Blood samples used to generate the DNAm CRP score were collected at the Generation Scotland baseline appointment (between 2006 and 2011) and DNA methylation was profiled using the Illumina Human-MethylationEPIC BeadChip in two different sets. Pre-processing and quality control steps for both sets of methylation data have previously been fully reported (Madden et al., 2020 , Stevenson et al., 2020 (link)).
Full details of the calculation of the DNAm CRP score in GS have been reported previously (Stevenson et al., 2020 (link)). Briefly, methylation beta values were extracted for 6 CpG sites shown to have the strongest evidence of a functional association with serum CRP levels as shown by Lighthart and colleagues (n = 8863 and 4111 of European and African ancestries, respectively) (Ligthart et al., 2016 (link)). One of the 7 CpG sites (cg06126421) in the original study was unavailable in the GS dataset and was therefore not included in the current analysis resulting in 6 CpG sites. The beta values for the six CpG sites associated with serum CRP were then multiplied by their respective regression weights and summed to generate a single score for each STRADL participant (Stevenson et al., 2020 (link)). As all the EWAS regression weights were negative, a higher DNAm CRP score corresponds to a score closer to zero.
Full details of the calculation of the DNAm CRP score in GS have been reported previously (Stevenson et al., 2020 (link)). Briefly, methylation beta values were extracted for 6 CpG sites shown to have the strongest evidence of a functional association with serum CRP levels as shown by Lighthart and colleagues (n = 8863 and 4111 of European and African ancestries, respectively) (Ligthart et al., 2016 (link)). One of the 7 CpG sites (cg06126421) in the original study was unavailable in the GS dataset and was therefore not included in the current analysis resulting in 6 CpG sites. The beta values for the six CpG sites associated with serum CRP were then multiplied by their respective regression weights and summed to generate a single score for each STRADL participant (Stevenson et al., 2020 (link)). As all the EWAS regression weights were negative, a higher DNAm CRP score corresponds to a score closer to zero.
Free full text: Click here
