Title: Maternal circulating cytokine and chemokine levels during pregnancy and risk of autism spectrum disorder in offspring: a systematic review and meta-analysis.
Maternal immune activation (MIA) during pregnancy has been implicated in offspring neurodevelopment, but evidence linking maternal gestational immune mediators with autism spectrum disorder (ASD) remains heterogeneous. This systematic review and meta-analysis evaluated whether maternal circulating cytokine and chemokine levels during pregnancy are associated with autism in offspring.
Five databases were searched for human observational studies measuring maternal blood cytokines or chemokines during pregnancy and reporting autism outcomes in offspring. Eligible studies included autism versus control comparisons using maternal serum or plasma samples. Random-effects meta-analyses were conducted using odds ratios (ORs) as the primary effect measure. Standardised mean differences (SMDs) were also analysed where group-level concentration data were available. Heterogeneity was assessed using I² statistics, and small-study effects were explored using Egger's test where estimable.
Seven studies met the eligibility criteria. In OR-based analyses, nominal associations with autism were observed for IL-10 (OR 1.34, 95% CI 1.06-1.69), CXCL10 (OR 1.62, 95% CI 1.24-2.11), IL-1α (OR 1.42, 95% CI 1.23-1.65), IL-5 (OR 1.49, 95% CI 1.18-1.89), and IL-17A (OR 0.81, 95% CI 0.70-0.94). CCL4 showed a borderline association. Supportive SMD analyses indicated higher IL-6 and lower IL-1β concentrations in autism-associated pregnancies. Heterogeneity varied across analytes and was substantial for IL-10 and CCL4.
Maternal gestational immune profiles may differ in pregnancies later associated with autism, but current evidence does not support a single robust cytokine biomarker. Progress requires larger prospective cohorts with repeated sampling, harmonised assays, pathway-level analyses, and integrated placental-fetal neurodevelopmental phenotyping.
Five databases were searched for human observational studies measuring maternal blood cytokines or chemokines during pregnancy and reporting autism outcomes in offspring. Eligible studies included autism versus control comparisons using maternal serum or plasma samples. Random-effects meta-analyses were conducted using odds ratios (ORs) as the primary effect measure. Standardised mean differences (SMDs) were also analysed where group-level concentration data were available. Heterogeneity was assessed using I² statistics, and small-study effects were explored using Egger's test where estimable.
Seven studies met the eligibility criteria. In OR-based analyses, nominal associations with autism were observed for IL-10 (OR 1.34, 95% CI 1.06-1.69), CXCL10 (OR 1.62, 95% CI 1.24-2.11), IL-1α (OR 1.42, 95% CI 1.23-1.65), IL-5 (OR 1.49, 95% CI 1.18-1.89), and IL-17A (OR 0.81, 95% CI 0.70-0.94). CCL4 showed a borderline association. Supportive SMD analyses indicated higher IL-6 and lower IL-1β concentrations in autism-associated pregnancies. Heterogeneity varied across analytes and was substantial for IL-10 and CCL4.
Maternal gestational immune profiles may differ in pregnancies later associated with autism, but current evidence does not support a single robust cytokine biomarker. Progress requires larger prospective cohorts with repeated sampling, harmonised assays, pathway-level analyses, and integrated placental-fetal neurodevelopmental phenotyping.
Authors
Imran Imran, Islam Islam, Wright Wright, Williams Williams, Stephenson Stephenson
View on Pubmed