Increased regional gray matter atrophy and enhanced functional connectivity in male multiple sclerosis patients

Evidence suggests that sex/gender is an important factor for understanding multiple sclerosis (MS) and that some of its neuropathological consequences might manifest earlier in males. In the present study, we assessed gray matter (GM) volume and functional connectivity (FC) in a sample of female and male MS patients (MSp) and female and male healthy controls (HCs). As compared to female MSp, male MSp showed decreased GM volume in the bilateral frontal areas and increased FC between different brain regions. Because both sets of changes correlated significantly and no differences in cognitive performance were observed, we suggest that the FC increase observed in male MSp acts as a compensatory mechanism for their more extensive GM loss and that it promotes a functional convergence between maleand female-MSp.


Introduction
There is growing evidence that sex and/or gender might be relevant factors that affect the prevalence, course, disability and treatment response of multiple sclerosis (MS) patients. Indeed MS incidence and prevalence are higher in females than in males [1].
Although MS onset commences at earlier ages in females than in males [2], once the disease is established, its progression seems faster and results into a higher degree of physical disability in males than in females [3,4].
The possible effects of sex and/or gender on neurocognitive MS alterations are less clear. Although being male has been proposed as a risk factor for cognitive impairment in MS patients [5], the results on brain neuropathological MS indices are conflicting, some studies have suggested a higher degree of brain damage in males [6][7][8], while others have failed to replicate these findings [9,10]. As these conflicting results might derive from gender interacting with other relevant demographic and clinical variables, we explored whether female and male MS patients, carefully matched for age, years of disease, level of education, intelligence quotient (IQ), neuropsychological performance and physical disability, differed in two indices of MS progression: gray matter (GM) atrophy and resting state functional connectivity (FC) changes.

Participants
From the MS population of the Hospital General of Castellón, 34 female and 22 male relapsing-remitting (RR) right-handed patients were selected to make up two experimental subgroups (Female-and-Male MSp), who were matched for several demographical and clinical variables (see Table 1). All the MS patients were relapseand steroid-free for at least 2 months prior to this study. Patients were neuropsychologically assessed using the Brief Repeatable Battery Neuropsychology test (BRB-N), from which Z-scores for four attentional domains (attentional/ executive; visual memory; verbal memory and fluency) were calculated as previously described [11]. The healthy control subgroups (Female-and-Male HC) included 28 females and 35 males that were right-handed, with no presence or history of psychiatric or neurologic diseases.
The study was approved by the Ethics Standards Committees of the Hospital General and the Universitat Jaume I, both of Castellón. All the participants gave informed written consent prior to participating.

MRI acquisition and preprocessing
Anatomical and resting-state functional images were acquired with a 1.5 T scanner (Siemens Avanto, Erlangen, Germany). An optimized VBM protocol using the Diffeomorphic Anatomical Registration Through Exponential Lie Algebra (DARTEL), included in SPM8 (Wellcome Trust Centre for Neuroimaging, London, UK), was performed to explore regional GM volume differences. Resting-state fMRI scans were pre-processed using the DPARSFA tool as previously described (see Supplementary

Statistical Analyses
T-tests were run to assess dyadic differences between groups for all the demographical and clinical variables. While no differences among them were found for any variable, both male-and female-MSp differed from their respective HC subgroups in age, years of education and BPF terms (See Table 1). Therefore, these three variables were used as nuisance variables in all the subsequent statistical analyses.
Regional GM volume differences were assessed by a SPM8 full-factorial design (p< 0.05 FWE cluster-corrected for the multiple comparisons; cluster extent of k=50 voxels). For the FC analysis, individual pair-wise correlation connectivity matrices were entered into a GLM model and t-tests were performed to compare the four groups of interest (p< 0.05 FDR corrected for the multiple comparisons with 100 000 permutations using the Network-based statistic toolbox, NBC). The relationship between the FC changes and GM volume in MS patients was assessed by Pearson correlation indices and the relationship between these two variables and cognitive performance were assessed by means of linear regression analyses (significance level: p< 0.05).

Seed-based Functional Connectivity
Although gender-related FC differences were found in the HC and MS groups, they differed in both location and direction. While the gender-related differences in the MS group were found on three intra-hemispheric pathways (right Olfactory Cortex -right

Relationship between GM loss and increased Functional Connectivity in MS patients
As previous studies [12,13] have suggested that increased FC is a secondary response to primary MS-related neuropathological changes in GM, we explored the relationship between both variables. We found that the FC scores on the three pathways, where we observed gender differences within MS patients ( Figure 1B), were inversely related to the GM volume in the two frontal areas, where the Male-and Female-MS subgroups also differed ( Figure 1A). This inverse relationship was statistically significant in five (of six) pair-wise correlations ( Figure 1C), whereas the FC scores on these pathways did not correlate with global atrophy as estimated by BPF or TLV (data not shown). Finally, to further support a possible functional relationship between GM loss and increased FC in MS patients, we found that the gender differences previously observed in FC faded when GM volume measurements were used as covariates in all these dyadic comparisons.

Functional-cognitive significance
Increased FC resulting from GM atrophy in MS patients has been interpreted in two different ways: Either as maladaptive and, as such, a sign of MS neuropathological progression [14] or, to the contrary, as a compensatory mechanism promoting normal cognitive competence of MS patients at initial stages of the disease [12,13]. To find out which of these two alternatives applies to our observations, we analyzed the relationship between FC, GM volume and cognitive performance in Male-and Female-MSp.
Male-and female-MSp did not significantly differ in any test of the BRNB (table 1) and scores of both groups were ranged within normality boundaries. However, their attentional Z scores were differentially related to GM volume and FC in the brain areas for which gender-differences had been previously identified (figure 1). More specifically, linear regression analyses revealed that attentional Z scores of Male-but not of Female-MSp could be satisfactorily predicted from these two sets of parameters (r=0.72; p<0.05 and r= 0.53, p=0.11, respectively) and that remaining GM volume in frontal areas and FC values between the left Parahippocampal Gyrus and the left Inferior Frontal Gyrus were positively associated to cognitive performance (see supplementary Table 3 for details). Taken together, our results suggest that the increased FC observed in Male-MSp in this neuroanatomical pathway acts as a compensatory mechanism that allows them to retain similar cognitive performance to that observed in Female-MSp despite having a higher degree of GM atrophy.

Discussion
The findings of the present study show that, even when matched for age, years of disease, level of education, IQ, neuropsychological performance and physical disability, Male-MSp display a higher degree of GM atrophy than Female-MSp. These results confirm those of previous studies which have suggested that males are more prone to MS brain damage [7,8] and/or that they show faster MS progression than females [3,4].
On the other hand, according to the recently proposed classification of sex/ gender differences [15], this first observation should be considered as a contingent (e.g. MSdependent) average divergence.
We also showed that Male-MSp exhibit more FC between temporal and inferior frontal areas, and between the right amygdala and the right olfactory cortex than Female-MSp.
Increased FC has previously been understood as a secondary neuropathological sign of MS progresion [14] or as a compensatory mechanism for early MS neuropathological changes [12,13]. Only the later interpretation would be coherent with several additional findings of our study: 1) the direct correlation found between FC and GM atrophy; 2) the fading gender-related differences in FC when GM volume was introduced as a covariate; 3) lack of differences in neurological and neuropsychological performance (see Table 1) between Male-and Female-MSp despite their different degrees of GM atrophy in several brain regions; 4) the positive association of remaining GM volume and FC scores with attentional capabilities in Male-but not in Female-MSp. Therefore, in our study increased FC seems to act as a compensatory mechanism for GM atrophy that allows Male-MSp to retain normal cognitive performance. Further, it should be noted that the gender-related differences in FC and GM atrophy observed in this study functionally compensate each other and, following a recently proposed classification of sex/ gender differences [15], they are better described as a contingent average convergence between Male-and Female-MSp.
Taken together, the present results and previous ones highlight the importance of taking into account sex/gender in MS research. Additional studies are needed to clarify whether these differences are primarily related to progression speed or to the extent of MS neuropathology, and whether sex (i.e. chromosomic complement, steroid hormone levels, etc.), gender (i.e. behavioral and sociological factors askew distributed in males and females), or both are responsible for these differences between Female-and Male-

MSp.
This research has been supported by grant P1-1B2014-15 provided to Cristina Forn by the Universitat Jaume I.