In this study, the affordability of a healthy diet is measured by the number of people who can afford it. Under all three scenarios, the global GDP is expected to increase by 5 to 10 times by the end of this century (SI Fig.1), significantly increasing the number of people worldwide who can afford a healthy diet. However, climate change will impact crop yield and therefore the affordability of a healthy diet, which is the core focus of this study.
The climate-induced affordability loss is estimated by calculating the differences in the number of people who can afford a healthy diet via simulations with and without climate-induced yield reductions (Fig.1a, specific details can be found in the Methods section).
By the end of this century, under the SSP1-2.6, SSP2-4.5, and SSP5-unconstrained scenarios, the populations unable to afford a healthy diet are 122 million, 307 million, and 129 million, respectively, accounting for 1.74%, 3.37%, and 1.72% of the total population (SI Fig.2). Among these, the increases due to the impact of climate change are 7.6 million, 46 million, and 74 million, respectively (Fig.1b-e). In all scenarios, the spatial distributions of the losses in affordability due to climate change are similar; Africa, Asia, and South America are the most affected regions (Fig.1b-c). In contrast, North America, Europe, and Oceania are relatively less affected. This finding highlights a key issue: the food systems of developing regions are more vulnerable to climate change than are those of wealthier regions, which typically have greater impacts on climate change.
As global average temperatures rise, the disparity between developing and wealthier regions in terms of climate-induced losses in affordability will increase nonlinearly. In the SSP1-2.6 scenario, 94.7% of the affordability loss occurs in Africa, Asia, and South America, while in the SSP5-unconstrained scenario, this proportion increases to 97.3% (Fig.1e). Additionally, under the SSP5-unconstrained scenario, regions such as Indonesia are particularly severely impacted, with an additional 1158.1% of the population unable to afford a healthy diet (Extended Fig.1). These results indicate that as temperatures increase, the gap in the affordability of a healthy diet between economically underdeveloped regions and wealthier regions will widen, and the differences in the vulnerability of food systems to climate change across regions will become more pronounced.
Globally, the population unable to afford a healthy diet is projected to increase due to climate change-induced escalation in the prices of various food items, consequently increasing the CoHD (Fig.2). Across all nations, more pronounced climate change is anticipated to increase expenditures on a healthy diet. In the SSP1-2.6 scenario, which is the lowest warming scenario of the three scenarios, the CoHD increases by 2.1% to 11.9% (median of 3.5%) relative to CoHD under no climate change (Fig.2a). For the SSP2-4.5 scenario, the increase in price for a healthy diet range from 2.1% to 31.6% (median of 5.81%) (Fig.2b). Compared to under SSP1-2.6, under SSP2-4.5, some regions experience similar price changes, while others see increases that are twice as large. Under the SSP5-unconstrained scenario, climate change is projected to increase the local CoHD by 11.9% to 1159% compared with that under the no climate effect scenario (with a median increase of 29.5%), which is significantly greater than that under the other two scenarios, with some regions experiencing extremely large increases (Fig.2c). Overall, the increase in CoHD under the high-temperature scenario is significantly greater than that under the low-temperature scenario. However, regional disparities are also evident. For example, in Africa, the relative changes in CoHD range from 3.0%-10.7% (SSP1-2.6) and 42.5%-329.6% (SSP5-unconstrained) (right panels in Fig.2a, Fig.2c).
The change in CoHD is driven by the changes in the food prices of the six food groups. Among the six food groups constituting a healthy diet, 'Animal source foods,' 'Fruits,' and 'Vegetables' are relatively less affected by climate change in all three SSP scenarios. In contrast, 'Starchy staples', 'Legumes, nuts and seeds' and 'Oils and fats' are the most sensitive to climate change. Under low warming scenarios, the price response ratio of these two food groups is comparable to the overall trend. However, under high warming scenarios, significant outliers emerge, with locals needing to spend up to 30 to 50 times the price to meet their caloric requirements for these two food groups from a healthy diet (Fig.2c) .
Notably, under the high warming scenario, North America, particularly the United States and Canada, exhibits the lowest price response ratios across all food groups. More intense climate change poses a substantial impact on the food systems of some developing countries and regions, leading to a sharp increase in the prices of certain food items and a consequent substantial rise in the CoHD. For wealthier regions, while climate change poses a challenge to food production, the impact is not as critical as that faced by developing nations. These findings underscore the urgency for targeted policy interventions that can enhance the resilience of food systems, particularly in developing regions that are most vulnerable to the adverse effects of climate change. The detailed trend of changes in CoHD with and without climate change impacts across regions from 2015 to 2100 under these three scenarios are provided in the extended data (Extended Fig.2).
According to data from the FAO and the World Bank, as regional GDP per capita increases, the proportion of the malnourished population tends to decrease. However, our study's simulation results reveal an important phenomenon: the positive impact of per capita GDP growth on improving the affordability of a healthy diet diminishes with the intensification of climate change. Specifically, the yield loss in agriculture caused by climate change largely offsets the positive effects of economic development in reducing the number of people who are unable to afford a healthy diet.
We estimated the years of per capita GDP growth that have been offset from the perspective of improving the affordability of a healthy diet. Figure 3a is a conceptual diagram showing how much affordability loss and how many years of per capita GDP growth are offset, by comparing the proportions of the different regional populations that can afford a healthy diet at the end of the century with those in the scenario without climate change impacts.
The offset years for the regions are smallest under SSP1-2.6 (0.7-3.9 years), followed by SSP2-4.5 (1.1-11.1 years), and largest under SSP5-unconstrained (3.9-63.1 years) (Fig.3b). Under SSP5-unconstrained, Europe has the least offset years, followed by North America, Australia, and South America, while Asia and Africa have the most offset years. The regions where GDP growth is most significantly offset are Indonesia, Western Africa, and Eastern Africa, with offsets of 63.1 years, 41.3 years, and 34.8 years, respectively, reaching only the economic development levels of 2036-2065 without the impact of climate change by the end of the century (Fig.3c). Our results indicate that the benefits of economic development on healthy diet affordability are significantly offset by climate change, particularly in Africa and Asia. To mitigate these losses, additional per capita GDP growth will be needed. We tested the impact of different global GDP annual growth rates on affordability loss. Taking SSP2-4.5 as an example, we found that under varying GDP growth rates, climate change robustly leads to a significant number of people being unable to afford a healthy diet (Extended Fig.3). Furthermore, if economic growth falls short of expectations, the number of people unable to afford a healthy diet will increase substantially.
We compared the impact of climate change on the affordability of healthy diets in different regions by the end of this century and contrasted it with the cumulative carbon emissions of these regions (Fig.4). The results show a significant mismatch between the loss of affordability due to climate change and the corresponding climate responsibilities of these regions. The top 8 regions experiencing loss of affordability suffer more than 80% of the global loss in all scenarios, but their cumulative carbon emissions are only approximately 20%. In all scenarios, these top 8 regions are in Africa, South Asia, and Southeast Asia, all of which are developing regions (Fig.4). This finding demonstrates that these regions are the most vulnerable to the impact of climate change on the affordability of a healthy diet.
In the SSP1-2.6 scenario, which is the lowest warming scenario among the three scenarios, the top eight regions suffering the most from climate change-induced affordability loss account for 83% of the global total affordability loss but only 20% of the global cumulative emissions. Conversely, regions accounting for 80% of the global cumulative CO2 emissions suffer only 7% of the global total affordability loss (Fig.4a). The inequitable burden on developing nations is evident. However, this scenario is the most balanced. In the SSP2-4.5 and SSP5-unconstrained scenarios, these regions bear an even larger share of the affordability loss while contributing even less to carbon emissions (Fig.4a, Fig.4b). Considering all the scenarios, our results reveal the unfair trend of "less emission, more penalty," where residents of some developing regions cannot afford a healthy diet due to the emissions of other regions.