|
|
||||||||
Instituto Nacional de Salud Pública, Centro de Investigación en Nutrición y Salud. Cuernavaca, Morelos, MEXICO
Address reprint requests to: Fabiola Mejia-Rodriguez, Centro de Investigación en Nutrición y Salud, Instituto Nacional de Salud Pública, Av. Universidad No. 655, Sta. Ma. Ahuacatitlán, Cuernavaca, Mor. CP 62508, MEXICO. E-mail: fmejia{at}insp.mx
| ABSTRACT |
|---|
|
|
|---|
Methods: Information was collected by questionnaire on the frequency and duration of supplement use in the previous 6 months (n = 17,794). Dietary intakes by 24-hour recall were determined in a representative sub-sample (n = 2,599). Frequency of use and available information on the nutrient content of supplements was used to estimate daily equivalent intakes.
Results: 17.6% of women reported to have used supplements. The majority of these took supplements once daily (71%) and for
2 months (75%). While nutrient intakes from diet alone did not differ between users and non-users, the proportion with intakes of Vitamins A, B6, B12, and C, folate, iron, and zinc < EAR were significantly greater among the supplement non-users when intakes from supplements were also considered. The proportion of women with intakes > UL was greater among supplement users than non-users for iron, folate and Vitamin B6.
Conclusion: Supplement use contributes to the adequacy of nutrient intakes but may also increase the possible risk of toxic intakes of some nutrients among Mexican women.
Key words: diet, Mexico, minerals, supplements, vitamins, women
| INTRODUCTION |
|---|
|
|
|---|
Nutritional supplement use is often limited to sectors of the population with higher levels of education and higher socioeconomic status [57]; these are also often the groups at lower risk of dietary inadequacy. Indeed, it has been observed that individuals who use supplements tend to already have greater intakes of micronutrients from the diet than non-users [68]. Therefore, the impact of voluntary supplement use on reducing the prevalence of inadequate dietary intakes may be limited if the majority of people who use them are those who least require them.
Concern has also been raised about the possible risk of toxic effects from chronic, excess nutrient intakes because of voluntary supplement use [9, 10]. It is therefore of interest to public health to evaluate and monitor to what extent this practice contributes to risk of nutrient toxicity in different populations.
In Mexico, data from the National Nutrition Survey of 1999 (NNS-99) indicate that 18% of all women use nutritional supplements [11]. However, further quantitative information on supplement use and their contribution to the adequacy of nutrient intakes and to possible toxic intakes has not yet been reported. The objectives of this study are to describe the practice of nutritional supplement use among Mexican women and to estimate the contribution of supplements to the adequacy of nutrient intakes and to toxic intakes of select nutrients.
| SUBJECTS AND METHODS |
|---|
|
|
|---|
Dietary intakes were determined for one female 1249 years of age in a subsample (20%) of 4200 selected households; complete dietary and demographic information were available for 2,599 women. Of the women with dietary intake information, complete data on supplement use were obtained from 2,257 women. Fig. 1 summarizes the sample used for the present analysis.
|
Confounding variables included age, body mass index, socioeconomic status level, hemoglobin concentration, level of education, access to public health care, and participation in food assistance programs. Body mass index was categorized as underweight, normal, overweight, and obese according to Must et al for women
18 y and to Rolland-Cachera et al for girls <18 y [13, 14]. Hemoglobin concentration was classified as anemia (<12.0 g/dL) or non-anemia (
12.0 g/dL) after adjusting for altitude [15]. Socio-economic index was calculated as a continuous variable according to Bronfman et al [16], using principal components factor analysis. Selected variables were flooring material, availability of potable water, and ownership of household electrical appliances (washing machine, refrigerator, television, radio, and stove). This index explained 51.6% of the generalized variance of the set of variables included. Subsequently, the index was used to form tercile groups to classify socioeconomic status into high, medium, and low strata.
Nutrient Intakes from Supplements
As part of the present analysis, a database was constructed of the nutrient content of supplements reported to be used during the survey. Nutrients considered were Vitamins A, B6, B12, and C, folic acid, iron, and zinc. The nutrients chosen for this analysis are restricted to micronutrients, selecting among those that are more commonly deficient in developing countries, those for which nutrient intake from the diet was calculated, and those for which an Estimated Average Requirement (EAR) has been assigned.
Information on the nutrient content was derived from pharmacies and other stores selling nutritional supplements, a pharmaceutical dictionary, or directly from the manufacturer. When the specific content of a supplement was not found, it was matched with a similar product based on the ingredient list, and the same nutrient content was assumed. All nutrient contents were converted to the same units for which dietary intakes are expressed. Beta-carotene and retinol were converted to Retinol Activity Equivalents (RAE) and folic acid was converted to Dietary Folate Equivalents (DFE), as recommended [17, 18].
As intakes from supplements were reported in terms of frequency of use per week, we estimated the average daily equivalent of the dose for each person. For example, if it was reported that a supplement was taken 7 days per week, then the daily equivalent intake was calculated as one dose per day. If the frequency was reported as 1 to 2 times per week, the daily equivalent was calculated as 1.5 (mean of 1 to 2) times per week divided by 7 days per week, or 0.21 doses per day [17, 18].
Dietary Intake Data
The NNS-99 obtained information on dietary intakes using a single 24-h recall applied to the randomly selected subsample of participants. The data collection procedures and methods to calculate nutrient intakes from food intakes were previously described [11, 12]. An estimate of the inadequate intakes was determined using the proportion of women with intakes below the EAR [17, 18]. Similarly, the prevalence of possibly toxic intakes was estimated from the proportion of women with intakes greater than the Upper Limit (UL) for each nutrient [1719]. These prevalences (< EAR and > UL) were compared between users and non-users of supplements both for diet alone and for diet plus the daily equivalent derived from supplements.
Statistical Analysis
All data were analyzed using the statistical software package Stata for Windows version 8.0 (Cop College Station, Texas, USA, Stata 8.2, Corp., 2003). Descriptive statistics were used to summarize information on the supplements reported to be used and their nutritional content. For all analyses of the nutrient intake data we used the SVY module of Stata to take into account the survey design and weights. Dietary inadequacy (<EAR) for each nutrient studied was compared between users and non-users using logistic regression analysis, controlling for confounding variables found to be statistically significant (p < 0.05) explanatory variables for use or non-use of supplements.
| RESULTS |
|---|
|
|
|---|
The distribution of supplements reported by category is shown in Fig. 2. The multi-vitamin/mineral supplements were most commonly reported (33%), followed by multiple vitamins (13%) and single vitamins (10%). The nutrient content per dose of each supplement (median, minimum, maximum) and the proportion of supplements whose content exceeds the Recommended Dietary Allowances (RDA) and the UL are shown in Table 1. The daily dose for several micronutrients exceeded the RDA in one-quarter to nearly one-half of the supplements. The nutrients that most frequently exceeded the UL per dose were iron, folate, and Vitamin A.
|
|
Contribution of Supplements to Nutrient Intakes
The median (25th, 75th percentiles) nutrient intakes from the diet alone were compared for supplement users and non-users (Table 2). The prevalences of women with intakes below the EAR are also shown. The prevalence of intakes below the EAR from diet alone did not differ significantly between the two groups, with the exception of vitamin A, where a greater proportion of supplement non-users had low intakes (p = 0.04). However, when the contribution of supplements to nutrient intakes is considered, the prevalence of low intakes was significantly greater among the supplement non-users than users for all nutrients (Table 2). When considering only supplement users, the estimated prevalence of intakes below the EAR was anywhere from 10 to 23 percentage points lower when the contribution from supplements was included. On a national level, not including the contribution of supplements to nutrient intakes resulted in an overestimate of about 2 to 4 percentage points of nutrient intake inadequacy (Table 3).
|
|
|
| DISCUSSION |
|---|
The majority of supplement users reported to take supplements on a daily basis, however most often these were taken for no more than two months, suggesting a transient practice. This, combined with the fact that only 18% of women used the supplements, suggests that the practice is not yet widespread enough to have a major impact on national level estimates of inadequate intakes. Nonetheless, it is still recommended to continue monitoring supplement use in national surveys; if trends for increased supplement use occur in Mexico, as they appear to be in the United States [1], the quantification of supplements as a source of nutrients may take on greater importance in future dietary surveys. This may be of particular interest in Mexico given that the prevalence of low nutrient intakes and nutrient status according to biochemical indicators is relatively high [2022].
On the other hand, the assessment of relative intakes of nutrients among those who do use supplements is greatly affected by the inclusion of intakes from supplements. The proportion of intakes below the EAR was reduced by at least 25%, as in the case of Vitamin B12, and up to 57%, as in the case of folate, when the estimated nutrient intake from supplements was added to dietary intakes. Therefore, for studies or surveys that focus on subpopulations of Mexican women more likely to use supplements, the estimates of dietary adequacy will be inaccurate, as suggested by our data and associations between dietary intake and biochemical or functional indicators of nutritional status may be attenuated.
Other studies have reported that women who use nutritional supplements have greater nutrient intakes from the diet than their counterparts who do not use supplements, even after controlling for social and demographic characteristics associated with dietary adequacy [8,2124]. This tendency may occur because those who use supplements are more health conscious and select more nutritious diets as well as chose to use nutritional supplements. However, this trend was not observed among supplement users and non-users in Mexico.
In this population of non-pregnant, non-lactating women in Mexico, women who were of higher socioeconomic status (p = 0.049) and higher education level (p = 0.004), of urban residence (p = 0.040), with social security and a health care plan (p = 0.010), or with low body mass index (p < 0.001) were more likely to use supplements (p values of the coefficients determined by multiple logistic regression; Mejía-Rodríguez F, unpublished data). Based on biochemical data from this same population, the likelihood of deficiencies of Vitamin C, iron, and zinc were higher among women with lower socioeconomic status [20, 21]. Therefore, although women of higher socioeconomic status are less likely to be deficient and are among those more likely to use supplements, the dietary intake data suggest that there is still a relatively high risk of inadequate micronutrient intakes in this group. This suggests that voluntary use of supplements may still be beneficial for the socio-demographic group that is currently using supplements in Mexico.
The NNS-99 was not designed to determine the reasons for nutritional supplement use in the population. Surveys in other countries have indicated that women use supplements when they perceive that their diets are poor, when under conditions of stress, and when they seek to correct or reduce their risk of health problems [25, 26]. A survey carried out in Mexico City indicated that 50% of the population was not aware of the health functions of many nutrients [27]. In the same survey, 31% of respondents indicated that the main source of nutrition information was the mass media. This suggests that there is a great need for public health workers to improve the nutrition knowledge of the public so that informed decisions about the use of supplements may be made. The importance of this is further suggested by the statement that appears on supplement labels in Mexico: the consumption of this product is the responsibility of the one who recommends it and the one who consumes it.
For some nutrients, the difference between the recommended daily allowance level and the Upper Limit is small. Therefore, while attempting to use nutritional supplements to meet nutrient requirements, the risk of consuming toxic levels of nutrients also increases. Indeed, our data suggest that when the estimated contribution of nutrients from supplements is taken into account, the proportion of women with intakes of iron, folate, and Vitamin B6 >UL increased significantly. The daily dose recommended by the supplement manufacturers in some cases exceeds the UL (Table 1). Many of these supplements as used by non-pregnant and non-lactating women included in this analysis were marketed to athletes or for women during pregnancy and they may not be appropriate for the general population for long-term, unsupervised use.
Evidence for the effects of chronically high intakes of some nutrients is still limited, and the scientific basis for setting ULs for some nutrients is controversial [28]. On the other hand, strong evidence exists for the deleterious effect of high-dose supplemental Vitamin E on the risk of mortality [29], and for the teratogenic effects of high doses of Vitamin A during pregnancy [30]. Recent evidence suggests an increased risk of mortality from cardiovascular disease among postmenopausal, diabetic women with high Vitamin C intakes from supplements [31]. Intakes of Vitamin E, Vitamin A during pregnancy, or Vitamin C in postmenopausal diabetic women were not evaluated in the present analysis. However, Vitamin E is a constituent of some supplements reportedly used and 14.5% of supplements contained doses of Vitamin A greater than the UL which, when combined with an adequate intake of Vitamin A from the diet, approach levels that may lead to teratogenicity (> 10,000 IU) [30]. Given that the majority of women in this analysis (75%) reported to have used supplements for 2 months or less, there may not be too much cause for concern at present regarding the manifestation of toxic effects, although this should continue to be monitored in this population. The risk of toxic intakes may also be minimized by ensuring that supplements with excessively high nutrient content are used only under medical supervision and do not reach the open market.
The interpretation of the results of this analysis is limited for a few reasons. We calculated the daily equivalent of nutrient intakes from supplements based on the frequency of use reported and the manufacturer's recommended daily dose. The true intake of supplements on the day of the 24-h recall was not determined and therefore the additional nutrient intake is only an estimate of the true intake. The intake from supplements may also have been underestimated as supplements for which nutrient content could not be found or imputed were not included. Further, caution must be used in interpreting the prevalence data. The estimations of the prevalence of inadequate intakes presented here are likely overestimates of the true prevalence because the distribution of intakes was not corrected for day-to-day variation [32]. This correction can be made when more than one day of intake data is available and results in a narrower intake distribution that is more representative of the true distribution of intakes in the population [32]. Nonetheless, the results of this analysis are still useful for estimating the relative contribution of supplements to total nutrient intakes in Mexico.
In summary, although the use of supplements by women in Mexico is not as high as in some other countries, we showed that failure to consider supplement use results in a subestimation of the intake of many micronutrients. At this time, voluntary supplement use does not seem to pose much concern regarding toxic intakes at the population level, but this should continue to be monitored. This analysis will be useful in informing the design of future dietary surveys in Mexico and will serve as a starting point for assessing trends in supplement use in this population.
Received February 14, 2006.
Revised April 21, 2006.
| REFERENCES |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |