Dual Diagnosis: A Problematic Construct When Applied to Persons with Intellectual Disabilities

Abstract

The term dual diagnosis can refer to the co-occurrence of an intellectual disability and a mental disorder. While such a term may have some advocacy rationale aimed at facilitating improved mental health care for those with intellectual disabilities, it is proposed that the construct has flawed underpinnings, and its application may problematize mental health service delivery. A core concern is the promotion of categorical diagnostic models, whereas dimensional models may more accurately reflect underlying continuums for both cognitive and mental health challenges. A categorical diagnostic approach may also contribute to questionable dichotomization of mental health difficulties in persons with intellectual disabilities into “problem or challenging behaviours” versus “mental disorders.” Organizing services based on beliefs that such distinctions and categorical classifications are accurate may contribute to unnecessary and inappropriate fractionation of interventions and create additional service barriers for a vulnerable population. It is proposed that the term dual diagnosis be abandoned and replaced by systematic use of a dimensional approach to help facilitate assessment, intervention evaluation, and equitable service access.

Dual diagnosis refers to the co-occurrence of select combinations of mental health conditions in the same person. In the United States, this often refers to the co-occurrence of substance abuse and severe mental disorders, ¹ although that particular combination has been labelled concurrent disorders in Canada. ² This article will consider the term dual diagnosis as it is applied to persons who have both an intellectual disability (ID) and a mental disorder. While seemingly well intended, assumptions underlying this construct and aspects of its application may undermine service delivery to address mental health challenges in persons with cognitive difficulties.

Questioning the Pursuit of Categorical Mental Disorders

Advocacy intentions behind the term dual diagnosis include increasing awareness that persons with ID experience mental disorders similar to persons with otherwise typical cognitive development and may experience such at significantly higher prevalence rates but receive interventions at lower rates. ^3,4 In addition, mental disorders manifested by persons with ID are at risk for being missed due to diagnostic overshadowing. ⁵ This advocacy appears to take as a given the robust existence of categorical mental disorders. An indication of this confidence is the publication of the Diagnostic Manual–Intellectual Disability, ⁶ which is based on the Diagnostic and Statistical Manual of Mental Disorders (DSM). ⁷ This initiative modifies existing DSM criteria, which may not be readily discernable among those with more severe ID, particularly criteria requiring the person to articulate certain internal mental states (e.g., feeling depressed, experiencing hallucinations).

Unfortunately, timing of this categorical pursuit and promotion is decidedly off given that empirical evidence supporting underlying dimensions for mental health problems is expanding, and categorical classifications are increasingly questioned. It is now evident that many mental “disorders” have patterns more consistent with extreme ends of trait distribution in the population, rather than being qualitatively distinct entities that are discontinuous with trait distribution in the population (i.e., there is a lack of evidence for both zones of rarity and of identifiable taxons). ^{8 –10} For example, multiple lines of evidence have found that attention-deficit/hyperactivity disorder (ADHD) is much more consistent with a dimensional model versus existence as a discrete entity. ^9,11 ADHD is a commonly identified comorbidity with ID and occurs at a greater frequency in this population. ^{12 –14} However, the likelihood that ADHD exists as a discrete disorder for persons with ID but not those without ID, would seem remote. Rather than arguing that persons with ID can have ADHD too, it may be more informative to expand efforts to measure ADHD symptoms dimensionally to aid in determining aspects such as symptom cluster distribution in subpopulations of persons with ID and to determine relative responsiveness to interventions. While there is as dearth of studies examining the extent of evidence for categorical versus dimensionality underpinnings of mental health difficulties in persons with ID, it would seem reasonable to explore such hypotheses before prematurely embracing a categorical structure that is increasingly contested. ^{15 –18}

Questioning the Dichotomy between Problem Behaviours and Mental Disorders in Persons with ID

While challenges to the veracity of categorical mental disorders are not specific to those with ID, there may be unique concerns in applying a categorical approach to persons with ID. More specifically, there is a related notion that there exists a manifestation of “problem behaviours” (sometimes labelled “challenging behaviours”) distinct and separate from “mental disorders” in persons with ID. This particular dichotomization may be driven in part by concerns of overuse of psychotropic medications in persons with ID. This concern is sometimes exemplified by noting that only a fraction of persons with ID on psychotropic medications have been diagnosed with a mental disorder. ¹⁹ This exemplification may be further specified by emphasizing the low rates of psychotic disorders among those on antipsychotic medication. ^20,21

While there are compelling concerns over excessive and inappropriate use of psychotropic medications for persons with ID, conflating this concern with the belief that the use of psychotropic medications ought to align with categorical mental disorders is problematic. Otherwise, this might convey a belief that psychotropic medications are automatically deemed appropriate when a person with ID receives a mental disorder diagnosis.

The corollary concern in this dichotomization is that phenomena labelled problem behaviours are not a manifestation of a mental disorder and that psychotropic medications ought not to be used. This latter argument may be underpinned by legitimate concerns that persons with ID with problem behaviours are at risk for psychotropic medication(s) being the primary, and perhaps only, intervention attempt. The failure to consider the role of first-line functional behavioural assessments and behavioural and environmental interventions for problem behaviour would be inconsistent with best practice recommendations. However, reliance on this questionable dichotomization might result in premature narrowing of intervention options.

Intentional or not, a dichotomized model may underlie a service structure in which a presentation that is labelled as problem behaviour funnels persons with ID to a behavioural specialist (perhaps housed within developmental services funded by a social service ministry), while presentations labelled as mental disorders may funnel persons to psychiatric services (perhaps housed within medical services funded by a health ministry). A more nuanced model that moves beyond this simplistic bifurcation describes supplemental roles for (1) psychotropic medications for problem behaviours in the absence of a diagnosed mental disorder and (2) behavioural modification interventions for those diagnosed with a mental disorder. ²² While an advance, this model still builds on a belief in the existence of distinct entities, “problem behaviours” and “mental disorders,” and seems to retain an approach tracking the former to predominant behaviour interventions and the later to predominant psychotropic medication interventions. In contrast, if the pursuit of distinct categorical mental disorders is abandoned (except in those few cases for which there is compelling evidence for such) and there is instead a concerted effort to more systematically measure dimensions of specific behaviours (and clusters of behaviours), it is anticipated that the continuities between what are currently labelled problem behaviours and mental disorders will be more readily apparent. It is acknowledged that some in the field may approach various challenges in more nuanced and comprehensive ways, not confined by these dichotomized constructs. It is surmised that the dual-diagnosis concept is dispensable for those taking such an approach.

Questioning Intellectual Disability Thresholds for Informing Mental Health Interventions

To this point, a key argument has been to challenge the use of categorical mental disorders as a core component of the dual-diagnosis construct. However, other categorizations in this construct can be challenged, more specifically, the dichotomization of the population into those with and without ID. While there has been an effort to move away from excessive reliance on results from standardized testing for determining whether to classify someone as having an ID, it can be informative to consider the intelligence quotient (IQ) distribution in the population. Before DSM-5, specific IQ scores informed the attainment of ID (or what was then labelled mental retardation) and its subgroups. More specifically, an IQ of <70 was a threshold proposed as one criterion for meeting the diagnosis of mental retardation. That a point at approximately –2 standard deviations (SD) from a mean value should connote some meaningful clinic line is hard to justify. Such a cutpoint is not supported by discontinuities or a zone of rarity. At a clinical level, it would not seem to separate, in any meaningful way, a person with IQ results falling at –1.9 versus –2.1 SD from populations norms.

Considering the concept and impact of borderline intellectual functioning (BIF) may further question such artificial cutpoints. Within the DSM-IV, BIF had been operationalized as falling between –1 and –2 SD below the population IQ mean (an IQ of approximately 71-84; although DSM-5 has dropped such specific criteria). ²³ Various studies have identified that those falling within this IQ range may be at higher risk for negative social and mental health outcomes. For example, a population-based study in Australia found higher rates of mental health problems in those with BIF relative to a non-ID population and intermediate between non-ID and ID populations. ²⁴

It may be argued that since more contemporary diagnostic schemes have decreased excessive weighting of the role of IQ in categorizing someone as having an ID, an argument emphasizing IQ cutpoints is moot. Recent diagnostic schemes have increasingly emphasized the importance of levels of adaptive functioning and levels of needed support. ⁷ However, it is hard not to assume that adaptive functioning and levels of support also fall on continuums with no zones of rarity or clinically meaningful cutpoints. If it is accepted that the cutpoints on dimensions of cognitive functioning and intellectual disability are arbitrary, as they are for many categorical mental disorders, then one must conclude that attaining thresholds for a dual diagnosis is a function of crossing 2 (or more) arbitrary cutpoints.

Concerns about Adverse Service Impacts from Employing a Dual-Diagnosis Categorization

Arguments as to whether there is greater empirical support for dimensional versus categorical models for mental health and intellectual difficulties might be dismissed as primarily an academic exercise; however, there may be serious, and potentially adverse, real-world impacts on service delivery. This risk was articulated more than 30 years ago in the early days of this “new, faddish, innocent, ‘buzz word’”, dual diagnosis.” ^25(p156) These authors identified that “the most important potential implication of the ‘dual diagnosis’ classification is its exclusionary effect.” ^25(p156) This could take the form of agencies responsible for those with ID excluding those who are also identified with mental disorders, arguing that such persons are outside of their mandate or expertise, while mental health providers may exclude those mental health patients with identified ID with similar claims. ²⁵

Advocacy intentions behind promoting dual diagnosis may be to increase access to mental health specialists and associated interventions for those with ID. Understanding and/or labelling some problem behaviours as manifestations of underlying mental disorders might lead to hoped-for access to mental health services and specialists. This may be the case if access to mental health services is conditional on having a mental disorder, not just “problem behaviours.” To what degree such a formulation would increase services access is poorly known.

A more evident barrier may be exclusionary practices in some mental health services whereby persons with ID are explicitly excluded, making the attainment of a dual diagnosis irrelevant. Little systematic information appears available on the extent of such practices, although the clinician and administrator reader may well be aware of specific examples in their regions. In one study on child mental health waitlist management barriers in Canada, 11.5% of participating agencies acknowledged restricting services to children without developmental delays. ²⁶ In contrast, a study in the United Kingdom did not find significant differences in the receipt of child mental health services based on having an intellectual disability. ²⁷ Further studies on service access are needed.

There is also a lack of data on the extent to which persons with ID who have serious mental health problems might be excluded from ID services. The special designation as “dual diagnosis” might signal that the person will be too complicated for both typical mental health services and developmental disability systems. In such cases, a dual diagnosis patient may experience disproportionately longer wait times to access rare specialized dual-diagnosis mental health and developmental disability services. Such an approach would be inconsistent with other societal initiatives for greater integration and for prioritizing those with the most complex needs. Exclusionary practices and systematic exposure to longer service wait times would be functionally discriminatory.

Shifting to a Dimensional Approach

Given the proposed problems with the dual-diagnosis construct, it is recommended that such a designation be abandoned. Instead, challenges, be they cognitive and/or mental health, should be approached dimensionally. A more deliberate pursuit of a dimensional approach may lead to more consistent use of dimensional metrics in practice (not just research), which could facilitate clearer understandings of intervention impact. Some dimensional metrics are already available, for example, the irritability subscale of the Aberrant Behavior Checklist, a metric that has been extensively used to evaluate psychopharmacological interventions. ²⁸ Another dimensional measure, the modified version of the Yale-Brown Obsessive Compulsive Scale, attempts to quantify compulsive behaviours in persons with autism spectrum disorder (ASD). ²⁹ A dimensional anxiety measure for persons with ASD and ID is under development. ³⁰ Alternatively, or additionally, direct frequency measures of problem behaviours (e.g., frequency of self-injurious behaviour [SIB]) is another metric for assessment and intervention evaluation that is also not dependent on categorical diagnoses. Potentially, such dimensional measures might be incorporated at service intake to capture the severity of immediately relevant constructs (e.g., aggression). Additional measures may aim to capture constructs that may directly inform aspects of anticipated assessment and intervention, such as the nature and extent of functional communication, which would be more strategic than the questionably informative designation of having an ID. A key needed next step is to develop, refine, and evaluate the use of such dimensional tools for routine clinical use.

A dimensional approach may also facilitate better cross-field (e.g., behavioural and psychopharmacological) evaluations, particularly for concerning targets such as SIB and aggression. This contrasts with interpreting such manifestations as possible symptoms of a mental disorder (e.g., equivalents of a depressed mood as part of a depressive disorder, an interpretation that is increasingly questioned ³¹ ), an approach that may obscure behavioural dimensions. A dimensional approach may also equip the field to more coherently and systematically scrutinize the patterns under which psychotropic medications (e.g., antipsychotics) are employed and their resulting impact, positive and negative. A de-emphasis on categorical disorders might also address a concern that framing problem behaviours as mental disorders in those with ID might be imposing a “medical model,” which may too readily attribute problem behaviours to intrinsic aspects of the individual with a concomitant de-emphasis on a contextual understanding of such behaviour. ³²

A hesitation with such a direction may be underpinned by a belief that categorical diagnoses are critical for intervention decision making. This is clearly not the case for behaviour-based interventions. This is also not the case for a wider range of interventions that are increasingly recognized as having transdiagnostic applications. ³³ While it is recognized that a clustering of some symptoms may inform selection of specific interventions, it does not follow that categorical boundaries are necessary nor necessarily informative. For example, there is evidence of effectiveness of stimulant medication specific to ADHD symptom clusters for those with ID. ^34,35 However, there is no compelling evidence that the effectiveness of stimulants are tied to proposed cutpoints for those with or without ID. This pattern almost certainly extends to thresholds for other mental disorders.

In conclusion, it is proposed that a dimensional approach may better facilitate (1) triage and assessment, (2) intervention outcome monitoring and evaluation, (3) determination of the extent of equitable service access, and (4) investigations of more complex relationships between cognitive and mental health problems, no longer impeded by the use of artificial categories. In contrast to the potential benefits, this dimensional direction might lead to the inability to easily exclude groups based on categories and cutpoints, which may increase already high demands on limited services.