Home
Diabetologia November 2016, Volume 59, Issue 11, pp 2369–2377
Abstract
Aims/hypothesis
Epidemiological studies have identified several traits associated with CHD, but few of these have been shown to be causal risk factors and thus suitable targets for treatment. Our aim was to evaluate the causal role of a large set of known CHD risk factors using single-nucleotide polymorphisms (SNPs) as instrumental variables.
Methods
Based on published genome-wide association studies (GWASs), we estimated the associations between the established risk factors (blood lipids, obesity, glycaemic traits and BP) and CHD with two complementary approaches: (1) using summary statistics from GWASs to analyse the accordance of SNP effects on risk factors and on CHD; and (2) individual-level analysis where we constructed genetic risk scores (GRSs) in a large Finnish dataset (N = 26,554, CHD events n = 4016).
We used a weighted regression-based method for summary-level data to evaluate the causality of risk factors. The associations between the GRSs and CHD in the Finnish dataset were evaluated with logistic and conditional logistic regression models.
Results
The summary-level data analysis revealed causal effects between glycaemic traits (insulin and glucose) and CHD. The individual-level data analysis supported the causal role of insulin, but not of glucose, on CHD. The GRS for insulin was associated with CHD in the Finnish cohort (OR 1.06 per SD in GRS, 95% CI 1.02, 1.10, p = 0.002).
Conclusions/interpretation
These results support the causal role of insulin in the pathogenesis of CHD. Efficient treatment and prevention of insulin resistance is essential to prevent future CHD events.
Link to this latest information here https://link.springer.com/article/10.1007/s00125-016-4081-6
Abstract
Aims/hypothesis
Epidemiological studies have identified several traits associated with CHD, but few of these have been shown to be causal risk factors and thus suitable targets for treatment. Our aim was to evaluate the causal role of a large set of known CHD risk factors using single-nucleotide polymorphisms (SNPs) as instrumental variables.
Methods
Based on published genome-wide association studies (GWASs), we estimated the associations between the established risk factors (blood lipids, obesity, glycaemic traits and BP) and CHD with two complementary approaches: (1) using summary statistics from GWASs to analyse the accordance of SNP effects on risk factors and on CHD; and (2) individual-level analysis where we constructed genetic risk scores (GRSs) in a large Finnish dataset (N = 26,554, CHD events n = 4016).
We used a weighted regression-based method for summary-level data to evaluate the causality of risk factors. The associations between the GRSs and CHD in the Finnish dataset were evaluated with logistic and conditional logistic regression models.
Results
The summary-level data analysis revealed causal effects between glycaemic traits (insulin and glucose) and CHD. The individual-level data analysis supported the causal role of insulin, but not of glucose, on CHD. The GRS for insulin was associated with CHD in the Finnish cohort (OR 1.06 per SD in GRS, 95% CI 1.02, 1.10, p = 0.002).
Conclusions/interpretation
These results support the causal role of insulin in the pathogenesis of CHD. Efficient treatment and prevention of insulin resistance is essential to prevent future CHD events.
Link to this latest information here https://link.springer.com/article/10.1007/s00125-016-4081-6
