Free Sugars Consumption in Canada

Free sugars (FS) are associated with a higher risk of dental decay in children and an increased risk of weight gain, overweight and obesity and type 2 diabetes. For this reason, Canada’s Food Guide recommends limiting foods and beverages that contribute to excess free sugars consumption. Estimating FS intakes is needed to inform policies and interventions aimed at reducing Canadians’ consumption of FS. The objective of this study was to estimate FS intake of Canadians using a new method that estimated the free sugars content of foods in the Canadian Nutrient File, the database used in national nutrition surveys. We define FS as sugars present in food products in which the structure has been broken down. We found that 12% of total energy (about 56 g) comes from FS in the diet of Canadians 1 year of age and older (≥1 year). The top four sources were: (1) sugars, syrups, preserves, confectionary, desserts; (2) soft drinks; (3) baked products and (4) juice (without added sugars), and accounted for 60% of total free sugars intake. The results show that efforts need to be sustained to help Canadians, particularly children and adolescents, to reduce their FS intake.


Introduction
Sugars are a type of carbohydrate found naturally in fruit, vegetables and milk. Sugars are added to foods to impart or increase sweetness, for preservation, or confer several functional characteristics such as viscosity, texture and color [1]. Free sugars (FS) are naturally present in honey, syrups, fruit juices and fruit juice concentrates [2]. FS do not include the naturally occurring sources of sugars found in intact or cut fruit and vegetables and (unsweetened) milk [2]. Monosaccharides and disaccharides added to foods and drinks by the manufacturer, cook or consumer are also considered free sugars ( Figure 1) [3]. In order to captures all sources of free sugars in the diet of Canadians, we used an adapted FS definition for this study; see Figure 1 for details. Our definition considers foods in which the structure has been broken down as foods that contain free sugars. Using this methodology will enable the estimation of FS consumption of Canadians in a consistent manner for the purpose of informing healthy eating policies.
Evidence on the links between diet and health show that the intake of FS-particularly in the form of sugary drinks (including 100% fruit juice)-has been associated with a higher risk of dental decay in children [4].
FS are highly prevalent in the Canadian food supply. However, it is difficult for consumers to identify their food sources, as they are not listed on food labels. This is because there is no analytical method to distinguish FS from other sugars that also occur in foods. A study by Bernstein et al. showed that, on average, FS represents 62% of the total sugars amount in prepackaged foods in the Canadian food supply [9].
A study by Bernstein et al. shows that on average FS represent 62% of the total sugars 42 amount in prepackaged foods in the Canadian food supply. (9) 43 Guidance on the consumption of sugar differs around the world. In 2015, the World 44 Health Organization (WHO) published its Sugar intake for adults and children Guideline rec-45 ommending a FS intake of less than 10% of total energy intake to reduce the risk of non-46 communicable diseases. (2) The United States' Dietary Guidelines Advisory Committee 47 recommends "Americans should keep their intake of added sugars to less than 10% of 48 their total daily calories as part of a healthy diet". (7) This recommendation for added sug-49 ars allows a larger amount of free sugar consumption as part of the daily intake because 50 it excludes 100% fruit juice. The United Kingdom's Scientific Advisory Committee on Nu-51 trition recommends that "Free sugars should account for no more than 5% daily dietary 52 energy intake". (4) In line with the guidance based on a strong recommendation by the 53 WHO, Canada's Dietary Guidelines support a FS intake of <10%E.(10) 54 The main objective of this study was to estimate FS intake in various Canadian age-55 sex groups using data from 2015 CCHS-Nutrition. A methodology that could be applied 56 to estimate free sugars content of foods in the 2015 CNF was developed for this purpose. 57 The 2015 CNF database is a standard reference food composition database reporting the 58 amount of nutrients in composite foods commonly consumed in Canada.(11) Composite 59 foods report a national sales weighted average of nutritional information derived by com-60 bining multiple products (e.g. multiple skim milk brands) into a single composite profile 61 (e.g. skim milk) that represents skim milk available in the Canadian food supply. The secondary objectives of this study were: 64 1. To describe and compare FS intakes among various sociodemographic groups 65 and by body mass index (BMI). 66 2. To identify the top food sources of FS for various age-sex groups. 67 Monosaccharides and disaccharides added to foods by the manufacturer, cook or consumer, plus sugars naturally present in honey, syrups, fruit juices and fruit juice concentrates (WHO definition) + our adaptation (the sugars from foods whose structure has been broken down), such as:  100% and concentrated fruit juice as stand-alone products and as an ingredient;  Vegetable juices, fruit and vegetable purées, pastes;  Extracts and artificial flavourings;  Alcoholic drinks, including barley malt.
This definition was adapted from WHO and Public Health England (PHE) definitions of free sugars.

World Health Organization (WHO) [2]:
"The term ''free sugars'' refers to all monosaccharides and disaccharides added to foods by the manufacturer, cook or consumer, plus sugars naturally present in honey, syrups, fruit juices and fruit juice concentrates".

Public Health England (PHE) [3]:
"This definition of free sugars for the UK excludes sugars naturally present in milk and dairy products, fresh and most types of processed fruit and vegetables and in cereal grains, nuts and seeds. It includes :  all added sugars in any form;  all sugars naturally present in fruit and vegetable juices, purées and pastes and similar products in which the structure has been broken down  all sugars in drinks (except for dairy-based drinks);  lactose and galactose added as ingredients." Guidance on the consumption of sugar differs around the world. In 2015, the World Health Organization (WHO) published its 'Sugar intake for adults and children' Guideline recommending a FS intake of less than 10% of total energy intake to reduce the risk of noncommunicable diseases [2]. The United States' Dietary Guidelines Advisory Committee recommends "Americans should keep their intake of added sugars to less than 10% of their total daily calories as part of a healthy diet" [7]. This recommendation for added sugars allows a larger amount of free sugar consumption as part of the daily intake, because it excludes 100% fruit juice. The United Kingdom's Scientific Advisory Committee on Nutrition recommends that "Free sugars should account for no more than 5% daily dietary energy intake" [4]. In line with the guidance based on a strong recommendation by the WHO, Canada's Dietary Guidelines support a FS intake of <10%E [10].
The main objective of this study was to estimate the FS intake in various Canadian agesex groups using data from 2015 Canadian Community Health Survey (CCHS)-Nutrition. A methodology that could be applied to estimate the free sugar contents of foods in the 2015 Canadian Nutrient File (CNF) was developed for this purpose. The 2015 CNF database is a standard reference food composition database reporting the amount of nutrients in composite foods commonly consumed in Canada [11]. Composite foods report a national sales weighted average of nutritional information derived by combining multiple products (e.g., multiple skim milk brands) into a single composite profile (e.g., skim milk) that represents skim milk available in the Canadian food supply.
The secondary objectives of this study were: 1.
To describe and compare FS intakes among various sociodemographic groups and by the body mass index (BMI).

2.
To identify the top food sources of FS for various age-sex groups.

Materials and Methods
Dietary intake data from the 2015 CCHS-Nutrition were used along with the food composition data from the 2015 CNF database.
The 2015 CCHS-Nutrition was a cross-sectional survey conducted by Statistics Canada, with data collected from January to December 2015 [12]. The target population is Canadian household residents aged 1 year and older living in one of Canada's 10 provinces.
Their sociodemographic, health-related information was collected through a questionnaire. Participants were also asked to report all foods and drinks consumed in the 24 h prior to the interview. The Automated Multiple Pass Method (AMPM) was used to collect the dietary intake data. The AMPM is a questionnaire that guides the interviewer to maximize the respondents' recall of all foods consumed in the previous 24 h [12] A subsample of approximately 30% of respondents was randomly selected for a second interview between 3 to 10 days after the first one. More information on the 2015 CCHS-Nutrition is available elsewhere [12].

Methodology for Use with CNF
We reviewed the existing methodologies and explored options to assess the FS in foods [13][14][15][16][17]. Most of these methods require ingredient lists of foods to estimate the FS. Consequently, they could not be used directly on foods in the CNF, because this database is comprised of composite foods and, therefore, does not contain this information. A 6-step methodology was thus developed to estimate the FS content of the 2015 CNF food composition database in the absence of ingredient lists (see Figure 2). Pre-step: Two registered dietitians assessed the presence of naturally occurring sugars, added sugars, free suga sugar substitutes of all foods in the CNF database prior to applying the 6-step process.
*Categories of foods (used in Step 3) where 100% of total sugars are considered free sugars  Cake mixes, Coffee Cake, Sponge Cakes  All types of Cookies Step 1: The food or drink contains 0 g total sugars.
Step 2: Are all sugars in the food or drink free sugars as per the adapted free sugars definition (see Fig.1)?
Step 3: Does the food belong to a food category* where 100% of total sugars are considered free sugars? [13,15] Step 4: The food or drink contains both naturally occurring and free sugars. Is there an unsweetened version available? [15] Step 5: Is the food or drink part of a food category with existing estimated free sugars content using another method?
Step 6: • If the product is a sweetened yogurt or milk OR a processed fruit item, then 50% of total sugars are free sugars.
• If the product contains milk or fruit, then 75% of declared total sugars are free sugars.  Pre-step: Two registered dietitians assessed the presence of naturally occurring sugars, added sugars, free sugars and sugar substitutes of all foods in the CNF database prior to applying the 6-step process.
* Categories of foods (used in Step 3) where 100% of total sugars are considered free sugars [9,13] Consensus was reached for all decisions when discrepancies occurred. The FS estimates from the 6-step methodology were then applied to the food intake data from the 2015 CCHS-Nutrition. The contribution of FS and %E from FS for the Canadian population was estimated. Meal replacement drinks and powders were excluded from these estimates, as these products fall under therapeutic dietary regimens. The Supplementary Materials include a separate file with our FS estimates of CNF foods.

Subjects
In the 2015 CCHS-Nutrition, 19,218 Canadians aged ≥ 1 year reported intakes of food. Participants with missing values of FS for at least one food (n = 1), those who were pregnant (n = 114), or breastfeeding (n = 185) were excluded from all the analyses. Height and weight were only measured for a subsample of participants aged ≥ 2 years. After previous exclusions, 5791 participants were identified with missing BMI due to missing height, weight or both measures and were excluded from BMI-related analyses.

Statistical Analysis
Statistical analyses were performed using the shared file through Statistics Canada using SAS EG version 7.1 (SAS Institute Inc, Cary, USA) [18,19]. Means and percentages were estimated using the first 24-h recall only. The PROC SURVEYREG procedure was used to estimate means of FS and %E from FS, overall, by categories within sociodemographic characteristics and by the BMI. The Wald's F test was used to perform overall comparisons, while t-tests were used for comparisons between categories. Bonferroni corrections were applied to adjust for multiple comparisons. To meet the normality assumption of the models, a Box-Cox transformation of the outcome variable was implemented (i.e., FS, energy from FS). In order to estimate the percent of the population below 10%E from FS, respectively, the NCI method (amount-only model) was used [20]. The univariate approach was used in this study for consistency purpose with the rest of the analyses.
To identify the top sources of FS, the PROC SURVEYMEANS procedure, along with the RATIO statement, was used. To account for the complex sampling design and nonresponse, all estimates were weighted to be representative of the Canadian population.
In analyses involving BMI, special survey and bootstrap weights (also available through the survey) were used to account for missing height or weight responses. For the variance estimations, the Balanced Repeated Replication method was used in all analyses. Table 1 shows estimates of FS intakes as %E from FS and in grams by sociodemographic characteristics and by BMI. In the overall population (i.e., ≥1 year), the mean intake of FS was 12%E (or 56 g). Men reported significantly higher absolute average intake of FS (i.e., in grams) than women, 64 g to 48 g, respectively. However, their relative average intakes (i.e., %E from FS) show similar values, with 12%E in both groups (p-value = 0.53). Our data showed that, overall, approximately 55% of the total sugars intake came from FS. See Appendix A Table A1 for results of t-tests that show between group differences. The analysis by age groups shows that the highest average intake of FS was by adolescents. The 9-13 years and 14-18 years reported the highest relative average intakes (about 15%E from FS), followed by the 1-8 years (14%E) and adults ≥19 years (11%E). Similarly, when looking at the absolute intake, the highest consumers were adolescents aged 14-18 years (80 g), followed by children aged 9-13 years (75 g). Children 1-8 years reported FS intake of 54 g and adults ≥19 years reported 53 g.

Free Sugars Intake Overall by Sociodemographic Groups and BMI
In the overall population, 40% of Canadians ≥1 year reported <10%E from FS in line with the WHO recommendation (results not shown). In children 1-8 y, this fell to 22%. Children and adolescents aged 9-13 years and 14-18 years were the least likely to meet the recommendation with 14% and 18%, respectively, consuming <10% from the FS. However, these results should be interpreted with caution due to the large variability in the estimates. 45% of adults ≥19 years meet the <10%E from FS recommendation.
The analysis by BMI groups did not suggest that there were any differences in the relative intake (p-value = 0.023) or the absolute average intake (p-value = 0.17) of FS. The %E from FS for people living in both urban and rural areas was approximately 12%. Individuals living in urban centers reported higher average intakes of FS (62 g) than those living in rural areas (55 g). Although the p-value for intake in grams suggests that the differences between groups is significant (p-value = 0.02), the difference appears to be small (<1%). Individuals with a university degree or higher reported a lower mean intake in both absolute and relative intakes of FS (53 g, 11%E) compared to those without a university degree or diploma (58 g, 12%E) with p-values of 0.005 and <0.001, respectively. The analysis of the FS intake in grams by income groups showed no significant difference (between 54 g and 57 g in all groups), with a p-value = 0.58. When considering the %E from FS, there was a difference from the highest to the lowest income quartile (from 12.3% to 10.7%) with a p-value ≤ 0.001.
Overall, in all groups the mean estimated %E from FS exceeds the WHO recommendation of <10%E from FS.

Top Sources of Free Sugars by Age-Sex Groups in 2015
The top 10 sources of FS in the diet of Canadians accounted for 81% of the total intake of FS in the overall population. The top four sources were: (1) sugars, syrups, preserves, confectionary, desserts; (2) soft drinks; (3) baked products and (4) juice (without added sugars). These top four sources accounted for 60% of the total free sugars intake. Together, sugary drinks (including soft drinks, juice without added sugars, beverages, e.g., tea or coffee with added sugars, energy drinks and alcohol) accounted for 40% of the FS intake from all sources (See Figure 3).
Comparisons by sex within the same age groups show very similar results for the top sources of FS (results not shown). In all age groups, as well as in the overall population, there was little difference in the relative amount of FS consumed from the top 10 sources of FS between men and women. The main difference was that men ≥ 19 y reported 6% more FS from sugary drinks than women (17% and 11%, respectively). In addition, the top four sources are almost identical in all groups, although in a different order. Women consumed more FS from sources such as "sugars, syrups, preserves, confectionary, desserts' and "baked products" than men.
Nutrients 2021, 13, x FOR PEER REVIEW

Discussion
To our knowledge, this study is the first to estimate the FS consumption of Cana in all age-sex groups using data from 2015 CCHS-Nutrition combined with the Can food supply data.
The results from this study illustrate the FS intake in the Canadian populati age-sex, BMI and sociodemographic characteristics. Overall, our analyses of CCH trition (2015) show a mean intake of free sugars of 12%E in the Canadian populati y), which is higher than WHO recommendation. Usual intake distributions reveale only 40% of Canadians consume <10%E. In order to verify the impact of our adapte sugars definition, we also compared the mean intakes of FS when applying the WH inition vs. our adapted WHO definition. As expected, due to our comprehensive d tion that included all sources of FS, the absolute intake in grams was higher. How the relative intakes (%E) remained unchanged for the most part (See Appendix A A2 for a comparative data table).
Further analyses indicate slight differences in the FS consumption by age an Men appear to consume more FS in grams; however, their %E from FS is sim women. This can be explained by the fact that men have higher energy intakes. One key findings of our analyses, by age groups, is that adolescents (9-13 years and years) are the highest consumers of FS, with a mean of about 15%E from FS. Les 14% of 9-13 years and 18% 14-18 years consumed <10%E from FS. The proportion o dren (1-8 years) meeting the <10%E recommendation is also low (22%). One of the re for this high intake of FS is the advertising of highly processed foods and beverages, tend to be high in FS, to this age group, which are an important driver of childhood o [21].
Other Canadian-based studies showed similar results [14,15,22,23] A study b geron et al. 2019 in French-speaking adults ≥18 years in the province of Québec rep an average of 12%E consumption from FS, which is close to our results (11%E in a [14] Additionally, a study by Wang et al. published in 2020, using 2015 CCHS-Nu data as well, estimated an average consumption of 9.9%E from FS, for Canadian >19 years [23]. Veugelers et al., 2020 also found similar estimates (13.3%E) of FS for dians using 2015 CCHS Nutrition data [22].

Discussion
To our knowledge, this study is the first to estimate the FS consumption of Canadians in all age-sex groups using data from 2015 CCHS-Nutrition combined with the Canadian food supply data.
The results from this study illustrate the FS intake in the Canadian population by agesex, BMI and sociodemographic characteristics. Overall, our analyses of CCHS-Nutrition (2015) show a mean intake of free sugars of 12%E in the Canadian population (≥1 y), which is higher than WHO recommendation. Usual intake distributions revealed that only 40% of Canadians consume <10%E. In order to verify the impact of our adapted free sugars definition, we also compared the mean intakes of FS when applying the WHO definition vs. our adapted WHO definition. As expected, due to our comprehensive definition that included all sources of FS, the absolute intake in grams was higher. However, the relative intakes (%E) remained unchanged for the most part (See Appendix A Table A2 for  a comparative data table).
Further analyses indicate slight differences in the FS consumption by age and sex. Men appear to consume more FS in grams; however, their %E from FS is similar to women. This can be explained by the fact that men have higher energy intakes. One of the key findings of our analyses, by age groups, is that adolescents (9-13 years and 14-18 years) are the highest consumers of FS, with a mean of about 15%E from FS. Less than 14% of 9-13 years and 18% 14-18 years consumed <10%E from FS. The proportion of children (1-8 years) meeting the <10%E recommendation is also low (22%). One of the reasons for this high intake of FS is the advertising of highly processed foods and beverages, which tend to be high in FS, to this age group, which are an important driver of childhood obesity [21].
Other Canadian-based studies showed similar results [14,15,22,23] A study by Bergeron et al. 2019 in French-speaking adults ≥18 years in the province of Québec reported an average of 12%E consumption from FS, which is close to our results (11%E in adults) [14] Additionally, a study by Wang et al. published in 2020, using 2015 CCHS-Nutrition data as well, estimated an average consumption of 9.9%E from FS, for Canadian adults >19 years [23]. Veugelers et al., 2020 also found similar estimates (13.3%E) of FS for Canadians using 2015 CCHS Nutrition data [22].
In general, the approaches used to estimate FS intake in all these studies are similar but appear to have a few differences [14,15,22,23]. One of the factors influencing the estimates is what researchers included in their definition of FS. As well, since an ingredient list is not available in the CNF, professional judgement was required when going through the different steps to estimate FS. We also used the University of Toronto's FLIP database that has nutritional information of food and drink products in the Canadian food supply, rather than databases from other countries which other Canadian studies have used [9,14,22,23]. These methodological nuances could lead to different decisions when it comes to identifying food sources of FS and, thus, the impact population-level estimates. We also observed differences in statistical methodologies that could lead to slightly different estimates from ours. For example, in our study we used the univariate approach from the NCI methodology to estimate %E from FS (i.e., usual intake of ratios), while other researchers used the bivariate approach (i.e., ratio of usual intakes). In some cases, these two approaches could yield to different results. Additionally, our study applied certain exclusion criteria to the population, as noted in the methodology section, which differed from the exclusion criteria used by others. While there are differences in the definitions used for FS and variations in the methodologies used to estimate FS intake, these studies show similar results. The estimates based on our methodology and those of others demonstrate that, overall, Canadians consume FS above the WHO recommendation of <10%E.
Other interesting data that are useful to guide population health interventions are analyses from various age-sex and demographic groups. Overall, we found that the top for four food sources of free sugars are similar between the various age-sex groups except for adult men (≥19 years), who reported consuming more FS from sugary drinks compared to adult women.
Our results also showed that there were no important differences in the intake of FS (from both %E and g) between different BMI groups in Canadian population ≥1 year (<1% difference in %E). While these results may be unexpected, as one would expect that a high FS intake may lead to high BMI, they may be explained by under-reporting in higher BMI groups [24]. Furthermore, we found that there was no important difference between Canadians living in urban settings compared to rural settings in %E from FS (<1% difference in %E); however, the absolute amount showed a significant difference at 61.6 g vs. 54.5 g (p < 0.001). This was interesting given that Canadian living in urban centres reported consuming a mean caloric intake of 1836 calories (95% CI: 1809.1-1863.4) compared to rural respondents consuming a higher mean intake of 1970 calories (95% CI: 1913.2-2027.5). One potential factor explaining the higher absolute FS estimate in urban populations may be the greater availability of foods containing FS in urban settings, which may contribute to higher intakes of FS in urban populations compared to rural settings [25][26][27]. Our study is the first to estimate intakes of FS of various sociodemographic groups in Canada.
Canadians' FS intakes are similar to international estimates. A national populationlevel health survey from Australia found that Australians consumed about 11%E from FS (about 60 g) [28]. The Australian data showed that about 52% of total sugars intake came from FS [17]. Similar to Canada, a New Zealand population-level survey showed that men consumed more FS than women [29]. Population-level surveys from the United Kingdom found British children consumed between 12-15%E from FS compared with 14-16%E for Canadian children [16,30]. These comparisons show much work remains to be done globally to meet the WHO FS recommendation in order to help reduce the risk of non-communicable diseases.
The 2019 Canada's Food Guide (CFG) recommends limiting the consumption of highly processed products which are high in FS, sodium and saturated fats. Most foods in the top 10 sources of FS in the diet of Canadians aged ≥1 year are highly processed foods that do not meet CFG recommendations. Encouraging the consumption of foods recommended as part of the CFG and limiting the foods high in FS, sodium and saturated fats can contribute to improved dietary intakes of Canadians, while meeting the WHO's FS recommendations.

Limitations of the Study
There were a few limitations in this study. The use of the 2015 CNF to assess the FS intake of Canadians is challenging, since the composite database does not contain ingredient lists. Food intakes in CCHS were self-reported and, therefore, could be prone to recall bias or potential under-reporting because of social desirability. Comparison with 2004 CCHS reported data was not possible as FS estimation was not done, thus changes in consumption between 2004 and 2015 for FS in Canadians is unknown. As well, there is inherent subjectivity in this methodology even though knowledgeable Registered Dietitians applied extensive knowledge of nutrition and food composition at each of the six steps.

Conclusions
We developed a methodology to estimate FS intakes of Canadians and applied this methodology to CCHS-Nutrition (2015). We found Canadians consume about 12%E from FS; this varied between 11%E to 16%E by various age-sex groups. Adolescents aged 9-13 years and 14-18 years appear to consume the highest amount of their energy intake from FS. Similar amounts of FS intakes were observed between men and women; however, the main sources of FS differed. Only 40% of Canadians ≥ 1 year consume <10%E from FS in line with the WHO recommendation. The age groups of most concern are adolescents, since they show the lowest proportion with intakes below the <10%E from FS recommendation (less than 14% for 9-13 years and less than 18% for 14-18 years).
The main sources of FS in the diet of Canadians were "sugary drinks", "sugars, syrups, preserves, confectionary, desserts and "baked products". Altogether sugary drinks accounted for most of the FS intakes (40%E) among Canadians (≥1 year).
These results reinforce the need for initiatives to help Canadians, particularly children and adolescents, to reduce their FS intake. It supports Health Canada's promotional and educational efforts on the CFG for which children and youth are priority target audiences. Encouraging the consumption of foods recommended as part of the CFG and limiting the foods high in FS can contribute to the improved dietary intakes of Canadians while meeting the WHO FS recommendations.