Environmental and nutritional assessment of young children’s diets in Norway: comparing the current diet with national dietary guidelines and the EAT-Lancet reference diet

Current dietary patterns have a negative impact on people’s health and on the natural environment, globally as well as locally [1‐3]. In the Nordic countries, unhealthy diets are a leading risk factor for poor health and contribute to negative environmental impacts domestically and abroad [3]. A widespread shift in dietary patterns towards healthy diets from sustainable food systems is considered necessary to prevent malnutrition and diet-related non-communicable diseases, as well as reducing environmental pressure from the food system [2, 4]. Moreover, evidence suggests that food preferences and dietary habits are already established in childhood [5‐8]. Introduction of healthy and sustainable dietary patterns early in life could therefore build the foundation for lifelong healthy dietary patterns with a low impact on the environment. Thus, it is important with knowledge about the environmental sustainability of current diets among young children and possible areas for improvements.

Healthy and sustainable dietary patterns are often described as predominantly plant-based diets [2, 9‐11]. To aid transformations of dietary patterns towards more sustainable healthy diets, the EAT-Lancet Commission on Healthy Diets from Sustainable Food Systems described a universal healthy and environmentally sustainable reference diet, including a diversity of plant-based foods and low amounts of animal-sourced foods, highly processed foods and added sugars [2]. The reference diet should be suitable for adults and children and enable local adaptations to dietary preferences and food cultures of different populations.

In Norway, the national food-based dietary guidelines (FBDG) describe a healthy eating pattern tailored to Norwegian food culture [12]. While the guidelines are primarily based on health concerns, a separate assessment found its focus on increasing vegetable intake and limiting red meat consumption aligned with an environmentally sustainable diet [13].

To date, most research on the environmental sustainability of diets has focused on dietary patterns of adult populations [14‐23] with a few exceptions [24‐27]. In a Dutch study, the greenhouse gas emissions and blue water use of current diets were assessed among children aged 1–8 years [26] and, in a study from Italy, the carbon and ecological (land use) footprints from diets of 8- to 10-year-old children were assessed [25]. Among adolescents in Sweden, Colombo et al. assessed the carbon footprint of current diets and various optimised dietary patterns [27]. These studies all found unique environmental impact patterns for children and adolescents compared with adults, including which food groups contributed most to the different impact categories, showing the value of investigating children’s diets separately [25‐27].

The environmental impacts and nutritional content were estimated for three dietary patterns for 2-year-old children in Norway: the current habitual diet, a scenario diet based on the Norwegian FBDG and a scenario diet based on the EAT-Lancet reference diet [2, 11]. The diets were adjusted to 5.3 MJ/day, which is the reference value for energy intake for children aged 2–5 years [28].

To the authors’ knowledge, this was the first study to assess the environmental impact across six impact categories from diets among 2-year-old children. We observed that in the current diet the intake of dairy products was the most important contributor to five out of six environmental impact categories, followed by red meat and grain consumption. Furthermore, the environmental impacts of the current diet were found to be higher than impacts from the modelled scenario diets based on the Norwegian FBDG and the EAT-Lancet reference diet. The highest reductions in environmental impacts were found when shifting from the current diet to the EAT-Lancet scenario diet. Last, we showed that nutritionally adequate scenario diets for young children could be constructed within the recommendations of the Norwegian FBDG and the intake ranges of the EAT-Lancet reference diet.

On average, the current diet of Norwegian 2-year-olds contributed with 2.1 kg CO₂-equiv., 0.5 g P-equiv., 2.3 g N-equiv. and 24.5 g SO₂-equiv. and used 0.4 m³ of water and 2.1 m² a crop-equiv. of land per day. Most other studies that have assessed the environmental impact of diets have focused on global warming potential. Among older children in Italy and the Netherlands, the daily global warming potential per person was similar and slightly higher than observed in the present study: 2.2 and 3.0 kg CO₂-equiv., respectively [25, 26]. Among adolescent and adult populations in Europe, the per person daily dietary global warming potential ranged between 3.4 and 6.0 kg CO₂-equiv. [15, 20, 23, 26, 27]. The lower observed environmental impacts among the 2-year-olds compared with older populations were partly due to a lower energy intake. When adjusting for energy intake, the average global warming potential for Norwegian 2-year-olds was 0.4 kg CO₂-equiv./MJ compared with 0.4 and 0.5 kg CO₂-equiv./MJ among children in Italy and the Netherlands, respectively [25, 26], and between 0.4 and 0.6 kg CO₂-equiv./MJ among adolescents and adults in Europe [15, 20, 23, 26].

The studies among older children in Italy and the Netherlands assessed land use and water use, respectively, in addition to global warming potential [25, 26]. Direct comparison between studies is however problematic due to common methodological differences between studies, such as the system boundary applied and LCA data source. The average water use among children in the Netherlands was found to be 0.1 m³ per day (compared to 0.4 m³ per day among Norwegian children) [26], while the method used to assess land use in the study from Italy was not comparable to the method used in the present study [25]. Further, the studies on environmental impacts of children’s diets in Italy and the Netherlands [25, 26] both included impacts from home preparation, which was excluded in the present study. However, when impacts from home preparation were included in the estimated environmental impacts from the current diet among 2-year-olds in Norway, the total impacts per impact category were less than 1% higher, indicating a very low contribution from home preparation to the total environmental impacts,

Similar to findings among other population groups, animal-sourced foods contributed the most to global warming potential of current diets among young children in Norway. However, intake of meat is often the most important contributor to global warming potential from adult diets [15, 20, 23, 26], while consumption of dairy products was the most important contributor from the diets among 2-year-olds in Norway. Dairy products contributed with 46% of global warming potential whereas meat (red meat and poultry combined) contributed with 16%. In European adult populations, the contributions to global warming potential from dairy products consumption were between 13% and 23% whereas the contributions from meat consumption were between 30% and 38% [15, 20, 23, 26]. Vellinga et al. similarly found that the contribution of dairy products to global warming potential was more prominent among young children (1–8 years) compared with older children and adults [26].

In the present study, the intake of dairy products was the most important contributor to the other impact categories as well, except marine eutrophication. For marine eutrophication, grain consumption was the largest contributor. The content of fruit and vegetables is commonly found to strongly contribute to the water use of diets [9, 26]. However, we found that, in the current diet among 2-year-olds in Norway, fruit and vegetables contributed with only 9% and 4%, respectively.

Changing from the current diet to a diet in line with the Norwegian FBDG reduced the overall environmental impact, whereas changing to a diet aligned with the EAT-Lancet reference diet reduced most of the environmental impacts even more. A notable dietary change from the current diet to the scenario diets was the inclusion of less dairy products. A high consumption of milk is a distinct characteristic of diets among young children compared with older children and adults [26, 28, 44‐48]. Among Norwegian 2-year-olds, the average daily intake of milk and yoghurt was 460 g, and dairy products (in total 489 g including milk, yoghurt, cheese and butter) contributed with 28% of the total daily energy intake. The scenario diets included much less milk and yoghurt, i.e. 250 g in the FBDG scenario diet and 126 and 252 g in the EAT-Lancet scenario diets, as well as lower energy contribution from total dairy, 12% and 5%, respectively. Due to this large reduction of milk and dairy products content, the scenario diets were effective in reducing the environmental impacts of the diet especially for global warming potential, freshwater eutrophication, terrestrial acidification and water use. However, marine eutrophication was increased in the EAT-Lancet scenario compared to the other diets. In the EAT-Lancet scenario diet, the three main sources of energy were grains, vegetables oils and legumes and nuts—food groups that showed relatively high impacts on marine eutrophication, especially the vegetable oils. Further, land use was almost similar for all three diets, as the reductions in land use from dairy products in the EAT-Lancet scenario diet was outweighed by the increase in land use from vegetable oils, legumes and nuts.

Other potential reductions in environmental impacts of the current diet that could have been explored were to choose lower-impact foods within a food group [10, 41, 49‐52]. For instance, there are large variations in the environmental impacts of meat, especially between red meat and poultry, but also between different types of red meat such as beef and pork [10, 41, 49]. Seafood is another food group in which the environmental impacts vary widely depending on the species, production and harvesting techniques, because in particular crustaceans and farmed fish have higher environmental impacts than many wild-caught fish species [10, 41, 49‐52], although some wild fish species are under the pressure of over-fishing [2, 51].

Young children have unique nutritional needs to support rapid growth and development and require higher nutritional density in their diets than adults [53]. Plant-based diets generally result in lower environmental impacts compared with omnivorous dietary patterns [9, 27] and have been linked to a more favourable intake of many micronutrients and fatty acids among children aged 1–3 years [54]. However, if not carefully constructed, diets without animal-sourced foods may increase the risk of nutritional deficiencies among young children [54‐56]. A study applying the EAT-Lancet reference diet to a Danish setting found that the dietary content of vitamin D, calcium, iron, iodine and zinc could be a concern among children [21]. In the present study we similarly observed challenges with reaching adequate levels of vitamin D, calcium, selenium and iodine in the EAT-Lancet scenario diet when strictly applying the target values of the EAT-Lancet reference diet. In the current diet, milk and other dairy products contributed with about 80%, 35% and 60% of the total calcium, selenium and iodine intake, respectively [29]. In addition, fish intake contributed with 23% and 17% of the selenium and iodine intake, respectively. The target values of the EAT-Lancet reference diet include a low amount of both dairy products and fish; however, when the amounts of these food items were increased within the recommended ranges to 252 and 28 g, respectively, the EAT-Lancet dietary scenario reached the RIs for calcium, selenium and iodine. The EAT-Lancet Commission did emphasise that some population groups, including young children, could benefit from higher consumption than the target values [2]. Our study demonstrates the importance of actively using the intake ranges proposed for the EAT-Lancet reference diet in addition to the target values, and not simply “downscale” according to energy intake from scenarios that have been created for adults. The FBDG scenario diet reached the RI for all nutrients (except vitamin D), including calcium, selenium and iodine.

The content of vitamin D was below the RI for children aged 2–5 years in all diets analysed in the present study. The challenge to reach adequate levels of vitamin D intake from the diet alone has been widely recognised, and Norway currently implements voluntary vitamin D fortification of low-fat milk, butter and margarine [57, 58]. Furthermore, the Norwegian health authorities recommend supplementation with vitamin D for infants from birth [59] and continued supplementation for children and adults with low intake of vitamin D-rich foods [57]. In the present study, intake of micronutrient supplements was not included in the analyses, whereas micronutrients fortified in foods were captured. The current diet included a mix of vitamin D-fortified and unfortified milk, according to the average food intake reported in the dietary survey. The scenario diets included vitamin D-fortified milk only. Inclusion of fortified plant-based beverages as a substitution for cow milk could potentially decrease the environmental impacts of the diets [10] while at the same time provide valuable micronutrients in diets with no or low cow milk intake. The scenarios in the present study did not include fortified plant-based beverages, and inclusion of such products in the EAT-Lancet scenario diet could have increased the content of e.g. calcium, selenium and iodine.

Additionally, we found that with the food composition of the scenario diets the dietary content of saturated fat was reduced and the dietary content of iron was increased compared with the current diet. This was perhaps due to a lower content of processed meat products, often high in saturated fat, and a higher content of lean meat, whole grain products and legumes contributing to the higher content of iron.

As dietary patterns seem to persist from early childhood into adolescence and adulthood [5‐8], a preference for environmentally sustainable foods at an early age can lead to dietary choices with lower environmental impact later in life as well. This highlights the importance of establishing healthy and sustainable dietary patterns early in life. It can be argued that the role of the environmental impact of the diet of the youngest children is less important than among adults due to their lower energy intake. However, the present study has demonstrated that a dietary shift for young children from current diets towards diets aligned with the Norwegian FBDG and the EAT-Lancet reference diet can potentially reduce certain environmental impacts by up to 45% without compromising the nutritional adequacy. Hence, even in diets with a total energy intake of around 5 MJ/day, dietary shifts can contribute to overall reduced environmental pressure.

It should be noted that dietary change towards more environmentally sustainable diets is only one of several measures to reduce the environmental pressure from the food system. Other important measures that have not been addressed in the present study include improved food production practices and food waste reduction [2].

A strength of the present study was the simultaneous consideration of food-based dietary guidelines, nutrient adequacy and environmental impacts. This provides a more comprehensive picture, highlighting challenges of meeting adequacy on all dimensions at the same time.

Another strength of the present study was the comprehensive, updated and context-specific LCA database incorporated into the national food and nutrition calculation system, which allowed for automatic and simultaneous estimates of nutrient content and environmental impact of food items and diets. Moreover, it is a strength that we included estimation of six different environmental impact categories, because both similarities and specificities of the different impact categories were revealed.

Another limitation was the uncertainty in the environmental impact data from different LCA studies which applied different methodologies estimating the impact category values. Due to lack of data, avoidable and unavoidable food loss at retail and at home has not been considered at the present stage of the database.

As with all dietary assessment methods, there are limitations with FFQs. Portion sizes and frequency of consumption may be difficult to recall. Moreover, specifications of foods are limited due to the closed food list in the FFQ. The FFQ used to assess the current diet in the present study was based on an FFQ that was validated against 4-days weighed food records [30]. The validation indicated both over- and underestimation in intake of certain food groups and nutrients in the FFQ, although most of the food groups and nutrients did not show significant differences between the two methods.

Transforming current diets among young children in Norway towards diets in line with the Norwegian FBDG or the EAT-Lancet reference diet could reduce the environmental impacts by up to around 50%. In the current diet, consumption of milk and other dairy products contributed the most to the environmental impacts, and the reduction in milk and dairy products content from the current diet to the scenario diets was driving the reduction in overall environmental impacts. Although young children may have specific nutritional needs, we found that nutritionally adequate diets could be constructed within the Norwegian FBDG as well as within the intake ranges of the EAT-Lancet reference diet.

If Norway or other countries wish to create healthy and environmentally sustainable FBDG for children in the future, the present study has exemplified some challenging topics such as the role of dairy in the diet. We have also shown that focusing on global warming potential as the only environmental impact factor may ignore stress on other environmental impact categories, such as marine eutrophication and water use.