DNAm was profiled using the Illumina Infinium EPIC Beadchip array (i.e. the ‘850 K array’) (98 (link)). Raw 850 K array data were extracted using GenomeStudio. Methylation data preprocessing was performed using the minfi R package (99 (link)). Minfi quality controls were applied to ensure that each sample had high median intensities in the methylated and unmethylated channels, had mean P-values access all probes that was less than 0.01 and that the sex of the patient sample was accurately predicted from methylation data. WateRmelon (100 (link)) was used to ensure that all samples had high (>94%) bisulfite conversion rates. CpGs were discarded from the analysis that had single nucleotide polymorphisms within either the interrogated CpG or at the single nucleotide extension, that represented cross-reactive probes that are listed within the maxprobes R package (101 (link)), or that had detection P-values of greater than 0.01 in more than 10% of samples. CpGs within sex chromosomes were excluded from all analyses in order to exclude biases associated with the analysis of sex chromosome DNAm in samples of mixed sexes.