Establishing cut-off values for apolipoprotein B and non-HDL-C according to LDL-C values in a South European population

Background: Low-density lipoprotein cholesterol (LDL-C) remains the primary target of therapy in most strategies of dyslipidaemia management focused on cardiovascular disease prevention. Different guidelines have identified specific LDL-C cut-off points as targets for therapeutic intervention. Many clinical situations characterised by dyslipidaemia and elevated triglycerides are common in our environment and in overall industrialised countries. Thus, lipid goals based only on LDL-C could misclassify an important percentage of subjects. The objective of the present study was to establish cut-off point values for apoB and non-HDL-C in relation to the identified LDL-C cut-off points for cardiovascular risk in a South European population. Methods: We performed a cross-sectional study including 1501 subjects (770 women and 731 men) between 18 and 80 years of age. Samples were collected after 1214 h of fasting. Cholesterol, HDL-C, triglycerides and apoB levels were measured using direct methods. LDL-C was calculated by the Friedewald formula. Non-HDL-C was calculated as total cholesterol minus HDL-C. Results: The Spearmans rank correlations between apoB and LDL-C (r 0.86, p < 0.0001), and between apoB and non-HDL-C (r 0.91, p < 0.0001) were both significant. The proposed cut-off points for apoB, according to LDL-C goals (70, 100, 130 and 160 mg/dl) in our population are 70, 80, 100 and 115 mg/dl respectively. The proposed cut-off values for non-HDL-C are 100, 120, 150 and 190 mg/dl respectively. Conclusion: The established LDL-C cut-off values could not be accurate to estimate cardiovascular risk in subjects with mild hypertriglyceridaemia, as frequently occurs in our Mediterranean population. To take into consideration the burden of atherogenic particles and better classify patients at risk we propose cut-off values for apoB or the equivalent for non-HDL-C. Prospective trials including cardiovascular variables are needed to validate our assumption.