Examining the causal association of fasting glucose with blood pressure in healthy children and adolescents: a Mendelian randomization study employing common genetic variants of fasting glucose

Abstract

The aim of the study was to determine whether genetically raised fasting glucose (FG) levels are associated with blood pressure (BP) in healthy children and adolescents. We used 11 common genetic variants of FG discovered in genome-wide association studies (GWAS), including the rs560887 single-nucleotide polymorphism (SNP) located in the G6PC2 locus found to be robustly associated with FG in children and adolescents, as an instrument to associate FG with resting BP in 1506 children and adolescents from the European Youth Heart Study (EYHS). Rs560887 was associated with increased FG levels corresponding to an increase of 0.08|[thinsp]|mmol|[thinsp]|l|[minus]|1 (P|[equals]|2.4 |[times]| 10|[minus]|8). FG was associated with BP, independent of other important determinants of BP in conventional multivariable analysis (systolic BP z-score: 0.32 s.d. per increase in mmol|[thinsp]|l|[minus]|1 (95|[percnt]| confidence interval (CI) 0.20–0.44, P|[equals]|1.9 |[times]| 10|[minus]|7), diastolic BP z-score: 0.13 s.d. per increase in mmol|[thinsp]|l|[minus]|1 (95|[percnt]| CI 0.04–0.21, P|[equals]|3.2 |[times]| 10|[minus]|3). This association was not supported by the Mendelian randomization approach, neither from instrumenting FG from all 11 variants nor from the rs560887, where non-significant associations of glucose with BP were observed. The results of this study could not support a causal association between FG and BP in healthy children and adolescents; however, it is possible that rs560887 has pleiotropic effects on unknown factors with a BP lowering effect or that these results were due to a lack of statistical power.