The autism spectrum disorder ‘epidemic’: Need for biopsychosocial formulation

Introduction

In DSM-IV autism spectrum disorder (ASD) was defined as a cluster of lifelong neurodevelopmental disorders consisting of autistic disorder, Asperger’s syndrome and pervasive developmental disorder not otherwise specified (PDD-NOS). ASDs can be disabling disorders and warrant multi-faceted assessments and interventions that are often required in some form across the lifespan.

The prevalence of ASDs has climbed dramatically in recent years and an over-diagnosis ‘epidemic’ has been suggested (Frances and Batstra, 2013). Partly to reduce over-diagnosis, ASDs have been consolidated in DSM-5, such that the subtypes are no longer used but all are referred to as ASD and severity is specified. ASD in DSM-5 is characterised by deficits in social communication and social interaction plus restricted repetitive behaviours, interests and activities (RRBs). Subjects without RRBs can be given a new DSM-5 diagnosis: social communication disorder. However, the DSM-5 field trials suggest most individuals currently diagnosed with ASDs will still have ASD under the new criteria. This allayed the fears of some advocacy groups.

We argue that there is an over-diagnosis ‘epidemic’ of ASDs, particularly in the paediatric population, and doubt that DSM-5 criteria will minimize this. The underlying drivers of this and similar over-diagnostic ‘epidemics’ is a combination of the descriptive symptom-focused and context-deficient nosology inherent in the DSM model combined with needs for services and other psychosocial gains associated with the use of diagnostic labels. A false ASD diagnosis can provide practical benefits such as financial and educational assistance, but it comes with side effects that may take years to be fully apparent.

Rising prevalence of ASDs

The prevalence of ASDs has dramatically increased over recent decades. Most studies conducted from 1960 to 1980 show a prevalence ranging from 2 to 5 in 10,000. However, these studies assessed narrowly defined autistic disorder (Fombonne, 2009). Studies published in the early 2000s reported prevalence ranging from 30 to 60 in 10,000, rising in recent years to 50 to 114 in 10,000 children (Baird et al., 2006). Prevalence rates for different subtypes of ASD have varied considerably. In most studies, the number of children with Asperger’s syndrome, Rett’s syndrome, and particularly PDD-NOS, has outnumbered children with autistic disorder by almost 2 to 1 (Rosenberg et al., 2009).

Postulated driving factors for the ASD ‘epidemic’

Possible factors include changing diagnostic criteria, differing study methodologies, the coexistence of the disorder with a range of other conditions; heightened awareness of both professionals and the general public; an increase in services to meet the growing numbers of identified children which may further facilitate increased diagnosis; the ongoing investigation of possible aetiological factors; media publicity; and a professional and societal shift towards a biomedical perspective for human emotional and behavioural problems (Rutter, 2009). All of these can result in increased case identification, diagnostic substitution and diagnostic accretion.

To what extent the increase in ASD diagnoses is ‘real’ is a matter of debate. Reviews (Fombonne, 2009) suggest methodological differences between studies, referral patterns, diagnostic substitution, changes in the availability of services, the possible effects of migration into the area, and changes in public and professional awareness, make meaningful interpretation of prevalence rates and time trends across studies problematic.

There is evidence that a prime driver of diagnostic ‘upcoding’ is the coupling of ASD diagnoses to extra resource allocation from education and welfare services. In Queensland, specialist medical clinicians reported that, in the face of diagnostic uncertainty, they would provide an ASD diagnosis on a service provider’s form even when diagnostic criteria had not been met (Skellern et al., 2005). This practice occurred more frequently in regard to meeting Queensland’s access requirements for special education services. The authors caution, therefore, that numbers as high as 1 in 50 students should not be considered a true reflection of the ASD prevalence rate (Skellern et al., 2005).

Diagnostic upcoding has been implicated particularly amongst those with intellectual and learning disabilities, both in Australia (Skellern et al., 2005) and in Canada (Thivierge, 2008). However, it can occur amongst individuals of average or above average IQ as well.

Diagnostic instability

A systematic review of the diagnostic stability of ASD diagnoses found that whilst autistic disorder was relatively stable, Asperger’s disorder and PDD-NOS were highly unstable. There was large variation in outcome studies but some showed up to 53% of children with Asperger’s disorder and PDD-NOS later moved off the ASD spectrum (Woolfenden et al., 2012). It is unclear as yet whether collapsing the subcategories to a single ASD label as per DSM-5 will or won’t assist diagnostic stability.

Need for diagnostic clarity

As in most conditions in child psychiatry the clinical presentation of individuals with ASD is marked by complex comorbid conditions and challenging psychosocial issues. There are several conditions that can mimic the symptoms of ASD especially in the absence of true restricted and repetitive behaviour (RRB), which is one of the key features for an ASD diagnosis; this has been emphasized strongly in DSM-5. A major challenge in the future would perhaps be how to differentiate technological addictions from the RRB of ASD. As many individuals with other neurodevelopmental conditions like speech and language disorders with social phobia can indulge in excessive gaming and internet use, which in turn can worsen their social and communication skills and mimic the symptoms of ASD. This is a diagnostic and therapeutic challenge with the potential to become an even greater issue when it is likely that some conditions would be funded under the National Disability Insurance Scheme and other conditions that are equally distressing for the sufferers won’t.

Another issue in cases presenting with ASD features is the role of developmental trauma and insecure attachment (Alverez, 2004). This issue has been emotive and controversial given historical psychoanalytic theories that postulated a ‘refrigerator mother’ parenting style as a cause for autism. The paradigm shift to a more neurobiological perspective of ASD has been helpful but in light of the current ASD epidemic it seems the pendulum has swung too far. On the one hand there has been progress and benefits from a more neurobiological perspective in ASD which include reduction of parental guilt, focused evidence-based treatment, better understanding of the condition, better ways of dealing with the behaviour, etc. However, the potential disadvantages of a misdiagnosis or a narrow perspective in true cases are significant. An erroneous or too narrowly understood ASD diagnosis can have a self-fulfilling deleterious effect on a child’s psychosocial development via self, family, teacher, peer and others’ reduced expectations – the ‘Pygmalion effect’ (Batstra and Frances, 2012). The denial or ignoring of trauma, maltreatment and insecure attachment is widespread in society and even in the child psychiatric literature when that literature has a narrow DSM symptom focus (Parry and Levin, 2012).

Costs of over-diagnosis

Over-diagnosis creates the personal costs of: (i) stigma; (ii) reduced self and family expectations; and (iii) having to undergo unnecessary treatment and educational interventions. Symptoms of otherwise treatable psychiatric conditions can be conceptualized as a hallmark of ASD, potentially giving rise to therapeutic nihilism. The diagnosis can become a rationale to explain social withdrawal and justify continuing repetitive playing of online games. Social withdrawal can be further entrenched in the high functioning individuals who receive benefits from social services. The long-term future in such a situation might be an individual with a doubtful diagnosis of ASD, on a disability pension, socially isolated, spending hours on a computer immersed in virtual reality. This scenario must be borne in mind before diagnosing a high-functioning individual with ASD if there is any doubt.

The societal costs of over-diagnosis are the diversion of scarce educational, therapeutic and welfare resources away from those who most need them; and parental anxiety and confusion occasioned by the false ‘epidemic’. Whilst funding announcements geared to ASD continue, society tends to respond eventually to false diagnostic epidemics. The Victorian public education system now mandates a multidisciplinary assessment including speech assessment. ASD funding is no longer provided in the absence of severe speech impairment. In such a situation children with genuine Asperger’s disorder are likely to miss out.

Concluding remarks

Epidemiology, anecdotal reports and educational and service provider statistics reveal a growing ‘epidemic’ of ASDs. Changes to DSM-5 have partly been drafted to address over-diagnosis. However, the neo-Kraepelinian paradigm that underpins the DSM since DSM-III is part of the problem. Whilst symptom complexes can be more reliably described, individual developmental narrative and relative contribution of biopsychosocial contextual factors have generally been lost.

The checklist approach of the DSM has faced a barrage of criticism in recent years including an online petition from the American Psychological Association, British Psychological Society and 50 other professional organisations (American Psychological Association, 2012). In particular it can be argued that the DSM model neglects attachment theory, maltreatment effects and developmental psychology and is more problematic in child and adolescent psychiatry (Dignam et al., 2010).

We argue that the bestowing of a DSM label of ASD, however useful and at times justified, is not a risk-free intervention. Rather, the gold standard of assessment and therapy in such cases is a thorough and often lengthy process of engagement with the child, family and school, leading to a comprehensive biopsychosocial formulation that considers all developmental and contextual factors in a child’s life.

A tsunami of ASD diagnosed children and adolescents are entering adulthood. This will pose challenges for adult mental health services. Many young adults, maturing and moving out from their families, may find themselves being undiagnosed. ‘Undiagnosed’ may be welcomed by some but can be traumatic to an individual’s self-concept. This delicate issue warrants a careful consideration (Patfield, 2011).