PREDICTING HIGH-RISK CHOLESTEROL LEVELS

The pattern of longitudinal changes in cholesterol levels has importan t implications for screening policies and for understanding the role o f cholesterol as a risk factor for coronary heart disease. We explored a variety of longitudinal models to predict changes in cholesterol ov er several years, emphasizing the probability that an individual will develop a cholesterol level that requires further diagnostic tests or treatment. The first question was whether measured cholesterol is Mark ovian. A chi-square statistic based on the bootstrap and motivated by the Chapman-Kolmogorov equations established that it is not. Related b ootstrap-based tests indicate that the probability structure of measur ed cholesterol is not that of a low order autoregressive moving averag e (ARMA) model. We then tested several alternative models to predict f uture cholesterol levels from the pattern of previous measured values, using receiver-operating characteristic (ROC) curves to summarize the sensitivity and specificity of the resulting rules for predicting hig h risk values. One method was based on the Gaussian assumption that th e logarithms of cholesterol levels are jointly Gaussian; a second was based on ordinary least squares regression; a third was based on logis tic regression. We developed a bootstrap technique for finding confide nce regions for points on the ROC curves. Bootstrap simulations were u sed in three different ways in computing the regions: one to bias corr ect each point on a curve, a second to find the bootstrap distribution of points for each threshold that defines a particular value of sensi tivity and specificity, and a third to find the volume of the (ellipso idal) regions. The results of our analyses suggest that the models can be used to identify subgroups of individuals who are unlikely to deve lop very high risk levels of cholesterol. The models also can be used to help formulate schedules for screening individuals.