There is an extensive body of literature indicating that a subset of mothers who have children with autism carry brain-reactive antibodies - which affect the function of brain cells. There is a smaller, but quite compelling, body of literature indicating that these maternal antibodies can alter fetal brain development and subsequent behavior in the offspring if given intravenously to a gestating mouse or monkey. Betty Diamond and Peter Gregersen propose that women who carry these antibodies have a higher number of autoimmunity susceptibility genes compared with controls. The maternal brain is protected from antibody-mediated damage by the blood-brain barrier, whereas the fetal brain is exposed to all maternal antibodies. If brain-reactive antibodies of the mother can lead to autism spectrum disorders in the offspring, then approximately 15 to 20 percent of autism cases could be explained by this mechanism. In support of the researchers' hypothesis, a large population-based study has shown that mothers of children with autism are more likely than average to have rheumatoid arthritis or celiac disease. This study strengthens the conclusion of several smaller studies showing an increased incidence of autoimmune disease in families with a history of autism. This hypothesis is also appealing because of the potential ability to identify at-risk pregnancies and develop strategies for preventing antibody-mediated brain damage.