Triglyceride-rich lipoproteins cholesterol, 10-year atherosclerotic cardiovascular disease risk, and risk of myocardial infarction and ischemic stroke.

ABSTRACT

BACKGROUND: Triglyceride-rich lipoproteins cholesterol (TRLs-C) has been associated with atherosclerotic cardiovascular disease (ASCVD), even among individuals with low-density lipoprotein cholesterol in the targeted range. We assessed whether the associations of TRLs-C with myocardial infarction (MI) and ischemic stroke (IS) vary by the burden of traditional cardiovascular risk factors, as reflected by predicted 10-year risk for ASCVD. MATERIALS AND METHODS: Included were 327,899 participants from the UK Biobank who were free of MI or IS and did not receive lipid-lowering treatment at baseline. Ten-year risk for ASCVD was estimated by the Pooled Cohort Equations and was grouped as low (<7.5%), intermediate (7.5% to <20%), and high risk (≥20%). Multivariable Cox regression models were used to examine the associations of TRLs-C and triglycerides with risk of MI and IS, overall and by the 10-year risk categories. RESULTS: During a median of 12.3 years of follow-up, 8,358 incident MI and 4,400 incident IS cases were identified. Overall, higher TRLs-C was associated with a higher risk of MI (p-trend <.0001) but not IS (p-trend = 0.074). Triglycerides and non-HDL-C levels provide generally similar results. There was evidence for interactions between TRLs-C, triglycerides, non-HDL-C and 10-year ASCVD risk on risk of MI. However, the interaction was only between TRLs-C, triglycerides and10-year ASCVD risk on risk of IS. Hazard ratios (95% CIs) of MI comparing the highest with the lowest quartiles of TRLs-C, triglycerides, non-HDL-C were 2.10 (1.23-1.30), 2.02 (1.80-2.27) and 2.17 (1.93-2.44) in the low-risk group. The corresponding estimates for IS were 1.24 (1.05-1.45) 1.25 (1.06-1.47) and 1.08 (0.92-1.27) respectively. CONCLUSIONS: The associations of TRLs-C with MI and IS were significant in the low-risk group. Triglycerides and non-HDL cholesterol are roughly equivalent to TRLs-C in determining risk. These findings may have implications for more detailed risk stratification and early intervention.