Influence of intermediate-density lipoproteins on the accuracy of the Friedewald formula

Values of low-density lipoprotein (LDL) cholesterol (C) according to the Friedewald formula (Clin Chem 1972;18:499-502) were compared with those obtained by lipoprotein fractionation in 98 healthy subjects (control group), 135 specimens from patients with peripheral vascular and cerebrovascular disease (atherosclerotic group), and 45 with chronic renal failure on hemodialysis (CRF group). All had concentrations of total cholesterol between 3.23 and 7.76 mmol/L (1.25-3.00 g/L) and triglycerides less than 3.39 mmol/L (less than 3.00 g/L). The percentage error of calculated LDL-C was 4% in controls with a cholesterol/triglycerides (C/TG) ratio for very-low-density lipoprotein (VLDL) of 0.20, but greater than 60% in those with a (C/TG)VLDL ratio of 0.40. The percentage of error in sera of patients with atherosclerosis and chronic renal failure was higher than in controls with a similar mean (C/TG)VLDL ratio. The percentage of error of calculated LDL-C increases progressively with the increase in the C/TG intermediate-density lipoprotein (IDL) ratio, both in controls and in the atherosclerotic and CRF groups. Similar findings are observed when the mean percentage of error of measured LDL-C is evaluated. The percentage of error from calculated LDL-C in the atherosclerotic and CRF groups is significantly lower than that obtained by comparison of LDL-C separated by ultracentrifugation when the "broad cut" LDL (IDL plus LDL, both by ultracentrifugation) was used. The high percentage of errors found in the groups of patients studied underlines the need for caution when assessing the reliability of the Friedewald formula, particularly in cases in which disturbances in IDL composition are suspected.