Background and aim: Dyslipidaemia is the main risk factor for coronary heart disease (CHD). Plasma lipid levels are conven­tionally used to predict coronary risk globally, but further studies are required to investigate whether the lipoprotein ratios are superior to conventional lipid parameters as predictors for CHD.

Methods: A hospital-based case-control study consisting of 738 CHD patients and 157 control subjects was conducted in a Chinese Han population. Demographic characteristics and plasma lipid or apolipoprotein data were collected. Univariate and multivariate logistic regression analyses were carried out to examine the relationship between the lipoprotein ratios and CHD risk.

Results: The CHD group had significantly higher age, non-high-density lipoprotein cholesterol (non-HDL-C), lipoprotein (a) [Lp(a)], triglyceride (TG)/HDL-C, total cholesterol (TC)/HDL-C, low-density lipoprotein cholesterol (LDL-C)/HDL-C, non-HDL-C/HDL-C, very low-density lipoprotein cholesterol (VLDL-C)/HDL-C, and apolipoprotein B100/apolipoprotein AI (apoB100/apoAI) than the control group (p < 0.05 for all). Moreover, the prevalence of male sex, smoking, and hypertension in the CHD group was significantly higher than in the control group. The results from univariate logistic regression analysis showed that the ratios of TC/HDL-C (OR 1.135, 95% CI 1.019–1.265), LDL-C/HDL-C (OR 1.216, 95% CI 1.033–1.431), non-HDL-C/HDL-C (OR 1.135, 95% CI 1.019–1.265), and apoB100/apoAI (OR 1.966, 95% CI 1.013–3.817) significantly increased the risk for CHD. By multivariate logistic regression analysis, the results were not materially altered and each of the four ratios was independently associated with CHD after adjustment for non-lipid coronary risk factors. ApoB100/apoAI showed the strongest association with CHD in both the univariate and multivariate logistic regression analyses.

Conclusions: Our data indicate that the lipoprotein ratios are superior to conventional lipid parameters as predictors for CHD. Of the ratios, apoB100/apoAI is the best to predict CHD risk.