Till startsida
University of Gothenburg
To content Read more about how we use cookies on gu.se

Overdiagnosis and underdiagnosis of autism spectrum disorders -- causes and effects

Lena Nylander is a Chief Physician and specialist in general psychiatry, researcher at GNC and our interviewee in this edition of the Researcher’s Corner. In 2011 she put forward her dissertation on ASD and ADHD in adult psychiatric patients, at the University of Gothenburg. Her research is primarily concerned with autism spectrum disorders (ASD) and other cognitive developmental deviations and how these can manifest in adult psychiatric patients. In this interview she talks about causes and effects of overdiagnosis and underdiagnosis of ASD.

You stress that an inadequate diagnostic process can result in both underdiagnosis and overdiagnosis. Can you give us some examples of this?

- What I see as a physician at a unit for adult psychiatry is primarily an overdiagnosis of autism with normal intelligence/Asperger syndrome. Professionals within adult psychiatry seem to have a tendency to think of ASD as soon as a patient shows any kind of problem with social interaction. Perhaps this is partly due to Asperger syndrome having become somewhat of a ”trendy diagnosis” – there are films and books about individuals, who are sometimes explicitly said to have ASD, and sometimes, are thought by readers to ”have” ASD. But these characters are fictional, and can therefore have symptoms and traits, which are both compatible and incompatible with ASD. Additionally, symptoms of Asperger syndrome in the eyes of the public have become almost synonymous with high intelligence and even ingenuity. There is a tendency to think that ability to learn by heart and good memory for facts equal high intelligence, something that we know is not the case. The Asperger syndrome diagnosis has a higher ”status” than most psychiatric diagnoses. Among psychiatric staff there is still some reluctance to giving a diagnosis of for example schizophrenia, since this diagnosis is considered more stigmatising than an ASD-diagnosis.
- You then forget, or choose to overlook, that virtually all patients with mental illness have some form of difficulties with social interaction, at least while they are ill. Regarding for example personality disorders or other conditions with long-term courses the difficulties can have started relatively early on, but it is still not autism. The criteria for deficits in social interaction are rather vague and broadly defined in DSM. This causes confusion as to the nature of the described deficit, which is a fundamental deficit in the ability/understanding of social interaction rather than a deficit in the execution of/desire for social interaction.

- The other conditions that I find are often prevalent in those patients thought to have ASD within adult psychiatry are ADHD, borderline intellectual functioning, personality deviations and psychoses. But I have also encountered for example depressions (where “the Asperger syndrome disappeared with electroconvulsive therapy”!), social phobia and anorexia. The tricky part of course is that individuals with ASD can have, and often do have, these conditions as well. Extremely high intelligence in a young person has also potentially been misdiagnosed as ASD. For a very highly intelligent person the social functioning in a surrounding where the general level is lower is of course not easy, (for example in a small school). There will be cognitive dissonance until the context matches the individual and the person can fit in and find friends on his/her level, for example at the university.

- Underdiagnosing is something I see among adults with an intellectual disability, where, in some cases, professionals have settled for that particular diagnosis without paying closer attention to whether autism too could be an appropriate diagnosis. In some cases this underdiagnosis has resulted in a failure to adequately understand the person with intellectual disability, who in turn has not got the pedagogical support and interventions, which might have contributed to making his/her life easier. Sometimes such shortcomings have led to challenging behaviours, sometimes to the person getting large doses of several different types of medicine.

- Old people constitute another group that is probably underdiagnosed – statistically there should be more elderly people with ASD than the rather few individuals who so far have got this diagnosis. Among older people with ASD there are surely also un-met support needs. For example some of them probably have difficulty getting access to adequate health care, due to lacking ability to assess their own condition and difficulty knowing when and how to contact health services.

You have written about both overdiagnosis and underdiagnosis of autism in adults, advocating a diagnostic process based on a thorough medical work-up rather than, for instance, self-assessment questionnaires only. Why is this important?

- I am sceptical about self-assessment questionnaires, especially in regard to ASD, where the individuals often have great difficulty assessing their own behaviour. On numerous occasions I have met individuals with clear and classical ASD, who have considered their own behaviour to be completely normal (which it is to them of course) and assessed their own level of function to GAF 90 - 100. I have also encountered a number of individuals with psychopathy, who achieve high scores on self-assessment scales – they know how to fill out the forms as well as how to behave in conversation to get the desired diagnosis. To realise that the person in front of you does not have Asperger syndrome but instead is a case of psychopathy with high intelligence and very good mentalising ability is virtually impossible unless you have two experienced professionals (who are familiar with psychopathy) interviewing at the same time.

What problems do you identify as causes of misdiagnosing of ASD? How can the risk of underdiagnosis and overdiagnosis be decreased?

- I think a problem with the diagnosing of ASD is relying too much on questionnaires and isolated tests, and inexperienced people conducting “assessments” of ASD as solitary work. The diagnostic criteria are formulated in a very broad way in ICD and DSM, leaving a lot to the clinician to figure out. Hence the diagnostic process requires both substantial knowledge and experience. And the only way of gaining knowledge is to work for an extended period of time and to see a lot of patients/clients with ASD. Within adult psychiatry it is still not common practice to take a careful developmental and medical history or to involve family and the patient as collaborative partners in assessing the problems that the patient experience in daily life. Also, members of staff are not accustomed to thinking in terms of “input” – what does the patient understand and experience – rather than “output” – symptoms and behaviours.
- I think the best way of getting it as right as possible is to always have two people – a doctor and a psychologist – performing the assessment together, i.e. both an interview and an assessment of the patient. This readily takes place as a conversation in a ”team” consisting of the patient, his/her family (if possible) and the two professionals, at least one of whom should be experienced. Through a thorough medical work-up the clinical need is identified and appropriate support measures can be given. The best aids in such a work-up in my opinion are semi-structured interviews like the DISCO and cognitive tests that provide a broad overview of strengths and weaknesses and at the same time an opportunity to observe the patient’s cognitive style, for example WAIS. When conducting research, the diagnostic process is somewhat different, but in a way more uncertain. All psychiatric diagnostic processes contain a more or less significant portion of uncertainty, and part of professionalism is to handle this.

Can you name some potential consequences of overdiagnosing ASD?

- Overdiagnosing has consequences of creating tighter boundaries when it comes to who should be eligible for support – resources are not enough for everyone, and thus rules limiting who is eligible for support are established. There is also an apparent risk of ”watering down” of the concept – the notion of ”everyone having Asperger syndrome”, even well-functioning individuals whose surroundings cannot detect and perhaps have never noticed any form of disability, and then it becomes something that we don’t need to pay attention to. Furthermore, in a long-term perspective, it is frightening if all deviations from the mean are to be diagnosed and pathologised – ”the normal range” must be allowed to be broad and varied. Variation is not pathology, but what has spurred the evolution.

Are there other factors than the diagnostic process itself that affect how often a diagnosis of ASD is given?

- One factor influencing the risk of overdiagnosing is the difficulty to get access to support without a diagnosis, or, more accurately, without a certain diagnosis. In the Swedish Act concerning Support and Service for Persons with Certain Functional Impairments (LSS), the right to support is clearly stated – if you are in need of it! – if you have autism or autism-like conditions. The municipalities’ interpretation of this statement makes a person with an autism diagnosis more likely to receive support and interventions than a person with for example schizophrenia, regardless of whether the person with schizophrenia has the same or more extensive needs.

By Nanna Gillberg

Photo: Lena Baravägen

Page Manager: Anna Spyrou|Last update: 2/26/2015

The University of Gothenburg uses cookies to provide you with the best possible user experience. By continuing on this website, you approve of our use of cookies.  What are cookies?