Anorexia nervosa and autism: a prospective twin cohort study

Matt Kempen
Marketing Manager for ACAMH

Posted on

Video abstract from Lisa Dinkler on ‘Anorexia nervosa and autism: a prospective twin cohort study’.

Authors: Lisa Dinkler, Mark J. Taylor, Maria Råstam, Nouchine Hadjikhani, Cynthia M. Bulik, Paul Lichtenstein, Christopher Gillberg, Sebastian Lundström
First published: 04 June 2020

Anorexia nervosa (AN) and autism spectrum disorder (ASD) may be phenotypically and etiologically linked. However, due to the absence of prospective studies, it remains unclear whether the elevation of autistic traits in AN is evident in early childhood. Here, we prospectively investigated autistic traits before and after the first diagnosis of AN.

Lisa Dinkler
Lisa Dinkler

Lisa Dinkler is a psychologist who studied in Jena and Frankfurt am Main, Germany, and a full-time PhD student at the Gillberg Neuropsychiatry Centre. Her PhD project investigates the development and course of eating disorders in association to neurodevelopmental disorders (NDDs), and the mechanisms underlying the overlap of eating disorders and NDDs. Applying twin methodology Lisa studies the extent to which genetic and environmental factors are shared between eating disorders and NDDs. In a large birth cohort of Japanese preschool children she investigates the prevalence and course of a wide range of childhood eating problems, including Avoidant Restrictive Food Intake Disorder (ARFID), and their association with NDDs. Lisa’s research also explores neuropsychological factors underlying the social difficulties common to eating disorders and NDDs with the help of eye-tracking methodology.

Bio via University of Gothenburg


Hi, my name is Lisa Dinkler and I’m a PhD student at the University of Gothenburg in Sweden. Today, I will give you a brief summary of a study my colleagues and I recently published in the Journal of Child Psychology and Psychiatry, titled ‘Anorexia nervosa and autism, a prospective twin cohort study’.

Anorexia and autism have been reported to co-occur more than expected by chance. While the prevalence of autism in the general population is about one per cent, studies have found a prevalence of up to 50 per cent in individuals with anorexia. As autism is a neuro developmental disorder, it needs to be assessed whether symptoms have been present from early childhood. Some studies have done so retrospectively. That means, when assessing autism in adolescents or adults with anorexia, they also collected parental reports of autistic traits in childhood. As retrospective reports are prone to recall bias, prospective designs are generally preferable. We therefore wanted to see if it would be possible to capture the link between anorexia and autism prospectively by measuring autistic traits already in childhood and then following up these individuals for later diagnosis of anorexia.

We have used data from almost 6,000 individuals from the Child and Adolescent Twin Study in Sweden. Autistic traits were measured by parent report at age nine, and then again at age 18, to ascertain diagnosis of anorexia. We retrieved diagnosis from the Swedish National Patient Register, and we also used parent reports of treatment for anorexia.

We then compared whether individuals with and without anorexia differed in autistic traits before the first diagnosis for anorexia at age nine and after the first diagnosis of anorexia at age 18. Assuming that there is an increased prevalence of autism and anorexia, so that there is a sub-group that has both, we expected to see some elevation of childhood autistic traits in individuals with later anorexia.

We found that children later diagnosed with anorexia did not show elevated autistic traits at age nine compared to children who did not develop anorexia. In the figure, autistic traits are split into the two domains, social communication problems in blue and restrictive and repetitive behaviour and interests in yellow. You can see that at age nine, all confidence intervals of the estimated increase in risk overlap with the baseline of one, which indicates no increase in risk. At age 18, however, both social communication problems and restrictive and repetitive behaviour and interests were significantly increased in girls with anorexia. In boys with anorexia, restrictive and repetitive behaviour and interests but not social communication problems were significantly increased.

We then split the girls with anorexia into two groups. One group with ongoing acute anorexia during the assessment of autistic traits at age 18, and one group who previously had anorexia. We did not split the group of boys with anorexia as it was already very small. What we saw then was that the elevation of autistic traits at age 18 was confined to the sub-group with acute anorexia, while girls with previous anorexia had the same level of autistic traits as girls who never had anorexia.

These results were surprising at first sight as they seem to contradict previous findings of an over-representation of autism in anorexia. There are, however, many possible explanations. For example, successful camouflaging behaviour of children with autism and anorexia might have contributed to that we did not find elevated autistic traits in children who later developed anorexia. Camouflaging describes coping strategies to socially fit in by either hiding behaviours associated with autism or purposely performing behaviours deemed to be neuro typical. Camouflaging can be very demanding and might, together with other problems in adolescence, such as the emergence of anorexia, lead to a depletion of resources, which in turn exacerbates autistic traits and makes them more noticeable. Especially during acute anorexia, when brain functioning is severely affected by starvation, the ability to camouflage might be impaired, leading to an exacerbation of the previously subtle autistic traits. After weight recovery, the ability to camouflage might be restored, however. And this might also explain why, at age 18, we only found increased autistic traits in girls with acute anorexia, but not in those with previous anorexia.

A second but related explanation for our findings is that girls show a somewhat different autism phenotype and their autistic traits are often more subtle and more difficult to detect with the existing assessment instruments that were developed based on the male autism phenotype. Previous studies also show that subtle social difficulties in girls only become more overt in adolescence where the social context become more complex. These factors might have further contributed to that we did not find elevated autistic traits in children who later developed anorexia.

Our results imply that the screening assessment of autism by parent report may not be sufficient to detect potential elevations of autistic traits in girls who later develop anorexia, and that we need more specific instruments to capture autism in young girls, especially the cognitively-able ones. We also need to be extremely careful with the assessment of autism in the acute phase of anorexia, and collecting information on autistic traits in the early development is indispensable.

Thank you for watching this video abstract. And if you want to know more about this study, you can find it online. It is published with open access in the Journal of Child Psychology and Psychiatry.

Add a comment

Your email address will not be published. Required fields are marked *