Examining longitudinal associations between prenatal exposure to infections and child brain morphology

Abstract

Background: Maternal infection during pregnancy has been identified as a prenatal risk factor for the later development of psychopathology in exposed offspring. Neuroimaging data collected during childhood has suggested a link between prenatal exposure to maternal infection and child brain structure and function, potentially offering a neurobiological explanation for the emergence of psychopathology. Additionally, preclinical studies utilizing repeated measures of neuroimaging data suggest that effects of prenatal maternal infection on the offspring's brain may normalize over time (i.e., catch-up growth). However, it remains unclear whether exposure to prenatal maternal infection in humans is related to long-term differential neurodevelopmental trajectories. Hence, this study aimed to investigate the association between prenatal exposure to infections on child brain development over time using repeated measures MRI data.

Methods: We leveraged data from a population-based cohort, Generation R, in which we examined prospectively assessed self-reported infections at each trimester of pregnancy (N = 2,155). We further used three neuroimaging assessments (at mean ages 8, 10 and 14) to obtain cortical and subcortical measures of the offspring's brain morphology with MRI. Hereafter, we applied linear mixed-effects models, adjusting for several confounding factors, to estimate the association of prenatal maternal infection with child brain development over time.

Results: We found that prenatal exposure to infection in the third trimester was associated with a slower decrease in volumes of the pars orbitalis, rostral anterior cingulate and superior frontal gyrus, and a faster increase in the middle temporal gyrus. In the temporal pole we observed a divergent pattern, specifically showing an increase in volume in offspring exposed to more infections compared to a decrease in volume in offspring exposed to fewer infections. We further observed associations in other frontal and temporal lobe structures after exposure to infections in any trimester, though these did not survive multiple testing correction.

Conclusions: Our results suggest that prenatal exposure to infections in the third trimester may be associated with slower age-related growth in the regions: pars orbitalis, rostral anterior cingulate and superior frontal gyrus, and faster age-related growth in the middle temporal gyrus across childhood, suggesting a potential sensitive period. Our results might be interpreted as an extension of longitudinal findings from preclinical studies, indicating that children exposed to prenatal infections could exhibit catch-up growth. However, given the lack of differences in brain volume between various infection groups at baseline, there may instead be either a longitudinal deviation or a subtle temporal deviation. Subsequent well-powered studies that extend into the period of full brain development (∼25 years) are needed to confirm whether the observed phenomenon is indeed catch-up growth, a longitudinal deviation, or a subtle temporal deviation.

Keywords: Maternal immune activation; Neurodevelopment; Pediatric neuroimaging; Population-based cohort; Prenatal maternal infection.

Fig. 1

Fig. 1A depicts the prevalence per infection type per trimester Fig. 1B depicts the distribution of the infection sum score across trimesters, and Fig. 1C depicts the rate of infection per mother across gestation (including datapoints; and the thickness of the line indicates the number of women that had a similar infection rate). In Fig. 1B, the dotted line indicates the quartiles. Of note, Fig. 1B-trimester 3 only shows two lines as the median and third quartile have the same infection score (the second dotted line depicts the median).

Fig. 2. Flowchart depicting the selection of study participants.

Fig. 3

Prenatal infection and changes in child brain morphology; significant (p_fdr < 0.05) results. Plot 3A depicts the neuroanatomical position of the significant regions in trimester 3. Plot 3B shows a forest plot for the significant brain regions in trimester 3, but also the effect estimates and confidence intervals of the same regions in trimesters 1 and 2. In plot 3C, for visualization purpose only, prenatal exposure to infection is dichotomized into ‘+1 SD’ and ‘-1 SD’ groups, and timepoints 1, 2 and 3, indicate child’s mean age 8, 10 and 14, respectively.

Fig. 4. Brain anatomy plot with significant (p_uncorrected < 0.05) cortical and subcortical brain regions.