VOLUME 9 | NUMBER 10 | OCTOBER 2006 NATURE NEUROSCIENCE
CHILDHOOD DEVELOPMENTAL DISORDERS
Time to give up on a single explanation for autism
Francesca Happé, Angelica Ronald & Robert Plomin
We argue that there will be no single (genetic or cognitive)
cause for the diverse symptoms defining autism. We
present recent evidence of behavioral fractionation
of social impairment, communication difficulties and
rigid and repetitive behaviors. Twin data suggest largely
nonoverlapping genes acting on each of these traits. At the
cognitive level, too, attempts at a single explanation for the
symptoms of autism have failed. Implications for research
and treatment are discussed.
Autism, and autism spectrum disorders (ASD), such as Asperger
syndrome, are neurodevelopmental conditions diagnosed on the basis of
a triad of behavioral impairments: impaired social interaction, impaired
communication and restricted and repetitive interests and activities1.
These core diagnostic features, which reflect Kanner’s first reports of
autism with an emphasis on “autistic aloneness” and “insistence on
sameness”, mark out a highly recognizable and yet richly heterogeneous
group of children and adults2. Yet, despite half a century’s research into
ASD, there is little evidence regarding the unity of the three core areas of
impairment. Indeed, an early epidemiological study3, which set in place
the notion of the triad of impairments, remains the only full examina-
tion of this issue; these authors found some evidence of clustering of the
three impairments. The key importance of this issue is that researchers
have, for the last half century and with only a few notable exceptions4–6,
been searching for the causes and cures for autism as a whole. Research
on ASD at the behavioral, cognitive and genetic levels has proceeded
on the assumption that the three impairments that define autism must
be explained together. In this article, we question this basic assumption
and present evidence at each of these levels that suggests the triad of
impairments can be fractionated and should be studied separately.
Behavioral features of autism: integral or fractionable?
One of the challenges in establishing whether the triad of ASD features
requires a unitary explanation or is, instead, fractionable, lies in the
circularity of examining diagnosed populations. Because the diagnosis
of autism (and even Asperger syndrome) requires impairments in each
of the three key areas, examination of diagnosed populations cannot
establish the potential fractionation of the triad. One way through this
impasse is to explore the relationship between social, communicative
and rigid/repetitive traits in the general population.
Recent work suggests that autistic-like traits can be measured in
the general population7–9. We can ask, for example, whether a child
joins in playing games with other children easily, can keep a two-way
conversation going or likes to do things over and over again in the same
way all the time. The distribution of such traits supports a smooth
continuum (at least at the behavioral level) between individuals meeting
diagnostic criteria for ASD and individuals in the general population10.
Importantly, there is no evidence of a bimodal distribution, or ‘hump’
at the extreme, separating clinical from nonclinical levels of difficulty11.
We can therefore ask the question, do autistic-like social impairments,
communication difficulties and rigid/repetitive behavior cluster strongly
in the general population, suggesting the same underlying causes?
In our population-based studies, using data from over 3,000 twin
pairs assessed between the ages of 7 and 9, we have found modest-to-
low correlations between autistic-like behavioral traits in the three
core areas9. Somewhat to our surprise, even social and communication
impairments, which are often seen as almost indistinguishable in
real life and have been suggested to result from a single cognitive
deficit12, were only modestly related (with correlations (r) in the range
of 0.2 to 0.4). This relationship was no stronger than that between
communicative difficulties and rigid/repetitive behavior (r = 0.3–0.4),
whereas social impairments and rigid/repetitive behaviors were the
least strongly linked (r = 0.1–0.3). The modest correlations between
the three areas of autistic-like traits were found both across the gen-
eral population and when only children with relatively extreme scores
were considered13. It thus appears that, in middle childhood at least,
the degrees of social difficulty, communicative impairment and rigid/
repetitive behavior are only modestly related.
Indeed, within our large population-based sample, a considerable
number of children showed isolated difficulties in only one area of
the autistic triad. For example, 59% of children who showed social
impairments showed only social impairments. Around 10% of
all children showed only social impairment, only communicative
difficulties or only rigid and repetitive interests and behavior, and
these problems appeared to be at a level of severity comparable to
that found in children with diagnosed ASD in our sample. We did
find that children who showed one impairment were at increased risk
of showing a second or third area of autistic-like difficulty, but the
risks were relatively low, emphasizing the separability of the three
impairments. For example, only 32% of all children who showed
social impairments also showed communication difficulties. Thus,
although the three areas of autistic-like behavior and impairment did
occur together at above-chance rates, there was considerable evidence
for fractionation of the three aspects of the triad.
There is other evidence for the relative independence of the three core
features of autism. Studies of the development of children with autism
Francesca Happé, Angelica Ronald and Robert Plomin are at the Institute of
Psychiatry, Kings College London, De Crispigny Park, London SE5 8AF, UK.
Published online 26 September 2006; doi: 10.1038/nn1770
© 2006 Nature Publishing Group http://www.nature.com/natureneuroscience
NATURE NEUROSCIENCE VOLUME 9 | NUMBER 10 | OCTOBER 2006
suggest different developmental trajectories for different parts of the
triad. Rigid and repetitive behavior and interests, for example, emerge
later than social and communicative difficulties, are less good markers
of autism in infancy, improve less from infancy to early childhood14,
are poorly predicted from early measures of imitation or language15,16
and respond less well to some intervention programs17. There are some
measures, however, most notably the Social Responsiveness Scale (SRS),
which is heavily weighted toward the social domain, that do not show
fractionation of the triad at the level of behavior7,18. It is unclear at this
stage to what extent different findings are due to differences in measures.
Unlike diagnostic tools, for which ‘gold standard’ measures exist that are
used across studies, measures of autistic traits are still in development.
Genes ‘for autism’ or genes for independent traits?
Is the behavioral or phenotypic separability of the triad of autis-
tic-like traits mirrored at the genetic level; that is, are there separate
genes contributing to social impairment, communicative difficulties
and rigid/repetitive behavior? Our research suggests that the answer
is yes. Comparison of roughly 3,000 monozygotic and dizygotic twin
pairs at ages 7 and 8 years (Twins Early Development Study Institute
of Psychiatry) suggests that each aspect of the triad is highly heritable,
both across the range of individual differences and at the extreme9,13.
However, model-fitting analyses of cross-twin, cross-trait correlations
suggest that more than half the genes that contribute to variation in,
say, social (dis)ability are independent from those that contribute to
variation in communicative skills or rigid/repetitive tendencies9,19. Thus,
most of the genetic effects (at least in middle childhood) are specific,
acting on just one part of the triad.
This new conclusion from two studies using a large normative twin
sample fits with results from family studies of individuals with ASD.
Family and twin studies have shown that it is not only autism itself
that is heritable, but that relatives show increased rates of the “broader
autism phenotype”, which refers to subclinical manifestations of all
or part of the triad of autistic features. Importantly, some relatives
show only isolated traits, for example communication difficulties
without social impairment or rigidity20,21. This suggests that the genes
that contribute to autism segregate among relatives and have distinct
influences on the different parts of the phenotype.
No single explanation at the neurocognitive level
In the name of parsimony, cognitive accounts have traditionally
aimed to explain all three key features of autism. It is our belief,
however, that the accumulating evidence of behavioral and genetic
fractionation of the autistic triad is paralleled by a failure to find a
single cognitive account for the three core features of autism. Indeed,
we would argue that the evidence suggests we should abandon the
attempt to find a single cognitive explanation, in favor of good
accounts for each distinct aspect of the triad.
Current cognitive accounts of autism can be divided into those
that posit a primary deficit in social cognition (theory of mind12,
emotion processing22 or social orienting23) and those that posit a
primary deficit in nonsocial or domain-general processing (executive
dysfunction24, enhanced processing of local features25,26 or abnormal
attentional processes27). To date, no primary deficit has been proposed
that can plausibly account for the full triad of social, communicative
and rigid/repetitive difficulties28. ‘Social first’ accounts easily explain
why social and communicative difficulties might develop, but deriving
rigid/repetitive behavior as a downstream effect has proved unsatis-
factory. Repetitive behavior and interests do not appear to be simply
a reaction to an incomprehensible social world; they occur in high-
functioning individuals with ASD who have some social insight as well
as in lower-functioning ‘mind-blind’ people with autism, and they serve
self- stimulatory as well as calming functions29.
Nonsocial or domain-general accounts of autism, on the other
hand, struggle to explain why social insight and communication are
so particularly impaired, often alongside average or high intelligence
and good reasoning in nonsocial areas. A general appeal is often
made to complexity, with the claim that social interactions are
simply the hardest thing our brains have to process. But without a
metric for complexity, this claim seems empty; after all, computers
struggle to solve 3D segregation of objects in vision but need
little processing power to multiply large numbers. By contrast,
intellectually impaired children do not struggle to see in 3D, but may
never master multiplication. The distinction is presumably that our
brains evolved to solve some types of problems and not others; hence
some tasks are ‘easy’ and some ‘difficult’ for a typically developed
brain. Interpreting social stimuli may be one of the tasks for which
our brains are adapted, thus complexity explanations of specific
social impairments in autism seem unsatisfactory. Nor do the data,
such as they are, suggest that the degree of cognitive impairment
in, say, reading others’ minds, is strongly related to the degree of
detail-focus or repetitive behavior30. Instead, different cognitive
accounts appear to explain the distinct features of autism well, and
correlations can be found between specific test performance and
symptom severity within each area of impairment31,32.
Neuroimaging studies of autism also appear to offer support for the
independence of the cognitive substrates for social, communicative
and rigid/repetitive impairments. Imaging in healthy and ASD adult
volunteers suggests that social cognition relies upon a specific network
of brain regions, including the medial frontal cortex, temporoparietal
junction, superior temporal sulcus and temporal poles33. Lack of
preferential attention to speech in autism has been associated with
specific abnormalities in encephalographic (EEG) studies34, and lack of
activation of superior temporal sulcus voice-selective regions in response
to vocal sounds has been reported35. Rigid and repetitive behavior has
been linked to caudate abnormality in ASD36. Future developmental
neuroimaging will clearly be critically informative regarding the relative
independence of brain substrates for key cognitive functions.
In light of the above research, we suggest that it is time to give up on the
search for a monolithic cause or explanation for the three core aspects
of autism, at the genetic, neural and cognitive levels. Clearly a question
remains of why these three features co-occur at above-chance rates. At
the genetic level, although the majority of genes appear to be symptom
specific, there is evidence for a minority of overlapping genes between
domains. At the cognitive level, impairments in more than one domain
may interact; compensatory strategies may be reduced in the face of
multiple impairments. Given the widespread comorbidity generally
found in developmental psychopathology, what is most remarkable is
the extent of fractionation among the three core aspects of autism.
The implications of our argument are as follows. First, behaviorally it
would seem useful to measure the three aspects of the triad separately,
rather than rely on global ratings of autism severity, or ratings that focus
exclusively on social functioning.
Secondly, molecular genetic studies, which have resulted in little by
way of replicated linkage, should abandon the search for genes ‘for
autism’ as a whole. Instead, we suggest approaches that will allow
identification of genes contributing specifically to social, communica-
tive or rigid/repetitive traits, as we believe that the majority of genes
relevant to ASD will have symptom-specific action. Indeed, recent stud-
ies that have focused on subgroups within ASD, such as those showing
© 2006 Nature Publishing Group http://www.nature.com/natureneuroscience
VOLUME 9 | NUMBER 10 | OCTOBER 2006 NATURE NEUROSCIENCE
high levels of insistence on sameness37 or those with delayed onset of
phrase speech38, have shown stronger linkage signals.
Heterogeneity within the autism spectrum is perhaps the biggest
single obstacle to research at all levels39. A third implication is that
heterogeneity in ASD, on our account, is not simply due to noise or the
complex unfolding of development, but is an unavoidable consequence
of variation along at least three largely independent (although of course
interacting) dimensions of impairment.
Fourth, our argument and our findings within a large twin sample
suggest that there may be many individuals with isolated impairments
in one aspect of the triad, who do not meet diagnostic criteria for
any recognized disorder, but show difficulties of comparable severity
to those with autism. How we identify and meet the needs of these
children and adults is a key challenge for the future.
Lastly, if different features of autism are caused by different genes,
associated with different brain regions and related to different core
cognitive impairments, it seems likely they will respond to different
types of treatment. Abandoning the search for a single cause for a
single entity of autism may also mean abandoning the search for a
single ‘cure’ or intervention.
We would like to thank all the families who participate in Twins Early Development
Study, which is supported by Medical Research Centre grant G0500079 to R.P.
COMPETING INTERESTS STATEMENT
The authors declare that they have no competing financial interests.
Published online at http://www.nature.com/natureneuroscience
Reprints and permissions information is available online at http://npg.nature.com/
1. American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders
(American Psychiatric Association, Washington, DC, 1994).
2. Kanner, L. Autistic disturbances of affective contact. Nervous Child 2, 217–250
3. Wing, L. & Gould, J. Severe impairments of social interaction and associated abnor-
malities in children: epidemiology and classification. J. Autism Dev. Disord. 9, 11–29
4. Goodman, R. Infantile autism: a syndrome of multiple primary deficits? J. Autism Dev.
Disord. 19, 409–424 (1989).
5. Bishop, D.V. Autism, Asperger’s syndrome and semantic-pragmatic disorder: where are
the boundaries? Br. J. Disord. Commun. 24, 107–121 (1989).
6. Bishop, D.V. in Autism: Neural Basis and Treatment Possibilities 213–234 (Wiley,
7. Constantino, J.N., Przybeck, T., Friesen, D. & Todd, R.D. Reciprocal social behavior in
children with and without pervasive developmental disorders. J. Dev. Behav. Pediatr. 21,
8. Baron-Cohen, S., Wheelwright, S., Skinner, R., Martin, J. & Clubley, E. The autism-spec-
trum quotient (AQ): evidence from Asperger syndrome/high-functioning autism, males
and females, scientists and mathematicians. J. Autism Dev. Disord. 31, 5–17 (2001).
9. Ronald, A. et al. Genetic heterogeneity between the three components of the autism
spectrum: a twin study. J. Am. Acad. Child Adolesc. Psychiatry 45, 691–699 (2006).
10. Constantino, J.N. & Todd, R.D. Autistic traits in the general population: a twin study.
Arch. Gen. Psychiatry 60, 524–530 (2003).
11. Skuse, D.H., Mandy, W.P. & Scourfield, J. Measuring autistic traits: heritability, reliability
and validity of the Social and Communication Disorders Checklist. Br. J. Psychiatry 187,
12. Baron-Cohen, S., Tager-Flusberg, H. & Cohen, D. Understanding Other Minds:
Perspectives from Autism and Developmental Cognitive Neuroscience (Oxford Univ.
Press, Oxford, 2000).
13. Ronald, A., Happé, F., Price, T.S., Baron-Cohen, S. & Plomin, R. Phenotypic and
genetic overlap between autistic traits at the extremes of the general population.
J. Am. Acad. Child Adolesc. Psychiatry (in press).
14. Charman, T. & Swettenham, J. Repetitive behaviors and social-communicative impairma-
nets in autism: implications for developmental theory and diagnosis. in The Development
of Autism: Perspectives from Theory and Research (eds. Burack, J.A., Charman, T.,
Yirmiya, N. & Zelazo, P.R.) 325–345 (Lawrence Erlbaum, Hillsdale, New Jersey,
15. Charman, T. et al. Predicting language outcome in infants with autism and pervasive
developmental disorder. Int. J. Lang. Commun. Disord. 38, 265–285 (2003).
16. Lord, C. & Pickles, A. Language level and nonverbal social-communicative behaviors
in autistic and language-delayed children. J. Am. Acad. Child Adolesc. Psychiatry 35,
17. Aldred, C., Green, J. & Adams, C. A new social communication intervention for chil-
dren with autism: pilot randomised controlled treatment study suggesting effectiveness.
J. Child Psychol. Psychiatry 45, 1420–1430 (2004).
18. Constantino, J.N. et al. The factor structure of autistic traits. J. Child Psychol. Psychiatry
45, 719–726 (2004).
19. Ronald, A., Happé, F. & Plomin, R. The genetic relationship between individual differ-
ences in social and nonsocial behaviours characteristic of autism. Dev. Sci. 8, 444–458
20. Pickles, A. et al. Variable expression of the autism broader phenotype: findings from
extended pedigrees. J. Child Psychol. Psychiatry 41, 491–502 (2000).
21. Piven, J., Palmer, P., Jacobi, D., Childress, D. & Arndt, S. Broader autism phenotype: evi-
dence from a family history study of multiple-incidence autism families. Am. J. Psychiatry
154, 185–190 (1997).
22. Moldin, S.O. & Rubenstein, J.L.R. Understanding Autism: From Basic Neuroscience to
Treatment (Taylor & Francis, Boca Raton, Florida 2006).
23. Klin, A., Volkmar, F.R. & Sparrow, S.S. Autistic social dysfunction: some limitations of
the theory of mind hypothesis. J. Child Psychol. Psychiatry 33, 861–876 (1992).
24. Hill, E.L. Executive dysfunction in autism. Trends Cogn. Sci. 8, 26–32 (2004).
25. Happé, F. & Frith, U. The weak coherence account: detail-focused cognitive style in
autism spectrum disorders. J. Autism Dev. Disord. 36, 5–25 (2006).
26. Mottron, L., Dawson, M., Soulieres, I., Hubert, B. & Burack, J. Enhanced perceptual
functioning in autism: an update, and eight principles of autistic perception. J. Autism
Dev. Disord. 36, 27–43 (2006).
27. Belmonte, M.K. & Yurgelun-Todd, D.A. Functional anatomy of impaired selective atten-
tion and compensatory processing in autism. Brain Res. Cogn. Brain Res. 17, 651–664
28. Happé, F. Cognition in autism: one deficit or many? Novartis Found. Symp. 251, 198–
29. Turner, M. Annotation: repetitive behaviour in autism: a review of psychological research.
J. Child Psychol. Psychiatry 40, 839–849 (1999).
30. Pellicano, E., Maybery, M., Durkin, K. & Maley, A. Multiple cognitive capabilities/ deficits
in children with an autism spectrum disorder: ‘weak’ central coherence and its relation-
ship to theory of mind and executive control. Dev. Psychopathol. 18, 77–98 (2006).
31. Lopez, B.R., Lincoln, A.J., Ozonoff, S. & Lai, Z. Examining the relationship between
executive functions and restricted, repetitive symptoms of Autistic Disorder. J. Autism
Dev. Disord. 35, 445–460 (2005).
32. Briskman, J., Happé, F. & Frith, U. Exploring the cognitive phenotype of autism: weak
“central coherence” in parents and siblings of children with autism: II. Real-life skills
and preferences. J. Child Psychol. Psychiatry 42, 309–316 (2001).
33. Amodio, D.M. & Frith, C.D. Meeting of minds: the medial frontal cortex and social cogni-
tion. Nat. Rev. Neurosci. 7, 268–277 (2006).
34. Kuhl, P.K., Coffey-Corina, S., Padden, D. & Dawson, G. Links between social and linguis-
tic processing of speech in preschool children with autism: behavioral and electrophysi-
ological measures. Dev. Sci. 8, F1–F12 (2005).
35. Gervais, H. et al. Abnormal cortical voice processing in autism. Nat. Neurosci. 7, 801–
36. Sears, L.L. et al. An MRI study of the basal ganglia in autism. Prog. Neuropsychopharmacol.
Biol. Psychiatry 23, 613–624 (1999).
37. Shao, Y. et al. Fine mapping of autistic disorder to chromosome 15q11-q13 by use of
phenotypic subtypes. Am. J. Hum. Genet. 72, 539–548 (2003).
38. Buxbaum, J.D. et al. Evidence for a susceptibility gene for autism on chromosome 2 and
for genetic heterogeneity. Am. J. Hum. Genet. 68, 1514–1520 (2001).
39. Newschaffer, C.J., Fallin, D. & Lee, N.L. Heritable and nonheritable risk factors for autism
spectrum disorders. Epidemiol. Rev. 24, 137–153 (2002).
© 2006 Nature Publishing Group http://www.nature.com/natureneuroscience