Autism Risk May Be Buried in the Past

Childhood abuse of mothers and autism risk appear linked though reasons unclear

/ Author:  / Reviewed by: Joseph V. Madia, MD

(RxWiki News) The causes of autism increasingly appear to come from a variety of different risk factors. One of those risk factors may be related to the childhood experiences of the children's mothers.

A recent study found that women who had been abused as children were more likely to have children with autism.

The reason for this link is not clear, however. It could involve a biological change in the mother.

Or the link could be a genetic one where mental illness in the mother or grandparents is linked to the child's autism risk.

"Report suspected child abuse."

The study, led by Andrea L. Roberts, PhD, of the Department of Social and Behavioral Sciences at Harvard School of Public Health, looked at whether the risk of autism appeared linked to whether a pregnant woman had been abused when she was a child.

The researchers relied on data from 116,430 women involved in the Nurses' Health Study II. Within this group, the researchers found 451 mothers of children with autism and then randomly selected 52,498 mothers of children without autism for a comparison group.

The researchers found a higher percentage of women with autistic children had been abused in their youth, compared to the women with non-autistic children.

The link between a woman's childhood abuse and her own child's autism appeared to grow stronger as the severity of the abuse was higher.

Among women with the highest levels of abuse, for example, 1.8 percent of them had children with autism, compared to 0.7 percent of the women with non-autistic children.

The researchers calculated women who had been severely abused as children were about 3.7 times more likely to have a child with autism than mothers who had not been abused as children.

The researchers also looked at a variety of pregnancy complications in the women. They found women who had been abused as children were more likely to have toxemia, gestational diabetes, intimate partner abuse and a pregnancy shorter than 37 weeks (full term).

Abused women were also more likely to be taking antidepressants, to use alcohol and to smoke during pregnancy. Therefore, the researchers made adjustments in their calculations to account for all these other high risk factors in women who had been abused.

After adjusting for these factors, the researchers still found a three times higher risk for autism in the children of women who had been abused.  

The researchers suggest a variety of possibilities to account for these findings. For one thing, there may be other factors, such as diet or illness during pregnancy, that were not studied in this research but which may be related to women's higher risk of having a child with autism.

The authors also suggest that the abuse in a woman's childhood may have had biological effects that influenced the risk of autism in her child. Or, they propose that women who were abused might be more sensitive to stress during her pregnancy, which may influence a child's risk of autism.

Another possibility is that a child is genetically predisposed to have a higher risk of autism if the child's mother was abused and that abuse resulted from mental illness in one of her parents. Past research has shown there is some overlap in risk for some mental illnesses and risk for autism.

A related possibility is that women who were abused in childhood had undiagnosed autism or another mental illness which may have been linked to the abuse and may also have meant a child is genetically more likely to have autism.

Finally, although the researchers attempted to account for all the other risk factors in pregnancy that many of the abused women had, it's unclear whether statistical calculations could account for such a significant number of additional risk factors in that population.

The study was also limited by the fact that the child's autism, the mother's childhood abuse and the mother's pregnancy behaviors and illnesses were all reported by the mothers in the study.

A related limitation is that women reported whether they had been abused after already having had a child with autism, which may have influenced their answers. This is less likely, however, since the questions about a child's autism and the mother's abuse were asked in separate questionnaires four years apart.

The study was published March 20 in the journal JAMA Psychiatry. The research was funded by the Department of Defense through the US Army Medical Research and Materiel Command and the National Institutes of Health. The authors declared no conflicts of interest.

Reviewed by: 
Review Date: 
March 19, 2013
Last Updated:
March 21, 2013