Authors: Philipp G Sand
Publish Date: 2014/06/15
Volume: 63, Issue: 9, Pages: 797-797
Abstract
In their recent contribution to this journal Li and coworkers investigated the pooled effects of two IL18 gene variants in periodontitis Based on data from 576 cases and 458 controls they claim significant associations for both variants with the phenotype under study At closer inspection however this claim is not tenableFirst the literature search conducted by the authors misses key investigations including the largest study published 1 Of the three studies that were retrieved by Li et al two clearly feature overlapping control populations yet were counted as separate samples 2 3Second information from earlier studies is grossly misinterpreted Thus the risklowering effect of the minor allele at rs187238 4 was mistaken for a riskenhancing effect causing further bias in the allelic and genotypic models see forest plots in 5 When the authors refer to the “C” allele at rs187238 this corresponds to the “G” allele on the forward DNA strand reference by consensus as in 6 as the primers in all three investigations align to the reverse strandFinally the authors have merged information from rs187238 and rs1946519 for sensitivity and publication bias analyses in Figs 4 and 5 5 This is not legitimate as the two SNPs are not in tight linkage disequilibrium and cannot be considered surrogate markers
Keywords: