The aim of this study was to evaluate and compare the role of lipid markers including total, low-density lipoprotein (LDL) and high-density lipoprotein (HDL) cholesterol with lipid indices (total/HDL cholesterol, LDL-cholesterol/HDL-cholesterol and non-HDL-cholesterol) as predictors of cardiovascular outcomes in adults over 30 years.Research design and method
In a nested case–control study, 207 cardiovascular events among participants of the Tehran Lipid and Glucose Study (TLGS) were documented during 3 years of follow-up. Those cases that were free of cardiovascular disease at baseline (132 subjects) were matched to 264 controls for age and sex. In all subjects, demographic and clinical data including blood pressure and anthropometric measurements as well as serum lipids, fasting and 2-h plasma glucose were obtained from the database of the TLGS. We estimated the relative risk for each lipid parameter in a multiple stepwise regression model after adjustment for family history of premature coronary heart disease, smoking, systolic and diastolic blood pressure, fasting and 2-h plasma glucose and waist-to-hip ratio.Results
The relative risks associated with an increase of approximately 1 SD of independent lipid predictors in the multivariate model were as follows: total cholesterol, 1.6 (1.2–2.1), SD=1.3 mmol/l; LDL-cholesterol 1.5 (1.1–2.0), SD=1 mmol/l; non-HDL-cholesterol 1.6 (1.2–2.1), SD=1.2 mmol/l and cholesterol/HDL-cholesterol 1.5 (1.1–2.0), SD=1.8. Comparison of these four independent variables with receiver-operating characteristic curve analysis showed no significant difference in their predictive power for cardiovascular outcome. There was no association between HDL-cholesterol, triglyceride and LDL/HDL cholesterol and cardiovascular disease outcomes in multivariate analysis.Conclusion
It seems that for short-term prediction of cardiovascular disease outcome, serum total cholesterol is the preferred lipid parameter to measure in the Iranian population.