Risk of misclassification with a non-fasting lipid profile in secondary cardiovascular prevention

Aims: Routinely fasting is not necessary for measuring the lipid profile according to the latest European consensus. However, LDL-C tends to be lower in the non-fasting state with risk of misclassification. The extent of misclassification in secondary cardiovascular prevention with a non-fasting lipid profile was investigated. Methods and results: 329 patients on lipid lowering therapy for secondary cardiovascular prevention measured a fasting and non-fasting lipid profile. Cut-off values for LDL-C, non-HDL-C and apolipoprotein B were set at < 1.8 mmo1/1, < 2.6 mmo1/1 and < 0.8 g/l, respectively. Study outcomes were net misclassification with non fasting LDL-C (calculated using the Friedewald formula), direct LDL-C, non-HDL-C and apolipoprotein B. Net misclassification < 10% was considered clinically irrelevant. Mean age was 68.3 8.5 years and the majority were men (79%). Non-fasting measurements resulted in lower LDL-C (- 0.2 0.4 mmo1/1, P < 0.001), direct LDL-C (-0.1 0.2 mmo1/1, P = 0.001), non-HDL-C (- 0.1 0.4 mmol/l, P = 0.004) and apolipoprotein B (-0.02 0.10 g/1, P = 0.004). 36.0% of the patients reached a fasting LDL-C target of < 1.8 mmo1/1 with a significant net misclassification of 10.7% (95% CI 6.4-15.0%) in the non-fasting state. In the non-fasting state net misclassification with direct LDL-C was 5.7% (95% CI 2.1-9.2%), 4.0% (95% CI 1.0-7.4%) with non-HDL-C and 4.1% (95% CI 1.1-9.1%) with apolipoprotein B. Conclusion: Use of non-fasting LDL-C as treatment target in secondary cardiovascular prevention resulted in significant misclassification with subsequent risk of undertreatment, whereas non-fasting direct LDL-C, nonHDL-C and apolipoprotein B are reliable parameters.