The scientific interest in food consumption has evolved in recent decades, given its social and human health implications, direct and indirect pressure on domestic and global environmental systems, and contribution to the wellbeing of individuals and societies (1, 2). A sustainable food system ensures food security and nutrition for all while considering present and future economic, social, and environmental implications (2, 3). Nevertheless, providing a growing global population with healthy and sustainable diets is an immediate challenge (1, 3).

The Mediterranean diet (MED) is well known for its health benefits and has been identified for its environmental benefits as well (4). The diet is characterized by a high intake of plant-based foods; moderate to high intake of fish; moderate to low consumption of poultry, meat, and dairy; high intake of monounsaturated fatty acid (mainly from olive oil); and a moderate amount of wine (1–2 portions per day). As such, the MED's use of natural resources and environmental footprint have been revealed to be low (4, 5). The EAT-Lancet Commission has defined a reference “planetary health diet” based on both sustainability and health. The diet outlines a combination of food groups and ranges of food intake that could optimize human health and the environment (6). In a previous study, the Sustainable and Healthy Diet (SHED) index was developed and validated against both the MED and EAT-Lancet reference diet, which are both considered healthy and sustainable dietary regimens (7).

However, as dietary patterns are shaped by combined social and environmental factors, there is a need to examine not only recommended diets, but also those that are actually practiced, which is critical for advancing healthy and sustainable diets. Few recent studies have examined actual consumption patterns of individuals, households, and societies, their socio-demographic drivers, and their environmental and health implications (4, 5). One of these studies was performed among Italian adults and suggested that the adoption of healthy dietary patterns involves less use of natural resources and greenhouse gas (GHG) emissions (4). Similar results were shown among graduate students in Spain (5).

This paper joins this emerging direction of research in analyzing actual consumption patterns and exploring the gap between recommended diets and actually practiced ones, as well as the implications for the environmental impact of consumption of such diets. To this end, the study explores the differences in the environmental footprint (land, water, and GHGs) of different consumed diets based on local life cycle assessment (LCA) analysis, including unique aspects of the food system in Israel. For this purpose, we created an integrated database within the Food Frequency Questionnaire (FFQ), which includes environmental coefficients for environmental footprints. We analyzed our results with respect to the SHED index (7), the MED (4), and the EAT-Lancet dietary patterns (6) and evaluated the contribution of specific food groups to the environmental footprints using population-based cross-sectional data.

The sample included 525 young adult participants aged 20–66 years. The study participants were educated (82% had an academic education) and physically active, and only 13% were smokers. Higher SHED scores were associated with older age, women, higher education, non-smokers, normal weight, flexitarians, and vegetarian/vegans (Table 1). The same trends were observed for MED and EAT-Lancet scores (not shown).

The mean footprint use of the total sample was 5.7 ± 3.8 m² for land, 422 ± 229 liters for water, and 2.84 ± 1.32 kg/CO₂ for GHGs. The food groups' contributions to the different environmental footprints differ across factors (Figure 1). The main contributor to water use was fruits (40%), followed by vegetables (12%) and dairy (11%). The main contributor to land use was meat (30% for beef and 14% for poultry), and the main contributor to GHG emissions was dairy products (26%), followed by meat (14%) and vegetables (14%).

Next, the environmental factors were calculated according to adherence to MED and EAT-Lancet dietary patterns and according to tertiles of the SHED scores (Figure 2). Higher adherence to MED was associated with lower land use (high vs. low adherence: 4.07 ± 2.63 vs. 6.6 ± 4.3 m², p < 0.001), GHG emissions (high vs. low adherence: 2.32 ± 0.9 vs. 3.21 ± 1.32 kg/CO₂, p < 0.001), and higher water use (high vs. low adherence: 560 ± 281 vs. 360 ± 203 liters, p < 0.001). The same trend was observed for EAT-Lancet adherence and SHED scores (all p < 0.001). Figure 3 shows the mean contribution to the environmental footprints by food groups for different dietary patterns. At each adherence level of the dietary scores, there was a similar ranking of the food groups to that in Figure 1. The main contributor to GHG emissions was dairy products, followed by meat. The main contributor to land use was meat, and the main contributor to water use was fruit intake at each adherence level.

To further explore the relationship between the food groups and the environmental coefficients, we calculated the average consumption of the food-group items for different tertiles of GHG emissions, land use, and water use. Table 2 shows the mean daily intake of the different food groups by tertiles of environmental footprints. The consumption of plant food groups (fruits, vegetables, nuts/seeds, legumes/soy, and grains/potatoes) did not differ for different land tertiles with the exception of plant oils. The consumption of animal-based food increased for different land tertiles, demonstrating high use of land resources with increased consumption.

The meat intake was 12 times higher in the upper tertile of land use than in the lowest tertile, while poultry intake was 4 times higher, and egg intake was 2.5 times higher. Regarding GHG emissions, only the consumption of nuts/seeds and legumes/soy did not differ between tertiles. In the case of water, only poultry and dairy did not differ between tertiles. These results indicate a narrow range of consumption and smaller influence on water use and GHG emissions.

The need to assess both human health and environmental footprint of diets has been widely acknowledged. Nevertheless, most studies have focused on theoretical models of dietary guidelines and their potential contributions to sustainability and human health rather than actual consumption (3). The analysis presented in this paper joins a limited number of studies that have identified this gap (4, 5, 24). By analyzing data from a diverse sample of 525 people in Israel, we evaluated real consumption data in regard to adherence to healthy and sustainable diets. These findings are the integrative product of both the studied population's consumption habits and the environmental factors of each studied commodity.

Our findings indicate that the main contributors to water use were fruits, vegetables, and dairy. The main contributors to land use were meat and poultry, and the main contributor to GHG emissions was dairy products. We found that the highest tertiles of adherence to the MED and the EAT-Lancet reference diet (6, 9) were associated with the lowest GHG emissions and land use. On the other hand, the highest tertiles of adherence to the MED and EAT-Lancet were associated with higher water use. To expand our view on sustainability beyond the environmental footprint, adjoining sociocultural, economic, and health aspects we used the SHED index (7). The need to develop methods to include all 4 dimensions of sustainability of the diet (in their case MED) were acknowledged in a recent review by Portugal et al. (24). The results of the SHED index analysis were similar to those obtained for the MED and EAT-Lancet dietary pattern.

The health value of the MED is well established. During the last 20 years MED was shown to benefit health and function, reducing mortality rates (9, 10, 25). The EAT-Lancet as a theoretical dietary pattern targets both health and the environment using evidence based data, indicating that healthy and sustainable diet is achievable (11, 26). In a review by Aleksandrowicz et al. (1), the authors calculated the potential shift in footprint values associated with different dietary patterns. They conclude that shifting from Western diets to more environmentally sustainable dietary patterns can reduce above 70% of GHG emissions and land use, and 50% of water use.

Our findings parallel those of prior studies. In studies from Italy and Spain that used real consumed diet, high adherence to MED pattern was associated with lower GHG emissions and land use (4, 5). The contribution of animal products (meat, poultry, dairy, egg, and fish) constituted the greatest contributor to GHG emissions (19, 21, 22, 27). Higher adherence to the MED or the EAT-Lancet recommended diet was associated with lower GHG emissions in other studies In Spain (the SUN cohort) better adherence to the Spanish MED was associated with decreased environmental pressures in all assessed dimensions including GHG, land and water (5). Likewise, a cross-sectional study among Italian adults showed that omnivorous dietary choices or low adherence to the MED correlated with higher GHG emissions, land, and water use (4, 25). Results in the same direction were found in a non-MED country. Data from the European Prospective Investigation into Cancer and Nutrition—Netherlands (EPIC-NL) (28) cohort showed that the WHO and Dutch dietary guidelines lower the risk of all-cause mortality and moderately lower the environmental impact, while the DASH diet (Dietary Approaches to Stop Hypertension diet), despite leading to similar health outcomes, was associated with higher GHG emissions due to high dairy product consumption in the Netherlands (29). Another study conducted on Italian adults showed that omnivorous dietary choices generated worse carbon, water, and ecological footprints than other plant-based diets, while no differences were found for the environmental impacts of ovo-lacto-vegetarians and vegans (28). Unlike the above studies, in our findings water footprint was higher in the third tertiles of MED, EAT-Lancet and SHED and was connected tofruit intake.

This unique aspect revealed in our analysis needs further discussion. Since fruits and vegetables are more dependent on irrigation than animal-based foods, reducing animal-based foods and increasing plant-based foods do not always correspond with lower water use, as shown by Harris et al. (30). Plant-based foods were major contributors to dietary blue-water footprints. Grosso et al. (4) also found that higher adherence to the MED was not linearly associated with lower water consumption. Higher fruit consumption was also associated with higher water footprints in the United States, particularly the blue-water footprint (17).

However, theoretical dietary models show different results indicating that shifting to a more plant-based diet would reduce the water footprint (31). Other studies also demonstrate that the contribution of fruits and vegetables does not exceed the meat contribution for both blue and green water (32). Indeed, we found that sustainable diets like MED or the EAT-Lancet reference diet are characterized by higher water consumption, but several considerations need to be taken into account. One is that most fruits and vegetables consumed in Israel are grown locally. It follows that given the climatic conditions, most are irrigated, so they would have relatively high rates of blue-water footprints.

Nevertheless, it is important to note that not all blue water is the same as most of the fruit-related water footprint relies on treated wastewater, which reduces environmental pressure (33). Other solutions such as the use of desalinated water and efficient and cost-effective irrigation techniques already exist in Israel, but our findings emphasize the importance of further development of water management, including advanced technologies, reducing water losses, and improving data quality and monitoring for water–food system linkages (32, 33).

Management of the local food systems can also reduce water consumption. There are substantial differences of 2–10-fold in water consumption between different fruit crops (34). Prioritizing certain types of crops that are less burdensome in terms of water requirements while considering their health benefits could be another future direction to increase the health and sustainability of food systems.

As for dairy intake in Israel, the mean intake in our study is 402 ± 385 g per day. This value is higher than the estimated intake in Europe and North America, where the average daily intake is 364.8 g per day according to the PURE study (35). Since most dairy products consumed in Israel are domestically produced, and the production system is based on non-grazing cows, production can occur in a relatively small area (19, 27). In addition, the productivity of Israeli dairy cows is very high, which reduces the footprint per unit of milk (20). Nevertheless, the high demand for dairy products identified in our analysis led to high rates of dairy-related footprints.

According to the EAT-Lancet commission (6) and in accord with the national dietary recommendations, the requirement for different food groups is calculated based on healthy dietary intake within global boundaries. For example, the reference intake is 29 g per day for poultry, 300 g per day for vegetables, 200 g per day for fruits, and 250 g per day for dairy. In our data (Table 2), the actual consumption in the lowest tetiles of land GHG and water was nearly similar to the EAT-Lancet recommended diet. The main difference was fruit consumption. Thus, a shift, toward less animal based and more plant-based diets, is beneficial for both health and the environment. The Isocaloric Substitution of Plant-Based and Animal-Based Protein was related with Aging-Related Health Outcomes (36). A recent paper by Eisen and Brown, (37) show that, following a phaseout of livestock production will independently provide persistent drops in atmospheric methane and nitrous oxide levels, and slower carbon dioxide accumulation. This reduction through the end of the century, have the same cumulative effect on the warming potential of the atmosphere as a 25 gigaton per year reduction in anthropogenic CO2 emissions. This level of reduction will provide half of the net emission reductions necessary to limit warming to 2C.

Based on our data, which originate from the FFQ results of 525 participants, it seems that there is no conflict between a healthy and sustainable diet, but there is a need to adjust and optimize dietary patterns in light of recommendations for various populations with different dietary needs. For example, Israel is characterized with mixed Jewish and non-Jewish population, locals and new and established immigrants; and a significant young population alongside a growing share of elderly population. It is important to note that the EAT-Lancet reference diet stems from a theoretical model for a healthy and sustainable diet, whereas our data represent actual dietary patterns of the Israeli population. The results may be a proof of concept that the EAT-Lancet reference diet is indeed feasible.

Our analysis of consumed diets revealed that animal protein is associated with the highest GHG emissions and land use, while fruits and vegetables were associated with the highest water consumption. Nevertheless, most of them are grown using treated wastewater, which reduces environmental pressure. The differences in water consumption for different fruit crops support the need to prioritize certain types of crops, which should be less burdensome in terms of water requirements while considering their health benefits. Given these findings, we suggest that adherence to MED and EAT-Lancet dietary patterns should be included in national dietary guidelines and encouraged for consumption by all. Furthermore, our data could be used as a database to create healthy and sustainable diet recommendations while adjusting for nutritional needs and health status, as well as maintaining diversity within dietary patterns.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the Ethics Committee of Tel-Hai College (#09/2017-2). The patients/participants provided their written informed consent to participate in this study.

ST planned and conducted the research and wrote the first draft of the manuscript. MK calculated the environmental footprints and co-authored the manuscript. DS created the combined database and co-authored the manuscript. KA analyzed the data and co-authored it. All authors contributed to the article and approved the submitted version.

This study was supported by Ben-Gurion University's internal fund for nutritional research, MIGAL - Galilee Research Institute grant, and Tel-Hai College research fund.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.