Hikikomori and the false positives challenge: Comment on Amendola's “Clarifying the position of hikikomori in mental health”

Abstract

Amendola's review raises the question of whether some variants of hikikomori-type social withdrawal are not mental disorders but rather psychologically normal-range attempts to cope with a changing social environment that is unlike the one for which human beings were biologically designed. To place this question in context, I survey issues regarding the prevention of false-positive diagnoses of problematic non-disordered conditions as disorders. I offer a series of examples of other categories in which such errors have occurred and sometimes been corrected, drawing parallels to and implications for the consideration of hikikomori. I argue that hikikomori offers an opportunity for reflection on how novel problematic behaviors can be assessed for disordered versus non-disordered variants.

Simone Amendola's (2024) comprehensive and meticulous review of proposed definitions of hikikomori raises many fundamental questions about how to evaluate proposals to add newly emerging categories of disorder to the DSM and ICD. I will expand on an issue raised by Amendola but often ignored (although of recently increased salience; e.g., see Kato et al., 2024), namely, distinguishing disordered from non-disordered hikikomori and preventing false-positive diagnoses of non-disorder as disorder. In an era of social change, individuals may develop novel adaptive strategies that may seem disordered when they are merely unusual and socially problematic. Some subset of individuals displaying hikikomori are likely to fit this description. I will offer some examples from other categories that illustrate how disorder versus non-disorder judgments can go wrong as cautionary exemplars in considering hikikomori.

Psychiatry remains ignorant of the specific etiology of almost all mental disorders, and so uses descriptive syndromes to identify instances of disorders. Yet, most symptoms of disorder can also occur as normal-range responses. Thus, nosologists must judge when criteria likely indicate disorder versus non-disorder. As in all of medicine, this distinction is based on whether it can be plausibly inferred that a condition is caused by a dysfunction—a failure of some psychological mechanism to be capable of performing its biologically designed function (Wakefield, 1992). With no knowledge of etiology, this can be an extraordinarily difficult judgment to make.

Although mental health professionals treat a variety of human problems that are not disorders (see ICD's and DSM's “Z Codes”), psychiatric diagnostic criteria are specifically designed to distinguish mental disorders from other problems of living by identifying conditions plausibly caused by dysfunctions (First & Wakefield, 2013). This is not just a philosophical nicety or an irksome insurance requirement; the disorder/non-disorder distinction can have clinical prognostic, treatment-planning, and informed consent implications. It is also highly meaningful to various constituencies; most of the major disputes about diagnostic categories and criteria in the contentious run-up to DSM-5 concerned worries about false-positive diagnoses that misclassify normal-range problematic conditions as psychiatric disorders (Wakefield, 2016).

It is surprisingly easy to confuse contextually determined socially problematic normal-range reactions to novel social environments with mental disorders because we tend to read our society's values and expectations into human nature. An example is “Identity Disorder” in DSM-III-R, a diagnostic version of the “identity crisis” popularized by the psychoanalyst Erik Erikson. The turmoil commonly experienced by adolescents during this difficult period as they confront important life choices regarding personal identity in multiple domains including goals, religion, career, sexuality, morality, and friendship, was labeled a disorder. However, it was distinctive of modern Western cultures and almost always resolved by early adulthood. It was soon concluded that this condition was due not to a dysfunction but to a culturally imposed period of unusual psychological demand. Without fanfare, the condition was moved in DSM-IV to the non-disorder “Z-Code” section. Similarly, some subset of hikikomori could easily be a way of coping with a social context that imposes novel social demands on young people that did not exist in the environment that shaped human nature.

Similarly, adolescent antisocial behavior (“delinquency”) often qualifies for diagnosis as conduct disorder based on behavioral symptoms. Yet, it has been persuasively argued that this behavior is often a response to the lengthy time interval between physical maturity and social independence in modern social environments that did not exist in our evolutionary environments (Moffitt, 1993). This novel mismatch between developmental readiness and social constraints can yield violation of social rules as a socially maladaptive and problematic but psychologically normal-range reaction.

DSM incorporates a contextual exception to conduct disorder for another non-dysfunction etiology. It notes that the diagnosis may be misapplied to individuals in settings where disruptive behavior is normative (e.g., in very threatening, high-crime areas or war zones), so context must be considered. Similar considerations of social context play a role in determining dysfunction versus non-dysfunction in hikikomori. For example, most hikikomori researchers recognized that the COVID pandemic, in which it was socially acceptable to withdraw and working and schooling from home became common and socially supported, made subsequent choice of social withdrawal for non-dysfunction motives more likely and more personally acceptable. Outing criteria for a suspected disorder require continued reexamination in this regard.

DSM attempted to limit false positives by adding the clinical significance requirement (CSC), that the condition causes distress or role impairment, to the symptom criteria for most disorders. However, this strategy has limitations. Here is an example. The DSM-5-TR symptom criterion for childhood “selective mutism” is “Consistent failure to speak in specific social situations in which there is an expectation for speaking (e.g., at school) despite speaking in other situations.” In an attempt to avoid false positives, DSM adds a CSC requiring impairment: “The disturbance interferes with educational or occupational achievement or with social communication.” However, the CSC cannot do its job of eliminating symptomatic false positives because it is too broadly formulated. Anyone whose condition causes consistent failure to speak in social situations where speaking is expected is experiencing interference with social communication.

This vacuousness also afflicts the CSC for major depressive disorder (MDD). Anyone satisfying the 5-symptom requirement for MDD will thereby be impaired or distressed (Wakefield et al., 2010). Yet, ample false positives fall under the MDD criteria due to reactions to acute stressors (Wakefield & Schmitz, 2014). DSM-5-TR acknowledges this, advising the clinician to use clinical judgment to decide whether the patient's condition is intense normal-range sadness or pathological depression.

Similar problems with impairment emerge with hikikomori. The CSC usually requires distress or impairment, but studies suggest that distress is often not present in hikikomori. Amendola notes that Nonaka et al. (2022) found that impairment as well as distress are often not measured in hikikomori studies. However, from the perspective of social functioning, hikikomori is surely role impairing because one is not engaging in work, school, or in-person social interactions. The problem is that this makes impairment redundant with the main symptom of social withdrawal, and so does not help in avoiding symptomatic false positives, which is a problem if one accepts that in current social contexts social withdrawal in itself does not imply psychological dysfunction. It is important not to confuse social role impairment with the requirement that there be internal dysfunction causing the impairment to constitute pathology, even though “dysfunction” in a social sense is sometimes used for role impairment. For example, when Kato et al. (2024) say, “physical isolation itself is not pathological, but when dysfunction and distress are present, rapid mental health support should be provided” (p. 177), they seem to be referring to the CSC's required distress and impairment caused by the condition, but that must be distinguished from the dysfunction as an inferred pathological cause of the condition.

Amendola observes that different comorbidities and etiologies are suggested by subtle variations in the way hikikomori is defined (e.g., anxious avoidance of social interaction versus lack of interest in social interaction versus replacing face-to-face social interaction by internet gaming). His approach of comparing and elaborating on differences among possible definitions is a useful strategy. For cases of hikikomori that do not plausibly fall under existing categories, there may be multiple pathogenic and non-pathogenic etiologies revealed by such differences in subtle aspects of the nature of the motivation for social withdrawal. As Amendola also observes, comparative critique and analysis of variations in the definition can yield some insight into and protection from false positives.

Like a Rorschach test for nosologists, hikikomori presents an ambiguous stimulus into which pathological and normal-range psychological causes can be projected. By approaching this diagnosis in a conceptually careful way, we can learn how better to address the challenges of distinguishing mental disorders from novel but normal-range socially divergent reactions to social contexts and demands for which human beings were not biologically designed. Amendola's insightful review is a helpful step toward such self-reflection.