Added sugars and nutrient density in the diet of Danish children

The relation between added sugars and density of dietary fibre and micronutrients was examined by analysing diets from 983 Danish girls and boys aged 4-14 years. The average intake of added sugars was 13.7 E% ranging from 3.4 to 38 E%. One fourth of the diets were below the recommended maximum level of 10 E%. No differences between sex and age groups were found. The diets were ranked and divided into quintiles (fifths) according to percentage of energy from added sugars. A strongly significant decline in nutrient density for all nutrients, except vitamin C, was observed as sugar concentration rose across quintiles. Nutrient densities expressed relative to recommended values varied from 30% to 300% illustrating that the nutritional significance of the dilution effect of added sugars differs from one nutrient to another. The results support the concept of "empty calories" and restriction of added sugars should still be recommended. Key words: Added sugars, children, empty calories, micronutrients, nutrient density


Introduction
In the Nordic Nutrition Recommendations (NNR) it is recommended to limit the intake of refined sugars to achieve a sufficient intake of dietary fibre. For adults with a low energy intake as well as for children this limit is set to a maximum of 10% of the energy intake (1). This is a pragmatic level which will ensure an adequate supply of essential nutrients, reduce the risk of caries, and still allow space in the diet for the sweet taste. Refined sugars include sucrose, fructose, starch hydrolysates etc. as food components or added in food manufacturing (I). A popular phrase for these pure carbohydrates is "empty calories", because they contain no micronutrients and thus dilute the content of vitamins and minerals in the diet. The aim of this study was to examine the influence of refined or added sugars on the quality of children's diet with respect to the dilution hypothesis and the concept of "empty calories".

Methods Subjects
Data were drawn from the national dietary survey ''Dietary Habits in Denmark 1995" (2). In this survey the participants (or their parents) kept a 7-day food record with preprinted response categories supplemented with open-ended alternatives. The quantity of foods eaten was estimated from household measures. The diets of 983 children (483 boys and 500 girls) aged 4-14 years were examined. In the survey response rate was 82% in this age group. The age group 4-14 years was chosen because the recommendations concerning refined sugars are primarily directed towards children, and the level of consumption of added sugars in Denmark is approximately 50% higher in children compared to adults (2).

Definitions
The term "added sugars" refers to refined or industrially produced sugars (usually sucrose) used as ingredient in processed foods or added at home in the kitchen or at the table. Added sugars do not include naturally occurring mono-and disaccharides in fruits, juices and milk. Most other studies on sugars and nutrient intake describe effects of total sugars or non-milk extrinsic sugars (NMES

Calculations
Nutrient density expressed as amount of dietary fibre, vitamins and minerals per 10 MJ was calculated for each individual diet. The argument for expressing diet composition instead of absolute intakes is the wide variation in energy intake and expenditure between age and sex groups. Furthermore, the percentage of energy from added sugars was calculated. The analyses are based on ranking of diets in quintiles (fifths) according to energy percent from added sugars. In calculating nutrient contents corrections for losses in cooking are done (2). Nutrient data including added sugars are taken from the National Food Data Base (4). Results concerning differences related to age are grouped according to age groups in N N R (1).
Relative nutrient density is expressed in percent of recommended nutrient density, to be used for planning of diets to a heterogeneous group of boys and girls aged 4-14 years. The recommended densities for each nutrient are calculated according to NNR (1) and based on the age and sex category for which the highest nutrient density is required.  Table 2. Intake of energy and energy percent from added sugars in 983 Figure 1. Intake distribution of added sugars (E%); diets boys and girls aged 4-14 years; mean (standard error). Differences in from 983 boys and girls aged 4-14 years. added sugars are not significant. Differences in energy intake are significant in all age and sex groups (pe0.001).

Statistical analyses
Differences between the quintiles are assessed by one-way analysis of variance and differences between age and sex groups by unpaired t-test.

Results
More than half of the added sugars in the average diet derives from sweets, cakes, ice cream and table sugar, and more than one third from sweetened beverages equally divided between sweetened fruit juices and carbonated soft drinks (Table 1). The remaining minor part of the sugars comes from cereals, fruit and milk products.
Energy intake rises with increasing age, and boys have a significantly higher energy intake than girls in all age groups (Table 2). No differences are found in percentage of energy from added sugars between sex and age groups.
The level of added sugars is close to 14 E%. The intake distribution is shown in Figure 1. Approximately one fourth of the children's diets are below the recommended maximum of 10 E%. Since no differences in E% are found between sex and age groups, the diets of all children are analysed together in order to improve statistical power. Table 3 shows the major statistics of percent energy derived from sugar (E%) after ranking and dividing into quintiles. The E% distribution of the lower and upper quintiles are skewed to the left and to the right, respectively, while E% in each of the three quintiles in the middle are uniformly distributed.
The energy intake of each age group divided into quintiles of energy percent from added sugars is shown in Table 4. Mean energy intakes when all children are Scand J Nutr/N&ingsforskning 1/99 Percent of energy from added sugars grouped together tend to rise slightly as intake of added sugars increases, even though differences between quintiles are not significant. Within each age group no systematic changes across quintiles are found, and only the older age group shows significant differences.
The nutrient density across quintiles of energy percent from added sugars is shown in Table 5 for dietary fibre and 10 vitamins and 9 minerals. Figures 2-4 are typical representatives of the general picture. Except for vitamin C the nutrient density decreases with increasing content of added sugars in the diet. Relative nutrient densities expressed in percent of recommended values for planning of diets (1) are presented in Figure 5. The percentage of coverage is above or well above hundred for vitamin A, thiamin, riboflavin, niacin, vitamin B6, vitamin B,,, vitamin C and phosphorous in all quintiles. For folate, calcium, magnesium and zink full coverage is only reached in lower quintiles. Dietary fibre and the remaining nutrientsvitamin D, vitamin E, iron, iodine, selenium and potassiumdo not fulfil the recommended levelsnot even in the low sugar quintile.

Discussion
The results clearly show a general trend of declining nutrient density with increasing percentage of energy from added sugars. The negative association is much stronger than demonstrated in earlier studies (33-8). However, comparison is complicated by the fact that some of these studies deal only with total sugars (5,6) or NMES (3), which might explain why they show a weaker dilution effect. Another complicating factor is the common practice to estimate the effect of sugars on absolute intakes of nutrients rather than nutrient densities. Rugg-Gunn et al. (7) found a  ** p<O.Ol; quintiles that share a common letter (a,b,c) are not significantly different.  positive association between intake of added sugars and micronutrients reflecting that energy intake is the major predictor of nutrient intake. This is not the case in a study by Gibson (8) of the diet of British adults, which showed that as the percentage of energy from added sugars increased, then energy intake increased and intake of micronutrients decreased slightly. However, many nutrients showed a non-linear relationship with sugar energy with the highest nutrient intake among average sugar consumers. This finding could be explained by a general underreporting of food intake among consumers in the lowest quintile of sugar intake. In our study energy intake only increases slightly and insignificantly across quintiles of energy percentage from added sugars. So, differences in absolute intake of nutrients will follow the same pattern as nutrient density do, i.e. intakes of dietary fibre and micronutrients are negatively associated with intake of added sugars.
The on1,y exception from the marked negative association is vitamin C showing a non-significant (p=0,09) lower nutrient density in the upper quintiles. This finding reflects that some of the foods, rich in added sugars, are also sources of vitamin C. Sweetened fruit juice is most important in this respect.
Even if the results do support the concept of "empty calories", the nutritional significance of increasing added sugars in the diet differs strongly from one nutrient to another. The effect ranges from no significance at all to a higher probabi-lity of insufficient intake as sugar concentration increases. But since a surplus of one nutrient can not compensate a possible insufficient intake of other nutrients, recommendations for dietary changes must focus on those nutrients of borderline adequacy. It seems prudent to continue recommending moderation in intake of added sugars. In fact, one may conclude that further restriction of added sugars and foods supplying the majority of refined sugars could be an important means to increase concentrations of dietary fibre and marginal vitamins and minerals in the diet. This is certainly a must when energy intake is restricted, i.e. if energy expenditure is low or in slimming diets.