Utility of repeat serum cholesterol measurements for assessment of cardiovascular risk in primary prevention.

Abstract

Measurement of total cholesterol is part of cardiovascular disease risk assessment in primary prevention. Frequently, old total cholesterol values are available. We assessed whether previously measured total cholesterol values could be used to estimate cardiovascular disease risk in primary prevention. A longitudinal analysis using data of a population-based prospective cohort study (Study of Health in Pomerania) over a period of 10 years. Altogether, 1112 subjects (45% male, mean age 57.8 years, SD ± 6.8 years) without prior history of myocardial infarction or stroke were included. We calculated diagnostic test properties for SCORE-Germany estimates using 5- and 10-year-old total cholesterol values to predict high (≥5%) cardiovascular disease risk using estimates based on current total cholesterol as a gold standard. Mean total cholesterol decreased from 5.88 mmol/l (SD ± 1.21) at baseline to 5.73 mmol/l (SD ± 1.10) after 10 years. A high cardiovascular disease risk was found in 3.2% of subjects at baseline, in 4.9% after 5 years and in 16.2% after 10 years based on current total cholesterol values. SCORE-estimates using 5-year-old total cholesterol had a sensitivity of 90.9% (95% confidence interval 87.7-94.1) and specificity of 97.2% (95% confidence interval 96.4-98.0). For 10-year-old total cholesterol, sensitivity was 94.2% (95% confidence interval 91.6-96.8) and specificity 96.3% (95% confidence interval 95.3-97.2). We observed only minor changes of total cholesterol over time. SCORE-estimates using 5- or 10-year-old total cholesterol had a high sensitivity, specificity and test accuracy to identify subjects at high cardiovascular disease risk and resulted in low misclassification rates.