Abstract
BACKGROUND: Circulating biomarkers have been previously associated with atherosclerosis-related risk factors, but the nature of these associations is incompletely understood. METHODS: We performed multivariable-adjusted regressions and 2-sample Mendelian randomization analyses to assess observational and causal associations of 27 circulating biomarkers with 7 cardiovascular traits in up to 451 933 participants of the UK Biobank. RESULTS: After multiple-testing correction (alpha=1.3×10(-4)), we found a total of 15, 9, 21, 22, 26, 24, and 26 biomarkers strongly associated with coronary artery disease, ischemic stroke, atrial fibrillation, type 2 diabetes, systolic blood pressure, body mass index, and waist-to-hip ratio; respectively. The Mendelian randomization analyses confirmed strong evidence of previously suggested causal associations for several glucose- and lipid-related biomarkers with type 2 diabetes and coronary artery disease. Particularly interesting findings included a protective role of IGF-1 (insulin-like growth factor 1) in systolic blood pressure, and the strong causal association of lipoprotein(a) in coronary artery disease development (β, -0.13; per SD change in exposure and outcome and odds ratio, 1.28; P=2.6×10(-4) and P=7.4×10(-35), respectively). In addition, our results indicated a causal role of increased ALT (alanine aminotransferase) in the development of type 2 diabetes and hypertension (odds ratio, 1.59 and β, 0.06, per SD change in exposure and outcome; P=4.8×10(-11) and P=6.0×10(-5)). Our results suggest that it is unlikely that CRP (C-reactive protein) and vitamin D play causal roles of any meaningful magnitude in development of cardiometabolic disease. CONCLUSIONS: We confirmed and extended known associations and reported several novel causal associations providing important insights about the cause of these diseases, which can help accelerate new prevention strategies.