Home > News & Opinion > In Brief > 2013 > Clinical research: Attention deficit forecasts autism traits

Clinical research: Attention deficit forecasts autism traits

Popularity tracker
Laura Geggel
5 February 2013

Twin troubles: Communication problems in identical and fraternal twins are linked to attention problems at age 8 and to autism at age 12.

Children with attention difficulties early on may later show signs of autism, such as trouble holding a conversation, says a new study published 14 November in Psychological Medicine1.

The study, a survey of typically developing children, is potentially important because it suggests that treating early signs of attention deficit hyperactivity disorder (ADHD) may prevent autism traits, the researchers say.

More than 70 percent of people with autism also meet the diagnostic criteria for another psychiatric or developmental disorder, and ADHD is one of the most common. A number of studies have shown that individuals with autism and those with ADHD share certain mutations.

In the new study, researchers surveyed parents of more than 5,000 pairs of identical and fraternal twins in the U.K. about traits associated with ADHD and autism. Parents used the Childhood Autism Spectrum Test (CAST) and the Conners’ Parent Rating Scale for ADHD traits first when their twins were 8, and then four years later when they were 12. The study focused on this age bracket because the brain changes as children transition from childhood into adolescence.

Using the survey data, the researchers compared the traits for autism and ADHD in each twin pair to assess their genetic and environmental similarities. If both conditions have a common genetic basis, identical twins would be expected to share more traits of the disorders than fraternal twins do.

In a 2008 report on the twins at age 8, the researchers found that identical twins share more traits of autism and ADHD than fraternal twins do2. In the new study, the researchers again found that identical twins have more traits of both disorders in common.

They then analyzed data from both ages to see how traits of autism and ADHD influence one another over time.

At age 8, communication difficulties in identical and fraternal twins were often linked to attention problems, such as interrupting others or not listening when people spoke to them. Four years later, most of these children’s communication challenges were more reflective of autism than of ADHD.

The researchers also found that traits of autism can later lead to problems with attention, such as being easily distracted, although this trend is less significant. 

News and Opinion articles on SFARI.org are editorially independent of the Simons Foundation.


1: Taylor M.J. et al. Psychol. Med. Epub ahead of print (2012) PubMed

2: Ronald A. et al. J Child Psychol. Psychiatry 49, 535-542 (2008) PubMed


,  ,  ,   


Name: usethebrainsgodgiveyou
5 February 2013 - 6:03PM

You are forgetting Dyslexia, which is the broadest diagnosis to fit under the umbrella of "behavioral impairment" that is also often a communication impairment. I firmly believe 20% of the population is under possible scrutiny for behavioral impairment. It is related to right brain dominance, and about as "brilliant" of the specialists to earn a living trying to change it as it was of the experts 100 years ago who attempted to change the sinister child by tying his left hand behind his back...Not that I'm bitter or anything, but mostly amused.

Name: Anonymous
5 February 2013 - 8:18PM

As someone with autism who struggles with inattentiveness, I find these findings very interesting. I know the findings are traits, so I don't know how far they generalise to people like me, but I find that inattentiveness tends to interfere with my ability to pursue my interests. I have read the paper, and noticed that they find weak associations between so-called 'repetitive, restricted behaviors and interests' and thr two ADHD scales.

Does this article not misrepresent the paper slightly? I've read it, and I don't recall the authors claiming that treating ADHD will decrease later autism traits - in fact, they seem speculative and cautious about drawing strong conclusions, perhaps because these are traits, not diagnosed people.

Name: Laura Geggel
5 February 2013 - 10:05PM

Hi Anonymous,

Thanks for your comment. You are correct — the authors did not mention that treating ADHD may decrease later autism traits in the paper. However, when I spoke with one of the authors during my research, she mentioned it as a possibility.

Name: Anonymous
6 February 2013 - 5:11PM

I see - thanks Laura for this.

While there are obviously challenges associated with being autistic, I don't feel that a treatment or support with these challenges that then reduced autism would be good for me. Being autistic has given me many advantages as well as difficulties.

Still very interesting findings though, hope researchers can do more on co-occurring symptoms which can be hard to live with.

Name: Robert Jensen
6 February 2013 - 3:02PM

The Childhood Autism Spectrum Test (CAST) has produced some unusual findings in the past from this group.

In a 2006 study sample by the same group there were no sex differences in extreme model fitting for extreme BAP traits in general population MZ twins as measured by various CAST cutoffs. The was no sex differences in concordance rates for extreme BAP traits seen in opposite sex dizygotic (DZ) twin pairs in contrast to male/male DZ twin pairs using various CAST cutoff scores. Those findings stand in sharp contrast to the well-recognized high male female ratio in strictly diagnosed autism. Scroll down and see table 2 in the following PDF.


If there is no or little sex differences in extreme BAP traits as measured by CAST scores in contrast to strictly diagnosed autism with a high male female ratio, the implication is that extreme BAP traits are not a causal mechanism but rather is a background genetic effect that is always reliant on other genetic mechanisms (inherited or de novo), environmental risk factors and/or epigenetic events that follows a developmental trajectory to a strict autism diagnosis.

Name: Robert Jensen
6 February 2013 - 7:51PM

Skuse et al (2005) also found no sex differences in extreme autistic-like traits as measured by the Social and Communication Disorders Checklist (SDSC). 'Earlier analyses of these data showed no significant difference in the size of genetic influence in males and females and no evidence of separate non-additive genetic effects in females'.


Several diagnostic checklists used to meaure the heritabilty of autistic-like traits in general population trwins such as the ASQ and SRS may have a male gender bias and the diffulty in diagnosing autism and autistic-like traits is well recognized:


Name: Robert Jensen
6 February 2013 - 8:05PM

Finally, Robinson et al(2011) using the Social and Communication Disorders Checklist (SCDC) reported that there was no age effect in extreme autistic-like traits in the general population. Neither maternal nor paternal age was associated with extreme scores. This is inconsistent with the well recognized association of advancing paternal and maternal age and increased autism risk.


Name: Angelica Ronald
6 February 2013 - 4:51PM

Reply to Robert Jensen.
Thank you for your interest in our research. I agree that careful consideration of the psychometric properties of all measures is crucial. Just to clarify the point you make about our measure of autistic traits, the CAST. In fact the CAST shows higher male than female scores in the general population at large. The first paragraph of the results from the linked paper reports this. Furthermore, the proportion of males increases gradually with increasing scores on the CAST, as would be expected for a measure of autistic traits -- see Table 1 of the linked paper.

Table 2 presents the results from the twin modelling analyses. Differences here between male and female MZ and DZ groups would indicate the extent to which there are quantitative sex differences in the genetic and environmental influences on autistic traits. That is, are the estimates of heritability and shared and nonshared environmental influences different for males and females. In the literature to date, there is a mixed picture about whether heritability and environmental estimates are significantly different for males and females for autism and autistic traits (reviewed by Ronald & Hoekstra, 2011). You wouldn't necessarily expect them to be different even for a trait or disorder which shows a significant mean sex difference. But another challenge about comparing heritability and environmental estimates for males and females separately is that it is often difficult to have enough power in analyses to do this, even with very large sample sizes or for clinical samples.

As such, the data on mean sex differences from the CAST supports the hypothesis that there might be a continuum between autistic traits and diagnosed ASD. It does not prove it, but it is supportive. I won't respond further on this topic and I would like this to leave space in this comment section for discussion of the topic of the article (Taylor et al paper).

Add a Comment

You can add a comment by filling out the form below. Plain text formatting.

Question: What is 10 + 4 ?
Your answer: