In this essay, I approach the question about the BPSM from the perspective of a teacher of psychiatry, medical school dean responsible for (among other matters) student career advising, and clinician. In those capacities, my duties include fostering an understanding of psychiatric disorders among medical students and residents, instilling confidence in and respect for the discipline of psychiatry among students as well as nonpsychiatric colleagues, and explaining psychiatric diagnosis and treatment to patients and their families.
Presumably as a consequence of the complex and varied nature of our subject matter, psychiatry has had a more difficult time than the rest of medicine agreeing upon a conceptual framework. The closest the field has come to a consensus on this score is George Engel's biopsychosocial model (BPSM), whose explicit and implicit influences extend throughout psychiatric education, training, and practice.1 The question that needs to be examined is how those influences manifest and whether they aid or hinder understanding and teaching in psychiatry.
In this essay, I approach the question about the BPSM from the perspective of a teacher of psychiatry, medical school dean responsible for (among other matters) student career advising, and clinician. In those capacities, my duties include fostering an understanding of psychiatric disorders among medical students and residents, instilling confidence in and respect for the discipline of psychiatry among students as well as nonpsychiatric colleagues, and explaining psychiatric diagnosis and treatment to patients and their families. Why are those tasks so difficult? The obvious answer is that psychiatry is a complicated field. I also argue, however, that psychiatry's de facto philosophical structure-of which the BPSM is currently a prominent component--impedes rather than facilitates our ability to grasp, describe, and teach our complex topic and incorporate its expanding database in valid and useful ways.
The components of my argument are:
We need, therefore, to scrutinize this nexus of modes of thinking and speaking that reflect dualistic misconceptions and that, in turn, impair our capacity to conceptualize and communicate the complexities of our specialty.
Let us first examine the assertion that the theoretical structure of psychiatry is in disarray. One need not look hard to see evidence of its validity. On as basic a question as whether psychiatric disorders are a subset of medical disorders, or whether psychiatric explanations are a subset of medical explanations, psychiatrists' opinions appear incoherent. Thus, while statements such as "depression is a serious medical disorder" or "mental illnesses are brain diseases" are routinely mouthed as articles of faith, any psychiatrist who peruses his mail will see on a daily basis phrases such as "psychiatric care of the medically ill," "treatment of depression with coexisting medical illness," or "evaluation of medically unexplained symptoms." Any hope that the resurgent and anachronistically named subspecialty of "psychosomatic medicine" will bring coherence to this matter appears misplaced. With its dualistic talk of the "mind-body connection," one wonders whether psychosomatic medicine is simply Descartes without the pineal gland hypothesis.
In order to determine the effects that the BPSM (and related conceptual and linguistic conventions) has on our thinking in psychiatry, we must examine the arenas where it is typically applied and held to be valid and/or useful. I have identified 4 such domains:
For each of these domains of psychiatric thinking, I have tried to condense into a single statement the approach to it that is reflected in or dictated by the BPSM and/or the linguistic and conceptual conventions that are related to it. After presenting each statement, I analyze each statement's content, validity, and consequences.
Domain 1: Causes of illness
Statement: Diseases result from a summation or interaction of biological, psychological, and social influences. Since people commonly but erroneously conflate "biological" with "heritable" and "psychological" or "social" with "environmental," the above statement can be misunderstood as being equivalent to, "diseases result from the summation or interaction of genes and environments." If the BPSM entailed the latter statement, there would be no need to quarrel with it. The fact that it does not, however, can be confirmed by considering into which of the 3 categories of the BPSM (ie, biological, psychological, or social) the influence of radiation exposure on the development of disease, for example, would fall. Although radiation exposure is clearly an environmental factor, the BPSM would place it into its "biological" box. Thus, the BPSM separates some environmental influences on disease causation (eg, radiation and diet into the "biological" compartment) from other environmental influences (eg, neglect and witnessing violence) into the "psychological" or "social" compartments, thereby obscuring the point that different environments may simply be salient to different human structures and functions.
Specifically, the fact that some brain functions are exquisitely sensitive to events in the social environment to which the brain has access via the sense organs-one of the most important truths in all of psychiatry-is inconsistent with the framework provided by the BPSM. The imperative of the BPSM to dichotomize the various ways that environments affect brains (eg, toxic chemical agents: biological; sense organ input: psychological) is spurious and misleading and interferes with our ability to understand and talk coherently about the etiology and pathogenesis of psychiatric disorders.
The way the BPSM would have us conceptualize the causes of depression reveals how it fosters confusion of etiology and pathogenesis. Experience leads me to believe that most psychiatrists would agree with the statement, "some people are depressed because of abnormalities of neurotransmission in certain brain circuits, some people are depressed because of adverse events in their lives, and some people are depressed because of a combination of those factors." Yet, a sentence of identical logical structure, "some people have myocardial infarctions because of insufficient oxygen supply to their hearts, some people have myocardial infarctions because of shoveling snow, and some people have myocardial infarctions because of a combination of those factors," is readily seen as nonsensical. The fact that myocardial infarction necessarily entails insufficient oxygen supply to the heart, and that there is a variety of routes-including increasing the oxygen requirement of the heart by shoveling snow--by which that may occur, is well known and understood.
In psychiatry, however, the impulse, which is both reflected in and fostered by the BPSM, to see some environmental influences as somehow nonbiological obscures the crucial point (if one rejects mind-body dualism in favor of monism) that events in the social environment can be transduced into disease only to the extent that they affect the structure and function of the body. Far from being helpful in our attempts to understand and communicate information about the causes ofpsychiatric illness, the BPSM is, at least in this effort, a millstone around our collective neck.
Domain 2: Manifestations of illness
Statement: Illnesses may present with both psychological and physical (or biological) symptoms.
Here the categories of "biological" (or, alternatively, "physical") and "psychological" (or, alternatively, "mental") are applied not to putative etiologic factors but to the presentations of illness. In medical school, students are taught early to distinguish between the signs and the symptoms that comprise the manifestations of different diseases. They are taught that signs are abnormalities that are observed by the examiner (and, hence, are by definition objective or physical phenomena), while symptoms are problems that are experienced by the patient (and are, therefore, by definition subjective or psychological phenomena). In conventional usage in medicine (including psychiatry), however, the sign/symptom dichotomy is not congruent with the physical/psychological one. Thus, some subjective (psychological) experiences (eg, pain, dizziness) are separated from others (eg, anxiety, dysphoria) on the spurious basis that the former are (oxymoronically) "physical symptoms" while the latter are (redundantly) "psychological symptoms."
The BPSM-related convention of thinking in terms of physical or biological symptoms on the one hand and psychological symptoms on the other makes what might otherwise be a straightforward understanding of their commonalities unnecessarily convoluted. Thus, the exacerbating effect of anxiety on pain becomes a "psychological factor affecting a physical condition," or even a "mind-body" phenomenon, with the vaguely (and thoroughly unwarranted) mystical connotations that attend such a conceptualization.
Domain 3: Classification of illness
Statement: Some diseases are psychiatric, and some diseases are medical.
The nosologic separation of psychiatric disorders from all other maladies to which humans are prone became official with the adoption of the multiaxial diagnostic system of DSM-III, and it has been preserved in subsequent editions of DSM. The relation between this dualistic linguistic and conceptual convention and the BPSM is evidenced, among other ways, by the following explanation from one of its authors: "The main advantage of a multiaxial approach over a single-category (or axis) approach is that the former better reflects the complexity of and interrelationships between the various biopsychosocial aspects of the individual who is being evaluated."2 With no consensus on what makes a disorder "psychiatric" (ie, Axis I or Axis II), however, the justification for this practice is at best unclear.
The conceptual cost of dichotomizing the universe of disease in this fashion is left unaddressed. One drawback is that it obscures potential etiopathogenetic commonalities among conditions on opposite sides of the divide, and it creates the false impression that the mechanisms and/or manifestations of such diseases are distinct. A common scenario that illustrates this point is that of depression with mild hypothyroidism. Under the sway of the BPSM's progeny-the DSM multiaxial diagnostic system-such a clinical presentation is typically viewed as a problem of differential diagnosis: major depressive disorder versus depressive disorder due to hypothyroidism, with hypothyroidism (a "general medical condition" in the terminology of DSM) entered on Axis III. The diagnostician is thus misled into seeing these as essentially distinct disorders with distinct etiopathogeneses. The prevailing linguistic and conceptual conventions all but prevent students and practitioners from seeing thyroid dysfunction as a risk factor--along with others such as alcohol use, adverse life events, and family history of mood disorder-for depression.
Moreover, the DSM multiaxial diagnostic system causes descriptions of complex illnesses to be unnecessarily convoluted. Let us consider the example of a patient with depressed mood, bradykinesia, and cognitive dysfunction. Depending on the clinician's initial point of emphasis (psychiatric or "general medical"), the formulation may be either that the patient suffers from a mood disorder with associated psychomotor retardation and "depressive pseudodementia" (a misnomer), or that he or she has Parkinson disease with an associated depression "due to a general medical condition." The far more parsimonious (not to mention valid) conceptualization of depression and Parkinson disease as disorders that involve subcortical abnormalities with shared motor, cognitive, and affective manifestations is obscured when the dualism of the DSM multiaxial diagnostic system is allowed to guide our thinking. Far from carving nature at its joints, the multiaxial system appears to have applied its knife to a vital structure that should remain intact.
Domain 4: Treatment of illness
Statement: Psychiatric disorders may be (or should be) treated with biological, psychological, and social therapies.
While the BPSM is used by some clinicians as a sort of checklist that helps them to remember to consider the variety of available therapeutic options, its liabilities in this context are similar to those exposed under Domain 3. Specifically, dividing up psychiatric treatments in this way obscures potentially common mechanisms among them. Thus Baxter and colleagues,3 writing about their positron emission tomography (PET) findings in obsessive-compulsive disorder, felt the need to explain that "the possibility of both a serotonin reuptake inhibitor and behavior modification treatments having the same neural effects is not as farfetched as it might seem to some at first glance."
Further, the BPSM as applied to the conceptualization of differential therapeutics creates a false expectation of correspondence between putative causes or manifestations of illness on the one hand and components of treatment on the other. Thus, an illness thought to have a "biological" (often conflated with heritable) origin, and/or one that presents with "biological" signs or symptoms (a categorization often applied, for example, to vegetative abnormalities), warrants a "biological" (ie, pharmacologic or electroconvulsive) therapy. In contrast, a condition conceived of as resulting from a "psychological" or "social" cause (often conflated with adverse life events), and/or presenting with "psychological" symptoms (eg, identity disturbance), calls for use of a "psychosocial" intervention. Because the BPSM encourages its users to conceptualize psychopathology as resulting from a combination of biological, psychological, and social forces-which, in turn, manifest as a combination of physical and psychological symptoms-it appears to be an inescapable conclusion that essentially all patients should be managed with a combination of treatments selected from each of the categories that the BPSM comprises.
Discouragingly, this ideologic (as opposed to empiric) imperative is codified as dogma by the American Board of Psychiatry and Neurology. As Morrison and Munoz4 advise their anxious readers who are preparing for the oral portion of the psychiatry specialty board examination, "psychiatric management requires attention to these potential interventions: biological, psychological, social. This biopsychosocial framework underlies modern treatment formulation. At all costs, you must discuss each of its parts in the treatment plan of any patient you examine." Candidates for board certification in psychiatry ignore that admonition at their peril!
As I have shown, the BPSM is a linguistic and conceptual convention whose influence (on its own and via related linguistic and conceptual conventions such as the DSM multiaxial diagnostic system) is felt across a range of domains of psychiatric thinking. Although I am certainly not the first to criticize the BPSM and its uses in psychiatry (see, for example, Ghaemi5), what I have outlined (following Goodman6) is how, as a fundamentally dualist model that forces its users to employ the categories of "bio," "psycho," and "social" in various misleading ways, it is both a manifestation of and a contributor to misunderstanding. Practical consequences of that misunderstanding almost certainly include separation of psychiatry from the rest of medicine, stigmatization of patients with psychiatric illnesses, and an inequitable distribution of resources (eg, managed care "carveouts").
What I have focused on in this essay, though, are its conceptual costs, including distortions of the differential diagnostic and differential therapeutic processes and-of paramount importance-impairment of our ability to communicate modern psychiatric thinking to our students, residents, colleagues, patients, and the public. After all, it should not come as a surprise that 21st century science does not fit well into the 17th century dualist framework of the BPSM.
Into what conceptual structure, then, does modern psychiatric science fit? The answer, I believe, is a mature biological, or "holobiological," model. This is in contrast to the crude and naive biological reductionism of an earlier era that discounted the role played by the interpersonal environment in the etiopathogenesis and treatment of illness and to which the BPSM was in some ways a response. This mature or holobiological model is one that accommodates within an explicitly monist/ materialist framework the broad range of inherited and acquired phenomena that are salient to human disorders. Its mode of explanation is gene and environment action and interaction in the production of disease phenotypes (see, for example, Pezawas and colleagues7) and therapeutic responses. Guided by such a model, we can be in a position to fashion the future findings of psychiatric science into a coherent account-for ourselves, our students, our colleagues, and our patients-of the causes, manifestations, classification, and treatments of psychopathology.
Dr Waterman is associate professor of psychiatry and associate dean for student affairs at the University of Vermont College of Medicine in Burlington. He reports that he has no conflicts of interest concerning the subject matter of this article.
References1. Engel GL. The clinical application of the biopsychosocial model. Am J Psychiatry. 1980;137:535-544.
2. Williams JB. The multiaxial system of DSM-III: where did it come from and where should it go? I. Its origins and critiques. Arch Gen Psychiatry. 1985;42:175-180.
3. Baxter LR Jr, Schwartz JM, Bergman KS, et al. Caudate glucose metabolic rate changes with both drug and behavior therapy for obsessive-compulsive disorder. Arch Gen Psychiatry. 1992;49:681-689.
4. Morrison JR, Munoz RA. Boarding Time: The Psychiatry Candidate's New Guide to Part II of the ABPN Examination. 3rd ed. Washington, DC: American Psychiatric Press; 2003.
5. Ghaemi SN. The Concepts of Psychiatry: A Pluralistic Approach to the Mind and Mental Illness. Baltimore: Johns Hopkins University Press; 2003.
6. Goodman A. Organic unity theory: the mind-body problem revisited. Am J Psychiatry. 1991;148:553-563.
7. Pezawas L, Meyer-Lindenberg A, Drabant EM, et al. 5-HTTLPR polymorphism impacts human cingulate-amygdala interactions: a genetic susceptibility mech-anism for depression. Nature Neurosci. 2005;8:828-834.