Data

Where the numbers come from.

Food

Where the numbers come from

Where each table comes from

1,369

foods from 6 official tables, each value per 100 g

32,856

data points (24 per food) re-checked against the original files by a separate program

Foods by region

Number of foods. Solid: mainly eaten there. Striped: eaten widely, placed by the country of its data table.

  • North America370, 19 mainly eaten there, 351 placed by data table
  • South Asia297, 297 mainly eaten there, 0 placed by data table
  • Europe276, 113 mainly eaten there, 163 placed by data table
  • East Asia196, 196 mainly eaten there, 0 placed by data table
  • South America122, 122 mainly eaten there, 0 placed by data table
  • Africa94, 94 mainly eaten there, 0 placed by data table
  • Southeast Asia10, 10 mainly eaten there, 0 placed by data table
  • Middle East4, 4 mainly eaten there, 0 placed by data table

How a food gets its region. Each food sits in the region where it is mainly eaten, such as natto in East Asia or feijoada in South America. Foods eaten all over the world, like apples or chicken, sit in the region of the table their numbers come from.

6 official tables

Number of foods taken from each table.

  • United States: USDA SR28USDA National Nutrient Database, Release 28 · US Department of Agriculture561
  • India: IFCT 2017Indian Food Composition Tables 2017 · ICMR-National Institute of Nutrition224
  • United Kingdom: CoFID 2021UK Composition of Foods Integrated Dataset 2021 · UK Government (McCance and Widdowson)217
  • Japan: MEXT 2020Standard Tables of Food Composition in Japan 2020 · Ministry of Education, Culture, Sports, Science and Technology163
  • Brazil: TACO 2011Brazilian Food Composition Table (TACO), 4th edition · NEPA, State University of Campinas117
  • West Africa: WAFCT 2019West African Food Composition Table 2019 · FAO and INFOODS87

Every value checked

Result of the checks, by number of foods.

  • Verified, matches the source1,211
  • Verified, with a note140
  • Corrected, documented18
18 corrections, and 1 column not used
  • Coconut oil, Sesame oil (gingelly), Groundnut oil, Mustard oil, Palm oil, Rice bran oil, Safflower oil, Soybean oil, Sunflower oil, Ghee, Vanaspati (hydrogenated fat)
    Source lists 0 kcal for a pure fat, which is impossible. Energy calculated from fat at 9 kcal per g.
  • Duck (lean meat)
    IFCT labels this "meat, with skin", but its values match skinless duck meat (USDA 05141), so it is shown as lean duck meat.
  • Rambutan
    IFCT fatty-acid breakdown adds up to more than total fat. Saturated, mono and poly values are not used.
  • Quinoa
    USDA lists DHA (a long-chain omega-3) for quinoa; plants do not make DHA. Not used.
  • Fuba (fine cornmeal), Doce de leite
    TACO fatty-acid breakdown adds up to more than total fat. Saturated, mono and poly values are not used.
  • Gari (toasted fermented cassava)
    WAFCT lists vitamin B12 for gari; cassava does not contain meaningful B12. Not used.
  • Fufu flour (dry fermented cassava)
    WAFCT lists vitamin B12 for fufu flour; cassava does not contain meaningful B12. Not used.
  • IFCT vitamin D not used (224 foods)
    IFCT 2017 lists vitamin D in plant foods at levels plants cannot reach (for example 109 µg in pomegranate) and in egg white, which has none. The whole column is treated as unreliable and is not used.

How processed

Number of foods in each NOVA processing group.

  • 1: Fresh or minimally processed875
  • 2: Cooking ingredient49
  • 3: Processed264
  • 4: Ultra-processed181

Ultra-processed foods lose 6 points from the longevity score. A large 2025 study linked them to lower odds of healthy ageing.

Air & UV

Where the Air & UV numbers come from

  • Open-Meteo Air Quality: UV index, PM2.5, PM10 and air quality indices, hourly, modelled from the Copernicus Atmosphere Monitoring Service (CAMS) and ECMWF data.
  • Open-Meteo Geocoding: place names and coordinates, based on GeoNames.
  • OpenAQ: the latest PM2.5 readings from government and research monitoring stations.