Introduction
Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder characterized by qualitative impairments in social interaction and communication skills, accompanied by repetitive and stereotyped behaviors and interests (APA
2000). Family and twin studies have shown that the risk for autism is largely determined by genetic factors (>90%) (Geschwind and Levitt
2007). However, the pattern of inheritance is not straightforward and may include additive effects of common risk alleles, rare gene mutations (e.g., fragile-X) chromosomal abnormalities (e.g., copy number variants or CNVs), and possibly environmental insults (Mendelsohn and Schaefer
2008). Despite the significant heritability, finding the cause has been daunting due to genetic complexity and phenotypic variation (Szatmari et al.
2007; Abrahams and Geschwind
2008). Such heterogeneity in autism has led researchers to seek for reliable diagnostic tools to create genetically homogenous subgroups.
The study of head circumference and other morphological characteristics has begun to merge into autism research to stratify ASDs into more homogenous subgroups. Excessive head growth found in the first year of life, in children later diagnosed with autism, has been one of the most promising quantitative traits (Miles et al.
2000; Sacco et al.
2007). With respect to other morphological characteristics, an excess of minor physical anomalies (MPAs) in autistic individuals received specific attention (Ozgen et al.
2010). Although in the psychiatric literature the term “MPAs” is generally accepted, morphological features encompass major abnormalities as well as minor variants (Aase
1990; Merks et al.
2008). Minor variants are defined as slight morphological deviations that have no serious medical or cosmetic significance, however, they are of great value to the clinician because (1) they can be utilized as indicators of underlying disease susceptibility or markers of disturbed development (e.g., they are found to be more common in individuals with an obvious major embryonic defect; Aase
1990; McGrath et al.
2002); (2) They serve as external indicators of neurodevelopment because of the close relationship between cerebral and craniofacial development (Hammond et al.
2008); (3) Topography of morphological features could provide clues to the process of ASDs as they reflect adverse events during critical periods of development, usually within the first and second trimester (Lane et al.
1997). Minor variants that are observed in 4% or less of the normal population are termed minor anomalies, and those occurring in more than 4% of the normal population are termed common variants.
There is now robust evidence for the association between morphological abnormalities and autism (Ozgen et al.
2010). Existing studies, however, are limited by a number of methodological constrains: a. lack of standardization of the nomenclature and the absence of uniform diagnostic criteria; b. patients with different ethnic backgrounds were included; c. patients were not physically examined by the investigators specifically for the study; d. lack of control data or use of biased populations (e.g., hospital-based cohorts); e. relatively small sample sizes; f. no reports on interrater reliability; g. lack of consideration on gender effect. Moreover, the majority of studies that have assessed the incidence of morphological abnormalities in autism used the Waldrop scale (vide infra), with occasional modifications and omissions of items (Waldrop et al.
1968; Ismail et al.
1998). While the Waldrop scale is able to distinguish patients from controls, it has been criticized for inherent limitations regarding content and form; i.e., its restricted range of 18 items and low sensitivity (Lane et al.
1997).
With these caveats in mind, here we present an approach to study morphological features in a large sample of children with autism and a matched-case control group by using detailed definitions of the phenotype and an internationally accepted classifying list of 683 morphological features (Merks et al.
2003).
Discussion
We have demonstrated a strikingly high prevalence of morphological abnormalities in a large cohort of patients with ASD without mental retardation, compared with controls. The results support existing research findings from a meta-analysis by Ozgen et al. (
2010), as well as those of previous reports in the literature (Steg and Rapoport
1975; Walker
1977; Campbell et al.
1978; Gualtieri et al.
1982; Rodier et al.
1997; Hardan et al.
2006; Miles et al.
2008). In this present study, we tried to overcome the methodological limitations that occurred in previous studies. The main strengths of our study are careful clinical assessments using stringent morphological characterization, robust statistical methods and adequate power relative to previous studies. This sample size of 224 matched patient-control pairs is the largest to be studied to date for morphological features in ASD. Furthermore, in the present study all patients underwent a clinical morphological examination, using a list of 683 well-defined morphological abnormalities, assessed by the same investigator, and co-assessed blindly by a second investigator in about 10%.
Several interesting findings emerged from this study. First, our study demonstrated that the frequency of morphological anomalies is highly elevated in ASD without mental retardation, that not only minor anomalies but also major abnormalities and common variants are significantly more prevalent in children with autism versus controls. Some of the features may be direct malformations whereas others might be indirect e.g., open mouth expression, dry skin, pes planus. This finding resembles that of an association of autism with CNVs, which are common under normal circumstances, but also are an important aspect of genomic disorders, i.e., diseases caused by an alteration of the genome causing complete loss of copy, gain of copy or disruption of a dosage-sensitive gene (Sebat et al.
2007). In fact, Engels et al. (
2007) showed a direct association between the severity of physical anomalies and the gene content of the microdeletions/microduplications.
Second, males with autism showed a trend for more morphological abnormalities scored than females, i.e., males with ASD without mental retardation were notably having significantly more morphological abnormalities than female patients. Although one has to be cautious in interpreting these gender differences because of large differences in sample size, there are many plausible explanations for the observed differences. An explanation might be that the human male brain is generally more vulnerable to endogenous or exogenous factors which are poorly understood (Mizuno
2000). In addition, sex differences in brain structure and development are well known (Baron-Cohen et al.
2005). These results may also substantiate that some genetic risk factors influencing liability to autism in males and females might act differently. Furthermore, in schizophrenia literature there are findings parallel to ours regarding a gender effect on the occurrence of morphological characteristics (McGrath et al.
1995; Akabaliev and Sivkov
2007).
Third, the pattern of changes in morphological features was not straightforward. For example, we found that both smaller and larger mouth size were significantly elevated in patients with ASD compared to controls. Moreover, in contrast to some previous studies (Hardan et al.
2006; Rodier et al.
1997), we did not find that the presently studied patients with ASD had smaller ICD. On the contrary, we found that our patients had larger ICD and OCD, a finding that is consistent with the findings of Walker (
1977) and Bailey et al. (
1995). This observation may suggest either that the direction of changes in the orbital distance may be a random outcome of a common neurodevelopmental defect or that these opposite effects may underlie separate neurodevelopmental defects which might eventually lead to a better characterization of subgroups of patients with autism.
Fourth, in a recent meta-analysis in schizophrenia, a higher prevalence of morphological abnormalities was also established (Weinberg et al.
2007) with similar overlapping markers (e.g., brachycephaly; McGrath et al.
2002). Indeed, there is strong indication that individuals with schizophrenia may display evidence of craniofacial dysmorphology. Some evidence for such an overlap comes from the observation that individuals with ASD may also be at greater risk for schizophrenia (Mouridsen et al.
2008).
There are certain limitations that should be borne in mind when interpreting the results. First, although our sample size provides the necessary statistical power for quantitative analyses, yet some categorical analyses may suffer from power limitations due to the relatively small number of females in the sample. Notably, our sample was predominantly male. The potential difference in morphological features found between males and females require further confirmation.
Second, as morphologic examination requires in-person examination, it is generally not possible for the raters to be blind to diagnosis. However, 11% of controls and 9% of patients were scored independently by two observers, resulting in high kappa scores. Additionally there were no prior hypotheses as of which morphological abnormalities should be associated with autism.
Third, we limited our study population to Caucasian patients and controls as ethnicity can influence the prevalence of morphological abnormalities. Future studies are needed to establish similar norms for other ethnic groups. Likewise, we restricted ourselves to non-mentally retarded ASD patients. Therefore, we cannot generalize our findings to mentally retarded ASD patients or patient subgroups.
Forth, it is typically quite a challenge to determine whether these features are risk factors. The possible causal relevance of the association between disturbed prenatal development and subsequent autism has stimulated considerable discussion. In the present consideration of causality, we do not suggest that abnormal morphology in itself causes autism. In addition, we acknowledge that there are many disorders with similar morphological characteristics. e.g., increased head size. The same feature has already been shown to be related to a particular SNP variant in people with autism (Conciatori et al.
2004) as well as in people without a diagnosis (Muscarella et al.
2007). On the other hand, people with a very different etiology for autism, in utero exposure to valproic acid, also exhibit increased head size (Rasalam et al.
2005). Further studies are needed to discover the specificity, sensitivity and predictive value of these features for ASD. Moreover, some of the morphological characteristics among families may not be related to the ASD; demonstrating a genetic component independent of any syndrome and condition. Therefore, we recommend sib comparisons as future studies.
Last, the goal of this analysis was to provide a detailed analysis for the association of morphological features and ASD without mental retardation. We have not been able to address other intriguing questions such as clustering of the morphological features in different subgroups of ASD. These associations probably exist but are beyond the scope of this study. While we did not intend to establish specificity of single phenotypic characteristics to specific subtypes of ASD, we consider it possible that sets of physical anomalies might be related to severity of illness, or those specific phenotypic behavioral characteristics may define even more subtypes of ASD.
The current findings may have implications for future research, psychiatric nosology and clinical practice. With regard to future research implications, our findings help to reconcile previous puzzling evidence about the relationship between morphological features and ASD such as the inconsistency of this association across topographical areas. With regard to psychiatric nosology, our study adds to a growing body of research that suggests that morphological features might identify subgroups of individuals with ASD. It is possible that a subgroup of individuals with a higher number of unusual morphological findings or specific combinations of morphological findings is at highest risk of developing ASD without mental retardation. With regard to implications for clinical practice, our results support the importance of measuring morphological features in individuals with ASD. We suggest that routine assessment of quantitative and qualitative measurements could provide clinicians with necessary information to homogenise the ASD. In turn, the early recognition of the risk associated with morphological abnormalities might help to address pressing needs for the care of individuals with ASD.