The international journal Science of the Total Environment has just published a compelling study from the Republic of Korea, where autism prevalence is high. The study identifies a strong relationship between prenatal and early childhood exposure to mercury and autistic behaviors in five-year-olds.
Lead author Jia Ryu and coauthors acknowledge mercury’s potential for neurotoxicity straight away but choose to characterize previous findings on the mercury-autism relationship as “inconsistent.” They attribute the seeming lack of consistency, in part, to methodological issues, especially flagging problematic cross-sectional study designs that measure autistic behaviors and mercury levels (in either blood or hair) at a single point in time. To rectify these methodological weaknesses, Ryu and coauthors report on data from a multi-region longitudinal study in the Republic of Korea called the Mothers and Children’s Environmental Health (MOCEH) study.
The ongoing MOCEH study examines environmental exposures during pregnancy and childhood and their effects on children’s growth and development. A unique feature is that it includes five different blood samples: maternal blood from early and late pregnancy; cord blood; and samples from children at two and three years of age. In addition, the study asks mothers to complete three follow-up surveys and—when their child reaches age five—the 65-item Social Responsiveness Scale (SRS), which assesses autistic behaviors.
Ryu and colleagues present available data for 458 (26%) of the 1,751 mother-child pairs originally recruited into the MOCEH study. What are their key findings?
–First, the researchers report a significant linear relationship between mercury exposure and autistic behaviors (as indicated by a scaled score called an SRS T-score). Strikingly, they find that with a doubling of blood mercury levels at four time points (late pregnancy, cord blood, and at two and three years of age), SRS T-scores are significantly higher.
–Ryu et al. also look specifically at SRS T-scores greater than or equal to 60. Sixty and above is the accepted threshold for detecting “mild to moderate” deficits of social behavior related to autism; scores of 76 or more are in the “severe” range. In these analyses, the same linear relationship holds for late pregnancy and birth (i.e., cord blood). With a doubling of blood mercury levels at these two time points, there is a 31% and 28% increase, respectively, in the risk of an SRS T-score of 60 or more.
–Finally, the researchers identify a stronger association between late-pregnancy mercury exposure and autistic behaviors in five-year-old boys versus five-year-old girls, perhaps due to mercury’s endocrine-disrupting properties.
Although the Korean study has many methodological strengths, it displays some curious omissions and contradictions. For example, the authors note, almost in passing, that they measured total mercury “instead of methylmercury,” but they discount any potential relevance of ethylmercury from thimerosal-containing vaccines. In support of this dismissive stance, they cite a 2014 meta-analysis that failed to find any “material associations” between thimerosal exposure and ASD or attention-deficit hyperactivity disorder (ADHD). They do not mention that the authors of the meta-analysis actually urged readers to interpret their results with caution “since the number of epidemiological studies on this issue [is] limited and still at an early stage.”