Maternal immune activation and autism in the offspring-what is the evidence for causation?

The maternal immune activation hypothesis has gained attention over the past two decades as a potential contributor to the etiology of autism. This hypothesis posits that maternal conditions associated with inflammation during pregnancy may increase the risk of autism in offspring. Autism is highly heritable, and causal environmental contributors to autism largely remain elusive. We review studies on maternal conditions during pregnancy, all associated with some degree of systemic inflammation; namely, maternal infections, autoimmunity, and high BMI. We additionally review studies of inflammatory markers in biological samples collected from the mother during pregnancy or from the neonate and their relationship with autism assessed in children later in life. Recent reports indicate familial clustering of autism, autoimmunity and infections, as well as genetic correlations between autism and aspects of immune function. In light of this literature, there is an apparent risk of confounding of the reported associations between inflammatory exposures and autism by familial genetic factors in both clinical and epidemiological cohort studies. We highlight recent studies that have attempted to address potential confounding to assess evidence of causal effects of inflammation during early life in autism.