The National Cholesterol Education Program has recently
published guidelines for the assessment of cardiovascular risk
and goals for laboratory accuracy. To test the impact of biologic
and analytic variability on the ability of a single lipid measurement to assess risk accurately, lipids were measured on three
occasions in 51 volunteers. Notable day-to-day variability of total
cholesterol (5%), triglyceride (20%), high-density lipoprotein
cholesterol (10%), and calculated low-density lipoprotein cholesterol (8%) levels was found. Analytic variability contributed significantly to total variability of high-density lipoprotein cholesterol levels and calculated low-density lipoprotein cholesterol
levels. Confidence intervals constructed around National Cholesterol Education Program cutoff points suggested that classification was reliable from a single measurement if total cholesterol
value was below 4.78 (<185 mg/DL), between 5.56 and 5.81 (215
and 225 mg/DL), or above 6.59 mmol/L (>225 mg/dL). Low-density lipoprotein cholesterol value classification from a single measurement was only accurate at below 3.00 (<1166 mg/dL) or above
4.50 mmol/L (>174 mg/dL). This study documents significant
day-to-day variability of serum lipids and suggests that patients
near the National Cholesterol Education Program cutoff points
may require repeated measurements to assign risk accurately.
(Arch Intern Med. 1990;150:1653-1657)