The authors have declared that no competing interests exist.
Food production is a major driver of greenhouse gas (GHG) emissions, water and land use, and dietary risk factors are contributors to non-communicable diseases. Shifts in dietary patterns can therefore potentially provide benefits for both the environment and health. However, there is uncertainty about the magnitude of these impacts, and the dietary changes necessary to achieve them. We systematically review the evidence on changes in GHG emissions, land use, and water use, from shifting current dietary intakes to environmentally sustainable dietary patterns. We find 14 common sustainable dietary patterns across reviewed studies, with reductions as high as 70–80% of GHG emissions and land use, and 50% of water use (with medians of about 20–30% for these indicators across all studies) possible by adopting sustainable dietary patterns. Reductions in environmental footprints were generally proportional to the magnitude of animal-based food restriction. Dietary shifts also yielded modest benefits in all-cause mortality risk. Our review reveals that environmental and health benefits are possible by shifting current Western diets to a variety of more sustainable dietary patterns.
There is an urgent need to curb the degradation of natural resources and to limit global warming to less than 2°C, while providing a nutritious diet to a growing and changing world population [
The Rockefeller Foundation-Lancet Commission on Planetary Health suggested that there is major potential for dietary changes to improve health and reduce the environmental impacts of food production [
However, widespread policy action is lacking on integrating environmental and nutritional priorities [
We systematically review the evidence of the impacts of adopting sustainable diets on GHG emissions, agricultural land requirement, and water use, and compare the environmental and health effects between various types of sustainable dietary patterns. Our analysis aims to substantially expand on two previous reviews [
We conducted a systematic review of studies measuring the environmental impacts of shifting current average dietary intake to a variety of proposed sustainable dietary patterns, and our review is current as of 10th June 2016. We followed PRISMA quality guidelines [
Inclusion criteria for studies were as follows: quantifying changes in GHG emissions, land use, or water use, between average population-level dietary intake and proposed sustainable dietary patterns; using dietary or consumer expenditure surveys, or food balance sheets to inform the baseline diets; and, using baseline dietary data from 1995 onwards. The three environmental indicators were selected based on an initial screening of available indicators in the literature. Studies were excluded if they evaluated the impacts of single food items or meals rather than dietary patterns, or used alternative diets targeting meat or dairy reduction without compensating for this decrease in energy intake with intake of other foods. Our literature search identified a related theme of research on carbon taxes, which have been proposed as a tool to reduce GHG emissions through influencing consumer food choice and therefore dietary patterns. We did not include these studies in our main analysis as the resulting diets did not fully align with the common dietary patterns found across all other retrieved studies. However, the discussion section summarises findings from the studies that investigated the effect of carbon taxes on dietary GHG emissions.
The following parameters were extracted from studies: country or region, year of baseline diet, methods and sources of environmental impact data, type of sustainable diet(s) measured, environmental impacts of baseline and sustainable diets, if GHG emissions included those from land use change, health impacts, degree of change for the sustainable diet (e.g., amount of meat reduction), whether sustainable dietary patterns were self-selected within studies (dietary patterns as eaten by study participants, as opposed to modelled or designed by study authors), and energy content of baseline and sustainable diets.
Average population-level intakes in the reviewed studies were taken as the baseline diet, with each comparison between a baseline diet and a given sustainable diet categorised as an individual scenario. In each scenario, differences in environmental impacts between baseline and sustainable diets were quantified as the relative differences in carbon dioxide-equivalent GHG emissions (kg CO2eq/capita/year, which is an adjusted indicator including CO2, N2O, and CH4), land use (m2/capita/year), and water use (L/capita/day). Where studies reported impacts in absolute amounts, we converted these to relative differences. Impacts were stratified by sustainable dietary pattern type, and by environmental indicator. Environmental impact data using life cycle analysis (LCA) often do not include measures of variance, and therefore the reviewed studies did not provide confidence intervals for environmental impacts. Impacts did also not include systemic environmental feedbacks. Differences in environmental impacts between diet types were assessed using medians, and visualised using box and whisker blots. We converted any health effects originally reported in absolute terms to relative changes, by using appropriate population totals from the Global Burden of Disease Study [
Study quality was assessed through three requirements: modelling the baseline diet on dietary intake surveys rather than food availability or expenditure; a description of the source and methods of the environmental impact data used; and that differences in the energy content of baseline and sustainable diets were within 5%. This latter cut-off was used as as some studies aimed for an isocaloric design between compared diets, but due to modeling logistics, some minor caloric differences remained. These quality measures were selected since food balance sheets or expenditure-based surveys may differentially under- or over-estimate consumption of certain food groups [
The review protocol, with additional information and specific search terms, is available in
A total of 210 scenarios were extracted from 63 studies. Of these, 204 scenarios were modelled on national-level diets in HICs, one on a city in a middle-income country, and five on global dietary patterns (
Sustainable diet type | Environmental impact | ||
---|---|---|---|
GHG emissions | Land use | Water use | |
Vegan | 14 | 6 | 1 |
Vegetarian | 20 | 7 | 9 |
Ruminants replaced by monogastric meat | 6 | 3 | 1 |
Ruminants replaced by monogastric + no dairy | 1 | - | - |
Meat partially replaced by plant-based food | 8 | 4 | - |
Meat partially replaced by dairy products | 3 | 1 | - |
Meat partially replaced by mixed food | 7 | 1 | - |
Meat + dairy partially replaced by plant-based food | 5 | 3 | 3 |
Balanced energy intake | 6 | 2 | 1 |
Healthy guidelines | 21 | 10 | 9 |
Healthy guidelines + further optimisation | 16 | 5 | 4 |
Mediterranean | 8 | 5 | 4 |
New Nordic Diet | 3 | 1 | - |
Pescatarian | 6 | 4 | 2 |
124 | 52 | 34 |
Of the 210 scenarios, 197 showed a reduction in environmental impacts when switching from baseline to alternative dietary patterns (sign test: p<0·0001), while thirteen scenarios showed an increase or no impact. The median changes in GHG emissions, land use, and water use, across all sustainable diet types, were -22%, -28%, and -18%, respectively. The largest environmental benefits across indicators were seen in those diets which most reduced the amount of animal-based foods, such as vegan (first place in terms of benefits for two environmental indicators), vegetarian (first place for one indicator), and pescatarian (second and third place for two indicators).
The ranking of sustainable diet types showed similar trends for land use and GHG emissions, with vegan diets having the greatest median reductions for both indicators (-45% and -51%, respectively), and scenarios of balanced energy intake or meat partly replaced with dairy, having the least benefit. Although the water use scenarios had smaller sample sizes, they showed somewhat similar trends across sustainable diet types, with vegetarian diets having the largest benefit (median -37%), though with the notable exception of the single vegan scenario showing an increase in water use (+107%) (Figs
Note: n = number of studies, mdn = median.
Note: n = number of studies, mdn = median.
Note: n = number of studies, mdn = median. The lower and upper bounds of the boxes represent the 1st and 3rd quartiles, respectively, and the line within is the median. Whiskers show the minimum and maximum range, excluding outliers, which are shown as dots, and represent values more than 1.5 times the 1st and 3rd quartiles.
We assessed the sensitivity of our findings to study quality. Excluding papers that did not meet the three quality criteria resulted in minor differences in findings. The overall direction of impact did not change (sign test: p = 0·5), and the ranking of sustainable diet types had strong correlation with the full list of studies for GHG emissions and land use (Spearman’s rho: 0·93, p<0·0001; 0·83, p = 0·003, respectively). The correlation between rankings was not significant for water use (Spearman’s rho: 0·20, p = 0·8); this was likely due to the number of scenarios decreasing from 34 to 4 when removing lower-quality studies (
Analyses of the health effects of sustainable diets were limited. Within the seven studies reporting health effects of adopting sustainable diets, 11 out of the 14 sustainable diet types were modelled, with a single estimate of all-cause health impacts for all but two of the 11 diet types. Most studies assessed the reduction in mortality risk from adopting a sustainable diet, either by all-cause or cause-specific mortality (
Study | Country | Sustainable diet type | Health indicator | Change in health indicator (95%CI) |
|
---|---|---|---|---|---|
Sabate 2015 | US/Canada | Vegan | All-cause mortality rate | 19.2% | |
Soret 2014 | US/Canada | Vegetarian | All-cause mortality risk | 9% (0–17) | |
Tilman 2014 | Globally | Vegetarian | All-cause mortality risk | <1% (0–2) |
|
Sabate 2015 | US | Vegetarian | All-cause mortality rate | 15.9% | |
Aston 2012 | UK | Meat partially replaced by mixed food | CHD risk (men) | 9.7% (-3.5–22) | |
Aston 2012 | UK | Meat partially replaced by mixed food | CHD risk (women) | 6.4% (-1.8–14.3) | |
Aston 2012 | UK | Meat partially replaced by mixed food | Diabetes mellitus risk (men) | 12% (-4.5–22.7) | |
Aston 2012 | UK | Meat partially replaced by mixed food | Diabetes mellitus risk (women) | 7.5% (0.5–14.5) | |
Aston 2012 | UK | Meat partially replaced by mixed food | Colorectal cancer risk (men) | 12.2% (6.4–18.0) | |
Aston 2012 | UK | Meat partially replaced by mixed food | Colorectal cancer risk (women) | 7.7% (4.0–11.3) | |
Soret 2014 | US/Canada | Meat partially replaced by mixed food | All-cause mortality risk | 14% (4–23) | |
Sabate 2015 | US/Canada | Meat partially replaced by mixed food | All-cause mortality rate | 7.2% | |
Biesbroek 2014 | Netherlands | Meat partially replaced by plant-based food | All-cause mortality risk | 10% (3–16) | |
Biesbroek 2014 | Netherlands | Meat partially replaced by dairy | All-cause mortality risk | 6% (-4-14) | |
Tilman 2014 | Globally | Mediterranean | All-cause mortality risk | 18% (17–19) | |
Sabate 2015 | US/Canada | Pescatarian | All-cause mortality rate | 17.6% | |
Milner 2015 | UK | Healthy guidelines | Years of life lost |
6% | |
Milner 2015 | UK | Healthy guidelines + further optimisation | Years of life lost |
7% | |
Scarborough 2012 | UK | Meat, dairy partially replaced by plant-based food | Deaths averted | 6% | |
Scarborough 2012 | UK | Ruminants replaced by monogastric | Deaths averted | <1% |
*Percentages refer to reductions in health indicators, except for deaths averted
**Mortality risk reduction by cause: cancer 10%, coronary heart disease 20%, type 2 diabetes 42%
+Years of life lost, at year 30 (after adoption of the sustainable diet scenario)
Our review showed that reductions above 70% of GHG emissions and land use, and 50% of water use, could be achieved by shifting typical Western diets to more environmentally sustainable dietary patterns. Medians of these impacts across all studies suggest possible reductions of between 20–30%. This review is the most recent and comprehensive to date, and the first to compare impacts across GHG emissions, land use, and water use. This work supports the conclusions of previous reviews in this area[
Underlying environmental data in the studies (where shown) on the land use, GHG emissions, and water use impacts from the production of food items showed decreasing impacts, from greatest to least, across ruminant meat, other meat, dairy, and plant-based foods [
Studies modelling the health impacts of shifts from typical Western diets to sustainable dietary patterns showed modest health gains from reductions in mortality rates and risks [
This review had several limitations. The available studies were from a narrow range of HICs with different baseline dietary patterns, and used largely HIC-specific environmental data sources. The results may therefore only be generalizable to HICs. The data on environmental impacts did not provide measures of variance, and we were limited to graphical and non-parametric statistical methods to assess the differences between sustainable dietary patterns. We were also unable to rule out any effects of publication bias in the literature. The use of environmental indicators varied across studies, such as whether blue, green or grey water (or a combination) was used, and whether GHG emissions included the often significant emissions from land use change. Our use of relative differences in the analysis helped to accommodate some of the differences in methodology across studies, and despite this heterogeneity, our resulting median impacts produced internally consistent and plausible trends; for example, vegan diets having greater reductions in GHG emissions than vegetarian; greater benefits from reducing meat and dairy consumption compared to meat alone; and replacing meat with dairy having little benefit.
There is an increasing body of evidence on which to base the integration of environmental priorities into dietary recommendations. Several of these dietary patterns are already promoted through public health efforts, such as the healthy dietary guidelines, the Mediterranean diet [
Several considerations regarding environmentally sustainable eating are worth noting. Firstly, the production of food (i.e. the growing of crops and raising of livestock) is the primary driver of environmental impacts, as opposed to later stages such as transport and processing [
However, complete removal of animal-source foods is not realistic in many cultures and may have important health implications. Meat and dairy are high-quality sources of protein and micronutrients, and ensuring adequate bioavailable supply of these is essential for public health [
Lastly, shifts to sustainable diets must be affordable and desirable for consumers. Studies have shown that large reductions in GHG emissions are possible without complete exclusion of animal products [
Our estimates would benefit greatly from more comprehensive data, and further work should generate regional and food-specific environmental impacts, including for fisheries and aquaculture, as well as measures of variance. A limited number of studies calculated a reduction in nitrogen and phosphorus water contamination from sustainable eating patterns [
The impacts of sustainable diets are linked to a number of SDGs, including goals on sustainable agricultural practices, health, water use, and climate change. Promotion and uptake of these diets could therefore offer a route, along with other strategies, to achieving several of the SDGs.
Across a large and heterogeneous set of studies, several policy implications are clear: environmental benefits are possible in HICs from shifting current diets to a variety of more sustainable dietary patterns; environmental benefits are largely proportional to the magnitude of meat (particularly from ruminants) and dairy reduction; and a redoubling of efforts to promote the uptake of diets that support these changes could bring environmental and health benefits.
(DOCX)
(XLSX)
LA, RG and AH designed the study protocol. LA analysed the data and drafted the paper. LA and EJ reviewed the literature. All authors were involved in data interpretation, critical revisions of the paper, and approved the final version.