OBJECTIVE To assess the validity of calculated low-density lipoprotein cholesterol by the Friedewald formula for management of lipoprotein abnormalities in patients with diabetes mellitus.

RESEARCH DESIGN AND METHODS Calculated LDL cholesterol by the Friedewald formula was compared with measured LDL cholesterol after separation by ultracentrifugation in 61 patients with type I diabetes, 50 patients with type II diabetes, and 116 healthy control subjects.

RESULTS Calculated LDL cholesterol coincided with measured LDL cholesterol, with < 10% error, in 54 (49%) patients with diabetes mellitus, and 85 (73%) control subjects. Calculated LDL cholesterol was overestimated, with an error of > or = 10% of measured LDL cholesterol in 39% of patients and 26% of control subjects, and underestimated in 13 and 1%, respectively. Despite a good correlation between calculated and measured LDL cholesterol, the intraclass correlation coefficients demonstrated a poor concordance between calculated and measured LDL cholesterol, both in patients and control subjects. When comparing the mean differences of calculated and measured LDL cholesterol for diabetic subjects versus control subjects, significantly greater differences in type II (but not type I) diabetic subjects were seen.

CONCLUSIONS Calculation of LDL cholesterol by the Friedewald formula may be inaccurate for assessment of cardiovascular risk in patients with type II diabetes and may not be appropriate for management of lipoprotein abnormalities in those diabetic patients.