ABSTRACT
Estimated intake of six low and no-calorie sweeteners (LNCS) (acesulfame potassium, aspartame, cyclamate, saccharin, steviol glycosides and sucralose) from processed foods and beverages, as well as from tabletop sweeteners uses, by the Brazilian population were derived and compared to the respective Acceptable Daily Intake (ADI). The estimates were based on dietary consumption data from a nationwide cross-sectional survey conducted by the Instituto Brasileiro de Geografia e Estatística (IBGE) in 2008–2009 and LNCS use rates and associated market share information based on data provided by Brazilian industry members. Two intake scenarios were considered: a conservative brand loyal consumer scenario that assumes all LNCS-containing foods and beverages and tabletop sweeteners contain the maximum reported LNCS concentrations (Scenario A) and a scenario representative of the general consumer population that uses a market share weighted average of the reported concentrations (Scenario B). Intake estimates were derived for the total Brazilian population (age 10 + y), and for the subpopulations of adolescents (10–18y), adults (19–59y), and older adults (60 + y). Intake of LNCS up to the 95th percentile did not exceed their respective ADI for all subpopulations considered, in either the general consumer scenario or the brand loyal scenario. Among consumers age 10 + y, the 95th percentile intakes of the various LNCS ranged from 6.8% to 54% of their respective ADI for the brand loyal scenario and from <1% to 6.0% of their respective ADIs for the general consumer scenario.
Introduction
In response to concerns about the potential association between sugar intakes and high obesity rates, and guidelines from the World Health Organisation that recommend reducing the level of sugar in the diet to less than 10%, or even less than 5% (conditional recommendation based on very low and moderate evidence in the case of dental health) (WHO Citation2015), low and no-calorie sweeteners (LNCS) may be used in lieu of sugar as one possible alternative to decrease sugar intake. Increased LNCS intakes are suggested to be associated with such substitutions (Sylvetsky and Rother Citation2016), leading to questions about the risk of exceeding the respective LNCS Acceptable Daily Intake (ADI) (Serra-Majem et al. Citation2018).
A comprehensive review of the global intake data on the major LNCS (acesulfame potassium, aspartame, saccharin, sucralose, cyclamate, and thaumatin) during a 10-year period from 2008 to 2018 did not show any exceedance of individual sweetener ADIs among the general population (Martyn et al. Citation2018). It was however noted in this review that LNCS intake data for Latin America were limited to a single study conducted in Brazil (1990–1991) by Renwick (Citation2008), in which the ADIs were not exceeded for the evaluated sweeteners (aspartame, cyclamate and saccharin). Martyn et al. (2018) did not identify any study examining sweetener intake in Brazil using available national food consumption data and recommended future LNCS intake assessments make use of such valuable information.
In the current study, food consumption data from a nationwide cross-sectional survey conducted by the Instituto Brasileiro de Geografia e Estatística (IBGE) in 2008–2009, along with LNCS use rates and associated market share information provided by Brazilian industry members, were relied upon to develop estimates of dietary exposure to six LNCS (acesulfame potassium, aspartame, cyclamate, saccharin, steviol glycosides and sucralose) among the Brazilian population. The LNCS intake estimates were compared to their respective Joint FAO/WHO Expert Committee on Food Additives (JECFA) ADI for the following two exposure scenarios: (A) a conservative brand loyal consumer scenario that assumes all LNCS-containing foods and beverages and tabletop sweeteners contain the same LNCS at the maximum reported concentrations and (B) a scenario representative of the general consumer population that uses a market share weighted average of the reported concentrations.
Materials and methods
Food consumption data
IBGE database
Estimated intake of LNCS was based on food consumption records collected by the IBGE as part of the Brazilian Household Budget Survey in 2008–2009. A randomly selected sub-sample of 13,569 households from the 55,970 private households participating in the Household Budget Survey participated in the Individual Food Intake component. A total of 34,003 individuals provided individual food intake data via dietary records.
Each participant in the dietary survey component recorded all foods consumed on two non-consecutive days and maintained a record of each food consumed, along with the method of preparation, the amount of food consumed in common measures, and the time and place of intake (at home or away from home). In addition, respondents completed a separate question on whether they frequently add “açúcar” (sugar), “adoçante” (sweeteners), “açúcar e adoçante” (sugar and sweeteners), or “não utiliza” (do not use [sugar and/or sweeteners]) to their foods and beverages. The survey records were reviewed by trained IBGE survey staff who probed about any additional foods and beverages that may have been consumed to confirm that the information recorded by the participants was correct.
The current analysis is based on the first day of food records as not all participants completed two days of dietary recall and as the recollection of foods and amounts consumed on the first day of dietary records/recalls is usually more complete than on subsequent days (Subar et al. Citation2003).
Identifying foods and beverages that may contain LNCS
Foods reported consumed in the dietary survey component were identified and classified as containing, or potentially containing, LNCS, based on their names or the name of the food that was used to obtain the nutrient composition associated with each food code (provided with the IBGE survey documentation, typically in English), and their sugar content. Foods and beverages with the terms “adoçante”, “sweetener”, “diet” or “light” were identified, and the nutrient compositions of the identified foods and beverages were also reviewed to confirm that those foods identified as “diet” or “light” were low sugar as opposed to low fat.
The identified foods and beverages were grouped into 26 categories, e.g., “Non-nutritive sweetener, liquid”, “Non-nutritive sweetener, powder”, “Diet carbonated soda, cola type”, “Diet carbonated soda, fruit flavoured”, “Diet jello”, etc. Using this list of foods and beverages, members of the Associação Brasileira da Indústria de Alimentos para Fins Especiaise Congêneres (Brazilian Association of the Industry of Food for Special Purposes and Similar, or ABIAD) were surveyed to obtain the use rates of the various LNCS and the associated market share for each food/beverage category.
Use of LNCS tabletop sweeteners
While the IBGE survey collected information of whether participants frequently add sugar or sweeteners to their foods and beverages, it did not collect information on the amount of tabletop sugar or sweeteners that are typically added.
Zanini Rde et al. (Citation2011) reported that in a survey of adults age 20 + y in the Pelotas, Rio Grande do Sul State, Brazil, the mean per serving intake of liquid and powdered sweeteners added to beverages was 6 drops and 1.5 packets, respectively. Data from sweetener producers that participated in the ABIAD survey indicated that depending on the LNCS, 6–10 drops of liquid LNCS, and 600–800 mg of powder LNCS, provide an equivalent of 2 teaspoons of sugar. Based on this information, it was assumed that the amount of sweetener added to commonly sweetened beverages was equivalent to the amount corresponding to two teaspoons of sugar per cup for typical users of sweeteners (or the amount equivalent to one teaspoon of sugar per cup for typical users of a combination of sugar and sweeteners).
LNCS use rates and market share
LNCS use rates
Fifteen surveyed ABIAD members provided information on the type and amount (mg kg−1) of LNCS added to the foods and beverages they manufacture. The provided LNCS use rates did not always match the form in which the foods or beverages were reported consumed in the IBGE database. For example, the reported consumption of diet fruit drink, tea, and Jello gelatin in the IBGE database was for the prepared beverages/food while the ABIAD provided use rates for these foods were for the dry mix powder. On the other hand, for the diet nutritional drinks, the reported consumption in the IBGE database was for the powder, while the use rate provided by ABIAD was for the prepared, ready-to-drink, beverage. Default adjustment factors based on U.S. Department of Agriculture (USDA) recipes for similar foods/beverages were used (USDA Citation2018) to reconstitute or dilute beverages to the food form of interest. Specifically, adjustment factors of 5%, 4%, and 36% were used to convert the LNCS concentrations in the diet fruit drink, tea and Jello dry powder mixes into an as consumed basis, while an adjustment factor of 20 was used to convert LNCS concentrations in the ready to drink diet nutritional beverage to a dry mix powder basis. Information provided by ABIAD on recommended dilution rates for some of the products indicates that the adjustment factors used in the current study are likely conservative, resulting in higher concentrations of LNCS in the finished products, but given that the dilution information was not provided for all applicable products, the USDA-based default concentration factors were used in the current analysis.
LNCS market share information
Estimates of market share were provided by the surveyed ABIAD members either as point estimates or as ranges. It was assumed that the range in the market share reflects uncertainty in the actual market share and the midpoint of that range was used. Individual products in categories where market share information was missing for all products were assumed to have equal market share within that category. Hence, if a category had four products, each product was assigned a 25% market share. For categories where market share information was missing for some, but not all products within the category, it was assumed that the individual products with missing market share have a market share equal to the average market share of the products with information. Thus, if a category had three products, one of which (Product X) was missing market share information, and the average reported market share of the two other products was 25%, then it was assumed that the market share for Product X was 25%. For categories where market share information was available on all products, the reported market share was used. Finally, it was assumed that the information provided by ABIAD represented the overall LNCS uses in processed foods and beverages in Brazil. Therefore, the market shares were standardised to sum to 100% within each food/beverage category.
The market share information was used to derive an average LNCS concentration per food/beverage category for use in the general consumer population assessment (Scenario B). summarises the approach and assumptions used to assign market share for the products with missing information.
Table 1. Market share assignment strategy
LNCS concentrations assigned to each food/beverage category
LNCS use rates were provided by ABIAD members either as point estimates or as ranges. It was assumed that the ranges could be reflecting an actual use range for various products manufactured by the ABIAD member. Therefore, for each product category and each LNCS, two use rate values were derived as follows for use in the intake assessment (see ):
The maximum of the use rates (or upper limit of the use rate if the rate was provided as a range), across all products within a category. This estimate was not adjusted for market share and was used in Scenario A, representing the “brand loyal” consumer.
A weighted average of the use rates (or midpoint of the use rate if the rate was provided as a range), where the weights reflect the market share of the individual products within a category. This estimate was used in Scenario B, representing the “general” consumer.
Table 2. LNCS use rate derivation for each scenario
summarises the resulting LNCS use rates assigned to each food/beverage category. Potential exceedances of the LNCS use rates reported in over regulatory use limits are most likely due to the use of default conservative dilution/concentration factors derived from USDA recipes.
Table 3. LNCS use rate per food category and LNCS
Statistical methods
The LNCS use rates in were assigned to all food/beverage products within a category and combined with the corresponding consumption amounts reported by survey participants to obtain LNCS intake estimates from processed foods/beverages. For the tabletop sweeteners, the amount of a sweetener added to a cup of beverage was assumed to be equivalent to two teaspoons of sugar for survey participants reporting “always adding” sweetener (or the amount equivalent to one teaspoon of sugar for survey participants reporting “adding sugar and sweetener”). Finally, a total LNCS intake from processed foods/beverages and from tabletop sweetener uses was derived by adding the intakes form either source at the participant level.
Mean and associated standard errors (SE) and upper percentile (90th and 95th) LNCS intake estimates (on an mg kg-bw−1day−1 basis) for consumers were derived for the total population (10 + y) and three subgroups of adolescents (10–18y), adults (19–59y), and elderly (60 + y) and compared to the respective ADI (). The age subgroups were selected to be comparable to those presented in the summary intake reports generated by (IBGE Citation2011), except that the two younger age groups used in the IBGE database (10–13y and 14–18y) were combined in the current analysis due to the relatively small number of LNCS consumers in these two age groups. Pregnant women were excluded from the analysis, as their intake patterns may be different from those of the general population.
Table 4. JECFA Acceptable Daily Intake (ADI) for the six LNCS
All estimates were calculated with sampling weights and adjusted for the complex sampling survey design. Statistical analyses were carried out in the statistical software package STATA version 12.1.
Results
Study population
Of the 34,003 survey participants reporting consumption data in the survey, 390 were pregnant females and were excluded from the current assessment. The study population in the current analysis consisted of the remaining 33,613 survey participants.
The percent of the study population consuming LNCS, whether associated with a processed food/beverage or associated with addition of sweeteners to commonly sweetened beverages such as coffee, tea and fruit-based drinks is shown in .
Table 5. Study population by age group and consumption of LNCS
Estimated daily intake of LNCS
Estimated daily intakes of the six LNCS by the 10 + y population for the conservative brand loyal scenario are summarised below and estimates for both scenarios for all population groups are presented in and compared to their respective ADI in .
Table 6. Estimated daily intake of acesulfame potassium (mg kg-bw−1 day−1); Brazilian population
Table 7. Estimated daily intake of aspartame (mg kg-bw−1 day−1); Brazilian population
Table 8. Estimated daily intake of cyclamate (mg kg-bw−1 day−1); Brazilian population
Table 9. Estimated daily intake of saccharin (mg kg-bw−1 day−1); Brazilian population
Table 10. Estimated daily intake of steviol glycosides (mg kg-bw−1 day−1); Brazilian population
Table 11. Estimated daily intake of sucralose (mg kg-bw−1 day−1); Brazilian population
Figure 1. Estimated Daily Intake of Acesulfame Potassium as % of ADI; Brazil Population
Figure 2. Estimated Daily Intake of Aspartame as % of ADI; Brazil Population
Figure 3. Estimated Daily Intake of Cyclamate as % of ADI; Brazil Population
Figure 4. Estimated Daily Intake of Saccharin as % of ADI; Brazil Population
Figure 5. Estimated Daily Intake of Steviol Glycosides as % of ADI; Brazil Population
Figure 6. Estimated Daily Intake of Sucralose as % of ADI; Brazil Population
Acesulfame potassium
Acesulfame potassium intakes from processed foods and beverages, tabletop sweeteners additions to beverages, and from the combined uses were derived and compared to the ADI of 15 mg kg-bw−1 day−1. For the brand loyal consumer, the mean, 90th, and 95th percentile estimated intake of acesulfame potassium for the 10 + y population were 0.81, 1.50 and 2.04 mg kg-bw−1 day−1, respectively, for consumers of acesulfame potassium-containing processed foods and beverages, and 0.23, 0.48, and 0.56 mg kg-bw−1 day−1, respectively, for tabletop users of acesulfame potassium sweeteners in beverages, while the combined (total) acesulfame potassium intake for consumers of acesulfame containing processed foods and beverages or tabletop sweeteners, was 0.35, 0.71, and 1.02 mg kg-bw−1 day−1, respectively (). The mean, 90th and 95th percentile total acesulfame potassium intake for the 10 + y population brand loyal consumers and the general consumers were 2.3%, 4.7% and 6.8%, and 0.3%, 0.5%, and 0.9%, respectively, of the ADI of 15 mg kg-bw−1 day−1 ().
Aspartame
Aspartame intakes from processed foods and beverages, tabletop sweetener additions to beverages, and from the combined uses were derived and compared to the ADI of 40 mg kg-bw−1 day−1. For the brand loyal consumer, the mean, 90th, and 95th percentile estimated intake of aspartame for the 10 + y population were 1.99, 3.48, and 4.49 mg kg-bw−1 day−1, respectively, for consumers of aspartame-containing processed foods and beverages, and 0.78, 1.61, and 1.89 mg kg-bw−1 day−1, respectively, for tabletop users of aspartame sweeteners in beverages, while the combined (total) aspartame intake for consumers of aspartame-containing processed foods and beverages or tabletop sweeteners, was 1.02, 1.94, and 2.70 mg kg-bw−1 day−1, respectively (). The mean, 90th and 95th percentile total aspartame intake for the 10 + y population brand loyal consumer and the general consumer were 2.6%, 4.9% and 6.8%, and 0.2%, 0.4%, and 0.5%, respectively of the ADI of 40 mg kg-bw−1 day−1 ().
Cyclamate
Cyclamate intakes from processed foods and beverages, tabletop sweetener additions to beverages, and from the combined uses were derived and compared to the ADI of 11 mg kg-bw−1 day−1. For the brand loyal consumer, the mean, 90th, and 95th percentile estimated intake of cyclamate for the 10 + y population were 3.59, 6.34 and 8.57 mg kg-bw−1 day−1, respectively, for consumers of cyclamate containing processed foods and beverages, and 0.47, 0.96, and 1.14 mg kg-bw−1 day−1, respectively, for tabletop users of cyclamate sweeteners in beverages, while the combined (total) cyclamate intake for consumers of cyclamate containing processed foods and beverages or tabletop sweeteners, was 0.93, 2.57, and 3.85 mg kg-bw−1 day−1, respectively (). The mean, 90th and 95th percentile total cyclamate intake for the 10 + y population brand loyal consumer and the general consumer were 8.5%, 23% and 35%, and 1.5%, 3.1%, and 4.1%, respectively, of the ADI of 11 mg kg-bw−1 day−1 ().
Saccharin
Saccharin intakes from processed foods and beverages, tabletop sweetener additions to beverages, and from the combined uses were derived and compared to the ADI of 5 mg kg-bw−1 day−1. For the brand loyal consumer, the mean, 90th, and 95th percentile estimated intake of saccharin for the 10 + y population were 0.48, 0.93, and 1.10 mg kg-bw−1 day−1, respectively, for consumers of saccharin containing processed foods and beverages, and 0.48, 0.99, and 1.16 mg kg-bw−1 day−1, respectively, for tabletop users of saccharin sweeteners in beverages, while the combined (total) saccharin intake for consumers of saccharin containing processed foods and beverages or tabletop sweeteners, was 0.50, 1.00, and 1.19 mg kg-bw−1 day−1, respectively (). The mean, 90th and 95th percentile total saccharin intake for the 10 + y population brand loyal consumer and the general consumer were 10%, 20% and 24%, and 2.4%, 5.0%, and 6.0%, respectively, of the ADI of 5 mg kg-bw−1 day−1 ().
Steviol glycosides
Steviol glycosides intakes from processed foods and beverages, tabletop sweetener additions to beverages, and from the combined uses were derived and compared to the ADI of 4 mg kg-bw−1 day−1. For the brand loyal consumer, the mean, 90th, and 95th percentile estimated intake of steviol glycosides for the 10 + y population were 0.63, 1.58, and 1.83 mg kg-bw−1 day−1, respectively, for consumers of steviol glycosides containing processed foods and beverages, and 0.89, 1.83, and 2.15 mg kg-bw−1 day−1, respectively, for tabletop users of steviol glycosides sweeteners in beverages, while the combined (total) steviol glycosides intake for consumers of steviol glycosides containing processed foods and beverages or tabletop sweeteners, was 0.89, 1.83, and 2.17 mg kg-bw−1 day−1, respectively (). The mean, 90th and 95th percentile total steviol glycosides intake for the 10 + y population brand loyal consumer and the general consumer were 22%, 46% and 54%, and 1.3%, 2.3%, and 2.8%, respectively, of the ADI of 4 mg kg-bw−1 day−1 ().
Sucralose
Sucralose intakes from processed foods and beverages, tabletop sweetener additions to beverages, and from the combined uses were derived and compared to the ADI of 15 mg kg-bw−1 day−1. For the brand loyal consumer, the mean, 90th, and 95th percentile estimated intake of sucralose for the 10 + y population were 0.89, 1.49, and 2.02 mg kg-bw−1 day−1, respectively, for consumers of sucralose containing processed foods and beverages, and 0.27, 0.55, and 0.65 mg kg-bw−1 day−1, respectively, for tabletop users of sucralose sweeteners in beverages, while the combined (total) sucralose intake for consumers of sucralose containing processed foods and beverages or tabletop sweeteners, was 0.39, 0.80, and 1.13 mg kg-bw−1 day−1, respectively (). The mean, 90th and 95th percentile total sucralose intake for the 10 + y population brand loyal consumer and the general consumer were 2.6%, 5.3% and 7.5%, and 0.4%, 0.6%, and 0.8%, respectively, of the ADI of 15 mg kg-bw−1 day−1 ().
Estimated LNCS intake for consumers of LNCS sweetened processed foods and beverages and tabletop sweeteners
An additional analysis was conducted to evaluate intake of the various LNCS for the consumers with joint exposure to LNCS from both processed foods/beverages and tabletop sources, that is for IBGE participants who reported consuming LNCS-containing foods and beverages as well as LNCS-containing tabletop sweeteners in beverages. Given the small number of such LNCS consumers in the IBGE survey, ranging from 24 (0.1%) for steviol glycosides to 176 (8.4%) for acesulfame potassium (), this analysis was conducted on the total 10 + y population only.
Table 12. Estimated daily intake of LNCS from combined processed foods/beverages and tabletop sources (mg kg-bw−1 day−1); Brazilian consumers of both processed foods/beverages and tabletop sweeteners (10 + y)
Even in the extreme unlikely scenario that survey participants who consume LNCS-containing foods/beverages and tabletop sweeteners in beverages consume products that contain the maximum reported LNCS concentrations (scenario A), the combined LNCS intake for these extreme brand loyal consumers remains below the respective ADI at the mean as well as the upper 90th and 95th percentiles of intake.
Discussion
Results of the present study provide nationally representative intake estimates of six LNCS by the Brazilian population based on dietary records collected from 33,613 individuals aged ≥10 years by the IBGE as part of the Brazilian Household Budget Survey (2008–2009). The IBGE survey did not collect information on the amount and the sweetener types in the LNCS foods and beverages reported consumed. In addition, while the IBGE collected information on whether participants typically add sweeteners to their beverages, it did not collect information as to what type of sweetener or how much sweetener is typically added. To fill this data gap, we analysed LNCS usage rates and market share data obtained from a survey completed by ABIAD members and published data on typical amounts of sweeteners added to beverages.
The IBGE food/beverage categories identified as containing LNCS in the current studies overlap the categories identified in a recent study by Figueiredo et al. (Citation2018) that summarised information on the presence of LNCS in packaged foods and beverages sold in a supermarket in Brazil. Most of the food/beverage categories with the highest prevalence of LNCS usage in the Figueiredo et al. (Citation2018) study – i.e., baked goods (including cakes and granola bars), dessert mixes, powdered drinks, soft drinks, jelly, and candy – were also identified in the current study as containing LNCS, thus lending support to the use of the IBGE survey data to assess intake of LNCS by the Brazilian population.
The estimated prevalence of LNCS use in the current study is comparable to estimates reported by others (Zanini Rde et al. Citation2011; Silva Monteiro et al. Citation2018; Arrais et al. Citation2019). Zanini Rde et al. (Citation2011) reported that 19% of adults (>20y) surveyed in Pelotas, the third most populous city in the southern state of Rio Grande do Sul, add sweeteners to their beverages, with prevalence increasing from 8.7% among the 20–29y subgroup to 32% among the 60 + y subgroup. Arrais et al. (Citation2019) reported that 13.4% of surveyed adults living in urban areas reported they use sweeteners in their diet, with a higher prevalence among the elderly 60 + y (29.5%) as compared to younger adults, ages 20–39y (10.3%) and 40–59y (20.3%). Using data from the IBGE survey, Silva Monteiro et al. (Citation2018) estimated the prevalence of tabletop LNCS consumers among adults and elderly to be 12.4% and 28.6%, respectively. The percent consumers from these studies are comparable to the percent consumers of tabletop sweeteners presented in for adults 19–59y (10.9%) and 60 + y (26.5%). On the other hand, the prevalence of LNCS consumption in a study conducted by Geraldo and Pinto-e-Silva (Citation2016) in 2016 among staff of two public universities in the state of São Paulo, Brazil was 54%, higher than the prevalence estimated in the current study and those estimated by Zanini Rde et al. (Citation2011), Arrais et al. (Citation2019), and Silva Monteiro et al. (Citation2018). Geraldo and Pinto-e-Silva (Citation2016) attributed the higher estimated prevalence of LNCS consumers in their study as compared to those obtained by Zanini Rde et al. (Citation2011) to regional, income, and male to female ratio differences between the two studies.
Estimated intakes were derived using the first day of food records, similar to the approach used in other studies that analysed the IBGE data (Souza Ade et al. Citation2013; Pereira et al. Citation2014; Murphy et al. Citation2016; Silva Monteiro et al. Citation2018). Single 24-hr dietary records/recalls have been shown to provide reliable estimates of population means (Dodd et al. Citation2006), estimates of upper percentile intakes, and even mean estimates among consumers, based on a single day record/recall are typically higher than estimates based on multiple days (Lambe and Kearney Citation1999; Hoffmann et al. Citation2002). Thus, the estimated intakes reported in the current study are conservative estimates of consumers’ long-term LNCS intakes. Despite this likely overestimation, the estimated intakes for both the general consumer scenario and the brand loyal scenario were below the respective LNCS ADIs even at the 95th percentile of the intake distributions for all LNCS and subpopulations considered. Further, even when the analysis was restricted to the subset of consumers with joint exposure to LNCS in tabletop uses and processed foods/beverages, the estimated intakes were below the respective ADIs for all six LNCS.
The intake of LNCS associated with the consumption of LNCS-containing processed foods and beverages among consumers of these products was generally higher than the intake of LNCS associated with the consumption of LNCS-containing tabletop sweeteners among consumers of these sweeteners. A similar finding was observed by Toledo and Ioshi (Citation1995) who surveyed a random sample of Brazilian adults who consume high-intensity sweeteners (or LNCS).
The estimated total LNCS intake for consumers of LNCS from either processed foods and beverages or from tabletop sweeteners in the current study is comparable to the intake of LNCS from tabletop usage, as the number of consumers of tabletop sweeteners was much higher than the number of consumers of LNCS-containing processed foods/beverages. These findings are also similar to the findings reported by Toledo and Ioshi (Citation1995) who found that the main source of high-intensity sweetener intake was tabletop uses and driven by the high percent of consumers of tabletop sweeteners as compared to the percent of consumers of LNCS-containing foods/beverages. The percent consumers of LNCS from tabletop sweeteners ranged from 68% to 99% among the LNCS consumers that were included in the Toledo and Ioshi (Citation1995) study. In the current study, the estimated percent consumers of LNCS from tabletop usage among those reporting LNCS use from either source were comparable to the findings in the Toledo and Ioshi (Citation1995) study and ranged from 89% (for acesulfame potassium and sucralose) to 98% (for steviol glycosides) (data not shown).
The main strengths of this study include the use of a nationally representative sample to estimate consumption of LNCS-containing foods/beverages and tabletop sweeteners and use of actual LNCS use rates obtained for a survey of members of a national organisation of producers of LNCS foods/beverages in Brazil. Potential limitations of this study include the age of the consumption data (2007–2008); however, as noted above the estimates of the per cent consumers of tabletop sweeteners obtained in this study are consistent with estimates reported by Arrais et al. (Citation2019) which were based on data collected in 2013–2014, indicating that the prevalence of LNCS-containing tabletop sweeteners has remained relatively stable in Brazil.
LNCS intake estimates for Latin America remain limited and available national food consumption data should be utilised to generate better information (Martyn et al., 2018). The current study, which uses nationally representative food consumption data fills that gap and demonstrates that intakes of the six LNCS considered in this study, namely acesulfame potassium, aspartame, cyclamate, saccharin, steviol glycosides and sucralose, were well below their respective ADI, not only for the general consumers (Scenario B) but also for the brand loyal consumers (Scenario A) who were assumed to always consume foods and beverages and tabletop sweeteners that contain the same LNCS at the maximum use levels reported by ABIAD members.
Correction Statement
This article has been republished with minor changes. These changes do not impact the academic content of the article.
Additional information
Funding
References
- Arrais PSD, Vianna MPN, Zaccolo AV, Moreira LIM, Thé PMP, Quidute ARP, Fontanella AT, Pizzol TDSD, Tavares NUL, Oliveira MA, et al. 2019. Utilização de adoçantes no Brasil: uma abordagem a partir de um inquérito domiciliar [Use of artificial sweeteners in Brazil: a household survey approach]. Cad Saude Publica. 35(11):e00010719. Portuguese. doi:10.1590/0102-311X00010719.
- Dodd KW, Guenther PM, Freedman LS, Subar AF, Kipnis V, Midthune D, Tooze JA, Krebs-Smith SM. 2006. Statistical methods for estimating usual intake of nutrients and foods: a review of the theory. J Am Diet Assoc. 106(10):1640–1650. doi:10.1016/j.jada.2006.07.011.
- Figueiredo LDS, Scapin T, Fernandes AC, Proença RPDC. 2018. Where are the low-calorie sweeteners? An analysis of the presence and types of low-calorie sweeteners in packaged foods sold in Brazil from food labelling. Public Health Nutr. 21(3):447–453. doi:10.1017/S136898001700283X.
- Geraldo APG, Pinto-e-Silva ME. 2016. Nonnutritive sweeteners in Brazil: current use and associated factors. J Hum Growth Dev. 26(3):297–306. doi:10.7322/jhgd.122760.
- Hoffmann K, Boeing H, Dufour A, Volatier JL, Telman J, Virtanen M, Becker W, De Henauw S. 2002. Estimating the distribution of usual dietary intake by short-term measurements. Eur J Clin Nutr. 56(Suppl 2):S53–S62. doi:10.1038/sj.ejcn.1601429.
- [IBGE] Instituto Brasileiro de Geografia e Estatística. 2011. Pesquisa de Orçamentos Familiares, 2008–2009: análise do Consumo Alimentar Pessoal no Brasil. Rio de Janeiro, Brazil: Instituto Brasileiro de Geografia e Estatística. https://www.ibge.gov.br/does not include information on State
- Lambe J, Kearney J. 1999. The influence of survey duration on estimates of food intakes–relevance for food-based dietary guidelines. Br J Nutr. 81(Suppl 2):S139–S142. doi:10.1017/s0007114599000999.
- Martyn D, Darch M, Roberts A, Lee HY, Yaqiong Tian T, Kaburagi N, Belmar P. 2018. Low-/No-calorie sweeteners: A review of global intakes. Nutrients. 2018 Mar 15;10(3):357. doi:10.3390/nu10030357. PMID: 29543782; PMCID: PMC5872775.
- Murphy MM, Barraj LM, Toth LD, Harkness LS, Bolster DR. 2016. Daily intake of dairy products in Brazil and contributions to nutrient intakes: a cross-sectional study. Public Health Nutr. 19(3):393–400. doi:10.1017/S1368980015001780.
- Pereira RA, Duffey KJ, Sichieri R, Popkin BM. 2014. Sources of excessive saturated fat, trans fat and sugar consumption in Brazil: an analysis of the first Brazilian nationwide individual dietary survey. Public Health Nutr. 17(1):113–121. doi:10.1017/S1368980012004892.
- Renwick AG. 2008. The use of a sweetener substitution method to predict dietary exposures for the intense sweetener rebaudioside A. Food Chem Toxicol. 46(Suppl 7):S61–S69. doi:10.1016/j.fct.2008.05.009.
- Serra-Majem L, Raposo A, Aranceta-Bartrina J, Varela-Moreiras G, Logue C, Laviada H, Socolovsky S, Pérez-Rodrigo C, Aldrete-Velasco JA, Meneses Sierra E, et al. 2018. Ibero American consensus on low- and no-calorie sweeteners: safety, nutritional aspects and benefits in food and beverages. Nutrients. 10(7):818. doi:10.3390/nu10070818.
- Silva Monteiro L, Kulik Hassan B, Melo Rodrigues PR, Massae Yokoo E, Sichieri R, Alves Pereira R. 2018. Use of table sugar and artificial sweeteners in Brazil: National Dietary Survey 2008-2009. Nutrients. 10(3):295. doi:10.3390/nu10030295.
- Souza Ade M, Pereira RA, Yokoo EM, Levy RB, Sichieri R. 2013. Most consumed foods in Brazil: National Dietary Survey 2008-2009. Rev Saude Publica. 47(Suppl 1):190S–9S. doi:10.1590/s0034-89102013000700005.
- Subar AF, Kipnis V, Troiano RP, Midthune D, Schoeller DA, Bingham S, Sharbaugh CO, Trabulsi J, Runswick S, Ballard-Barbash R. 2003. Using intake biomarkers to evaluate the extent of dietary misreporting in a large sample of adults: the OPEN study. Am J Epidemiol. 2003 Jul 1;158(1):1–13. doi: 10.1093/aje/kwg092. PMID: 12835280.
- Sylvetsky AC, Rother KI. 2016. Trends in the consumption of low-calorie sweeteners. Physiol Behav. 164(Pt B):446–450. doi:10.1016/j.physbeh.2016.03.030.
- Toledo MC, Ioshi SH. 1995. Potential intake of intense sweeteners in Brazil. Food Addit Contam. 12(6):799–808. doi:10.1080/02652039509374372.
- [USDA] U.S. Department of Agriculture, Agricultural Research Service. 2018. USDA Food and Nutrient Database for Dietary Studies 2015-2016. Food Surveys Research Group Home Page. http://www.ars.usda.gov/nea/bhnrc/fsrg.
- [WHO] World Health Organization. 2015. Guideline: sugars intake for adults and children. Geneva (Switzerland): WHO.
- [WHO] World Health Organization. 2017. Evaluations of the joint FAO/WHO Expert Committee on Food Additives (JECFA). Steviol Glycosides. https://apps.who.int/food-additives-contaminants-jecfa-database/chemical.aspx?chemID=267.
- [WHO] World Health Organization. 2019. Evaluations of the joint FAO/WHO Expert Committee on Food Additives (JECFA). List of chemicals in functional class SWEETENER. https://apps.who.int/food-additives-contaminants-jecfa-database/search.aspx?fc=66.
- Zanini Rde V, Araújo CL, Martínez-Mesa J. 2011. Utilização de adoçantes dietéticos entre adultos em Pelotas, Rio Grande do Sul, Brasil: um estudo de base populacional [Use of diet sweeteners by adults in Pelotas, Rio Grande do Sul State, Brazil: a population-based study]. Cad Saude Publica. 27(5):924–934. doi:10.1590/s0102-311x2011000500010.