Carbon footprint and nutritional quality of different human dietary choices

Carbon footprint and nutritional quality of different human dietary choices Sara González-García, Xavier Esteve-Llorens, María Teresa Moreira, Gumersindo Feijoo Accepted Mansucript How to cite: González-García, S., EsteveLlorens, X., Moreira, M., & Feijoo, G. (2018). Carbon footprint and nutritional quality of different human dietary choices. Science Of The Total Environment, 644, 7794. doi: 10.1016/j.scitotenv.2018.06.339 Copyright information: © 2018 Elsevier Ltd. This manuscript version is made available under the CC-BY-NC-ND 4.0 license (http://creativecommons.org/licenses/by-nc-nd/4.0/) Carbon Footprint and Nutritional Quality of different human dietary choices: A methodical review Sara González-García*, Xavier Esteve-Llorens, María Teresa Moreira and Gumersindo Feijoo Department of Chemical Engineering, School of Engineering, University of Santiago de Compostela. 15782Santiago de Compostela, Spain. * Corresponding author: E-mail address: [email protected] Abstract The production of food products presents an outstanding contribution to the global greenhouse gases emission. Since nutrition is a basic human need and humans are able to select the foodstuffs that constitute their diets, dietary choices have a remarkable effect on climate change. In this study, differences on the carbon footprint and nutritional quality of different dietary choices selected from a methodical review have been analysed. In addition, methodological gaps with relevant importance for the outcomes have been also identified. The review includes 21 peer-review journal studies (after a specific selection criteria) deriving on 66 dietary scenarios located in different countries all over the world. We identified that so-called recommended diets such as the Mediterranean and the Atlantic ones present both high nutritional scores (NRD9.3, a composite nutrient score of a diet) and low carbon footprints. Dietary choices identified in North and West Europe as well as in USA, present the highest carbon footprints. In both European regions, dairy products constitute a basic source of nutrients and high-quality protein. In addition diets adhering to food based dietary guidelines (i.e., healthy diets) do not always derive on lower GHGs emission rates. In general lines, dietary choices rich on plant based products (e.g., vegan, vegetarian as well as Indian and Peruvian) have a better environmental profile than these rich on meat (mostly, ruminant meat). In line with these findings, shifting from ruminant meat to chicken, pork and poultry meat, the iso-caloric substitution of animal based protein by other alternative foodstuffs and the promotion of olive oil ingestion may be coherent with more environmental and healthy diets. We conclude that what we eat plays an important role in the evaluation of the sustainability of people's lifestyles. Thus, although meat and dairy products are the most GHGs-intensive foodstuffs, their complete removal from the daily diet could not be realistic in many cultures deriving also on no-healthy habits due the supply of some micronutrients (e.g., calcium and vitamin D) below the recommended daily levels. Limitations were identified in the consulted studies based on the consideration of different system boundaries as well as background uncertainties linked to data sources. Therefore, efforts should be paid into the development of consistent and agreed methods to estimate both carbon footprint and nutritional quality based scores to avoid discrepancies. Keywords: Diet; Greenhouse gases emission; Healthy diet; Life Cycle Assessment; NRD9.3; Sustainable diet 1. Introduction The food system is considered one of the most important responsible issues of negative environmental impacts in Europe, mainly in terms of Greenhouse Gases (GHGs) emission, water requirements and land use (Friel et al., 2009; Tukker et al., 2011; Wolf et al., 2011). Moreover, it is expected an increment of impacts from food consumption due to the population and wealth growth, deriving the latter to higher impact diets richer in meat and dairy products (Tukker et al., 2011). Nutrition is a basic human necessity and food consumption involves a complex system including steps such as food production (e.g., agricultural and farming activities), processing, distribution, final consumption and disposal of waste (Duchin, 2005; Friel et al., 2009). On the other hand, human diets are more than the sum of individual food items. They are complex combinations of different food ingredients, influenced by cultural and regional preferences (de Ruiter et al., 2014). The access to an appropriate, healthy and suitable nutrition as well as food consumption trends depend on factors such as lifestyle, marketing and, on political and economic aspects (Hawkesworth et al., 2010; Heller et al., 2013). Over the last seventy years, food consumption patterns have considerably changed (Vranken et al., 2014). In this sense, two different stages could be identified in the nutrition transition: the expansion stage, in which consumers increased their energy intake through an augmented intake of vegetable based foodstuffs and, the substitution stage (the current one in industrialised and emerging countries), where carbohydrate rich ingredients (e.g. cereals, roots) are being replaced by vegetable oils, sugar and animal based products (Vranken et al., 2014). The relevance of the link between the choice of diet, longevity and health is well known and in recent years it is receiving special attention due to the rising awareness by the society (Friel et al., 2009; Thaler et al., 2015). Dietary habits can contribute or prevent diseases such as diabetes, cancer and cardiovascular diseases (WHO, 2003; Wolf et al., 2011). Diets rich on salt, saturated fat and free sugars are example of unhealthy dietary choices (Hawkesworth et al., 2010). In this sense, the low incidence of cardiovascular diseases in the countries located around the Mediterranean Sea has partially been attributed to their dietary habits (Menotti et al., 1990; Duchin, 2005; Estruch et al., 2006). The well-known Mediterranean diet is predominantly a plant based diet rich on fruits, vegetables and nuts and low in meat, added sugars, saturated fatty acids and salty snacks (Castañé and Antón, 2017). Olive oil (rich in monounsaturated fatty acids) is the main source of fat that may beneficially influence the risk for these diseases (Estruch et al., 2006). However, it is true that the dietary pattern is only one factor influencing human health amongst others such as physical activity (Wolf et al., 2011). Moreover, differences on dietary habits can be found between countries (Van Kernebeek et al., 2014). In this sense, Western diets are based on a high intake of meat, dairy products and eggs deriving on higher ingestions of saturated fat exceeding the dietary recommendations (Tukker et al., 2001; Westhoek et al., 2014). Another topic gaining increasing consideration is the relationship between dietary pattern, resources consumption and environmental impacts. According to Friel et al. (2009), about the 50% of all food system derived GHGs emission is linked to farming activities mainly due to nitrous oxide (from feeding crops production), methane (from enteric fermentation) and carbon dioxide (from agriculturally-induced change in land use) emissions. In this line, Garnett (2011) and Smil (2002) reported that meat and dairy products are the foodstuffs carrying the greatest environmental burdens and depleting resources. In addition, according to the current population growth trend and foodstuffs intake, it is projected an increment of livestock based products demand of up to 70% by 2050 (Ran et al., 2017). Thus, plant based diets (lactoovovegetarian diets) are considered more environmentally friendly in comparison with these containing high resources-intensive products, i.e., diets rich on meat (Baroni et al., 2007; Risku-Norja, 2011; Van Kernebeek et al., 2014). In this sense, research studies focused on designing more environmentally sustainable food production chains remark the necessity of promoting more sustainable dietary patterns (Stehfest et al., 2009; Röös et al., 2015). Besides, dietary changes could be attractive not only from an environmental perspective but also in terms of human health and life expectancy. Health recommendations clearly highlight that less animal based food (specifically beef and pork meat) and more plant based food should be consumed (Stehfest et al., 2009; Thaler et al., 2015). Therefore, the achievement of sustainable food security in the future is linked to a dietary shift from a meat to a plant based diet (Godfray et al., 2010). Regarding the estimation of the environmental impact of dietary habits and/or daily diets, it is generally quantified based on data from life cycle assessments (LCAs), which has emerged as a dominant methodological framework. Although LCA method follows international guidelines (ISO 14040, 2006), it presents flexibility allowing the application to a widespread range of systems (Heller et al., 2013). However, the methodological approach followed in the analysis must be clearly defined since it can have a decisive effect on the quality of the results. System boundaries definition, data managed, functional unit and other inherent uncertainties in the method are factors that must be noticeably established specifically in comparative studies. Numerous studies can be found in the literature considering the environmental footprint (mostly carbon footprint) of European food consumption patterns following an LCA framework (Heller et al., 2013; Van Kernebeek et al., 2014) and even environmental friendly strategies (e.g., refusal of air-transported products, promoting organic production, partial meat replacement by plantbased food, dairy products or mixed food) have been proposed in some cases (Jungbluth et al., 2000; Hallström et al., 2015). However it is important to take in mind the nutritional quality of the diets as well as the recommended intake levels of proteins and energy when dietary patterns are compared and alternative modifications on the dietary habits are proposed (Röös et al., 2015). In this study, the carbon footprint, nutritional quality (in terms of Nutrient Rich Diet 9.3 score, NRD9.3) and daily energy intake (i.e. kcal) of different European dietary choices (daily diets) are assessed and compared for several reasons: 1) to determine the relationship between these items (when possible), 2) to identify differences in the same dietary pattern (i.e., vegan and vegetarian diets) between European regions, 3) to answer the question if eating less meat is more environmental friendly maintaining always the nutritional recommendations and 4) to demonstrate the potential environmental benefits of introducing alternative foodstuffs (e.g., superfood) in human nutrition to replace meat consumption. It is important to take in mind that the number of calories that an average person needs daily depends on several factors, such as the minimum and average dietary energy requirements (Vázquez-Rowe et al., 2017), level of activity, gender, age, weight, geographical location and cultural issues (EFSA, 2009). The carbon footprint has been considered as a proxy for environmental impact in line with the studies checked and taking into consideration the awareness from the research community and society on the reduction of GHGs emission and counteracting the climate change effects. To do so, a detailed review has been performed considering LCA based studies available in the literature. To address the comparison, different LCA methodological aspects (boundary and scope of assessments and defining diet) specific for food from a consumption perspective have been established following the recommendations from Heller et al. (2013). 2. Material and methods 2.1. Literature search strategy It is well-known the remarkable differences between European countries and the consideration of one representative European diet is not realistic. Different data sources are available giving information regarding food consumption in the European Union either by surveys, household economic expenditure data and food balance sheets from national statistics considering imports and exports (Tukker et al., 2011). A literature review of related studies measuring environmental impacts (specifically carbon footprint estimation) of current dietary patterns that is, considering only daily diets (i.e., complete diets) and excluding the ones focussed on single food items or meals has been performed to ensure scientific quality and to increase the comparability between similar diets as well as following a consumption perspective (Heller et al., 2013). To do so, Scopus, Web of Knowledge (ISI) and Google scholar have been managed considering only English language peer-review studies. According to it, although more than 50 studies have been found assessing environmental impacts of country diets, in total 21 peer-review studies (Wallén et al., 2004; Risku-Norja et al., 2009; Muñoz et al., 2010; Pathak et al., 2010; Tukker et al., 2011; Vieux et al., 2012; Meier and Christen, 2013; Sáez-Almendros et al., 2013; Saxe et al., 2013; Vieux et al., 2013; Werner et al., 2013; Wilson et al., 2013; Scarborough et al., 2014; van Dooren et al., 2014; Röös et al., 2015; van Dooren et al., 2016; Castañé and Antón, 2017; Pairotti et al., 2017; Vázquez-Rowe et al., 2017; Esteve-Llorens et al., 2018; van de Kamp et al., 2018) have been identified fulfilling the selection criteria. 2.2. Food consumption scenarios Food has multiple functions for humans: i) supplying nutrients (i.e., proteins, and vitamins) and energy as well as ii) offering pleasure, culture, and social identity. The function of the dietary patterns under evaluation is to satisfy both items. In this study, 12 different countries have been considered for analysis after the literature search not with the aim of directly comparing the countries, but with highlighting how diets and habits can considerably differ between them, even following the same style diet (e.g., vegetarian, vegan or Mediterranean). Therefore, the different countries have been previously classified in four different zones or clusters (Tukker et al., 2011; Vanham et al., 2013): Western Europe (Germany, France and The Netherlands), Northern Europe (United Kingdom, Sweden, Finland and Denmark), Southern Europe (Spain and Italy) and other countries (Perú, India, USA and New Zealand). Within each of the three European regions or zones there are similarities due to related climatological conditions which can be important in terms of background involved processes (e.g., agricultural or farming production), per capita food consumption ratios and energy intake as well as consumption behaviour for remarkable food groups (e.g., olive oil as fat source in the diet). Finally, to only capture the effect linked to changes in dietary habits and diets composition, additional differences based on foodstuffs production methods (i.e., organic vs conventional) have not been considered for evaluation in this study. Using the literature search, it has been defined 66 European examples of dietary choices which are mentioned as nutritionally healthy diets, current dietary patterns at national level, balanced diets, lifestyle diets and/or environmental sustainable ones. Table A1 provides an overview of the main characteristics of the LCA dietary scenarios proposed for assessment including the identification of the corresponding system boundaries. A detailed description of scenarios per cluster of countries in reported below. 2.2.1. Dietary scenarios in Southern Europe Mediterranean diet is characteristic of the Mediterranean climate area due to the resources available and it is typical of people from Spain, Italy, Northern Africa and Greece (Pairotti et al., 2015). The Southern Europe cluster has included in our classification Spain and Italy as representative countries since no studies have been found in the literature search regarding the remaining involved ones. It is considered an example of healthy and nutritious dietary pattern and recognised by the UNESCO as an intangible heritage of humanity (Castañé and Antón, 2017). Mediterranean diet does not exclude animal based food but rather admits to low ingestions of them. It is a diet based on a high vegetable intake as well as cereals, fruits and fish. The ingestion of other food ingredients such as meat, eggs, dairy products and sweets is limited (Sáez-Almendros et al., 2013; Pairotti et al., 2015). However it is important to consider that current diet in Mediterranean countries are departing from the traditional one, deriving in different quantities of food groups as consequence of food production globalisation (SáezAlmendros et al., 2013; Pairotti et al., 2015). In this line, Mediterranean diet in Italy is richer on pasta and flour products than in Spain. However, e.g. potatoes are consumed in higher amount in Spain than in Italy. According to our literature review, not only the Mediterranean type diet receives special attention on Spain (Muñoz et al., 2010; Sáez-Almendros et al., 2013; Castañé and Antón, 2017) and Italy (Pairotti et al., 2017), but also other dietary patterns that currently exit in these countries must be considered due to the population dietary behaviour. The term Mediterranean type diet refers to one with the characteristics previously mentioned, but they could be realized through alternative choices for individual foods. Therefore, it has not the rights on healthy food. The Atlantic dietary pattern is common in Northern Portugal and Galicia (North-western Spain). It is characterised by the abundant consumption of plant based food as well as local and fresh products (seasonal food) minimally processed, being considered as an example of healthy diet (Vaz Velho et al., 2016). The consumption of meat (mostly beef and pork) and eggs is also moderate. Olive oil is used for seasoning and cooking and constitutes the main source of fat (Calvo-Malvar et al., 2016). The main differences with regard to the Mediterranean diet is the largest consumption of fish, red meat, pork, milk, potatoes, fruit and vegetables (GuallarCastillón et al., 2013). Nevertheless, Mediterranean and Atlantic diet are not the only types of diet as example of healthy ones. Thus, additional dietary behaviours have also been taken into consideration. Vegan and vegetarian dietary patterns are also extended in the Mediterranean area (and in the other clusters under study). The vegetarian diet is based on the consumption of vegetables (any), fruits, dairy products, flour based products and eggs excluding any type of animal based food (Baroni et al., 2007; Pairotti et al., 2015). Regarding the vegan dietary pattern, is also a plant based diet but excluding any foodstuff of animal origin that is, eggs, dairy products, fish phosphorus, potassium, iron and sodium) described by Verger et al. (2012), in this study it has been considered the Nutrient Rich Diet (NRD9.3) score mainly due to the lack of existing information regarding all the parameters required in the previous mentioned scores and supported by published studies available in the literature (Drewnowski, 2009; Fulgoni et al., 2009; van Kernebeek et al., 2014; Röös et al., 2015; Castañé and Antón, 2017). Thus, a total of nine nutrients to encourage (protein, fibre, calcium, iron, magnesium, potassium, vitamin A, vitamin C, vitamin E) and three nutrients to limit (sodium, saturated fat and total sugar) have been considered for the estimation of the score taking into account the recommend intake for each nutrient to encourage and the maximum intake for each nutrient to limit reported by the European Food Safety Authority (EFSA, 2010; 2012). In addition, having in mind the trends in Europe towards a Mediterranean diet on a global scale as well as healthy effects derived from Atlantic diet, NRD9.3 scores have also been computed considering the recommended daily average nutrients intake in both diets (Fundación Española de la Nutrición, 2004; Castañé and Antón, 2017). Table 1 displays the recommend intakes for each nutrient to encourage and the maximum ones for each nutrient to limit considering the recommendations previously reported. Table 1. Recommended nutrients daily intake (RDV). Acronyms: A - RDV from EFSA (2010; 2012); BRDV from Castañé and Antón (2017); C - RDV from Fundación Española de la Nutrición (2004). Nutrients to encourage Nutrients to limit Protein Fiber Vit A Vit C Vit E Ca Fe K Mg Saturated fat Total sugar Na g g µg mg mg g mg g mg g g g A 57 25 800 80 12 0.80 14 2.0 375 20 90 2.4 B 50 25 1850a 1030b 210b 1.0-2.5 18-45 3.5 400 20 50 1.5-2.4 C 91 21 1404 179 13 1.01 13.8 3.5 237 28 77 1.9 a 700-3000 µg; b 60-2000 mg; c 20-1000 mg 3. Results and discussion 3.1. Carbon footprint of selected dietary scenarios In this study, GHGs emission (in kg CO2eq·person-1·day-1) have been used for the comparison of the environmental impact of different food consumption habits. However, there is a limitation that must be considered and thus, the carbon footprints reported in Table A2 must be interpreted with prudence. The limitation is linked to a methodological approach considered for the estimation of this environmental impact that is, the system boundaries. In LCA studies the system boundaries define the processes or activities included in or excluded from the analysis. Although a consumption perspective was considered in all of the scenarios and only daily diets were selected, the consumption phase (which could include activities beyond the time of purchase such as transport from retailer to the house, home storage, cooking, management of end-of-life phases and even human excretion) has been only considered in 39% of the scenarios. Even when this phase is computed, not always the activities included are the same. Human excretion, waste water treatment and solid waste management have been considered by Muñoz et al. (2010) although they were not computed in this review (they represent the 17% of total GHGs emissions) in S7. On the other hand, cooking in household has not been considered in seven scenarios (S8-S11 and S20-S22). Tukker et al. (2011) excluded from their analysis the consumption phase assuming that its effect (i.e., kg CO2eq derived) on the global carbon footprint value is the same regardless the dietary pattern. However, it could not be totally right mainly comparing diets from different countries (Heller et al., 2008; Berlin and Sund, 2010). Moreover, the consumption phase could have an interesting effect in these types of food that derive on small amounts of GHGs throughout their production chain. The cultivation to retailer approach has been considered by 33% of the scenarios and the cultivation to farm gate approach, in 28% of total. Since in the comparison between different diets the system boundaries must be comparable, the evaluations have been firstly performed between studies with similar methodological assumptions. 3.1.1. Cultivation-to-consumer approach The impact of dietary change on GHGs emissions is depicted in Figure 1 considering a cultivation-to-consumer perspective (26 of the 66 scenarios). The diets under this approach have been classified in eight groups such as Atlantic diet (1), Mediterranean diets (2), vegan diets (2), vegetarian diets (5), healthy diets (3), diets based on current dietary consumption patterns (6), lifestyle diet (1) and omnivorous diets with different ratios of dairy products consumption (6). According to the results depicted, remarkable differences can be identified not only between diet groups but also within each group. Diets reducing or completely avoiding the ingestion of animal based products such as the vegetarian and vegan derive on the best carbon footprint profiles that is, they are more sustainable in terms of GHGs emissions. This is in line with other studies (Werner et al., 2013; Hallström et al., 2015; Vázquez-Rowe et al., 2017). Figure 1. Comparative profile in terms of GHGs emission of reviewed scenarios based on a cultivation-to-consumer approach. For references and detailed information about each scenario see Table A1. Regarding the Mediterranean diet based scenarios (S2 and S10) a huge difference has been identified between the estimated GHGs emission being 1.8 times higher for S10 with regard to S2. In both scenarios the assessment of the diet was based on Mediterranean diet recommendations (i.e., recommended foodstuffs and servings) following the corresponding food pyramid. However, S10 corresponds to an Italian variance where differences can be identified in the pyramid regarding the food proportions and consumption frequencies (e.g. .potatoes). Moreover, although in both studies the cultivation-to-consumer perspective was established, 0 1 234567 S1 S2 S10 S3 S24 S9 S23 S62 S65 S66 S11 S20 S45 S7 S8 S21 S44 S63 S64 S22 S25 S26 S27 S28 S29 S30 kg CO 2 eq·person -1 ·day -1 Dietary scenarios Omnivourous diets -effect from dairy products Lifestyle diet Dietary patterns Healthy diets Vegetarian diets Vegan diets Mediterranean diets Atlantic diet some differences have been identified on this issue. A LCA-Input Output Analysis based model (hybrid method) was applied in S10 including in the estimations the environmental impact of the whole supply chain of food products that is, from production to the disposal of packaging .On the contrary, S2 followed a standard LCA approach. Thus, some impacts from the life cycle of food products were not accounted for in S2 as difference to S10. The Atlantic diet (S1) reported a carbon footprint score very close (3% lower) to the Spanish Mediterranean diet (S2). As previously indicated, both diets present the same philosophy regarding the consumption of fruits, vegetables and olive oil. However, Atlantic diet prioritizes seasonal, fresh and local products consumption, deriving on a reduction of GHGs emission. Regarding the vegan (S3 and S24) and vegetarian (S9, S23, S62, S65 and S66) diets available in the literature, a huge range of fluctuations on the carbon footprint estimations can be identified in Figure 1. It is well-known that diets which most reduce the amount of meat consumption derive on the best environmental profiles (Sonesson et al., 2009; Pathak et al., 2010; Aleksandrowicz et al., 2016) in favouring of vegan and vegetarian diets. It can be identified in S3, S24, S62, S65 and S66 (see Figure 1). However, two exceptions have been detected with S9 and S23. The former corresponds to an Italian vegetarian diet including dairy products and eggs consumption but no fish. In addition, cooking stage is not computed. The latter is a Danish vegetarian diet that as difference to S9 includes fish and food preparation in household. Surprisingly, the carbon footprint reported for S9 is up to 1.6 times higher than S23. However, it can be explained due to differences on the method selected for analysis. S23 followed the standard LCA approach for the estimation of GHGs emission. In the case of S9, the previously mentioned hybrid method was considered, accounting for impacts corresponding to waste disposal and management. Six daily diets based on current dietary consumption patterns have been also identified and correspond to S7 (Spain), S8 (Italy), S21 (Sweden), S44 (The Netherlands), S63 (India) and S64 (India). According to the revised studies, diets based on current dietary consumption patterns involve a wide array of GHGs emission ranging from 0.68 kg CO2eq·person-1·day-1 (S64) to 5.80 kg CO2eq·person-1·day-1 (S44). These diets are mainly based on statistical data of actual food purchases not in recommended food intakes which is expected to derive on worse environmental profiles and higher ratios of food amounts (Castañé and Antón, 2017). It is important to have in mind that large amounts of losses and waste can occur in the consumption phase (e.g., cooking step, inedible parts of the food such as bones or peels), which are initially not accounted for in studies where diets are based on recommended intakes. Surprisingly, S63 and S64 – scenarios focused on Indian non-vegetarian dietary patterns including the consumption of mutton and chicken meat, respectively, involve carbon footprints up to 9 times lower than the ones for dietary patterns in other countries such as in The Netherlands (S44), Italy (S8), Sweden (S21) or Spain (S7). In fact, the Indian diets reported the lowest carbon footprint scores in line with vegetarian diets (see Figure 1) in spite of including meat as well as the lowest energy intake per day (1,739 kcal). One of the most important issues responsible of these outstanding differences between the carbon footprints is the dietary habit which is specifically different between European countries and India as well as only three greenhouse gases were accounted for (methane, nitrous oxide and carbon dioxide). The average quantity of food ingestion per adult person is surprisingly low in Indian studies in comparison with available data found in the literature for other diets (i.e., European ones). According to Pathak et al. (2010), although India has a wide diversity of foodstuffs, Indian consumption habits are based on wheat, rice and pulses as staple foods (225g, 250g and 65g per day, respectively) together with vegetables. Meat from mutton is consumed as well as chicken, but in a minor extent (e.g., 30 g per day). However, European dietary patterns involve larger amounts (on average) of meat consumed per day – mostly red meat (Westhoek et al., 2014; Pairotti et al., 2015; Röös et al., 2015; van de Kamp et al., 2018). Differences also exist regarding the consumption of dairy products (100g per day) being higher in European countries (Westhoek et al., 2014). It is well known that the intensity of GHGs emission from livestock based products that is, meat (and meat based products) and dairy products, is higher in comparison with plant-based products (Garnett, 2011; Westhoek et al., 2014). Therefore, having in mind the Indian habits, it could be expected lower GHGs emission associated to these dietary patterns. Comparing S63 and S64, a reduction on the carbon footprint is reported in S64, which could be expected in line with the results from Westhoek et al. (2014) since mutton meat is replaced by chicken. According to Westhoek et al. (2014), substituting meat consumption from ruminants by monogastric animals contributes significantly to reduce carbon footprint mostly due to the outstanding reduction of methane emissions from enteric fermentation. Regarding the remaining scenarios, S44 reported the highest carbon footprint (5.80 kg CO2eq·person-1·day-1) and corresponds to current Dutch consumption patterns. According to that study (van de Kamp et al., 2018), near 130g of meat (mainly red meat) and 400g of milk based products (including cheese) are consumed, on average per day and person, in that country. Consumption of grain products (including bread), potatoes and pulses is around 230g, 119g and 3g respectively. In the case of S21 (Röös et al., 2015) the carbon footprint is lightly smaller (~10%) in spite of not so outstanding differences on the consumption trends: 110g of meat (mainly pork), 375g of dairy products, 280g of grain products, 190g of potatoes and 16g of legumes. Moreover, both diets report similar daily energy intakes. Thus, the minor intakes of meat (being mostly pork) and dairy products in S21 is behind the best profile in terms of GHGs emission. Regarding S7 and S8, both correspond to current consumption trends in Spain and Italy respectively, being the carbon footprint ~13% higher in S8. Both are Mediterranean countries (belong to the same cluster), follow the Mediterranean diet recommendations and not so outstanding differences could be expected. The consideration of the mentioned hybrid method in S8 (with an accepted level of uncertainty), different data sources and expected differences regarding the consumption of some foodstuffs (e.g., with regard to bread and pasta and, meat based products the average ingestion is of 1.4 and 10 times more portions per capita respectively than average Mediterranean diet - Pairotti et al., 2015) between both countries are behind the worse carbon footprint in the Italian scenario. Regarding the Swedish and Dutch dietary patterns (S21 and S44 respectively) report 1.1 and 1.2 times higher GHGs emission than S7. The different consumption habits based on higher dairy products consumption are responsible of these results. Six Danish omnivorous diets were identified in the literature (Werner et al., 2013) based on different dairy products ingestion, all of them supplying the same daily energy intake to the consumers (2,197 kcal). The interest on these scenarios was based on the analysis of dairy products’ role have on the GHGs emission since their ingestion is recommended due to their supply of high-quality protein, calcium and other valuable nutrients. The carbon footprint values range from 3.62 to 4.83 kg CO2eq·person-1·day-1. The lowest value reported corresponds to the scenario where dairy products are totally substituted by unfortified soy drinks as alternative to milk. It must be highlighted that diets based on soy drinks do not supply the recommended daily levels of some micronutrients such as calcium (up to 40% lower) and vitamin D (~60% lower) between others (Werner et al., 2013). The highest carbon footprint score corresponds to S28, where cheese products are consumed but any other type of dairy product is excluded by means of the ingestion of soft drinks. This option was assessed due to evidences regarding the inverse association between milk and soft drinks consumption (Werner et al., 2013). The production of the consumed amount of soft drinks involves higher GHGs emission than milk based products manufacture (around 1.8 times higher). Moreover, a similar trend as S30 was also identified with regard to the intake of micronutrients, being under the recommended values. On the other hand, three healthy diets were identified for comparison. An Italian diet (S11), a Swedish diet (S20) and a Dutch diet (S45), following recommended nutrients and food intakes in these countries. The carbon footprint scores reported were 1.4 and 1.3 times higher in S11 with regard to S20 and S45, respectively. As previously mentioned, the hybrid method considered for analysis in S11 is the major responsible of these results and it is not possible to identify in which extent. Moreover, these diets belong to different country clusters (SE, NE and WE respectively for S11, S20 and S45) so that inherent differences on foodstuffs consumption and quantities are a reality. Italian diet prioritises bread and pasta as well as vegetables consumption. Regarding the other two, dairy products are of the main nutritional source. Finally S22 represents a lifestyle diet widespread in Sweden and reported the worse carbon footprint score as depicted in Figure 1 (6.30 kg CO2eq·person-1·day-1). This diet is based on the low ingestion of carbohydrates and the high intake of fat. According to Röös et al. (2015), it includes an outstanding daily consumption of eggs (133g), meat –especially pork (230g), butter (32g), cheese (41g) and cream (160g) and a surprising reduction of fruits intake (20g). These values (mostly meat and dairy products) are the responsible of its amazing carbon footprint score. 3.1.2. Cultivation-to-retailer approach The consideration of the system boundaries in daily diets assessment till retailers has been assumed in 22 of the 66 reviewed studies. Thus, consumer stage contribution (transport from retailer to household, cooling and food preparation at home) was excluded from analysis. The impact of dietary change on GHGs emissions is depicted in Figure 2. The diets under this approach have been classified in seven groups such as Mediterranean diets (2), diets based on current dietary consumption patterns (8), vegetarian diets (2), vegan diets (2), diets based on different ratios of meat consumption (4), sustainable diet (1) and healthy diets (3). Figure 2. Comparative profile in terms of GHGs emission of reviewed scenarios based on a cultivation-to-retailer approach. For references and detailed information about each scenario see Table A1. 012345678910 11 12 13 14 15 S4 S33 S5 S6 S18 S31 S38 S39 S56 S57 S12 S42 S13 S43 S14 S15 S16 S17 S19 S32 S40 S41 kg CO2eq·person-1·day-1 Dietary scenarios Meat-based diets Sustainable diet Dietary patterns Healthy diets Vegetarian diets Vegan diets Mediterranean diets According to the results depicted, light differences can be identified not only between diet groups but also within each group although a specific scenario (S56) reports the worst profile. Once again, diets based on the ingestion of plant based products (vegetarian and vegan) derive on the best carbon footprint profiles (S12, S13, S42 and S43). The diet defined as sustainable (S19) derived on the second best score of GHGs emission, close to the ones corresponding to vegan and vegetarian diets. Finally, one of the Mediterranean diets (S4) can be identified as the most environmentally sustainable within the set in terms of GHGs emission. Having a look into the selected scenarios based on current dietary patterns, they correspond to very different dietary habits: Swedish (S18), Peruvian (S57), American (S56), Danish (S31), French (S38), German (S39) and Spanish (S5 and S6). The reported values of carbon footprint range from 2.48 to 13.43 kg CO2eq·person-1·day-1 (S18 and S56, respectively). This huge rank of values considerably depends on the consumption habits, the daily energy intakes as well as carried uncertainties regarding the use of different data sources and possible assumptions (e.g., consideration or not of beverages). In the case of S18, it corresponds to consumption trends in period 1998-2000. Meat based products and dairy products were identified as environmental hotspots due to their outstanding daily consumption (330g and 460g per person, respectively). The authors proposed sustainable food consumption (S19), involving a reduction of 36%, 50% and 30% of meat, cream and cheese intakes, respectively. Therefore, around 6% of improvement in carbon footprint was achieved (S19). Regarding the American dietary pattern (S56), it is based on a large intake of meat (365g) and dairy products (700g) – it is the scenario with the largest ingestions of both foodstuffs, which is responsible of the surprising GHGs emission. Thus, its worse environmental profile could be expected. Danish and German dietary patterns (S31 and S39) are line in terms of GHGs emission (5.56 and 5.62 kg CO2eq·person-1·day-1) although belong to different clusters (NE and WE) and reported different daily energy intakes (1.4 times higher in S31 than S39). Dairy products (378g) and meat (205g) are the most consumed food products in S31 and the main responsible of carbon footprint. In the German diet, both are consumed in a minor amount (253g and 103g, respectively). Thus, differences on the score are mainly affected by background processes. French dietary pattern (S38) presents an environmental behaviour and daily energy intake close to the Spanish one (S6) based on data from consumption surveys from the Spanish Ministry of Agriculture, Food and Environment (Sáez-Almendros et al., 2013). Consumers from both countries reported similar consumption habits according to the consulted statistics. Differences of up to 43% were identified between the Spanish diets designed from consumption patterns, S5 and S6. The consideration of different consumption data is belong that outstanding variance. Finally, Peruvian consumption habits were also included in the comparison (Vázquez-Rowe et al., 2017), reporting a value aligned to those corresponding to vegan diets in spite of including meat and fish consumption. However, Peruvians consume meat and dairy products in a minor extent than Europeans and Americans (100g and 73g, respectively), promoting the ingestion of fruits, vegetables and grains. With regard to vegan and vegetarian diets, two different scenarios (one corresponding to UK and the other to Germany) were identified per diet group. Variations within each group were around 10% and the main rationale behind them could be associated to different foodstuffs ingestion (i.e., doses) and habits but mostly, to the use of different data sources. Three healthy diets were identified under a cultivation-to-retailer approach corresponding to case studies of Denmark (S32) and Germany (S40 and S41). In all of them, sanitary recommendations were followed for their design and differences on the corresponding carbon footprints lower than 4% were identified as depicted in Figure 2. S41 derived on the best profile because besides healthy recommendations, ecological and social constrains were also considered prioritizing organic food. However, the improvement associated to the profile was almost negligible. Finally, four diets were included in the assessment considering different food consumption habits identified in United Kingdom regarding the ingestion of meat. High-, medium-, lowand no-meat consumers were analysed (S14, S15, S16 and S17). As expected, the GHGs emission increased in line with the reduction of meat portions ingestion. 3.1.3. Cultivation-to-farm gate approach (Saxe et al., 2013) in Denmark. A detailed identification of diets and corresponding caloric intake is shown in Table A2. Outstanding differences can be also reported between clusters of countries, being the highest energy intakes identified in NE (2,197-3,095 kcal per person and day). These variations between regions are mainly affected by their consumption habits. Therefore, the relatively high consumption of vegetables, legumes and grains as well as low consumption of animal based products (mainly animal fat since it is substituted by olive oil) characteristic of SE-countries is behind of their lower caloric values (Tukker et al., 2011). However, for the direct comparison between daily diets, all of them have been adjusted to an equalised daily intake of 2,000 kcal. This perspective has also been considered by other authors (Van Dooren et al., 2014; van de Kamp et al., 2018) giving insight into differences in carbon footprint directly derived from differences on the diet composition. Figure 6 displays, per country clusters, the carbon footprints of the dietary choices selected for analysis adjusted to 2,000 kcal. Seven scenarios (S7-S11, S18-S19) have been excluded from the comparison due to the lack of information regarding their caloric intakes per day. Figure 6. Carbon footprint of adjusted diets to 2,000 kcal. Bars in black correspond to recalculations under a cultivation-to-farm gate approach. It is important to bear in mind that although the effect of both the system boundaries (cultivation to consumer, retailer or farm gate) and other methodological assumptions (e.g., hybrid method and data sources) is noticeable on the results, as previously discussed, the composition of the diets (and amount of foodstuffs intake) is the main responsible issue of differences on the carbon footprints. According to Berlin and Sund (2010), contributions to the total GHGs emission from consumption stage (that is, transportation from retailer to household and home cooking) in cultivation-to-consumer approach is around 10%. Esteve-Llorens et al. (2018) reported an average contribution of around 14% from that phase and around 3% for the distribution phase which includes transportation till retailers. Thus, the contribution from both distribution and consumption phases have been removed from the carbon footprints to facilitate the comparison 012345678 S1 S2 S3 S4 S5 S6 S12 S13 S14 S15 S16 S17 S20 S21 S22 S23 S24 S25 S26 S27 S28 S29 S30 S31 S32 S33 S34 S35 S36 S37 S38 S39 S40 S41 S42 S43 S44 S45 S46 S47 S48 S49 S50 S51 S52 S53 S54 S55 S56 S57 S58 S59 S60 S61 S62 S63 S64 S65 S66 kgCO 2 eq/day/person Southern Europe - SE Northern Europe - NE Western Europe - WE 13.02 kgCO 2 eq·day-1·person-1 13.43 kgCO2eq·day-1·person-1 Other countries - O (cultivation-to-farm gate approach in all scenarios) and in line with Tukker and colleagues (Tukker et al., 2011) who reported that their effect is similar in all diet scenarios. Figure 6 also displays the new estimated GHGs based profiles. Applying that restriction the average carbon footprint per day and person is 3.33±1.87 kgCO2eq taking into account 59 dietary scenarios. As expected, dietary choices from SE and other countries such as India and Peru together with vegan and vegetarian diets are linked to the best environmental profiles. Regardless the country cluster, vegan diets do not derive on outstanding deviations on the carbon footprint (2.23±0.51kgCO2eq·day-1·person-1). On the contrary, differences can be highlighted in the vegetarian diets (2.16±1.57kgCO2eq·day-1·person-1). The rationale behind it is related to different dietary guidelines depending on the countries. Dietary choices from NE and WE countries reported around 3.50±1.10 kgCO2 eq and 3.53±0.8kgCO2 eq per day and capita respectively. In both clusters, the results are affected by the largest ingestion of dairy products and red meat, excluding vegan and vegetarian diets. According to the literature review, estimations for carbon footprint of diets can considerably vary even up to four times (e.g.,S4 and S5) even considering the same system boundaries to calculate the GHGs emission. The consideration of different approaches regarding foreground and background data sources (i.e., data from statistical consumptions or data according to sanitary recommendations), transport mode and distances as well as production systems can derive on underestimations. Thus, the use of appropriate system boundaries as well as the definition of methods that allow a correct trade-off between available data sources and accurate results is mandatory mostly in decision making strategies focused on diets. According to the results displayed in Figure 6 and Table A1, carbon footprint can decrease till 30% when changing from average dietary pattern to a healthy diet (e.g., S20 vs S21). In this line, reduction of 60% could be achieved changing from diets rich on meat based products to vegan diets (e.g., S13 vs S14). This large decrease can be explained by the fact that vegan diet is considered also an example of healthy diet (Pernollet et al., 2017). 3.5. Decision support for sustainable and healthy diets The comparative analyses previously reported show that the composition of diets is the main responsible of variations identified in carbon footprints. Thus, possibilities to reduce GHGs emission based on changes of foodstuffs and dietary habits could be proposed. In this sense, diet and sustainability are closely connected. Shifting from animal based to plant based products is supported by numerous studies (Garnett, 2011; Dagevos and Voordouw, 2013; Hoolohan et al., 2013; Westhoek et al., 2014) as the best choice to reduce GHGs emission and from a sustainability viewpoint. However, this option must be analysed in detail since some plant based foodstuffs such as rice could emit surprising GHGs emission. If consumption of rice is presented as an alternative ingredient, special attention should be paid into the cultivation strategy (Pathak et al., 2010; Fusi et al., 2017). Regarding meat consumption, the type of animal considerably affects the GHGs emission and dietary patterns including meat can be also considered as an example of sustainable and healthy diets. In this sense, monogastric meats have smaller carbon footprints than ruminant based meats (Westhoek et al., 2014). Thus, the consumption of red meat is a decisive parameter of the carbon footprint (Hallström et al., 2015). Substitution of animal based protein by alternative foodstuffs such as pulses and legumes could also be considered as an improvement (Pathak et al., 2010). In this sense, attention can be paid into the so-called superfoods such as quinoa, a food product rich on protein of which ingestion (while removing meat) could derive on outstanding reductions of GHGs emissions (Vázquez-Rowe et al., 2017). Satisfying daily energy requirements by increasing the ingestion of fruits and vegetables (isocaloric substitution of animal based protein) have not only environmental but also healthy benefits. However, in case of considering any of these proposals, further attention should be paid into the intake of micro-nutrients, the corresponding nutritive values and getting a balanced diet. A balanced diet supplies the energy and nutrients by means of the ingestion of different and valuable foods in adjusted quantities. Finally, it is also important to bear in mind that only eating the necessary amounts of food is a priority measure to reduce GHGs emission from dietary habits (Garnett, 2011). Therefore, several considerations regarding environmentally and healthfully sustainable eating are worth noting always taking in mind that consumer preferences depends not only on environmental and health issues but also on lifestyles and trends. Public procurement policies are required to raise awareness not only of society but also food manufacturers and food businesses. Social campaigning for promoting the benefits from environmental sustainable diets should also require advises and cultural formation based activities. Attention should be paid into the incorporation of food taxes (Westhoek et al., 2014) although it should be analysed with care. Vinnari and Tapio (2012) proposed a direct taxation (consumption taxes) in livestock based products (i.e., increasing the price of meat and dairy products) motivated either by environmental or ethical issues. It could address not only to reduce these products consumption but also to the optimisation of their manufacturing processes with the aim of achieving the reduction of their environmental profiles. Taxing the environmental effects derived from foodstuffs production was also proposed as an efficient policy instrument to deal with GHGs emission (Wirsenius et al., 2011). Promoting the reduction of meat and dairy products by means of increasing consumption of plant based products is an interesting alternative to mitigate GHGs emission. However, it could mean that the final goal should be a society that would give up animal based products eating to be a vegan society. In this sense, it is well-known that definition of vegan diet (even vegetarian one) is in conflict with the FAO definition of sustainable diet since it is neither nutritionally adequate (some required micro-nutrients are not supply) nor culturally accepted by general society (van Dooren et al., 2014). Therefore, “a sustainable” reduction should be based on the consumption of alternative products which could contribute to satisfy the recommended daily values of nutrients. It is the case of protein rich foodstuffs such as quinoa and pulses, whole grain cereals and shifting from animal to vegetable fats (e.g., olive oil). 4. Conclusions What we eat plays an important role in the evaluation of the sustainability of people's lifestyles. It is well-known that meat and dairy products consumption contributes significantly to GHGs emission being the most GHGs-intensive foodstuffs. However, complete removal of animal based products from the daily diet is not realistic in many cultures and could derive on nohealthy habits. This review has largely focused on the comparative assessment of both carbon footprint and nutritional quality of reported dietary choices (daily diets) available in the literature. This study shows that diets with high nutritional score based on NRD9.3 score such as Mediterranean and Atlantic as well as vegan diets also have high sustainability scores (i.e., low carbon footprints). Indian and Peruvian diets report the best carbon footprint scores mainly due to their characteristic composition rich on the remarkable ingestion of plant-based products such as pulses, grains and vegetables as well as the low intake of livestock based products. Dietary choices from Northern and Western Europe countries derive on the largest carbon footprints mostly due to outstanding consumption of dairy products. Moreover, diets adhering to food based dietary guidelines (i.e., healthy diets) did not always derive on lower GHGs emission rates. In line with the results, shifting the intake of ruminants based meat to a moderate consumption of chicken, pork and poultry meat, the iso-caloric substitution (partial no total) of animal based protein by other foodstuffs such as quinoa and legumes as well as the promotion of olive oil ingestion may be consistent with a more environmentally and healthfully friendly diet. Many of these consulted studies use different system boundaries, ignoring important life cycle stages and underestimating GHGs emissions. Thus, challenges should be beard in mind to improve environmental impact of the food system without reducing the health impact of nutrition. Efforts should be also paid into the development of consistent and agreed methods to estimate both carbon footprint and nutritional quality based scores to avoid discrepancies and to reduce uncertainties in the comparisons. Moreover, policies could be developed to support consumers in making both more environmental sustainable and healthier dietary choices. Acknowledgements This research has been supported by a project granted by Xunta de Galicia (project ref. ED431F 2016/001). Dr. S. González-Garcia would like to express her gratitude to the Spanish Ministry of Economy and Competitiveness for financial support (Grant reference RYC-201414984). The authors belong to the Galician Competitive Research Group GRC 2013-032 as well as to CRETUS (AGRUP2015/02), co-funded by Xunta de Galicia and FEDER. References ADA, 2009. Position of the American dietetic association: vegetarian diets. J. Am.Diet. Assoc. 109, 1266–1282 Aleksandrowicz, L., Green, R., Joy, E.J.M., Smith, P., Haines, A., 2016. The Impacts of Dietary Change on Greenhouse Gas Emissions, Land Use, Water Use, and Health: A Systematic Review. Plos One 11(11), e0165797. doi:10.1371/journal.pone.0165797 Amcoff, E., Edberg, A., Enghardt Barbieri, H., Lindroos, A.K., Nä lsén, C., Pearson, M., Warensjö Lemming, E., 2012. Livsmedelsoch nä ringsintag bland vuxna i Sverige. Riksmatenvuxna 2010–11. (The food and nutrient intake among adults in Sweden. Riksmaten-adults 2010–11).Swedish Food Agency, Uppsala, Sweden. Astrup, A., Andersen, N.L., Stender, S., Trolle, E., 2005. The Danish Dietary Recommendations 2005. Denmark: Ernæringsrådet og Danmarks Fødevareforskning; 2005 Baroni, L., Cenci, L., Tettamanti, M., Berati, M., 2007. Evaluating the environmental impact of various dietary patterns combined with different food production systems. Eur. J. Clin. Nutr. 61, 279–286. doi:10.1038/sj.ejcn.1602522 Berlin, J., Sund, V., 2010. Environmental Life Cycle Assessment (LCA) of Ready Meals, SIK-Report No. 804. The Swedish Institute for Food and Biotechnology AB Blauert, E., Holt, M.H., Mithril, C., Meyer, C., 2010. Basis of the New Nordic Diet. University of Copenhagen. http://foodoflife.ku.dk/opus/english/wp/nordic_diet/basis_nnd/ Calvo-Malvar, M. del M., Leis, R., Benítez-Estévez, A.J., Sánchez-Castro, J., Gude, F., 2016. A randomised, family-focused dietary intervention to evaluate the Atlantic diet: the GALIAT study protocol. BMC Public Health 16, 820. doi:10.1186/s12889-016-3441-y Carlsson-Kanyama, A., Gonzalez, A.D., 2009. Potential contributions of food consumption patterns to climate change. Am J Clin Nutr. 89, 1704S_9. 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. doi:10.1016/j.jclepro.2017.04.121 Dagevos, H., Voordouw, J., 2013. Sustainability and meat consumption: is reduction realistic?. Sustainability: Science, Practice & Policy 9 (2), 1-10. Dahlqvist, A., Höglund, M., 2009. Doktor Dahlqvists och Airams LCHF-kokbok. (Dr Dahlqvist’s and Airam’s LCHF-cookbook).Pagina Förlags AB/Optimal Förlag, ISBN: 978-917241-725-0. De Ruiter, H., Kastner, T., Nonhebel, S., 2014. European dietary patterns and their associated land use: Variation between and within countries. Food Pol. 44: 158-166. https://doi.org/10.1016/j.foodpol.2013.12.002 DGE, 2008. Vollwertig essen und trinken nach den 10 Regeln der DGE; aid infodienst und Deutsche Gesellschaft fu r Erna hrung DGE: Bonn, Germany. Drewnowski, A., 2009. Defining nutrient density: Development and validation of the nutrient rich foods index. J. Am. Coll. Nutr. 28, 421S–426S. doi:10.1080/07315724.2009.10718106 Duchin, F., 2005. Sustainable Consumption of Food - A Framework for Analyzing Scenarios about Changes in Diets. J. Ind. Ecol. 9, 99–114. doi:10.1162/1088198054084707 EFSA, 2009. Review of labelling reference intake values Scientific Opinion of the Panel on Dietetic Products , Nutrition and Allergies on a request from the Commission related to the review of labelling reference intake values for selected nutritional elements 1 Adop. EFSA, 2010. EFSA panel on dietetic products, nutrition, and allergies (NDA); scientific opinion on dietary reference values for carbohydrates and dietary fibre. EFSA J. 8, 1462 [77 pp.]. EFSA, 2012. EFSA panel on dietetic products, nutrition and allergies (NDA); scientific opinion on dietary. Reference values for protein. EFSA J. 10, 2557 [66 pp.]. ENAPREF, 2012. Perú: Consumo per cápita de los principales alimentos 2008-2009. Encuesta Nacional de Presupuestos Familiares (ENAPREF). Dirección Técnica de Demografía e Indicadores Sociales. Instituto Nacional de Estadística e Informática (INEI). [in Spanish] Esteve-Llorens, X., Darriba Ferradás, L.C., Moreira, M.T., Feijoo, G., González-García, S., 2018. Towards an environmentally sustainable Atlantic dietary pattern: life cycle carbon footprint and nutritional quality. Environmental Research. under review. Estruch, R., Martínez-González, M.A., Corella, D., Salas-Salvadó, J., Ruíz-Gutierrez, V., Covas, M.I., Flol, M., Gómez-Gracia, E., López-Sabater, M.C., Vinyoles, E., Arós, F., Conde, M., Lahoz, C., Lapetra, J., Sáez, G., Ros, E., 2006. Effects of a Mediterranean-Style Diet on Cardiovascular Risk Factors. A Randomized Trial. Ann. Intern. Med.145 (1), 1-11. FAO, 2010. Biodiversity and Sustainable Diets United Against Hunger. International Scientific Symposium, FAO Headquarters, Rome. FAO, 2014. Necesidades Nutricionales del Ser Humano, FAO. Friel, S., Dagour, A., Garnett, T., Lock, K., Chalabi, Z., Roberts, I., Butler, A., Butler, C.D., Waage, J., McMichael, A.J., Haines, A., 2009. Public health benefi ts of strategies to reduce greenhouse-gas emissions: food and agriculture. The Lancet 374 (9706): 2016-2025. doi:10.1016/S0140-6736(09)61753-0 Fulgoni, V.L., Keast, D.R., Drewnowski, A., 2009. Development and Validation of the Nutrient-Rich Foods Index: A Tool to Measure Nutritional Quality of Foods. J. Nutr. 139, 1549– 1554. doi:10.3945/jn.108.101360 Fundación Española de la Nutrición, 2004. Consumo de alimentos en Galicia. La dieta atlántica. Madrid. Fusi, A., González-García, S., Moreira, M.T., Fiala, M., Bacenetti, J., 2017. Rice fertilised with urban sewage sludge and possible mitigation strategies: an environmental assessment. J.Clea.Prod. 140, 914-923 Garnett, T., 2011. Where are the best opportunities for reducing greenhouse gas emissions in the food system (including the food chain)? Food Policy 37, 463–466. Guallar-Castillón, P., Oliveira, A., Lopes, C., López-García, E., Rodríguez-Artalejo, F., 2013. The Southern European Atlantic Diet is associated with lower concentrations of markers of coronary risk. Atherosclerosis 226, 502-509. Hallström, E.,Carlsson-Kanyama, A., Börjesson, P., 2015. Environmental impact of dietary change: a systematic review. J. Clean. Prod. 91, 1-11. https://doi.org/10.1016/j.jclepro.2014.12.008 Hawkesworth, S., Dangour, A.D., Johnston, D., Lock, K., Poole, N., Rushton, J., Uauy, R., Waage, J., 2010. Feeding the world healthily: the challenge of measuring the effects of agriculture on health. Phil. Trans. R. Soc. B 365: 3083–3097. doi:10.1098/rstb.2010.0122 Health Council, 2006. Guidelines for Good Nutrition 2006 (in Dutch). Gezondheidsraad, The Hague, Netherlands. Helakorpi, S., Patja, K., Prättälä, R., Aro, A., Uutela, A., 2003. Health Behaviour and Health among Finnish Adult Population, Spring 2003, National Public Health Institute, Helsinki Heller, M. C., Keoleian, G. A., 2003. Assessing the sustainability of the U.S. food system: A life cycle perspective. Agric. Syst. 76 (3), 1007−1041 Heller, M.C., Keoleian, G.A., Willett, W.C., 2013. Toward a life cycle-based, diet-level framework for food environmental impact and nutritional quality assessment: a critical review. Env. Sci Technol 47, 12632–12647. doi:10.1021/es4025113 Hoolohan, C., Berners-Lee, M., McKinstry-West, J., Hewitt, C.N., 2013. Mitigating the greenhouse gas emissions embodied in food through realistic consumer choices. Energy Policy 63, 1065-1074. http://dx.doi.org/10.1016/j.enpol.2013.09.046 ISO 14040, 2006. Environmental management — Life cycle assessment — Principles and framework, Iso 14040. Switzerland. doi:10.1136/bmj.332.7550.1107 Jungbluth, N., Tietje, O., Scholz, R.W., 2000. Food purchases: Impacts from the consumers’ point of view investigated with a modular LCA. Int. J. Life Cycle Assess. 5, 134– 142. doi:10.1007/BF02978609 Li, Q., Li, T., Liu, C., Chen, J., Zhang, R., Zhang, Z., Dai, T., McClements, D.J., 2016. Potential physicochemical basis of Mediterranean diet effect: Ability of emulsified olive oil to increase carotenoid bioaccessibility in raw and cooked tomatoes. Food Research International 89, 320–329