LDL polygenic risk score predicts silent atherosclerosis
F F Staeger, D M Cayero, F J G Sanchez, F S Cabo, K K Iversen, M Boettcher, K F Kofoed, J Q Thomassen, B Ibanez, H Bundgaard, J J F Ortuno, R Frikke-Schmidt, R HasselbalchAbstract
Background
Atherosclerosis is a progressive condition characterized by a long asymptomatic phase. This period of silent atherosclerosis represents a potential window for screening and early treatment. However, most genetic investigations have focused on overt clinical events rather than on identifying individuals at risk during the preclinical stage, when preventive strategies may be most effective.
Purpose
To investigate the predictive power of polygenic risk of LDL cholesterol for silent atherosclerosis.
Methods
We included Danish individuals between the age of 18 and 70 years with no history of myocardial infarction, stroke, and atherosclerotic cardiovascular disease from the REACT study. We used 3D vascular ultrasound to measure the volume of atherosclerotic plaque in the carotid and femoral arteries bilaterally as a measure of silent atherosclerosis. We compared the predictive ability of polygenic risk scores for LDL cholesterol, PGSLDL, to the measured concentration using linear models for the peripheral plaque volume (PPV) in mm3 on a logarithmic scale, log-PPV, adjusted for age and sex.
Results
We included 1065 individuals (median age 54 years (IQR 46-61), 48% female) in this analysis. The model showed a significant effect of PGSLDL on log-PPV (beta=0.09 (0.016;0.17), p-value=0.018), whereas the measured LDL-cholesterol did not significantly predict log-PPV (beta=0.06 (-0.003; 0.14), p-value=0.059). We also showed that adding PGSLDL to the model (full model) compared to only having measured LDL-cholesterol (reduced) significantly improved the model fit (F-test; p-value=0.0183, F=5.58).
Conclusion
In this initial analysis including 1065 of a planned 16000 included, a PGS for LDL cholesterol significantly predicted silent atherosclerosis independently of measured LDL levels and demonstrated stronger predictive power than circulating LDL concentration per se. This could be an important element in future diagnostic algorithms for atherosclerosis prevention.