Familial confounding on the ability to read minds: A co-twin control study.
Poor mind-reading scores may reflect family genetics more than autism itself, so check family history before blaming the diagnosis.
01Research in Context
What this study did
Isaksson et al. (2019) looked at twins to test if autism really causes poor mind-reading. They gave mind-reading tasks to identical and fraternal twin pairs. Some twins had autism, some did not.
The trick: compare each twin to their own co-twin. This wipes out shared genes and home life. If the autism twin does worse only when compared with strangers, family background—not autism—may be the real culprit.
What they found
Across all twins, kids with autism or more autistic traits scored lower on mind-reading tests. The gap vanished when they compared twin with twin inside the same family.
Result: the link between autism and poor mind-reading disappears once you control for family factors. The deficit may ride along with family genetics or environment, not with the autism label itself.
How this fits with other research
Zubizarreta et al. (2025) used the same co-twin trick on camouflaging stress. Camouflaging looked harmful in the general crowd, but twin-with-twin comparisons erased the danger. Both papers show that family confounding can fake an autism problem.
Jolliffe et al. (1999) found worse Strange Stories performance in autistic adults and took it as proof of a theory-of-mind deficit. Johan’s findings say that result could be driven by shared family traits, not by autism alone.
Kupzyk et al. (2011) showed clumsiness and autistic traits share 63 % of their genes in twins. This older twin study set the stage for Johan’s message: many autism-linked features are family-wide, not diagnosis-specific.
Why it matters
Before you write “poor perspective-taking due to autism” in a report, ask about the client’s siblings and parents. If the whole family struggles with social cues, target broader social-cognition training, not just autism-skills drills. Use tasks that separate individual skill from family style, and share this family context with teachers so they avoid autism-only explanations for social mistakes.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Ask parents to complete the same mind-reading task; compare their errors to the client’s to see if the pattern runs in the family.
02At a glance
03Original abstract
Alterations in social cognition are hypothesized to underlie social communication challenges in autism spectrum disorder. However, the etiologic underpinnings driving this association, as well as the impact of other psychiatric conditions on the association, remain unclear. Using a co-twin control design, we examined n = 308 twins (mean age = 16.63; 46% females) with autism spectrum disorder, attention-deficit/hyperactivity disorder, affective disorders, or typical development using the Reading the Mind in the Eyes Test to operationalize social cognition ability. Clinical diagnosis of autism spectrum disorder, as well as the extent of quantitative autistic traits, as measured by parental reports using the Social Responsiveness Scale-2, predicted fewer expected responses on the Reading the Mind in the Eyes Test across the pairs. The association remained when adjusting for other diagnoses and IQ. In addition, male sex, lower age, and lower IQ predicted poorer performance on the Reading the Mind in the Eyes Test. The associations between autism and social cognition ability were lost within pairs in both the full sample and the monozygotic subsample. We conclude that the association between autism and social cognition across the sample highlights the importance of social cognition alterations in autism spectrum disorder when compared with other conditions. The attenuation of the association in the within-pair models indicate familial confounding, such as genes and shared environment, influencing both autism and social cognition.
Autism : the international journal of research and practice, 2019 · doi:10.1177/1362361319836380