Food Composition Tables and Databases

Food composition tables and databases are curated collections of representative nutrient and component values for foods, organised so that they can be searched, linked to dietary intake data, and used in nutrition research, labelling, and policy. They are the bridge between laboratory analysis of individual samples and the population-level estimation of what people eat.

Definition

A food composition table or database is a compiled, documented set of representative values for the energy, nutrient, and component content of foods, structured for retrieval and for linkage to food consumption data.

Scope

This topic covers how composition databases are built — sampling foods, selecting analytical values, aggregating and documenting them, and harmonising across sources — and how they are used in dietary assessment and regulation. It treats databases as reference infrastructure and does not provide dietary advice or specific intake figures.

Core questions

• How are representative composition values selected and aggregated from analytical data?
• How is the provenance and quality of each value documented?
• How are databases harmonised so values are comparable across countries and studies?
• How are composition databases linked to dietary intake data for nutrition research?

Key concepts

• Representative (compiled) values
• Value documentation and metadata
• Data aggregation conventions
• Harmonisation across databases
• Food description and classification systems
• Missing values and imputation
• Linkage to dietary intake data

Mechanisms

A composition database is assembled by sampling representative foods, analysing or borrowing nutrient values, and compiling them into a single value per food-nutrient pair using documented aggregation rules. Each value ideally carries metadata on its source, analytical method, and quality, so users can judge its reliability. Harmonisation frameworks such as EuroFIR define how values are described and how aggregated values are documented, allowing databases from different countries to be compared and combined; web-based management systems support this curation in practice.

Clinical relevance

Because dietary assessment converts reported food intake into nutrient intake through these databases, the completeness and quality of the tables directly affect estimated intakes used in research and surveillance. This topic explains how that reference infrastructure is built; it is descriptive and not a source of individual dietary guidance.

Evidence & guidelines

The FAO/INFOODS programme and the EuroFIR network provide the principal conventions for producing, documenting, and harmonising composition data, extending the foundational guidelines of Greenfield and Southgate. Recent EuroFIR guidance specifies how aggregated and compiled values should be documented to improve cross-database harmonisation.

History

National food composition tables proliferated through the twentieth century as nutrition planning expanded. From the 1980s the FAO/INFOODS programme promoted international standards for data quality and interchange, and from the 2000s EuroFIR built networked, documented, and harmonised European databases, shifting the field from isolated printed tables toward interoperable digital systems.

Debates

Borrowed versus originally analysed values

Compilers often import values from other databases or the literature to fill gaps, which improves coverage but can propagate errors and obscure provenance; how much borrowing is acceptable, and how it should be documented, remains a curation judgement.

Key figures

• Heather Greenfield
• David A. T. Southgate
• Paul Finglas
• Mirjana Gurinovic

Seminal works

• greenfield-southgate-1992
• southgate-1988
• westenbrink-2021

Frequently asked questions

Are food composition database values exact for a specific food I buy?

No. Database values are representative estimates aggregated across samples, so the actual content of a particular item can differ depending on cultivar, origin, season, storage, and processing.

Why do nutrient values differ between national databases?

Different countries sample different food supplies and may use different analytical methods, aggregation rules, and food descriptions, which is why harmonisation conventions exist to make values more comparable.