Novel findings from 2,838 Adult Brains on Sex Differences in Gray Matter Brain Volume

www.nature.com/scientificreports OPEN Received: 26 June 2018 Accepted: 18 December 2018 Published: xx xx xxxx Novel findings from 2,838 Adult Brains on Sex Differences in Gray Matter Brain Volume Martin Lotze 1, Martin Domin 1, Florian H. Gerlach1, Christian Gaser2, Eileen Lueders3,4, Carsten O. Schmidt5 & Nicola Neumann1 There is still disagreement among studies with respect to the magnitude, location, and direction of sex differences of local gray matter volume (GMV) in the human brain. Here, we applied a state-of-the-art technique examining GMV in a well-powered sample (n = 2,838) validating effects in two independent general-population cohorts, age range 21–90 years, measured using the same MRI scanner. More GMV in women than in men was prominent in medial and lateral prefrontal areas, the superior temporal sulcus, the posterior insula, and orbitofrontal cortex. In contrast, more GMV in men than in women was detected in subcortical temporal structures, such as the amygdala, hippocampus, temporal pole, fusiform gyrus, visual primary cortex, and motor areas (premotor cortex, putamen, anterior cerebellum). The findings in this large-scale study may clarify previous inconsistencies and contribute to the understanding of sex-specific differences in cognition and behavior. Many brain structures present themselves as similar in men and women, especially when properly accounting for total brain size. It has therefore been proposed that both, brain structure and behavior, is largely overlapping between sexes1. However, one cannot assume a total absence of sex differences with regard to brain features as evidenced by both single studies and meta-analyses. In a gaussian-process regression coordinate-based meta-analysis including 16 voxel-based morphometry (VBM) studies, altogether comprising of 2,186 brain scans, Ruigrok and colleagues2 reported larger gray matter volumes (GMV) in women within the frontal pole, inferior/middle frontal gyrus, planum temporale/parietal operculum, anterior cingulate gyrus, right insular cortex, Heschl’s gyrus, thalamus, precuneus, parahippocampal gyrus, and lateral occipital cortex. In men, GMV was larger for the amygdala, hippocampus, parahippocampal gyrus, precuneus, putamen and temporal poles, the cingulate gyrus, as well as cerebellum2. While meta-analyses have an enormous advantage over single studies in terms of statistical power, they are not immune to other pitfalls3,4, such as related to data acquisition, image analysis, and the (often manual) transfer of peak coordinates. Therefore, a study comparable in scale to meta-analyses, but not weakened by the typical methodological confounds would be desirable. On a similar note, studies investigating the reproducibility of sex effects in two independent cohorts using identical measurements and evaluation methods seem imperative. Recently, in an UK biobank study on 5216 participants Ritchie and colleagues5 presented data on sex differences in total brain volume (preselected subcortical structures), cortical thickness (cortical structures), white matter, resting state connectivity and cognitive testing. Although this is the first study on such a large dataset, including different characteristic measurements of brain structure, connectivity and cognition, the investigated sample was not representative (mean age 62, range 44–77) years, higher education over-represented). Roughly, Ritchie and colleagues described larger brain volume in all preselected subcortical areas (except n. accumbens) of both hemispheres in men and increased cortical thickness in women. Men showed larger variance of brain volume measures than women. However, an investigation of sex differences in the brain of a large representative cohort with (1) a broad age range, (2) correction for the total brain volume (TBV), (3) inclusion of a number of confounds known to affect brain volume, (4) using voxel-based morphometry for differentiating subcortical subregions is still lacking. 1 Functional Imaging Unit, Department of Radiology, University Medicine Greifswald, Greifswald, Germany. Department of Psychiatry, University of Jena, Jena, Germany. 3School of Psychology, University of Auckland, Auckland, New Zealand. 4Laboratory of Neuro Imaging, School of Medicine, University of Southern California, Los Angeles, USA. 5SHIP, Institute for Community Medicine, University Medicine of Greifswald, Greifswald, Germany. Correspondence and requests for materials should be addressed to M.L. (email: ) 2 Scientific Reports | (2019) 9:1671 | https://doi.org/10.1038/s41598-018-38239-2 1 www.nature.com/scientificreports/ For the general differences between sexes, such as larger TBV, GMV and WMV in men6 and local cortical differences between men and women in the fronto-parietal (women > men) and occipito-temporal cortex (men > women), there seems to be increasing support1,2,7. In contrast, for subcortical structures, such as the hippocampus, inconsistent results have been reported. This inconsistency might result from differences in methodological approaches (lack of correction for TBV, global structure volume assessment in comparison to regional volume changes in VBM), but also from differences in cohort selection (age, sample size). For instance Neufang et al.8 found that testosterone levels predicted hippocampal size in younger females having larger hippocampi. Whereas sex differences in puberty and early adulthood may be particularly modulated by hormonal factors, in older adulthood environmental factors may have a greater impact. Therefore differences in younger collectives might well be absent in older cohorts, and vice versa. Large cohorts with a broad age range (21–90 years) like the current one may have the power to detect sex differences in small structures, such as the hippocampus, as a function of age. Here, we compared male and female brains with respect to local (voxel-wise) gray matter volume in a large representative sample. First, we tested for reproducibility of effects in two independent cohorts (n1 = 967; n2 = 1,871). Since both showed highly reproducible results, the methods of collecting data were identical and the cohorts were not overlapping, we were able to combine both to perform one unified analysis. For this purpose, we applied a state-of-the-art brain mapping approach9 and analyzed 2,838 T1-weighted scans obtained from these two general population cohorts10. In addition, we investigated the interaction of sex and age for the GMV in the hippocampus, since previous findings with regard to this structure were highly controversial and an impact of age on the hippocampal GMV can be assumed. Results Cohort 1 (n = 967) and Cohort 2 (n = 1,871). Analyzing Cohort 1, in women (women > men), on average larger GMV was prominent in bilateral prefrontal areas, such as the ventrolateral prefrontal cortex (vlPFC, BA 47), the medial and lateral orbitofrontal cortex (OFC), the anterior cingulate cortex, the frontal pole, and t (...truncated)