Abstract
Symptoms are vital representations of human disorders, are key to understanding disorders, and may be the focus of specific therapeutic efforts. Symptoms are imperfect, and there are many influences on how they are described and understood. Are they hand servants of diagnosis or important in their own right? The answer seems to be both, but diagnosis is typically the way in which communication about psychopathology occurs internationally in many clinical disciplines. Diagnosis is also the basis of knowledge of the natural history of psychopathology and its treatment and therapy. Investigations into the nature and meaning of symptoms can helpfully focus on emerging disorders, of which gaming disorder is provided as an example.
Researchers in recent decades have focused on symptoms as being of primary importance in understanding psychopathology and making diagnoses of mental health disorders. The nature and status of mental health symptoms are extensively reviewed in an article published by Wilshire and her colleagues in this issue (p. 323). In the present commentary, I will provide some context from the perspective of addictive disorders (as a distinct group of mental health disorders) and from personal involvement in the development of the following international diagnostic systems: the Diagnostic and Statistical Manual of Mental Disorders, currently in its fifth edition (DSM-5; American Psychiatric Association [APA], 2013); the International Classification of Diseases (ICD), the current revision being the 10th (ICD-10); and the 11th version of the ICD (ICD-11) approved by the World Health Organization (WHO) in mid-2019 (WHO, 1992, 2019).
The Purpose of Diagnosis
A central thrust of the article by Wilshire et al. (2021) is that symptoms are too often considered of subsidiary importance to diagnoses. They criticized, reasonably, the symptoms 1 in DSM-5 (APA, 2013) as being “thin” (p. 329). In addition, many symptoms in DSM-5 overlap with each other and/or combine multiple experiences (“portmanteau” symptoms). Wilshire et al. contrasted this with the symptoms network model (Borsboom, 2017), in which the network of symptoms is the diagnosis, and the Cambridge model (Marková & Berrios, 2009), in which symptoms are accorded primary importance for the understanding of psychopathology. The authors noted more briefly the Research Domain Criteria (RDoC) approach (Insel et al., 2010), in which psychopathology is explicitly hypothesized as a disorder of brain circuitry and symptoms contribute to but are not at the vanguard of understanding disorders and the diagnostic process.
How important, therefore, is diagnosis? Diagnosis is the central intellectual discipline of medicine, including psychiatry, and many forms of professional clinical practice. Its history dates back to at least Hippocrates (Kleisiaris et al., 2014), 2,400 years ago. Diagnosis has several purposes, which in broad terms can be summarized as follows:
to combine the person’s symptoms, experiences, reports from collateral sources, findings on mental state and physical examination, results of laboratory and imaging studies into a coherent understanding (the “diagnosis”) of what ails the person (which may be a disease, a disorder, a syndrome or a health risk factor); 2
in doing so, to ensure clarity of thinking by the practitioner and impose intellectual discipline;
to provide a convenient summary of observations and findings to communicate with the person affected;
to select appropriate treatment and therapy;
to indicate the natural course of the disorder and the expected outcome from treatment;
to communicate to referring practitioners and to colleagues providing ongoing advice, treatment, and therapy;
to provide a basis for the education and training of health professionals such that a common language is achieved; and
to provide a basis for scientific endeavor to develop and evaluate new therapies, new medications, and new procedures directed at cure or symptom relief.
At best, the diagnoses so defined have universal meaning and value throughout the world. Specifying diagnoses is vital for the development of new approaches for intervention, not least by ensuring that there is a clearly defined sample of participants in randomized controlled trials and cohort and treatment outcome studies.
Diagnostic Approaches in Mental Health
For some disorders, there is a known pathology: An example is traumatic brain injury. For some disorders, there is a known etiology at least in part—alcohol in the case of alcohol use disorder. For the majority of mental disorders, the etiology is uncertain, and mechanisms by which the disorder develops are complex and variably understood. Over several decades, there has been much effort directed toward the genetic underpinnings of mental disorders, gene–environment interactions, and the development of laboratory tests, physiological challenge tests, and neuroimaging studies. The limitations of these latter sources of information have become evident over this period.
As an example, functional neuroimaging studies identify distinct responses to alcohol-related cues in persons with alcohol use disorder/alcohol dependence (Bühler & Mann, 2011). This might seem promising in terms of making diagnoses in clinical practice. Currently available technologies enable pooled images (from multiple individuals) to demonstrate characteristic responses (e.g., suppression of uptake of externally administered radiolabeled opioids and increased oxygen utilization in the mesolimbic system) in persons with alcohol dependence and other addictive disorders who are cued (Bühler & Mann, 2011; Tanabe et al., 2019). However, the degree of variation from individual to individual is such that this type of imaging does not allow for precise individual diagnosis. An example of a physiological diagnostic test is the dexamethasone suppression test, which offered promise in the diagnosis of major depression and its subtypes (Gaudiano et al., 2009) but has now largely fallen into disuse. Laboratory tests can be helpful to suggest certain diagnoses, for example blood or breath alcohol concentrations in the diagnosis of alcohol intoxication.
The primacy of symptoms in mental health
For these and other reasons, a descriptive approach to diagnosis has held sway for classifying psychopathology (Compton & Guze, 1995). Notwithstanding the above comments on the key importance of diagnosis, accuracy in understanding and expressing symptoms is essential. This is both where they contribute to a diagnosis and when the symptom may be of such importance that it forms the focus of therapeutic action.
The way in which symptoms are elicited varies widely. In medical training, one is taught to listen to the patient’s narrative history and then ask direct questions to elicit experience of specific symptoms. The total experience of symptoms (and other diagnostic information as available) is then compared with the known features of a series of disorders, including what is considered the most likely diagnosis and differential diagnoses. In some cases and in some professions, an interview schedule is used to structure the assessment, and these have varying degrees of structure; in highly structured schedules, the question is posed in an unvarying manner.
Questions about symptoms
However symptoms are elicited and recognized, they may have different meanings for the person and depend on the individual and societal contexts. This is the first question posed by Wilshire et al (2021). An example from addictive disorders is “craving.” This is the strong desire or urge experienced to use a substance. It can occur in various contexts, including when a person is cued by (a) the sight, smell, or taste of a substance; (b) the consumption of the substance; or (c) more general cues such as exposure to the environment associated with substance use. Craving may or may not have physiological accompaniments such as increases in pulse rate and blood pressure. It may also be regarded as an undesirable experience to report, one that might lead to a diagnosis the person would prefer to avoid (e.g., alcohol dependence or alcoholism), with their attendant stigmas.
The second question identified by the authors is the level of abstraction that is necessary to understand the symptom. Tolerance is a case in point. It implies neuroadaptation to the substance consequent on changes in the activity of certain neurotransmitter systems, for example, mesolimbic reward systems. It is elicited by identifying whether the substance results in a lesser effect than before (e.g., a desired sensation) or when a person uses a larger amount of the substance to achieve that effect. This needs to be assessed in longitudinal perspective. Thus, there are symptoms and also a time frame to be considered.
The third question posed relates to the importance of descriptive models of each symptom, with the proposition that multiple levels of analysis (behavioral, phenomenological, neurophysiological, and physiological) are relevant. Some symptoms are inherently cognitive, for example, craving. Some are clearly behavioral, for example, continued use of a substance despite harm. Analyzing all symptoms along these lines is an interesting intellectual pursuit, but some may naturally align with one of these levels of analysis.
Causal explanations (the fourth question) introduce another level of complexity. A given symptom could arise from distinctly different causes; alternatively, the same disorder may be expressed through a variety of different symptoms. This may be less of an issue when the symptom is considered as part of the narrative history. Symptoms due to substance use may be conflated with those of an independent mental health disorder. This may not be resolved by an analysis of the possible cause of the symptom at a single point in time and may require successive evaluations. This is recognized by the requirement for persistence of symptoms beyond the stages of intoxication and withdrawal for a diagnosis of a substance-induced mental disorders to be made and for 1 to 6 months to elapse before an independent mental disorder can be diagnosed with confidence.
In their fifth question, Wilshire and colleagues asked what role symptoms should play in furthering the understanding of mental illness and whether they should be the primary target of research. This provides an opportunity to draw conclusions about the comparative research needs of understanding symptoms and understanding broader entities such as disorders and the diagnostic process.
Research priorities
The research effort required to assess the hundreds of symptoms that apply to all mental, behavioral, and addictive disorders would be huge. It is undeniable that certain symptoms such as those in DSM-5 (APA, 2013) have been essentially unchanged for some decades. More and more sophisticated analyses have been undertaken to group them in different combinations to produce disorders in DSM-5 (Hasin et al., 2013; Saunders et al., 2007). The ICD-11 has taken a different approach of aggregating features and symptoms that formed diagnostic guidelines for substance dependence in ICD-10 into “addiction constructs” that form the ICD-11 guidelines (Saunders et al., 2019), using a “conceptual-pragmatic-confirmatory” approach. This may seem to remove diagnoses further from a focus on individual symptoms. How much research and over how many years would it take to reexamine individual symptoms along the lines suggested?
Priorities need to be identified. In DSM-5 (APA, 2013), Internet gaming disorder has been introduced as a provisional diagnosis, and in ICD-11, gaming disorder is a new diagnosis (WHO, 2019). Much research into gaming disorder has been based on generic addiction constructs. What is missing at present is an understanding of the totality of symptoms and experiences among persons with different levels of gaming involvement. Impaired control is a feature, as is prioritization of gaming over other activities. How is prioritization operationalized in persons at different life stages or in different cultures in which the time spent on an activity may not be a defining issue. What is the place of “immersion” in distinguishing people who game without problems and people with a gaming disorder? Is there an equivalent of tolerance? Does withdrawal exist? What are the experiences in gamers that are still to reach the surface?
An important step from the work of Wilshire et al. (2021) is to identify where focusing on the nature and meaning of symptoms would generate new data on key disorders, for example on gaming when it becomes problematic. In turn, it would provide a platform for a raft of societal and clinical responses for the prevention and management of emerging disorders.
Footnotes
Acknowledgements
The views expressed in the present commentary are those of the author and are not intended to represent any policy or decisions of any other individual or organization.
Transparency
Action Editor: Kenneth J. Sher
Editor: Kenneth J. Sher
Author Contributions
J. B. Saunders is the sole author of this article and is responsible for its content.