Serum cholesterol in patients undergoing cardiac catheterization for suspected coronary artery disease: diagnostic and prognostic implications.

The Duke Data Bank for Cardiovascular Diseases is an observational data bank capturing baseline, treatment, and follow-up information on patients referred for evaluation of suspected coronary artery disease. From 6829 patients referred between 1969 and 1990 for diagnostic angiography, baseline cholesterol levels were determined as part of the standard clinical procedure. Baseline total cholesterol values were found to be related to the presence of coronary disease in men and women, although in the elderly little evidence for a significant relationship was found. Formal tests for interactions revealed a significant interaction between age and cholesterol but not between gender and cholesterol. In subgroup analyses using the Cox proportional hazards model, no relationship was found between cholesterol and either cardiovascular death or total cardiac events (death or nonfatal myocardial infarction) in 2038 medically treated patients with 5 to 20 years of follow-up. In the subgroup of 1798 surgically treated patients, there was also no relationship between baseline cholesterol level and these end points. Finally, detailed lipid analysis was done at baseline and 6-month angiography in a second population of 759 patients undergoing percutaneous coronary angioplasty between 1986 and 1989. No substantial relationship between baseline or follow-up cholesterol, low-density or high-density lipoproteins, or triglyceride levels and restenosis was found.

Duke Scholars

Author Kerry L. Lee Biostatistics & Bioinformatics, Division of Biostatistics