Prenatal Maternal Thyroid Antibody, Immune Context and Risk of Autism in a Finnish Prenatal Birth Cohort

Background: Prenatal maternal autoimmunity has been associated with an increased risk of autism in the offspring, for example through studies showing a higher incidence of autoimmune disease diagnoses in mothers of children with autism versus those of controls. This has not been tested using clinical biomarkers of autoimmune disorders. Additionally, the role of prenatal maternal autoimmunity in relation to autism risk has not been examined in the context of additional biomarkers of immune dysfunction or other potentially related factors.

Objectives: To assess: a) the association between maternal thyroid peroxidase (TPO) autoantibodies in prenatal serum samples and the subsequent diagnosis of childhood autism in the offspring; and b) additional immune related factors that may influence this association.

Methods: The Finnish Prenatal Study of Autism (FIPS-A) is a case-control study nested in a national birth cohort. Cases with childhood autism (ICD-10 F84.0) born from 1987-2005 were identified using national registries and matched with controls by birth date, sex, location of birth, and residence in Finland. Prenatal maternal serum samples from 967 pairs of cases and matched controls were analyzed for thyroid peroxidase antibodies (TPO-Ab), a marker of autoimmune thyroid disease, using a chemiluminescent microparticle immunoassay. Data were analyzed using conditional logistic regression. Stratified analyses were conducted by offspring sex, the presence or absence of intellectual disability in the case, and birth year. Heterogeneity of the association across strata was determined using tests of interaction. We will further test this association in the context of additional immune related factors such as biomarkers of inflammation and exposures/conditions potentially related to inflammation/immune dysfunction.

Results: TPO antibody positivity was associated with a significantly increased odds of autism (OR (95% CI) =1.78 (1.16-2.75)) after adjusting for potential confounders. Log transformed TPO antibody defined as a continuous variable was also positively associated with the odds of autism (OR (95% CI) =1.09 (1.01-1.17)).  The association between TPO antibody positivity and autism was stronger for cases without intellectual disability (OR (95% CI) = 2.35 (1.33-4.14)) than for cases with intellectual disability (OR (95% CI) = 1.13 (0.57-2.27)), although the difference between groups was not statistically significant. A lack of association between thyroid hormone levels and autism suggested that altered hormone levels secondary to autoimmunity did not mediate the relationship. We will also present results of analyses assessing the joint role of TPO antibody and additional immune related factors.

Conclusions: These findings provide biomarker-based evidence for the association of prenatal maternal autoimmune thyroid disease with autism. The results suggest that the association may be related to underlying maternal autoimmunity or immune factors rather than thyroid dysfunction. We will also present results on the relationship of maternal inflammation to TPO autoantibodies and whether the association is mediated by inflammation.