• Yiqiang Zhan, Ida K Karlsson, Robert Karlsson, Annika Tillander, Chandra A Reynolds, Nancy L Pedersen, and Sara Hägg.
• From the Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden (Y.Z., I.K.K., R.K., A.T., N.L.P., S.H.); and Department of Psychology, University of California, Riverside (C.A.R.).
• Circ. Res. 2017 Jul 21; 121 (3): 214-219.

RationaleObservational studies have found shorter leukocyte telomere length (TL) to be a risk factor for coronary heart disease (CHD), and recently the association was suggested to be causal. However, the relationship between TL and common metabolic risk factors for CHD is not well understood. Whether these risk factors could explain pathways from TL to CHD warrants further attention.ObjectiveTo examine whether metabolic risk factors for CHD mediate the causal pathway from short TL to increased risk of CHD using a network Mendelian randomization design.Methods And ResultsSummary statistics from several genome-wide association studies were used in a 2-sample Mendelian randomization study design. Network Mendelian randomization analysis-an approach using genetic variants as the instrumental variables for both the exposure and mediator to infer causality-was performed to examine the causal association between telomeres and CHD and metabolic risk factors. Summary statistics from the ENGAGE Telomere Consortium were used (n=37 684) as a TL genetic instrument, CARDIoGRAMplusC4D Consortium data were used (case=22 233 and control=64 762) for CHD, and other consortia data were used for metabolic traits (fasting insulin, triglyceride, total cholesterol, low-density lipoprotein cholesterol, fasting glucose, diabetes mellitus, glycohemoglobin, body mass index, waist circumference, and waist:hip ratio). One-unit increase of genetically determined TL was associated with -0.07 (95% confidence interval, -0.01 to -0.12; P=0.01) lower log-transformed fasting insulin (pmol/L) and 21% lower odds (95% confidence interval, 3-35; P=0.02) of CHD. Higher genetically determined log-transformed fasting insulin level was associated with higher CHD risk (odds ratio, 1.86; 95% confidence interval, 1.01-3.41; P=0.04).ConclusionsOverall, our findings support a role of insulin as a mediator on the causal pathway from shorter telomeres to CHD pathogenesis.© 2017 American Heart Association, Inc.

31 keywords...

hide…