A profile and review of findings from the Early Markers for Autism study: unique contributions from a population-based case-control study in California
Kristen Lyall,Jennifer L Ames,Michelle Pearl,Michela Traglia,Lauren A Weiss,Gayle C Windham,Martin Kharrazi,Cathleen K Yoshida,Robert Yolken,Heather E Volk,Paul Ashwood,Judy Van de Water,Lisa A Croen
DOI: https://doi.org/10.1186/s13229-021-00429-7
2021-03-18
Abstract:Background: The Early Markers for Autism (EMA) study is a population-based case-control study designed to learn more about early biologic processes involved in ASD. Methods: Participants were drawn from Southern California births from 2000 to 2003 with archived prenatal and neonatal screening specimens. Across two phases, children with ASD (n = 629) and intellectual disability without ASD (ID, n = 230) were ascertained from the California Department of Developmental Services (DDS), with diagnoses confirmed according to DSM-IV-TR criteria based on expert clinical review of abstracted records. General population controls (GP, n = 599) were randomly sampled from birth certificate files and matched to ASD cases by sex, birth month and year after excluding individuals with DDS records. EMA has published over 20 papers examining immune markers, endogenous hormones, environmental chemicals, and genetic factors in association with ASD and ID. This review summarizes the results across these studies, as well as the EMA study design and future directions. Results: EMA enabled several key contributions to the literature, including the examination of biomarker levels in biospecimens prospectively collected during critical windows of neurodevelopment. Key findings from EMA include demonstration of elevated cytokine and chemokine levels in maternal mid-pregnancy serum samples in association with ASD, as well as aberrations in other immune marker levels; suggestions of increased odds of ASD with prenatal exposure to certain endocrine disrupting chemicals, though not in mixture analyses; and demonstration of maternal and fetal genetic influence on prenatal chemical, and maternal and neonatal immune marker and vitamin D levels. We also observed an overall lack of association with ASD and measured maternal and neonatal vitamin D, mercury, and brain-derived neurotrophic factor (BDNF) levels. Limitations: Covariate and outcome data were limited to information in Vital Statistics and DDS records. As a study based in Southern California, generalizability for certain environmental exposures may be reduced. Conclusions: Results across EMA studies support the importance of the prenatal and neonatal periods in ASD etiology, and provide evidence for the role of the maternal immune response during pregnancy. Future directions for EMA, and the field of ASD in general, include interrogation of mechanistic pathways and examination of combined effects of exposures.