A statistical approach to study the influence of sex and anger on the brain structure

The human brain is still not fully understood. Analysing sex differences in grey matter volume (GMV), may help explain sex-variability on the prevalence in neuropsychiatric disorders. Dis-orders such as major depression disorder (more frequent in women) and intermittent explo-sive disorder (more frequent in men) are partially characterised by emotional symptoms such as sadness or anger. Understanding whether, emotions, sex and GMV are connected in healthy adults could provide a better understanding of possible influences of emotion and sex on brain structure. The aim of this thesis was to examine the effects of sex, Anger/Hostility, and their interaction on GMV. Data from the Human Connectome Project (young adults), were analysed. We used Subsample that only contained unrelated participants (Subsample A: n=424, 228 females, 196 males, age 22-36 years). GMV was estimated from T1-weighted magnetic resonance images and was averaged over 68 cortical parcels. The influence of a confounder, the estimated total intracranial volume (eTIV), was removed from each brain re-gion using linear regression. Both raw and residuals GMV data were analysed. After testing the normality of the data (Shapiro-Wilk-Test), linear regression models with sex, Anger/Hos-tility and their interaction as predictors were performed for each brain region. Multiple com-parisons were corrected with Bonferroni. To account for a possible bias from floor values of Anger/Hostility, analyses were repeated on a second Subsample (Subsample B: n=375, 204 females, 171 males) in which subjects with the lowest value of Anger/Hostility were excluded. Sex had a significant main effect, for raw data throughout the brain cortex (67/68 regions), with higher volume in men, than in women. After accounting for the eTIV, four regions re-mained significant for main effects of sex (bilateral Insula, right Inferiorparietal, right Lateral-occipital). No significant main effect of Anger/Hostility, or its interaction with sex were found, for raw or residuals data. In Subsample B, only the main effect of sex was significant, with the same regions as Subsample A for raw data, and only one region remained significant for the main effect of sex (right Lateraloccipital) for residual data. The findings indicate widespread sex-differences in raw GMV, while a sex-related GMV difference beyond the brain size ap-peared only in four regions. Moreover Anger/Hostility showed no measurable effect on GMV, or interaction with sex on GMV, in healthy adults. This suggests that associations may only be found in a clinical group. Future studies should examine the effect of other emotional vari-ables and other brain properties.