点击显示 收起
1 From the Office of Dietary Supplements, National Institutes of Health, Bethesda, MD
2 Presented at the conference "Multivitamin/Mineral Supplements and Chronic Disease Prevention," held at the National Institutes of Health, Bethesda, MD, May 15–17, 2006. 3 Address reprint requests to EA Yetley, Office of Dietary Supplements, National Institutes of Health, 6100 Executive Boulevard, Room 3B01, MSC 7517, Bethesda, MD 20892-7517. E-mail: yetleye{at}od.nih.gov.
ABSTRACT
Although multivitamins, multiminerals, and similar terms (eg, multis or multiples) are commonly used, they have no standard scientific, regulatory, or marketplace definitions. Thus, multivitamins-multiminerals refers to products with widely varied compositions and characteristics. Multivitamin-multimineral composition databases use label values as surrogates for analyzed values. However, actual vitamin and mineral amounts often deviate from label values. Vitamin and mineral bioavailability for dietary supplements also lacks a standard scientific and regulatory definition and validated in vitro and animal models that accurately reflect human bioavailabilities. Systematic information on the bioavailability and bioequivalence of vitamins and minerals in marketed products and on potential drug interactions is scarce. Because of limited information on product characteristics, our ability to directly compare results across studies, estimate changes in usage patterns or intakes over time, and generalize from published results to marketed products is problematic.
Key Words: Vitamin and mineral supplements • bioavailability • bioequivalence • drug interactions • supplement composition • supplement definitions
INTRODUCTION
Product definitions and other product characteristics (eg, composition, bioavailability, and bioequivalence and potential for drug interactions) are important for assessing the scientific quality of published research, making comparisons among study results, evaluating time trends in consumer use patterns and intakes, and generalizing results from studied to marketed products. However, these important product-related methodologic details are often inadequately considered or described or inconsistently dealt with in research and population monitoring settings and publications. This manuscript focuses on these product-related issues.
MULTIVITAMIN-MULTIMINERAL DEFINITIONS
Although multivitamin-multimineral (MVM) and similar terms (eg, multis or multiples) are commonly used, they have no standard or defined meaning and can refer to products with widely varied compositions and characteristics.
Monitoring and research definitions
Since the 1980s, several nationally representative surveys have collected dietary supplement information (Table 1). Questionnaire differences within a survey series and across surveys, however—such as the types of supplement products for which information was sought; whether vitamin- and mineral-containing drugs, nonnutrient supplements, or supplements in forms other than pills and tablets were included; the duration of supplement usage; and the availability of product composition information—make it difficult to track changes in supplement use and intakes over time or across surveys (1-12). Even when a survey series asks the same question over time (eg, National Health Interview Survey, 1987–2000), the results are confounded because of the constantly evolving composition and intended use of marketed products.
View this table:
TABLE 1. Dietary supplement information in national surveys1
Product categorization differences also occurred among surveys and associated publications. The heterogeneity in the MVM category definitions that have been developed by users of national surveys is described in Table 2 (1-12). When respondents rather than data users are asked to identify their use of MVMs, errors can occur. For example, some respondents answered questions about MVM use by describing non-MVMs (eg, botanicals, sports drinks, and amino acids), which suggests that respondents' perception of MVMs may not match that of scientists (5, 13, 14). In some cases, nonnutritionist interviewers incorrectly recorded MVM products (13, 15). In one survey, the interview questionnaire asked respondents whether they ever used vitamins for health and treatment, and then respondents were shown a card that listed not only MVMs but non-vitamin-mineral supplements (eg, DHEA, coenzyme Q10) as possible responses, making interpretation of results difficult (8).
View this table:
TABLE 2. Examples of multivitamin-multimineral categorization in national surveys1
Reports from published studies also exhibited either a lack of standardization of MVM supplement definitions and categorizations or a lack of sufficient information to ascertain how the study defined or categorized MVMs (16). Categorizations varied from systems that focused on the product's intended use (eg, body building, weight loss), ingredients (eg, vitamin and mineral content), sources of ingredients (eg, antacids as a source of calcium) (16), disease perspectives (eg, antioxidants and cancer) (4), or consumer behaviors (eg, herbal versus MVM users) (17).
Regulatory definitions
No regulatory definitions exist for MVMs. In the United States, dietary supplements may contain multiple ingredients, including vitamins, minerals, herbs or other botanicals, and amino acids; dietary substances for use by humans to supplement the diet by increasing the total dietary intake; concentrates, metabolites, constituents, and extracts; or combinations of one or more of these ingredients (18). The European Commission proposed to define food supplements as concentrated sources of nutrients (primarily vitamin and mineral salts) marketed in dose form (eg, capsules, tablets, sachets, etc) to supplement the nutrient intake in a normal diet (19). The United Nation's Codex Alimentarius completed similar international standards for food supplements (20).
Marketplace definitions
A search of MVM products marketed on the Internet reveals a variety of products described as multivitamin, multimineral, multis, multiple nutrients, and multivitamin/mineral. These include "one-a-days" and specialized products (eg, MVMs for men, senior women, menopause, persons with diabetes, daytime or nighttime, performance, energy, menopause, hair). Many products do not include MVM terms in their names but contain similar types and amounts of vitamins and minerals as MVM-labeled products.
MVMs vary in the types, numbers, and amounts of vitamins and minerals they contain and whether they contain other nonvitamin and nonmineral ingredients (eg, dietary fiber, botanicals, glucosamine, lycopene). The same brand name with modifiers (eg, Brand X calcium plus) is used for products that differ in composition. Some products bearing the same brand name come in different forms (eg, pill or liquid) but differ in composition.
CHARACTERIZATION
Manufacturers determine both the types and levels of vitamins and minerals in MVMs (18). Information is scarce on the actual amounts of vitamins and minerals in these products; therefore, label declarations of vitamin-mineral content are often used as surrogates for actual levels. US regulations require that the amount of a vitamin or mineral is always equal to or greater than the label declaration after batch-to-batch variations and expected shelf life losses are taken into account.
Relatively few data are available that compare analyzed with declared label values in marketed products. Trade associations from the United States and the United Kingdom identified overages consistent with good manufacturing practices of 30–100% of declared value for vitamin A; 50% for vitamin B-12; 30–50% for ergocalciferol; 30% for cholecalciferol, folic acid, thiamine, biotin, β-carotene, vitamin K, riboflavin, niacin, vitamin B-6, and vitamin C; and 5% for vitamin E (21). However, underages have also been reported in US and Canadian products (22, 23). Comparisons of analyzed with label values showed selenium deviations from –19% to 23% (23) and up to 2.5 times label value (24) for products marketed in Canada and the United States, respectively. For Canadian products, vitamin E analyzed values deviated from label declarations by –41% to 57%, and vitamin D was within 15% of the stated dose (23).
Compositional databases are often developed by recording label values from respondent supplement containers because analytic data are not available or by using defaults based on simplified questionnaires or composition of commonly used products because of interview time or budget constraints. Comparisons of intake estimates derived from a simplified inventory (25) or default product categories (26) showed that the accuracy of results based on defaults or simplified questionnaires varied by nutrient.
BIOAVAILABILITY
Definitions: bioavailability and bioequivalence
The concept of vitamin and mineral bioavailability for dietary supplements lacks standard scientific and regulatory definitions. Commonly used definitions include concepts of absorption (27) and some also include utilization (eg, availability for use or storage) (28-30). For some nutrients, beneficial functions of unabsorbed nutrients (eg, binding of bile salts by calcium in the gut) would be missed by absorption-based definitions (27). Similar to the situation for drugs, definitions of nutrient bioavailability may benefit from a provision for functionality at sites of action not dependent on systemic blood circulation for delivery of the active moiety (31).
Bioequivalence is closely related to the concept of bioavailability (31, 32). For some nutrients, equal absorption does not mean equal biological effects because the nutrient sources are chemically different, resulting in differences in nutrient activity (27). Bioavailability and bioequivalence factors are the basis for adjustments for some nutrient reference values (Table 3) (33-35). For example, differences in the activity of different chemical sources of vitamin A and folate are converted to vitamin equivalents when setting Recommended Dietary Allowances (RDAs). The effect on intake estimates for folate equivalents versus micrograms is illustrated in Figure 1 (36). For other nutrients (eg, iron, niacin, vitamin B-6, and vitamin B-12), RDA or Adequate Intake (AI) values are adjusted on the basis of different bioavailabilities and bioequivalencies from mixed diets, making their use for dietary supplements problematic.
View this table:
TABLE 3. Examples of considerations of bioavailability and bioequivalency of the Dietary Reference Intakes (DRIs)
FIGURE 1.. Percentiles of total folate intakes (foods + supplements) of women aged 31–50 y from the National Health and Nutrition Examination Survey 2001–2002 expressed as amount (µg/d) or as dietary folate equivalents (DFEs)/d. Intake distributions are adjusted for intraindividual variations in intakes to represent usual intakes (36).
Factors affecting bioavailability
Vitamin and mineral bioavailability in MVMs is affected not only by product but also by host factors. Host factors include homeostatic mechanisms that regulate absorption or excretion depending on the nutrient status of the host (eg, the iron status of the host affects iron absorption) (27, 29). These factors vary by age, sex, and physiologic state (eg, pregnancy) (29, 30, 37). Homeostatic mechanisms may regulate circulating concentrations of nutrients within a tight range and are therefore insensitive to changes in ingested amounts or to utilization at the site of action (27). The size of the ingested load may affect bioavailability [eg, the absorption load of calcium varies inversely as the logarithm of the load size (27), and single high doses of folic acid exceed the metabolic capacity for reduction and methylation (38)]. These complex host factors are the basis for questions as to the validity of including utilization in definitions of nutrient bioavailability (27). They also give rise to concerns that higher bioavailability is not necessarily better (27, 29).
Hoag and Hussain (32) identified 2 categories of product formulation factors that affect bioavailability and bioequivalence: factors that affect product dissolution or release from the dosage form and factors related to excipients or inactive ingredients that may affect drug stability, drug absorption, and metabolic processes. Product dissolution or release is the basis of pharmacopoeial dissolution test methods (28). Different chemical forms of nutrients and nutrient-nutrient interactions may affect bioavailability, eg, different forms of inorganic iron or zinc vary in bioavailability (39, 40), vitamin C interacting with inorganic iron may enhance bioavailability of the iron (39), and decreasing the levels of magnesium and calcium increases the bioavailability of iron (41). The formulations, fillers, coatings, excipients, and surfactants of products affect the completeness or rate of release of calcium (27, 42), vitamin E (43, 44), pyridoxal phosphate (45), iron (46), folic acid (47), and vitamin B-12 (48).
Testing of bioavailability
Currently, in vitro and animal models do not accurately reflect human bioavailabilities and, therefore, human testing is the gold standard (27). Similar to their quality assurance standards for drugs, the United States Pharmacopeia (USP) has published disintegration and dissolution standards for evaluating MVMs (28). These standards assume that in vitro acid solubility is a surrogate for in vivo absorption. However, the usefulness of these standards for nutrients not dependent on acid solubility for absorption (eg, calcium) has been questioned (27). The failure of US marketed products to meet USP dissolution standards was reported for a prenatal MVM (22), folic acid in prenatal MVMs (49, 50), and renal MVMs (51). Similar failures of dissolution and disintegration were reported for MVMs marketed in Canada as compared with USP standards (52), for folic acid–containing products marketed in the United Kingdom as compared with British Pharamcopoieal standards (53), and for iron-containing products in Sweden (54). In the United States, the use of USP standards by manufacturers is voluntary and not required by regulation.
DRUG INTERACTIONS
Drug-nutrient interactions are the result of both host and nutrient-drug factors. These interactions can make a drug less effective, increase the action of a drug, or cause unexpected side effects. There are also reports of certain drugs decreasing the effectiveness of vitamins and minerals. Little research-based information is available on potential or actual drug-vitamin or drug-mineral interactions. Reported interactions include vitamin E and aspirin with the potential for an additive antithrombotic effect and between vitamin E and warfarin related to an increased risk of bleeding (55). An antioxidant supplement containing vitamins C and E, β-carotene, and selenium used in the treatment arm of a randomized controlled clinical trial blocked the beneficial response of HDL to simvastatin-niacin therapy in patients with coronary artery disease and low HDL (56). The vitamin-drug and mineral-drug interactions that have been reported in the Thomson Micromedex Healthcare Series are listed in Table 4 (57).
View this table:
TABLE 4. Reported interactions of vitamins and minerals with drugs1
Dietary supplement manufacturers have not been required to evaluate the potential for drug interactions, but legislation to mandate the reporting of serious adverse effects is currently under consideration. Reports of suspected or documented adverse events may be voluntarily submitted to the Food and Drug Administration's MedWatch program (58) or other organizations, such as poison control centers. If drug-MVM supplement interactions are documented, information identifying this would more likely be required for drug labeling than for supplement labeling. However, manufacturers and retailers may voluntarily place warning statements on MVMs. These statements do not require review or approval by the Food and Drug Administration.
CONCLUSION
The definitions and categorization of MVM products are not standardized. These differences can affect estimates of both prevalence and intakes and make it difficult to compare trends over time, make direct comparisons from one study or survey to another, or generalize from published results to marketed products. The variable composition of marketed MVMs and the inaccuracy of label declarations underscore the need for more systematic analysis of vitamin and mineral content of MVMs. Deviations of even 25–50% from label values, given the relatively high quantities of vitamins and minerals in marketed MVMs, can result in significant errors in estimates of dietary intakes from MVMs. It is unclear whether a single definition of bioavailability can apply to all vitamins and minerals because factors affecting nutrient absorption and the relation of circulating nutrient concentrations to functional effects at sites of action vary among nutrients. Although drug definitions of bioavailability are often applied to nutrients, inherent differences between drugs and nutrients require that drug definitions be modified if they are to be usefully applied to vitamins and minerals. Drug interactions with MVM ingredients are rarely systematically studied but may interfere with or augment the effects of some drugs in unexpected ways.
ACKNOWLEDGMENTS
I thank Susan Pilch, Biomedical Librarian/Informationist, National Institutes of Health Library, for her expertise and patience in conducting extensive bibliographic searches to identify hard-to-find references on the topic of this manuscript. I also thank Connie Hardy, Consumer Safety Office, Center for Food Safety and Applied Nutrition, Food and Drug Administration, for her skilled and timely assistance in identifying marketed MVM products.
REFERENCES