Normal cholesterol measurements in white collar workers still at cardiovascular risk?

Low serum total cholesterol (TC) in workers has been taken to indicate absence of cardiovascular (CV) risk. In angiographically confirmed coronary artery disease TC has been shown to be less poignant than compound indices of cardiovascular risk in separating patients from controls. The implications for prevention in an industrial medical setting of relying on TC measurements are discussed. Employees (n = 229) tested by an occupational health service in a non-manufacturing firm were dichotomized as low and high cardiovascular risk subjects either by the level of total cholesterol, or by two compound indices of blood lipid components. The compound indices were: the TC/HDLc-ratio, and an 'atherogenic index' (ATH-index) defined as ([TC-HDLc] x [apo B]) + ([HDLc] x [apo A]). (apo A = apolipoprotein A-I, apo B = apolipoprotein B). Cut-off values to separate between low- and high-risk subjects were defined as TC = 6.0 mmol/l, HDLc = 0.9 mmol/l, apo A = 1.8 milligrams and apo B = 1.3 milligrams, based on clinical guidelines in Norway. These individual cut-off values gave TC/HDLc and ATH-index cut-off values of 6.7 and 4.1, respectively. Assuming a more correct discrimination between persons at low- and high-risk, using compound lipid indices, both the number of persons given unnecessary advice on lifestyle changes or urged to take TC reducing medication, and the number of persons not treated on the basis of normal TC levels, would be reduced. Percentages of persons classified as TC-level-low risk, were adjusted using empirical data on sensitivity and specificity of the compound indices. Among employees with TC < 6.0mmol/l, 15% and 23% of women, and 12% and 19% of men would be classified as high-risk persons using the TC/HDLc-ratio or the ATH-index, respectively. Lack of prospective data on compound indices suggests cautious interpretation. TC values in spurious testing, as often applied in occupational health service without due regard to other lipid fractions, would increase the probability of unnecessarily treating persons not at CV risk and withholding people at probable CV risk from treatment. Although prospective studies are needed to confirm findings, the changes observed suggest avoiding measurements of some single lipid factors.