Fig. 2
From: Towards robust and replicable sex differences in the intrinsic brain function of autism
Sex-by-diagnosis interaction effect, its robustness and replicability. (a) On the right, surface maps show the cluster with a significant (Z > 3.1, P < 0.01) sex-by-diagnosis interaction for voxel-mirrored homotopic connectivity (VMHC) resulting from discovery analyses in the ABIDE sample using the component-based noise reduction (CompCor) pipeline. The statistical Z maps are overlaid on inflated brain maps generated by BrainNet Viewer. (b) The upper panels show the pattern of VMHC group means in males and females by each diagnostic group (ASD and NT) extracted from the same cluster in data pre-processed following two alternative denoising pipelines, Global Signal Regression (GSR, left) and Independent Component Analysis-Automatic Removal of Motion Artifacts (ICA-AROMA, right). Results show a pattern similar to the those observed in discovery analyses with small to moderate effect sizes (ηp2 range = 0.01–0.07). (c) The lower graph shows replicability in two independent samples: the Gender Explorations of Neurogenetics and Development to Advance Autism Research (GENDAAR) and the EU-AIMS Longitudinal European Autism Project (LEAP). The pattern of results was replicable in the EU-AIMS LEAP (N = 309) with a small effect size (ηp2 = 0.01) and had a negligible effect size in GENDAAR (N = 196; ηp2 < 0.01). For all graphs VMHC data are shown as residuals obtained after regressing out mean framewise displacement and age effects. L left, R right, A anterior, P posterior