点击显示 收起
1 From the Carolinas Medical Center Department of Family Medicine (MFD) and Center for Cardiovascular Health (LFH and TAB), Charlotte, NC, and The R Stuart Dickson Institute for Health Studies, Charlotte, NC (HCS)
2 Supported by a grant from The Charlotte-Mecklenburg Health Services Foundation. The medication and placebo were provided by Garuda International (Lemon Cove, CA). 3 Address reprint requests to M Dulin, PO Box 32861, Charlotte, NC 28232-2861. E-mail: michael.dulin{at}carolinashealthcare.org.
ABSTRACT
Background: Policosanol is one of the fastest growing over-the-counter supplements sold in the United States. The use of policosanol to treat elevated cholesterol is based on clinical trials conducted in Cuba, which showed sugar cane–derived policosanol to be similar in efficacy to statins. Recent studies have challenged these findings, but there have been no trials conducted in North America that have examined the ability of sugar cane–derived policosanol to lower cholesterol.
Objectives: This study investigated the efficacy of sugar cane–derived policosanol in healthy adults with mild hypercholesterolemia. The primary outcome was the percentage change in LDL cholesterol after 8 wk of therapy. Secondary outcome measures included changes in total cholesterol, HDL cholesterol, triacylglycerols, C-reactive protein, and nuclear magnetic resonance-determined lipoprotein profile. Dietary habits, weight, and blood pressure were also monitored.
Design: Ambulatory, community-dwelling healthy adults with mild hypercholesterolemia (n = 40) were assigned to receive oral policosanol (20 mg) or placebo once daily for 8 wk. This was a double-blind, randomized controlled trial conducted from January through August 2005.
Results: No significant differences in the change in LDL cholesterol were observed between the placebo (n = 20) and policosanol (n = 20) groups. Also, no significant changes in secondary outcome measures, including total cholesterol, HDL cholesterol, triacylglycerol, C-reactive protein, and nuclear magnetic resonance spectroscopy–determined profiles were observed. Policosanol was well tolerated, and no significant adverse events were noted.
Conclusion: Policosanol does not alter the serum lipid profile over an 8-wk period in adults with mild hypercholesterolemia.
Key Words: Cholesterol • policosanol • aliphatic alcohols • LDL • HDL • nuclear magnetic resonance spectroscopy
INTRODUCTION
Policosanol is the commonly used name for a mixture of long-chain aliphatic alcohols originally derived from purified sugar cane wax (1). The purported ability of this drug to lower cholesterol without significant side effects has made it one of the fastest growing over-the-counter supplements in the United States (2). It is widely available in health food stores and has even been added to a major multivitamin preparation (3). Policosanol has been used in Cuba since 1991, and until 2004 virtually all published medical literature on policosanol had been authored by one research group based in Havana (4-18). These initial studies indicated that sugar cane–derived policosanol had similar efficacy to statins in its ability to lower total and LDL cholesterol and even greater efficacy in raising HDL cholesterol without any significant side effects (19). The underlying mechanism of action by which policosanol lowers cholesterol has not been definitively elucidated but is proposed to include inhibition of cholesterol synthesis by down-regulating the cellular expression of hydroxymethylglutamyl coenzyme A reductase (20, 21). The cholesterol-lowering response appears to be dose dependent within the range of 2–40 mg (22). Recently, the beneficial effect of policosanol on cholesterol has come into question with the publication of several negative trials. In 2006, 2 studies were published that revealed no changes in lipid variables in subjects who received policosanol (23, 24). These studies examined both the Cuban-derived policosanol preparation in doses ranging from 10 mg to 80 mg (24) and a commercially available poicosanol supplement (Octa-60; MZL Investments CC, Cape Town, South Africa) in a 20-mg once daily dose (23, 24). In 2004, a Dutch research group tested a standard 20-mg dose of policosanol derived from wheat germ and also found no effect (25). Finally, 2 animal studies also showed no significant effects of sugar cane–derived policosanol on lipid concentrations (26, 27). However, more independent studies are needed to verify the inability of policosanol to lower cholesterol and offset the large number of positive studies that have been published to date. We conducted a randomized, placebo-controlled trial to evaluate whether a 20-mg dose of sugar cane–derived policosanol could lower cholesterol in healthy participants with mild hypercholesterolemia who maintained a typical American diet.
SUBJECTS AND METHODS
Participants
Healthy male and female volunteers aged 18 y with primary hypercholesterolemia were recruited from the Charlotte, NC metropolitan area between January and June 2005. Participants were recruited through the newly formed, practice-based research network, the Mecklenburg Area Partnership for Primary-care Research (MAPPR). The participants were required to have an LDL-cholesterol concentration between 3.37 and 5.19 mmol/L (130 to 200 mg/dL), with fasting triacylglycerol concentrations <3.39 mmol/L (300 mg/dL). Exclusion criteria included the following: any history of heart disease (history of abnormal EKG, abnormal stress test, myocardial infarction, angina, or heart failure), clinical atherosclerotic disease, Framingham risk score >20%, cerebral vascular disease, diabetes, renal impairment, uncontrolled hypertension (blood pressure consistently >140 systolic/90 diastolic), untreated or clinically evident thyroid disorder, women who were pregnant or breastfeeding, and use of any prescription medications or dietary supplements known to alter LDL cholesterol (specifically sterol and stanol products, psyllium-based fiber supplements, or red yeast rice) within 30 d before screening. Vitamin, mineral, and other food supplements such as fish oil were permitted provided the dose was small (<1 g n–3 fatty acids) and consumption was consistent. The study protocol was approved by the institutional review board at Carolinas Medical Center and registered with the US National Library of Medicine (Protocol Number NCT00255216). Procedures followed were in accord with the ethical standards of the Helsinki Declaration of 1975 as revised in 1983. All subjects were advised of the potential side effects of the study medication and possible harm in participation. Written informed consent was obtained from each participant before participation in the trial.
Study design
The participants were initially screened for eligibility with a phone call. Those who had a qualifying lipid panel that had been drawn in the last 6 mo were allowed to enter the study; otherwise, a repeat screening lipid profile was drawn before allowing enrollment in the study. The participants were stratified by sex and randomly assigned to take either placebo or policosanol once daily. The participants were assigned to either supplement A or B according to a computer-generated blocked randomization schedule that was stratified by sex. The study coordinator assigned the participants to the supplement specified by the schedule in chronological order as the participants enrolled. Assignment into each study group remained concealed until participant enrollment was completed. A safety monitor who was otherwise uninvolved with the study held the key to the blinded assignments in a sealed envelope until all results were obtained and analyzed.
Study assessments took place 1 wk before starting the medication (week –1), at baseline (week 0), week 8, and week 9. The study participants were asked to continue their usual dietary habits and not pursue any changes in food selections or exercise that could change their lipid concentrations. Diet stability was verified by using a standardized diet questionnaire (The Diet Habit Survey), which was administered at week 1 and at week 8 (28). This survey was designed to assess changes in food selections that could result in a change in plasma lipids. A phone call to the participants at 4 wk asked about medication side effects and compliance. During each study visit, the participants were weighed, asked about adverse events, had blood samples drawn, and their pulse and blood pressure were monitored. Due to cost limitations, the nuclear magnetic renosance (NMR) spectroscopy and alanine transaminase (ALT) measurements were preformed only once at the beginning and end of the study. The lipid profiles and C-reactive protein (CRP) measurements were performed twice at baseline and during the final 2 visits to help reduce variability in these measures. Pill counts were performed at the final 2 visits as a measure of compliance. A modest compensation was provided to the participants at the end of the study to compensate them for their time.
Study capsules
Each capsule was standardized to contain either 10 mg policosanol (intervention) or maltodextrin (placebo), and the participants were asked to take 2 capsules each day. As part of quality control, an independent laboratory (Eurofins Scientific, Petaluma, CA) was contacted to perform HPLC to verify that each caplet contained the described amount of policosanol or maltodextrin. This analysis revealed that the study caplets contained 63% octacosanol (C28), 14% triacosanol (C30), 6% hexacosanol (C26), and 8% other long-chain alcohols. A concurrent analysis of the Cuban policosanol product obtained from Dalmer Labs (La Habana, Cuba) showed that to contain 66% octacosanol, 13% triacosanol, and 7% hexacosanol. The supplement was provided by Garuda International Inc (Lemon Cove, CA). The placebo and policosanol caplets were identical in color, shape, texture, and taste. Masking was assessed by questionnaire at the conclusion of the study. The policosanol and placebo tablets were labeled only as medication A and B before shipment to the investigators. A sealed envelope with the identity of supplements A and B was shipped directly to the study safety monitor. Blinding was maintained until after the initial data analysis was performed.
Lipoprotein analyses
Screening of lipid panels was performed at the Carolinas Medical Center laboratory and was processed immediately to determine patient eligibility for the study. These data were not included in the study and were only used for enrollment purposes. All other samples were put into aliquots and frozen at –70 °C until all participants had completed the study. An accredited commercial laboratory (LipoScience Inc, Raleigh, NC) then processed all samples concurrently by using Center for Disease Control and Prevention standards. Plasma samples for lipid and lipoprotein variable measurements were collected after a 12-h fast and stored in EDTA-coated tubes. Plasma concentrations of total cholesterol, HDL cholesterol, and triacylglycerols were measured by conventional enzymatic method, and LDL cholesterol was estimated by using the Friedewald equation. Lipoprotein subclasses were determined by NMR spectroscopy with the use of previously described methodology (29, 30).
Assessment of adverse events and side effects
Diaries were provided along with the study medication to determine potential medication side effects. The participants were counseled about potential side effects and requested to call the study investigators or seek medical attention for major symptoms or adverse events. In addition, a pharmacist who served as the safety monitor was asked to review symptoms in consultation with the investigators to determine whether symptoms could potentially be related to the study medication, and, if needed, to break the blinded assignements. Otherwise, minor symptoms were recorded in the diary as they occurred. Study personnel reminded the participants to record their symptoms with a phone call midstudy and at the last 2 clinic visits. All significant adverse events were reported to the center’s institutional review board.
Statistical analysis
The hypothesis for the present study was that 20 mg policosanol taken daily for 8 wk would lower LDL cholesterol by 20%. The primary outcome of the study was percentage change in LDL-cholesterol concentration. The average of lipid measurements taken at baseline and week 1 and then repeat measurements at weeks 8 and 9 were used as the pre- and postintervention values, respectively. Secondary outcomes were percentage changes in total cholesterol; HDL cholesterol; triacylglycerols; CRP; NMR spectroscopy measurements of LDL, HDL, and VLDL particle size; and blood pressure. ALT was measured at baseline and at completion of the study as a marker of potential hepatotoxicity. Changes in body weight, blood pressure, and total scores from the Diet Habit Survey were monitored during the study to evaluate for potential behavior changes that might confound results.
The sample size for the study was based on anticipated LDL-cholesterol reductions of 20% in participants treated with policosanol and 4% in participants in the placebo arm. We estimated that a sample size of 20 per group would provide a minimum of 90% power to detect a difference between groups of at least this magnitude using a 2-tailed t test with significance level of 0.05.
Continuous variables were assessed for normality and analyzed by using the techniques described below. Between-treatment comparisons were done to verify that randomization had produced sufficiently homogenous groups. The average difference in LDL cholesterol and other variables in the 2 treatment arms before and after treatment and the average before-after change were assessed with a Student’s t test. The within-group change in LDL cholesterol and other variables from baseline to study completion was assessed with paired t tests or Wilcoxon’s signed-rank tests. A significance criterion of 0.05 was used in all cases, and all tests were two-sided. The SAS System (SAS Institute, Cary, NC) was used for all analyses.
RESULTS
Subject accrual and retention
During the 3-mo recruitment period, 81 persons responded to the study advertisements. Of these, 40 healthy persons met the entry requirements and were enrolled. Thirty-nine of the participants completed the study, with 1 dropout in the policosanol group. The participant voluntarily withdrew from the study shortly after enrollment. This resulted in at 97.4% completion rate for the study. The flow of study participants is illustrated in Figure 1.
FIGURE 1.. Flow diagram of subjects included in the study.
Baseline characteristics
Participant demographic and baseline clinical characteristics are shown in Table 1. Most enrollees were women and white. No significant differences in baseline demographics, medical history, medications, lipid variables, CRP, dietary or exercise habits, blood pressure, or body mass index were observed between the treatment arms.
View this table:
TABLE 1. Baseline characteristics of the study participants1
LDL cholesterol, lipoprotein concentrations, and particle size
The 8-wk changes in LDL cholesterol and lipoprotein concentrations and particle size are shown in Table 2. The median LDL-cholesterol percentage change was not significantly different between the policosanol and placebo groups.
View this table:
TABLE 2. Change in lipoprotein variables between baseline and week 81
No statistically significant differences in change in HDL cholesterol or triacylglycerols were observed between the randomization groups. This lack of effect was also identified in the median change in LDL particle number. Additionally, no significant differences in the changes in small LDL particle number, LDL particle size, or any other lipoprotein variable measured by NMR imaging were observed.
Nonlipid variables
Policosanol had no significant effect on CRP measurements before and after 8 wk of treatment (Table 2). In addition, there were no significant differences in weight, body mass index, and systolic and diastolic blood pressures between the study arms. Also, no significant changes in potential confounding variables, such as alcohol consumption, exercise frequency, or dietary habits as measured by the Diet Habit Survey, were noted (Table 3).
View this table:
TABLE 3. Changes in selected dietary and exercise variable by group between baseline and week 81
Safety and tolerability
Our study found that policosanol was well tolerated with a similar side-effect profile to placebo. No adverse events occurred throughout the study in either group. No significant change in liver function was noted by monitoring ALT before and after treatment. Overall, 90% of the participants in the treatment group and 84% of the participants in the control group recorded minor symptoms in their medication side-effect diary throughout the trial, and the side-effect profiles were similar between the 2 groups.
Adherence and masking
Adherence to the study therapies, as measured by pills per day multiplied by days on study, was 95% in both study arms. The participants were asked at the last visit to identify what they believed to be in the supplement caplets that they had been taking. In the placebo group, 45% believed they were receiving placebo compared with 32% in the policosanol group. No significant difference (P = 0.6, chi-square analysis) in the participants’ ability to identify the supplement they were receiving was observed; therefore the masking was adequate for the present study.
DISCUSSION
We found that supplementation with 20 mg/d of sugar cane–derived policosanol did not alter lipid concentrations when compared with placebo in participants with mild hypercholesterolemia. The lack of effect was seen in the primary endpoint (change in LDL cholesterol) as well as all secondary endpoints (change in total cholesterol, HDL cholesterol, triacylglycerols, and CRP and LDL particle analysis).
Our results contrast most existing published articles on policosanol. Indeed, a recent MEDLINE search showed >15 studies conducted before 2004 that uniformly showed the effectiveness of policosanol in lowering LDL cholesterol (4-6, 8-15, 17, 18, 31-37). These results have led to a number of review articles and a meta-analysis concluding that policosanol is a promising alternative to traditional lipid-lowering agents, especially given that many persons who need a reduction in LDL cholesterol do not tolerate statin therapy (1, 19, 22, 38). The positive studies would also appear to serve as the scientific support for the dramatic rise in policosanol use in this country as well as the use of this supplement in multivitamins, as demonstrated by the recent release of a Bayer pharmaceutical company product containing 20 mg policosanol per dose (2, 3). Despite the abundance of articles reporting significant cholesterol-lowering effects, it is extremely important to note that most of the positive studies have been published by one research group in Havana, Cuba, where the original product was developed by Dalmer Labs Inc (7). In addition, there have been no studies published that have examined clinical outcomes related to use of policosanol.
In contrast to the Cuban data, several recent animal and human studies from outside of Latin America have shown a lack of efficacy for policosanol to lower plasma LDL cholesterol (24-27, 39). The animal studies from Canada and Australia tested sugar cane–, sunflower seed-, and rice-derived policosanol and found no effect in lipid concentrations (26, 27). The first negative clinical trial in humans was published by a research group from the Netherlands; they showed no effect of wheat germ–derived policosanol on plasma cholesterol variables (25). This study tested a 20-mg standard dose of the wheat germ policosanol, although earlier findings indicated an increasing dose response effect from 2–40 mg (22). The second negative human trial was a study of rice-derived policosanol conducted in Croatia, which showed no significant changes in LDL cholesterol after 8 wk of therapy with a 10-mg dose (40). In 2006, another negative trial conducted in Germany was published. In this trail, Cuban supplied sugar cane policosanol was tested in doses of 10 to 80 mg; however, no significant change in cholesterol was noted (24). Finally, a study conducted in South Africa also showed the inability of a policosanol supplement, Octa-60, to affect cholesterol concentrations in humans (23). To counterbalance the large number of positive trials and confirm recent negative findings, additional clinical trials that used policosanol needed to be completed. Our finding that a 20-mg dose of sugar cane–derived policosanol lacks the efficacy to lower cholesterol augments the preexisting data from South Africa, Germany, and the Netherlands and conclusively shows that this supplement is not effective in lowering cholesterol in patients with mild hypercholesterolemia who are on a typical American diet.
The cause for the differences in outcomes between our trial and the prior publications from Cuba is uncertain. Differences between the typical North American diet and the dietary habits of the Latin American participants seem unlikely to be the cause of such a large variation in the cholesterol-lowering effects of this supplement. Ethnic differences could potentially underlie the different outcomes; however, this has not been shown by other lipid interventions such as statins (41). In addition, our study incorporated a diverse population, including African Americans and Hispanics.
Although there was no difference in any of the endpoint variables between the placebo and policosanol arms in our study, both groups had an 7% reduction in LDL cholesterol. This finding may merely illustrate the classic Hawthorne effect and serves to reiterate how essential it is for any valid scientific study to have a placebo control arm with evidence for adequacy of the double-blind. To assess the effectiveness of blinding, we asked the participants to guess the arm to which they were assigned at the completion of the study. Blinded participants should not guess correctly any more often than expected by chance, or 50% of the time, as occurred in the present study. This test of blinding adequacy is of particular importance in studies of herbal supplements where all too often it is not included.
Although the present study was adequately powered, one potential concern deserving discussion is the small sample size used. We anticipated an LDL-cholesterol reduction of 20% in participants treated with 20 mg policosanol on the basis of studies published before 2004. Originally designed to be a pilot study, our sample size calculation indicated that inclusion of 20 participants in each arm would provide 90% power to detect such a difference. The LDL-cholesterol change was almost identical in the 2 arms of the trial. The lack of even a small trend toward a difference, combined with similar negative results in most studies (human and animal) conducted outside Cuba, make it highly unlikely that the shown lack of efficacy was due to a type II error.
In conclusion, our study showed that a 20-mg dose of sugar cane–derived policosanol, with a chemical composition similar to that of the originally developed product from Cuba, did not lower LDL cholesterol. This finding corroborates the data from several other independently performed clinical trials. We would like to highlight the inconsistency between the negative results now published from 5 independent laboratories and the data published from Cuba, Argentina, and Russia before 2004. We hope to bring this discrepancy to the attention of healthcare providers and consumers of health supplements as a caution to use healthy skepticism when choosing supplemental products for lipid-lowering purposes. There is certainly no dearth of readily available supplements purporting such efficacy, but until the efficacy is verified through valid clinical trials performed by independent research groups, actively recommending the use of these products to persons at risk of coronary artery disease is negligent.
ACKNOWLEDGMENTS
Eric Schneider was the safety monitor for the study, and Laura Watson assisted with data entry. We thank the Charlotte-Mecklenburg Health Services Foundation for funding the study, and Garuda International for providing study medications free of charge. We thank the study subjects for their willingness to participate in the study.
MFD was the primary investigator and had full access to all data in the study and takes responsibility for the integrity of the data and the accuracy of the data analyses. MFD and TAB were responsible for study concept, design, and obtaining funding. Data collection was performed by MFD, TAB, and LFH. HCS performed the statistical analysis and interpretation of data. MFD, TAB, LFH, and HCS were responsible for initial writing and critical revision of the manuscript. None of the authors had any financial or personal conflict of interest in the companies or organizations sponsoring this research. The sponsors had no role in the collection, analysis, or interpretation of the data. The authors maintain full access to the study data and take responsibility for data integrity and the accuracy of the data analysis.
REFERENCES