The authors used a panel of monoclonal antibodies to study small bowel biopsies from children with autism in comparison to unaffected children with other diseases.

The abnormalities were distinct from previously reported conditions, pointing to the gut epithelium as a target of a specific immune response.

This finding suggests that autism may be an autoimmune disorder.

The work was performed by researchers at the Centre for Paediatric Gastroenterology, with the Inflammatory Bowel Disease Study Group, Royal Free & University College Medical School, London, UK; The IBD Research Unit, St. Mark's Hospital, Harrow, London, UK; Department of Medicine, Royal Free and University College Medical School, London, UK; Department of Histopathology, Royal Free and University College Medical School, London, UK.

It remains unclear whether autism is a single disease or a condition occurring as an end result of various abnormalities. Fundamental uncertainty remains about the relative input of genetic predisposition and environmental exposures.

Central to this uncertainty is the conflicting evidence concerning the incidence of autism. While there are several reports of rapid increase in incidence in Western countries -- suggesting an important environmental component -- others suggest that the increase is more apparent than real, and dependent on increased recognition, thus favoring a primarily genetic predisposition.

Most research has focused on the genetics of autism, and several genes have been implicated in classic autism.

This study is based on children with a form of autism characterized by regression in the second year of life, after apparently normal early development.

Most reports of immunological abnormalities in autistic children have been from this subgroup of affected children, and the authors cite the increasing body of evidence for abnormal immune regulation and autoimmunity in autism.

The initial observation of unexpected bowel pathology in autistic children came from the same group, and centered on pathology in the colon (Lancet 1998; 351: 637-641, American Journal of Gastroenterology 2000; 95: 2285-2295).

Use of immunohistochemical techniques had suggested a novel form of colitis, in which the epithelium of the colon was particularly affected (Journal of Pediatrics 2001; 138: 366-372), and, thus, possibly suggestive of autoimmunity.

In the current study, the authors report a highly unusual form of small bowel abnormality, subtle on routine staining, but with distinct and apparent autoimmune features.

Again the epithelium appears to be the focus of this immune response, with evidence of direct binding of self antibody to the surface of the epithelial cells (enterocytes), increased epithelial division and infiltration of T lymphocytes.

The authors used an extensive panel of monoclonal antibodies, and made comparisons to children with celiac disease, cerebral palsy and mental retardation, and apparently normal controls, none of whom showed this pattern of abnormalities.

The findings thus support an autoimmune basis for the unexpected bowel abnormalities in children with autism.

Many questions remain, in particular, the relevance of these findings to the general autistic population, as these children had more obvious bowel symptoms than commonly seen.

The authors avoid a direct suggestion that these bowel findings may be causal in autistic regression, but they cite reports that cognitive function is enhanced in some children with regressive autism when gut-based therapy is introduced.

They also raise the possibility that the bowel changes may be a manifestation of a primarily genetic condition affecting several systems in which disturbance of brain function is simply more apparent.

Further work thus clearly needs to be done to determine the role of the "gut-brain axis" in autism. However, confirmation that autoimmune responses really do contribute to cognitive regression in autistic children will raise the probability of a fundamental change in the treatment of autistic regression, based on modulating the immune response in children with early autistic symptoms.

An accompanying editorial further discusses this work and its implications ("Autoimmunity in Autism," by J Licinio, I Alvarado, and Ma-Li Wong, Molecular Psychiatry 2002 Volume 7, number 4, page 329.)

Molecular Psychiatry is published by the Nature Publishing Group. Its editor is Julio Licinio, M.D.

(Reference: "Small intestinal enteropathy with epithelial IgG and complement deposition in children with regressive autism;" F Torrente, P Ashwood, R Day, N Machado, RI Furlano, A Anthony, SE Davies, AJ Wakefield, MA Thompson, JA Walker-Smith and SH Murch, Molecular Psychiatry 2002 Volume 7, number 4, pages 375-382.)

[Contact: Dr. Simon Murch, Julio Licinio]

30-Apr-2002