The data was provided by the UK Biobank and was already subject to quality control59. We excluded individuals with IDP values outside a range of four standard deviations (SDs). Given the large sample size, we chose this lenient threshold as a non-trivial number of observations (97 for GM, 127 for WM) fall between three and four SDs away from the mean. The IDPs beyond the four SD range are likely the results of processing errors, or the corresponding individuals present severe brain irregularities (5 individuals for GM, 7 for WM). Note that excluding these outliers does not change the statistical significance or magnitude of our reported effects. The exclusion of individuals falling within three SDs of the mean does not change the results either.
