Pregnancy is Associated with Accelerated Atrophy in Key Social Processing Brain Regions in Women with MS
Ayushi Balan1, Amit Akula2, Annika Anderson3, Stephanie Hsu4, Manula Dombagahawatta2, Holly Zhao2, Roland Henry2, Riley Bove2
1UCSF, 2University of California, San Francisco, 3UCLA David Geffen School of Medicine, 4Weill Cornell Medicine
Objective:
To evaluate volumetric changes associated with pregnancy in the brains of women with multiple sclerosis (WwMS).
Background:
In the general population, brain volume decreases are reported after pregnancy, including in gray matter structures relating to social processing. Such changes are not well characterized in WwMS, who may experience postpartum inflammatory activity and neurogenerative changes.
Design/Methods:
From an observational MS pregnancy registry, n=33 WwMS identified to have at least one brain MRI in both the pre-pregnancy (0-12M) and post-live birth (0-3 years) phases, using similar acquisition protocols. These individuals (P-WwMS) were manually matched (by age, interval between MRIs, acquisition protocol) with 33 non-pregnant (NP-WwMS) participants in an ongoing observational study. Two validated pipelines, FreeSurfer and FastSurfer, were utilized to measure regional brain volumes, focusing on four social processing regions. Changes (percent and absolute) between the two timepoints were compared between the P- and NP- cohorts using paired t-tests.
Results:
The cohorts did not differ in age, race/ethnicity, MS type or duration (p>0.05 each). P-WwMS experienced significantly greater % reductions in volume in the hippocampus (p=0.0029), fusiform gyrus (p=0.017), and anterior cingulate cortex (p=0.018). Evaluating absolute changes across both processing pipelines revealed similar findings, in all four regions of interest including the inferior frontal gyrus (p=0.0105), and multiple exploratory structures. P-WwMS also experienced more inflammatory activity (T2 lesions) and tissue destruction within lesions (T1 hypointensities) (p<0.05 each). However, the observed differences between cohorts in the four social processing areas remained significant after adjusting for these (p<0.05 each).
Conclusions:
WwMS who had a pregnancy experienced accelerated volume loss across key structures associated with social processing, consistent with the general population. These changes remained, after accounting for MS-related factors (inflammatory activity, T1 hypointensities). Possible underlying mechanisms—including atrophy, pruning, or resolution of inflammation—require further analysis to better inform considerations around pregnancy-planning in WwMS.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.