A worldwide pandemic of obesity and metabolic diseases has raised concerns about the adverse effects of sugar intake above the recommended levels. There is epidemiological evidence available that links the consumption of added and/or free sugar and especially sugar-sweetened beverages (SSBs) with several undesirable health effects such as dental caries [1
], insulin resistance [6
], type 2 diabetes [7
], metabolic syndrome [7
], cardio-vascular disease [11
], fatty liver [15
], and dyslipidaemia [14
]. Additionally, several clinical dietary intervention studies have indicated that diets with commonly-consumed levels of added sugars impair glucose and lipid metabolism, and increase several other risk factors associated with metabolic diseases [19
]. However, some controversy remains as to whether these unfavorable health outcomes are a direct consequence of excessive consumption of free sugar, or rather a result of excessive energy intake [21
Several governmental and public health organizations have issued recommendations to reduce the intake of free sugars. The new WHO (World Health Organization) guidelines on sugar intake published in 2015 recommended a reduction in free sugar intake for children and adults to less than 10% of total energy intake (strong recommendation), while a further reduction of the intake of free sugar to a maximum of 5% of total energy intake was recommended conditionally for additional health benefits [23
]. In accordance with a recent European Society for Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) position paper, free sugar intake should be reduced to <5% energy intake in children and adolescents aged ≥2–18 years, and should be even lower in infants and toddlers <2 years [24
]. Similarly, The UK Scientific Advisory Committee on Nutrition (SACN) recommends a free sugar intake lower than 5% of total energy from two years old into adulthood [25
], while the American Heart Association (AHA) recommends children consume ≤25 g (100 kcal) of added sugar per day and to avoid added sugar completely in children <2 years of age [26
Free sugars are the prime target of potential nutrition interventions since they provide energy with little or no nutritional value. The term free sugar in this paper follows the updated WHO definition and refers to all monosaccharaides and disaccharides added to foods and beverages by the manufacturer, cook, or consumer, and sugars naturally present in honey, syrups, fruit juices, and fruit concentrates [23
]. This definition does not include naturally occurring or intrinsic sugars, which are stored within the cells of intact fruits, vegetables, or involve lactose in milk or unsweetened dairy products.
The review of European-based nation-wide representative studies on added and total sugar intake showed that sugars represent a significant proportion of the total daily energy intake. The reported consumption of added sugars was especially high among children, ranging from 11.0% to almost 16.8% of daily energy intake, while in adults it ranged between 7.3–11.4% [27
]. However, in recent years, the attitude of some consumers towards free/added sugar has changed. An increasing number of consumer groups are initiating a reduction of free sugar intake in their diets [28
]; however, doing so might prove difficult due to the lack of accessible information on the amount of added or free sugar in foods and beverages on the market [29
]. Major progress in the EU (European Union) has been made with the new requirements for mandatory nutrition labelling of pre-packaged foods, which must provide the information about the total sugar content, as instructed by the Regulation 1169/2011 in the Provision of Food Information to Consumers [30
]. This rule was enacted in December 2016; however, prior to the rule, nutrition declarations had been voluntary, except on foods carrying nutrition or health claims. This regulatory change has been particularly important for member states with the lowest penetration of nutrition declaration on food labels such as Slovenia [31
]. While the nutrition declaration now provides information on (total) sugar content, the EU regulation does not require labelling of added or free sugar content. One of the major challenges encountered with potential free sugar labelling is its chemical indistinguishability from naturally-occurring sugar; therefore, there is no simple and accurate analytical method available that would discriminate between naturally-occurring sugar and added or free sugar. Nevertheless, the US Food and Drug Administration (FDA) has updated the nutrition facts label requirements where all nutritional labels on food packages will have to provide information on the amount of added sugar by 2019 [33
In Slovenia, high sugar consumption and obesity prevalence are particularly concerning. According to the last available data, the mean intake of free sugar in Slovenian adolescents (aged 15–16) was 130 g/day in boys and 110 g/day in girls (16% and 17% of the daily energy intake, respectively) [34
]. Results of a cross-national HBSC study (Health Behavior in School-aged Children), published by WHO in 2017, showed a high prevalence of obesity among Slovenian adolescents aged 11, 13, and 15 years. In 2014, 3.1% of girls and 7.8% of boys were obese, ranking Slovenia 4th among the 31 European countries included in the survey. The data were especially worrisome for 11 years old boys, where the prevalence of obesity was as high as 10.4% [36
]. To tackle the problem of obesity on a national level, the Slovenian Ministry of Health has issued a strategic plan to gradually diminish the percentage of overweight and obese individuals. One of the main goals of The Slovenian Resolution on Nutrition and Physical Activity for Health (SRNPAH) 2015–2025 is to reduce the intake of free sugars by 15%, with an emphasis on soft drinks, sweets, and desserts [37
To enable more accurate studies in monitoring the intake of total and free sugar in the Slovenian population, this study aimed to establish a comprehensive database of estimated free sugar content in the Slovenian food supply, focusing on pre-packaged foods. Using a step-by-step approach adapted from Bernstein and colleagues [38
], free sugar contents were calculated by combining the data from the nutrition declarations and ingredient lists of foods and beverages in the food supply. Such data, currently unavailable, would serve as a reference point for future studies, as well as for monitoring the efficiency of dietary recommendations and public health interventions.
2. Material and Methods
2.1. Data Collection
As a part of our previously conducted study described in Reference [39
], nutritional composition data were collected between January and February 2015 from the supermarket stores of the three major grocery chains with the largest nation-wide network of shops (Spar, Mercator, and Hofer). Sampling was performed in Ljubljana, Slovenia. In agreement with the retailers, all pre-packaged products with a unique European/International Article Number (EAN) barcode were systematically photographed and recorded in an online Composition and Labelling Information System (CLAS) database [40
]. Such database was first compiled in 2011 and has been described elsewhere in References [32
]. The CLAS Content Management System is supported by a specially developed mobile application, which enables the photographing of foods, data transfer to the CLAS database, and digital recognition of EAN codes to speed up the database formation and to avoid duplicate entries. During data collection in 2015, information on a total of 10,674 unique items were collected, including the product name, company, brand, list of ingredients, nutritional values, packaging volume, price, and EAN barcode. Flour, spices, sugar, food supplements, as well as all alcoholic beverages were excluded from the data acquisition.
All data entered into the CLAS database was rechecked and each product assigned to one of the 49 predefined food categories using a classification system developed by Dunford and colleagues [42
], which was developed as part of the Global Food Monitoring Initiative. The categories used in this paper were as follows: baby foods; biscuits, bread; breakfast cereals; butter and margarine; cakes, muffins and pastry; canned fish and seafood; cereal bars; cheese; chewing gum; chilled fish; chocolate and sweets; coffee and tea; cooking oil; cordials; couscous; cream; crisps and snacks; desserts; eggs; electrolyte drinks; frozen fish, fruits; fruit and vegetable juices; honey and syrups; ice cream and edible ices; jam and spreads; jelly; maize (corn); mayonnaise/dressings; meal replacements; meat alternatives; milk; noodles; nuts and seeds; other; other salt; pasta; pizza; pre-prepared salad and sandwiches; processed meat and derivatives; ready males; rice; sauces; soft drinks; soups; spreads; unprocessed cereals; vegetables; waters; and yogurt products.
2.2. Identifying Free Sugars within the Ingredient Lists
Following the WHO definition of free sugars [23
], ingredient lists of all products were systematically inspected for free sugar ingredients (FSI). All identified FSI and their varieties are listed in Supplementary Materials Table S1
. Despite being used as food sweeteners, artificial sweeteners and sugar alcohols are neither monosaccharaides nor disaccharides and were therefore not considered as free sugars.
2.3. Assesment of Total and Free Sugar Content
Nowadays, nutrition declarations on pre-packaged products contain information on the (total) sugar content across the entire EU. However, labelling of the sugar content on pre-packaged foods sold across the EU market only became mandatory in 2017, despite the information having been voluntarily presented on most pre-packaged foods by then. The CLAS database was compiled using all available information from nutritional declarations on food labels. We also collected information provided in the lists of ingredients, which included all the ingredients in a certain product, in descending order of weight. By law, the indication of the quantity of an ingredient is required if the ingredient (a) appears in the name of the food; (b) is emphasized on the labelling in words, pictures or graphics; or (c) is essential to characterize a food and to distinguish it from products with which it might be confused because of its name or appearance. Considering that nutrition declarations do not provide information on the amount of free sugar content, such information can only be calculated or estimated based on the ingredient list and total sugar content of each product.
The free-sugar decision algorithm was established based on the previously published algorithm developed by Bernstein and colleagues for the Canadian food market analyses [38
] (Table 1
). In Steps 1 and 2, all products that contained 0 g of total sugar (e.g., cooking oils, eggs, plain rice) or contained no ingredients that would contribute to free sugar content were assigned 0 g of free sugar. In Step 3, all products that contained no ingredients with more than 0.5 g of naturally-occurring sugar per 100 g had all their total sugar assigned as free sugar. Among the remaining food products, those that had a complete ingredient list with percentages of FSI and/or ingredients with naturally-occurring sugars available, were assigned under Step 4. The equation used for calculating free sugar content is presented in Table 1
. While Steps 1–4 were considered to be objective, the decisions for Steps 5–7 had to be made subjectively. Thus, the confidence of the calculated/estimated free sugar value decreased with each step. Products that could not be assigned to Steps 1–4, but had an unsweetened comparison available within the same category (e.g., a fruit yogurt compared with a plain yogurt), had a free sugar value estimated based on the naturally-occurring sugar content of the unsweetened product (Step 5). When the unsweetened comparison was non-existent, two options remained. If a comparable product that had already been assigned a free sugar value was available, a value reflective of the proportion of free sugar in the comparison product was assigned to the remaining product (Step 6). If an estimation of free sugar content was impossible from Steps 1–6, the free sugar values were estimated based on the order of ingredients on the ingredient lists (Step 7). All calculations in the paper were based on food composition data available from the Slovenian food composition database [43
For certain food products such as condensed soups, the amount of total and free sugar was calculated for the “as consumed” form of the product to enable comparison with other food products within the same category. The calculations were based on the preparation instructions written on the product, while both the dilution factor and additional ingredients were considered.
While the proportions of products containing free-sugar ingredients were calculated with the full sample of eligible items (Figure 1
= 10.674), estimations of the contents of free sugar were calculated only for products labelled with both an ingredient list and nutrition declaration (Table 1
and Table 2
, Figure 2
, Figure 3
and Figure 4
2.4. Share in Free Sugar Sales
The CLAS database was further complemented with country-wide 12 months sales data. Ensuring proper data handling, we obtained sales data from retailers covering the majority of the national market. The sales data referred to the national market and presented sales of food products for the 12 months period before the data collection (January 2014–December 2014). This was arranged on the condition that the results would not reveal sales data of any particular retailer. In this way we were able to access sales data for 8620 (91.7%) out of 10,674 products in the CLAS database, which were then included into analyses as reported in Figure 4
and Figure 5
. The sales data were given in universal form, including EAN number, description of the product, the number of products sold per year, and the quantity of food (kg or L) per packaging. Matching foods between the databases was performed using EAN numbers. Using sales data from 2014 and the previously calculated free sugar content of each product, we evaluated the relative contribution of each food category. The final values representing shares in free sugar sales are presented as ratios between the total free sugar in all sold items in certain (sub) categories and the total free sugar in all sold items in the sample.
2.5. Sale-Weighted Total and Free Sugar Content
With the use of sales data acquired from the retailers and previously calculated free sugar content, we could analyze the correlation of total/free sugar content with sales data. Analyses were conducted using a sample of foods for which both the total/free sugar and sales data were available (N
= 6563). Sales-weighted total/free sugar contents were calculated on the food category level (in g per 100 g or mL; results provided in the Supplementary Materials Table S2
). Based on the priorities of the SRNPAH 2015–2025, five different food categories were chosen for in-depth analysis: (A) yogurts; (B) biscuits; (C) breakfast cereals; (D) cakes, muffins and pastry; and (E) soft drinks. All items within each selected food category were assigned to an appropriate group that represented a certain interval in the free sugar range. The intervals for each category were determined based on the free sugar range of all items within the category. Sales factors were calculated from the total number of products sold, and corrected by the number of items per interval. The final factor was given as the proportion of the average sales volume of the interval.
2.6. Statistical Analysis
The data were processed and evaluated using the computer programs Microsoft SQL Server Management Studio V13.0, Microsoft Analysis Services Client Tools 13.0, Microsoft Data Access Components (MDAC) 10.0, Microsoft Excel 2016 (Redmond, WA, USA), and the program tool CLAS V1.0 (Composition and Labelling Information System; Nutrition Institute, Ljubljana, Slovenia). For the purpose of statistical analyses, products’ nutritional values from the CLAS database were exported in the form of Microsoft Excel spreadsheets. Mean values, standard deviation (SD), and quartiles (min, 25th, 50th, 75th, max) were calculated for the total and free sugar content. Sales-weighted total/free sugar contents were given as an exact value and, therefore SDs are not provided. The proportion of free sugar-containing items was presented as a percentage (%) of all products within each category. The average amount of free sugar in an individual food category was expressed as a percentage (%) of the averaged total sugar content of the group. The impact of free sugar content on sales was analyzed using the analysis of variance.
In concordance with recommendations on free sugar consumption postulated by different national and international health organizations, researchers have attempted to estimate the added/free sugar content of pre-packaged foods and beverages in several countries [38
]. The precise estimation of free sugar content in the available food supply is a necessary first step towards more informative studies focusing on per-capita free sugar consumption and follow-up strategies. This study was carried out as a starting point for future projects investigating free and total sugar intakes as well as potential sugar-content fluctuations in the food supply. While this cross-sectional study did not cover the entire national food market, a systematically inventoried subset of all foods and beverages found in the three largest grocery chains accounted for the majority of the Slovenian food supply of pre-packaged foods and non-alcoholic beverages. Our data showed that more than 50% of all 10,563 investigated products on the market contained one or more free sugar ingredients. This was somewhat lower when compared to the percentage of free sugar ingredients used in the US (74%) and Canada (65.4%) [38
]. We identified 65 different types of free sugar ingredients listed on the ingredient lists, which, as already pointed out by Bernstein and colleagues [38
], presents a great challenge for consumers trying to avoid excessive sugar consumption. Moreover, in an average sugar-containing product, 57.5% of all the sugar can be attributed to free sugar. The median free sugar content of products containing free sugar ingredients was 4.5 g per 100 g/100 mL.
Beyond the calculations of free sugar content, we also calculated the relative contribution of each food category to the total amount of free sugar sold. Results showed that chocolate, sweets and soft drinks accounted for nearly 60% of all free sugar sold in the three main grocery chains, possibly representing major culprits for excessive free sugar intake in Slovenia.
Using the sales data, we also investigated a possible correlation between free sugar content and the sales data provided by the retailers. Focusing on five food categories, we observed rather specific consumer preference for each food category regarding free sugar content. Brands of breakfast cereals with the highest amounts of free sugar were, for example, a preferred choice when it came to Slovenian consumers. However, it should be noted that this observation cannot be attributed to the effect of free sugar content alone and does not necessarily point to a causation, as the consumers’ purchasing decisions are driven by a variety of very different factors including brand, price, and packaging. While the exact effect of free sugar content on sales needs to be further investigated, the reported results in some food categories, such as breakfast cereals and biscuits, remain rather concerning.
The data obtained in this study will serve as a foundation for calculations of free sugar consumption as part of an ongoing national dietary intake survey. The outcomes of this survey will provide the first reliable data on free sugar consumption amongst different age groups and genders in Slovenia. These results will be needed when preparing further interventions aiming at lower sugar consumption.
A major strength of this study is the extent of the data collection and employment of yearly product-specific sales data, which have provided a very interesting insight into free sugar content in market-leading brands. The compiled database will be useful for various purposes and will support efficient monitoring of the changes in the food supply, and more reliable dietary assessments in the population. One major limitation of our study was the use of sugar content data as reported on food labels (rather than chemical analyses). However, there is no simple accurate analytical method that would enable measurements of free sugar content. Even measurements of total sugar would not be feasible on such a large scale. Therefore, we had to rely on the information provided by food manufacturers written on food labels, which is, however, a subject of regular inspection by the food authorities. In relation to the source of the data, another limitation of the study was that we were not able to estimate the content of free sugar in all of the products. Some foods with free sugar (based on information provided in the mandatory list of ingredients) were not labelled with nutritional declarations and were therefore excluded from our analyses. However, less than 12% of the foods (N = 1269) in the total sample were excluded due to missing data on sugar levels. It should be also noted, that results on sugar content are provided in grams per 100 g, and not per portion size, because portion size measures are not standardized in Slovenia, nor in European Union. Additionally, such a presentation allows for an easier comparison of the data collected in different countries.