In V+, while dichotomized cut-offs did not discriminate lipid severity, the continuous 12-SNP score showed a positive trend with lipid severity in univariate analyses (Spearman ρ = 0.26 with non-HDL-C, p = 0.063; ρ = 0.23 with LDL-C calculated by Friedewald, p = 0.107).
Using the Friedewald equation, this association did not reach statistical significance in our cohort; however, when LDL-C was calculated using the Martin–Hopkins formula, the correlation with PRS became significant, suggesting that the choice of LDL-C estimation method may influence the detection of genotype-phenotype associations.