Prenatal Inflammation and Autism Risk in a Swedish Birth Cohort

Oral Presentation
Thursday, May 10, 2018: 1:57 PM
Arcadis Zaal (de Doelen ICC Rotterdam)
R. Gardner1, C. Dalman2, B. K. Lee3 and H. Karlsson4, (1)Karolinska Institutet, Stockholm, Sweden, (2)Department of Public Health Sciences, Karolinska Institutet, Stockholm, Sweden, (3)Epidemiology and Biostatistics, Drexel University, Philadelphia, PA, (4)Department of Neuroscience, Karolinska Institutet, Stockholm, Sweden

Mounting evidence from both animal and human studies indicates that maternal immune function during pregnancy can influence neurodevelopmental outcomes relevant to autism. Most human studies of inflammatory markers measured in biosamples collected during the perinatal period conduct analyte-by-analyte statistical analyses, which can be problematic in terms of the number of statistical tests conducted and the underlying correlations among immune markers.


We aimed to measure acute phase proteins and cytokines in maternal serum samples collected during early pregnancy, as markers of prenatal inflammation and immune function, and to determine associations with later risk of autism spectrum disorders (ASD).


We performed a case-control study of 430 ASD cases and 589 unaffected controls born 1998-2000 in Sweden, with case ascertainment as of December 2011. Maternal sera samples were collected from regional biobanks. Nine acute phase proteins (α-2 microglobulin, C-reactive protein, haptoglobulin, serum amyloid P, procalcitonin, ferritin, tissue plasminogen activator, fibrinogen, and serum amyloid A) and seventeen cytokines (IL-1beta, IL-2, IL-4, IL-5, IL-6, IL-7, IL-8, IL-10, IL-12p70, IL-13, IL-17, IFN-gamma, G-CSF, GM-CSF, MCP1, MIP-1beta, TNF- alpha) were measured using a magnetic bead-based multiplex panel. Markers were individually examined, as well as combined into an inflammatory risk score based on ridge regression coefficients. We used logistic regression models adjusted for sex, birth year, maternal age, maternal immigration (Swedish born or not), parity, gestational week at sampling, and season at sampling.


Our preliminary analyses indicate that higher maternal levels of two cytokines were individually associated at p < 0.05 with increased risk of ASD in continuous analysis of log2-transformed concentrations: IL-17 (OR 1.12, 95% CI 1.01-1.25) and G-CSF (OR 1.13, 95% CI 1.02-1.27). None of the nine APP markers were associated with offspring ASD risk. A one standard deviation increase in the inflammatory risk score (including information for all analytes measured) was associated with an 85% increase in odds of ASD, OR: 1.85, 95% CI 1.42-2.42.


The maternal immune system, assessed in early pregnancy, is associated with subtle changes in risk for offspring ASD. These risks were apparent when examining certain markers individually. The individual markers related to ASD are interesting for their previously documented roles. Animal studies have shown that maternal IL-17 is required for offspring behavioral changes after maternal immune activation. G-CSF promotes neurogenesis within the central nervous system. Risks were also apparent when examining a composite inflammatory risk score. The risk score approach moves beyond analyte-by-analyte testing in order to avoid multiple comparisons and over-fitting of the data, as well as to improve generalizability of the results and address the underlying correlations among analytes.

See more of: Prenatal Autism Risk Factors
See more of: Epidemiology