I want to be slightly careful when discussing the conclusion reached in the paper by Andrea Roberts and colleagues  that: "autism spectrum disorder risk was increased in children of women who reported fear of partner or sexual, emotional, or physical abuse in the 2 years before the birth year."
Careful not only because correlation has an uncanny habit of being translated into causation for some people (they are not one and the same thing) and how this combines when one interprets the concept of 'risk', but also because the results potentially, yet again, increase the likelihood of stigmatising parents of children with autism. Refrigerator mums turning into domestic violence dads is a headline that I don't think anyone wants to see with autism in mind...
That being said, I don't want to downplay the findings from Roberts et al particularly given that this is not the first time that this team have reported results in this area  (open-access). On that previous occasion as on this, the implication was that past events that produce serious psychological (as well as physical) harm might have an "intergenerational" effect. Indeed, even more research from this group has hinted that maternal posttraumatic stress disorder (PTSD) might also affect offspring autism risk . That is alongside the uncomfortable fact that having a child with autism seemingly does not protect mothers (or fathers) from being in abusive relationships.
The most recent study "calculated risk ratios for autism spectrum disorder associated with abuse in a population-based cohort of women and their children (54,512 controls, 451 cases)". They found that offspring risk of a diagnosis of autism spectrum disorder (ASD) was elevated for those mothers reporting "fear of partner or sexual, emotional, or physical abuse" sometime in the 2 years before the birth of their child, taking into account several other possible confounding variables. Interestingly, the risk of autism was not significantly elevated when it came to descriptions of "Physical harm from abuse during pregnancy" potentially indicating that any effect might be 'active' before conception.
This work potentially intersects with other research describing, for example, how parental experiences of trauma might have an intergenerational effect on future offspring. The paper by Duad and colleagues  suggesting that: "the children of tortured parents had more symptoms of anxiety, depression, post-traumatic stress, attention deficits and behavioural disorders" is not necessarily a direct comparator to the Roberts results but does provide a template for how parental experiences might have the ability to influence offspring outcomes. You could well suggest that the experiences of torture might have an effect on factors such as parenting style which could subsequently impact on offspring outcomes, but again it's difficult to prove causation. I might add that the Duad results focusing on "immigrant children whose parents have been tortured before coming to Sweden" may well also link in with other findings with an autism research slant (see here).
As to the mechanisms and without falling into any psychobabble explanations, I'm tempted to suggest that there may be quite a few areas requiring further investigation. Reiterating that correlation is not the same as causation and accepting that there may be as many different routes to a diagnosis of autism as there are 'autisms' [plural], I'm drawn to parallel these findings with other examples. I've mentioned the Dutch Hongerwinter before on this blog (see here) and how early nutrition might very well have an effect on later life health and wellbeing. Extending such effects to offspring of those subjected to the famine of 1944 (which also coincided with an unusually harsh winter) there is some evidence of a transgenerational effect for offspring in terms of the greater risk for being diagnosed with schizophrenia for example . I'm not necessarily saying that abuse is the same as starvation; merely that there may be similar biological mechanisms at work in terms of their respective effects on stress and also the concept of resilience. This area of work also introduces the idea that epigenetics might play some kind of role in potentially 'programming' offspring development for something like autism. Just in case this might seem a little outlandish, I might refer you to some work coming from Project Ice Storm  and how DNA methylation patterns *might* be tied into maternal 'appraisal' .
In a previous post on this blog, I've also covered some interesting work suggesting that trauma and PTSD *might* elevate the risk of autoimmune disorders (see here). I know that again we have a 'correlation not necessarily being the same as causation' state of affairs in discussing such a possible link, but given the rising tide of research suggesting that there may be something more to see when it comes to autoimmunity and [some] autism (see here) with speculation a-plenty (see here) I wonder whether some additional questioning about autoimmune disorders appearing in medical records or more direct evidence of autoimmune biology might be indicated in further work in this area? I say this accepting that the process of autoimmunity still has the ability to invoke head-scratching (see here).
That being said, there are caveats to all this. The variables which did not seem to account for the recent Roberts results - "gestation length, birth weight, maternal smoking or alcohol consumption during pregnancy, gestational diabetes, preeclampsia, or history of induced abortion" - cannot and do not cover the multitude of other confounding variables that might affect risk. Familial predisposition to autism as per the broader autism phenotype (BAP) perhaps also needs some further inspection, accepting my slightly furrowed brow when it comes to the BAP being tied into something like postpartum depression for example (see here). I would however, like to know a little bit more about other potential psychopathology / psychiatric diagnoses of parents and whether this might affect risk of autism outside of the history of exposure to abuse. Again, not to generalise or stigmatise, I'll link to some of the work on schizophrenia and violence  with some further analysis (see here) for example, and some suggestion that certain parental psychiatric diagnoses might elevate the risk of offspring autism .
Finally, I want to end with a line or two about the societal implications to come from the Roberts results. Abuse comes in many forms and is unfortunately an all-too-common occurrence these days (see here). Aside from the suggestion of a link with offspring behavioural and developmental outcomes, I'd be minded to suggest that we really should be doing a lot more to reduce the numbers of those suffering such acts and dedicate further resources to helping those who have been abused to try and mitigate not just the potential intergenerational effect but what such events might mean for them as individuals also .
 Roberts AL. et al. Maternal exposure to intimate partner abuse before birth is associated with autism spectrum disorder in offspring. Autism. 2015 Feb 6. pii: 1362361314566049.
 Roberts AL. et al. Association of maternal exposure to childhood abuse with elevated risk for autism in offspring. JAMA Psychiatry. 2013 May;70(5):508-15.
 Roberts AL. et al. Women's posttraumatic stress symptoms and autism spectrum disorder in their children. Res Autism Spectr Disord. 2014 Jun 1;8(6):608-616.
 Daud A. et al. Children in families of torture victims: transgenerational transmission of parents’ traumatic experiences to their children. International Journal of Social Welfare. 2005; 14: 23-32.
 Susser ES. & Lin SP. Schizophrenia after prenatal exposure to the Dutch Hunger Winter of 1944-1945. Arch Gen Psychiatry. 1992 Dec;49(12):983-8.
 Walder DJ. et al. Prenatal maternal stress predicts autism traits in 6½ year-old children: Project Ice Storm. Psychiatry Research. 2014; 219: 353-360.
 Cao-Lei L. et al. Pregnant women’s cognitive appraisal of a natural disaster affects DNA methylation in their children 13 years later: Project Ice Storm. Translational Psychiatry. 2015. Feb 24.
 Fleischman A. et al. Schizophrenia and violent crime: a population-based study. Psychol Med. 2014 Oct;44(14):3051-7.
 Daniels JL. et al. Parental psychiatric disorders associated with autism spectrum disorders in the offspring. Pediatrics. 2008 May;121(5):e1357-62.
 Dube SR. et al. Childhood abuse, household dysfunction, and the risk of attempted suicide throughout the life span: findings from the Adverse Childhood Experiences Study. JAMA. 2001 Dec 26;286(24):3089-96.
Roberts AL, Lyall K, Rich-Edwards JW, Ascherio A, & Weisskopf MG (2015). Maternal exposure to intimate partner abuse before birth is associated with autism spectrum disorder in offspring. Autism : the international journal of research and practice PMID: 25662292