Aims: We propose to perform a meta-analysis of approximately 6,000 European children to i) replicate GWA SNP associations with fasting glucose levels from adult populations in the ALSPAC, European Youth Heart study (EYHS), Raine, Gene-Diet Attica investigation (GENDAI) studies and two samples of French children and ii) to compare effects on FG between adults and children as a way of exploring the age-dependency of SNP effects.

Following the discovery of T2D and continuous trait associated variants key questions remain owing to the cross-sectional nature of GWAS. Firstly, at what point in life does the genetic susceptibility conferred by these variants start to operate? And secondly, do the effects of these variants differ by age?

Studies in children provide a unique opportunity to answer both of these questions as we are able to evaluate whether variants significantly associated with type 2 diabetes risk or continuous traits in adults also display effects in children and additionally whether the effect size of these variants are different between adults and children.

Type-2-diabetes in children is still a rare disease and obtaining a sufficient number of paediatric cases hinders the investigation of the time-dependency of type 2 diabetes associated variants meaning the analysis of continuous related traits in healthy children is a more feasible approach.

GWAS have identified common genetic variants associated with FG levels with variants in four genes being identified as the first to be associated with FG levels in non-diabetic, healthy adults. G6PC2 (glucose-6-phosphatase, catalytic, 2) [1], GCK (glucokinase) [2], GCKR (glucokinase regulator) [3] all encode proteins which are directly involved in glucose metabolism, whereas an unexpected association was observed for MTNR1B (melatonin receptor 1B) [4, 5, 6], a protein expressed in pancreatic beta-cells and implicated in mediating melatonin induced impaired insulin secretion [5].

To further this work the Meta-Analyses of glucose and insulin related traits Consortium (MAGIC) was constituted to conduct large-scale meta-analyses of genome-wide data for continuous diabetes-related traits. In a meta-analysis of 21 GWAS the consortium have recently published data on 9 novel new loci (in or near ADCY5, MADD, ADRA2A, CRY2, FADS1, GLIS3, SLC2A2, PROX1 and FAM148B), which were shown to have genome wide significant association with FG levels in healthy adults [7].

We plan to analyse SNPs in G6PC2, GCK, GCKR and MTNR1B, in addition to new SNPs identified by the MAGIC study, in this meta-analysis of children.

We currently have 5 samples of European children amounting to a total of approximately 5,000 children and inclusion of data from ALSPAC will increase this to 6,000 children and hence greatly improve the power of our meta-analysis.

Plan of investigation:

Analyses:

The association between SNPs and untransformed fasting glucose (mmol/L) will be tested using linear regression models assuming an additive effect. Associations between SNPs and log-normalised outcomes (for insulin (pmol/l), HOMA-IR, HOMA-B) will also be analysed. All models will be adjusted for age, sex and BMI.

Random effects meta-analyses will be performed to combine results across studies.

The MRC Epidemiology Unit will take the lead in writing up the main findings from these new data, and we have already contacted ALSPAC regarding which authors to include in the authorship of all publications that include the requested ALSPAC genotypes.