A gene variant near ATM is significantly associated with metformin treatment response in type 2 diabetes: a replication and meta-analysis of five cohorts

Abstract

Aims/hypothesis: In this study we aimed to replicate the previously reported association between the glycaemic response to metformin and the SNP rs11212617 at a locus that includes the ataxia telangiectasia mutated (ATM) gene in multiple additional populations.

Methods: Incident users of metformin selected from the Diabetes Care System West-Friesland (DCS, n = 929) and the Rotterdam Study (n = 182) from the Netherlands, and the CARDS Trial (n = 254) from the UK were genotyped for rs11212617 and tested for an association with both HbA(1c) reduction and treatment success, defined as the ability to reach the treatment target of an HbA(1c) ≤ 7 % (53 mmol/mol). Finally, a meta-analysis including data from literature was performed.

Results: In the DCS cohort, we observed an association between rs11212617 genotype and treatment success on metformin (OR 1.27, 95% CI 1.03, 1.58, p = 0.028); in the smaller Rotterdam Study cohort, a numerically similar but non-significant trend was observed (OR 1.45, 95% CI 0.87, 2.39, p = 0.15); while in the CARDS cohort there was no significant association. In meta-analyses of these three cohorts separately or combined with the previously published cohorts, rs11212617 genotype is associated with metformin treatment success (OR 1.24, 95% CI 1.04, 1.49, p = 0.016 and OR 1.25, 95% CI 1.33, 1.38, p = 7.8 × 10(-6), respectively).

Conclusions/interpretation: A gene variant near ATM is significantly associated with metformin treatment response in type 2 diabetic patients from the Netherlands and the UK. This is the first robustly replicated common susceptibility locus found to be associated with metformin treatment response.

Fig. 1

Association between rs11212617, metformin treatment response and treatment HbA_1c in different cohorts. The grey squares and horizontal lines indicate (a) the cohort-specific ORs and 95% CI for the ability to reach the treatment target of HbA_1c ≤7% (53 mmol/mol) and (b) the cohort-specific β-coefficients and 95% CI for the treatment HbA_1c as a continuous variable. The size of the squares is proportional to the weights of the studies. GoDARTS and UKPDS data were from Zhou et al [8]