The association of inflammation, triglycerides and lipoprotein(a) with cardiovascular events in the UK Biobank

Abstract Background Recent trial data has suggested that high-sensitivity C-reactive protein (hsCRP) is, at least, as strong a predictor of future cardiovascular events as LDL cholesterol in patients on lipid-lowering therapy1,2. There are limited data however on how hsCRP and other additional risk factors perform in patients at different levels of CVD risk. Purpose To evaluate the relationship of hsCRP, triglyceride and lipoprotein(a) [Lp(a)] levels with cardiovascular events in guideline-defined, clinically distinct cohorts. Such a study will improve the understanding of CVD risk prediction and provide essential information on the role of these pathways in CVD. Methods Data from the UK Biobank were used. In accordance with the 2021 ESC CVD prevention guidelines, four risk groups were derived based on clinical data and 10-year CVD risk, calculated from the recently derived SCORE2 model (Figure 1)3. Those not falling into one of the four cohorts were excluded as further risk stratification in these individuals would not influence management. For increasing quartiles (Q) of each biomarker, hazard ratios (HR), 95% confidence intervals (CI) and p-values were calculated relative to the lowest Q, using Cox proportional hazards models, adjusted for age, sex, smoking status, systolic blood pressure, HDL cholesterol and total cholesterol. Interactions between sex and biomarker were explored. The primary endpoint was major adverse cardiovascular events (MACE), defined as a composite of cardiovascular death, myocardial infarction, and ischaemic stroke. Secondary endpoints were the individual components of the primary endpoint. Results A total of 229,339 participants were included across the four risk groups, with a total of 24,133 primary endpoint events over the mean follow-up period of 14 years. In all four groups there was a robust linear relationship of hsCRP with MACE (p<0.001 [Figure2]) – strongest in the secondary prevention population (Q4 v Q1 HR=1.79 [95%CI:1.60, 2.02]). Of the secondary endpoints, hsCRP was most strongly associated with cardiovascular death (Q4 v Q1 secondary prevention cohort HR=1.97 [1.68, 2.30]). Lp(a) was associated with MACE in the primary prevention cohorts only (Q4 vs Q1 high-risk cohort HR=1.35 [1.27, 1.44]), driven by its strong association with myocardial infarction (Q4 vs Q1 HR 1.61 [1.48, 1.75]) but not other secondary endpoints. After adjustment, triglyceride levels were not associated with MACE in the four cohorts. There was no evidence of sex interactions in any analysis (p>0.05). Conclusions After adjustment for traditional CVD risk factors, hsCRP is a more robust predictor of cardiovascular events than Lp(a) or triglycerides in both primary and secondary prevention settings. Consideration should be given to evaluating whether hsCRP, Lp(a) and other ‘non-traditional’ risk factors can improve CVD risk reclassification and/or be used to guide allocation to targeted anti-atherosclerotic therapies.Figure 1:Participant FlowFigure 2:Main effects