Translational validity across neuroscience and psychiatry
Drozdstoj St. Stoyanov, Stefan J. Borgwardt, and Somogy Varga
In all fields of inquiry, be it in the humanities or the sciences, an important goal is to establish the validity of theories, methods, and lastly knowledge about the world. In this context, validity (derived from the Latin validus) means “well-grounded” and “sound.” Nonetheless, the criteria by which we judge whether knowledge is valid differ between the fields. For instance, philosophy, and particularly the tradition of hermeneutics, addresses the conditions under which valid understanding and valid interpretation can proceed. Hermeneutics focuses on meanings rather than facts. In hermeneutics, we can say that we are dealing with valid knowledge if we are able to place the phenomenon under investigation in that “space of reasons,” to use a term coined by the philosopher John McDowell.
The picture is different with scientific knowledge. Scientific knowledge aims to provide us with plausible explanations and trustworthy predictions of phenomena in the world. While different scientific disciplines share many features, such as the generation of explanatory theories, reliance on observable evidence, and testing of hypotheses by experimental studies, they fundamentally differ in their methods of validating theoretical constructs. This may partly be why the meaning of validity tends to remain ambiguous and it tends to elude a neat cross-disciplinary definition.
In addition, the answer to the question of validity in the sciences also hinges on whether one adopts an instrumentalist or realist attitude toward the methods and results of scientific inquiry. While proponents of both attitudes agree that science advances by trial and error and generates genuine knowledge, the question that divides them is the nature of the knowledge thus generated. On an instrumentalist conception, the knowledge that scientific investigations give us is understood as trustworthy quantitative predictions of phenomena. In this case, scientific knowledge is instrumental: it provides us with suitable information about some limited domain of phenomena, and it explains and solves problems associated with that domain.
On a realist conception, there is a general agreement that a valid theory is not merely explanatorily powerful, but in an important sense captures the nature of a mind-independent world and thus “cuts nature at its joints.” In other words, the knowledge that scientific investigations offer us should be comprehended as true discovery and accurate description of the world “as it really is,” independently of human perceptions, theories, and methods of measurements.
It is quite clear that these positions are committed to very different underlying metaphysical pictures. Realists claim that the world is independent of particular theories and that scientific statements are truth-evaluable, i.e., they can be confirmed or rejected by empirical research. On the instrumentalist view, scientific statements are not truth evaluable, and should therefore not be evaluated based on how accurately they describe objective reality; rather, they are assessable by their usefulness, thus by how successfully they explain and predict particular phenomena.1
In psychiatry, the question of validity and validation is particularly important and complex. Since its birth in the nineteenth century, psychiatry has occupied a unique position within the science of medicine (Gadamer 1996: 163). In spite of continued efforts to bring psychiatry back within the boundaries of neurology, it remains an amalgam discipline, located on the border between science and the humanities. After much work on improving the reliability of diagnostic criteria, the question of the validity in psychiatry is becoming especially prominent.
One of the challenges that makes the issue of validity in psychiatry extremely complex is the so-called brain–mind problem, which is a modern version of the well-known mind–body problem. Some argue that mental phenomena should be reduced to underlying brain processes, while others maintain that intentionality, or the subjective perspective, is irreducible to brain processes. Since the introduction of the DSM-III in 1980, American psychiatry in particular has sidestepped this issue by circumventing etiological theories in favor of a descriptive approach. Disorders are pigeonholed by sets of symptoms that are mainly elicited by patient report and observation. One advantage of adopting the descriptive approach to classification is its improved reliability over prior systems.
However, while the descriptive approach was able to improve reliability, it was not designed to establish the validity of classifications. The expectation was that identifying and descriptively grouping covarying symptoms in clinical populations would be a major step toward explaining them by a common underlying etiology. Robins and Guze (1970) predicted that the validity of descriptively defined syndromes could be incrementally improved through increasingly precise clinical description, laboratory studies, delimitation of disorders, follow-up studies of outcome, and family studies. “Once fully validated, these syndromes would form the basis for the identification of standard, etiologically homogeneous groups that would respond to specific treatments uniformly” (Kupfer et al. 2002: xviii). However, the goal of explaining these syndromes with reference to an underlying etiology has not been achieved.
Reflecting upon the noteworthy development of the neurosciences in the second half of the twentieth century, thinkers such as Christopher Boorse (1975) claimed that psychiatry would inevitably evolve into a form of applied biological science. Today some think that the future psychiatry is destined to be “clinical neuroscience” (Reynolds et al. 2009). Many believe that the scope of mechanisms active in mental disorders can be confined to biological mechanisms. When those mechanisms are discovered, it is believed that psychiatry will merit the same scientific status as other areas of medicine. Neuroscience, it seems to many, is the solution to the problems of validation.
However, at the beginning of the twenty-first century, despite the early optimism of the “Decade of the Brain” neuroscientists have not discovered any biomarkers or laboratory tests for the most common psychiatric disorders. Those disorders with a confirmed biological disease or genetic defect listed in the DSM-5 (for, e.g., Alzheimer’s and other forms of dementia) fall under the scope of neurology. Critical voices from prominent psychiatrists have diminished the widespread hope that neuroscience will soon provide solutions to psychiatric questions. One of the problems is, as Miller (2010: 718) puts it, that “a mental disorder need not be triggered by, due to, or explained by brain pathology any more than a software bug must be a consequence of hardware failure.” Even in a seemingly clear-cut case in which it could be demonstrated that the etiology of a disorder involves causally active brain mechanisms, it remains a possibility that the respective mechanisms are causally affected by psychological events. Expressing his own pessimism about a simple biological etiology of mental disorders, Frances argues that a wide variety of pathways likely lead to the development of a disorder such as schizophrenia (Frances 2010). Frances somewhat pessimistically declares that diagnostic classification is the result of historical accretion and accident, and is not grounded on scientific necessity: “Our mental disorders are not more than fallible social constructs (but nonetheless useful if understood and applied properly)” (Frances 2010).2
Given a widespread (but not universal) skepticism of biological reduction, and the status of psychiatry as a hybrid discipline that embraces both the sciences and the humanities, making the case for realism and realist approaches to validity in psychiatry is—at least currently—out of reach. In this chapter, we attempt to draw the contours of a concept of translational validity, which is a non-conventional and instrumentalist approach to validation.
In this section we claim that there are substantial differences between validation in the natural sciences and humanities, and because psychiatry is a hybrid discipline that embraces both the sciences and the humanities, a non-conventional approach to validation is called for. We name this approach translational validity. We will further argue that neuroimaging is an important instrument for establishment of translational validity under some conditions, e.g., simultaneous administration of the diagnostic assessment tools and brain scan.
The disciplines we will focus on are clinical psychology, psychopathology, and neuroscience. Clinical psychology is a discipline that studies the reports of patients using interviews and inventories, and that often relies on ideographic methods for understanding mental phenomena. Although contemporary clinical psychology is usually considered to be quantitative and thus scientific, in fact the items of its different assessment tools represent decontextualized narratives composed from excerpts of the patients and/or professional narratives (Stoyanov et al. 2012, 2013).
Clinical psychology and psychopathology are considered distinct fields of inquiry in this chapter. Clinical psychology operates within a predominantly humanistic framework and is therefore more dimensional, whilst psychopathology is understood as an attempt to impose medical categorization upon mental phenomena which are described as “symptoms” and “syndromes,” unified in nosological blocks.
To some extent this distinction is provisional, since contemporary clinical psychology has already been mixed up with psychopathology and vice versa. However, they are still regarded as two distinct fields of expertise.
Validation in clinical psychology is based on two kinds of comparisons. In the first, a score on the rating scale under validation is correlated with another score on a rating scale, which is already assumed to be valid. The prototype for this approach is the Minnesota Multiphasic Personality Inventory (MMPI). Stoyanov et al. (2012) argued that the MMPI items were extracted statements/questions from patients’ narratives as they emerged in qualitative psychiatric interviews. The items were empirically sorted into scales on the basis of their ability to distinguish between a specific diagnostic group (such as people with depression) and a non-psychiatric population. In this comparison the psychiatric diagnosis was presumed to be correct and uncontestable. In the second kind of comparison the clinical scales are administered to many people. The statistical index called Cronbach’s alpha measures the extent to which all the items are measuring the same construct. Cronbach’s alpha in particular is usually classified as a measure of internal consistency reliability, but it can also be considered as a measure of (factorial) validity as well whenever the rating scale is supposed to be homogenous.
Although psychological tests are quantified and represent generalizations across persons, their explanatory power is limited by the qualitative patient reports on which they are based. In this way psychological assessments tend to lack explanatory potential located outside the clinical measures to underpin them and could therefore benefit from independent cross-validation, especially with the methods of neuroscience.
Contemporary neuroscience encompasses genetics, physiology, and functional neuroimaging. As a robust natural science, it seeks to discover objective scientific evidence about the mechanisms underlying mental disorders. Yet neuroscience suffers from many methodological limitations in terms of its validity and clinical utility (Borgwardt 2012; Stoyanov et al. 2012). To address validation, neuroimaging parameters need to demonstrate discriminative power at the single-subject level. Moreover, MRI modalities have to be calibrated across different scanners and centers, and provide good test–retest, inter-subject, and cross-scanner reliability. After reliability has been established, to achieve internal validity it needs to be determined that what is being measured is actually a clinically relevant psychopathological process. Also, neuroimaging needs to be applicable beyond research laboratory settings to clinical psychiatric situations (Rusconi and Mitchener-Nissen 2013).
Validation in psychopathology occupies the very borderline area in between clinical psychology and neuroscience. It is operating with hybrid objects, called “phenomena” (Berrios 2011), which cannot be exactly identified in the conventional operational languages of clinical psychology or neuroscience. Over the past decades after DSM III-R (1973), psychopathology has been operationalized with structured clinical interview protocols such as the Structured Clinical Interview for DSM (SCID, First et al. 2012), the Positive and Negative Syndrome Scale (PANSS, Kay et al. 1989), and the Montgomery–Asberg Depression Rating Scale (MADRS, Montgomery and Asberg 1979). These protocols are also based on patient narratives. From an epistemological point of view there is no substantial difference in cognitive content between psychological tests and structured clinical interviews; nonetheless, they remain distinguished in clinical practice. This leads to certain epistemic circularity in which dimensional rating scales (like MMPI) are validated backwards on structured clinical interviews (SCID), which are themselves validated on tests like the MMPI. Another problem with both tests and structured interviews is that they have led to what Andreasen (2006) called the “death of phenomenology.” In effect, the careful description of patients’ experiences has been replaced with conventional lists of reported symptoms. Yet this agenda misses the rationale for successful interplay and integration of the psychopathological quantitative assessment with neurobiological measures.
Meta-language (Berrios 2006) is a methodological tool for integrating the divergent disciplinary languages of psychiatry. The fundamental problem in integrating the different sources of psychiatric knowledge (clinical psychology, psychopathology, and neurosciences) is termed the “explanatory gap.” The explanatory gap refers to the incommensurability of the nomothetic and ideographic disciplinary languages (Broome 2008). From a practical standpoint this means that the construct of “depression” in clinical psychology/psychopathology and in neuroscience are defined and measured in sometimes incompatible ways.
In the tests of the clinical psychologists, depression represents a dimensional measure. A high score (above a certain cut-point) on a dimensional depression scale is taken to be a valid indicator of a depressive disorder.
Structured interviews used in psychopathology differ from the above-mentioned scales mainly in their observational, therefore presumably “objective” segment. This observation typically describes both verbal and nonverbal behavior of patients. As it has already been argued, the latter represents just another kind of structured “professional” narrative.
Neuroscience attempts to identify biological bases of disorders (e.g., genome-wide association studies, biomarkers at the level of serotonin transport and receptors, etc.), which are in turn practically untranslatable into clinical reality, especially into the assessments of descriptive psychopathology. This is one reason why it has not been possible to incorporate data from neuroscience into diagnostic criteria and contemporary classifications. This means that from an epistemological perspective, validity and validation in psychiatry are left at mono-disciplinary levels, either neurobiological or psychopathological.
In other words, patient narratives are hermeneutic but not explanatory and the measures obtained in neuroscience are potentially explanatory but not hermeneutic/meaningful, and that is the so-called explanatory gap. To manage this gap (or rather to escape from it) and avoid inter-paradigm controversies, contemporary psychiatry has adopted an instrumentalist approach to clinical taxonomy.
Therefore, the major issue which complicates the dialogue across the different disciplinary languages constituting psychiatry appears to be translation among them.
As a strategy for bridging the explanatory gap we propose a program of translational validation, which would use neuroimaging as a tool for improvement of cross-disciplinary instrumental psychiatric validity.
As has been stated elsewhere (Stoyanov, Machamer, et al. 2012, 2013; Stoyanov, Stanghellini, et al. 2013), clinical and neurobiological measures are considered valid for different reasons inside their own domains. All disciplines concerned with mental health establish internal or intra-correlative validity, i.e., psychological measures are validated with other psychological measures, and neurobiological measures are validated with other neurobiological tests. What is still missing is the inter-correlative or inter-disciplinary validity, which entails consistent inter-domain translation. Since the issue of translation is involved, we prefer the term “trans-disciplinary.”
As a potential source of external validity for the scales of clinical psychology, neuroscience can contribute information from two major biological databases: (epi-)genetic risk factors and neuroimaging abnormalities. Unfortunately, most of the efforts to discover behavioral-genetic and epigenetic biomarkers in psychiatry are too inconsistent and unstable to underpin any translational validity (Yosifova et al. 2009; Betcheva et al. 2013).
One critical consideration against the implications of genetic markers in psychiatric diagnosis is their “state independence.” The latter has been incorporated into the “endophenotype strategy” and is defined as independence of the biomarkers from the current mental state, which means that endophenotypes are lifetime stable, present in both clinical episodes and remissions (Hasler et al. 2006). Whilst state independence might be a useful assumption for some retest stable mental phenomena such as the traits in the psycho-biological model of personality (Cloninger et al. 1993), it is less relevant for clinical states like bipolar depression which are in fact determined by instability of emotional regulation. In those cases, “state independence” would be a shortcoming rather than an advantage. On the contrary, we argue that “state dependence” should be rendered as an alternative and sounder approach to translation of the neurobiological mechanisms of mental disorders into clinical reality. State dependence means that certain correlations are directly relevant and specific to the current mental state. This is why the clinical and biological measures should be performed simultaneously in our paradigm.
Structural and functional neuroimaging as potential external validators are also exposed to critical queries about their validity and clinical utility (Fusar-Poli and Broome 2006; Borgwardt et al. 2012; Korf et al. 2011; Stoyanov, Stanghellini, et al. 2012). On one hand, functional MRI is considered in our perspective as a translation validation tool because of the following reasons:
1.In comparison with other imaging methods, fMRI can capture very close to real-time brain response to psychological stimuli of diagnostic significance.
2.It has enhanced spatial resolution as compared to other neuroimaging techniques (e.g., Positron Emission Tomography—PET) and can penetrate the substrate of mental function—the oxygen metabolism of cortical and sub-cortical neural substrates. Cortical regions are usually easier to reach with neuroimaging; however, there are advanced methods in structural/functional MRI (PET and Magnetic Resonance Spectroscopy) that can specifically address sub-cortical brain regions.
3.Modern upgrades in fMRI facilities allow multimodal imaging which can integrate further modalities like receptor expression and quantitative electro- encephalography (qEEG); the latter can substantially contribute to time resolution.
On the other hand, we have identified so far several major shortcomings, which seem to undermine the data translation from neuroimaging to psychiatry:
1.The psychological stimuli (such as emotional pictures processing) administered during functional brain scanning are specifically designed to study general psychological processes and not day-to-day diagnostic constructs in clinical psychiatry (such as International Affective Picture System—IAPS).
2.Clinical assessment inventories, both observational and self-assessment (such as MADRS, BDI), are administered outside the brain scanner and are thus discrepant from the imaging findings. This argument concerns bipolar depression first, since one of the cardinal features of bipolar disorder is the instability of the circadian rhythm of emotions, which may vary significantly, especially in depression.
3.Statistical correlations between neuroimaging measures and clinical assessment are performed post-hoc, are very often unstable (not replicated) in larger cohorts, and cannot be regarded as cross-validation operations. In this way no validity connections are traced across the explanatory and ideographic knowledge in psychiatry.
Those shortcomings, however, can be managed with some modifications in the experimental paradigm, in particular with simultaneous—concordant in real time—and full-length administration of clinical measures during a functional neuroimaging session. One way to do this would be to project one by one the items from the selected clinical evaluation scale (e.g., Depression-Scale by Von Zerssen 1986 or Beck Depression Inventory 1988, etc.) in real time on a screen above the patient inside the scanner during fMRI. The patient’s rating responses will be recorded by a button click of a four-button response panel that is installed in the fMRI system anyway. Should the two measures correspond, they are regarded as convergent translational validity operations.
The explicit objective of such protocol is aiming at cross-validation as a complementary approach for the establishment of bi-conditional rules for translation of the data of neuroscience to clinical psychology and psychopathology. Following the state dependence argument as exposed earlier, we set aside re-test stable personality traits as measured by MMPI, EPQ (Eysenck Personality Questionnaire), or TCI (Temperament and Character Inventory). Those are complex intentional structures which are difficult to assess simultaneously with fMRI. Besides, much progress has been achieved in the paradigm of C. R. Cloninger with TCI without time synchronization of the measures (Gusnard et al. 2003). This was our reason to focus on frequently employed in diagnostic practice brief state measures instead, such as the Beck Depression Inventory and the Von Zerssen Depression Scale.
It is suggested bi-directional cross-validation in our model: using the functional neuroimaging measures as external validator of psychological clinical test scores and using the psychological test scores as a way to bring a more hermeneutic dimension into the procedures of validation in psychiatric neuroimaging.
The translation takes place on two levels: empirical and meta-empirical. First, the corresponding empirical measures are cross-validated (e.g., depression clinical rating score on BDI and fMRI blood oxygenation level dependent (BOLD) activation); then the entire constructs and relevant theoretical models.
Those rules or “manual for translation” may provide a synergistic explanation for the mechanism of production of the disorder and facilitate the inter-domain dialogue.
Much hope in psychiatry has been directed toward functional neuroimaging approaches, which promise to identify core neurobiological alterations. A non- invasive technique that can be used repeatedly in a clinical population, neuroimaging could in principle support diagnosis and effective interventions in psychiatry (Borgwardt and Fusar-Poli 2012; Borgwardt 2013).
Functional brain imaging methods such as functional magnetic resonance imaging (fMRI), which allow the in vivo investigation of human brain function, have been increasingly employed to examine the neurophysiological substrate of cognitive processes and psychopathological features. As the signals of the human brain functions are universal, fMRI studies that explore the neural substrates of psychopathology theoretically no longer rely on subjective measures, resulting in numerous publications of fMRI studies employing task- and non-task-related paradigms.
Methodological factors may account for the considerable heterogeneity in findings across fMRI studies. These factors include differences in relevant acquisition design, lack of statistical power due to small sample sizes, different methods of image analysis (i.e., parametric versus non-parametric), differences in the demographic and sociodemographic group characteristics, and confounding effects of medication or illness chronicity. Analysis of the consistency and comparability of the results obtained using different fMRI acquisition and analysis methods on the same set of neuroimaging data is a crucial prerequisite for accurate localization of various brain functions. To reliably apply fMRI in clinical settings, stable and consistent results, irrespective of the particular image acquisition and analysis methods used, are needed. In addition, a cognitive frame shift is required from the empirical level of investigation to trans-disciplinary validation of the clinical assessment tools and imaging data.
For a program of translational validity to succeed, it is crucial that the results of psychiatric neuroimaging can become more reliable. In what follows we offer some practical guidelines to conduct or evaluate functional neuroimaging studies in a program of translational validity (Borgwardt et al. 2012):
♦ Implementation of an increased number of ways of pre-processing the data Regions of Interest (ROIs) studies (employing preselected masks or adopting Small Volume Corrections) should first report standard whole brain results and acknowledge if no significant clusters were detected at whole brain level before presenting the ROI findings;
♦ Both ROIs and whole brain studies should first report the results significant at p<0.05 corrected for multiple comparisons (i.e., FWE, FDR, Montecarlo) and then employ more liberal thresholds;
♦ When several ROIs are used, correction for multiple comparisons should be based on a mask which includes all of them rather than considering each ROI separately;
♦ Authors should be encouraged to blind the statistical analyses of the imaging datasets to avoid ROIs analyses being built post-hoc on the basis of the results;
♦ All studies should report a statistical analysis modeling an agreed set of possible confounding variables; these could include, for instance, gender, age, and handedness. In addition, studies would have the option of reporting further statistical analyses modeling additional study-specific confounding variables;
♦ All studies should acknowledge the number of analyses or brain correlations performed, giving a clear rationale for each, to avoid conducting exploratory analyses and reporting the most significant result;
♦ The potential overlapping of the patient and control group with previously published studies should be clearly acknowledged, and the spatial coordinates always reported, to assist future voxel-based meta-analyses in the field;
♦ Peer reviews should be as strict when assessing the methods of a study reporting abnormalities in expected brain regions, as when assessing the methods of a study not finding any expectable finding;
♦ Acceptance or rejection of a manuscript should not depend on whether abnormalities are detected or not, or on the specific brain regions found to be abnormal.
In summary, neuroimaging methods may help to understand the pathogenesis of brain changes to clarify the onset and dynamic neurobiological processes underlying psychiatric disorders. However, for neuroimaging to be a clinically useful and valid tool, a framework linking basic, clinical research and target-specific treatments for people with psychiatric disorders should be developed. Translational validation is one vehicle to enhance this link (Stoyanov 2009, 2011).
To support the feasibility of our theoretical model of translational validation, we will review the empirical findings in the paradigm of “high risk” for psychosis (Koutsouleris et al. 2012). Early clinical intervention in schizophrenia has become a major objective of mental health services, and the finding that structural and functional alterations in the cingulate cortex during a first episode of psychosis are related to outcomes is of great interest (Bora et al. 2011). However, cingulate function and structure has also been reported to be especially sensitive to remedial antipsychotic treatment in psychosis (Lahti et al. 2009; Stip et al. 2009) and there is evidence indicating that a few weeks of antipsychotic treatment modulate the functional response in this region (Lahti et al. 2004; Snitz et al. 2005).
Previous reviews and meta-analyses have shown that significant brain changes driven by antipsychotic exposure can play a prominent confounding role in psychiatric imaging, thus preventing its translational clinical application (Smieskova et al. 2010; Fusar-Poli et al. 2013, published online). One possible approach to circumvent this problem is to selectively analyze drug-naïve first-episode subjects. For example, in a recent meta-analysis of untreated first-episode subjects, structural alterations in the cingulate cortex appear to be present before the initiation of antipsychotic treatment (Fusar-Poli, Radua, et al. 2011).
An alternative option would be to endorse “close in” clinical high risk (HR) approaches to identify a group of individuals with higher transition rates to psychosis (18 percent after six months of follow-up, 22 percent after one year, 29 percent after two years, and 36 percent after three years [Fusar-Poli, Bonoldi, et al. 2011]) than those observed in the general population. This clinical strategy aims at identifying neural changes occurring prior to the onset of psychosis and may improve translational ability of neuroimaging to predict schizophrenia outcomes. The presence of individuals who are high risk but not psychotic is consistent with evidence that schizophrenia results from the interaction of environmental with both genetic and neurodevelopmental factors, with the latter associated with clinical, neurobiological, and neuropsychological features before the onset of psychosis.
In recent years, a broad range of functional imaging methods have rapidly developed as powerful tools to explore the neurophysiological basis of the HR (Fusar-Poli, Borgwardt, et al. 2011; Fusar-Poli et al. 2007). Overall, these studies have shown that several abnormalities in brain function and neurophysiology that are fundamental to schizophrenia are also present in people at HR of psychosis, and may therefore represent vulnerability markers (Fusar-Poli et al. 2007).
Meta-analyses of whole brain structural studies comparing HR subjects with controls have confirmed reduced gray matter volume in the HR as compared to controls in the cingulate cortices as well as in temporal, prefrontal, parahippocampal/hippocampal regions (Fusar-Poli, Borgwardt, et al. 2011; Fusar-Poli et al. 2013). Volumetric reductions in cingulate and temporal, insular, prefrontal cortex, and in cerebellum have been also associated with the development of psychosis over follow-up (Smieskova et al. 2010).
The largest study published to date showed that the non-converting HR group demonstrated significant improvement in attenuated positive symptoms, negative symptoms, and social and role functioning, with more than 50 percent of this non-converting sample no longer presenting with any HR symptoms (Addington et al. 2011). However, this group remained on average at a lower level of functioning than non-psychiatric comparison subjects, suggesting that initial HR categorization is associated with persistent disability for a significant proportion for at least two years (Addington et al. 2011). It would be very useful to address functional changes associated with remission status within the HR cohort to identify protective neurobiological markers of later development of illness. Additionally, there is evidence from functional imaging and neurochemical HR studies that the extent of abnormality at baseline is predictive of subsequent conversion to psychosis (Smieskova et al. 2010).
These neurofunctional abnormalities of the At-Risk-Mental-State (ARMS) were not only related to different duration of ARMS, but also to gray matter reductions (GMV) (Smieskova et al. 2011) and the GMV itself was positively correlated with clinical outcomes as global functioning, negative symptomatology, and hallucinations (Smieskova et al. 2011). In particular, MRS studies have revealed reduced neuronal density and increased membrane turnover in cingulate as well as in frontal and insular lobe in HR subjects who later developed psychosis. Overall, the burden of functional imaging research into the HR state for psychosis has progressed exponentially, sustaining preventive interventions in clinical psychiatry (Ruhrmann et al. 2010).
However, despite these promises, validity of HR criteria is still highly discussed and the problem of the high number of false positives undermines the benefits of preventive interventions. In order to transcend over mere reliability and convert into valid inter-disciplinary paradigm, AMRS needs to employ the rationale of the translational cross-validation of the methods for clinical assessment and functional neuroimaging. In particular, the temporal gap between acquisition of clinical and functional brain measures may still present a limitation for the translation of the research findings (Korf and Gramsbergen 2007; Stoyanov et al. 2013; Stoyanov et al. 2012). Therefore, the management of the temporal discordance by simultaneous application of the two methods in our paradigm may help to enhance the research designs of clinical neuropsychiatry (Stoyanov et al. in press).
There is thus an urgent need for psychiatric imaging to further develop linking neurobiological markers with longitudinal outcomes, including transition, remission, and response to preventative interventions.
The currently employed instrumental approaches are insufficient in terms of their validity and ability to integrate the domains of psychopathology and neuroscience in order to provide sound explanatory predictions of the mental phenomena. In this chapter we have explored the complex interdisciplinary structure of psychiatry as an amalgam of multifaceted sources of inquiry. The main sources, including psychopathological clinical assessment and neurobiological studies, have defined validity and validation procedures of their own and no particular approach to cross-disciplinary validation to foster the translation of data across the disciplines constituting psychiatry.
We proposed rules for making translations between clinical and neuroimaging data as one critical step forward to the introduction of the notion of translational validity.
The procedure of translational cross-validation entails simultaneous measurement of the brain activation measured by fMRI and self-report responses to psychological tests. The correlations between the two measures remain free of any ontological speculation about the mind–brain causation and are regarded as ispo facto real-time correspondence.
Further, we present substantial empirical and meta-empirical data from current neuroimaging investigations of at-risk-mental-state (AMRS). On one hand, this analysis provides critical insights into the limitations of functional brain imaging in psychiatry. On the other hand, the empirical findings of correlations between brain processes and clinical assessment support the thesis that if correlations are discovered with inert psychological stimuli (of no immediate diagnostic relevance), then they should exist between functional neuroimaging measures and clinical rating scales of diagnostic significance. The real-time diagnostic testing combined with Blood-Oxygenation-Level-Dependent (BOLD) fMRI may provide cross-disciplinary connections of translation and therefore back up sounder instrumental validation across mental health disciplines.
The critical analysis of the problem of validation as emerging in the interdisciplinary situation across psychiatry and neuroscience, as well as the current empirical results in other studies attempting to relate neurobiological data to clinical reality, do ascertain the feasibility of the model of translational validity.
1.This short introductory description is in many ways an oversimplification. The picture is more complex, and the question whether or not scientific statements are truth-evaluable will also depend on whether one embraces a correspondence theory of truth.
2.Of course, what we refer to here as a disappointment with neuroscientific progress is not a view that everybody shares. Indeed, it seems to be the case that the main tenor in current research is still the idea that mental disorders are to be understood and treated as brain disorders (Insel and Quirion 2005).
Addington, J., Cornblatt, B. A., Cadenhead, K. S., et al. (2011). At clinical high risk for psychosis: outcome for nonconverters. American Journal of Psychiatry, 168, 800–5.
Andreasen, N. C. (2006). DSM and the death of phenomenology in America: an example of unintended consequences. Schizophrenia Bulletin, 33(1), 108–12.
Beck, A. T., Steer, R. A., and Carbin, M. G. (1988). Psychometric properties of the Beck Depression Inventory: Twenty-five years of evaluation. Clinical Psychology Review, 8(1), 77–100.
Berrios, G. E. (2006). “Mind in general” by Sir Alexander Crichton. History of Psychiatry, 17(4), 469–86.
Betcheva, E. A. G., Yosifova, T., Mushiroda, M., et al. (2013). Whole genome association study in Bulgarian population reveals HHAT as schizophrenia susceptibility gene. Psychiatric Genetics, 23 (1), 11–19.
Boorse C. (1975). On the Distinction between Disease and Illness. Philosophy and Public Affairs, 5(1), 49–68.
Bora, E., Fornito, A., Radua, J., et al. (2011). Neuroanatomical abnormalities in schizophrenia: a multimodal voxel-wise meta-analysis and meta-regression analysis. Schizophrenia Research, 127, 46–57.
Borgwardt, S. (2013). 2989–Structural and functional neuroimaging in early schizophrenia: translating research evidence into clinical utility. European Psychiatry, 28, 1.
Borgwardt, S., and Fusar-Poli, P. (2012). Third-generation neuroimaging in early schizophrenia: translating research evidence into clinical utility. The British Journal of Psychiatry, 4, 270–2.
Borgwardt, S., Koutsouleris, N., Aston, J., et al. (2012). Distinguishing Prodromal from First-Episode Psychosis using Neuroanatomical Single-Subject Pattern Recognition. Schizophrenia Bulletin
Borgwardt, S., Radua, J., Mechelli, A., and Fusar-Poli, P. (2012). Why are psychiatric imaging methods clinically unreliable? Conclusions and practical guidelines for authors, editors and reviewers. Behavioral and Brain Functions, 8(1), 46.
Broome, M. R. (2008). Philosophy as the science of value: neo-Kantianism as a guide to psychiatric interviewing. Philosophy, Psychiatry & Psychology, 15(2), 107–16.
Cloninger, C. R., Svrakic, D. M., and Przybeck, T. R. (1993). A psychobiological model of temperament and character. Archives of General Psychiatry, 50(12), 975.
First, M. B., Spitzer, R. L., Gibbon, M., et al. (2012). Structured Clinical Interview for DSM-IVо Axis I Disorders (SCID-I), Clinician Version, Administration Booklet. Washington, DC: American Psychiatric Publishing.
Frances, A. (2010). Allen Frances Responds. Bulletin of the Association for the Advancement of Philosophy and Psychiatry, 17(1), 21–5.
Fusar-Poli, P., Bhattacharyya, S., Allen, P., et al. (2010). Effect of image analysis software on neurofunctional activation during processing of emotional human faces. Journal of Clinical Neuroscience, 17(3), 311–14.
Fusar-Poli, P., Bonoldi, I., Yung, A. R., et al. (2012). Predicting psychosis: meta-analysis of transition outcomes in individuals at high clinical risk. Archives of General Psychiatry, 69(3), 220–9.
Fusar-Poli, P., Borgwardt, S., Crescini, A., et al. (2011). Neuroanatomy of vulnerability to psychosis: a voxel-based meta-analysis. Neuroscience and Biobehavioral Reviews, 35(5), 1175–85.
Fusar-Poli, P., Borgwardt, S., and Valmaggia, L. (2008). Heterogeneity in the assessment of the at-risk mental state for psychosis. Psychiatric Services, 59(7), 813.
Fusar-Poli, P., and Broome, M. R. (2006). Conceptual issues in psychiatric neuroimaging. Current Opinion in Psychiatry, 19(6), 608–12.
Fusar-Poli, P., Broome, M., Barale, F., et al. (2009). Why is psychiatric imaging clinically unreliable? Epistemological perspectives in clinical neuroscience. Psychotherapy and Psychosomatics, 78(5), 320–1.
Fusar-Poli, P., Perez, J., Broome, M., et al. (2007). Neurofunctional correlates of vulnerability to psychosis: a systematic review and meta-analysis. Neuroscience and Biobehavioral Reviews, 31(4), 465–84.
Fusar-Poli, P., Radua, J., McGuire, P., et al. (2012). Neuroanatomical maps of psychosis onset: voxel-wise meta-analysis of antipsychotic-naïve VBM studies. Schizophrenia Bulletin, 38(6), 1297–307.
Fusar-Poli, P., Smieskova, R., Kempton, M. J., et al. (2013). Progressive brain changes in schizophrenia related to antipsychotic treatment? A meta-analysis of longitudinal MRI studies. Neuroscience and Biobehavioral Reviews.
Gadamer, H.-G. (1996). The Enigma of Health: The Art of Healing in a Scientific Age. Stanford, CA: Stanford University Press.
Gusnard, D. A., Ollinger, J. M., Shulman, G. L., et al. (2003). Persistence and brain circuitry. Proceedings of the National Academy of Sciences, 100(6), 3479–84.
Hasler, G., Drevets, W. C., Gould, T. D., et al. (2006). Toward constructing an endophenotype strategy for bipolar disorders. Biological Psychiatry, 60(2), 93–105.
Insel, T. R., and Quirion, R. (2005). Psychiatry as a clinical neuroscience discipline. JAMA: the journal of the American Medical Association, 294(17), 2221–4.
Ioannidis J. P. (2011). Excess significance bias in the literature on brain volume abnormalities. Archives of General Psychiatry, 68(8), 773–80.
Kay, S. R., Opler, L. A., and Lindenmayer, J. P. (1989). The Positive and Negative Syndrome Scale (PANSS): rationale and standardisation. The British Journal of Psychiatry.
Korf, J., and Gramsbergen, J. B. (2007). Timing of potential and metabolic brain energy. Journal of Neurochemistry, 103(5), 1697–708.
Koutsouleris, N., Borgwardt, S., Meisenzahl, E. M., et al. (2012). Disease prediction in the at-risk mental state for psychosis using neuroanatomical biomarkers: results from the FePsy study. Schizophrenia Bulletin, 38(6), 1234–46.
Kupfer, D. J., First, M. B., and Regier, D. A. (eds) (2002). A Research Agenda for DSM–V. Washington, DC: American Psychiatric Association.
Lahti, A. C., Weiler, M. A., Holcomb, H. H., et al. (2009). Modulation of limbic circuitry predicts treatment response to antipsychotic medication: a functional imaging study in schizophrenia. Neuropsychopharmacology, 34, 2675–90.
Miller, G. (2010). Mistreating psychology in the decades of the brain. Perspectives on Psychological Science, 5(6), 716–43.
Montgomery, S. A., and Asberg, M. A. R. I. E. (1979). A new depression scale designed to be sensitive to change. The British Journal of Psychiatry, 134(4), 382–9.
Radua, J., Mataix-Cols, D., Phillips, M. L., et al. (2011). A new meta-analytic method for neuroimaging studies that combines reported peak coordinates and statistical parametric maps. European Psychiatry
Reynolds, C. F., et al. (2009). The future of psychiatry as clinical neuroscience. Academic Medicine, 84(4), 446–50.
Robins, E., and Guze, S. B. (1970). Establishment of diagnostic validity in psychiatric illness: its application to schizophrenia. American Journal of Psychiatry, 126(7), 983–7.
Ruhrmann, S., Schultze-Lutter, F., Bechdolf, A., et al. (2010). Intervention in at-risk states for developing psychosis. European Archives of Psychiatry and Clinical Neuroscience, 260, Suppl 2, S90–4.
Rusconi, E., and Mitchener-Nissen, T. (2013). Prospects of Functional Magnetic Resonance Imaging as Lie Detector. Frontiers in Human Neuroscience, 7, 00594.
Simon, A. E., Velthorst, E., Nieman, D. H., et al. (2011). Ultra high-risk state for psychosis and non-transition: a systematic review. Schizophrenia Research, 132, 8–17.
Smieskova, R., Allen, P., Simon, A., et al. (2011). Different duration of at-risk mental state associated with neurofunctional abnormalities. A multimodal imaging study. Human Brain Mapping, 33(10), 2281–94.
Smieskova, R., Fusar-Poli, P., Allen, P., et al. (2010). Neuroimaging predictors of transition to psychosis—a systematic review and meta-analysis. Neuroscience and Biobehavioral Reviews, 34(8), 1207–22.
Snitz, B. E., MacDonald, A., Cohen, J. D., et al. (2005). Lateral and medial hypofrontality in first-episode schizophrenia: functional activity in a medication-naive state and effects of short-term atypical antipsychotic treatment. American Journal of Psychiatry, 162(12), 2322–9.
Stip, E., Mancini-Marie, A., Letourneau, G., et al. (2009). Increased grey matter densities in schizophrenia patients with negative symptoms after treatment with quetiapine: a voxel-based morphometry study. International Clinical Psychopharmacology, 24, 34–41.
Stojanov, D., Korf, J., de Jonge, P., et al. (2011). The possibility of evidence-based psychiatry: depression as a case. Clinical Epigenetics, 2(1), 7–15.
Stoyanov, D. (2009). The cross-validation in the dialogue of mental and neuroscience. Dialogues in Philosophy, Mental and Neuro Sciences, 2(1), 24–8.
Stoyanov, D. (2011). A linkage of mind and brain: towards translational validity between neurobiology and psychiatry. Biomedical Reviews, 22, 65–76.
Stoyanov, D., Machamer, P. K., and Schaffner, K. F. (2012). Rendering clinical psychology an evidence-based scientific discipline: a case study. Journal of Evaluation in Clinical Practice, 18, 149–54.
Stoyanov, D., Machamer P., and Schaffner. K. F. (2013). In quest for scientific psychiatry: towards bridging the explanatory gap. Philosophy, Psychiatry & Psychology, 20(3), 261–73.
Stoyanov, D., Stanghellini, G., and Broome, M. (2012). Conceptual issues in psychiatric neuroimaging. Current topics in medicinal chemistry, 12(21), 2348–56.
Stoyanov, D. S., Stieglitz, R. D., and Borgwardt, S. J. (in press). The translational validation as novel approach to integration of neuroscience and psychiatry. In A. L. Mishara, P. Corlett, P. Fletcher, A. Kranjec, and M. A. Schwartz (eds), Phenomenological Neuropsychiatry, How Patient Experience Bridges Clinic with Clinical Neuroscience. New York: Springer.
Von Zerssen, D. (1986). Clinical self-rating scales (CSRS) of the Munich psychiatric information system (PSYCHIS München). In Assessment of depression (pp. 270–303). Heidelberg: Springer Berlin.
Weinberger, D. R., and McClure, R. K. (date?). Neurotoxicity, neuroplasticity, and magnetic resonance imaging morphometry: what is happening in the schizophrenic brain? Archives of General Psychiatry, 59(6), 553–8.
Yosifova, A., Mushiroda, T., Stoianov, D., et al. (2009). Case-control association study of 65 candidate genes revealed a possible association of a SNP of HTR5A to be a factor susceptible to bipolar disease in Bulgarian population. Journal of Affective Disorders, 117(1), 87–97.