High-density lipoprotein cholesterol elevation with gemfibrozil: effects of baseline level and modifying factors

The effects of baseline level and modifying factors on gemfibrozil-induced high-density lipoprotein (HDL) cholesterol elevation were studied in 1028 participants with good compliance in the Helsinki Heart Study. The absolute (mmol/L) increment in HDL cholesterol was independent of baseline when the change in the placebo group (regression toward the mean) was considered. In contrast, absolute reductions in low-density lipoprotein (LDL) cholesterol and triglycerides correlated with their baselines, relative percentage changes being constant. Statistically, this could indicate differences in the mode of action of gemfibrozil: an independent and additive effect on HDL cholesterol and a multiplicative effect on LDL cholesterol and triglycerides. These differences may have a physiologic background because the main effect of gemfibrozil is in the stable HDL3 subfraction, rather than in the variable HDL2. Only 13% of the variation in gemfibrozil-induced HDL cholesterol changes were explained by modifying factors. The basic assumptions in the uses of absolute or relative changes as a measure of treatment effect are discussed.