The real role of exercise versus medication in lipid profile of patients referred to a cardiac rehabilitation program.

We enjoyed the letter by Lakusic and Granec; however, our response is delayed because of unawareness of their letter. First, as we have mentioned in our paper, we have considered all other antilipid therapies, and not only those statins, which Lakusic et al. have considered. In our study, medications that were prescribed to patients were categorized to (i) directly affect lipid level (clofibrate, gemfibrazil, lovastatine, simvastatin, atrovasta-tin, nicotinic acid, and cholesteramin), (ii) indirectly affect lipid levels (glucocorticoids, thiazide, b-blocker, valproate and related drugs, garlic, estrogen and progesterone), and (iii) having no effect on lipids. It does not seem that the normal range of lipids has significant difference between our study and similar others. Moreover, publication bias in the literature is inevitable. It is obvious that journals usually tend to publish manuscripts with positive results. Therefore, more published papers with a positive effect of lipid-lowering medications do not necessarily prove the effectiveness of statins in increasing high-density lipoprotein (HDL) even after the deletion of the positive effect of exercise. We suggest a study of Hill’s criteria to show that statins definitely increase HDL. In addition, there are some differences between our study and that of Lakusic et al. Our program has been Comprehensive Cardiac Rehabilitation whereas their program has been Stationary Cardiac Rehabilitation. Sample size and duration of the studies were also different: 547 cases for 8 weeks (24 sessions) in our study versus 444 patients for 3 weeks (18 sessions). We should add difference in cultures, habits, exercise and nutritional status, activity level, alternative agents’ effect on lipid level such as some nutritional factors, compliance of the patients, adherence to treatment, proper use of the medications, etc. It is true that the difference in HDL levels is not significant in the group of patients who have received medication (group 1) but the statistical significance by itself is not important because the primary and final level of HDL in group 1 is higher than in patients without antilipid drugs (group 2). Therefore, this can explain why the differences in group 1 are not significant. Medication has had more efficiency in increasing HDL levels than exercise. In contrast, higher figures at baseline level may fail to show an increase later. Our baseline HDL cholesterol level was higher, which can be a reason for the difference between our study and that of Lakusic et al. However, we did a new analysis according to the baseline situation of cholesterol (normal vs. abnormal), end-of-treatment comparison, which did not show any changes in our results. When we considered statins as the only lipid-lowering medications, it showed significant effect in increasing HDL (mean standard error) in patients with high cholesterol at baseline (39.97 0.44 vs. 41 0.43, P1⁄4 0.029, N1⁄4403). But there was no significant difference in patients with normal cholesterol at baseline, who received statins as their only treatment. Lakusic and Granec mentioned that ‘Sarrafzadegan et al. found no significant differences in the value of lipids at the end of rehabilitation between groups of coronary patients who did and those who did not receive antilipid medications’. It is not logical to compare only the end of treatment between two groups in our study, while the baseline values are not similar. We are also completely in agreement with Lakusic and Granec about the effectiveness of cardiac rehabilitation effect on patients with cardiac morbidity as our