Table 1.
Minimum and maximum recommended dietary intake of 9 food categories—Based on EAT-Lancet Dietary Guidelines and [42].
Fig 1.
Aggregation and computation method for dietary quality metrics—Overshoot of EAT-lancet dietary guidelines and undershoot of EAT-lancet dietary guidelines.
Table 2.
Global average GDP per capita and population values for the IMPACT model base year (2005) and four select SSP scenarios.
Based on data from IIASA public SSP database (https://tntcat.iiasa.ac.at/SspDb/).
Fig 2.
Percent overshoot and undershoot of EAT-Lancet recommended dietary guidelines for a global set of countries in the year 2010, based on FAO data.
Where no data exists for country-level diets, values from the IMPACT model run for SSP2 are used. Values closer to zero indicate more “ideal” dietary outcomes with respect to EAT-Lancet guidelines. Source: Author’s computation. Shapefiles for map generation sourced from the NaturalEarth project (naturalearthdata.com).
Fig 3.
2050 projected average “distance” (overshoot or undershoot) from ideal diet (zero = ideal).
Regional averages are a population-weighted average of country-level overshoot/undershoot as projected by the IMPACT model. Regions as follows: SSA = Sub-Saharan Africa, SAS = South Asian States, NAM = North America, MEN = Middle East/North Africa, LAC = Latin America and Caribbean, FSU = Former Soviet Union, EUR = Europe, EAP = East Asia Pacific.
Fig 4.
2050 projected percent of total diet that is satisfied through imported goods.
Region-level values are population-weighted country-level averages of the portion of average diet imported. Regions as follows: SSA = Sub-Saharan Africa, SAS = South Asian States, NAM = North America, MEN = Middle East/North Africa, LAC = Latin America and Caribbean, FSU = Former Soviet Union, EUR = Europe, EAP = East Asia Pacific.
Table 3.
Correlation between the change in the level of agricultural imports per capita and change in the dietary overshoot and dietary undershoot metrics between 2010–2050 for SSPs 2–5.
Statistically significant relationships are shown in bold.
Table 4.
Cross-country correlations tracking the change in country-level overshoot (arithmetic mean of food group category-level overshoot) between 2010 and 2050 and the change in imports per capita between the years 2010 and 2050.
Aggregation by income is via global income quartiles, and are roughly analogous to World Bank income groupings. Income quartiles groupings are static over time (i.e., the income designation of a given country in 2010 persists through 2050). Correlation coefficients are calculated at the income group level.
Table 5.
Cross-country correlations tracking the change in country-level undershoot (arithmetic mean of food group category-level overshoot) between 2010 and 2050 and the change in imports per capita between the years 2010 and 2050.
Income regions are constructed via global income quartiles, and are roughly analogous to world bank income groupings. Income quartiles groupings are static over time (i.e., the income designation of a given country in 2010 persists through 2050). Correlation coefficients are calculated at the income group level.
Fig 5.
Regional comparison of “pre-trade” (production-only) average diets and actual average diets in overshoot and undershoot terms.
Regional-level metrics are constructed as a population-weighted mean of average country-level dietary undershoot or overshoot.
Fig 6.
Regional percentages of dietary undershoot and overshoot represented by EAT-Lancet dietary category.
Regional-level metrics are constructed as a population-weighted mean of country-level overshoot (left), undershoot (middle), and import volume (right). As the percent decomposition is a relative metric, the absolute size of any dietary category is not comparable across regions, but should be observed against other dietary categories across time for a given region.