To the Editor: Although the status of plasma high-density lipoprotein (HDL) levels as a marker of protection against cardiovascular disease is well established, the value of more detailed analyses of this class of lipoproteins is less certain. Stampfer et al. (Aug. 8 issue)1 evaluated the predictive strength of HDL subfractions in 246 men who had myocardial infarction and 246 controls. The authors concluded that measurement of the HDL subfractions (HDL2 and HDL3) was no more predictive than measurement of total HDL. This finding contrasts with those of case–control studies in which HDL2 appeared to be a. . .