Human menstrual cycle variation in subcortical functional brain connectivity: a multimodal analysis approach

Abstract

Increasing evidence suggests that endogenous sex steroid changes affect human brain functional connectivity, which could be obtained by resting-state fMRI (RS-fMRI). Nevertheless, RS studies on the menstrual cycle (MC) are underrepresented and yield inconsistent results. We attribute these inconsistencies to the use of various methods in exploratory approaches and small sample sizes. Hormonal fluctuations along the MC likely elicit subtle changes that, however, may still have profound impact on network dynamics when affecting key brain nodes. To address these issues, we propose a ROI-based multimodal analysis approach focusing on areas of high functional relevance to adequately capture these changes. To that end, sixty naturally cycling women underwent RS-fMRI in three different cycle phases and we performed the following analyses: (1) group-independent component analyses to identify intrinsic connectivity networks, (2) eigenvector centrality (EC) as a measure of centrality in the global connectivity hierarchy, (3) amplitude of low-frequency fluctuations (ALFF) as a measure of oscillatory activity and (4) seed-based analyses to investigate functional connectivity from the ROIs. For (2)-(4), we applied a hypothesis-driven ROI approach in the hippocampus, caudate and putamen. In the luteal phase, we found (1) decreased intrinsic connectivity of the right angular gyrus with the default mode network, (2) heightened EC for the hippocampus, and (3) increased ALFF for the caudate. Furthermore, we observed (4) stronger putamen-thalamic connectivity during the luteal phase and stronger fronto-striatal connectivity during the pre-ovulatory phase. This hormonal modulation of connectivity dynamics may underlie behavioural, emotional and sensorimotor changes along the MC.

Predicting ovulation from brain connectivity: Dynamic causal modelling of the menstrual cycle

“…In women, frontal areas such as the IFG and MFG have shown enhanced activity during the pre-ovulatory phase and related to increased estradiol levels for implicit memory 61 and verbal encoding 62 . We recently found increased right frontrostriatal functional connectivity during resting state, related to higher estradiol levels during the pre-ovulatory phase 14 . The increased afferent connectivity we observed from the left hemisphere to bilateral MFG right before ovulation could reflect underlying structural changes, but in humans, the relationship between molecular mechanisms and neural functions are difficult to determine.…”

“…Saliva samples were collected from participants, stored and processed as described in Hidalgo-Lopez et al 14 . Estradiol and progesterone were assessed using the Salimetrics High Sensitivity salivary Estradiol assay (sensitivity of 1 pg/ml) and the DeMediTec Progesterone free in saliva ELISAs (a sensitivity of 10 pg/ml), respectively.…”

“…Although sparse, acute structural changes related to hormonal fluctuations have already been reported for both grey 9,10 and white matter 11 . Accordingly, connectivity patterns vary in women depending on the hormonal status, as shown by resting-state functional MRI (fMRI) [12][13][14] .…”

“…Within the DMN, intrinsic resting state connectivity has been shown to increase before ovulation 9,24 and decrease during the luteal phase to both left and right AG 14,25 . Within the SN, higher activity and connectivity have been consistently reported during the luteal cycle phase and related to higher progesterone levels, both at rest and during tasks 26,27 .…”

“…As several psychiatric and neurological disorders share an aberrant intrinsic organization of the aforementioned three "core" networks 19 , we consider it of the utmost importance to characterize their nonpathological directed organization in healthy women. In order to do so, we applied state-of-the-art effective connectivity analyses to a previously published longitudinal resting state-fMRI data set acquired during menses, pre-ovulatory and luteal cycle phase in a large sample of healthy young women 14 . Finally, we aimed to identify, whether changes in specific directed connections were able to predict which cycle phase a woman was in.…”