Neuroethics to date has tended to focus on social and ethical implications of developments in brain science, especially in functional neuroimaging. Within clinical neuroethics, the emphasis has been on ethical issues in clinical neuroscience practice, including informed consent to neuroimaging; the development of ethical research protocols for functional magnetic resonance imaging especially, and especially in children; and the ethical clinical management of incidental findings.1
See, for example, Illes J. Medical imaging: A hub for the new field of neuroethics. Academic Radiology 2004;11:721–3; Illes J, Kirschen MP, Karetsky K, Kelly M, Saha A, Desmond JE, et al. Discovery and disclosure of incidental findings in neuroimaging research. Journal of Magnetic Resonance Imaging 2004;20:743–7.
See, for example, Farah MJ, Illes J, Cook-Deegan R, Gardner H, Kandel E, King P, et al. Neurocognitive enhancement: What can we do and what should we do? Nature Reviews Neuroscience 2004;5:421–5; Farah MJ, Wolpe PR. Monitoring and manipulating brain function: New neuroscience technologies and their ethical implications. Hastings Center Report 2004;34(3):35–45; Racine E, Bar-Ilan O, Illes J. fMRI in the public eye. Nature Reviews Neuroscience 2005;6:159–64.
Some psychiatrists have claimed that the application of technologies from such domains as neuroscience, genetics, and genomics promise revolutionary advances in diagnosis, classification, and patient care. Such promises include the identification of “genes for” all kinds of psychiatric disease from bipolar disorder to schizophrenia and the construction of a new taxonomy of disease based on the underlying etiology of illness. Whether these putative advances are worth realizing remains to be determined on a case-by-case basis, given a certain skepticism about the likelihood that they in fact can be realized.3
Kendler KS. “A gene for …”: The nature of gene action in psychiatric disorders. American Journal of Psychiatry 2005;162:1243–52.
Robert JS. Schizophrenia epigenesis? Theoretical Medicine and Bioethics 2000;21:191–215.
See, for example, Pies R. Why psychiatry and neurology cannot simply merge. Journal of Neuropsychiatry and Clinical Neurosciences 2005;17:304–9; Kanner AM. When did neurologists and psychiatrists stop talking to each other? Epilepsy and Behavior 2003;4:597–601.
My guiding concern is for appropriate patient care (an ethical consideration), and my discussion is premised on a systems approach to understanding development, and so to understanding the role of genes in development (a conceptual concern in the philosophy of biology).6
Robert JS. Embryology, Epigenesis, and Evolution. New York: Cambridge University Press; 2004.
In what follows, I contend that whether a new taxonomy of psychiatric disease is adequate to patient care will depend largely on the approach taken in designing that taxonomy. Whether the approach is purely reductionistic (based solely or largely on the underlying molecular etiology or molecular characterization,) or rather more integrative (based on not only molecular data but also intermediate and macro-level information) will determine its value for patient care. Given alternative visions of good psychiatry (as of good medicine more generally), and given that these are open to evaluation on multiple fronts, I will stake a claim for one alternative vision over the considerably more popular view that a reductionistic taxonomy is good, worthwhile, and imminent in psychiatry as elsewhere in medicine.
A Molecular Taxonomy of Disease?
The National Human Genome Research Institute's “vision for the future of genomics research” includes the quest for “a new molecular taxonomy of illness, which would replace our present, largely empirical, classification schemes and advance both disease prevention and treatment” (p. 841).7
Collins FS, Green ED, Guttmacher AE, Guyer MS, U.S. National Human Genome Research Institute. A vision for the future of genomics research. Nature 2003;422:835–47.
Zerhouni EA. US biomedical research: Basic, translational, and clinical sciences. JAMA 2005;294: 1352–8.
See, for example, Tsuang MT. Genetics, epidemiology, and the search for causes of schizophrenia. American Journal of Psychiatry 1994;151:3–6; and Bearden CE, Reus VI, Freimer NB. Why genetic investigation of psychiatric disorders is so difficult. Current Opinion in Genetics and Development 2004;14:280–6.
Let us assume for the sake of argument that there are very likely a number of genes involved in the etiology of schizophrenia. Now, what role should this molecular information play in psychiatric disease classification? The leading approach, a reductionistic one, is evidently motivated by a belief in the centrality and primacy of genetic explanations of higher level phenomena, and so instances of this approach are characterized by a certain optimism about the prospects of molecular genetics in psychiatry. Indeed, there almost seems to be an imperative to recreate psychiatric classifications on the basis of genomics data.
A recent review suggests that genomics can help to eliminate (some) social and political factors from the nosology, thereby generating a more objective, scientific nosology.10
Krishnan KR. Psychiatric disease in the genomic era: Rational approach. Molecular Psychiatry 2005;10:978–84. Whether it is possible to eliminate these putatively extraneous factors is entirely debatable. But whether it is desirable to do so is equally open to dispute. Increasingly, the literature seems to suggest that it is only by embracing the social and political (and cultural and ecological) embeddedness of medical conditions that clinicians will be able to grasp the complexity of illness and to offer suitable treatment regimes to their patients. Of course, the latter approach is not developed in ignorance of science, but neither is it based in an overly reductionistic approach to science: Molecular markers, brain images, and biochemical assays are indeed meaningful, but not on their own; they acquire meaning only in an appropriate bio-psycho-social context.
Harris HW, Schaffner KF. Molecular genetics, reductionism, and disease concepts in psychiatry. Journal of Medicine and Philosophy 1992;17:127–53. It is noteworthy that Schaffner is now more circumspect in this regard; see, for example, Schaffner KF. Genes, behavior, and developmental emergentism: One process, indivisible? Philosophy of Science 1998;65:209–52; Schaffner KF. Complexity and research strategies in behavioral genetics. In: Carson RA, Rothstein MA, eds. Behavioral Genetics: The Clash of Culture and Biology. Baltimore: Johns Hopkins University Press; 1999:61–88.
Tsuang MT, Faraone SV, Lyons MJ. Identification of the phenotype in psychiatric genetics. European Archives of Psychiatry and Clinical Neuroscience 1993;243:131–42. Note that this usage of “empirical” implies the philosophical sense of validated through systematic investigation and observation, rather than the pejorative sense in which Collins and colleagues use it in the passage cited above; see note 7, Collins et al. 2003.
Tsuang's work is especially instructive. He endorses the creation of a psychiatric genetic nosology that “seeks to classify patients into categories that correspond to distinct genetic entities” (p. 4).13
See note 9, Tsuang 1994.
See note 12, Tsuang et al. 1993.
See, for example, note 3, Kendler 2005 and note 4, Robert 2000.
Kendler KS. Toward a scientific psychiatric nosology. Archives of General Psychiatry 1990;47:969–73.
Philosophers of science have long contended that no scientific theories are perfectly value free; this is certainly the case with biomedical theories, insofar as diagnoses of illness inherently make value judgments about the state of the patient's body as compared to a normal body. Psychiatry could perhaps be rendered less (ostensibly) subjective by appeal to biological markers—that is the point of using, for example, neurological testing as a means of determining from what sort of disorder a particular patient suffers. But insofar as biological markers are themselves normative, the hope of value freedom in psychiatric diagnosis is a false hope, in the same way that such a hope is false also in the case of other biomedical diagnoses—none of which is to say that biomedical diagnoses are subjective or otherwise unscientific.
Indeed, the move toward a psychiatric genetic nosology does not do away with preconceptions. Insofar as knowledge of the normal follows upon knowledge of the pathological, any values (e.g., about normalcy, health, and illness) that are present at the level of phenotypes are imported to the level of the genome. Because those same values abound in the genetic realm, the process of identifying the genotype of a psychiatric condition simply is no less value laden than the process of identifying the phenotype of a psychiatric condition. (In fact, the process of identifying the genotype may be even more value laden than the process of identifying the phenotype, in assuming that the condition is a genetic disorder rather than something else.) Genetic normality is no more value neutral than phenotypic normality. To “go genomic” is just to push the matter down a level. Thus, whether a scientific psychiatric nosology grounded in molecular data is more credible than earlier nosologies remains to be demonstrated.
There is little doubt that some molecular taxonomy of disease is eminently achievable. Insofar as genes are involved in pathways of disease (psychiatric or otherwise), they will eventually be identified. This is a function of the immense efforts expended searching for such genes, as facilitated by whole genome scans and other technical innovations. But whether these immense efforts are worthwhile remains unclear. It is plainly evident that there is no necessary link between knowing the genes associated with a disease and developing therapies for that disease; there are plenty of instances of therapies in the absence of detailed knowledge about etiology and of detailed knowledge about etiology in the absence of therapies. Granted, a molecular characterization of a disease may be, even if it need not be, enormously therapeutically useful. For instance, in the presence of other data, genetic data may support novel treatment strategies for conditions that are apparently distinct, such as autism and catatonia,18
Dhossche DM. Autism as early expression of catatonia. Medical Science Monitor 2004;10:RA31–9.
Fiona Miller and colleagues have recently studied in detail how new genetic knowledge may contribute to lumping, splitting, or otherwise refining diagnostic categories, focusing on hemophilia, Rett syndrome, and cystic fibrosis.19
Miller FA, Begbie ME, Giacomini M, Ahern C, Harvey EA. Redefining disease? The nosologic implications of molecular genetic knowledge. Perspectives in Biology and Medicine 2006;49:99–114; see also Miller FA, Ahern C, Ogilvie J, Giacomini M, Schwartz L. Ruling in and ruling out: Implications of molecular genetic diagnoses for disease classification. Social Science and Medicine 2005;61:2536–45.
It is not clear that a single taxonomy, or a single approach to generating taxonomies, will ever be completely adequate. So the promise of a revolutionary new molecular taxonomy of disease is probably too strong a promise, potentially collapsing or otherwise eliding the differences between these various senses of adequacy. Moreover, that we have genomics data at hand says nothing about whether those data should be used to revolutionize or even to refine diagnostic categories—such changes are optional.
Because it is not in dispute that genetics should play some role in psychiatric classification, clarifying this more limited role is important.20
For further elaboration, see Robert JS, Plantikow T. Genetics, neuroscience and psychiatric classification. Psychopathology 2005;38:215–8.
See note 19, Miller et al. 2006.
Integrating Neuroscience into Nosology
There is more to biology than genetics, and it should come as no surprise that the human brain might prove inimical to genetic reduction, even though genetic information is critical to understanding the brain. Human brains are among the most complex of all biological systems; the heterogeneity, networked connectivity, and complexity of the brain (not to mention its relative inaccessibility) have rendered it an exceedingly difficult organ to study.22
Lo D. Challenges to neuroscience in a post-genome world. Nature Neuroscience 2001;4:1153–4.
For present purposes, it is important to underscore several features of the relationship between genetics and neuroscience both for understanding psychiatric and neurological disease, but also for understanding how to approach the process of classifying such diseases. Research on the genetics of behavior and of psychiatric disorders has proceeded through two distinct stages23
Robert JS. Genetics and behavior. In: Mitcham C, ed. Encyclopedia of Science, Technology, and Ethics. Farmington Hills, MI: MacMillan Reference USA; 2005:849–54.
See, for example, note 3, Kendler 2005 and note 4, Robert 2000.
Plomin R, Crabbe J. DNA. Psychological Bulletin 2000;126:806–28; Plomin R, McGuffin P. Psychopathology in the postgenomic era. Annual Review of Psychology 2003;54:205–28.
Boguski MS, Jones AR. Neurogenomics: At the intersection of neurobiology and genome sciences. Nature Neuroscience 2004;7:429–33.
See note 22, Lo 2001.
This sort of systems approach is only partly reductionistic—meaningful reductions are permitted, but where reduction limits rather than enhances understanding, integration is the norm. An integrative, systems approach to the brain depends jointly on gene maps and brain scans and a whole suite of observational and experimental techniques at all levels of analysis from the molecular to the behavioral and back again. Psychiatric illness is itself complex, requiring attention not just to genes and neurons but also what is clinically observable in the prime matériel of the psychiatrist: unhappy and often ill people who behave bizarrely. The person, her/his behavior, context, and life circumstances, must all be part of our analysis.
And yet the clinic has been removed from psychiatric nosology, and concurrently patients have been replaced by their blood, sweat, urine—and DNA.28
See, for example, Young A. The harmony of illusions: Inventing post-traumatic stress disorder. Princeton, N.J.: Princeton University Press; 1995; Wilson M. DSM-III and the transformation of American psychiatry: A history. American Journal of Psychiatry 1993;150:399–410.
Berrios GE. Towards a new descriptive psychopathology: A sine qua non for neurobiological research in psychiatry. Brain Research Bulletin 1999;50:457–8.
See note 12, Tsuang et al. 1993.
An integrative, systems approach to the brain recommends an alternative approach to psychiatric classification, one that reflects input from all relevant sources, incorporating multidirectional feedback between genes and larger phenotypic and environmental elements. This integrative approach to nosology31
See also Rudnick A. The molecular turn in psychiatry: A philosophical analysis. Journal of Medicine and Philosophy 2002;27:287–96.
To be sure, the point of the foregoing is not to deny the importance of biological aspects of psychiatric disease; it is instead to begin to understand such disease at an appropriate level of complexity. But clinical phenomenology is of absolute necessity. Whether we are justified in either lumping together discrete phenotypes on the basis of (presumably) shared genotype or splitting homogeneous phenotypes on the basis of discrete genotypes cannot be decided a priori. But neither can it be decided empirically on either a strictly genetic or a strictly clinical basis. Hence the need for a dialectical, integrative approach between good phenomenology, scientific but nonmolecular aspects of nosology (such as biometric and neurobiological approaches), and molecular genetics, in an effort to accurately describe the clinical, personal, and biological reality of psychiatric disease—and, eventually, to understand its etiology, its variability, and its tractability to treatment. Without sustained attention to clinical descriptions of psychological phenomena, and without adequate funding for this sort of research, Berrios predicts that “current investigative techniques” in both neurobiology and genetics—“however powerful,” and however well funded—“will draw the proverbial blank” (p. 458).32
See note 29, Berrios 1999.
Endophenotypes and an Integrative Approach to Taxonomy
As we recognize the limits of genetic reductionism in psychiatry and in psychiatric nosology, an increasingly critical concept is the notion of an endophenotype. Endophenotypes are “measurable components along the pathophysiological pathway between aetiology and psychopathology” (p. 204).33
Weiser M, van Os J, Davidson M. Time for a shift in focus in schizophrenia: From narrow phenotypes to broad endophenotypes. British Journal of Psychiatry 2005;187:203–5; see also Almasy L, Blangero J. 2001. Endophenotypes as quantitative risk factors for psychiatric disease: Rationale and study design. American Journal of Medical Genetics, Part B: Neuropsychiatric Genetics 2001;105:42–4.
See note 33, Weiser et al. 2005.
In general, in the context of behavioral development, the description of an endophenotype may require attention at least to clinical phenomenology, various biological and biochemical assays, neuroanatomy, and neurocognitive processing—the latter being typically accessed clinically and then mapped through neuroscientific techniques such as positron emission tomography scans and (functional) magnetic resonance imaging. Clinically relevant endophenotypes may prove useful in the establishment of robust psychiatric diagnostic categories that integrate data from various levels of analysis—clinical observations, patient narratives, molecular biomarkers, and neuroimages, for instance. These endophenotypes may also serve to facilitate integrative research of the sort envisioned above.
The notion of an endophenotype was initially devised in relation to schizophrenia in 1972,35
Gottesman II, Shields J. Schizophrenia and Genetics: A Twin Study Vantage Point. New York: Academic Press; 1972.
Andreasen NC, Nopoulos P, O'Leary DS, Miller DD, Wassink T, Flaum M. Defining the phenotype of schizophrenia: Cognitive dysmetria and its neural mechanisms. Biological Psychiatry 1999;46:908–20.
Andreasen NC, Shore D, Burke JD Jr, Grove WM, Lieberman JA, Oltmanns TF, et al. Clinical phenomenology. Schizophrenia Bulletin 1988;14:345–63.
See note 4, Robert 2000.
Recent work on endophenotypes may help to redress these concerns. Nancy Andreasen's group, focusing at intermediate levels between genes and observable symptoms, have proposed that the endophenotype of schizophrenia is a neurodevelopmental cognitive abnormality—a timing aberration in a fundamental feedback loop that is highly sensitive during development. If this endophenotype is in fact the underlying type uniting phenotypic manifestations of schizophrenia, it may well be able to refocus genetic research in productive ways. But even if this particular endophenotype turns out to be not well supported,39
For a skeptical view, see Kaprinis GS, Fountoulakis KN, Kaprinis SG. Arguments against the cognitive dysmetria hypothesis of schizophrenia. Perceptual and Motor Skills 2002;94:975–84.
This defect in timing affects a basic neurodevelopmental process upon which memory, attention, and language are based. The result of disrupting this process is the severe heterogeneous symptoms characteristic of schizophrenia. Andreasen et al. develop their model of cognitive development and function/malfunction on an analogy with the development of motor activity.40
See note 36, Andreasen et al. 1999; see also Crespo-Facorro B, Paradiso S, Andreasen NC, O'Leary DS, Watkins GL, Boles Ponto LL, Hichwa RD. Recalling word lists reveals ‘cognitive dysmetria’ in schizophrenia: A positron emission tomography study. American Journal of Psychiatry 1999;156:386–92; Andreasen NC, O'Leary DS, Cizadlo T, Arndt S, Rezai K, Boles Ponto LL, et al. Schizophrenia and cognitive dysmetria: A positron-emission tomography study of dysfunctional prefrontal-thalamic-cerebellar circuitry. Proceedings of the National Academy of Sciences, USA 1996;93:9985–90.
See note 36, Andreasen et al. 1999.
To explain this consequence, Andreasen et al. argue that
[an] individual with mistimed information transfer may incorrectly connect perceptions and associations and misinterpret both external and internal processes, leading in turn to delusions or hallucinations (e.g., a neutral perception will be associated with a frightening affective association, internal thoughts or vocalizations may be attributed to others). Defects in coordinating language production will lead to “thought disorder”. In addition, the flow of information through the system may become paralyzed, leading to “negative symptoms” such as alogia or affective blunting (p. 916).42
See note 36, Andreasen et al. 1999.
So, Andreasen and her colleagues speculate that the clinical variability of schizophrenia might be interpreted as a function of various perturbations in the CCTCC, whereas a disturbed CCTCC “represents a basic neural phenomenon common to patients suffering from schizophrenia” (p. 391).43
See note 40, Crespo-Facorro et al. 1999.
Setting aside whether Andreasen et al. are right about cognitive dysmetria as the schizophrenia endophenotype, two key observations are in order. First, notice that though the focus of inquiry is not the genetic level, it is inappropriate to thereby infer that the genetics of cognitive dysmetria are unimportant or irrelevant. Cognitive dysmetria is worth investigating on its own terms, from all relevant perspectives, using all appropriate methods, for the reinterpretation of the symptoms of schizophrenia, as sharing an underlying neuro-phenotypic substrate opens up the possibility of more phenotypically well-grounded genetic research. Indeed, numerous researchers have expressly argued that focusing on endophenotypes can facilitate genetic investigation.44
See, for example, note 9, Bearden et al. 2004; note 33, Almasy, Blangero 2001; Merikangas KR, Risch N. Will the genomics revolution revolutionize psychiatry? American Journal of Psychiatry 2003;160:625–35; Gottesman II, Gould TD. The endophenotype concept in psychiatry: Etymology and strategic intentions. American Journal of Psychiatry 2003;160:636–45.
Additionally, notice that endophenotypes such as this one—identified on the basis of the integration of multiple indicators accessed through clinical observation, theoretical neurobiology and neuropathology, and neuroimaging techniques—could help to delimit a robust disease category that may put to rest long-standing debates over the appropriate interpretation of the schizophrenia construct.
An integrative approach to disease classification, facilitated by but not reducible to molecular data, is a tractable research strategy for advancing taxonomy along multiple lines of adequacy. I reiterate that no single taxonomy is likely to be useful for all purposes, but the open-ended, integrative approach advocated here is more amenable to pragmatic negotiation than reductionistic, nonempirical alternatives.
Conclusion
What is most important about psychiatric nosology is whether it serves patients well and whether it is amenable to therapeutic ends for patients and their healthcare providers. There is no reason to believe that a reductionistic taxonomy, whether based on gene maps or brain scans, will achieve this end in all cases, or even very often. Just the opposite may prove true; hence my advocacy for an integrative approach based on the notion of the endophenotype, within which the boundaries of psychiatric classifications may be determined by appeal to relevant knowledge including but not limited to molecular knowledge. Though the adequacy of any classification system must be demonstrated and not assumed, the integrative approach to taxonomy takes seriously the clinical phenomenology of psychiatric conditions and thus is predisposed to clinical adequacy for patient care. Given everything we know about the brain and about the complexity of psychiatric classification, an a priori commitment to a new molecular taxonomy in psychiatry is unjustified.
As neuroethics develops, and as it attends both to medicine and also to the many social contexts into which brain science flows, such as education, law, and national security, psychiatry and psychiatric ethics will increasingly demand sustained attention. The relationships between neuroscience and psychiatry, brain and mind, are complex and worthy of both conceptual and ethical investigation. In this paper, I have raised some preliminary considerations about psychiatric classification. But psychiatric diagnosis is also of central clinical concern, raising important questions about the appropriate use of genetic and neuroscientific data in diagnosing patients. Determining the relationship between classification and diagnosis, and assessing whether and how molecular, neuroscientific, and neurological markers should be used in diagnosis, must await future analysis in neuroethics and the philosophy of psychiatry.
