Artigos

INTRODUCTION
Excessive table sugar consumption is one of the public health concerns worldwide, considering its direct association with excessive weight gain, development of chronic diseases, such as diabetes, and general mortality 1,2,3,4. At the same time, natural sweeteners, including stevia, thaumatin, sorbitol, xylitol, erythritol, or artificial ones, chemically synthesized sweeteners such as aspartame, acesulfame potassium, sucralose and sodium cyclamate, have been increasingly used as an alternative to eliminate or reduce the consumption of table sugar in our diet 5,6,7. However, except for individuals with diabetes mellitus, the World Health Organization (WHO) recently recognized that sweeteners should not be used for weight control or to reduce the risk of chronic non-communicable diseases. Moreover, the recommendation also suggests that long-term use may bring undesirable effects, such as an increased risk of type 2 diabetes, cardiovascular disease and mortality in adults 8.
However, information on the use of table sugar and sweeteners as an added item in the diet is still incipient 9,10. Overall, the information is based on the percentage contribution of dietary table sugar consumption, which varies widely from 8% to 20% in countries in Latin America, Europe and North America 3,4. The prevalence of sweetener consumption has even wider variations, ranging from 15% to 41% for North-American adults 11,12,13,14.
In Brazil, specifically in relation to added table sugar, Pereira et al. 9 found that 7.2% of the total energy intake came from added table sugar, based on data from the first National Dietary Survey (NDS) in 2008-2009. Additionally, Monteiro et al.10 observed that 85.7% and 7.6% of Brazilians used table sugar and sweeteners to sweeten foods and drinks, respectively. Data from related to the second NDS, carried out between 2017-2018, showed that the caloric contribution of table sugar varied from 8% to 12%, and that 85.4% and 8.4% of Brazilians reported using exclusively table sugar or added sweetener, respectively 15. More recent, comparing data from both NDS, Alves et al. 16, found that the use of sweetener remained stable and the table sugar decreased by 8% in ten years.
Among the most frequently added items, scientific literature points to sweetened beverages as the main vehicles of table sugar and sweeteners in the diet of American and European children and adults 3,13, 17,18 and in the Brazilian population 9,10.
Currently, the caloric contribution of table sugar in the diet exceeds the 10% recommended by the (WHO) 2 worldwide 3,18,19 and in Brazil, this caloric contribution in 2008-2009 was 8% 20. In 2017-2018, individuals use of table sugar had more intake of energy 16. Moreover, the new rules for nutritional labeling of packaged foods in Brazil, with changes in the information table and nutritional data and with the adoption of front nutritional labeling 21, may contribute to an increase in the addition of sweeteners to traditional drinks, aiming at reducing the total table sugar content of sweetened drinks to avoid front label information about high table sugar density, similar to what happened in Chile 22. This procedure allows maintaining the sweet taste of table sugar in the drink, while it increases the amount of sweetener in it 21,22.
This manipulation of sweetened drinks sold by industries is not new. According to Russell et al.23, a study on a global analysis of the amount of added table sugar and sweeteners in packaged foods and drinks between 2007 and 2019, showed that, worldwide, the added table sugar volume to beverages reduced by 22% and that of sweeteners increased by 36% 23. In Brazil, approximately 9% of packaged foods and drinks contain at least one type of sweetener, and 83% of foods and drinks with sweeteners also contained sugar, in an analysis carried out in 2017 24.
Additionally, there is still little information on other foods containing added table sugar and sweeteners and there are gaps regarding the sociodemographic profile of the population exposed to high dietary consumption of added table sugar and sweeteners. A cross-sectional study carried out with data from the National Health and Nutrition Examination Survey (NHANES) between 2015 and 2018 showed that table sugar consumption was higher among adults when compared to the elderly, and among individuals with lower income and education. The authors found no difference between the sexes 25. Regarding sweeteners, data from a national survey in Australia showed that the highest consumption was identified among women, individuals with a higher Body Mass Index (BMI), people with diabetes and those on a weight loss diet. The authors found no association between sweetener consumption, income and education 26.
This information can help guide the creation of strategies aimed at reducing and regulating the consumption of table sugar and sweeteners in Brazil. Therefore, the aim of this study was to describe the frequency of use of table sugar and sweetener as added items and the foods with the most added table sugar and sweetener according to the sociodemographic profile and weight status of the Brazilian population.

METHODS
Study design and population
This was a cross-sectional study that used data from the personal food consumption assessment module, known as the National Dietary Survey (NDS), carried out in a subsample of households investigated in the last Household Budget Surveys (HBS) conducted between 2017 and 2018 by the Brazilian Institute of Geography and Statistics (IBGE). Details about sampling and data collection were published by IBGE 15.
The selection of households for HBS 2017-2018 was based on a two-stage cluster sampling. In the first stage, the census sectors were stratified geographically and according to the income of the heads of families and selected by sampling with a probability proportional to the number of households in each sector, considering the demographic base of the 2010 Demographic Census. In the second stage, the permanent households in each census sector were selected without replacement through simple random sampling. In the 2017-2018 HBS, 57,920 households were investigated and to compose the 2017-2018 NDS sample, a subsample of 34.7% of households (n=20,112) was selected with a total of 46,164 individuals who were at least 10 years old. After excluding pregnant and breastfeeding women (n=1,420), the sample comprised 44,744 individuals 15.

Assessment of food consumption
Information on food consumption was collected on two non-consecutive days from a 24h-food recall (R24h) during the collection period at home, and in the present study, information related to the first day of the R24h was used. In it, the individuals, through personal interviews, carried out by trained research agents, were asked about all food and drinks (including water), consumed inside and outside the home, on the day before each of the two interviews 15. These interviews were carried out using the multiple passage method 27, on a tablet with specific software. The computerized data entry program contained a database of 1,832 foods and beverages and the interviewers could add foods that were not included in this database 15.
In an unprecedented way, the participant was asked about the use of some food items added to the food or drinks reported in the R24h. Thus, when the individual reported the consumption of fruit juice, for instance, they were asked whether they added sugar or sweetener to this drink, in addition to other added items 15.
In total, 12 food item options were selected as added items: table sugar, sweetener, olive oil, cream, ketchup, mayonnaise, butter/margarine, honey, molasses, mustard, soy sauce and grated cheese. The choice of added items depended on the type of food that could be added, that is, for drinks, in general, it was possible to choose the addition of table sugar, sweetener and honey. For other foods, options could be available from the list of items described above. Questions about the use of added table sugar and sweeteners covered 218 and 98 foods, respectively. Some examples of foods that raised the question about added table sugar were tubers, fruit juices, artificial juices, coffee, milk, yogurt, milk-based preparations, breakfast cereals, teas and fruits. For the addition of sweetener, the same foods were used as table sugar, with the exception of fruits and teas 15. The foods that could have added table sugar and sweeteners were classified into groups.
This question about possible added items was purely qualitative, whether or not the individual added the item. Therefore, the amount of added sweetener and table sugar in the present analyses was not considered, only the frequency of the reporting of the added item and the frequency of added foods. In the present analysis, we chose to call added table sugar for the purposes of data interpretation and discussion. Table sugar consists of sucrose obtained by industrial processes from sugar cane or beetroot 28. In addition, the term adopted in this study, sweeteners, refers to natural and artificial sweeteners 29. This term was chosen because the NDS did not distinguish between the types of sweeteners.

Sociodemographic characteristics
In relation to sociodemographic variables, the following were considered: sex (male and female), age group defined as: adolescents, between 10 and 19 years old; adults between 20 and 59 years old; and elderly people, over 60 years old, per capita household income stratified according to the minimum wage established at the time of the research (R$ 954 stratified into < 0.5 minimum wage; ? 0.5 and < 1 minimum wage; ? 1 and < 2 minimum wages; ? 2 minimum wages), schooling in complete years according to the phases of the school cycle for adults and elderly individuals (elementary school, high school and higher education), household situation (urban and rural) and macro-regions of Brazil (North, Northeast, Midwest, Southeast and South) 15.
Anthropometric measurements, weight and height, were reported by the individuals at the time of the home interview. Based on these measurements, the Body Mass Index (BMI) was calculated to classify individuals according to weight status categories, without and with excess weight. Adolescents (10 to 19 years old) were considered as having excess weight when their BMI for age was greater than 1 z-score of the reference curve proposed by the World Health Organization 30. As for the adults (20 to 59 years old) and elderly individuals (over 60 years old), they were classified as having excess weight by using the BMI cutoff point ? 25kg/m 31.

Statistical analysis
Frequencies and 95% confidence intervals were estimated for individuals who reported the use of added table sugar and/or sweetener in at least one food and/or drink according to sex, age group, income, education, macro-regions of the Brazil, domicile status and weight status, as of the first day of R24h.
In the present analysis, we chose not to estimate the frequency of concomitant use of table sugar and sweetener, as this represented approximately 1% of the investigated sample. Thus, the frequency of table sugar use included approximately 1% of individuals who also used sweeteners.
Among the food groups that could have added table sugar and sweeteners, the frequency and 95% confidence intervals were estimated and, additionally, the frequencies and 95% confidence intervals were analyzed according to the age group.
Estimates were calculated using the Statistical Analysis System (SAS) software, a free online version on Demand for Academics, taking into account expansion factors and sample complexity. The comparison of frequencies was based on the overlapping of 95% confidence intervals, with the difference between frequencies being considered statistically significant when the 95% CI did not overlap.

Ethical aspects
This is a study that used a secondary database with free and unrestricted access in the public domain, in which the microdata is freely available on the institution's website (https://www.ibge.gov.br/estatisticas/sociais/saude/24786-pesquisa-de-orcamentosfamilies-2.html?=&t=microdados) without the need for approval by the research ethics committee. This study was carried out with the support of the Coordination for the Improvement of Higher Education Personnel – Brazil (CAPES) - Funding Code 001.

RESULTS
The prevalence of the use of table sugar and sweetener as an added item in the diet of the Brazilian population was approximately 63.4% and 8.5%, respectively, and 29% of the Brazilians did not use table sugar or sweetener as an added item. There was no difference in the prevalence of table sugar use by sex. However, there was a difference in the prevalence of sweetener use according to sex, with women (10.8%) using more sweeteners than men (6.2%). In relation to the age group, adults (65.9%) used more table sugar as an added item than adolescents (60.7%) and elderly individuals (57.4%). An increase in the use of sweeteners was observed with increasing age, with the prevalence among the elderly being 2.7 and 15 times higher when compared, respectively, to adults and adolescents. The addition of table sugar in the Northeast region (72.1%) was greater than in other Brazilian regions, and the greatest addition of sweetener was observed in the Southeast region (10.2%) when compared to the other regions. Individuals living in urban areas of the country used 2.4-fold more sweeteners (9.3% urban vs. 3.9% rural) and one-fold less table sugar (62.2% urban vs. 71% rural) as an added item when compared to residents of rural areas (Table 1).
Individuals with income less than 0.5 minimum wage used more table sugar as an added item when compared to those with income ? 2 minimum wages (70.2% vs. 55,6%). Regarding the level of schooling, the prevalence of table sugar use was higher among individuals with elementary education when compared to individuals with higher education (66.7% vs. 54.4%). A gradient of increasing prevalence of sweetener use was observed according to the increase in income and level of schooling categories (Table 1).
Among individuals who did not use table sugar or sweetener as an added item, the highest frequencies were found among adolescents (38.4%), individuals with higher education (32.2%), residents of the Midwest region (40.7%) and from the urban area (29.7%). No differences were found between the sexes and income strata (Table 1).
Elderly individuals of both sexes with excess weight added had less frequency added table sugar when compared to those without excess weight (Men: 57% with excess weight vs. 64.6% without excess weight and Women: 51.6% with excess weight vs. 59.9% without excess weight). Male adolescents with excess weight used two-times more sweetener as an added item when compared to those without excess weight. There was a higher prevalence of sweetener use among adults with excess weight of both sexes when compared to those without excess weight. There was no difference in the prevalence of non-use of table sugar and sweetener as an added item according to weight status (Table 2).
Fruit juices (90%; 95% CI: 87;93), artificial juices (89%; 95% CI: 82;95) and coffee (83%; 95% CI: 82;84) were the foods with the most added table sugar, while artificial juices (42%; 95% CI: 24;60), fruit juices (40%; 95% CI: 33;47) and alcoholic beverages (27%; 95% CI: 16; 70) were the ones with the most added sweeteners (Table 3). The frequency of adolescents who added table sugar to fruit juice (97%; 95% CI: 95;98) was higher than that observed among adults (89%; 95% CI: 85;93) and elderly individuals (75%; 95% CI: 64;87). Adolescents also added more table sugar to coffee (89%; 95% CI: 87;90) when compared to adults (84%; 95% CI: 93;85) and elderly individuals (74%; 95% CI:72;76). Regarding alcoholic beverages, 92% (95% CI: 76;108) of the adults added table sugar, whereas only 32% (95% CI: 12; 75) of elderly individuals did so (Figure 1). However, the frequency of use of sweeteners added to alcoholic beverages was insignificant for adolescents and adults, although it represented 78% (95% CI: 30;126) among the elderly. It is worth mentioning that the frequency of sweetener use among elderly individuals in fruit juices (65%; 95% CI: 49;81) was higher than that observed in adolescents (18%; 95% CI: 1;37) and adults (37%; 95% CI : 30;45), and also for artificial juices (83%; 95% CI: 51;116) when compared to adults (47%; 95% CI: 26;69) and adolescents (24%; 95% CI: 5;54) (Figure 2).

DISCUSSION
In this study, 63.9% of the sample were adults, 50.7% were female, 42.8% lived in the Southeast region, 85.6% in the urban area of the country, 51.8% had elementary education and 31.9% had a family income ? 1 and < 2 minimum wages. The findings showed that the prevalence of the use of added table sugar was higher among residents of rural areas of the country, individuals from the lowest income and level of schooling strata. While the use of sweeteners as an added item was greater among women, the elderly, individuals with excess weight and in the southeast region. The highest prevalence rates of non-use of table sugar and sweeteners as added items were found among adolescents, individuals with higher education and residents of urban areas.
The results corroborate the findings of a national systematic review carried out by Canuto et al.32 which showed that residents of rural areas, with lower income and level of schooling have diets with high levels of sugar, i.e., have a lower chance of having access to healthy foods. The data from this review also showed that table sugar consumption was inversely associated with income and level of schooling, and positively associated with rural areas 32. The difficulty in having access to large food distribution centers and the high cost of fresh and minimally processed foods may contribute to lower consumption of these foods in rural areas. In parallel, urbanization and globalization have led to changes in food consumption, which may include a greater consumption of ultraprocessed foods in the diet of residents in rural areas 33,34. It was also observed from international North-American data from the National Health and Nutrition Examination Survey (NHANES) that individuals in the lowest income quarter were the highest table sugar consumers 35.
Initially, some results may seem divergent from those in the current literature, since in the present study no difference was observed in the prevalence of table sugar use between men and women, and adolescents added less table sugar when compared to adults. However, when observing added table sugar in fruit juices and coffee, the prevalence was higher among adolescents than among adults and the elderly. The findings of two systematic reviews carried out with representative surveys of the population showed that younger individuals and women were those who consumed the highest amount of sugar 3,4. Moreover, according to data from the Latin American Study of Nutrition and Health (ELANS), Fisberg et al.36 also found that women and younger individuals had higher table sugar consumption. In Brazil, data from the NDS 2008-2009 revealed that adolescents used more table sugar to sweeten foods and drinks when compared to adults and the elderly. The authors also showed that the highest frequency of table sugar use was found among men (89%) when compared to women (82%) 10. More recent, data from the NDS 2017-2018, adolescents (91,3%) had more table sugar consumption when compared adults (79,9%) and elderly (64,6%). Furthermore, individuals by male sex consumption more table sugar than individuals by female sex (81,3% vs. 76,4%) 16.
In the present study, it is believed that adolescents did not have a high prevalence of added table sugar because in this age group there is a predominant consumption of foods and/or drinks that already have added table sugar, such as ultraprocessed foods, including soft drinks, artificial juices, cookies, chocolate powder with or similar in milk and sweets. Brazilian adolescents had higher frequencies of consumption of soft drinks (20.9%), dairy drinks (14.2%) and artificial juices (6.1%), when compared to adults and elderly individuals 15. These findings reinforce that adolescents constitute the group that consumes the most ready-to-eat sweetened beverages, as well as other ultraprocessed foods 15,37, and that the high total percentage contribution of table sugar in the diet may come mainly from these industrialized food products, and not necessarily from added table sugar 38,39. According to data from the National School Health Survey, around 97% of Brazilian adolescents aged 13 to 17 consumed ultraprocessed foods on the day before the survey 37. Furthermore, data from the NDS 2017-2018 showed that the consumption of ultraprocessed foods among adolescents was approximately 26.7% when compared to 19.5% and 15.1% of consumption among adults and the elderly, respectively 15. In addition, 88% of the Brazilians adolescents consumption at least one snack per day, and these snacks contributed of higher energy intake and added table sugar 40.
It is worth highlighting that the high table sugar content of Western diets is associated with excess weight, low income and level of schooling and may represent a risk factor for the development of chronic non-communicable diseases 2,41,42,43,44,45. A meta-analysis that evaluated the association between the consumption of added table sugar and sweetened beverages and the risk of cardiovascular disease among North-American women revealed that sugar consumption greater than 15% of the total energy intake and that intake ? 1 serving of sweetened beverages per day has been associated with a higher risk of cardiovascular disease 46. In Brazil, the use of table sugar was related with higher consumption of candies and desserts and less consumption of fruits and vegetables, leading to increased mean intake of energy and added table sugar and lower micronutrient intake 16. Other authors have also demonstrated that diets rich in table sugar are also poorer in nutrients and fresh and minimally-processed foods and have higher concentrations of ultraprocessed foods 47,48, which are associated with excess weight and other chronic diseases 49,50. However, the lack of global standardization of the definition of sugar makes it difficult to compare studies from different countries.
In this sense, with the aim of reducing the damage caused by this risk to the population's health, the World Health Organization (WHO) in 2015 limited table sugar intake to 10% of the energy intake, and even suggested that a reduction to 5% can bring beneficial effects to the population’s health 2. Concomitantly with the recommendation of limited table sugar use, the consumption of sweeteners has grown significantly in an attempt to replace sugars 5,6,8.
The findings on sweeteners were similar to the results of other international and national surveys with population-based representation. According to North-American data from NHANES in the period between 2009 and 2012, the consumption of sweeteners among North-Americans was also higher among women (46% vs. 36%), elderly individuals (50% vs. 30%) and individuals with excess weight (49% vs. 33%) when compared to their peers 14. Brazilian national data showed an increase in the use of sweeteners from 7.6% in 2008-2009 to 8.4% in 2017-2018, which was also higher among the elderly, women, urban residents and individuals with excess weight when compared to younger individuals, men, residents of rural areas and individuals without excess weight, respectively 10,15. Other population-based studies carried out with Brazilian adults over the age of twenty showed that women, elderly people and overweight and individuals with obesity were those with the highest consumption of sweeteners 51,52.
The higher prevalence of sweetener use as an added item among females was expected and may be related to a greater concern about health and weight control, which also happens with the female adult population with excess weight 53. Thus, the positive relationship between the consumption of sweeteners and excess weight can be explained by reverse causality, that is, the presence of a disease or health problem interferes with the individual's behavior; in this case, individuals with excess weight use more sweeteners, compared to individuals without excess weight, according to dietary advice for weight control 54. Similarly, it is believed that the higher prevalence of sweetener use found among elderly people with excess weight can be explained by the fact that aging is associated with the emergence of chronic diseases, with dietary guidelines to limit table sugar consumption, such as diabetes, excess weight and hypertriglyceridemia 55. Other international studies corroborate our findings on the positive relationship between sweetener consumption and age and excess weight 14,56.
Furthermore, the results of the present study showed a higher frequency of added sweeteners among individuals with higher income and level of schooling. NHANES data, carried out between 2009 and 2012, showed that individuals in the highest income and level of schooling strata had the highest prevalence of sweetener consumption 14. Drewnowski and Rehm 13 analyzed the relationship between sociodemographic factors and sweetener use trends among North-Americans and also found a positive association between sweetener consumption and income and level of schooling.
In Brazil, some studies reinforced the higher prevalence of sweetener use in socioeconomic markers. A population-based study with Brazilian adults 51 showed that sweetener consumption increased with income and level of schooling. Moreover, another population-based study showed that individuals with higher level of schooling and income consumed more sweeteners when compared to their peers 52. This relationship can be explained by both the purchasing power, as well as access to information, health services and diagnosis of diseases that may have as a dietary prescription the use of sweeteners to replace sugar, in addition to a greater concern with weight gain and aesthetics, especially among higher-income women.
It is important to say that the indiscriminate use of sweeteners, even in situations where there is no dietary recommendation for their use, is one of concern. Recently, the WHO consolidated robust scientific evidence showing that the use of sweeteners to reduce weight or the risk of chronic non-communicable diseases did not demonstrate the expected benefits. It is worth noting that this recommendation applies to everyone, except those with pre-existing diabetes 8.
It is worth highlighting that in this context, beverages seem to be the main foods that most contribute to an increase in the consumption of added table sugar and sweeteners 3,14,18,57. The findings corroborate previous national results 9,58,59 and showed that fruit juices, artificial juices and coffee were the foods with the most added table sugar and sweeteners. The results also showed that alcoholic beverages contributed in the third and fourth place to greater addition of sweeteners and table sugar, respectively.
The results also revealed that the highest prevalence of individuals who did not use table sugar and sweeteners as an added item were found among adolescents, individuals with higher education and residents of urban areas. No difference was found between the sexes, income strata and weight status. Another study with data from the Brazilian population found no differences between the non-use of sugar and sweeteners and sociodemographic variables 10.
One study limitation was the fact that the survey did not evaluate information on the amount of table sugar or sweetener used by the individuals when added to food and/or beverages. Although this quantitative information was considered in the survey methodology at the data analysis stage, adding 10% of table sugar to the volume or weight of the consumed food/beverage, this standardization probably underestimates the variability of consumption between individuals 15. In other words, the present study was able to measure the prevalence of table sugar and sweetener use in the Brazilian population but was unable to adequately capture the differences in the amounts added to foods. Another limitation of the present analysis refers to the fact that the question about added table sugar and sweetener was asked for only 12% and 5%, respectively, of food items available in the database, that is, some foods may have had added table sugar and/or or sweetener, but the information was not retrieved 15. In addition, other forms of addition were not considered, such as honey. Therefore, it is believed that some people may replace sugar or sweetener with honey. However, only a small portion of the sample mentioned the use of honey as an addition item (0.71%). Finally, a third limitation was the absence of a question about the type of sweetener consumed. It is known that the impact on health differs according to the type of sweetener used, i.e., natural or artificial, as they have different metabolic pathways and, in addition, alter the intestinal microbiota 8,60,61,62.
The strength of the study is related to the novelty of describing the use of table sugar and sweetener as an added item to foods and the main foods with added table sugar and sweeteners according to population characteristics using a representative sample of the Brazilian population. The results helped to understand which population groups most frequently used table sugar or sweetener to sweeten foods and beverages, as well as which foods and beverages are the main vehicles for table sugar and sweeteners in the Brazilian population. It is believed that these findings can contribute to the creation of strategies aimed at reducing and regulating the consumption of table sugar and sweeteners in Brazil.