A food name is only part of a food-composition record’s identity. Two search results can describe similar foods while drawing their values from different kinds of evidence. In FoodData Central, record the data type before using the numbers in a comparison.
USDA’s comparison identifies Foundation Foods with analytical sample data, FNDDS with foods and portions used in dietary surveys, Branded Foods with label information, and SR Legacy as historical reference data. Those categories serve different research tasks. Their appearance in one search system does not erase their provenance.
Match the record to the question
| Question | First distinction to examine |
|---|---|
| A basic food’s composition | Analytical source and exact food description. |
| A named packaged product | Brand, market and label-based record identity. |
| A survey food code | Survey period and corresponding database release. |
These are starting questions, not instructions to choose an entry without reading it. Preparation, enrichment and the food’s form can still make an otherwise plausible match unsuitable.
Keep a selection note
Save the identifier, complete description, data type, release and access date. Add one sentence explaining why the record fits the intended comparison. If a substitute is needed, preserve its relationship to the requested food rather than presenting it as an exact match.
A generic food description should not silently acquire a specific retail product’s identity merely because its marketing name contains the same food word. A brand-level claim needs evidence about the identified record and version.
Compare like measures
After choosing records, check nutrient names, units and quantity bases. A sound selection can still produce an invalid comparison if one value is per portion and another per 100 grams. Keep missing information visible.
No food entries were selected, measured or ranked for this article. It explains how to document a research-record choice, without a dietary prescription or a claim that one data type is universally best. The output should have identifiable sources and explainable selection decisions.
Official documentation read 21 October 2026. Workflow suggestions and labelled arithmetic examples are editorial. No clinical finding or individual dietary advice is provided.
Primary documentation: USDA FoodData Central data-type comparison. Related discussion: health information source quality.
Leave a Reply