Literature review: gastrointestinal symptoms in ASD, brain structure of identical twins with ASD.

03Original abstract

▪ Gastrointestinal Symptom Incidence in ASD [Ibrahim, Voigt, Katusic, Weaver, & Barbaresi, 2009] This study was designed to test the hypothesis that there are more gastrointestinal symptoms and/or disorders in individuals with ASD compared to epidemiological controls. This goal was achieved using the resources of the Rochester Epidemiology Project; more than 95% of medical care in Rochester is provided at the Mayo Clinic and the Olmsted Medical Center. Complete computerized medical records are available, which include developmental, pediatric, psychiatric, neurological, and psychological assessments as well as indexing of all gastrointestinal symptoms and diagnoses. The authors identified individuals with autism who lived in Olmsted County, Minnesota between 1976 and 1997 and who were less than 21 years of age; diagnosis was based on chart review identification of individuals who met DSM-IV criteria for autism. To increase power, each child with ASD was matched to two control individuals on age, sex, year of first registration, and duration of follow-up. Five research categories of gastrointestinal symptoms or diagnoses were defined before record review: (1) constipation; (2) diarrhea (e.g. enteritis, gastroenteritis, colitis, or loose stool); (3) gastroesophageal reflux/vomiting; (4) abdominal discomfort, irritability; and (5) feeding issues and food selectivity. The investigators found no difference in overall gastrointestinal symptoms between the ASD and control groups, but individuals with autism had a twofold increase in relative risk for (1) constipation (incidence by 20 years-old: autism 33.9%, control 17.6%) and (2) feeding issues and food selectivity (incidence by 20 years-old: autism 24.5%, control 16.1%). The authors noted that there was only one diagnosis of Crohn disease in the autism group. In addition, diagnosed food intolerance was rare and equally common in the autism and control groups. Feeding difficulties and food selectivity appeared related to the restricted interests associated with autism, but the authors did not provide specific analysis of the relationship between DSM-IV symptom domains and this category of gastrointestinal symptoms. Constipation was interpreted as being secondary to food selectivity, which the authors suggested may have contributed to a restricted diet containing less fiber. The authors acknowledged the limited generalizability of their findings since 98% of participants were Caucasian. This epidemiological study did not involve elicitation of a gastrointestinal history nor appropriate physical examination; neither was it possible to detect an increase in gastrointestinal symptoms at a specific time in development. As with many studies there are concerns about statistical power. The sample size of 121 individuals with ASD is relatively small. In summary, this study of a population-based sample with long-term follow-up found no general increase in gastrointestinal symptomatology in individuals with ASD compared to epidemiological controls. Nevertheless, individuals with ASD were twice as likely to have constipation and show feeding issues/food selectivity. Although this study was not designed to study the etiology of constipation, it was evident that children and adolescents with ASD are at increased risk for constipation, and diagnosis and treatment of this problem is an important part of pediatric care. ▪ Brain Structure in Identical Twins with ASD [Mitchell et al., 2009] The authors studied structural MRI scans of identical twin pairs with ASD to determine if phenotypic differences within identical twin pairs related to variability in brain structure. Additionally, the authors were interested in separating out genetic and environmental components of autism, with the rationale that what was shared between identical twins would relate to shared nuclear genomic variants and what was not shared would relate to environmental effects. Examples of potential nonshared environmental effects might include differential optimality of pregnancy and delivery, somatic mutations after splitting into two embryos, and differential epigenetic regulation. The authors studied 14 identical twin pairs in which one of the twins had autism, defined by the authors as meeting ADI-R criteria for autism and scoring above or within one point of the ADOS-G criteria for autism. The mean age of the twin pairs was 8.79 years. There were 12 male and 2 female pairs. Apparently, the twin with more severe autism was chosen as the index case and the other twin was assigned to the co-twin sample. The authors found that 36% of the co-twins had autism, 28% had pervasive developmental disorder not otherwise specified, and 36% had no ASD diagnosis. Details about the twin sample were provided in a previous publication, apparently describing these 14 twin pairs and an additional 2 twin pairs [Kates et al., 2004]. In the 2004 report, in all but one of 16 twin pairs, the co-twin had ASD, or a history of language or developmental delay, or mild impairment of communication and social skills consistent with the broader autism phenotype. The sample of 14 index twins and their 14 co-twins were compared to 14 typically developing age and sex-matched singleton children. Matching did not occur on full-scale IQ. The autism subjects had a mean full-scale IQ of 65, their co-twins had a mean full-scale IQ of 86, and the typically developing controls had a mean IQ of 121. Planned analysis were: (1) to compare the twin with autism to their co-twin with varying degrees of ASD severity; (2) to compare the autism group to the typically developing children; and (3) to study the correlation between autism severity and neuroanatomical measures within the twin pairs. There were no measured differences in brain regions between the index twin and their co-twins. The authors did find increased dorsal prefrontal cortical volumes, decreased size of the genu and anterior body of the corpus callosum, and reduced volumes of cerebellar vermal lobules VI and VII in the twin sample compared to the typically developing children. The authors found a positive correlation between dorsolateral frontal cortical volumes and ADI-R total scores in the index twin group, but no correlation between volumes and ADI score in their co-twins, who had more variable ASD severity. It is relevant that this correlation was the only reported ASD phenotype comparison and it would have been helpful to see if a similar correlation was observed with contemporaneous ADOS scores. The authors acknowledged the limitations of the study, including the small sample size and the cross-sectional design. It would also have been useful to look at the relationship between brain volumes and specific components of the ASD phenotype (e.g. ADI domain scores or factor scores, rather than ADI-R total scores). The paper highlights the potential for examining variability within monozygotic twin pairs, but the negative findings also illustrate the difficulty in establish links between cognition and macroscopic anatomy.

Autism research : official journal of the International Society for Autism Research, 2009 · doi:10.1002/aur.95