Issues of Equivalency in Determining Antioxidant Capacity?

About ORAC Database

Procedures Used to Generate the Database

In 2007, USDA released the first database of antioxidant activity for 277 selected foods using the Oxygen Radical Absorbance Capacity (ORAC) in vitro laboratory assay. These foods were analyzed for ORAC by Wu et al. (74). Release 2 by USDA in 2010 contained ORAC values for 326 food items. Since then additional scientific publications containing data on the ORAC content of foods have been published which have been included in the current database. Data through January 2013 have been reviewed and incorporated into this ORAC database. The data file in this website contains 564 items. The data were aggregated where possible to match the food descriptions and Nutrient Database Number in the USDA National Nutrient Database for Standard Reference.

The analytical method developed by Prior et al. (56) was used as the reference method for evaluating analytical methods from both other published and unpublished sources. This method uses fluorescein as the fluorescent probe and measures hydrophilic (water soluble) as well as lipophilic (Lipid soluble) antioxidants. Analytical data from literature based on earlier methods that used β-phycoerythrin (β-PE) as the probe were not used in this compilation as β-PE may produce inconsistent results in some foods, is not photostable, and may involve nonspecific protein binding with polyphenols (16). Data have also been included which were provided by food industry sources.

Estimation of Missing Values

Some analytical studies reported only H-ORAC or Total-ORAC values. If that was the only data source for that particular food, values for other ORAC components were not imputed. Similarly, missing values were not imputed from other forms of the food (cooked, canned etc.) or other similar foods due to large variability in the ORAC values. If only H-ORAC, was reported, that value was assumed for Total-ORAC, as the values for L-ORAC are, with the exception of a few foods, very low compared to the H-ORAC values. When data from more than one source for the same food were aggregated, the missing ORAC components were imputed by averaging the available values from other sources for that component. An imputed value was calculated only when two or more values were available for the particular component in the aggregate. Generally, imputing was necessary for the L-ORAC component and occasionally, for both H- and L-ORAC components when only Total-ORAC values were reported by a particular data source. If there were many missing values for L-ORAC for a particular food item, and only one value was available to impute a missing value, the imputing step was not performed. In these cases, the single available value for L-ORAC was relatively small compared to the level of the H-ORAC value, and therefore, is not reported. Where imputed values were used in the calculation of the summary statistics, an asterisk appears next to the appropriate parameter in the table.

Format of the Database

ORAC values are reported for hydrophilic-ORAC (H-ORAC), lipophilic-ORAC (L-ORAC), Total-ORAC, and total phenolics (TP). H-ORAC, L-ORAC, and Total-ORAC are reported in μmol of Trolox equivalents per 100 grams (μmol TE/100 g), while TP is reported in mg gallic acid equivalents per 100 grams (mg GAE/100 g). The mean value (mg/100g), standard error of the mean (SEM), minimum (Min), and maximum (Max) values are reported for each food item and ORAC component value if more than 3 samples were analyzed. Standard error was not calculated if the number of samples reported was less than three.  Minimum and maximum values were not reported when the number of samples equaled one. Other conditions where full statistical information is not provided are described above.

Each food item was given a Nutrient Data Bank (NDB) number, the five digit numerical code used in the USDA SR to identify each unique food entry if it matches a food in the USDA SR. As the data came from various sources both in the United States and other countries, there are a number of foods which are not included in the USDA SR database. NDB numbers, beginning with either “97” or “99”, were assigned to foods that are not included in SR. A reference number corresponding to a publication in the sources of data section of the documentation is included in the table for each food and component.