The paper by Andrea Roberts and colleagues  is offered up for brief discussion today, and a slightly uncomfortable but nonetheless potentially important set of results that "provide evidence that childhood abuse affects mate selection, with implications for offspring health."
I tread very carefully with this topic but was intrigued by the findings looking at whether 'offspring health' in the context of the presence of autism or autism spectrum disorder (ASD) might be at least sometimes mediated by the mate selection decisions made by women who "experienced childhood abuse." I might add that sweeping generalisations from this and related work from this authorship group (see here) are not required.
Some of the data for this study, I believe, once again comes from the Nurses’ Health Study II where initially autistic traits were measured using the Social Responsiveness Scale (SRS) for men and women. These results were stratified for various types and levels of abuse reported to be encountered during childhood. Using some statistical modelling, researchers observed that: "Maternal childhood abuse was strongly associated with high paternal autistic traits." In other words, it seemed like the experience of childhood abuse - particularly severe childhood abuse - encountered by women, seemed to correlate with later selection of partners who scored high when it came to the presence of autistic traits (as measured by the SRS). Results also took into account various potential confounding variables including socio-economic status (SES).
Bearing in mind previous discussions on assortative mating and autism  where "people who have certain “autistic” traits are increasingly meeting and marrying each other and having offspring who are more likely to be on the spectrum" the Roberts results potentially add an additional dimension. They suggest that early life experiences (particularly adverse experiences) may have quite a powerful effect when it comes to choices of mating partner. Whether this is driven by an active preference for the presence of autistic traits in a partner or is down to other factors such as 'availability' and/or 'accessibility' cannot be deduced from the current data. It is also important to add that looking at autistic traits alone does not necessarily mean that autistic traits drive any mating selection being undertaken. As I've mentioned before on this blog, various screens for autistic traits seem to have a generalisability issue (see here for example). I might also add that: "Maternal and paternal autistic traits accounted for 21% of the association between maternal abuse and offspring autism" so one has to be careful not to make too much of any single connection.
But still there is something rather interesting from these latest results from Roberts et al [with appropriate cautions] that requires further research inspection...
 Roberts AL. et al. Maternal Exposure to Childhood Abuse is Associated with Mate Selection: Implications for Autism in Offspring. J Autism Dev Disord. 2017 Apr 9.
 Baron‐Cohen S. Two new theories of autism: hyper‐systemising and assortative mating. Archives of Disease in Childhood. 2006;91(1):2-5.
Roberts AL, Lyall K, & Weisskopf MG (2017). Maternal Exposure to Childhood Abuse is Associated with Mate Selection: Implications for Autism in Offspring. Journal of autism and developmental disorders PMID: 28393290