The environmental and health impacts of diets and dietary change in 5,500 cities worldwide

The environmental and health impacts of diets and dietary change in 5,500 cities worldwide | Research Square Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Help Center Sign In Submit a Preprint Cite Share Download PDF Social Sciences - Article The environmental and health impacts of diets and dietary change in 5,500 cities worldwide Marco Springmann, Jack Rawden, Olivia Auclair, Sebastiano Caleffi This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8483305/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract The majority of the world’s population live in cities, making food environments in cities an important driver of global diets and the associated health and environmental impacts. Based on dietary data for urban residences and demographic information, we estimated the dietary intake in 5,500 cities with populations over 100 thousand inhabitants. Our estimates indicate that diets in most cities contained greater amounts of foods compared to a country’s average intake in 2020. As a result, cities in most regions were responsible for a larger share of food-related environmental resource use and pollution as their share of population, something that in our impact assessment was mostly driven by increased intake of animal source foods in cities. Cities were also responsible for a large share of diet-related health burden and an outsized share of health-related costs, in line with the generally higher cost levels in cities. Dietary changes to healthier and more sustainable diets could substantially reduce the environmental, health, and cost impacts associated with city diets, but are dependent on consistent policy approaches and support. Earth and environmental sciences/Environmental social sciences/Sustainability Earth and environmental sciences/Environmental social sciences/Environmental impact Earth and environmental sciences/Environmental social sciences/Socioeconomic scenarios Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Most diets are neither healthy nor sustainable 1,2 . Unhealthy diets that are low in fruits, vegetables, legumes, nuts, and whole grains, and high in red and processed meat and sugary drinks are responsible for one in four deaths globally 3 . The cost related to treating diet-related diseases is projected to exceed USD 1 trillion in 2030 4,5 . The environmental impacts of diets are similarly daunting, with about a third of all greenhouse gas (GHG) emissions attributed to food production 6 , as well as the majority of land use, biodiversity loss, water use and pollution 7,8 . Without dietary changes towards healthier and more sustainable diets, several planetary boundaries related to the safe operating space for humanity are likely to be crossed 9 . Cities play a key role for food systems and their impacts. More than half of the world’s population, about 4 billion people, live in cities – a proportion that is projected to increase to more than two thirds or 7 billion by 2050 10 . As a result, a substantial portion of diet-related ill-health and environmental impacts are generated by dietary intake in cities. Populations in cities often consume more food, including high-impact foods, which further exacerbate these impacts 11,12 . Cities form the immediate food environments of their inhabitants, influencing dietary behaviour through physical and social aspects related to local food availability, accessibility, and affordability 13,14 . Cities therefore have an important role in enabling and supporting dietary and food-system transformations 15,16 , something that is increasingly recognised in and endorsed by political decision-making 17,18 . In contrast to the large number of cities, only a few dietary and food-system assessments have been conducted at the city level, often with important shortcomings which limit consistent policy planning. The studies range from case studies of specific cities 19,20 to comparative assessments of selected and often larger cities 21 , leaving many smaller cities without estimates. Many of the studies covered one or several environmental indicators, such as GHG emissions 20 and ecological resource demand 19 , but few included health-related assessments 21 . Many followed an approach based on tracking a city’s flows of inputs and outputs, but their concrete implementations differed, which complicates comparisons 22 . Lastly, most studies scaled national estimates of food intake – either from food balances or surveys – to city populations, which does not account for the particular conditions of urban food environments and diets 23 . Here we address these research gaps and present findings from a comprehensive health and environmental impact assessment of diets in about 5,500 cities worldwide, covering urban agglomerations with more than 100,000 inhabitants. About 90% of the included cities are smaller than one million, and 85% are in low and middle-income countries, which substantially expands the evidence base for city-level decision-making. For each city, we estimated food-related impacts in terms of a comprehensive set of indicators, following a consistent methodology adapted to the city level ( Methods ). The environmental assessment was based on environmental footprints from life-cycle assessments covering food-related GHG emissions, land use, water use, and water pollution measured in terms of eutrophication potential 7 . The health assessment was based on a comparative risk assessment of diet and weight-related risks, their relationship with the incidence and mortality from non-communicable diseases 24,25 , and a health-related cost assessment based on the costs of illness in cities 5,26 . As basis for our assessments, we developed new estimates of dietary intake in cities. For this, we combined dietary data that for each country were differentiated by urban residence and demographic (age and sex) groups 27 with the demographic composition of cities in each country 28 . This method of estimating city-level food intake substantially improves current population-scaled estimates by accounting for differences in food intake in urban populations in each country, and for the demographic composition of each city. Whilst we do not have data on what each inhabitant of a specific city eats, we know the proportion of female and male children, adolescents, young adults, middle-aged adults, and senior adults in that city, and what those groups eat when they live in an urban environment in a specific country. Because diets differ in urban compared to rural environments and even more between demographic groups 29 , our estimates display variability across cities in the same country, in addition to capturing regional differences. To inform city-level policy planning, we estimated the health and environmental impacts both of current diets and of dietary changes towards healthy and sustainable diets in each city. For the latter, we adopted the dietary recommendations of the EAT-Lancet Commission on Healthy Diets from Sustainable and Just Food Systems for each city 1,2 . Our scenarios of healthy and sustainable diets account for national and demographic preferences for foods within general food groups (e.g., wheat, rice, and maize within grains; and beef, lamb, and pork within red meat), and for healthy intake levels for general food groups in each demographic group and country, which similarly to the estimates of intake 30,31 , we aggregated to average diets of specific cities within each country. We developed an interactive dashboard to share the full set of results and allow for custom analyses at global, regional, national, and city-specific levels 32 . Results Our analysis estimated the dietary intake and related environmental and health impacts of 5,489 cities worldwide (Figure 1). About 60% of the cities had populations between 0.1 and 0.3 million people, 30% between 0.3 and 1 million, 9% between 1 and 5 million, and 1% each between 5 and 10 million, and above 10 million. Across regions, about a third (31%) of cities were in East Asia, a fifth (22%) in South Asia, 10-15% each in Europe and Central Asia, Sub-Saharan Africa, Latin America and the Caribbean, and the Middle East and North Africa, and 4% in North America. The interactive dashboard 32 provides a full list of all cities included in the analysis. In 2020, about 42% of the global population lived in the cities included in our analysis (SI Table 3). Across regions, 66% of the population in North America lived in cities, 57% in Latin America and the Caribbean, 52% in the Middle East and Northern Africa, 48% in East Asia, 42% in Europe and Central Asia, 32% in Southern Asia, and 30% in Sub-Saharan Africa. The cities with the highest population in our analysis were Tokyo (Japan) with about 37 million people, Delhi (India) with 30 million, Shanghai (China) with 27 million, Sao Paolo (Brazil) with 22 million, and Cairo (Egypt) with 22 million (SI Table 4). Dietary intake Urban diets showed large regional variability (Figure 2, SI Table 5). In 2020, energy intake in cities ranged from 2150 kcal/d in South Asian cities to 2485 kcal/d in North American cities, in line with the distribution of weight levels. The intake of fruits and vegetables ranged from 210 g/d in Sub-Saharan African cities to 630 g/d in East Asian cities, and that of legumes and nuts from 25 g/d in European and Central Asian cities to 50 g/d in Sub-Saharan African cities. Sugar intake ranged from 30 g/d in East Asian cities to 90 g/d in North American cities, and oil intake from 20 g/d in South Asian cities to 50 g/d in North American cities. Lastly, meat intake ranged from 25 g/d in South Asian cities to 210 g/d in North American cities, and milk intake from 65 g/d in Sub-Saharan African cities to 590 g/d in European and Central Asian cities. Compared to the average food intake in a country, diets in cities contained greater amounts of most foods, except for grains and legumes (Figure 2, SI Table 5). This included more foods considered healthy such as fruits (+6%), and nuts and seeds (+12%), but also foods that are less healthy or have high environmental impacts such as sugar (+20%), meat (+3-4%) and dairy (+6%). Across regions and compared to the national average intake, calorie intake was greater in Latin America cities (+26 kcal/d) due to more sugar, oils, and animal source foods; in European and Central Asian cities (+11 kcal/d) due to more dairy and oils; in Middle Eastern cities (+43 kcal/d) due to more grains, fruits, and animal source foods; in South Asian cities (+65 kcal/d) due more sugar and dairy; in East Asian cities (+48 kcal/d) due to more grains, dairy, and pork; and in Sub-Saharan African cities (+110 kcal/d) due to more roots and sugar. In contrast, calorie intake was lower in North American cities (-8 kcal/d) due to lower intakes of grains, nuts, and fruits. Intake also differed across cities in the same country, in line with demographic differences (Figure 3, SI Figure 2). For example, in a large country like China, total calorie intake ranged from 2190 kcal/d in Lingshan whose inhabitants included relatively large shares of children and adolescents who, on average, consume less food than the general population in a Chinese city, to 2475 kcal/d in Beitang whose inhabitants included a relatively large share of young adults who, on average, consume more food (SI Table 13). Across countries, the range between maximum and minimum intake was largest in Ethiopia (range: 325 kcal/d, number of cities: n=55), Russia (285 kcal/d, n=137), and China (285 kcal/d, n=1117) where demographic differences across cities were large, and it was lowest in Costa Rica (3 kcal/d, n=4), Croatia (4 kcal/d, n=3), and Sweden (4 kcal/d, n=6) where cities were fewer and had relatively similar demographic compositions. Environmental impacts Globally, food demand in cities was responsible for a larger share of environmental resource use and pollution (47% on average) than their share of population (42%). This included 7.7 GtCO 2 -eq in food-related greenhouse gas (GHG) emissions (48% of the total), 26 Mkm 2 in land use (49%), 1,300 km 3 in water use (43%), and 34 MtPO 3 4- eq in eutrophication potential (47%) (SI Table 14). Animal source foods were responsible for the majority of GHG emissions (59%), land use (85%), and eutrophication potential (54%), whereas crops were responsible for most of the water use (74%) (Figure 4, SI Figure 3). The importance of cities in driving a country’s food-related resource demand differed by region (SI Table 14). Cities accounted for 66% of national impacts in North America, 60% in Latin America and the Caribbean, 43% in Europe and Central Asia, 55% in the Middle East and Northern Africa, 50% in East Asia, and 32% each in South Asia and Sub-Saharan Africa. The shares of environmental impacts exceeded the cities’ population shares in all regions, except for cities in North America, which was in line with the relative differences in the intake of animal source foods. City-level impacts can be compared based on absolute (Figure 5) and per-person (Figure 3, SI Table 15) contributions. The cities with the greatest absolute contributions to food-related environmental resource use (measured in thousands of percent or per cent mille, pcm) were Sao Paulo (Brazil, 610 pcm), Buenos Aires (Argentina, 580 pcm), Tokyo (Japan, 450 pcm), Shanghai (China, 430 pcm), and Rio de Janeiro (Brazil, 370 pcm). The ordering reflects not only the cities’ population shares, but also their dietary composition. For example, Tokyo had more than double the number of inhabitants as Buenos Aires (+22 million), but its inhabitants consumed less than a quarter the amount of beef per person (Figure 5). The cities with the greatest impacts per person (measured by the deviation from average per-person impacts across all cities) were Ulaanbaatar (Mongolia, +1280%), N’Djamena (Chad, +400%), Faya Largeau (Chad, +365%), Sarh (Chad, +365%), and Koumra (Chad, +365%), all of which had very high intake of beef, lamb, and dairy (Figure 3, SI Table 15). The cities with the lowest per-capita impacts were Kayanza (Burundi, -65%), Ngozi (Burundi, -65%), Gitega (Burundi, -65%), Bujumbura (Burundi, -65%), and Demba (Democratic Republic of the Congo, -60%), all of which had very low intake of animal source foods. Within countries, the differences in per-person impacts across all cities ranged from -11% in Kefole (Ethiopia), due a large share of low-consuming children, to +6% in Beihai (China), due to a large share of high-consuming young adults (SI Table 16). Health impacts Globally, about 5.6 million deaths were attributable to diets in cities in 2020 (44% of all diet-related deaths) (SI Table 17). Two thirds were from cardiovascular diseases (3.8 million), including coronary heart disease (43%) and stroke (24%), a fifth (1.2 million) from cancer, and lower amounts from type-2 diabetes (7%) and respiratory disease (4%). By risk factor (Figure 4), about two thirds were attributable to risks related to dietary composition (3.7 million), including low intake of whole grains (24%), fruits (10%), vegetables (7%), legumes (9%), nuts and seeds (3%), and high intake of red meat (8%) and processed meat (5%). The remaining third (2 million) was attributable to risks related to the overall amount of food intake that is associated with being underweight (4%), overweight (8%), and obese (23%). The rate of diet-related deaths in cities varied by region (Figure 4, SI Figure 4). It ranged from 610 diet-related deaths per million people in Sub-Saharan African cities to 3,100 per million in European and Central Asian cities, with differences driven by the demographic composition of cities (which tend to be younger in lower-income regions), mortality rates for non-communicable disease (which increase later in life), and risk factors (which are relatively more present in higher-income countries). The cities with the greatest diet-related mortality rates were several cities in Ukraine (e.g., Chernihiv, Cherkasy, and Kropyvnytskyi) with rates of 7,100-7,700 deaths per million people, whereas the cities with the lowest rates were cities in Mali (e.g., Senou), Uganda (e.g., Lira), and Ethiopia (e.g., Areka) with rates of 210-240 deaths per million (Figure 3, SI Table 18). The cities with the greatest absolute number of diet-related deaths (Figure 5) were Tokyo (Japan; 83,000), followed by Shanghai (China; 61,000), Moscow (Russia; 60,000), Cairo (Egypt; 49,000), and New York (USA; 48,000). The ordering reflects the cities’ population shares, their dietary composition that is associated with dietary risks, as well as demographic aspects. For example, Shanghai had more than double the number of inhabitants as Moscow (+14 million), but a similar number of diet-related deaths, also because its inhabitants consumed less unhealthy foods (e.g., a third less red and processed meat) and more health-promoting foods (e.g., three times more vegetables and six times more nuts). Cost impacts The diet-related deaths in cities incurred substantial healthcare-related costs, which due to increased income and cost levels in cities exceeded the cities’ population shares (SI Figure 5). Across all cities, the costs of illness amounted to $720 billion, representing 51% of all healthcare-related costs (SI Figure 6). Of those, close to half ($330 million) were from direct healthcare costs, and more than half ($380 million) from indirect costs associated with lost productivity and unpaid care by family members. The costs per person ranged from $30 per person (pp) in cities in South Asia to $870 pp in cities in North America (Figure 4). The per-person costs were particularly high in cities in Belarus (e.g., Zhodzina, Minsk, Barysaw) and the USA (e.g., San Rafael, Sarasota Brandenton, Stuart), which was due to high disease cases in the former, and high cost levels in the latter (Figure 3, SI Table 19). The per-person costs were lowest in cities in Niger (e.g., Gaya, Magaria, Zinder), Mali (e.g., Tahoua, Segou), and the Democratic Republic of the Congo (e.g., Lubao) due to low diet-related mortality rates and cost levels (Figure 3, SI Table 19). The greatest total costs (Figure 5) occurred in New York (USA, $19 billion), Tokyo (Japan, $14 billion), Moscow (Russia, $12 billion), Los Angeles (USA, $11 billion), and Chicago (USA, $8 billion), reflecting high cost levels in the US cities, the high diet-related disease cases in Moscow, and the high population share of Tokyo. Impacts of dietary changes Diets in most cities differed substantially from planetary health diets that are considered healthy and sustainable (Figure 2). Across all regions, diets in cities contained too few whole grains (295-625 kcal/d), legumes (150-210 kcal/d), healthy oils (25-380 kcal/d), fruits and vegetables (40-90 kcal/d), and nuts and seeds (115-160 kcal/d), and too much processed grains (345-1095 kcal/d), sugar (40-195 kcal/d), animal source foods (50-475 kcal/d), and starchy roots (5-480 kcal/d), but to different degrees. While all cities exceeded healthy limits for processed grains, those in Sub-Saharan Africa consumed particularly high amounts of roots, those in South Asia, North America, and Latin America high amounts of sugar, those in Europe and Central Asia and North America high amounts of dairy, and those North America, Latin America, Europe and Central Asia, and East Asia high amounts of meat. Dietary changes to planetary health diets reduced the average environmental resource use of cities by 37% (Figure 6, SI Table 20). The reductions included 2.7 GtCO 2 -eq in GHG emissions (-36%), 13 Mkm 2 in land use (-50%), 350 km 3 in water use (-26%), and 13 MtPO 3 4- eq in eutrophication potential (-38%). Across regions, the average city-level reductions ranged from 11% in cities in South Asia to 51% in Latin America, which was in line with the degree of changes in the consumption of animal source foods. In comparison, halving food loss and waste reduced the average food-related resource use by 15%, ranging from 10% in cities in Sub-Saharan Africa to 20% in Northern American cities (SI Figure 7), most of which was determined by the loss and waste of animal source foods. Dietary changes to planetary health diets reduced the number of diet-related deaths by 4.7 million (-84%) (Figure 6, SI Table 20). Across regions, the reductions in the diet-related disease burden ranged from 540 deaths per million (-84%) in Sub-Saharan African cities to 2,500 deaths per million people (-83%) in European and Central Asian cities, in line with the region’s relative disease burden. The health-related costs were reduced by $190 pp (-84%) and $600 billion in total. The reductions ranged from $20 pp (-73%) in South Asia to $750 pp (-86%) in Northern America. The cities with the greatest absolute reductions in diet-related deaths (see, e.g., the interactive dashboard 32 ) were Tokyo (Japan; 69,000), Shanghai (China; 51,000), Moscow (Russia; 47,000), Cairo (Egypt; 44,000), and New York (USA; 41,000). Discussion The majority of the world’s population live in cities, making food environments in cities an important driver of global diets and the associated health and environmental impacts. Based on dietary data for urban residences and demographic information, we estimated the dietary intake in 5,500 cities with populations over 100 thousand inhabitants. Our estimates indicate that diets in most cities contained greater amounts of foods compared to a country’s average intake in 2020. As a result, cities in most regions were responsible for a larger share of food-related environmental resource use and pollution as their share of population, something that in our impact assessment was mostly driven by increased intake of animal source foods in cities. Cities were also responsible for a large share of diet-related health burden and an outsized share of health-related costs, in line with the generally higher cost levels in cities. Dietary changes to healthier and more sustainable diets could substantially reduce the environmental, health, and cost impacts associated with city diets. Several policy options exist for city-level governments to improve diets and food systems 16,33,34 . For example, the Milan Urban Food Policy Pact (MUFPP), launched in 2015, lists 37 actions which have been mapped to 44 indicators 17,35 . Whilst comprehensive in scope, content analysis of city-level food policies amongst MUFPP signatories has shown that the most promoted practices related to where to source food 36,37 . In contrast, our results suggest practices related to what to eat would be associated with substantial benefits that warrant increased policy attention. Such practices could include school feeding programmes, public procurement standards, food and nutrition regulation, fiscal incentives and disincentives, and planning and zoning approaches 16 , with some supported by city-wide or topical initiatives such as the C40 Good Food Cities Accelerator 18 or the School Meal Coalition 38 . Our study can inform policy planning by providing estimates of current diets, of the potential impacts of dietary change, and of examples of diets and impacts across cities in different world regions and countries 32 . Our estimates advance the current literature on city-level assessments in several ways. First, our study extends the coverage of city-level assessments from single case studies 19,20 and selected groups of cities 21 to most cities above 100 thousand inhabitants, and our analytical framework is general enough to be applied to urban agglomerations of other sizes. Second, our comparative assessment of environmental, health, and cost indicators extends the more specialised foci, e.g., on GHG emissions or ecological footprints, of many current studies 19,20 , and provides a comprehensive basis for cross-municipal “whole of food” policy planning. Third, we improve existing estimates of dietary intake at the city level, most of which represented population-scaled national data 22,23 , by accounting for differences in dietary intake across demographic groups in urban residences and the demographic composition of cities, something that we show introduces substantial variation in city-level diets and impacts. Our study is also subject to caveats, many of which are related to data availability ( Methods ). For example, although our estimates of food intake in cities improve current, population-scaled estimates, they do not account for differences in city-level intake between the same demographic group in the same country. This meant that the differences between cities of the same country were solely driven by differences in the cities’ demographic (age and sex) composition. More detailed assessments would require a more comprehensive monitoring of local food systems. For environmental assessments, this could include compiling more detailed accounts of food-related inputs and outputs to and from cities, e.g., by recording freight and transport statistics, and improving the collection and sharing of data from food processors, whole-sellers, retailers, and waste collectors 19,23 . For health-related assessments, it would include local dietary surveys, preferably conducted with objective measures and in combination with collecting anthropometric measures (e.g., weight and height) 29,39 for independent verification and sensitivity checks. Lastly, close collaboration between municipal and national governments and with local and international organisation and initiatives could strengthen data exchange and comparability 40,41 . Methods We followed several steps to assess the environmental, health, and health-related cost implications of diets and dietary change at the city level. They included estimating dietary intake at the city level, combining the estimated intake with environmental footprints, and using the estimated intake in a comparative risk assessment of diet-related diseases and the costs of illness. Estimating dietary intake in cities For estimating dietary intake at the city level, we used a new proxy of dietary intake that is specified by country, age group, sex, and urban/rural residence 27 together with detailed data on the age and sex structure of cities 28,42,43 . The dietary proxy was derived by combining (i) estimates of the dietary composition of complete diets from food-availability statistics 44 that were adjusted for food waste at the retail and household levels 45 with (ii) harmonised data from dietary surveys on the variation of food intake per general food groups across sociodemographic groups in each country 46 , and (iii) estimates of overall energy intake by population group from estimated energy requirements to sustain measured levels of body weight, height, and physical activity 29 . SI Tables 1-2 provide overviews of the dietary proxy used for the analysis. For deriving the age and sex composition of cities, we used a globally gridded dataset of population structures differentiated by age and sex 28,42 , and overlaid this with geographic information on the location of urban centres 43 . The data on population structures were from WorldPop 42 and included estimates of the age and sex distribution within 100 x 100 m cells across the world. The data on the location of urban centres were from the Global Urban Polygons and Points Dataset (GUPPD) 43 from which we used the information on agglomerations with more than 100,000 people. We used Google Earth Engine to process the two datasets. We harmonised the estimated total population and city names of each agglomeration to those of the World Urbanization Prospects (WUP) 10 where possible (for agglomerations over 300,000), and otherwise used population estimates from the GUPPD. SI Tables 3-4 provide an overview of the coverage of our estimates. For estimating the average food intake in a city, we combined (i.e., population-weighted) the estimates of food intake for urban residences differentiated by age and sex with the population estimates differentiated by age and sex for each urban agglomeration. This resulted in a dataset of food intake in 5,489 urban agglomerations, each covering 28 food groups. SI Table 5 provides an overview of the estimates of food intake in cities. For identification (see, e.g., SI Table 4), we used the names assigned in the GUPPD 43 , which sometimes summarised several cities within the same urban agglomeration (e.g., NewarkNewYork) or expressed city names in several languages (e.g., MoskvaMoscow). When no alphanumerical name was provided, we coded the agglomeration as NAN and appended to this the country code and population ranking within the country. We followed the same approach when a city occurred more than once (e.g., London in the UK and Canada) to retain a unique city identification. Scenarios of dietary change In addition to estimating current dietary intake in cities, we also constructed scenarios of healthier and more sustainable dietary intake. Our diet scenarios were based on the dietary recommendations of the EAT-Lancet Commissions on Healthy Diets from Sustainable and Just Food Systems 1,2 (SI Table 6). We followed several steps to adapt the general recommendations for each city 30 . First, we adapted the dietary reference values developed for adults to each country’s urban population by implementing them as ceilings to the estimates of current intake for discouraged food groups (e.g., red meat and sugar) and as floors for encouraged food groups (e.g., fruits and vegetables), without penalising greater than recommended intake of encouraged foods or lower than recommended intake for discouraged foods. Because the dietary reference values are expressed for general food groups, we were able to preserve regional preferences within those, e.g., of the types of grains, red meat, fish, and fruits. Second, we constructed nutritionally complete diets for each demographic group amongst the urban population, building on previous analyses 30 . For that, we first balanced energy intake amongst adults by adjusting grain intake and, if grain intake reached the recommended bounds, adjusting oil intake within recommended bounds to fulfil the recommended energy intake of urban adults. We then scaled overall food intake to fulfil the recommended energy intake across all other demographic groups. We ensured that the diets were nutrient balanced by adjusting the intake of nutrition-sensitive foods within general food groups (e.g., of green-leafy vegetables and soybeans for iron) when estimated nutrient intake fell below recommended values 47 . Lastly, we aggregated the urban diets for each demographic group to an average healthy and sustainable diet in a city by weighing intake by the demographic composition of each city in a country, similarly to how we estimated baseline dietary intake in a city.  SI Table 7 provides and overview of the dietary scenarios. Estimating the food-related environmental impacts of cities We calculated the environmental impacts of city diets by pairing estimates of food demand with food and region-specific environmental footprints. To calculate food demand, we added estimated portions of food waste 45 to our estimates of food intake, multiplied the per-person values by the estimated population in the specific demographic group 48 , converted daily values to an annual total, and aligned the food-group detail with those of the environmental footprints 44 . To estimate the associated environmental impacts, we used environmental footprints from a comprehensive meta-analysis of life cycle assessments (LCAs) covering GHG emissions, land use, freshwater use, and soil and water pollution as measured by eutrophication potential 7 . FAO data on food production and yields were used to scale and regionalise estimates 44 , and FAO data on trade were used to derive consumption-based footprints by commodity and region. SI Table 8 provides an overview of the footprints and food-group detail used. In addition to reporting changes in environmental impacts for each domain, we also calculated an overall indicator of food-related environmental impacts. Because the importance of diets and dietary changes for reducing environmental impacts varies for each domain, we used a weighing scheme that accounts for that, with weights assigned based on the needed contribution of dietary changes for staying within food-related environmental limits (or planetary boundaries) while also considering contributions from other food-system measures, including changes in technologies and management practices, food loss and waste, and socio-economic development (SI Figure 1) 9 . This meant weighing the changes in GHG emissions relatively more (46%) than changes in eutrophication potential (30%), and land and water use (12% each). Using simple averages yielded similar resulted, and we included those and the detailed set of estimated impacts in the interactive dashboard 32 . Estimating the diet-related health and cost impacts of cities We assessed the health and the associated cost burden of diets and dietary change in cities by using a comparative risk assessment framework with eight diet and weight-related risk factors and five disease endpoints 24 . For parameterizing the comparative risk assessment, we used data on cause-specific mortality from the Global Burden of Disease project 49 , body weight from the NCD Risk Factor Collaboration 50 , and relative risk estimates that relate change in risk factors to changes in disease mortality from meta-analyses of epidemiological cohort studies (SI Table 9). In line with the meta-analyses, we included non-linear dose-response relationships for fruits, vegetables, and nuts and seeds, and assumed linear dose-response relationships for the remaining risk factors. The selection of risk-disease associations was supported by available criteria used to judge the certainty of evidence (SI Table 10). In line with burden of disease assessments 51 , we estimated the number of deaths attributable to diet and weight-related risks in comparison to a counterfactual scenario in which disease risk is minimised. We adopted the intake values and weight levels associated with minimal risk from the same meta-analyses of cohort studies we used to parameterise the relative risk values (SI Table 11), supported by established criteria for defining minimal risk exposure levels 52 . As our analysis was primarily focused on mortality from chronic diseases, we focused on adults aged 20 year or older, and we adjusted the relative-risk estimates for attenuation with age based on a pooled analysis of cohort studies focussed on metabolic risk factors, 53 in line with other assessments 52,54 . We calculated uncertainty intervals associated with changes in mortality based on standard methods of error propagation and the confidence intervals of the relative risk parameters. For calculating the health-related costs of diets and dietary change, we paired the cause-specific estimates of attributable deaths with a set of cost-of-illness estimates 5,55 . Cost-of-illness estimates capture both the direct and indirect costs associated with treating a specific disease, including medical and health-care costs (direct), and costs of informal care and from lost working days (indirect). For our calculations, we used a global set of country-specific cost-of-illness estimates 5 and adapted them to the city level by accounting for within-country differences in per-capita GDP based on downscaled estimates 56 . We harmonised costs to 2020-USD by using changes in the consumer price index by country, and used costs adjusted for purchasing power parity. SI Table 12 provides an overview of the cost-of-illness estimates used in the analysis. Declarations Data Availability All results produced in this study are available via an interactive dashboard (https://livedataoxford.shinyapps.io/cities_dashboard/) and will be uploaded to a public repository upon publication. Acknowledgements MS, SC, and JR acknowledge funding from Wellcome Trust through a Career Development Award (Award number: 225318/Z/22/Z). MS and OA acknowledge funding from the EU Horizon Programme through the CATALYSE project (Grant agreement number: 101057131). MS also acknowledges funding from the EU Horizon Programme through the BrightSpace project (Grant agreement number: 101060075) and the ACT4CAP project (Grant agreement number: 101134874). The authors thank the C40’s food team (Christina Xiao, Anri Manderson, Stefania Amato, Artemis Tomadaki Balomenou) for helpful discussions. Contributions MS, JR, OA, and SC designed the study. MS conducted the main analysis, with contributions from JR, OA, and SC. JR conducted geospatial processing, OA contributed to the health assessment, and SC contributed to the environmental assessment and developed the interactive dashboard. MS wrote the manuscript. All authors provided inputs to the analysis, discussed results, commented on the manuscript, and approved the submission. Competing interests The authors declare no competing interests. Additional Information Supplementary Information is available for this paper. Correspondence and requests for materials should be addressed to Marco Springmann ( [email protected] or [email protected] ). References Willett, W. et al. Food in the Anthropocene: the EAT–Lancet Commission on healthy diets from sustainable food systems. The Lancet 393 , 447–492 (2019). Rockström, J. et al. The EAT–Lancet Commission on healthy, sustainable, and just food systems. The Lancet 406 , 1625–1700 (2025). Springmann, M., Mozaffarian, D., Rosenzweig, C. & Micha, R. Chapter 2: What we eat matters: Health and environmental impacts of diets worldwide. in 2021 Global Nutrition Report 34–50 (Development Initiatives, Bristol, UK, 2021). Springmann, M. Valuation of the Health and Climate-Change Benefits of Healthy Diets: Background Paper for The State of Food Security and Nutrition in the World 2020 . (FAO, Rome, Italy, 2020). doi:10.4060/cb1699en. Springmann, M., Godfray, H. C. J., Rayner, M. & Scarborough, P. Analysis and valuation of the health and climate change cobenefits of dietary change. Proceedings of the National Academy of Sciences 113 , 4146–4151 (2016). Crippa, M. et al. Food systems are responsible for a third of global anthropogenic GHG emissions. Nat Food 2 , 198–209 (2021). Poore, J. & Nemecek, T. Reducing food’s environmental impacts through producers and consumers. Science 360 , 987–992 (2018). IPBES. Global Assessment Report on Biodiversity and Ecosystem Services . (IPBES Secretariat, Bonn, Germany, 2019). Springmann, M. et al. Options for keeping the food system within environmental limits. Nature 562 , 519–525 (2018). United Nations Department of Economic and Social Affairs. World Urbanization Prospects: The 2018 Revision . (United Nations, 2019). doi:10.18356/b9e995fe-en. Miller, V. et al. Global dietary quality in 185 countries from 1990 to 2018 show wide differences by nation, age, education, and urbanicity. Nat Food 3 , 694–702 (2022). Miller, V. et al. Global, regional, and national consumption of animal-source foods between 1990 and 2018: findings from the Global Dietary Database. The Lancet Planetary Health 6 , e243–e256 (2022). Food and Agriculture Organization of the United Nations. Influencing Food Environments for Healthy Diets . (FAO, 2016). Herforth, A. & Ahmed, S. The food environment, its effects on dietary consumption, and potential for measurement within agriculture-nutrition interventions. Food Sec. 7 , 505–520 (2015). Food and Agriculture Organization of the United Nations. FAO Framework for the Urban Food Agenda . (FAO, Rome, Italy, 2019). doi:10.4060/ca3151en. Cohen, N. Roles of Cities in Creating Healthful Food Systems. Annual Review of Public Health 43 , 419–437 (2022). Carey, J. & Cook, B. The Milan Urban Food Policy Pact Monitoring Framework. A Practical Handbook for Implementation . (FAO, Rome, Italy, 2021). C40 Cities. C40 Good Food Cities Accelerator: How Cities Are Achieving the Planetary Health Diet for All . (2023). Goldstein, B., Birkved, M., Fernández, J. & Hauschild, M. Surveying the Environmental Footprint of Urban Food Consumption. Journal of Industrial Ecology 21 , 151–165 (2017). Ghaemi, Z. & Smith, A. D. A review on the quantification of life cycle greenhouse gas emissions at urban scale. Journal of Cleaner Production 252 , 119634 (2020). C40 Cities, Arup & University of Leeds. The Future of Urban Food Consumption in a 1.5 C World: C40 Cities . (2019). Guibrunet, L. & Sánchez Jiménez, A. The current and potential role of urban metabolism studies to analyze the role of food in urban sustainability. Journal of Industrial Ecology 27 , 196–209 (2023). Guibrunet, L., Hoekman, P., Bortolotti, A. & Battersby, J. Data requirements for a systematic analysis of urban food flows and their sustainability outcomes. Environment and Planning B: Urban Analytics and City Science 52 , 1444–1463 (2025). Springmann, M. et al. The healthiness and sustainability of national and global food based dietary guidelines: Modelling study. The BMJ 370 , 2322 (2020). Springmann, M. et al. Health and nutritional aspects of sustainable diet strategies and their association with environmental impacts: a global modelling analysis with country-level detail. The Lancet Planetary Health 2 , e451–e461 (2018). Springmann, M. The Diet Impact Assessment Model: A Tool for Analyzing the Health, Environmental and Affordability Implications of Dietary Change . https://www.who.int/europe/publications/i/item/WHO-EURO-2023-8349-48121-71370 (2023). Springmann, M. The Global Dietary Database for Impact Assessments (GDD-IA): facilitating diet-related analysis at global, regional, national, and sociodemographic levels. Lloyd, C. T., Sorichetta, A. & Tatem, A. J. High resolution global gridded data for use in population studies. Sci Data 4 , 170001 (2017). Springmann, M. Estimates of energy intake, requirements and imbalances based on anthropometric measurements at global, regional and national levels and for sociodemographic groups: a modelling study. BMJ Public Health 3 , (2025). Springmann, M. Many diets for many people: planetary health diets and their health and environmental impacts at global, regional, national, and demographic levels [preprint]. Preprint at https://doi.org/10.5281/zenodo.17079404 (2025). Springmann, M. et al. Health and nutritional aspects of sustainable diet strategies and their association with environmental impacts: a global modelling analysis with country-level detail. The Lancet Planetary Health 2 , e451–e461 (2018). Caleffi, S., Rawden, J., Auclair, O. & Springmann, M. Diets in cities and their impacts on the environment and health -- an interactive dashboard. https://livedataoxford.shinyapps.io/cities_dashboard/ (2025). Barbour, L., Lindberg, R., Woods, J., Charlton, K. & Brimblecombe, J. Local urban government policies to facilitate healthy and environmentally sustainable diet-related practices: a scoping review. Public Health Nutr 25 , 471–487 (2022). Westbury, S. et al. The influence of the urban food environment on diet, nutrition and health outcomes in low-income and middle-income countries: a systematic review. BMJ Glob Health 6 , (2021). FAO. The Milan Urban Food Policy Pact Monitoring Framework . (FAO, Rome, Italy, 2019). Candel, J. J. L. What’s on the menu? A global assessment of MUFPP signatory cities’ food strategies. Agroecology and Sustainable Food Systems 44 , 919–946 (2020). Barbour, L. R., Woods, J. L. & Brimblecombe, J. K. Local government policy to facilitate healthy and sustainable diets and the broader policy hierarchy: insights from Milan Urban Food Policy Pact cities. Health Res Policy Syst 21 , 35 (2023). Springmann, M., Hansoge, M. P., Schulz, L., Pastorino, S. & Bundy, D. A. P. The health, environmental, and cost implications of providing healthy and sustainable school meals for every child by 2030: a global modelling study. The Lancet Planetary Health 0 , (2025). Micha, R., Coates, J., Leclercq, C., Charrondiere, U. R. & Mozaffarian, D. Global Dietary Surveillance: Data Gaps and Challenges. Food Nutr Bull 39 , 175–205 (2018). Sonnino, R., Tegoni, C. L. S. & De Cunto, A. The challenge of systemic food change: Insights from cities. Cities 85 , 110–116 (2019). Blay-Palmer, A., Sonnino, R. & Custot, J. A food politics of the possible? Growing sustainable food systems through networks of knowledge. Agric Hum Values 33 , 27–43 (2016). WorldPop. Open Spatial Demographic Data and Research. WorldPop https://www.worldpop.org/. Earth Science Data Systems, N. Global Urban Polygons and Points Dataset (GUPPD), Version 1 | NASA Earthdata. Earth Science Data Systems, NASA (2025). Food and Agriculture Organization of the United Nations. FAOSTAT Statistical Database . (2022). Gustavsson, J., Cederberg, C., Sonesson, U., Van Otterdijk, R. & Meybeck, A. Global Food Losses and Food Waste: Extent, Causes and Prevention . (FAO Rome, 2011). Miller, V. et al. Global Dietary Database 2017: data availability and gaps on 54 major foods, beverages and nutrients among 5.6 million children and adults from 1220 surveys worldwide. BMJ Global Health 6 , e003585 (2021). Bajaj, S. & Springmann, M. A review of the quality of evidence of nutrient reference values. Lancet Planet Health 9 , 101308 (2025). United Nations Department of Economic and Social Affairs, Population Division. World Population Prospects 2024: Data Sources . https://population.un.org/wpp/assets/Files/WPP2024_Data_Sources.pdf (2024). GBD 2021 Causes of Death Collaborators. Global burden of 288 causes of death and life expectancy decomposition in 204 countries and territories and 811 subnational locations, 1990-2021: a systematic analysis for the Global Burden of Disease Study 2021. Lancet 403 , 2100–2132 (2024). NCD Risk Factor Collaboration (NCD-RisC). Worldwide trends in underweight and obesity from 1990 to 2022: a pooled analysis of 3663 population-representative studies with 222 million children, adolescents, and adults. Lancet 403 , 1027–1050 (2024). Murray, C. J. L., Ezzati, M., Lopez, A. D., Rodgers, A. & Vander Hoorn, S. Comparative quantification of health risks: conceptual framework and methodological issues. Population Health Metrics 1 , 1 (2003). Micha, R. et al. Etiologic effects and optimal intakes of foods and nutrients for risk of cardiovascular diseases and diabetes: Systematic reviews and meta-analyses from the Nutrition and Chronic Diseases Expert Group (NutriCoDE). PLOS ONE 12 , e0175149 (2017). Singh, G. M. et al. The Age-Specific Quantitative Effects of Metabolic Risk Factors on Cardiovascular Diseases and Diabetes: A Pooled Analysis. PLOS ONE 8 , e65174 (2013). Forouzanfar, M. H. et al. Global, regional, and national comparative risk assessment of 79 behavioural, environmental and occupational, and metabolic risks or clusters of risks in 188 countries, 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013. The Lancet 386 , 2287–2323 (2015). Springmann, M., Clark, M. A., Rayner, M., Scarborough, P. & Webb, P. The global and regional costs of healthy and sustainable dietary patterns: a modelling study. The Lancet Planetary Health 5 , e797–e807 (2021). Kummu, M., Kosonen, M. & Masoumzadeh Sayyar, S. Downscaled gridded global dataset for gross domestic product (GDP) per capita PPP over 1990–2022. Sci Data 12 , 178 (2025). Additional Declarations There is NO Competing Interest. Supplementary Files citiesSI123025.docx Supplementary Information Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information

Related documents

• Environmental Impact of Diets Based on Established and Newly Proposed Dietary Guidelines: DASH, Mediterranean Diet and Dietary Guidelines for Americans 2025-2030-Style Diet#253699
• Current Evidence on Dietary Patterns and Consumer Food Choices in the Context of Climate and Environmental Change: A Global Perspective-A Report from the Agriculture and Diet: Value Added for Nutrition, Translation, and Adaptation in a Global Ecology (ADVANTAGE) Project Working Group 2#215383
• Worldwide dietary patterns and their association with socioeconomic data: an ecological exploratory study#486922
• Current evidence on dietary patterns and consumer food choices in the context of climate and environmental change: a global perspective—a report from the “Agriculture and Diet: Value Added for Nutrition, Translation, and Adaptation in a Global Ecology” (ADVANTAGE) Project Working Group 2#124742
• Dietary exposure-outcome associations do not systematically differ by risk of bias in observational studies: a meta-epidemiological study#619802