Maternal Immune Activation and Child Brain Development: A Longitudinal Population-Based Multimodal Neuroimaging Study

Background: Maternal immune activation (MIA) has been hypothesized to have an adverse effect on child neurodevelopment, but only a few neuroimaging studies have been performed to date, mostly in neonates. In this population-based cohort study, we investigated the association between MIA and multiple neuroimaging modalities depicting brain development from childhood to adolescence. Methods: We used data of mother-child pairs from the Generation R Study. To define our exposure, we measured interleukin (IL) 1β, IL-6, IL-17a, IL-23, interferon gamma, and C-reactive protein at 2 time points during pregnancy. Because levels of these 5 cytokines were highly correlated, we were able to compute a cytokine index. We used multiple brain imaging modalities as outcomes, including global and regional measures of brain morphology (structural magnetic resonance imaging, volume; n = 3295), white matter microstructure (diffusion magnetic resonance imaging, fractional anisotropy and mean diffusivity; n = 3267), and functional connectivity (functional magnetic resonance imaging, graph theory measures, and network-level connectivity; n = 2914) in the children at ages 10 and 14 years. We performed mixed effects models using child's age as a continuous time variable. Results: We found no significant effect of time on any neuroimaging outcomes in children over time, and there was no time × MIA interaction. These associations were similar for the cytokine index, C-reactive protein, and individual cytokines. We observed no evidence for differential effects of timing of prenatal MIA or child sex after multiple testing correction. Conclusions: In this longitudinal population-based study, we found no evidence for an association between MIA and child brain development in the general population. Our findings differ from previous research in neonates that have shown structural and functional brain abnormalities after MIA.