Novel lipid profiles and atherosclerotic cardiovascular disease risk: insights from a latent profile analysis.

Abstract

Background: Dyslipidemia is a key contributor to atherosclerotic cardiovascular disease (ASCVD). Despite the well-established correlation between abnormal lipid metabolism and ASCVD, existing diagnostic and predictive models based on lipid indices alone or in combination often exhibit suboptimal sensitivity and specificity. There is an urgent need for improved lipid indicators or novel combinations thereof.

Methods: The study included 898 cardiology inpatients who underwent coronary angiography (CAG). A latent profile analysis (LPA) was utilized to delineate lipid profiles on the basis of four routine lipid indices (total cholesterol (TC), low-density lipoprotein (LDL), high-density lipoprotein (HDL), and triglycerides (TG)) and the triglyceride‒glucose (TyG) index as a proxy for the TG. Logistic regression models were used to assess the correlations between lipid profiles and the occurrence and severity of coronary artery stenosis (CAS and severe CAS), as well as the occurrence of coronary heart disease (CHD). Predictive modeling subsequently validated the predictive power of the lipid profiles for cardiovascular outcomes.

Results: The LPA delineated four distinct lipid profiles: Profile 1 (relatively high HDL with the lowest TC, LDL and TyG, 41.20%), Profile 2 (relatively high TC, LDL, and TyG with the lowest HDL, 36.42%), Profile 3 (relatively low TC, LDL and TyG with relatively high HDL, 18.93%), and Profile 4 (the highest TC, LDL, and TyG with the highest HDL, 3.45%). Profile 1 was associated with the lowest ASCVD risk, whereas Profile 2 posed the highest risk for all adverse outcomes. The risk associated with Profile 3 and Profile 4 varied depending on the outcome. Profile 4 presented a lower odds ratio (OR) for CAS than did Profile 3, whereas Profile 3 presented a lower OR for severe CAS and CHD. The lipid profile variable substantially outperformed individual lipid indices or their combinations in predicting cardiovascular outcomes.

Conclusions: Four distinct lipid profiles were identified among patients, with a particular profile characterized by lower levels of TC, LDL, and TyG, as well as a lower HDL, emerging as the most predictive of adverse cardiovascular outcomes. This underscores the critical need for a thorough lipid profile analysis to pinpoint individuals at heightened risk for adverse cardiovascular outcomes.