Co-benefits of the EAT-Lancet diet for environmental protection in the framework of the Spanish dietary pattern

Co-benefits of the EAT-Lancet diet for environmental protection in the framework of the Spanish dietary pattern Cristina Cambeses-Franco ⁎, Gumersindo Feijoo, María Teresa Moreira, Sara González-García CRETUS, Department of Chemical Engineering, School of Engineering, Universidad de Santiago de Compostela, Rúa Lope Gómez de Marzoa s/n, 15782 Santiago de Compostela, (Spain) HIGHLIGHTS •The Spanish dietary pattern has beencompared with the EAT-Lancet diet from carbon and water perspectives. •Starch-based products and oils andfatsare the main contributors to carbon and water footprints in the EAT-Lancet diet. •Meat and dairy are responsible for the greatest carbon and water footprints in the Spanish dietary pattern. •Major reductions in meat and dairy are needed to transitioning to low carbon and water footprint eating habits in Spain. GRAPHICAL ABSTRACT ABSTRACTARTICLE INFO Editor: Deyi Hou The immediate need to build resilient food systems with lower greenhouse gas (GHG) emissions and protection of water resources is a global challenge. To address this, the EAT-Lancet Commission described the global reference diet with principles of nutritional quality and environmental sustainability. With this in mind, the carbon and water footprints of the current Spanish dietary pattern have been compared with the EAT-Lancet global dietary recommendations, taking into account deviations in food intake. To provide additional context, differences between the average Spanish dietary pattern and dietary guidelines applied in other countries in Europe (Italy, the Netherlands and the Mediterranean region) and America have also been analyzed and discussed from a sustainability approach. We found that the EAT-Lancet diet requires less water resources (3056 L·person −1 ·day −1 ) and lower level of GHG emissions (2.13 kgCO 2 eq·person −1 ·day −1 ) in comparison with the Spanish dietary pattern (3732 L·person −1 ·day −1 and 3.62 kgCO 2 eq·person −1 ·day −1 , respectively). Starch-based products and oils and fats were identified as largest contributors to both environmental indicators in the EAT-Lancet diet. On the other hand, meat and dairy were the environmental hotspots in the Spanish dietary pattern. Comparison with other food-based dietary patterns also raisesenvironmental concerns about the high meat consumption in Spain. Overall, this analysis suggests that reducing the consumption of beef meat and dairy to a level in line with the global environmental targets set by the EAT-Lancet Commission would ensure a shift in Spanish dietary habits towards more environmentally sustainable food consumption patterns. Keywords: Water footprint Carbon footprint Sustainable healthy diet Spanish eating habits Dietary recommendations 1. Introduction Unsustainable food systems are posing a global danger to the health of the planet (Willett et al., 2019). Primary food production is the main driver of biodiversity loss and puts constant pressure on land, soil and water (Benton et al., 2021). A drastic and sustainable transformation of conventional food systems is required to achieve the UN Sustainable Development Goals and the Paris Agreement (Funabashi, 2018). With these needs and challenges in mind, initiatives have been developed to lay the foundation for public food and nutrition at national and global levels. National foodbased dietary guidelines (FBDGs) are government-endorsed policy Science of the Total Environment 836 (2022) 155683 ⁎Corresponding author. E-mail address: [email protected] (C. Cambeses-Franco). http://dx.doi.org/10.1016/j.scitotenv.2022.155683 Received 4 February 2022; Received in revised form 22 April 2022; Accepted 30 April 2022 Available online 5 May 2022 0048-9697/© 2022 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/). Contents lists available at ScienceDirect Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv documents aimed at educating consumers in their respective countries about healthy diets and lifestyles (UNICEF, 2021;WHO, 1998). For instance, in the European context and in line with the FOOD 2030 policy framework (European Commission, 2020), FBDG actions have been promoted such as the “Linee guida per una sana alimentazione”in Italy (CREA, 2019) or the Dutch dietary guidelines in the Netherlands (Health Council, 2015). In the United States, the Dietary Guidelines for Americans, published jointly by the Health and Human Services (HHS) and the Department of Agriculture (USDA), are the cornerstone for federal nutrition programs (USDA and HHS, 2020). Globally, the EAT-Lancet Commission addresses the need to feed the world's growing population with a healthy diet while defining sustainable food systems that minimize damage to our planet in terms of greenhouse gas(GHG), phosphorus and nitrogen emissions,water scarcity, biodiversity loss and land use change (Springmann et al., 2018a, 2018b). The EAT- Lancet reference diet is characterized by a high intake of fruits, vegetables, whole grains, legumes, nuts, and unsaturatedoils, and a drastic reduction in consumption of red meat, sugar, and refined grains (Willett et al., 2019). Recent studies documented the consequences of adherence to this diet in relation tonutritional quality, health, environmental and economic criteria. The Cancer and Nutrition (EPIC)-Oxford study reported the association of this reference diet with a lower risk of ischaemic heart disease and diabetes (Knuppel et al., 2019)andHirvonen et al. (2020) compared its affordability in different economic contexts. From a nutritional perspective, Blackstone (2020) conducted a comparison between the EAT Lancet diet and the Dietary Guidelines for Americans and Sharma et al. (2020) identified deviations with the Indian diet in terms of caloric intake and dietary composition. From an environmental perspective, Kesse-Guyot et al. (2021) explored the link between the adherence to the EAT-Lancet diet among French adults and three associated environmental pressures (GHG emissions, cumulative energy demand, and land occupation), and Kovacs et al. (2021) modelled the carbon footprint (CF) of seven food-based dietary guidelines (Germany, India, Netherlands, Oman, Thailand, Uruguay and the United States) and compared them with the EAT-Lancet diet. However, little is known about the magnitude of divergence in the environmental impacts of current diets compared to the EAT Lancet recommendations. This question is particularly challenging in Spain, where two well-ranked omnivorous dietary recommendations under economic, health and environmental criteria (González-García et al., 2020): the Mediterranean diet (MD) and the Spanish strategy for nutrition, physical activity and prevention of obesity (NAOS) have a strong influence on dietary habits. The Mediterranean diet is a predominantly plant-based diet, which emphasizes plenty of fruits, vegetables, nuts and olive oil as a source of fat (Trichopoulou and Vasilopoulou, 2016). On the other hand, NAOS diet emerged as a result of a Spanish public health planner whose goal was to reverse the increasing obesity prevalence through the promotion of healthy food habits and physical activity (Neira and Onis, 2006). An open debate is how media and educational campaigns to promote these win-to-win diets have had an impact on an increasingly westernized Spanish consumption pattern in terms of carbon and water footprints. To our best knowledge, two papers have examined the effect of thedietary patterns of the different climatic zones and autonomous communities of Spain on CF and other nutritional and socio-economic criteria (Esteve-Llorens et al., 2021, 2020). Food habits associated with each climatic zone could be explained based on geo-climatic conditions, food culture and tradition, and economic levels. The assessment of the CF of food habits in northern Spain revealed higher CF values in Northern Spain in comparison with those obtained from Southern Spain due to the major consumption of animal-based products and more caloric dietary patterns in the North (Esteve-Llorens et al., 2021). Moreover, seven autonomous communities showed sustainable dietary patterns (Andalusia, Community of Madrid, Basque Country, Navarre, Catalonia, Balearic Islands and Canary Islands) (Esteve-Llorens et al., 2020). On the other hand, a study on the CF of the average Spanish dietary pattern in 2006 and 2016 has been performed without considering water consumption (Batlle-Bayer et al., 2019). Thus, the main objectives of this research were: i) to translate the EAT- Lancet recommendations to the Spanish socio-cultural scenario, ii) to compare the environmental indicators (CF and WF) of the EAT-Lancet and the average Spanish dietary patternand, iii) to discuss the environmental implications of the adoption of FBDGs from Spain (MD and NAOS) and other regions of the world (Netherlands, United States and Italy) by the Spanish population. From a public perspective, this may shed light on the changes that may be necessary to modify the current Spanish dietary pattern to a healthy and sustainable universal reference diet. 2. Materials and methods 2.1. Description of dietary scenarios 2.1.1. Average Spanish dietary pattern The Spanish average dietary pattern was based on the 2019 Household Consumption Survey of the Ministry of Agriculture, Fisheries and Food (MAPA, 2019). This survey is conducted every year and covers a representative sample of more than 12,000 households. The most recent data prior to COVID-19 were used, as conditions of confinement such as reduced social activity, could have driven changes in dietary preferences. The average consumption per capita and year for each food category was obtained from the database considering that 82% of food consumption takes place at home and 18% outside of the home (Ministerio de Agricultura Pesca y Alimentación, 2020). Therefore, the average Spanish dietary scenario was designed considering the information of this database regarding the intake frequencies per food category and the amount of the most consumed foodstuffs in g that constitutes each food group within Spanish population. The associated caloric intake to the Spanish dietary pattern was calculated multiplying the number of grams of each foodstuff by the amount of energy per gram of each one. The BEDCA food composition database was used to find the caloric content of each food product (BEDCA, 2017). Thus, the energy supply estimated considering in-home and out-of-home consumption was estimated on 2214 kcal·person −1 ·day −1 . The scenario includes a wide variety of 59 representative foods, grouped into 14 food categories as can be seen in Table 1 according to the classification established by González-García et al. (2020). 2.1.2. EAT-Lancet diet This scenario corresponds to the EAT-Lancet dietary recommendations defined by Willett et al. (2019) for an individual intake of 2500 kcal/day. The reference diet proposes flexible intake ranges (weight and caloric intake) for the eight defined food groups (wholegrains, tubers or starchy vegetables, vegetables, fruits, dairy, protein sources, added fats and added sugars) and could be adopted for different cultures. To provide a direct comparison with the Spanish average dietary pattern, the eight food groups defined by the EAT-Lancet recommendations were re-categorized according with the classification established by González-García et al. (2020). The EAT-Lancet dietary scenario has been adjusted to an intake of 2500 kcal/day considering the flexible intake ranges for each food category based on EAT-Lancet recommendations for a healthy reference diet. EAT-Lancet recommendations give advice regarding the macronutrient intake (g·day −1 ) and caloric intake (kcal·day −1 ) per food category. However, it does not pay attention to the amount of each foodstuff in g that constitutes each group. This gap in guidelines was used as a strategy to adapt the EAT-Lancet recommendations to the Spanish socio-cultural scenario. Therefore, the representative foodstuffs of each food category were ascertained from the most consumed food products within the Spanish population gathered in the Household Consumption Survey by the Spanish Ministry of Agriculture, Fisheries and Food (MAPA, 2019). Table 1 shows the average daily recommended intake per food group in the Spanish adaptation of the EAT-Lancet recommendations. C. Cambeses-Franco et al. Science of the Total Environment 836 (2022) 155683 2 2.2. Environmental impact assessment 2.2.1. Carbon footprint TheLife Cycle Assessment (LCA) approach was followed to allow a comprehensive assessment of the CF, the best-known indicator of climate change and the focus of many sustainable policies in the food sector. LCA is a standardized method for assessing the environmental aspects and potential impacts associated with a product (ISO, 2006). The steps in LCA are (1) goal and scope definition, (2) life cycle inventory analysis, (3) life cycle impact assessment and (4) interpretation of the results. 2.2.1.1. Goal and scope. As aforementioned, the goal of this study was to compare the environmental profile of two dietary patterns from a cradle-to- consumer perspective, the average Spanish dietary patterns and the EAT- Lancet diet. To do so, it is necessary to choose a functional unit. According to the ISO standards, the functional unit (FU) is a measure of the function of the system expressed in quantitative terms and it provides a reference to which all the inputs and outputs to the product system are calculated. The average daily reference dietary intake per person was chosen as functional unit. 2.2.1.2. System description. The description of the system under study implies the definition of the system boundaries. The two dietary patterns were analyzed with the scope limited to the cradle-to-consumer phases along the life cycle (Fig. 1). i) Agricultural or industrial production: This subsystem includes all agricultural and industrial activities involved in the transformation of raw agricultural products into food products ready for distribution. Table 1 Daily-recommended intake (g·person −1 ·day −1 ) per food group for EAT-Lancet diet and Spanish Dietary Pattern (SDP). Nuts Lancet SDP Olives - 12.6 Almonds 6.8 1.3 Walnut 18.2 3.4 Peanut 26.0 1.5 Pulses Lancet SDP Chickpeas 20.7 4.7 Beans 13.9 3.2 Lentils 14.5 3.3 Meat Lancet SDP Beef 9.5 27.8 Chicken 35 71.0 Pork 9 55.3 Processed - 13.5 Dairy Lancet SDP Milk 153.9 282.5 Yogurt 21.8 40.0 Cheese 17.3 31.8 Ice-Cream - 12.0 Soy foods Lancet SDP Flour 12.5 - Soybean oil 12.5 - Starch-based Lancet SDP Bread 242.3 107.4 Rice 29.9 13.3 Pasta 32.3 14.3 Potatoes 53.4 95.5 Fish Lancet SDP Hake 9.1 17.6 Mackerel 1.1 2.0 Salmon 4.5 8.7 Pilchard 4.5 8.7 Cod 3.2 6.2 Tuna 1.7 3.4 Prawns 6.5 12.5 Squids 4.8 9.2 Mussels 3.7 7.1 Eggs Lancet SDP Eggs 12.5 27.9 Fruits Lancet SDP Oranges 68.2 79.3 Mandarine 24.5 28.4 Banana 51.6 59.9 Apple 39.9 46.4 Pear 19.7 22.9 Melon 32.7 38.0 Watermelon 34.0 39.5 Ready meals Lancet SDP Soups and creams - 24.2 Pizza - 10.3 Preserved food - 12.3 Sauces - 11.2 Sweets Lancet SDP Honey 3.4 1.4 Sugar 26.6 10.7 Vegetables Lancet SDP Tomatoes 125.7 97.3 Onion 67.0 51.9 Peppers 45.5 35.3 Lettuce 37.4 29.0 Carrots 31.8 24.6 Courgette 37.4 28.9 Cucumber 19.2 14.8 Oils and fats EAT- Lancet SDP Palm oil 6.7 - Olive oil 39.9 39.0 Lard or tallow 5.0 - Pastry Lancet SDP Biscuits - 25.5 Cereal breakfast -7.6 Nougat - 3.7 Chocolate - 17.0 Snacks - 2.7 1 Agricultural or industrial production 3 Household distribution Wholesale and retail distribution 2 Cradle-to-consumer gate Food losses Food waste Food losses Food waste Food waste Fig. 1. System boundaries considered in the carbon footprint assessment. C. Cambeses-Franco et al. Science of the Total Environment 836 (2022) 155683 3 ii) Wholesale and retail distribution: This subsystem includes the delivery of the food products to the wholesale and retail buyers. The calculation of the CF corresponding to this subsystem was made considering the geographical origin of all foods included in the dietary scenarios. Food waste and losses in the wholesale and retail distribution subsystem were taken into account. iii) Household distribution: This subsystem encompasses the transportation from supermarkets to households. Household food waste was considered in this stage. Food preparation at home was not taken into account because both dietary scenarios were adapted to the Spain's food culture and therefore both use similar cooking techniques, avoiding complex cooking methods. Moreover, the exclusion of cooking from the system boundaries was also considered in other similar studies available in the literature for other guidelines in Europe and America, facilitating the comparison (Cambeses-Franco et al., 2021;González-García et al., 2020). 2.2.1.3. Inventory data. Life cycle inventories for each foodstuff were compiled through a systematic literature review. Priority criteria were: i) scientific peer-reviewed LCA articles, ii) Spain as the country of reference, iii) life cycle studies from a cradle-to-gate perspective (where no or low-quality studies using this LCA approach were available, additional stages have been discarded). A detailed summary of the GHG emissions associated with each food item was gathered in Supplementary Table SM1.1. Carbon footprint associated tothe transportation of a food product from its production site to households was calculated according to the methodology proposed by González-García et al. (2020). The Multidimensional Datacomex platform provided free access to the origin and volumes of imports and exports per food item and country (for details, see Supplementary Table SM1.2). For international logistics, transoceanic vessels (18.60 mg CO 2 eq·kg −1 ·km −1 ) and trucks (92.10 mg CO 2 eq·kg −1 ·km −1 ) were considered (González-García et al., 2020). For national logistics, transport by truck to wholesalers and retailers was assumed to be 400 km by Euro 5 diesel freight trucks (>32 tons) (Castañé and Antón, 2017). A diesel vehicle (106 g CO 2 eq·km −1 )(Batlle-Bayer et al., 2019) was chosen as a means of transport to travel from homes to the nearest supermarket with a shopping frequency of once a week (average distribution distance of 3.3 km) (González-García et al., 2020). To address the challenge of finding a balance between food waste and food policy and security (Chen et al., 2020), the proportions of food losses and food waste during wholesale and retail transport (Gustavsson et al., 2013) and factors of food waste at the household consumer level (Garcia-Herrero et al., 2018) were incorporated into the CF assessment. 2.2.2. Water footprint Similar to LCA, the Water Footprint Assessment (WFA) methodology is developed in four stages, namely goal and scope, WF accounting, WF sustainability assessment and WF response formulation (Muthu, 2020). The first two steps of WF assessment were addressed in this study and the three key water components (green, blue and grey) were tracked using the comprehensive study of the agricultural and livestock WFs as a reference dataset (Mekonnen and Hoekstra, 2011, 2012). As climatic and production conditions are the main factors influencing green and blue WF, to ensure the geographical basis of food products, the methodological approach proposed by González-García et al. (2020) was adopted. Therefore, the WF of the category j for the food item F (m 3 ·t −1 )nationally produced (NWF j,F )(m 3 ·t −1 ) was obtained directly from the reported values in the Mekonnen and Hoekstra dataset for Spain (Mekonnen and Hoekstra, 2011, 2012). On the other hand, the imported total WF j of a food item F (IWF j,F )(m 3 ·t −1 )wasestimatedbasedontheWFofeachcategory j and country c, considering a representative number of importing countries (n) (the ones that supply a percentage of the total volume of imports superior to 70%), and being R c,F , the recalculated importing ratio for a country c taking into account only the representative countries (see Eq. (1)) IWF j,F¼∑ n c¼1 WFj,c,FRc,F(1) The WF per food product F (m 3 ·person −1 ·day −1 ) was calculated based on Eq. (2). WFF¼∑ 3 j¼1 IWF j,FXI,FþNWF j,FXN,F  MF10 6(2) where X I,F and X N,F are the average ratios of food imported and nationally produced, respectively, and M F (g·person −1 ·day −1 ) is the food intake by dietary scenario of the food item F. Food losses and food waste along the distribution chain and at the household consumer level were considered as for the CF assessment. Finally, the WF per dietary scenario was estimated based on Eq. (3). WFdiet ¼∑ k F¼1 WFF(3) For aquaculture species, we estimated the WF related to commercial feeds (Pahlow et al., 2015) and for non-aquaculture, a WF of zero was assumed (Blas et al., 2019;Harris et al., 2017). 3. Results 3.1. Carbon footprint assessment The GHG emissions value attributed to the current Spanish dietary pattern was 3.62 kgCO 2 eq·person −1 ·day −1 . On the other hand, the CF of the EAT-Lancet recommendations was approximately 1.7 times lower (2.13 kgCO 2 eq·person −1 ·day −1 ). The agricultural or industrial production stage (MAPA, 2019) is responsible for the largest share of GHG emissions in both scenarios, covering 91% and 96% of the total CF in EAT Lancet and the current Spanish dietary pattern, respectively. As a proportion of overall GHG emissions, the remaining stages of the supply chain represented a smaller contribution in the two diets analyzed. While the EAT- Lancet diet accounted for 0.13 kgCO 2 eq·person −1 ·day −1 in wholesale and distribution, the GHG emissions derived from this stage in the current Spanish dietary pattern were 0.09 kgCO 2 eq·person −1 ·day −1 . The higher levels of peanut consumption per capita in the EAT-Lancet diet, together with the high distribution distances (truck and ship) for this food, could explain the differences in the CF results for the wholesale and the distribution between both dietary patterns. The CF associated with the household distribution phase was 0.06 kgCO 2 eq·person −1 ·day −1 regardless of the scenario. In terms of food waste, 0.13 kgCO 2 eq·person −1 ·day −1 and 0.21 kgCO 2 eq·person −1 ·day −1 were attributed to food losses and food waste for the EAT Lancet and the current Spanish dietary pattern, respectively. The highest CF relative to food waste forthe Spanish dietarypattern was associatedwith the mostsignificant contribution of dairy, meat, fish and their associated food waste and loss. The daily CF attributed to both dietary scenarios, broken down by food categories, is shown in Fig. 2. The CFs for dairy products, eggs, meat, fish and seafood for the average Spanish dietary pattern are higher than the CF values calculated for the macronutrient intake ranges recommended by EAT-Lancet for these food groups. Foods of animal origin were the main dietary contributors to this environmental metric, accounting for 47% in the EAT-Lancet recommendation and 72% in the current Spanish dietary pattern. Meat and dairy werethemainhotspotsintheSpanishdietarypattern(1.42and 0.70 kgCO 2 eq·person −1 ·day −1 , respectively). On the other hand, the CF of starch-based products were quantitatively the most important in the EAT-Lancet recommendations (0.45 kgCO 2 eq·person −1 ·day −1 ), C. Cambeses-Franco et al. Science of the Total Environment 836 (2022) 155683 4 although closely followed by meat (0.39 kgCO 2 eq·person −1 ·day −1 )and dairy (0.35 kgCO 2 eq·person −1 ·day −1 ). EAT-Lancet does not publish gram ranges for pastries and ready meals. In this diet, the CF scores related to total vegetables, pulses, soy foods, starch-based products, nuts, and oils and fats were higher than those of the Spanish pattern. The fact that the EAT-Lancet diet emphasizes the consumption of plant-based foods is behind the higher production-based GHG emissions. The CF for fruits and sweets was similar between the EAT-lancet pattern and the Spanish pattern. 3.2. Water footprint assessment Our analysis reported the average WF of the Spanish dietary pattern (3732 L·person −1 ·day −1 ) and that of the EAT-Lancet diet (3057 L·person −1 ·day −1 ). The dietary green WF represented by far the largest fraction in the total WF for both diets (74% and 79% for the EAT-Lancet diet and the Spanish dietary pattern, respectively). Data on the contribution of food categories to dietary WFs of each dietary pattern are depicted in Fig. 3 and Supplementary Table SM1.3. Fig. 2. Carbon footprint estimation (kgCO 2 eq·person -1 day -1 ) per food groups for EAT-Lancet diet and average Spanish dietary pattern. Fig. 3. Water footprint estimation (L·person -1 day -1 ) per food groups for EAT-Lancet diet and average Spanish dietary pattern. C. Cambeses-Franco et al. Science of the Total Environment 836 (2022) 155683 5 Dairy products, meat, and oils and fats were the main contributors to dietary WF inthe averageSpanish dietarypattern. However, large differences in the interpretation of WF scores for these three main contributing food groups have been explored for both diets. The WF scores for dairy and meat in the Spanish mean dietary pattern (834 and 1024 L·person −1 ·day −1 , respectively) were significantly higher than the EAT-Lancet recommendations (423 and 297 L·person −1 ·day −1 , respectively), while those of oils and fats were slightly lower (679 and 745 L·person −1 ·day −1 for Spanish dietary pattern and EAT-Lancet, respectively). These results could be explained on the basis of the daily intake amounts of these food groups in the definition of the dietary patterns. Plant-based products, such as starch-based products and nuts, were also presented as large contributors to the dietary WF of the EAT-Lancet recommendations (538 and 470 L·person −1 ·day −1 , respectively). However, they were among the lowest WF scores on the average Spanish dietary pattern (253 L·person −1 ·day −1 for starch-based products and 119 L·person −1 ·day −1 for nuts). The reason behind this finding is that the reference diet is largely plant-based. Other plant-based products with lower WF contributions, legumes and soy foods, also showed higher WF values in the EAT-Lancet reference diet (213 and 90 L·person −1 ·day −1 , respectively) compared to those of the Spanish average dietary pattern (49 and 0 L·person −1 ·day −1 , respectively). However, no significant differences were found for the fruits and vegetables categories between the EAT-Lancet recommendation (123 L·person −1 ·day −1 for fruits and 60 L·person −1 ·day −1 for vegetables) and the Spanish food consumption pattern (143 L·person −1 ·day −1 for fruits and46L·person −1 ·day −1 for vegetables). Although fruits and vegetables contributed in a different proportion to thedailyintakeintheEAT-Lancetdiet (271 and 364 g·person −1 ·day −1 for fruits and vegetables, respectively) compared to the Spanish dietary patterns (314 and 282 g·person −1 ·day −1 for fruits and vegetables, respectively), their associated water requirements are not so high to account for significative differences. Compared to the reference diet, the WF of other foods of animalorigin (eggs, fish and seafood) of the current average Spanish dietary pattern was higher. The associated WF of pastries and prepared foods was not considered in the EAT-Lancet reference diet. 4. Discussion of results 4.1. Comparison with literature The adoption of the EAT-Lancet recommendations in Spain led to changes in the CF of the Spanish dietary pattern. In this sense, foodrelated GHG emissions could be reduced by 41% by adopting the recommendations in the Spanish dietary pattern, most of which were driven by stricter limits on animal-based products (meat, fish and seafood, and dairy) and were offset by higher consumption of plant products (whole grains, vegetables, nuts, seeds and legumes). This is in agreement with the findings of the modelling study by Springmann et al. (2020), who reported large net reductions in GHG emissions of 42% by adopting the EAT-Lancet recommendations. Furthermore, if the CF score for the EAT-Lancet diet is compared with those estimated in the literature, some differences arise. Kesse-Guyot et al. (2021) observed through a prospective cohort study that participants with higher adherence to the EAT-Lancet diet (ranked by quintiles) were associated with lower GHG emissions. The mean CF for individuals with the highest level of adherence to the EAT-Lancet diet was 2.73 kg CO 2 eq·person −1 ·day −1 . Different research boundaries (their results were referred at the farm level), calculation methods and data source were behind the differences in the CF score between both studies. Kovacs et al. (2021) also estimated the GHG emissions associated with the reference EAT Lancet diet,whoseCFwason1.36kgCO 2 eq·person −1 ·day −1 . This score is considerably lower than that reported here (2.13 kgCO 2 eq·person −1 ·day −1 ). The fact that their assessment was scaled to a 2000-kcal diet and only accounted for GHG emissions from resource extraction to the factory or the farm gate (cradle-to-farm gate) could explain these considerable differences. Regarding the CF for the current Spanish dietary pattern, comparisons should be made with caution. Our result (1321 kg CO 2 eq·person −1 ·year −1 ) differs from the ones reported by Esteve-Llorens et al. (2020) (1024 kg CO 2 eq·person −1 ·year −1 )andBatlle-Bayer et al. (2019) (1120 kg CO 2 eq·person −1 ·year −1 ). Esteve-Llorens et al. (2020) and Batlle-Bayer et al. (2019) used the same database as in our study (MAPA, 2019), albeit taking 2018 and 2017, respectively, as reference years. On the other hand, even if Esteve-Llorens et al. (2020) used similar data sources, food consumption outside home was not accounted and the production phase was only considered within the system boundaries of analysis. On the other hand,Batlle-Bayer et al. (2019) chose a cradle-to-consumerapproach. Regardless of the study, beverages were included as a food category. Secondly, a switch to the EAT-Lancet diet may reduce the WF for the Spanish dietary habits as also reported by Vanham et al. (2021),who estimated a consumptive WF score (green plus blue WF) for the Spanish dietary pattern of 4238 L·person −1 ·day −1 , considerably high compared to the EAT-Lancet diet (2414 L·person −1 ·day −1 ). Significant differences were found with our results in terms of the consumptive WF for the EAT- Lancet (2660 L·person −1 ·day −1 ) and the Spanish dietary pattern (3375 L·person −1 ·day −1 ). The rationale behind these results is that Vanham et al. (2021) used the average annual FAO Food Balance Sheets to estimate food supply data for the period 2011–2013, and we used the 2019 Household Consumption Survey (MAPA, 2019) to obtain national food intake data. Moreover, we considered as reference the Multidimensional Datacomex platform regarding the corresponding volumes of imports and exports, and Vanham and colleagues considered proportions of domestic production and imports to the total domestic supply in the FAO Food Balance Sheets. 4.2. Comparison with other food-based dietary guidelines A comparative analysis of the current Spanish dietary pattern with other FBDGs in terms of dietary intake, CF and WF was performed, with the aim of contributing information to the literature in this area of research. In particular, the FBDGs selected were the Mediterranean Diet (MD), the Italian Dietary Guidelines (CREA, 2019), the Spanish Strategy for Nutrition, Physical and Obesity Prevention (NAOS) (Neira and Onis, 2006), the Dutch Dietary Guidelines (Health Council, 2015) and the American Dietary Guidelines (USDA and HHS, 2020). These guidelines are not the only ones that include environmental considerations; we chose a purposive sample to prioritize studies conducted with the same methodology, system boundaries and food categorization as ours. In addition, these diets cover a sample of countries representing a wide range of world's population from the northern and southern regions of Europe and America. CF and WF results for these FBDGs were taken from Cambeses-Franco et al. (2021) and González-García et al. (2020). For comparison, our results were adapted to a cradle-to-gate approach (paying attention only to the production stage). In addition, the designed FBDG scenarios do not include the pastry and prepared meals subgroup in the mass-based recommendations. 4.2.1. Dietary intake The Spanish dietary pattern differs from the recommended intake amounts for each food subgroup in the existing dietary guide. Compared to the Southern European FBDGs (MD, NAOS and Italian), the Spanish dietary pattern included, on average, more meat (range among the Southern European FBDGs 2.7 to 3.9 times higher) and fewer fruits (0.6 to 0.7 times lower), vegetables (0.40 to 0.80 times lower) and starch-based products (0.5 to 0.7 times lower) (see Fig. 4a). The MD contained lower amounts of dairy (270 g·person −1 ·day −1 ) in comparison with the Spanish dietary pattern (366 g·person −1 ·day −1 ). In contrast, NAOS and the Italian dietary guidelines do not suggest limiting dairy intake to a higher percentage than the Spanish dietary pattern. Compared to the Dutch dietary guidelines, the Spanish dietary pattern included less starch-based products (0.5 times) and dairy products (0.8 times), and more meat (2.4 times), fruits (1.6 times) and fish and seafood (5.4 times). On the other hand, the main difference between the American FBDGs and the Spanish dietary pattern was the consumption of C. Cambeses-Franco et al. Science of the Total Environment 836 (2022) 155683 6 dairy products.The Spanish population consume lower dairy products (366 g·person −1 ·day −1 ) than the recommended servings in the American Dietary Guidelines (712 g·person −1 ·day −1 ). The comparison of the Spanish dietary profile not only with the EAT- Lancet recommendations, but also with the rest of the dietary guidelines, showed the convenience of increasing the consumption of starch-based products among the Spanish population while reducing the consumption of meat, pastries and ready meals. On the other hand, dairy consumption remained too high, both in the current Spanish pattern and in the other FBDGs (except MD) compared to the global recommendations of the EAT- Lancet diet. 4.2.2. Carbon footprint The GHG emissions associated with the dietary recommendations analyzed ranged from 1.95 kg CO 2 eq·person −1 ·day −1 for the EAT-Lancet diet to 2.98 kg CO 2 eq·person −1 ·day −1 for the American dietary guidelines (considering only the production phase). Therefore, the adoption of some of these FBDGs would allow a reduction in CF related to average Spanish food intake in a range between 14% (adoption of the American dietary guidelines) and 44% (adoption of the EAT-Lancet diet). The CF profiles (estimated as the difference between the associated CF of the current Spanish dietary pattern and the corresponding one for the FBDG by food group) were very similar for all FBDGs as can be seen in Fig. 4b. Fig. 4. Differences betweenSpanish dietary pattern and global and national food based dietary guidelines: EAT-Lancet recommendations, Mediterranean Diet(MD), Spanish Strategy for Nutrition, Physical Activity and Obesity Prevention (NAOS), Italian Dietary Guidelines, Dutch Dietary Guidelines and American Dietary Guidelines. Positive values indicate greater intake (a), CF (b) and WF (c) in Spanish dietary pattern and negative ones indicate lower. C. Cambeses-Franco et al. Science of the Total Environment 836 (2022) 155683 7 The adoption of FBDGs (national or global) led to significant reductions in the CF for meat (ranging from 83% in the Italian Dietary Guidelines to 27% in the American dietary guidelines) and fish and seafood (ranging from 28% in the Dutch Dietary Guidelines to 6% in NAOS) subgroups. The reductions in GHG emissions from meat for the Italian Dietary Guidelines could be explained not only by their limitations on meat consumption (43 g·person −1 ·day −1 ) with a higher CF, but also because they prioritize white meat over red meat, as does the EAT-Lancet diet. On the other hand, the FBDG recommendations were associated with a higher intake of starch-basedproducts (ranging from 315 g·person −1 ·day −1 in MD to 465 g·person −1 ·day −1 in NAOS) in comparison with the Spanish dietary pattern (231 g·person −1 ·day −1 ). Therefore, adoption of the global EAT-Lancet recommendations or some of the other national or regional FBDGs would result in a moderate increase in CF for this food subgroup. Comparing the Spanish dietary pattern and the FBDGs in terms of dairy was not straightforward. Negative deviations around the CF for dairy using the Spanish dietary pattern as a reference could be seen for NAOS, the Italian, American and Dutch dietary guidelines and positive deviations for the EAT-Lancet and MD (see Fig. 4b). This result suggests that reforming the national FBDG with a moderate consumption of dairy products as outlined in the EAT-Lancet recommendations could be beneficial for the transition to more sustainable diets. In general, limiting meat consumption as recommended by all the FBDGs analyzed, and moderating dairy consumption in accordance with the global EAT-Lancet recommendations, would increase the environmental sustainability of the current Spanish dietary pattern. 4.2.3. Water footprint Taking into account that WF depends not only on recommended daily intakes, but also on geographical considerations, local agricultural production methods, and climatic conditions, comparison with the Spanish reference scenarios is the most accurate. We conclude that adherence to the EAT-Lancet, MD or NAOS diets by the Spanish population would result in reductions in water resource demand ranging from 10% in NAOS, which reported the highest WF (3181 L·person −1 ·day −1 )to20%inMD,withaWF of 2826 L·person −1 ·day −1 . As a first important observation, higher scores were shown for legumes (ranged from 79 L·person −1 ·day −1 in MD to 186 L·person −1 ·day −1 in EAT- Lancet), starch-based products (ranged from 362 L·person −1 ·day −1 in NAOS to 504 L·person −1 ·day −1 in EAT-Lancet), and nuts (ranged from 209 L·person −1 ·day −1 in MD to 419 L·person −1 ·day −1 in EAT-Lancet) in the Spanish reference dietary recommendations in comparison with their corresponding values in the Spanish dietary pattern (42, 237 and 104 L·person −1 ·day −1 , respectively). In contrast, meat showed positive deviations as can be seen in Fig. 4c. For a Spanish citizen, the reduction in WF associated with meat when adopting the Spanish reference dietary recommendations would range from 57%, if the choice is NAOS, to 71%, if EAT-Lancet diet were the one chosen. On the other hand, considering that EAT-Lancet and MD were associated with more ambitious reductions in dairy consumption than NAOS, the adoption of some of these dietary guidelines would allow reductions in dairy WF (26% for MD and 48% for NAOS), which would not happen if adherence were to the NAOS dietary pattern. Comparison with the other dietary guidelines not framed in Spain should be made with caution. The minimum water required was associated with the adoption of the Italian dietary guidelines (1761 L·person −1 ·day −1 ), with a lower WF associated with for oils and fats (250 L·person −1 ·day −1 ), meat (150 L·person −1 ·day −1 ) and dairy (470 L·person −1 ·day −1 ) due to their lower consumption compared to the Spanish dietary pattern. Moreover, the national average values of the total WF of olive oil (9103 m 3 ·t −1 ), milk (805 m 3 ·t −1 ) and beef (11,500 m 3 ·t −1 ) in Italy were lower than those in Spain (13,190, 1847, 13,608 m 3 ·t −1 , respectively) (Mekonnen and Hoekstra, 2011). Compared to the Dutch FBDGs, the Spanish dietary pattern showed, on average, a higher WF impact for meat (3.9 times higher) and dairy (2.3 times higher), and a lower impact for starch-based products (0.5 times lower) and oils and fats (0.8 times lower). Although dairy products contributed a high percentage to the daily intake in the Dutch FBDG (471 g·person −1 ·day −1 ) in comparison with the Spanish dietary pattern (366 g·person −1 ·day −1 ), a positive difference could be explained in the comparative WF analysis. The total WF of milk at the national level, considering the Netherlands instead of Spain as the reference country, explains the resulting deviation profile (Mekonnen and Hoekstra, 2011). The Spanish dietary pattern represented a higher WF related to meat and oils and fats compared to American FBDGs. The average meat consumption per capita (168 g·person −1 ·day −1 ) and the large volumes of water used for beef cattle (13,608 m 3 ·t −1 ) in Spain explain the deviations from the American FBDGs. Otherwise, soybean oil, with a relatively small WF per gram of fat, may be an olive oil substitute in the American guidelines, which explains the lower WF for this food category. 4.3. Key findings and their policy implications This work is based on the proposal that the EAT-Lancet dietary recommendations can help to enable sustainable food transitions in Spain and other European and American countries, reaching children, families and communities. In this sense, favouring plant-based over animal-source meals can strength the potential of dietary patterns to enable sustainable food systems. However,reducing emissions from meat and dairy toachieve EAT-Lancet dietary guidelines do not seem plausible without further progress in planning and policy-making. First, governments should include, in addition to health and nutritional issues, sustainability considerations into food-based dietary guidelines. The approach discussed in this paper could be a good starting point for formulating national food-based dietary guidelines updated to perform better in terms of environmental implications. Second, school canteens provide a great opportunity to improve education on food sustainability with a future impact on food systems. School meals can influence the broader community by providing a wider understanding of nutritional learning experiences to children and their families. Particularly, school gardens are a powerful environmental educational tool with potential cross-sectoral implications for environmental awareness, economic development, health and wellness across the local community. Third, the development of sustainable, nutritious and healthy meat analogue products that reduce reliance of livestock production and meet the EAT-Lancet and United Nations SDGs. However, plant-based protein meat analogues lead to poor nutritional intake, mainly due to amino acid, vitamins and n-3 PUFAs deficiency. Formulation of alternative hybrid meat analogues could be a promising solution for climate change mitigation and water scarcity, without compromising health and nutritional issues. Therefore, government should fund academic research on meat substitutes. 5. Conclusions The EAT-Lancet diet is a universal reference diet for people and planetary health. Based on the environmental analysis made, it is evident that current Spanish dietary pattern differs significantly from EAT-Lancet targets on adopting healthy diets produced in sustainable ways. As discussed in the paper, a shift towards EAT-Lancet recommendations by Spanish people will require a significant reduction in animal source foods (dairy products, meat, fish and seafood), pastry and ready-meals, and an increase in plant-based foods (vegetables, pulses, soy foods, starch-based products, nuts, and oils and fats). Particularly, beef meat and dairy reductions have a significant potential for transitioning to low carbon and low water footprint eating habits in Spain. When compared the Spanish dietary pattern to other European and American recommended diets, the largest divergences in terms of CF and WF were in meat category. Although compared with Spanish current intake, the FBDGs included, on average, more plant-based products, such as vegetables or starch-based products, their relative impact on CF and WF was not as significative as meat, since plant-based products are associated with lower environmental emissions per unit weight. C. Cambeses-Franco et al. Science of the Total Environment 836 (2022) 155683 8 Understanding the limitations of the study is important to place research findings in context and interpret the validity of the results. The wide diversity of LCA studies, which were taken into consideration for the attribution of a CF for each foodstuff, makes difficult to known uncertainty for each value. Therefore, a reliability uncertainty analysis in the CF impact could not be conducted. On the other hand, lacking information in the quantification of the WF of non-aquaculture fish species and assumptions made for other foodstuffs included in pastry and ready meals food categories could have generated deviation in the WF assessment. To summarize, the results of this study can contribute to give information about the environmental implications (GHG emissions and water demand) of adopting the EAT-Lancet recommendations. Future research could focus on accounting deviations from EAT-Lancet targets in other dimensions: diet-healthrelations, nutrients need, economicfeatures, or cultural acceptance. Ensuring compliance with the global targets on achieving sustainable healthy diets outlined by EAT-Lancet will enable achievement of the Sustainable Development Goals and the Paris Agreement. CRediT authorship contribution statement Conceived and designed the experiments: C.C.-F., S.G.-G., G.F., M.T.M.; Performed the experiments: C.C.-F.; Analyzed the data: C.C.-F., S.G.-G., M.T.M; Contributed materials/analysis tools: G.F., M.T.M; Writing of the original draft: C.C.-F.; Review &Editing: C.C.-F., S. G.-G., G.F., M.T. M. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements C.C.-F. would like to thank the Ministry of Science, Innovation and Universities for financial support (Grant reference FPU 19/06648). C.C-F, S.G.-G., G.F. and M.T.M. belong to the Galician Competitive Research Group (GRC ED431C 2017/29) and to the Cross-disciplinary Research in Environmental Technologies (CRETUS Research Center, ED431E 2018/ 01). All these programs are co-funded by FEDER (EU). Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi. org/10.1016/j.scitotenv.2022.155683. References Batlle-Bayer, L., Bala, A., García-Herrero, I., Lemaire, E., Song, G., Aldaco, R., Fullana-i- Palmer, P., 2019. The spanish dietary guidelines: a potential tool to reduce greenhouse gas emissions of current dietary patterns. J. Clean. Prod. 213, 588–598. https://doi. org/10.1016/j.jclepro.2018.12.215. BEDCA, 2017. Agencia Española de Seguridad Alimentaria y Nutrición. Base de Datos Española de Composición de Alimentos. https://www.bedca.net/. Benton, T., Bieg, C., Harwatt, H., Pudassaini, R., Wellesley, L., 2021. Food system impacts on biodiversity loss. Energy, Environment and Resources Programme. Blackstone, N.T., 2020. Comparing the recommended eating patterns of the EAT-lancet commission and dietary guidelines for Americans : implications for sustainable nutrition. Curr. Dev. Nutr. 4. https://doi.org/10.1093/cdn/nzaa015. Blas, A., Garrido, A., Unver, O., Willaarts, B., 2019. A comparison of the Mediterranean diet and current food consumption patterns in Spain from a nutritional and water perspective. Sci. Total Environ. 664, 1020–1029. https://doi.org/10.1016/j.scitotenv.2019.02.111. Cambeses-Franco, C., González-García, S., Feijoo, G., Moreira, M.T., 2021. Driving commitment to sustainable food policies within the framework of American and European dietary guidelines. Sci. Total Environ. 807, 150894. https://doi.org/10.1016/j.scitotenv. 2021.150894. Castañé, S., Antón, A., 2017. Assessment of the nutritional quality and environmental impact of two food diets: A Mediterranean and a vegan diet. J. Clean. Prod. 167, 929–937. https://doi.org/10.1016/j.jclepro.2017.04.121. Chen, C., Chaudhary, A., Mathys, A., 2020. Nutritional and environmental losses embedded in global food waste. Resour. Conserv. Recycl. 160, 104912. https://doi.org/10.1016/j. resconrec.2020.104912. CREA, 2019. Linee Guida per una sana alimentazione. Centro di Ricerca Alimenti e Nutrizione. Esteve-Llorens, X., Martín-Gamboa, M., Iribarren, D., Moreira, M.T., Feijoo, G., González- García, S., 2020. Efficiency assessment of diets in the spanish regions: a multi-criteria cross-cutting approach. J. Clean. Prod. 242. https://doi.org/10.1016/j.jclepro.2019. 118491. Esteve-Llorens, X., Van Dooren, C., Álvarez, M., Moreira, M.T., Feijoo, G., González-García, S., 2021. Environmental and nutritional profile of food consumption patterns in the different climatic zones of Spain. J. Clean. Prod. 279, 123580. https://doi.org/10.1016/j.jclepro. 2020.123580. European Commission, 2020. FOOD 2030 Pathways for Action. Healthy, Sustainable and Personalised Nutrition. https://doi.org/10.2777/907055. Funabashi, M., 2018. Human augmentation of ecosystems: objectives for food production and science by 2045. npj Sci. Food 2, 1–11. https://doi.org/10.1038/s41538-018-0026-4. Garcia-Herrero, I., Hoehn, D., Margallo, M., Laso, J., Bala, A., Batlle-Bayer, L., Fullana, P., Vazquez-Rowe, I., Gonzalez, M.J., Durá, M.J., Sarabia, C., Abajas, R., Amo-Setien, F.J., Quiñones, A., Irabien, A., Aldaco, R., 2018. On the estimation of potential food waste reduction to support sustainable production and consumption policies. Food Policy 80, 24–38. https://doi.org/10.1016/j.foodpol.2018.08.007. González-García, S., Green, R.F., Scheelbeek, P.F., Harris, F., Dangour, A.D., 2020. Dietaryrecommendations in Spain –affordability and environmental sustainability? J. Clean. Prod. 254, 120125. https://doi.org/10.1016/j.jclepro.2020.120125. Gustavsson, J., Cederberg, C., Sonesson, U., Emanuelsson, A., 2013. The Methodology of the FAO Study : “Global Food Losses and Food Waste - Extent, Causes and Prevention”-FAO, 2011, SIK Report No. 857. Harris, F., Green, R.F., Joy, E.J.M., Kayatz, B., Haines, A., Dangour, A.D., 2017. The water use of indian diets and socio-demographic factors related to dietary blue water footprint. Sci. Total Environ. 587, 128–136. https://doi.org/10.1016/j.scitotenv.2017.02.085. Health Council, 2015. Dutch Dietary Guidelines 2015. Gezondheidsraad, The Hague, Netherlands. Hirvonen, K., Bai, Y., Headey, D., Masters, W.A., 2020. Affordability of the EAT–Lancet reference diet: a global analysis. Lancet Glob. Health 8. https://doi.org/10.1016/S2214-109X (19)30447-4 e59-e66. ISO, 2006. ISO 14040:2006. Environmental Management-Life Cycle Assessment. International Organization for Standarization. Kesse-Guyot, E., Rebouillat, P., Brunin, J., Langevin, B., Allès, B., Touvier, M., Hercberg, S., Fouillet, H., Huneau, J.F., Mariotti, F., Lairon, D., Pointereau, P., Baudry, J., 2021. Environmental and nutritional analysis of the EAT-lancet diet at the individual level: insights from the NutriNet-Santé study. J. Clean. Prod. 296, 126555. https://doi.org/10.1016/j. jclepro.2021.126555. Knuppel, A., Papier, K., Key, T.J., Travis, R.C., 2019. EAT-lancet score and major health outcomes: the EPIC-Oxford study. Lancet 394, 213–214. https://doi.org/10.1016/S0140- 6736(19)31236-X. Kovacs, B., Miller, L., Heller, M.C., Rose, D., 2021. The Carbon Footprint of Dietary Guidelines Around the World : A Seven Country Modeling Study. 20, pp. 1–10. https://doi.org/10. 1186/s12937-021-00669-6. MAPA, 2019. Spanish Ministry of Agriculture, Food and Environment. https://www.mapa. gob.es/app/consumo-en-hogares/consulta.asp. Mekonnen, M.M., Hoekstra, A.Y., 2012. A global assessment of the water footprint of farm animal products. Ecosystems 15, 401–415. https://doi.org/10.1007/s10021-011-9517-8. Mekonnen, M.M., Hoekstra, A.Y., 2011. The green, blue and grey water footprint of crops and derived crop products. Hydrol. Earth Syst. Sci. 15, 1577–1600. https://doi.org/10.5194/ hess-15-1577-2011. Ministerio de Agricultura Pesca y Alimentación, 2020. Informe del Consumo de Alimentacón en España 2019. Minist. Agric. Pesca y Aliment, p. 674. Muthu, S.S., 2020. Calculating the water and energy footprints of textile products. Assess. Environ. Impact Text. Cloth. Supply Chain, pp. 77–93 https://doi.org/10.1016/b978-0-12- 819783-7.00004-1. Neira, M., Onis, M., 2006. The Spanish strategy for nutrition, physical activity and the prevention of obesity. Br. J. Nutr. 96, S8–S11. https://doi.org/10.1079/bjn20061690. Pahlow, M., van Oel, P.R., Mekonnen, M.M., Hoekstra, A.Y., 2015. Increasing pressure on freshwater resources due to terrestrial feed ingredients for aquaculture production. Sci. Total Environ. 536, 847–857. https://doi.org/10.1016/j.scitotenv.2015.07.124. Sharma, M., Kishore, A., Roy, D., Joshi, K., 2020. A comparison of the indian diet with the EAT-lancet reference diet. BMC Public Health 20, 1–13. https://doi.org/10.1186/ s12889-020-08951-8. Springmann, M., Clark, M., Mason-D’Croz,D.,Wiebe,K.,Bodirsky,B.L.,Lassaletta,L.,deVries,W., Vermeulen,S.J.,Herrero,M.,Carlson,K.M.,Jonell, M., Troell, M., DeClerck, F., Gordon, L.J., Zurayk,R.,Scarborough,P.,Rayner,M.,Loken,B.,Fanzo,J.,Godfray,H.C.J.,Tilman,D., Rockström, J., Willett, W., 2018a. Options for keeping the food system within environmental limits. Nature 562, 519–525. https://doi.org/10.1038/s41586-018-0594-0. Springmann, M., Spajic, L., Clark, M.A., Poore, J., Herforth, A., Webb, P., Rayner, M., Scarborough, P., 2020. The healthiness and sustainability of national and global food based dietary guidelines: modelling study. BMJ, 370 https://doi.org/10.1136/bmj. m2322. Springmann, M., Wiebe, K., Mason-D’Croz, D., Sulser, T.B., Rayner, M., Scarborough, P., 2018b. Health and nutritional aspects of sustainable diet strategies and their association with environmental impacts: a global modelling analysis with country-level detail. Lancet Planet. Health 2. https://doi.org/10.1016/S2542-5196(18)30206-7 e451-e461. Trichopoulou, A., Vasilopoulou, E., 2016. Mediterranean diet. Encycl. Food Health 711–714. UNICEF, 2021. Food-Based Dietary Guidelines. A Review of National Guidance for Children, Adolescents and Women. USDA, HHS, 2020. Dietary Guidelines for Americans 2020-2025. Vanham, D., Guenther, S., Ros-Baró, M., Bach-Faig, A., 2021. Which diet has the lower water footprint in Mediterranean countries? Resour. Conserv. Recycl. 171, 105631. https://doi. org/10.1016/j.resconrec.2021.105631. C. Cambeses-Franco et al. Science of the Total Environment 836 (2022) 155683 9