Background: Very few, if any, studies have assessed the ability of apolipoproteins to predict new-onset of congestive heart failure (HF) in statin-treated patients with coronary heart disease (CHD).
Aims: To employ the Incremental Decrease in End points Through Aggressive Lipid Lowering Trial (IDEAL) study database to assess the association of on-treatment lipoprotein components with prediction of HF events and to compare their predictive value with that of established risk factors such as hypertension and diabetes.
Methods: We used Cox regression models to study the relationships between on-treatment levels of apolipoproteins A1 and B to subsequent HF. Chi square information value from the log likelihood was used to compare the predictive value of lipoprotein components with established risk factors of HF.
Findings: In the IDEAL study, on-treatment apolipoproteins proved to be associated with the occurrence of new-onset HF. Variables related to low-density lipoprotein cholesterol (LDL-C) carried less predictive information than those related to high-density lipoprotein cholesterol (HDL-C), and apoA-1 was the single variable most strongly associated with HF. LDL-C was less predictive than both non-HDL-C (total cholesterol minus HDL-C) and apoB. The ratio of apoB to apoA-1 was most strongly related to HF after adjustment for potential confounders, among which diabetes had a stronger correlation with HF than did hypertension. ApoB/apoA-1 carried approximately 2.2 times more of the statistical information value than that of diabetes. Calculation of the net reclassification improvement index revealed that about 3.7% of the patients had to be reclassified into more correct categories of risk once apoB/apoA-1 was added to the adjustment factors. The reduction in risk by intensive lipid-lowering treatment as compared to usual-dose simvastatin was well predicted by the difference in apoB/apoA-1 on-treatment levels.
Interpretation: The on-treatment ratio of apoB/apoA-1 was the strongest predictor of HF in CHD patients of both IDEAL treatment arms combined, mostly driven by the strong association with apoA-1, whereas LDL-C and non-HDL-C were less able to predict HF outcome. The predictive information value contained within apoB/apoA-1 was about 2.2 times more than that of diabetes. Between-treatment group differences in HF were to a significant extent explained by on-treatment differences in apoB/apoA-1, mostly through the changes in apoB. We argue therefore, on-treatment lipoprotein components contribute to the overall future risk of HF in statin-treated patients with CHD.