We appreciate Dr. Ernst's comments. His study of the effects of gallopamil [1] was unfortunately not included in our meta-analysis [2]. The report does not appear on MEDLINE, and it was apparently published too recently to have been mentioned in the bibliographies of the papers that we reviewed. However, the results of his study are not inconsistent with our conclusion that calcium antagonists had little effect on plasma lipids. Dr. Ernst noted apparent improvements in total cholesterol, LDL cholesterol, HDL cholesterol, and triglyceride levels during the uncontrolled, open-treatment phase of the study [1]. During the subsequent placebo-controlled phase, however, no statistically significant differences in total, LDL, or HDL cholesterol levels were noted between the placebo and gallopamil groups; only triglyceride levels showed any change.

In Dr. Ernst's study, the failure of the placebo-controlled phase to confirm the lipid changes that occurred during the open-treatment phase could be caused by several factors. The run-in period may not have negated a regression-to-the-mean effect because the period was short (2 weeks), and patients were apparently excluded at the end of this run-in period if they did not have high plasma lipid levels [1]. Study patients may have shown subsequent improvements in lipid values during the uncontrolled phase of study because they were selected on the basis of high lipid levels, and the negative results of the placebocontrolled phase may more accurately reflect the lack of effect of gallopamil on lipids. Alternatively, gallopamil may have significant effects on lipids that were not detected in the placebo-controlled phase because of the small study sample. Large, controlled clinical trials are needed to determine whether any calcium antagonists significantly affect serum lipids and thus differ from the agents and studies included in our meta-analysis.