Prevalence of Dietary Supplement Use in Patients with Proven or Suspected Cardiovascular Disease

Systematic search of bibliographic databases was conducted to describe the prevalence of dietary supplement use in cardiac patients. Included for review were studies that investigated supplement use in people with cardiovascular risk factors or proven cardiovascular disease. Databases searched were Medline, EMBASE, CINAHL, AMED, Meditext, H&S and IPA. Over five hundred articles were retrieved and twenty studies met the criteria for this review. Dietary supplements were taken by a median 36% (interquartile range: 26–42%) of cardiac patients; 36% (IQR 18–43%) reported taking a vitamin/mineral supplement and 12% (IQR 7–21%) used herbal supplements. Many users indicated that supplements were taken specifically for heart health and 16–64% of users reported using supplements alongside prescription medications. However 39–95% of treating physicians were unaware of patients' supplement use. Dietary supplement use in patients with cardiovascular disease appears common, as does the concurrent use of supplements with prescription medicines. This information is often not communicated to doctors and treating physicians may need to be more proactive in asking about supplement use.


Introduction
A dietary supplement is defined as "any product intended for ingestion as a supplement to the diet" [1]. These substances include amino acids, charcoal, choline salts, essential oils, plant or herbal material, homeopathic preparations, nonvaccine microorganisms, minerals, nonhuman animal material, lipids, substances from bees, and vitamins or provitamins. In this paper, we use the term "dietary supplement" to indicate any of these substances used as complementary and alternative medicine (CAM) and taken orally for the improvement of health or the prevention of illness.
The use of dietary supplements is common; in the general population, dietary supplements are the second most commonly used type of CAM after prayer [2]. Dietary supplements can interfere with the biotherapeutic action of prescription medications, and this is of particular concern in cardiac patients, many of whom are on long-term medications and are at increased risk of acute life-threatening events. Chronic conditions such as arthritis, cancer, depression, and anxiety have been consistently associated with CAM use [3], so it is probable that patients with chronic cardiovascular disease are also likely to use CAM. We conducted a systematic review of the literature to find the prevalence of dietary supplement use by cardiac patients and to identify commonly used supplements. We used the MeSH terms "cardiovascular diseases" with the qualifier "/therapy" to find articles relevant to treatment of cardiovascular disease. To find articles on dietary supplement use, we used the MeSH terms "medicine, traditional", "dietary supplements", "vitamins", "minerals", each qualified with "/utilization". All MeSH terms were exploded to include subheadings. The "cardiovascular disease" search results were then intersected with each of the dietary supplement searches. For databases not indexed using MeSH, searches were performed using the same terms without the qualifiers. In these databases, additional searches were conducted using the terms "cardiovascular diseases", "cardiology", and "cardiac" intersected with each of "complementary and alternative medicine", "dietary supplement", and "herbal medicine". Search results were limited to articles in English and those that dealt with humans where these options were available.
The titles and abstracts of articles were read and refined to include only those studies that (i) reported the prevalence of dietary supplement use, (ii) involved a sample of patients either attending for cardiac care or those reporting a cardiovascular condition, and (iii) indicated the types of supplements used. If abstracts were not available or were ambiguous with regard to these criteria, the full-text article was obtained for inspection. Articles were excluded if (i) study methods were not described, (ii) prevalence data on dietary supplement use were not reported (or these could not be calculated from reported data), and if (iii) they were not in English. Citations in the relevant papers were also used to locate articles not found by the above methods. Figure 1 shows the selection of studies from the search results.

Data Extraction.
Two data tables were designed for systematic extraction of study information. Methodological details were extracted and compared in Table 1, while study content and results were extracted and summarised in Table 2. The most common supplements are listed in Table 3, and for the sake of brevity, this only includes supplements used by ≥1% of patients and which were reported in 2 or more of the included studies.

Statistical Analyses.
Prevalence of overall supplement use was obtained from each study. For studies that did not report the overall prevalence, the figure corresponding to the most common supplement or most common supplement category was used. This method is expected to underestimate rather than overestimate the total number of supplement users. The median prevalence across studies and the corresponding interquartile ranges (IQR) were calculated for (i) any supplement use, (ii) the use of vitamin/mineral supplements, and (iii) the use of herbal supplements. Where studies reported use of individual supplements, data was extracted and pooled figures are presented in Table 3.
Evidence-Based Complementary and Alternative Medicine 3
Thoroughness of methodological reporting varied between studies (Table 1). Only three of the 16 clinical studies commented on the representativeness of their sample compared to the greater clinical population [7,19,23]. Six did not report the survey response rate [9,10,14,19,21,24]. For those that did, responding ranged from 26%-100% (mean 74%). Only four of the clinical studies piloted their survey instrument [7,19,24] and two trained multiple interviewers to an equivalent standard [12,23]. Four performed a sample-size calculation to determine whether there was sufficient statistical power to detect differences between supplement users and nonusers [8,14,23,25]. Many studies did not explicitly state what was meant by "dietary supplement" or CAM and three studies did not define the time period over which supplement use was surveyed [9,10,23].
Six out of 20 studies surveyed the use of CAM as well as dietary supplements [6,8,10,12,20,23], which included physical therapies such as chiropractic and yoga, and mindbody therapies such as relaxation and hypnosis. Another four included the use of nonprescription medications as well as dietary supplements [7,9,11,19]. Only results specific to dietary supplements are reported in this paper.
We also analysed results with respect to time periods surveyed. Seven studies looked at ever or yearly use, finding a median 25% of respondents used supplements (IQR 17%-45%). Three or 6-month use was surveyed by four studies with median 49% users (IQR 37%-61%). Median use within the last month in six studies was found to be 39% (IQR 37%-40%). Only two studies investigated the duration of supplement use, which averaged 39 months [24], with 88% of herbal users taking the supplement for a month or more [13].
Supplement users tended to use more than one supplement at a time: two studies reported an average of 2 supplements per user [14,24], while another recorded an average of 3 supplements per user [21]. Common supplements used by cardiac patients are summarized in Table 3 with the most common by far being multivitamins and minerals.

Concomitant Use of Supplements and Prescription Medications.
Five studies reported data on the concomitant use of prescription medication and dietary supplements [11,13,17,20,22,23]. The study of Indian patients indicated 64% were using CAM in conjunction with prescription antihypertensives, although 86% had started conventional treatment first, while 9% had began CAM first and 5% had started both alternative and prescription treatments simultaneously [20]. One population survey in the US indicated that 36% of vitamin/mineral users and 16% of other supplement users used prescription medication for cardiovascular health [11], while another found that 63% of those with hypertension and 64% of those with high cholesterol were found to also take dietary supplements with their medications for cardiovascular disease [13]. Thirteen percent of Canadian patients with diagnosed coronary artery disease (CAD) had used a dietary supplement in conjunction with warfarin, amiodarone, sotalol, or digoxin [23], while 26% of Chinese patients on warfarin had taken a herbal medicine in the previous week [22]. Supplement users were as likely as nonusers to be taking aspirin, betablockers, ACE inhibitors, statins, and warfarin [17]. There was no difference between herb users and nonusers in terms of indication for warfarin, and the duration and dosage of warfarin therapy [22].
Three studies looked at the impact of supplement use on compliance [10,15,16]. Of the two which surveyed hypertensive patients [10,16], one found no difference between supplement users and nonusers on adherence to antihypertensive medication [10], while the other found that female supplement users had lower adherence to prescription medications [16]. The third study that investigated outpatients with heart failure reported that 1 of 22 herbal supplement users reduced or interrupted heart failure medications while taking herbal medicine [15].

Cardiovascular Health and Supplement Use.
In patients with heart failure, up to 82% of supplement users were taking a supplement specifically for cardiovascular health [25]. In patients with diagnosed cardiovascular disease, 18 out of 42 supplements (43%) were taken for a cardiac reason [14]. Supplement users were more likely than nonusers to be under specialist cardiology care [9], although the use of herbal supplements was not associated with previous myocardial infarction or previous percutaneous coronary intervention in a Turkish sample [24]. A US study [21]       Note: Only supplements with reported ≥1% use in 2 or more studies were included. Some figures were calculated from reported percentages. did find higher supplement use in those with significant CAD (defined as prior myocardial infarction or previous percutaneous coronary intervention), and the studies of Ai and Bolling [8,26] showed that use of herbs and folk remedies were significantly associated with arrhythmia and left main disease. Six studies specifically investigated patients with congestive heart failure (CHF) [7,14,15,19,23,25]. In this population, supplement use was reported to be 40% [23] and 82% [25] although use appeared unrelated to the severity of CHF [7] or left ventricular ejection fraction [9,21].
Four studies reported data on measurable outcomes. In Chinese patients taking warfarin, INR was significantly lower in herbal users than nonusers [22]. In two separate studies [10,21] of clinic outpatients, blood pressure control did not differ between supplement users and nonusers although herbal use was associated with lower blood pressure in elderly patients with heart failure [9]. In one clinic population, lowdensity lipoprotein (LDL) levels also did not differ between supplement users and nonusers but high-density lipoprotein (HDL) was higher among the supplement users [21]. [6,9,14,17,20,23,24]. The most common reasons for supplement use were: to promote or maintain health (up to 47%), for overall wellbeing (up to 42%), and for energy (up to 25%) [9]. Dissatisfaction with conventional treatment was cited by 17% of patients in one study as a reason for supplement use [23] while in another, 59% cited fear of adverse drug reactions as the reason for using CAM [20]. Three times more supplement users than nonusers thought supplements were safe (75% versus 26%) [14], with 45% of users believing supplements were safer than prescription medications [15], 30% indicating that supplements were less likely to interact with other medicines [17], and 45%-47% thinking that supplements would result in fewer side effects compared to prescription medications [14,17]. Two studies found that 47% [17] and 81% [24] of users, respectively, were unaware herbal medicines could impact negatively on prescription medications. Dietary supplements were viewed as effective by 70%-80% of supplement users [6,14].

Patient Perceptions of Dietary Supplements. Seven studies investigated patient perceptions of dietary supplements
Only one study uncovered negative views of CAM [20], finding that 56% of patients were dissatisfied with CAM. Two thirds of these patients thought CAM was too expensive and the rest believed that CAM caused adverse effects. The same study reported exacerbated hypertension made up 74% of the adverse effects attributed by patients to CAM use [20] and all patients who reported adverse effects subsequently discontinued CAM use.

Physician Awareness of Patients' Supplement Use.
Six studies asked whether patients informed doctors about supplement use and found physicians were not notified 44% (IQR 40%-53%) of the time [6,14,15,17,20,24]. One study reported that 67% of supplement users did not disclose supplement use because their doctors did not ask [14], but other studies did not investigate the reasons behind nondisclosure.

Discussion
This is the first systematic review of dietary supplement use in cardiac patients and the findings have strong implications for clinical practice. The results indicate that supplement use is common in cardiac patients (26%-42%) and that the concomitant use of dietary supplements and prescription medicine also appears to be frequent (16%-64%). Sicker patients may be more likely to seek out alternative treatments [8,13,21], and drug interaction could play a causal role in the lower INR observed among supplement users on warfarin [22] or in the lower blood pressure among cardiac failure patients [9].
The potential for negative interaction in supplement users is high. Substances such as fish oil, hawthorn, garlic, ginseng, ginkgo, glucosamine, and parsley have antiplatelet properties [27,28] and may interact with prescription antiplatelets or anticoagulants. Supplements such as capsicum and ginseng have been shown to affect blood pressure [29]. Supplemental potassium was taken by 1 in 5 respondents in one study [7]; this may result in adverse outcomes when used in combination with commonly prescribed cardiovascular medications such as angiotensin converting enzyme inhibitors, aldosterone receptor antagonists, or angiotensin receptor blockers. Other researchers have thoroughly reviewed the effects of herbal medicines and dietary supplements for cardiovascular health [30] as well as their potential for drug interaction [31] and we refer clinicians to these resources for more details.
It is worth noting that patients are likely to be using more than one supplement at a time [21], for periods of months to years [13,24]. This raises concerns of possible ongoing adverse interaction between supplements and prescription medications, as well as the potential for negative effects on compliance with conventional therapy. Many patients believe supplements are safe, effective, and produce few side effects. Many are also unaware that supplements can negatively impact prescription medications [14,15,24]. Moreover, a large proportion of physicians (39%-95%) are unaware of supplement use by their patients. Although patients may be reluctant to disclose supplement use, the main reason for nondisclosure appears to be physicians not asking about supplement use [14]. This suggests physicians can produce significant change in this aspect by being more proactive and asking about supplement use.
There are inconsistent results with regard to the association between supplement use and compliance. There is also limited evidence on whether supplement use affects clinical outcomes such as blood pressure and cholesterol control. The available evidence suggests that supplement use is associated with differences in health status of cardiac patients, although it is unknown whether supplement use causes these differences or whether differences in health cause patients to self-manage through supplementation.
The aggregated results presented here should be treated with caution. Our results show wide variability in the prevalence of dietary supplement use, which is consistent with other reviews of the CAM literature [32,33]. Some of these variations may be due to the country and year in which the study was conducted. For the North American studies, the range of figures is smaller than for all studies but still large. For vitamin/mineral use it is 38% (IQR 26-49), for herbals 10% (IQR 7-18), and for other supplements 7% (IQR 5%-13%). In terms of variations over time, rates of supplement use seem to be decreasing over time, the overall rates being negatively correlated to the publication year in the North American studies (r = −0.30). However, there are also severe methodological limitations in the quality and comparability of the reviewed studies. For instance, even though many of the included studies involved clinical samples, few provided objective data on the health status of the patients. Clinical diagnoses of cardiovascular disease were reported in only seven studies [7,9,10,14,19,23,25], and only six compared supplement users and nonusers on clinical outcomes [7,9,10,17,21,22]. Study comparison was also difficult due to heterogeneity in the definitions of dietary supplements and supplement use. For instance, some studies did not include common vitamins and supplements in their surveys, which would radically reduce the rates of supplement use. Some studies provided a checklist of supplements with no room for respondents to list other substances they may have used. Across studies, patients were asked whether they had used dietary supplements in the past year [6,8,12,14,16], in the past month [13], in the past 3 and 6 months [7,17,19,25], as well as whether ever used in lifetime [15,18]. Two studies defined regular use as "at least once weekly" [7,19] and only one asked for the duration of use [24]. An obvious result of using dissimilar definitions are the disparate prevalence rates even in the largest population studies, all conducted in North America [6,11,13], which found supplement use to be 36%, 61%, and 22%, respectively.
Future studies should focus on developing three main areas. The first is to standardise and make explicit the definitions of dietary supplements and supplement use, which will facilitate cross-study comparison [33]. The second is to examine the impact of supplement use on prescription medications. Our paper found inconsistent associations between supplement use and compliance with prescription medications. Given that long-term compliance is essential to cardiac care, investigation into whether supplement use disrupts compliance is warranted. This is particularly important in population groups susceptible to noncompliance. In addition, the reluctance of patients to disclose use of dietary supplements may require patient and physician education. Lastly, studies are needed to examine both health and financial costs of supplement use. Only one of the reviewed studies examined the relative cost of CAM, finding that patients of an Indian clinic were spending the same amount on CAM and antihypertensive treatment, although the authors note that the cost of mainstream care can be reimbursed in most cases whilst CAM treatment cannot [20]. Prospective studies can assess whether concomitant use of prescription medications and supplements in cardiac patients affects outcomes such as morbidity, mortality, and quality of life, while economic evaluation can determine whether supplement use is cost effective, especially given the wide availability, increasing accessibility, and now common use of dietary supplements.
In summary, the use of dietary supplements is common in patients with cardiovascular conditions. There are no conclusive findings with respect to the health factors associated with this use although many commonly used supplements have the potential to interfere with the intended action of prescription medications. A substantial number of treating physicians are unaware of patients' supplement use, and consequently the effects of supplementation on conventional treatment may be overlooked. Disclosure of supplement use can be facilitated if medical practitioners are more proactive in questioning patients.