Structural and Functional Brain Remodeling during Pregnancy with Diffusion Tensor MRI and Resting-State Functional MRI

Abstract

Although pregnancy-induced hormonal changes have been shown to alter the brain at the neuronal level, the exact effects of pregnancy on brain at the tissue level remain unclear. In this study, diffusion tensor imaging (DTI) and resting-state functional MRI (rsfMRI) were employed to investigate and document the effects of pregnancy on the structure and function of the brain tissues. Fifteen Sprague-Dawley female rats were longitudinally studied at three days before mating (baseline) and seventeen days after mating (G17). G17 is equivalent to the early stage of the third trimester in humans. Seven age-matched nulliparous female rats served as non-pregnant controls and were scanned at the same time-points. For DTI, diffusivity was found to generally increase in the whole brain during pregnancy, indicating structural changes at microscopic levels that facilitated water molecular movement. Regionally, mean diffusivity increased more pronouncedly in the dorsal hippocampus while fractional anisotropy in the dorsal dentate gyrus increased significantly during pregnancy. For rsfMRI, bilateral functional connectivity in the hippocampus increased significantly during pregnancy. Moreover, fractional anisotropy increase in the dentate gyrus appeared to correlate with the bilateral functional connectivity increase in the hippocampus. These findings revealed tissue structural modifications in the whole brain during pregnancy, and that the hippocampus was structurally and functionally remodeled in a more marked manner.

Fig 1. Histogram comparison of the tissue structural changes in the global brain between three days before (baseline) and seventeen days after mating (G17) measured by DTI in the pregnancy (a) and control (b) groups.

Mean diffusivity (MD), fractional anisotropy (FA), axial diffusivity (AD) and radial diffusivity (RD) were evaluated in the whole brain (WB), gray matter (GM) and white matter (WM). They were generally observed to increase during pregnancy (a). In contrast, slight decreases were seen in the control group (b).

Fig 2. The comparison of the tissue structural changes in the global brain between the baseline and G17 in the pregnancy (top) and control (bottom) groups.

In the pregnancy group, MD, AD and RD increased globally, indicating tissue microstructural remodeling (that facilitated water molecular diffusion) in the global brain. Two-way ANOVA was applied, followed by post-hoc Bonferroni’s test. *, ** and *** denote p<0.05, p<0.01 and p<0.001, respectively. Error bars indicate the standard deviation.

Fig 3. The results of the voxel-based analysis followed by multiple testing corrections via false discovery rate in MD (a) and FA (b) of the pregnancy group. The hot voxels indicated significant increase during pregnancy, and the threshold applied was p = 0.05. With reference to (a), (b) and the rat brain atlas, regions of interest (ROIs) were defined (c) and employed for quantitative analysis. The ROI-HP covers the dorsal hippocampus, and the ROI-DG covers the dorsal dentate gyrus (DG). (d) The summary of the local tissue structural changes between the baseline and G17 in the ROI-DG and ROI-HP of both the pregnancy and control groups without any normalization. There was significant increase in FA in the ROI-DG, and significant increase in MD, AD and RD in the ROI-HP during pregnancy, but not in the controls. After normalization with global GM changes (e), MD, AD and RD changes in ROI-HP became smaller but remained increased, while the increase in FA in the ROI-DG was similar. These results indicated the presence of more pronounced tissue structural remodeling in the dorsal hippocampus, including the dorsal dentate gyrus during pregnancy. Two-way ANOVA was applied, followed by post-hoc Bonferroni’s test.

* and ** denote p<0.05 and p<0.01 and, respectively. Error bars indicate the standard deviation.

Fig 4. (a) The mean correlation coefficient maps of rsfMRI obtained using seed based analysis (SBA) with seeds (crosses) in the right or left dorsal hippocampus. (b) The mean z-score maps obtained using independent component analysis (ICA).

The mean correlation coefficient maps and mean z-score maps demonstrate the presence of the bilateral hippocampal rsfMRI connectivity in both pregnancy and control groups at baseline and G17. Both types of connectivity maps show that the rsfMRI connectivity in the hippocampus became stronger during pregnancy. These results indicated that the bilateral hippocampus became more functionally connected during pregnancy. The maps are overlaid on T2-weighted anatomical image. Two-way ANOVA was applied followed by post-hoc Bonferroni’s test. *** denotes p<0.001. Error bars indicate the standard deviation.

Fig 5. The relationship between FA in dorsal dentate gyrus and functional connectivity in hippocampus of individual rats from baseline to G17 in the pregnancy group (upper panel) and control group (lower panel).

These results indicated that DTI-based structural changes in the dorsal dentate gyrus and the bilateral rsfMRI connectivity changes in the hippocampus during pregnancy were correlated and might be coupled. In contrast, no such correlation was observed in the control group.