Biostatistical Theory Research Articles

Reference-Class Problems Are Real: Health-Adjusted Reference Classes and Low Bone Mineral Density.

Elselijn Kingma argues that Christopher Boorse's biostatistical theory (the BST) does not show how the reference classes it uses are objective and naturalistic. Recently, philosophers of medicine have attempted to rebut Kingma's concerns. I argue that these rebuttals are theoretically unconvincing, and that there are clear examples of physicians adjusting their reference classes according to their prior knowledge of health and disease. I focus on the use of age-adjusted reference classes to diagnose low bone mineral density in children. In addition to using the BST's age, sex, and species, physicians also choose to use other factors to define reference classes, such as pubertal status, bone age, body size, and muscle mass. I show that physicians calibrate the reference classes they use according to their prior knowledge of health and disease. Reference classes are also chosen for pragmatic reasons, such as to predict fragility fractures.

Can machine learning make naturalism about health truly naturalistic? A reflection on a data-driven concept of health

Through hypothetical scenarios, this paper analyses whether machine learning (ML) could resolve one of the main shortcomings present in Christopher Boorse’s Biostatistical Theory of health (BST). In doing so, it foregrounds the boundaries and challenges of employing ML in formulating a naturalist (i.e., prima facie value-free) definition of health. The paper argues that a sweeping dataist approach cannot fully make the BST truly naturalistic, as prior theories and values persist. It also points out that supervised learning introduces circularity, rendering it incompatible with a naturalistic perspective. Additionally, it underscores the need for pre-existing auxiliary theories to assess results from unsupervised learning. It emphasizes the importance of understanding the epistemological entanglements between data and data processing methods to manage expectations about what data patterns can predict. In conclusion, the paper argues against delegating the final authority for defining complex concepts like health to AI systems, as it necessitates ethical judgment and capacities for deliberation that AI currently lacks. It also warns against granting creators and deployers of AI systems the discretionary authority to determine these definitions outside the wider social discussion, advocating for ongoing public engagement on normative notions. Failure to do so risks limiting individuals and collectives’ ability to shape a just digital future and diminishes their fundamental epistemic agency.

Mental disorder: An ability-based view

What is it to have a mental disorder? The paper proposes an ability-based view of mental disorder. It argues that such a view is preferable to biological dysfunction views such as Wakefield’s Harmful Dysfunction Analysis and Boorse’s Biostatistical Theory. According to the proposed view, having a mental disorder is basically a matter of having a certain type of inability (or: an ability that is not sufficiently high): the inability to respond adequately to some of one’s available reasons in some of one’s reasons-sensitive attitudes or actions, where the threshold of inability is determined by one’s being harmed. The relevant concepts of inability, reasons, and harm are sketched. The paper argues that the proposed view evades some problems of biological dysfunction views by remaining neutral on questions of causation and the evolution of the mind. Furthermore, it can capture better what is distinctively “mental” about mental disorder. On the proposed view, it is the rational relations among an individual’s attitudes and actions that are “disordered” and the relevant norms in mental disorder are the norms of reasons. As further merits, the view can account for degrees of disorder, incorporate biological as well as social aspects, and elucidate the relations among disorders, symptoms, and their causes.

Osteoporosis and risk of fracture: reference class problems are real

Elselijn Kingma argues that Christopher Boorse’s biostatistical theory does not show how the reference classes it uses—namely, age groups of a sex of a species—are objective and naturalistic. Boorse has replied that this objection is of no concern, because there are no examples of clinicians’ choosing to use reference classes other than the ones he suggests. Boorse argues that clinicians use the reference classes they do because these reflect the natural classes of organisms to which their patients belong. Drawing on a thorough exploration of how the disease osteoporosis is defined in adults, I argue that clinicians do indeed make choices about which reference classes to use in diagnosis. Clinicians use young adult reference classes to diagnose osteoporosis in elderly patients. They also use young female reference classes to diagnose osteoporosis in elderly males. Clinicians adjust their reference classes so that the diagnosis of osteoporosis reflects a person’s risk of sustaining a fragility fracture. The ethical intuition that people with the same risk of fracture should receive the same diagnosis overwhelms the naturalistic intuition that reference classes should reflect natural classes of organisms of uniform functional design. Clinicians construct a variety of reference class types, including pathological reference classes and epidemiological population-specific reference classes, to serve this ethical intuition. I show how clinicians use several reference classes at once so that they can more accurately predict risk of fracture. Ultimately, the reference classes chosen and used in medical practice are quite different from those proposed in naturalistic philosophy of medicine.

Naturalism, Disease, and Levels of Functional Description.

The paper engages Christopher Boorse's Bio-Statistical Theory (BST). In its current form, BST runs into a significant challenge. For BST to account for its central tenet-that lower-level part-dysfunction is sufficient for higher-level pathology-it must provide criteria for how to decide which lower-level parts are the ones to be analyzed for health or pathology. As BST is a naturalistic theory, such choices must be based solely on naturalistic considerations. An argument is provided to show that, if BST is to be preserved, such parthood choices are based on non-naturalistic considerations. We demonstrate that even when parthood choices are based on the best way to preserve BST, there are counterintuitive results which bring the central tenet of BST into question.

Distinguishing Health from Pathology

This essay provides an account of how to distinguish between health and pathology of trait tokens in medical theory. It (1) proposes to distinguish between two health/pathology concepts—health/pathology pertaining to survival and health/pathology pertaining to reproduction. It (2) defines measures for survival-efficiency and reproduction-efficiency of performances of physiological functions. It (3) provides an account of how, using the efficiency measures, to draw the line between health and pathology. The account draws, but seeks to improve, on Christopher Boorse’s biostatistical theory. In relation to that theory, the suggested account has the advantages (1) that it defines efficiency and (2) that it harmonizes with judgments in medical theory in cases of common diseases and “normal aging.” Furthermore, the essay argues against a competing idea of how to improve on the biostatistical theory, advocated by Peter Schwartz and Daniel Hausman.

A Universal Definition of Mental Disorder: Neither Necessary nor Desirable

A Universal Definition of Mental Disorder:Neither Necessary nor Desirable G. Scott Waterman (bio) Psychiatry's relation to the rest of medicine is ambivalent. Its legitimacy as a specialty is often conceived as being closely linked to its fidelity to the fundamental paradigms of medicine, especially the centrality of diagnosis and the association of diagnosis with treatment indications. However, as Gagné-Julien notes, a major impetus behind the quest for a solution to the demarcation (i.e., disorder vs. not-disorder) problem in psychiatry is "growing concerns regarding over-medicalization" (Gagné-Julien, 2021, p. 353). Although it could appear that these two considerations point in opposite directions, both arguably would be served by what Gagné-Julien labels "a satisfying definition" of mental disorder. Such a definition would, the conventional thinking goes, rely on the notion of function/dysfunction while minimizing, if not eliminating, the role of values in distinguishing disorder from normality. It would, thereby, purportedly be more like diagnosis in the rest of medicine and reduce the risk of labeling as pathological phenomena that are merely socially disvalued. Gagné-Julien reviews the two most prominent representatives of what Varga has dubbed "natural function objectivism" (Varga, 2011): Boorse's biostatistical theory and Wakefield's harmful dysfunction analysis (HDA). Agreeing with several previous commentators, she finds biostatistical theory wanting with respect to the traditional desideratum of being value-free, while HDA presents different shortcomings related to its apparent reliance on greater knowledge of human evolutionary history than is available to us. She nevertheless maintains optimism that a definition of mental disorder that relies on functional language can be salvaged. The road she travels to an account of mental disorder as dysfunction that is sufficiently objective, albeit still value-laden, is serpentine but well worth taking with her. It traverses Cummins' "causal role" theory of function in biology, which circumvents the primary objection to HDA. She concedes its inability to distinguish function from dysfunction in a way that is independent of human interests, but argues that augmentation of causal role theory with explicit acknowledgement of the values-dependent nature of all of science might yield a suitable framework within which to develop a satisfying definition of mental disorder. The details of how that can be accomplished are beyond the scope of her paper, but perhaps the most intriguing element of Gagné-Julien's prescription [End Page 377] for formulation of such a definition is her invocation of Alexandrova's work on "social objectivity," which, as a systematic and democratized approach to identifying and scrutinizing normative assumptions that might otherwise remain implicit and/or parochial, resonates with recent work in critical psychiatry (e.g., Lewis, 2008). As much as I agree with much of Gagné-Julien's argument, in the remainder of this commentary I wish to challenge a couple categories of assumptions that seem to underlie and motivate the development (and belief in the feasibility) of a definition of mental disorder founded on the notion of function/dysfunction. I will then use Gagné-Julien's recommendation of democratization of the process of definition formulation as a point of departure to suggest what might appear to be a radical solution, but one I believe to be conceptually defensible and pragmatically desirable. In the introduction to the article in which he coins the expression "natural function objectivism," Varga recites—as one of the motivations behind the effort, begun in the 1970s, of the American Psychiatric Association to define mental disorder—the Szaszian argument that the "problems in living" that constitute the subject matter of psychiatry are "really non-medical in nature" (Varga, 2011). The implied presupposition of an essentialist account of what distinguishes a medical disorder from a problem in living is Szasz's, not Varga's. More to the point, it is false—both with respect to the somatic vs. mental distinction with which Szasz was primarily concerned and with the disorder vs. not-disorder one that is the focus of the present discussion. Readers in my age bracket will attest that advertisers on the broadcast television shows we watch know their demographic, as evidenced by the high proportion of commercials for treatments of a range of maladies. In one of them...

Comparative Analysis and Application of Deep Neural Networks in Covid-19 Prediction

Machine learning specifically deep or reinforcement learning is a growing field. With the advent of the SARS COV 2 pandemic, its application for Covid 19 prediction has continued to evolve as traditional prediction approaches are rendered inaccurate or outdated with the challenges paused as we globally try to understand the trend of this global pandemic. Deep learning coupled with bio-statistical theory and approaches provide a new way of tackling the prediction challenge in Covid 19.This new approach factors in complex variation in variables that is non-linear, multivariate and with multiple independent variables. We assess the promise entailed in automated machine learning, SIR & Hybrid SIR Models referred to as SEIR (Susceptible-Exposed-Infected-Recovered) Models and LSTM RNN (Long-Short Term Recurrent Neural Networks. These three approaches directly inform CDC, WHO, major entities sharing prediction results on the pandemic and individual government health organs globally. We explore why they exhibit efficiency in arriving at predictions as the variables, geography and demographics fed into each keep varying. The criticality of the assessment we arrive at is rigorously tested and validated using K-Fold validation, Mean Accuracy Prediction Error (MAPE) and we also plot receiver operating characteristic (ROC) curves the results are later on exhibited showing Auto-ML.

Theoretical and clinical disease and the biostatistical theory

Although concepts of disease have received much scrutiny, the benefits of distinguishing between theoretical and clinical disease—and what is meant by those terms—may not be as readily apparent. One way of characterizing the distinction between theoretical and clinical conceptions of disease is by relying on Boorse's biostatistical theory (BST) for a conception of theoretical disease. Clinical disease could then be defined as theoretical disease that is diagnosed. Explicating this distinction provides a useful extension of the BST. The benefits of this approach are clearly and non-normatively demarcating disease from non-disease, while allowing for values and purpose to determine what criteria are used in clinical practice to represent a disease's underlying dysfunction. Through discussion of a variety of medical conditions, including polycystic ovary syndrome and type 2 diabetes mellitus, I explore how the relationship between BST-based theoretical and clinical disease could make sense of various features of clinical practice and medical theory. It could do this by lending focus to a nuanced understanding of the pathophysiological defects present in disease and the means by which they are assessed. This could contribute to making sense of revised nosologies and diagnostic criteria.

Health and Reference Classes.

In this article, I address two objections developed by Kingma against Boorse’s (1977) bio-statistical theory of health, the objections that choice of reference classes renders the theory both circular and problematically value-laden. These objections not only apply to the bio-statistical theory of health but also to other naturalistic theories, like the dispositional theory of health. I present three rejoinders. First, I argue that the circularity objection arises from excessive methodological demands. Second, I argue that naturalists can resist the normativist claim that health and pathology are differentiated on the basis of personal or cultural values. Finally, I show that it is possible to justify choices between rival theories of health without the interference of evaluative commitments. With these rejoinders, I conclude that the bio-statistical theory, as well as other naturalistic theories of health utilizing reference classes, is not undermined by Kingma’s arguments.

Survival, Reproduction, and Functional Efficiency

The article examines the relationship between a trait’s effect on survival and reproduction and the notion of functional efficiency underlying the biostatistical theory of health (BST). BST faces the problem of how to measure a trait’s joint effect on survival and reproduction in its account of function. If one measures the joint effect by means of the biological notion of fitness, examples such as the hereditary breast and ovarian cancer syndrome do not count as a disorder. If one does not invoke biological fitness, it is unclear how to measure the joint effect while keeping to BST’s naturalist credentials.

Epistemic Authority, Philosophical Explication, and the Bio-Statistical Theory of Disease

Christopher Boorse’s (Philos Pub Aff 5:49–68, 1975; J Theory Soc Behav 6:61–84, 1976; Philos Sci 44:542–573, 1977; in: VanDeVeer and Regan (eds) Health care ethics: an introduction, Temple University Press, Philadelphia, pp 359–393, 1987; in Humber, Almeder, Totowa (eds) What is disease?, Humana Press (Biomedical Ethics Reviews), New York City, pp 1–134, 1997; J Med Philos, 39:683–724, 2014) Bio-Statistical Theory (BST) comprehends diseases in terms of departures from natural norms, which involve an objectively describable deviation from the proper physiological or psychological functioning of parts of the human organism. I argue that while recent revisions and additional considerations (Boorse 2014) shield the BST from a number of issues raised by critics, they give rise to significant new challenges. These are related to the attribution of epistemic authority, the possibility of multiple concepts in the authoritative literature, and the framing of BST as “philosophical explication.” The conclusion is that, in its current form, the BST is unable to provide a robust framework for a naturalist classification.

Boorse's Theory of Disease: (Why) Do Values Matter?

There has been much debate about whether the concept of disease articulated in Boorse's biostatistical theory is value-neutral or value-laden. Here, I want to examine whether this debate matters. I suggest that there are two basic respects in which value-ladenness might be important: it could threaten either scientific legitimacy or moral permissibility. I argue that value-ladenness does not threaten the scientific legitimacy of our disease-concept because the concept makes little difference to the formulation and testing of scientific hypotheses. Likewise, even if our disease-concept is value-laden, this does not show it is morally impermissible. To determine its permissibility, we must regard it as a tool to structure social institutions. Seen in that way, whether our disease-concept is morally permissible is a consequentialist matter and depends exquisitely on unanswered empirical questions.

Time for a Change: Topical Amendments to the Medical Model of Disease

There is a conceptual crisis in the biomedical sciences that is particularly salient in psychopathology research. Underlying the crisis is a controversy that pertains to the current medical model of disease that largely draws from causal-mechanistic explanations. The bedrock of this model is the analysis of biological part-dysfunctions that aims at unequivocally defining a pathological condition and demarcating it from its neighboring entities. This endeavor has led to a quest for physiological, biochemical, and genetic signatures. Yet, so far there is little evidence for reliable biomarkers for any mental disorder. The contemporary biomedical paradigm largely ignores historical, dynamic, and system-level aspects—a view that has contributed much to the conceptual disjunction of the patient as a person from his/her disease. Notwithstanding the impressive progress in the biomedical sciences, increasingly more critics question whether the constituting framework is sufficient to convey a comprehensive understanding of illnesses, especially mental illnesses. Thus, the medical model urgently requires an update. But rather than revamping it by methodological advancements, it will be necessary to critically review its philosophical roots. The most problematic issues that require reworking are: the preponderance of the biostatistical theory; the undue decoupling of physiological from evolutionary explanations of function; the clinging to the Modern Synthesis (if etiological aspects are ever considered); and the neglect of dynamic and system-level properties. The proposed overhaul requires the heeding of historical explanations that draw from the Extended Evolutionary Synthesis as well as systems biology approaches for tackling multilevel and dynamic phenomena of complex systems.

Defining “mental disorder” in legal contexts

Despite its widespread use, the term “mental disorder” has proven remarkably difficult to define. This is of grave concern in those legal contexts which crucially depend on the classification of an individual’s mental health condition as a “mental disorder”. It creates a likelihood that individuals will be treated in an inconsistent and unprincipled fashion, which is unacceptable in a legal system which purports to operate under the rule of law. This article examines the way in which the term “mental disorder” should be defined in legal contexts. It critiques the approach taken by the two main psychiatric manuals currently in use (the DSM-5 and the ICD-10), as well as four suggested alternatives: Boorse’s biostatistical theory; Wakefield’s harmful dysfunction theory; Jaspers’ lack of meaningful connections approach; and the psychological view of mental disorder as maladaptive behavior. It recommends the adoption of a context-specific, purpose-based approach that focuses on the normative concerns of the law.

Current Dilemmas in Defining the Boundaries of Disease.

Boorse's biostatistical theory states that diseases should be defined in ways that reflect disturbances of biological function and that are objective and value free. We use three examples from contemporary medicine that demonstrate the complex issues that arise when defining the boundaries of disease: polycystic ovary syndrome, chronic kidney disease, and myocardial infarction. We argue that the biostatistical theory fails to provide sufficient guidance on where the boundaries of disease should be drawn, contains ambiguities relating to choice of reference class, and is out of step with medical processes for identifying disease boundaries. Although proponents of the biostatistical theory might regard these practical issues as irrelevant to the aim of providing a theoretical account of disease, we take them to indicate the need for a theoretical account that is adequate for current needs-including limiting new forms of medicalization that are driven by the identification of disease based on dysfunction. Our processes for determining the boundaries for disease need to recognize that there is no value-free method for making these decisions.

Epidemiology and the bio-statistical theory of disease: a challenging perspective.

Christopher Boorse's bio-statistical theory (BST) of health and disease argues that the central discipline on which theoretical medicine relies is physiology. His theory has been much discussed but little has been said about its focus on physiology or, conversely, about the role that other biomedical disciplines may play in establishing a theoretical concept of health. Since at least the 1950s, epidemiology has gained in strength and legitimacy as an independent medical science that contributes to our knowledge of health and disease. Indeed, it not only provides important information about disease distribution and aetiology, but the risk-factor approach it employs seems to challenge BST's binary conception of health and disease. The objective of the article is to show, firstly, how important information deriving from descriptive and analytical epidemiology forms part of our contemporary medical concepts of health and disease, and secondly, that these elements are not taken into account by BST in a satisfactory way. The article's central contention, therefore, is that if the project of defining the theoretical concept of health is to be maintained, more importance should be accorded to the contribution made by epidemiology--alongside physiology--in defining health.

Ορίζοντας την υγεία και την ασθένεια: Η αυτογνωσία ως κριτήριο υγείας

A central issue in the philosophy of medicine is to solve the problem of determining the nature of the concepts of health and disease. Some theorists claim that health and disease are purely value-free and descriptive concepts that are discoverable and grounded in the biological and medical sciences. Others claim that health and disease are essentially value-laden concepts, i.e. healthy states are those states we (individuals, groups, societies) desire or value and diseased states are those we want to avoid or disvalue.Considering the longtime disagreement among physicians, theorists of health and philosophers about the essential characteristics of health and disease, the aim in the first part of this paper is to briefly present four different philosophical approaches of health and disease in an attempt to introduce the readers to the debate. This article begins with the Biostatistical theory of Christopher Boorse. It turns to the Malady theory of Danner Clouser, Charles Culver and Bernard Gert, the holistic theory of Lennart Nordenfelt and finally the adaptation theory of József Kovács.In the second part of the paper, some objections will be stated to each theory. In the third and last part, a completely different account of health will be given, which holds that self-knowledge is a criterion of human health. Health, unlike disease, is not a state or a condition of the organism but an endless process. Its essence can be better captured in philosophical rather than medical terms and it should be applied only to human beings.

Situation-Specific Disease and Dispositional Function

In ( Kingma [2010] ), I argued that Boorse's ([1977], [1997]) biostatistical theory of health (BST) is unable to accommodate diseases that are the normal result of harmful environments. Hausman ( [2011] ) disagrees: if the BST compares normal dispositional function against the whole population or reference class, rather than against organisms in similar circumstances as I proposed, then my ( [2010] ) challenge can be avoided. In this paper, I argue that Hausman's response fails: his proposal cannot accommodate a series of common physiological processes, such as sleep and those involved in reproduction. In the course of this argument, I also offer a detailed discussion of the concept of functional efficiency ( Hausman [2011] , [2012] ), and reveal a link between the problem of environments ( Kingma [2010] ) and that of reference classes ( Kingma [2007] ). 1 Introduction 2 Kingma on Boorse 3 Hausman on Kingma 4 Response to Hausman 5 Three Possible Replies 5.1 Reference classes 5.2 Temporary pathologies 5.3 Adequacy of function 6 Concluding Discussion

Health, homeostasis, and the situation-specificity of normality

Christopher Boorse's Biostatistical Theory of Health has been the main contender among naturalistic accounts of health for the last 40 years. Yet, a recent criticism of this theory, presented by Elselijn Kingma, identifies a dilemma resulting from the BST's conceptual linking of health and statistical typicality. Kingma argues that the BST either cannot accommodate the situation-specificity of many normal functions (e.g., digestion) or cannot account for many situation-specific diseases (e.g., mountain sickness). In this article, we expand upon with Daniel Hausman's response to Kingma's dilemma. We propose that recalling Boorse's specification that health is an intrinsic property of its bearers and explicating this intrinsic property in relation to the concept of homeostasis can illuminate how proponents of naturalistic accounts of health should deal with the situation-specificity of normal functions. We argue that beyond what Boorse and Hausman have delineated, the situation-specificity of normal function cannot be fully captured in a simple dichotomy between normal and abnormal environment or between relevant and irrelevant situations. By bringing homeostasis to the fore of the analysis of health, we set out a richer picture of what the various situations that affect living organisms' functional performance can be. Accordingly, we provide a broader classification of these various situations which, we contend, better accounts for the main intuitions that philosophers of medicine have sought to accommodate than previous naturalistic theories of health.