Abstract
Background:
Sources of fluoride exposure for Mexicans include foods, beverages, fluoridated salt, and naturally fluoridated water. There are no available data describing fluoride content of foods and beverages consumed in Mexico.
Objective:
To measure the content of fluoride in foods and beverages typically consumed and to compare their content to that of those from the United States and the United Kingdom.
Methods:
Foods and beverages reported as part of the Mexican Health and Nutrition Survey (n = 182) were purchased in the largest supermarket chains and local markets in Mexico City. Samples were analyzed for fluoride, at least in duplicate, using a modification of the hexamethyldisiloxane microdiffusion method. Value contents were compared to those from the US Department of Agriculture and UK fluoride content tables.
Results:
The food groups with the lowest and highest fluoride content were eggs (2.32 µg/100 g) and seafood (371 µg/100 g), respectively. When estimating the amount of fluoride per portion size, the lowest content corresponded to eggs and the highest to fast foods. Meats and sausages, cereals, fast food, sweets and cakes, fruits, dairy products, legumes, and seafood from Mexico presented higher fluoride contents than similar foods from the United States or the United Kingdom. Drinks and eggs from the United States exhibited the highest contents, while this was the case for pasta, soups, and vegetables from the United Kingdom.
Conclusion:
The majority of items analyzed contained higher fluoride contents than their US and UK counterparts. Data generated provide the first and largest table on fluoride content, which will be useful for future comparisons and estimations.
Highlights
Sources of fluoride exposure in Mexico include foods, beverages, fluoridated salt, and naturally fluoridated water.
The majority of foods and beverages consumed in Mexico City and analyzed in this study contain higher fluoride contents than their US and UK counterparts.
The data generated in this study will be useful to facilitate the monitoring of fluoride intake in the Mexican population.
Introduction
At appropriate levels, fluoride has been established as a safe and effective agent in the prevention of dental caries. 1 On the other hand, excessive intake of fluoride has been linked to detrimental effects, including the development of dental enamel and skeletal fluorosis, increased bone fractures, and, more recently, neurodevelopmental deficits in children. 2 -5
Since 1993, salt is fluoridated (either with sodium fluoride or potassium fluoride) as a measure to prevent dental caries in Mexico. The initial regulations established a fluoride concentration of 250 (50) ppm in salt, 6 but a later modification to the norm indicates a maximum concentration of 250 ppm of fluoride ion in the salt for human consumption (with the exception of municipalities where water contains >0.7 ppm fluoride). 7
Monitoring the intake of fluoride to ensure its protective effects is maximized while minimizing or preventing the detrimental effects caused by excessive exposure has become a need, particularly in areas where community fluoridation programs have been implemented. In addition to fluoride naturally present in water or added as part of community fluoridation programs to water or salt, the major sources of fluoride for humans are foods, beverages, dietary supplements, and for infants and toddlers, dental professional products and reconstituted formulae. 8 These sources must be taken into account when estimating the total fluoride exposure of an individual or population. 9
The US Department of Agriculture (USDA) coordinated the development of a National Fluoride Database of Selected 403 Beverages and Foods to be able to estimate fluoride intake through the diet. 10 A similar effort has been completed in the United Kingdom where a fluoride database containing 518 commercially available foods and drinks was compiled recently. 11
Results of observational studies in the United States, Canada, the United Kingdom, and Colombia designed to assess fluoride intake have shown that up to 70% of the total fluoride intake for children (including infants) may be derived from diet. 12 -18 In Mexico, a recent study analyzed samples of juices and carbonated products and reported the mean fluoride concentrations between 0.49 and 0.67 mg F/L. These findings suggest that fluoride ingested through bottled drinks may represent an important part of the total fluoride ingested by Mexican children. 19 In addition, variable fluoride concentrations were found in products like bottled water and other bottled drinks (0.08-1.42 mg F/L for juices, 0.07-1.30 mg F/L for fruit drinks, and 0.10-1.70 mg F/L for other bottled drinks). 20
In spite of the previous efforts to assess fluoride intake in Mexico, no comprehensive study has been conducted that includes a representative list of foods ingested by the Mexican population, as in other countries. Consequently, the aim of this study was first to estimate fluoride concentrations in the most frequently consumed foods and beverages in Mexico according to the National Health and Nutrition Survey (2012). As a second aim, the data were compared to 2 available databases for fluoride content: the USDA National Fluoride Database of Selected Beverages and Foods and the Fluoride Content of Selected Drinks and Foods in the UK tables, to establish the main differences between fluoride content of foods and beverages prepared with fluoridated water versus salt.
Methods
Selection of Foods
Using the data from the Mexican National Health and Nutrition Survey 2012 (ENSANUT), we selected the 100 more frequently consumed foods and beverages, reported in the 24-hour recall. 21 The list of the 100 more frequently consumed foods at the national level according to the 24-hour recall did not include all items of the Food Frequency Questionnaire, also applied in the same survey; therefore, we included the missing items (n = 25). Finally, qualitative information collected as part of an anthropological study was taken into account to include other foods frequently consumed by Mexico City adolescents participating in a cohort study (n = 57). 22 These foods were mostly candies and beverages that we considered useful to include to have a more complete list that will be crucial to develop future studies for the cohort Early Life Exposure in Mexico to ENvironmental Toxicants (ELEMENT), 23 which in turn may influence national policy.
In total, 182 different foods (n = 166) and beverages (n = 16) encompassed in 14 food groups were selected: (1) beverages; (2) meats, processed meat, and poultry; (3) cereals; (4) fast food; (5) Mexican food; (6) sweets, pastries, and snacks; (7) fruits; (8) egg; (9) dairy products; (10) legumes; (11) soups and pasta; (12) seafood; (13) vegetables; and (14) others.
Purchasing of Samples
Grocery stores and farmers’ markets in Mexico City were identified for the purchase of foods and beverages using the following criteria: (1) We bought fruits and vegetables in 3 different major markets in the city: La Central de Abastos Market and La Merced Market, as they distribute food to the majority of the smaller local markets and stores; one public market located in Mexico City (Mi Mercado Tacubaya). (2) We selected 4 large supermarket chains, the main distributors in Mexico City, with the largest geographical coverage: Walmart (Walmart and Bodega Aurrera), Soriana, Comercial Mexicana (Comercial Mexicana and Mega Comercial Mexicana), and Chedraui, all of which have multiple local stores across Mexico City (North, South, Center, West, and East areas). Meat and dairy products, as well as processed industrialized foods, were purchased from these supermarkets. (3) In the case of traditional prepared foods (corn-based foods such as tlacoyo, sopes, and tamales) and natural juices, streets vendors were chosen for purchasing these items. (4) Finally, other foods were purchased from designated stores, such as ice cream parlors for traditionally flavored waters (a traditional Mexican beverage usually prepared with water, fruit, and table sugar) and creameries for milk (ie, Liconsa milk, an item subsidized by the government).
Foods that are eaten raw or require minimal preparation were purchased in triplicate, including, zucchini squash, chayote squash, eggs, lettuce, lentils, mango, apple, orange, nopales, potato, papaya, cucumber, banana, watermelon, and carrots. For processed foods, at least 2 brand names and different presentations were included, such as sliced bread, breakfast cereals, precooked rice, and soda. For most foods, 3 different samples were purchased, selecting a different place of collection (for fruits and vegetables), or different brands or presentations (in the case of processed foods) to account for variability. All foods were bought between October 2014 and March 2017.
Fluoride Analysis
Foods and beverages were refrigerated at 4°C within 8 hours if they were not processed. Solid foods that are usually consumed cooked, such as meats, rice, pasta, and legumes, were boiled in the laboratory using the same brand of water (Evian) containing negligible amounts of fluoride (<0.01 mg/L) with no added salt. Solid foods that were purchased already cooked were thoroughly homogenized for approximately 2 minutes using a household blender and then processed for the analysis. In the case of beverages were directly saved for fluoride analysis in duplicate aliquots (approximately 20 mL).
Deionized water (DW) was added to each solid food sample in the laboratory, and the resulting weight of the food plus the water was recorded. The amount of DW to be added was decided depending on the consistency of each food. In the case of beverages or semiliquid foods, DW was not added.
Once the foods were homogenized, samples were placed in 50 mL scintillation vials. These vials had been previously labeled with the food code and the date on which the samples were processed. After the samples were processed, they were frozen at −30°C. All foods followed this procedure within the 24 hours after their purchase. Frozen aliquots were then shipped to the Indiana University School of Dentistry Oral Health Research Institute for analysis.
Analysis of food and beverages was conducted using a modification of the hexamethyldisiloxane (HMDS: Sigma-Aldrich, Indianapolis, Indiana) microdiffusion method as modified by Martinez-Mier et al as described elsewhere. 24 Briefly, known weights or volumes of samples were pipetted into a plastic Petri dish (Falcon 60 mm × 15 mm disposable Petri dishes, Fisher Scientific/BD Falcon); a sodium hydroxide (NaOH, A.R.: Fisher Scientific) trap solution was placed on the petri dish lid, and after the addition of sulfuric acid (H2SO4: Fisher Scientific) saturated with HMDS, each dish was immediately tightly sealed. During overnight diffusion, fluoride released by acid hydrolysis was trapped in the NaOH trap. The trap was then recovered and buffered to pH 5.2 with acetic acid (CH3COOH: Fisher Scientific). The recovered solution was adjusted to a final volume of 100 µL with DW. Fluoride content of each sample was obtained by comparison of the millivolt reading of the sample to a standard curve prepared from the data for diffused fluoride standard solutions analyzed at the same time.
Statistical Analysis
Descriptive analysis included the fluoride content of every sample, and the mean (standard deviation [SD]) for each food and beverage. Foods and beverages were organized into 14 different groups according to their type; mean and SD for each group were estimated to identify the food groups with the highest and lowest fluoride contents.
Comparisons between fluoride contents of foods, beverages, and food groups in the USDA and UK tables versus the fluoride content in the Mexican foods and beverages were performed. Each item was matched based on the Mexican foods list. Standard portion sizes of each food were also compared between the 3 databases.
Results
The fluoride content data (µg/g, µg/100 g, portion size, and µg/portion size) of foods and beverages within food groups are shown in Table 1. The compiled database shows a wide range in fluoride content between and within the different food groups (the full list of fluoride content of foods and beverages is shown in Supplementary Table 1).
Fluoride Content Data (µg/g, µg/100 g, Portion Size, and µg/Portion Size) of Foods and Beverages Within Food Groups Most Consumed in Mexico.
Abbreviation: SD, standard deviation.
The lowest fluoride contents were observed for vegetable shortening (0.24 µg/100 g), animal lard (0.28 µg/100 g), natural juice (0.50 µg/100 mL), sandwich (0.64 µg/100 g), canola oil (0.65 µg/100 mL), cabbage (0.67 µg/100 g), papaya (0.74 µg/100 g), and lemon (0.87 µg/100 g). Foods with the highest fluoride content were jelly (366.79 µg/100 g), precooked rice (434.38 µg/100 g), powder chocolate (521.16 µg/100 g), wholegrain bread/toast (588.44 µg/100 g), oysters (1461.58 µg/100 g), and fried/baked pork rinds (1465.40 µg/100 g). For the food groups, the ones with the highest median were seafood (371.29 µg/100 g), meats and poultry (191.47 µg/100 g), fast food (118.02 µg/100 g), and legumes (84.91 µg/100 g).
Once the fluoride content per portion size of food was calculated, fast foods were found to contain the highest content of fluoride (209.61 [182.30] µg/portion), followed by seafood (191.99 [305.57] µg/portion). Foods with the lowest fluoride content were eggs (1.27 [0.49] µg/portion) and fruits (5.08 [6.06] µg/portion).
When comparing fluoride content from foods or beverages among countries, differences were found for more than 40% (62 of 150) of equivalent items. When comparing the fluoride content of Mexican foods with USDA and UK data (by 100 g), some important differences by group were found (Table 2). In the case of meats and sausages, cereals, fast foods, candies and pastries, fruits, dairy products, legumes, and seafood, Mexican foods were found to contain more fluoride than their US and UK counterparts. In comparison, US beverages and eggs contained the most fluoride, whereas pasta, soups, and vegetables from the United Kingdom presented the highest fluoride content.
Fluoride Content From Foods or Beverages Among Composition Tables of Different Countries.
Abbreviations: NA, not applicable; SD, standard deviation; USDA, US Department of Agriculture.
Discussion
The present study aimed to develop a comprehensive fluoride database including 182 different foods and beverages, which is the first of its kind in Mexico. The creation of such a database will enable researchers and public health professionals to estimate the dietary fluoride intake in the Mexico City population, assuming consistent fluoride levels. Previous studies have evaluated fluoride intake from foods, 19,20 albeit on a limited scale not comparable to the present comprehensive study. Here, we included the most representative foods and beverages consumed at the national level, with the fluoride content of a total of 166 foods and 16 beverages being determined.
This effort is the first step in understanding the contribution of diet to the total fluoride intake of a Mexican sample. The complexity of understanding diet’s contribution to the total exposure implies taking into account both the fluoride content in foods and the portion eaten. For this reason, it is important to highlight that fast foods appear as the principal source of fluoride, per portion size. This could be explained principally due to the high amount of salt within fast foods. Also, the high levels of fluoride in fast foods could be derived in part by the type of food packing, as a source of fluoride (by contamination), as it has been previously shown that there is a migration of perfluorochemicals into food. 25,26 In the results of our analysis, also other foods with the highest fluoride content were jelly, precooked rice, and powdered chocolate, highly probable contaminated with fluoride by the wrapping materials. Whenever the source is fluoride, this finding is quite important in a country like Mexico where discretionary foods represent more than one-quarter of total energy intake. 27
The comparison of the fluoride content between countries showed multiple differences, which can be explained by previously documented factors that influence fluoride content such as water, soil, and pesticides 8 ; wrapping materials; and Teflon-coated containers. 11,28 The US Department of Agriculture tables show that the beverages have the highest fluoride content, which was expected because water is fluoridated in the United States. In Mexico, however, the food fluoride content is mostly associated with the content of salt. In the United Kingdom, fluoride content in vegetables is particularly high, compared with the other 2 countries; this can be explained by the high fluoride content in soil (ranging from 200 to 400 mg F/kg). 29
It is well known that the fluoride content of any sample presented in these tables could have varied due to intrinsic or extrinsic causes. Some natural (intrinsic) variations may depend on the seasons, soil (plants), and feeding regime (animal products, etc). 30 We had no control over this variation, except for the seasonality as we bought foods during all seasons for almost 3 years. In the case of extrinsic variations, one source could be the containers where food was prepared; when foods were prepared in the laboratory, we didn’t use Teflon-coated containers, 28 but in the case of food that was bought already prepared, we didn’t know the type of containers used. A second important extrinsic source of variation is the type of water used to prepare food. In Mexico, families can use tap water or bottled water for cooking. For the samples prepared in the laboratory, we used Evian water (with a negligible amount of fluoride) in order to capture the content of the food per se, trying to avoid this variation. A third source of variation could be the amount of salt (fluoridated salt) used when cooking the foods; to control for this source of variation, in the case of foods that we cooked in the laboratory, we followed the standardized recipes from the National Health and Nutrition Survey. In summary, we attempted to address the majority of the potential sources variation: measured in duplicate or triplicate every item, selecting different forms of food’s preparations, and selecting different brands in the case of industrialized foods, all of which afford a real-life representation of the true variability, that is, precooked rice was 10 times higher than rice prepared in the laboratory following a standardized recipe.
Our study has some limitations. Even though the list of foods selected for this study represents the national intake, all the samples were purchased in Mexico City, which limit the generalizability to the entire Mexican population. It would be ideal to have an estimate of fluoride content representative of the entire country, but in this first step, we needed to focus our resources on the largest city of the country. Mexico City, as an urban area, has many markets that receive supplies from all the country, especially the Market called “Central de Abastos” that we visited to buy many of the samples, and gets supplies from most of the states. In the case of industrialized foods, especially soda and beverages, our values can’t be taken as representative at the national level, as the industries use the local water to produce the beverages. For example in the north-central part of Mexico, fluoride in water could be as high as 4.5 ppm. 31 This has been shown in previous analysis, where sodas and bottled beverages also have high contents of fluoride. 32
Another limitation is that when comparing the different databases from the 3 countries, some foods lack an equivalent and were therefore not included in the analysis, including flour tortilla or Mexican-style Longaniza sausage. Thus, the averages calculated by food group had different numbers of items.
In conclusion, the information generated in this study has the potential to become a tool to facilitate the monitoring of dietary intake of fluoride, especially in vulnerable populations such as children and pregnant women. Its use in combination with estimates of intake from other sources like salt and toothpaste could give public health authorities a better picture of the total fluoride exposure in the Mexican population.
Supplemental Material
Supplemental_Material - Fluoride Content in Foods and Beverages From Mexico City Markets and Supermarkets
Supplemental_Material for Fluoride Content in Foods and Beverages From Mexico City Markets and Supermarkets by Alejandra Cantoral, Lynda Cristina Luna-Villa, Andres A. Mantilla-Rodriguez, Adriana Mercado, Frank Lippert, Yun Liu, Karen E. Peterson, Howard Hu, Martha M. Téllez-Rojo and Esperanza A. Martinez-Mier in Food and Nutrition Bulletin
Footnotes
Authors’ Note
Esperanza A. Martinez-Mier, Martha M. Téllez-Rojo, Karen E. Peterson, and Howard Hu contributed to conception and design and obtain funding; Lynda Luna-Villa and Alejandra Cantoral did acquisition of data; Andres Mantilla-Rodriguez, Frank Lippert, Lynda Luna-Villa, and Alejandra Cantoral analyzed and interpreted data; Alejandra Cantoral, Lynda Luna-Villa, and Andrés Mantilla-Rodriguez contributed to drafting the manuscript; and Martha Téllez-Rojo and Yun Liu performed statistical analysis; all the authors participated in critical revision and edition of the manuscript.
Acknowledgments
The authors thank Ana Benito for all her support in developing this project. A special thanks to Prithvi Chandrappa and Christine Buckley for their assistance with the laboratory analyses. The American British Cowdray Hospital provided facilities that were used for this research.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The study was supported by the grants R01ES021446, R01ES007821, P42-ES05947, and P30ES017885 from the US National Institute of Environmental Health Sciences (NIEHS), the grant P01ES022844/RD83543601 from NIEHS/US Environmental Protection Agency, by the National Institute of Public Health/Ministry of Health of Mexico.
Supplemental Material
Supplemental material for this article is available online
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
