and Sant’Andrea Hospital, Rome.
Dr. Mazzarini has disclosed no relevant financial relationships. Dr. Vieta has disclosed the following relevant financial relationships: AstraZeneca, Bristol-Myers Squibb, Eli Lilly, Forest Research Institute, GlaxoSmithKline, Janssen-Cilag, Jazz, Lundbeck, Novartis, Organon, Otsuka, Pfizer Inc., Sanofi-Aventis, Servier, and UBC: Advisor, consultant, research grant recipient, and/or member of speaker’s bureau; and Spanish Ministry of Health, Spanish Ministry of Science and Education, Stanley Medical Research Institute, and the 7th Framework Programme of the European Union: Research grant recipient.
Dr. Mazzarini and Dr. Vieta have disclosed that this research is supported and funded in part by the CIBERSAM (Instituto Carlos III, Madrid) and the Seventh European Framework Programme, European Network of Bipolar Research Expert Centers (ENBREC).
Dr. Mazzarini and Dr. Vieta have disclosed no relevant financial relationships.
Address correspondence to: Eduard Vieta, NESMOS Department, Sapienza University, 2nd Medical School, Sant’Andrea Hospital, Via di Grottarossa 1035–1039, 00189 Rome, Italy.
“There is nothing more likely to drive a man mad than an obstinate, constitutional preference for the true to the agreeable.”
In contrast with the other branches of medicine, in which classifications and diagnostic systems for phenotypic definitions are based on etiopathophysiology and/or biological measures (for example, blood glucose in diabetes mellitus), psychiatric taxonomy is suffering from the absence of a legitimate biomedical model because it depends on a descriptive approach. Thus, because valid biological markers for the definition of diagnostic criteria are absent, psychiatric nosology can only be established according to the following measures: clinical descriptions; clinical course and outcome; treatment response; and separation from other disorders. Currently, the classification of psychiatric disorders, as described in the Diagnostic and Statistical Manual of Mental Disorders, fourth edition, text revision (DSM-IV-TR) is mostly based on a categorical approach. This model may be helpful in terms of diagnostic agreement (reliability), ease of application in clinical practice, communication among practitioners or investigators, and academic teaching. However, it proved to be unsatisfactory in terms of diagnostic validity and insufficiently precise in considering the boundaries between diseases. This is the reason why symptom-centric diagnostic definitions, which are used in psychiatric nosology, often represent the result of an idealized abstraction rather than a reflection of the complexity and the heterogeneity of mental disorders.
The dimensional approach is an alternative to the categorical one. Compared with the latter, the model on which the dimensional approach is based, identifying psychopathological dimensions through factor analysis of symptoms, may be closer to clinical reality and may provide greater utility in research contexts. A dimensional perspective, which underlines differences of degree rather than differences of kind, may better explain natural boundaries between disorders, compared with artificial boundaries. Despite these advantages, this approach may carry reliability problems and may be difficult to apply in clinical settings.
To date, the debate between proponents and opponents of those two approaches is in full swing. The traditional Kraepelinian dichotomy represents the major nosologic conundrum in the classification of psychiatric disorders. There has been an intensive discussion if schizophrenia and bipolar disorder are distinct or related and potentially overlapping illnesses.1 We will try to clarify this controversy utilizing a slightly modified version of the criteria of Robinson and Guze,2 who 30 years ago introduced a biomedical approach to psychiatric nosology. As formulated by the authors, the validity of psychiatric diagnosis may rely on several domains:
Content validity, involving basically symptoms and clinical diagnostic criteria;
Concurrent validity, defined by neurobiological correlates, such as laboratory findings, neuroimaging, neuropsychology, and genetics, and also, from our perspective, family studies and treatment response;
Predictive validity, which has mainly to do with diagnostic stability over time, and
Discriminant validity, which involves tracing the border from other disorders.
This formulation may have considerable scientific effect in reframing the boundaries and overlaps between schizophrenia and bipolar disorder (BPD), as well as in unraveling this nosologic knot.
The Validity of BP and SZ as Two Distinct Diagnostic Categories
Content-Discriminate Validity: Problems of Current Definitions of BPD vs. Schizophrenia
Currently, by not including psychotic symptoms in the core criteria of the definition of BPD, leaving them as mere correlates of impairment or severity (fourth digit or specifier), the DSMs have indirectly reinforced the traditional misconception that psychosis is pathognomonic of schizophrenia. This may lead to misdiagnosing BPD patients as schizophrenia patients, thus entraining them into a path of inappropriately adopted therapeutic strategies. At least 60% of BPD individuals have experienced at least one lifetime psychotic episode.3 BPD and schizophrenia can have similar cross-sectional presentations, including Schneiderian first-rank, bizarre or mood-incongruent symptoms, catatonia, and disorganization of behavior.3 Affective symptoms and syndromes are often seen in patients with schizophrenia, and the so-called “negative symptoms” of schizophrenia, such as avolition, lack of motivation, anhedonia, and flat affect, may actually overlap with symptoms of depression.4 These data suggest that there are no discrete demarcations in the distribution of psychotic and affective symptoms between BPD and schizophrenia. Thus, in terms of psychiatric diagnoses, this phenomenological overlap shows that the use of disease-pathognomonic symptoms, considered discriminant validators between these two disorders, has to be questioned. Another basic assumption, which may be too simplistic, is that course and outcome, as stated by Kraepelin, differentiate schizophrenia from BPD. Compared with schizophrenia, BPD is generally viewed with a less severe course and with a better prognosis. Although it is true that most patients with schizophrenia present a severe deficit syndrome as an outcome that is not as common in bipolar patients, prognosis in schizophrenia is not invariably bad as it is not always benign in BPD.5,6 Therefore, distinguishing BP from schizophrenia just on severity in terms of course or outcome appears to be more a theoretical paradigm than a clinical fact based on scientific data inasmuch as “positive” psychotic symptoms and chronic unremitting course, considered as pivotal for schizophrenia, may be found also in severe BP cases.
A dimension-based approach may lead one to consider disorder spectra on the grounds of presence/absence of psychotic features, relative high/low severity, and more or less chronicity (time ill/time healthy), or to consider a single disorder as one with defined quantities of psychotic symptoms of a given severity and exacerbation-remission patterns. Thus, psychotic mood disorders may represent an intermediate state between schizophrenia (more severe) and non-psychotic mood disorders (less severe). In summary, clinical evidence suggests that a descriptive approach, considering the heterogeneity and complexity of schizophrenia and BPD, cannot represent alone the key feature for distinguishing those entities.
Concurrent Validity: Endophenotypes as Potential Diagnostic Validators
As a result of 30 years of scientific progress, today it seems possible to reconsider the meaning of neurobiological correlates, as formulated by Robinson and Guze,2 into a more extensive sense: the concept of endophenotype. Even if no laboratory marker proved to be specific in identifying any of the DSM-defined syndromes, there has been substantial progress in using biological findings as diagnostic validators, including genetics, family data, biomarkers, and therapeutic response. With regard to genetics, schizophrenia and BPD occur together in the same families more frequently than by chance.7 In genome linkage analyses of these disorders, at least five distinct loci have been implicated as being linked to susceptibility for schizophrenia and BPD.7 Recent molecular genetic studies suggest an overlap in the genes that predispose to both disorders, as G72, DiSC1, COMT, BDNF, and DSTNBP1.8 However, this overlap has a small effect size, and the magnitude of relative risk mediated by sequence variants in each specific susceptibility gene is quite low.9 Therefore, if it is desirable that psychiatric genetics may be used as diagnostic validator concurrently with clinical criteria in the near future, to date, it seems premature to base a new nosology of schizophrenia and BP just on these findings. As formulated by Robinson and Guze, family studies were originally excluded from the elements considered to be necessary for defining concurrent validity. Therefore, we believe that family data represent a “conditio sine qua non” for the definition of an endophenotype.
According to Gottesman and Gould,10 endophenotypes are neurophysiological, biochemical, neuroimaging, neurocognitive, and endocrinological components, associated with the target disorder. The endophenotype construct, borrowing from entomology, represents the effort to fill the gap between genotype and phenotype as a hypothetical “bridge” between DNA and behavior. Endophenotypes are easily and reliably measured, they are heritable, state-independent, they cosegregate with the illnesses within families and are also present in unaffected relatives. Therefore, endophenotypes represent biological deficits nearer to the genetic effects than to the end-stage disease state. Thus, these potential biomarkers are quantitative traits common to patients (individuals with a clarified diagnosis), as well as to their healthy first degree-relatives. In fact, they are not only indices of illness but also indices of a risk of illness. Some endophenotypes may be associated to more than one DSM-IV-TR defined diagnostic category, while some others may be more specific. Studying endophenotypes in schizophrenia and BP may inform if, how, and where these disorders diverge.
Several candidate endophenotypes in schizophrenia and BPD patients and in their unaffected relatives are neurophysiological markers. Neurophysiological tests have shown that schizophrenia and BPD patients and their healthy first-degree relatives present abnormalities in smooth pursuit eye movements, sensory gating P50, sensory-motor gating prepulse inhibition (PPI), and P300 evoked response latency.11 These electrophysiologic disturbances may be more intense in schizophrenia.
Neuroimaging studies comparing patients with schizophrenia and BP and their relatives provide several neuroanatomical endophenotypic markers specifically or generally associated with schizophrenia and BPD. Genetically mediated gray-matter volume deficits in prefrontal-striatum-thalamic and temporal areas may be found only in schizophrenia, whereas the genetic risk for bipolar disorder may be associated with gray matter volume loss in a more restricted set of regions, principally the right anterior cingulated gyrus and ventral striatum. However, a white matter variation in the left frontal and temporoparietal areas is an endophenotypic marker generally associated with genetic risk for both disorders, suggesting that dysconnectivity between these regions may be linked with risk for psychosis in general.12 Although individual studies may report abnormalities in brain structure shared by schizophrenia and BPD,13 such as temporal horn enlargement, it seems that when brain imaging data are pooled and meta-analyzed, the two disorders are separated.14 Again, if the disorders may clearly separate from one another, this may not apply to individual patients.
There is growing interest to investigate neurocognition as a putative endophenotype for subjects with schizophrenia and BPD.15 At first sight, neurocognitive deficits in bipolar patients and their unaffected relatives appear to be less severe than the impairments associated in patients with schizophrenia. Currently, verbal learning memory and verbal working memory seem to be potential endophenotypes for BPD, whereas executive functions, verbal fluency, and working memory are candidate endophenotypes for SZ.16,17 However, there is also evidence for shared endophenotypes because set shifting and verbal memory impairments exist in both disorders. This suggests that schizophrenia and mood disorders are on a neurobiological continuum, where psychotic mood disorder lies in the middle.18 To date, none of these endophenotypes is present in classification systems because none of these findings is specific enough to be used as a diagnostic test in clinical practice. It might be argued that severity or illness duration might be related with specific neurocognitive impairments, but data are still controversial.19 Factors related to drug treatment may contribute to type of deficit found,20 but the pattern of deficits between the disorders still differs.16,21 However, approaching BP and schizophrenia “endophenotypically” can provide some support to diagnostic boundaries, overcoming genetic and phenotypic complexities. Moreover, endophenotypes represent the first stage toward the identification of biomarkers of risk for SZ and BP. This may allow us to predict a possible development of either disorder in unaffected relatives.
With regard to treatment response, the efficacy of many psychotropic medications cuts across the DSM-defined categories. In fact, the utility of some anti-psychotics in schizophrenia as well as in BP (at least as adjunctive agents) shows an important limitation of “drug categorization,” indicating the involvement of dopamine dysregulation as a common key mechanism in their etiology.22 At the same time, the poor responsiveness of schizophrenia to lithium and anticonvulsants is consistent with a categorical difference between these disorders.23 Treatment response may offer insights into commonalities and differences between BPD and schizophrenia, with some data suggesting spectrum and overlapping patterns, whereas other data support lack of overlap or even reciprocal patterns.
Discriminant Validity of Separating BPD from Schizophrenia
As in every branch of medicine, the separation of one disorder from another (for example, BPD from schizophrenia), implies the knowledge of etiopathogenesis. In DSM-IV-TR nosology, this concept does not appear in the formulation of any diagnosis, except in the exclusion criteria “not be due to the direct physiological effects of a substance nor a general medical condition.” This hierarchical diagnostic system uses a descriptive approach despite the lack of biologically valid measures for the phenotype definition. On the one hand, the ambiguous schizoaffective disorder (a trick to avoid possible comorbidity of SZ and BP) and, on the other, the growing numbers of patients that fulfill criteria over more than two different conditions (artifactual comorbidity) may simply be the result of trying to trace boundaries where actually there is continuity. Along this path, we may reconsider the diagnoses for those patients who may be diagnosed as Schizoaffective Disorder at one time and Post-psychotic Depression in Schizophrenia at another. The introduction of a Psychotic Mood Disorder category may place these patients more on the bipolar side and allow more appropriate treatment that impairs less their cognition.
Temporal Stability: The Longitudinal Aspect of BPD and Schizophrenia
The temporal stability of a diagnosis is a measure of its predictive validity. The longitudinal course of major mental disorders may be endorsed as a criterion for distinguishing two entities (with the downside of having to wait-and-see to make it right). The temporal stability of BPD and schizophrenia has generally been reported to be high, ranging between 80% and 90%.24 However, the phenomenon of syndrome shift during long-term course can occur in both disorders.25 When it happens, DSM-IV-TR is not able to answer adequately to the question: What type of disorder do patients have? For instance, post-psychotic depression in schizophrenia may be evaluated as comorbidity between two different disorders, as a cosyndromatology26 in the same disorder, or as something new resulting from the mixture of two. Therefore, a longitudinally based diagnostic process, mostly in first-episode psychosis,27 and the use of potential course specifiers, such as the predominant polarity in BP,28 are needed.
There is no doubt that the DSM-IV-TR, as used currently to diagnose psychosis, lacks an etiological and pathophysiological basis. Consequently, it is not valid in strictly scientific terms. To date, if it is clear on one hand that the traditional paradigm of the Kraepelinian dichotomy has been challenged first by Kraepelin himself29 and then by research showing phenomenological and biological overlap between schizophrenia and BPD, it seems premature on the other hand to dismiss completely Kraepelin’s initial construct. Actually, more reliable, objective, sensitive, and specific laboratory assessments are needed to inform on if, why, how and where these disorders diverge. The endophenotypic approach can potentially contribute to enhance the understanding of etiopathophysiologic pathways in both disorders and to clarify the qualitative and quantitative differences between schizophrenia and BPD. Decomposing psychiatric complex phenotypes into biomarkers (endophenotypes) may not only overcome the problem of genetic heterogeneity in each individual disorder but also help to define homogeneous etiological subgroups of schizophrenia and BPD. Thus, the use of a “psychiatric toolbox,” built on an endophenotypic perspective, along with the identification of specific environmental risk factors and clinical features, will settle the issue whether Kraepelin was right or not. Meanwhile, with the expectation of further advances in biopsychiatric research, we have suggested a modular approach for the classification of mental disorders. This may integrate categorical and dimensional issues, laboratory data, nonpsychiatric medical conditions, psychological assessment, and social issues in a comprehensive yet practical approach.30 The advantage of this modular diagnostic system is represented by the possibility of a more exhaustive therapeutic intervention, which includes psychoeducation and neuropsychological rehabilitation, in addition to pharmacological treatment.31,32 Obviously, integrating all new information on dimensions and categories in new diagnostic systems will not affect current psychiatric practice dramatically, but it might be that a smooth transition is already on the way.33 We believe that a modular approach might better explain current clinical variance in the domain of the psychoses.30 Our views are changing and this affects our practices, allowing better quality of life, earlier detection of psychosis, perhaps better management, and specialized, more individualized programs (personalized medicine). Actually, now it may be the right time to refine the current nosology in format and content, making it more comprehensive, logical, and data-driven.
- Craddock N, Owen MJ. The beginning of the end for the Kraepelinian dichotomy. Br J Psychiatry. 2005;186:364–366. doi:10.1192/bjp.186.5.364 [CrossRef]
- Robins E, Guze SB. Establishment of diagnostic validity in psychiatric illness: its application to schizophrenia. Am J Psychiatry. 1970;126(7):983–987.
- Keck PE, McElroy SL, Havens JR, et al. Psychosis in bipolar disorder: phenomenology and impact on morbidity and course of illness. Compr Psychiatry. 2003;44(4):263–269. doi:10.1016/S0010-440X(03)00089-0 [CrossRef]
- Ketter TA, Wang PW, Becker OV, Nowakowska C, Yang Y. Psychotic bipolar disorders: dimensionally similar to or categorically different from schizophrenia?J Psychiatr Res. 2004;38(1):47–61. doi:10.1016/S0022-3956(03)00099-2 [CrossRef]
- Mason P, Harrison G, Glazebrook C, Medley I, Croudace T. The course of schizophrenia over 13 years. A report from the International Study on Schizophrenia (ISoS) coordinated by the World Health Organization. Br J Psychiatry. 1996;169(5):580–586. doi:10.1192/bjp.169.5.580 [CrossRef]
- Martinez-Aran A, Vieta E, Torrent C, et al. Functional outcome in bipolar disorder: the role of clinical and cognitive factors. Bipolar Disord. 2007;9(1–2):103–113. doi:10.1111/j.1399-5618.2007.00327.x [CrossRef]
- Berrettini W. Bipolar disorder and schizophrenia: not so distant relatives?World Psychiatry. 2003;2(2):68–72.
- Craddock N, O’Donovan MC, Owen MJ. Genes for schizophrenia and bipolar disorder? Implications for psychiatric nosology. Schizophr Bull. 2006;32(1):9–16. doi:10.1093/schbul/sbj033 [CrossRef]
- Kendler KS. Reflections on the relationship between psychiatric genetics and psychiatric nosology. Am J Psychiatry. 2006;163(7):1138–1146. doi:10.1176/appi.ajp.163.7.1138 [CrossRef]
- Gottesman II, Gould TD. The endophenotype concept in psychiatry: etymology and strategic intentions. Am J Psychiatry. 2003;160(4):636–645. doi:10.1176/appi.ajp.160.4.636 [CrossRef]
- Thaker GK. Neurophysiologicalal endophenotypes across bipolar and schizophrenia psychosis. Schizophr Bull. 2008;34(4):760–773. doi:10.1093/schbul/sbn049 [CrossRef]
- McDonald C, Bullmore ET, Sham PC, et al. Association of genetic risks for schizophrenia and bipolar disorder with specific and generic brain structural endophenotypes. Arch Gen Psychiatry. 2004;61(10):974–984. doi:10.1001/archpsyc.61.10.974 [CrossRef]
- Roy PD, Zipursky RB, Saint-Cyr JA, Bury A, Langevin R, Seeman MV. Temporal horn enlargement is present in schizophrenia and bipolar disorder. Biol Psychiatry. 1998;44(6):418–422. doi:10.1016/S0006-3223(98)00105-X [CrossRef]
- Kempton MJ, Geddes JR, Ettinger U, Williams SC, Grasby PM. Meta-analysis, database, and meta-regression of 98 structural imaging studies in bipolar disorder. Arch Gen Psychiatry. 2008;65(9):1017–1032. doi:10.1001/archpsyc.65.9.1017 [CrossRef]
- Daban C, Martinez-Aran A, Torrent C, et al. Specificity of cognitive deficits in bipolar disorder versus schizophrenia. A systematic review. Psychother Psychosom. 2006;75(2):72–84. doi:10.1159/000090891 [CrossRef]
- Balanzá-Martínez V, Rubio C, Selva-Vera G, et al. Neurocognitive endophenotypes (endophenocognitypes) from studies of relatives of bipolar disorder subjects: a systematic review. Neurosci Biobehav Rev. 2008;32(8):1426–1438. doi:10.1016/j.neubiorev.2008.05.019 [CrossRef]
- Snitz BE, Macdonald AW, Carter CS. Cognitive deficits in unaffected first-degree relatives of schizophrenia patients: a meta-analytic review of putative endophenotypes. Schizophr Bull. 2006;32(1):179–194. doi:10.1093/schbul/sbi048 [CrossRef]
- Martinez-Aran A, Torrent C, Tabares-Seisdedos R, et al. Neurocognitive impairment in bipolar patients with and without history of psychosis. J Clin Psychiatry. 2008;69(2):233–239. doi:10.4088/JCP.v69n0209 [CrossRef]
- Depp CA, Moore DJ, Sitzer D, et al. Neurocognitive impairment in middle-aged and older adults with bipolar disorder: comparison to schizophrenia and normal comparison subjects. J Affect Disord. 2007;101(1–3):201–209. doi:10.1016/j.jad.2006.11.022 [CrossRef]
- Bora E, Yucel M, Pantelis C. Cognitive endophenotypes of bipolar disorder: a meta-analysis of neuropsychological deficits in euthymic patients and their first-degree relatives. J Affect Disord. 2009;113(1–2):1–20. doi:10.1016/j.jad.2008.06.009 [CrossRef]
- Hill SK, Harris MS, Herbener ES, Pavuluri M, Sweeney JA. Neurocognitive allied phenotypes for schizophrenia and bipolar disorder. Schizophr Bull. 2008;34(4):743–759. doi:10.1093/schbul/sbn027 [CrossRef]
- Vieta E, Goikolea JM. Atypical antipsychotics: newer options for mania and maintenance therapy. Bipolar Disord. 2005;7(Suppl 4):21–33. doi:10.1111/j.1399-5618.2005.00212.x [CrossRef]
- Soares JC, Gershon S. The psychopharmacologic specificity of the lithium ion: origins and trajectory. J Clin Psychiatry. 2000;61(Suppl 9):16–22.
- Schimmelmann BG, Conus P, Edwards J, McGorry PD, Lambert M. Diagnostic stability 18 months after treatment initiation for first-episode psychosis. J Clin Psychiatry. 2005;66(10):1239–1246. doi:10.4088/JCP.v66n1006 [CrossRef]
- Marneros A, Roettig S, Roettig D, Tscharntke A, Brieger P. The longitudinal polymorphism of bipolar I disorders and its theoretical implications. J Affect Disord. 2008;107(1–3):117–126. doi:10.1016/j.jad.2007.08.009 [CrossRef]
- Möller HJ. Systematic of psychiatric disorders between categorical and dimensional approaches: Kraepelin’s dichotomy and beyond. Eur Arch Psychiatry Clin Neurosci. 2008;258(Suppl 2):48–73. doi:10.1007/s00406-008-2004-3 [CrossRef]
- Benabarre A, Vieta E, Colom F, Martínez-Arán A, Reinares M, Gastó C. Bipolar disorder, schizoaffective disorder and schizophrenia: epidemiologic, clinical and prognostic differences. Eur Psychiatry. 2001;16(3):167–172. doi:10.1016/S0924-9338(01)00559-4 [CrossRef]
- Colom F, Vieta E, Daban C, Pacchiarotti I, Sánchez-Moreno J. Clinical and therapeutic implications of predominant polarity in bipolar disorder. J Affect Disord. 2006;93(1–3):13–7. doi:10.1016/j.jad.2006.01.032 [CrossRef]
- Kraepelin E. Die Erscheinungsformen des Irreseins. Zeitschrift für die gesamte Neurologie und Psichiatrie. 1920;62(1):1–29. doi:10.1007/BF02887354 [CrossRef]
- Vieta E, Phillips ML. Deconstructing bipolar disorder: a critical review of its diagnostic validity and a proposal for DSM-V and ICD-11. Schizophr Bull. 2007;33(4):886–892. doi:10.1093/schbul/sbm057 [CrossRef]
- Colom F, Vieta E, Martinez-Aran A, et al. A randomized trial on the efficacy of group psychoeducation in the prophylaxis of recurrences in bipolar patients whose disease is in remission. Arch Gen Psychiatry. 2003;60(4):402–407. doi:10.1001/archpsyc.60.4.402 [CrossRef]
- Martínez-Arán A, Vieta E, Reinares M, et al. Cognitive function across manic or hypomanic, depressed, and euthymic states in bipolar disorder. Am J Psychiatry. 2004;161(2):262–270. doi:10.1176/appi.ajp.161.2.262 [CrossRef]
- Ketter TA, Wang PW, Becker OV, Nowakowska C, Yang Y. Psychotic bipolar disorders: dimensionally similar to or categorically different from schizophrenia?J Psychiatr Res. 2004;38(1):47–61. doi:10.1016/S0022-3956(03)00099-2 [CrossRef]