How PairIQ chooses its evidence
A finding only goes into PairIQ if we can point to where it came from. If we cannot, it stays out.
- 457foods
- 80research-backed pairing rules
- 902pairing effects mapped
- 52studies cited by DOI
The order of evidence
- Human studies. Absorption studies, trials and reviews in peer-reviewed journals come first.
- Reference data. USDA FoodData Central, NIH Office of Dietary Supplements fact sheets and national nutrient tables.
- Lab and animal studies. Used only when nothing better exists, and the finding says so in its own words.
How claims are checked
Citations are checked against the paper itself: that the DOI resolves to a real paper, that the paper's findings match the claim, and that the numbers match what the paper reports rather than a press-release version. When a "3 to 9 times" carotenoid claim turned out to be 3 to 8 times in the underlying Kim et al. (2015) paper, it was corrected. Where a claim cannot be backed honestly, it is hedged or removed.
Popular claims get the same treatment. The often-quoted "2000%" boost from black pepper and turmeric comes from one small supplement-dose study over a near-zero baseline, so PairIQ keeps the mechanism and drops the number. Read the turmeric finding.
How the iron score works
PairIQ's iron estimate runs Hallberg and Hulthén's published absorption algorithm on each food's own iron, vitamin C, calcium and phytate values, mostly from USDA FoodData Central. It is portion-aware: small, medium and large servings change it. The score compares one meal with another; it does not measure anyone's iron status.
Where the data comes from
6,252 nutrient values come from these USDA datasets:
- McCance and Widdowson's Composition of Foods Integrated Dataset 2021 (Public Health England). Food code 12-174. gov.uk
- USDA FoodData Central, Branded Foods
- USDA FoodData Central, FNDDS 2021-2023 (October 2024 release)
- USDA FoodData Central, Foundation Foods (April 2026 release)
- USDA FoodData Central, SR Legacy (2018)
- USDA FoodData Central, SR Legacy (2018), RAW basis. Raw elderberries are unsafe to eat; cooking reduces vitamin C substantially
- USDA FoodData Central, SR Legacy (2018), cooked
- USDA FoodData Central, SR Legacy (2018), dill
- USDA FoodData Central, SR Legacy (2018), raw
- USDA FoodData Central, SR Legacy (2018), raw. Largely oxalate-bound, availability not established
- USDA FoodData Central, SR Legacy (2018), sour pickles
- USDA FoodData Central, SR Legacy (2018). Range spans sweet (457) to sour (1208)
- USDA Nutrient Data Set for Fresh Pork, Release 1.1 (2007). DOI 10.15482/USDA.ADC/1409036
- USDA Nutrient Data Set for Retail Beef Cuts, Release 3.0 (2013). DOI 10.15482/USDA.ADC/1409036
- USDA Nutrient Data Set for Retail Lamb Cuts (2017). DOI 10.15482/USDA.ADC/1409036
- USDA Nutrient Data Set for Retail Veal Cuts (2014). DOI 10.15482/USDA.ADC/1409036
- Williams JR, et al. (2017). Poultry Science, 96(10):3586-3592. DOI 10.3382/ps/pex165
Recommended intakes come from the NIH Office of Dietary Supplements. Packaged-food data for the barcode scanner comes from Open Food Facts, under the Open Database License.
What PairIQ is not
PairIQ is general education about food and nutrient absorption. It is not a medical device and does not diagnose, treat or prevent any condition. It does not count calories. If you are pregnant, managing a condition or taking medication, talk to your doctor before changing how you eat or supplement.