Skip navigation

Calls to Action

Study Using AGRE Sample Finds Autism Genetic Link Influenced by Sex

October 14, 2007

Autism is a disorder that affects every race, nation and creed. However, there is a clear male bias in susceptibility to autism and Pervasive Developmental Disorders. Recent reviews have indicated that males are four to ten times more likely to be affected by autism than females. Interestingly, other neurodevelopmental disorders such as Attention Deficit Disorder and dyslexia reflect a similar male:female bias.

As a result, many researchers expected to identify susceptibility genes on the sex chromosomes. The first genome scans left researchers baffled because these signals were not evident. Researchers found it difficult to reconcile why there would be a bias in affected males that did not involve the sex chromosomes.

Scientists reasoned that one possible explanation for this phenomenon was that genes contributing to the susceptibility of autism are modified by sex hormones. In a report published in the September 2004 issue of the American Journal of Human Genetics, a group of scientists at UCLA and Columbia University identified a clever strategy to discover genes modified by sex-specific factors.

Jennifer Stone and colleagues used the existing genome scan data performed on AGRE families in 2003 (Yonan et al., 2003). They divided families into two groups: those containing only males affected with autism, and those containing both affected males and females. The rationale for this division was based on the idea that some autism susceptibility genes will be influenced by sex-specific risk factors. One can assume that the genes of male-only families are enriched for having male risk factors, while those of families containing both males and females have risk factors that affect both sexes.


A genome scan is carried out by performing many genetic tests on every available family member in a large collection of families (e.g., AGRE). The genetic tests determine which form or variant of a genetic marker (a piece of DNA that is used to tag specific areas of the genome) is passed from parents to children. When the forms of all the genetic markers are tabulated, statistical analyses can be made of this information to determine if certain areas in the DNA are shared among individuals with autism. This sharing is the foundation of how human geneticists identify regions in the DNA that play a role in diseases.


Amazingly, a genetic region on chromosome 17 identified in the original genome scan was seen as highly shared among affected individuals in the male-only families. More importantly, this trend toward higher sharing was much stronger than in the original genome scan despite the reduction in the number of families examined, because the families were divided into two nearly equal groups (58% male only families; 42% female containing families). Interestingly, when the genome scan was performed on female containing families, the chromosome 17 region did not indicate any sharing, suggesting that all the statistical signals seen in the original scan were generated almost exclusively from the male-only families.


Aside from the chromosome 17 region, 24 other genetic regions were found to be different between this genome scan and the original 2003 scan, but none reached the significance of the region found on chromosome 17. Some scientists may question whether the division of families into male-only and female-containing families was arbitrary and therefore question the meaning of the findings. To avoid this type of criticism, the scientists developed a method whereby all families in the initial analyses were randomly assigned to two groups, regardless of whether they were an affected male-only family or a mixed male-female family. This process was repeated 1000 times to try to replicate the amount of DNA sharing at particular genetic regions without the sex-specific bias. This procedure also served to create a baseline of what could occur by chance. Interestingly, the replicates did not produce any signals close to that identified on the chromosome 17 region in the sex-specific samples.


Because of its identification as a region likely to harbor susceptibility genes in the original genome scan, the region on chromosome 17 is actively being examined as a potential area of future research. Several groups have already examined the serotonin transporter gene also found in the chromosome 17 region. However, the evidence is not entirely conclusive, and a more thorough examination of this candidate genes as well as the surrounding region is being carried out.