PREPRINT
Waterhouse L. Autism: Two symptoms among many. OA Autism 2014 Jul 18;2(2):13.

Lynn Waterhouse, PhD
Professor, Global Graduate Programs
Director Emeritus, Child Behavior Study
The College of New Jersey, USA
Email: lynwater@tcnj.edu
Abstract
Introduction
Autism is diagnosed by the criteria for Autism Spectrum Disorder (ASD): impairments in social interaction and communication, and repetitive and restricted behaviours. Although for over seventy years autism was seen as one disorder, autism as ASD is currently understood to comprise an unknown set of subgroups called “the ASDs.” This paper argues that discovering the unknown set of biologically valid ASD subgroups requires studying ASD symptoms separately because four lines of evidence demonstrate that the two core ASD diagnostic symptoms do not identify a bounded entity. These four lines of evidence are briefly reviewed here. One, all genetic and environmental ASD risk factors are found with non-ASD syndromes and non-ASD symptoms, and no validating unique link has been found between an ASD risk factor and just the core ASD diagnostic symptoms. Two, the ASD diagnosis lacks neurobiological validity because many varied brain dysfunctions have been found with the ASD diagnosis. Three, the two ASD diagnostic criteria form two factors and have no consistent empirical association with one another. Four, each of the two core ASD criteria alone has been significantly linked to many varied brain dysfunctions.

Conclusion
This brief review of these four lines of evidence concludes that discovering subgroups of valid etiopathophysiologies requires exploring core ASD diagnostic symptoms separately in samples not defined by the ASD diagnosis. Finding biologically valid subgroups of neurodevelopmental social impairment will depend on studying samples defined by an ASD risk factor, or an ASD brain dysfunction, or defined by a specific risk factor linked to a specific brain dysfunction.

Introduction

For over seventy years autism was seen as one disorder.1,2 Autism is currently defined by the diagnostic criteria for DSM-5 Autism Spectrum Disorder (ASD).3 No unitary neurobiological basis for ASD has been found, 4,5,6 making behavioural intervention the only consistently effective treatment.7 Between 200 and 1,000 gene variants,8,9 and 35 to 60 environmental risk factors10,11,12,13 have been identified as possible contributing causes for ASD. Although most heritability estimates suggest that genetic risk factors predominate, two recent studies reported that genetic and environmental risk factors each accounted for approximately 50% of ASD variance.14,15

The vast number of ASD risk factors, 8,9,10,11,12,13 the extensive heterogeneity within any ASD sample,16,17,18 the shared risk factors between ASD and other psychiatric and medical diagnoses, 8,9,10,11,12,13,19,20 the ballooning prevalence of ASD,21,22 and the missing ASD neurobiological validity,5,6,16,17,18,23,24,25,26 together have led to a general understanding that the ASD diagnosis includes multiple subgroups, often referred to as “the autisms” or “the ASDs.”16,17,18, 23,24,25,26, 28 Parellada et al.23 stated that autism is now seen more as a “phenotypic endpoint of multiple deviant developmental pathways than a disorder” (p. 16). Williams et al.32 proposed that “as understanding of the causes of ASD advances, the collective term will be replaced over time by subgroups that are clearly defined…. by genetic and/or neurological…findings” (p. 5).

Discussion

This paper argues that four lines of evidence demonstrate that the ASD diagnosis is not a sound basis for discovering biologically valid subgroups of neurodevelopmental social impairment. First, no ASD risk factors have been shown to produce only ASD diagnostic symptoms.16,17,18,23,24,25,26,28,32 Second, ASD diagnostic symptoms are found with a wide range of brain dysfunctions, 5,6,16,17,18,23,24,25,26 thus the ASD diagnosis lacks neurobiological validity. Third, ASD diagnostic criteria form two factors and have no consistent behavioural associations.26,27,28,29,30,31 Fourth, each of the two core ASD diagnostic criteria– impairments in social interaction and communication, and repetitive and restricted behaviours—are found with many varied brain dysfunctions.16,17,18,23,24,25,26

No ASD risk factors have been shown to produce only ASD diagnostic symptoms

Although much evidence suggested ASD was essentially a genetic disorder, 2,8,9,19,20 two recent studies reported that genetic and environmental risk factors contributed equally to ASD.14,15 Butler and colleagues33 identified 629 genes linked to ASD, and more than 40 distinct environmental factors have been found to confer risk for ASD.10,11,12,13 However, the total number of ASD risk factors continues to rise because research continues to discover new genetic and environmental risks for autism symptoms.8,9

How genetic and environment risk factors mediate ASD symptoms is not yet well understood. Despite the evidence that over 600 gene variants have been found with ASD symptoms, 33,34 risk variants have only been identified in approximately 20% of individuals with ASD, 8,9 and no gene variant “accounts for more than 1% of ASD cases—a pattern consistent with extreme genetic heterogeneity among cases” (p. 75). 9

Froehlich-Santino and colleagues35 reported that an ASD twin study yielded no evidence for any interaction between genetic risk and the identified pre-and peri-natal risk factors, respiratory distress, jaundice, needing oxygen, and hypoxia. Nonetheless, it is likely that many gene-environment interactions contribute to the risk for autism, including advanced parent age causing gene mutations in sperm and ova, and foetal gene mutations conferring vulnerability to specific environmental insults.10,11

The range of environmental risk factors linked to ASD is extensive. A meta-analysis11 identified 16 significant pre- and perinatal risk factors for ASD, such as foetal distress, low birth weight, meconium aspiration, ABO or Rh incompatibility, and hyperbilirubinemia. Grabruker10 reviewed 15 other significant environmental risk factors for ASD, including maternal drug use, maternal influenza, rubella, and cytomegalovirus infections and other maternal diseases, as well as advanced paternal and maternal age.
ASD genetic and environmental risk factors are linked to similar non-ASD syndromes and non-ASD symptoms. Jeste and Geschwind 9 observed that genetic risk factors for ASD yielded many non-diagnostic symptoms including intellectual disability, seizures, sleep disorders, and motor impairments. Grabruker10 reported that environmental risk factors for ASD resulted in a similar set of non-diagnostic symptoms, including intellectual disability, seizures, sleep disorders, gastrointestinal problems, and motor impairments.

Most importantly, all ASD risk factors yield ASD symptoms and non-ASD symptoms, including attention disorders, language disorders, specific cognitive impairments, intellectual disability (ID), hearing problems, vision problems, motor problems, psychiatric disorders, seizures, atypical EEG findings, sleep disorders, eating disorders, and other medical disorders.15, 9,10,16,17,32 The standard explanation for non-ASD symptoms is that they result from disorders comorbid with, but distinct from ASD, such as epilepsy, ID, or ADHD. Individual non-ASD symptoms, such as hypotonia, are also explained as extrinsic to ASD. However, contrary to this standard assumption, State and Levitt34 proposed that many individual genetic risk factors for ASD must have a wide variety of effects on brain development because ASD genetic risk factors also result in other disorders. They asserted, “the observation that identical mutations may lead not only to ID (intellectual disability) and ASD, but to schizophrenia and possibly other neuropsychiatric disorders …would seem to argue for a model based on the pleiotropy of genes underlying fundamental neuronal processes” (p. 257).34

The many varied non-ASD symptoms and non-ASD syndromes consistently found with ASD genetic and environmental risk factors16,17,18,19,20,24,25,26,32 argue against the assumption that ASD is a bounded disorder of a natural convergence of just the two ASD diagnostic symptoms. Instead, the evidence that ASD genetic and environmental risk factors are consistently found with a combination of one or both of the two ASD criterial symptoms and a varied set of non-ASD symptoms16,17,18,19,20,24,25,26,32 demonstrates only that social impairment and repetitive and restricted behaviours are each symptoms with a wide cross-aetiology prevalence. As illustrated in figure 1, many symptoms of impaired neurodevelopment such as ID, ADHD, epilepsy, and motor impairments occur along with ASD diagnostic symptoms.

[FIGURE1 below references]

The ASD diagnosis occurs with varied brain dysfunctions and thus lacks neurobiological validity

ASD symptoms have been found in individuals with varied atypical brain chemistry and with varied atypical brain structures.5,6,16,17,18,23,24,25,26 This variation has included atypical levels of many different neurotransmitters, and varied brain structure impairments including atypically larger heads, smaller heads, atypically increased and decreased brain volume, atypical grey and white matter, with abnormalities in brainstem, cerebellum, amygdala, hippocampal, hypothalamus, cingulate cortex, corpus callosum, insula, fusiform gyrus, caudate and thalamus abnormalities, and with atypical organization or deficit in frontal, parietal, temporal, and occipital lobe regions. 5,6,16,17,18,23,24,25,26

Although no unitary brain dysfunction basis for ASD has been found, 4,5,6,16,17,18,23,24,25,26 hundreds of single brain dysfunction models of ASD have been proposed.16,17,24 For example, Stoner and colleagues36 reported neuropathology findings for discrete patches of abnormal laminar organization in prefrontal and temporal cortex in 10 of 11 ASD case samples but in only 1 of 11 control samples. On the basis of these data, Stoner et al.36 proposed that autism symptoms result from “macroscopic early brain overgrowth in the majority of cases” (p. 1210) that causes cortical patches of abnormal laminar cytoarchitecture and disorganization resulting from the “dysregulation of layer formation and layer-specific neuronal differentiation at prenatal development” (p. 1218).
Given the evidence that the ASD diagnosis is found with many varied brain impairments, the Stoner et al.36 atypical cortical patches model cannot provide a unified account of autism brain dysfunction. For example, isolated cerebellar deficit or isolated corpus callosum deficit are inconsistent with the cortical patches model.16,17,24,25
If ASD diagnostic criteria are understood to be two symptoms among many symptoms of neurodevelopmental impairment, there is no unitary ASD disorder, and thus, there is no need to find a single brain dysfunction to define ASD as a disorder.

ASD criterial symptoms have no consistent empirical association with one another and occur separately in individuals

The two ASD diagnostic criteria have not been consistently correlated. Harrop and colleagues29 found no association in an ASD sample between total restricted and repetitive behaviours and three standard social interaction impairment scores, or with two standard communication impairment scores. By contrast, however, Frazier et al.27 reported that social interaction impairment was significantly intercorrelated with insistence on sameness and repetitive mannerisms. Shuster et al.30 reviewed 36 factor analytic studies of autism symptoms in diagnosed individuals and concluded that autism social impairment symptoms were a separate factor from a factor that included restricted and repetitive behaviors.

In typically developing young children, severe social impairment is rare but restricted and repetitive behaviours are common.29 This disparity may be why Happé et al.28 reported finding only weak correlations between autism social impairment symptoms and repetitive and restricted behaviours in typical 7- and 8-year-old twins. Restricted and repetitive behaviours are understood to support typical developmental mastery of sensory-motor skills, and the rate of restricted and repetitive behaviours in typical toddlers can equal the rates found for children with ASD.29

Not all individuals who have been clinically diagnosed with autism have met the DSM-5 ASD criteria for restricted and repetitive behaviours.37,38 Kulage et al. 37 studied DSM-5 ASD diagnostic coverage and reported finding a subset of individuals who met the DSM-5 ASD criteria for social impairment but who did not express restricted and repetitive behaviours.37 Another comparison of DSM-5 ASD diagnostic coverage with earlier autism diagnoses conducted by Kim et al.38 found that individuals with a prior autism diagnoses who did not meet both DSM-5 ASD criteria, however, did meet criteria for social impairment but nonetheless, did not express restricted and repetitive behaviours.
These behavioural data taken together do not support a natural coherence of the two diagnostic criterial symptoms of ASD.

Each of the two core ASD diagnostic criteria is linked to distinct cognition impairments and distinct brain dysfunctions

Different cognitive impairments have been linked to each of the two criterial ASD symptoms. Brunsdon and Happé 26 concluded that deficits in theory of mind contributed to ASD social interaction impairment, but not to repetitive and restricted behaviours. Brunsdon and Happé 26 also reported the opposite finding that executive dysfunction deficits contributed to ASD repetitive and restricted behaviours, but did not contribute to social interaction impairment.
Separate brain dysfunctions have also been linked to each of the two ASD criteria. Pina-Camacho et al.18 reviewed a large set of published studies and reported that autism social impairment was found in association with aberrant functioning of the face-fusiform area, fronto-temporal cortical networks, including the mirror neuron system, the anterior cingulate cortex, and the subcortical amygdala-hippocampal network. By contrast, Pina-Camacho et al.18 reported that restricted and repetitive behaviours were found in association with aberrant functioning and connectivity within fronto-cerebellar network, fronto-striatal system, anterior and posterior cingulate, posterior parietal regions, posterior regions of corpus callosum, cerebellar vermis and peduncles.

In sum, the two core ASD diagnostic symptoms have been found to occur separately in typically developing children, and social impairment does occur without restricted and repetitive behaviours in a subset of individuals clinically diagnosed with autism. ASD diagnostic symptoms have no consistent empirical association with each other and form two separate factors. ASD diagnostic symptoms appear to result from different brain dysfunctions and different cognitive dysfunctions. The preponderance of this evidence provides no support for the existence of a natural convergence of the two core ASD diagnostic symptoms.

Research continues to look for a unitary biological basis for ASD

Despite the current view that ASD is heterogeneous and includes many subgroups, nonetheless, much ASD research continues to look for a unification of genetic risk factors or a unified brain dysfunction. In addition, individual variation is not explored, and research continues to report only ASD diagnosis group findings.36, 39 For example, Pinto and colleagues39 studied de novo and inherited rare chromosome mutations called copy number variants (CNVs) in 2,446 ASD individuals and their parents, and in 2,640 unrelated controls. The research group39 noted finding an “extreme degree of etiological heterogeneity (36 different genetic loci were found among 82 individuals with pathogenic CNVs)” (p. 689). They reported, “genes involved in ASD participate in a wide array of processes, from neuronal development and axon guidance to MAPK and other kinase signaling cascades…to chromatin modification and transcription regulation” (p.691). 39 Despite noting extensive ASD genetic heterogeneity, Pinto et al.39 concluded, “many genes converge in interconnected functional modules, providing diagnostic and therapeutic targets” (p. 691).

Although the discovery of gene networks of interconnected function modules has scientific value, the value of a gene network module for diagnosis is extremely limited, and, given the range and interaction of genetic and environmental risk factors for ASD, the explanatory power of a gene network module for ASD medication is unclear. Because the individual variation in genetic risk factors in ASD is considerable, this is likely to reduce the potential value of an interconnected module as a basis for exploring therapeutic targets.

A similar problem limits the value of Stoner and colleagues’36 finding of abnormal cell layers and cell disorganization in patches of prefrontal and temporal cortex in 10 of 11 ASD case samples. The researchers36 claimed that most autism cases resulted from genetic mutations causing macroscopic early brain overgrowth that resulted in abnormal cortical patches. The discovery of abnormally organized patches in 10 ASD cases has scientific value. However, given all the varied brain dysfunction findings for ASD, the explanatory power of the patches model for ASD is minimal.

Belief in a unitary biological basis for ASD has blocked research progress
Collins and Tabak40 noted that in all biomedical research, “there remains a troubling frequency of published reports that claim a significant result, but fail to be reproducible” (p. 612). To date, no unified biological model of autism has been replicated or validated.24 The heterogeneity of autism and the need to find valid subgroups is acknowledged in many published research reports. 2, 8-13, 16-18, 23-26 But many researchers still believe that despite the existence of heterogeneity and the need for subgroups, autism must have one unifying biological feature.17, 24-25 Studies based on the belief that autism has one unifying biological feature do not withstand replication, possibly because heterogeneity in autism makes each research sample sufficiently different that no unitary cause is consistently reproducible. Consequently, the field has been in a state of constant theory replacement.24

Because no unified biological model of autism has been validated, the diagnosis of autism cannot be confirmed with a genetic test, a blood test or a brain scan. As a consequence, no medical treatment has been developed, even as increasing ASD prevalence has intensified the need to find effective medical treatments and prevention. The ASD diagnosis best functions as a means to direct children who express ASD symptoms (along with other non-ASD symptoms) into early behavioural intervention treatment programs.7

Willingham,41 a science journalist and parent of a teenager with autism, concluded, “We won’t find one single cause for autism, no matter how much money we throw at [it] trying to do so. Map the last decade of autism research, and you’ll find cul-de-sac after cul-de-sac of dead-end ideas or streets lined with echo chambers reverberating with results from the same groups, saying the same things, tracing the Rube-Goldbergian contraptions built within the framework of their pet hypotheses”(p.1). Willingham41 proposed that although studying her son would only be a case study, single case studies would be better than current unproductive autism research.

Willingham41 is justified in being angry, and is right that the quest for a unified biological basis for autism has resulted in little direct medical benefit to those with autism.42

Conclusion

Although many children and adults do express the two diagnostic ASD symptoms, the “gold standard” clinical application of ASD diagnostic criteria has not been validated by a unitary brain dysfunction or unitary risk factor. No validating unique link has been found between any single ASD risk factor and just the core ASD diagnostic symptoms. Risk factors for autism are risk factors for other syndromes and for non-ASD symptoms. The two core ASD diagnostic criteria form two factors. Research has linked each of the two ASD criteria alone to a wide range of different brain dysfunctions. Given these data and given the heterogeneity in ASD symptoms, non-ASD symptoms, and brain dysfunctions, finding individuals who meet ASD criteria is not sufficient evidence to support the belief that the ASD criteria identify a coherent behavioural or neurobiological disorder. Instead, the totality of current evidence argues that the two core ASD criteria are two symptoms among many frequently occurring symptoms of varied neurodevelopmental etiopathologies. In fact, the two ASD criteria of social interaction impairments and repetitive and restricted behaviours are just like the complex symptoms of attention problems, language disorders, intellectual disability, motor disability, seizures, sleep disorders and other medical disorders. These symptoms all occur with impaired neurodevelopment.

Autism researchers have reached a consensus that the ASD diagnosis includes many unknown subgroups that must be discovered for the development of effective medical treatments. Because the totality of evidence argues that the two core ASD criteria are two symptoms among many frequently occurring symptoms of varied neurodevelopmental etiopathologies, there is a pressing need for research that is independent of the ASD diagnosis. It is true that the study of large ASD samples has uncovered significant numbers of genetic and environmental risk factors. However, biologically valid subgroups will be found through the study of micro-groups of individuals with social impairment who share a specific brain dysfunction, a specific risk factor or both. Then, biologically valid subgroups can provide the basis to discover effective medical treatment and valid risk rates. For this to happen, researchers will have to abandon the belief that the autism diagnosis identifies a bounded behavioural or biological entity. They will have see that most non-ASD symptoms expressed with ASD symptoms are integral to an individual’s diagnostic phenotype. Finally, researchers will have to relinquish the goal of establishing one comprehensive explanation for autism.

References
1. Kanner L. Autistic disturbances of affective contact. The nervous child. 1943;2:217–250.
2. Rutter M. Addressing the issue of fractionation in autism spectrum disorder: A commentary on Brunsdon and Happé, Frazier et al., Hobson and Mandy et al. Autism. 2014;18(1):55-57.
3. American Psychiatric Association. Diagnostic and Statistical Manual of Mental
Disorders. 5th ed. Arlington, VA: American Psychiatric Association; 2013.
4. Dickstein DP, Pescosolido MF, Reidy BL, Galvan T, Kim KL, Seymour KE, Laird AR, DiMartino A, Barrett RP. Developmental meta-analysis of the functional neural correlates of Autism Spectrum Disorders. Journal of the American Academy of Child and Adolescent Psychiatry. 2013;52(3):279-289.
5. Silver WG, Rapin I. Neurobiological basis of autism. Pediatric clinics of North America. 2012;59(1):45-61.
6. Vasa RA, Ranta M, Huisman TA, Pinto, PS, Tillman, RM, Mostofsky, SH. Normal rates of neuroradiological findings in children with high functioning autism. Journal of autism and developmental disorders. 2012;42(8):1662-1670.
7. Peters-Scheffer N, Didden R, Korzilius H, Sturmey P. A meta-analytic study on the effectiveness of comprehensive ABA-based early intervention programs for children with Autism Spectrum Disorders. Research in autism spectrum disorders. 2011;5(1):60-69.
8. Berg JM, Geschwind DH. Autism genetics: searching for specificity and convergence. Genome biology. 2012;13(7):247.
9. Jeste SS, Geschwind DH. Disentangling the heterogeneity of autism spectrum disorder through genetic findings. Nature reviews neurology. 2014;10(2):74-81.
10. Grabruker A. Environmental factors in autism. Frontiers in psychiatry. 2013;18(3):118.
11. Gardener H, Spiegelman D, Buka SL. Perinatal and neonatal risk factors for autism: a comprehensive meta-analysis. Pediatrics. 2011;128(2):344-355.
12. Maramara LA, He W, Ming X. Pre-and perinatal risk factors for Autism Spectrum Disorder in a New Jersey cohort. Journal of child neurology. 2014;0883073813512899.
13. Schieve LA, Tian LH, Baio J, Rankin K, Rosenberg D, Wiggins L, et al. Population attributable fractions for three perinatal risk factors for Autism Spectrum Disorders, 2002 and 2008 Autism and Developmental Disabilities Monitoring Network. Annals of epidemiology. 2014;24(4):260-266.
14. Hallmayer J, Cleveland S, Torres A, Phillips J, Cohen B, Torigoe T, et al. Genetic heritability and shared environmental factors among twin pairs with autism. Archives of general psychiatry. 2011;68(11):1095-1102.
15. Sandin S, Lichtenstein P, Kuja-Halkola R, Larsson H, Hultman CM, Reichenberg A. The familial risk of autism. JAMA. 2014;311(17):1770-1777.
16. Coleman M, Gillberg C. The autisms. Oxford: Oxford University Press; 2012.
17. Waterhouse L. Rethinking autism: variation and complexity. New York:
Academic Press; 2013.
18. Pina-Camacho L, Villero S, Fraguas D, Boada L, Janssen J, Navas-Sánchez FJ, et al. Autism Spectrum Disorder: does neuroimaging support the DSM-5 proposal for a symptom dyad? A systematic review of functional magnetic resonance imaging and diffusion tensor imaging studies. Journal of autism and developmental disorders. 2012;42(7):1326-1331.
19. McCarthy SE, Gillis J, Kramer M, Lihm J, Yoon S, Berstein Y, et al. De novo mutations in schizophrenia implicate chromatin remodeling and support a genetic overlap with autism and intellectual disability. Molecular psychiatry. 2014;19(6):652-658.
20. Stefansson H, Meyer-Lindenberg A, Steinberg S, Magnusdottir B, Morgen K, Arnarsdottir S, et al. CNVs conferring risk of autism or schizophrenia affect cognition in controls. Nature. 2014;505(7483):361-366.
21. Blumberg SJ, Bramlett MD, Kogan M, Schieve L, Jones J, Lu M. Changes in prevalence of parent-reported Autism Spectrum Disorder in school-aged U. S. children: 2007 to 2011–2012. National Health Statistics Reports. 2013;65:1–11.
22. Developmental, Disabilities Monitoring Network Surveillance Year, and 2010 Principal Investigators. Prevalence of autism spectrum disorder among children aged 8 years-autism and developmental disabilities monitoring network, 11 sites, United States, 2010. Morbidity and mortality weekly report. Surveillance summaries (Washington, DC: 2002). 2014;63:1.
23. Parellada M, Penzol MJ, Pina L, Moreno C, González-Vioque E, Zalsman G, & Arango C. The neurobiology of autism spectrum disorders. European psychiatry. 2014;29(1):11-19.
24. Waterhouse L. Autism overflows: increasing prevalence and proliferating theories. Neuropsychology review. 2008;18:273-286.
25. Waterhouse L, Gillberg C. Why autism must be taken apart. Journal of autism and developmental disorders. 2014;1-5.
26. Brunsdon VE, Happé F. Exploring the ‘fractionation’ of autism at the cognitive level. Autism. 2014;18(1):17-30.
27. Frazier TW, Ratliff KR, Gruber C, Zhang Y, Law PA, Constantino JN. Confirmatory factor analytic structure and measurement invariance of quantitative autistic traits measured by the Social Responsiveness Scale-2. Autism. 2014;18(1):31-44.
28. Happé F, Ronald A, Plomin R. Time to give up on a single explanation for autism. Nature neuroscience. 2006;9(10):1218-1220.
29. Harrop C, McConachie H, Emsley R, Leadbitter K, Green J. Restricted and repetitive behaviors in autism spectrum disorders and typical development: cross-sectional and longitudinal comparisons. Journal of autism and developmental disorders. 2014;44(5):1207-1219.
30. Schuster J, Perry A, Bebko J, Toplak ME. Review of factor analytic studies examining symptoms of Autism Spectrum Disorders. Journal of autism and developmental disorders. 2014;44(1):90-110.
31. Williams DM, Bowler DM. (2014) Autism Spectrum Disorder: Fractionable or coherent? Autism. 2014;18(1):2-5.
32. Williams K, Woolfenden S, Roberts J, Rodger S, Bartak L, Prior M. Autism in context 1: Classification, counting and causes. Journal of paediatrics and child health. 2014;50(5):335-340.
33. Butler MG, Rafi SK, Manzardo AM. Clinically relevant candidate and known genes for Autism Spectrum Disorders (ASD) with representation on high resolution chromosome ideograms. OA Autism. 2014;Mar 25;2(1):5.
34. State MW, Levitt P. The conundrums of understanding genetic risks for autism spectrum disorders. Nature neuroscience. 2011;14(12):1499-1506.
35. Froehlich-Santino W, Londono Tobon A, Cleveland S, Torres A, Phillips J, Cohen B, et al. Prenatal and perinatal risk factors in a twin study of autism spectrum disorders. Journal of psychiatric research. 2014;54:100-108.
36. Stoner R, Chow, ML, Boyle MP, Sunkin SM, Mouton PR, Roy S, et al. Patches of disorganization in the neocortex of children with autism. New England journal of medicine. 2014;370(13):1209-1219.
37. Kulage KM, Smaldone AM, Cohn EG. How will DSM-5 affect autism diagnosis? A systematic literature review and meta-analysis. Journal of autism and developmental disorders. 2014;1-15.
38. Kim YS, Fombonne E, Koh YJ, Kim SJ, Cheon KA, Leventhal B. A comparison of DSM-IV PDD and DSM-5 ASD prevalence in an epidemiologic sample. Journal of the American Academy of Child and Adolescent Psychiatry. 2014;53(5):500-508.
39. Pinto D, Delaby E, Merico D, Barbosa M, Merikangas A, Klei L et al. Convergence of genes and cellular pathways dysregulated in autism spectrum disorders. The American journal of human genetics. 2014;94(5):677-694.
40. Collins FS, Tabak LA. NIH plans to enhance reproducibility. Nature. 2014;505:612–613.
41. Willingham E. Autism science, birth to thirteen. Forbes. May 3, 2014. http://www.forbes.com/sites/emilywillingham/2014/05/03/autism-science-birth-to-thirteen/.
42. Insel TR. The NIMH research domain criteria (RDoC) project: precision medicine for psychiatry. American journal of psychiatry. 2014;171(4):395-397.

Figure 1 Autism risk factors yield ASD and non-ASD symptoms.