Comparison of Label and Laboratory Sodium Values in Popular Sodium-Contributing Foods in the United States.

BACKGROUND

Nutrition labels are important tools for consumers and for supporting public health strategies. Recent, published comparison of label and laboratory sodium values for US foods, and differences by brand type (national or private-label) or source (store or restaurant [fast-food and sit-down]) is unavailable.

OBJECTIVE

The objective was to compare label and laboratory values for sodium and related nutrients (ie, total sugars, total fat, and saturated fat) in popular, sodium-contributing foods, and examine whether there are differences by brand type, and source.

DESIGN

During 2010 to 2014, the Nutrient Data Laboratory of the US Department of Agriculture collected 3,432 samples nationwide of 125 foods, combined one or more samples of the same food (henceforth referred to as composites), and chemically analyzed them. For this comparative post hoc analysis, the Nutrient Data Laboratory linked laboratory values for 1,390 composites (consisting of one or more samples of the same food) of 114 foods to corresponding label or website (restaurant) nutrient values.

MAIN OUTCOME MEASURES

Label and laboratory values and their ratio for each composite, for each of the four nutrients (sodium, total fat, total sugars, and saturated fat).

STATISTICAL ANALYSES PERFORMED

Nutrient Data Laboratory analysis determined the ratio of laboratory to label value for each composite, and categorized them into six groups: ≥141%, 121% to 140%, 101% to 120%, 81% to 100%, 61% to 80%, and ≤60%. For sodium, the Nutrient Data Laboratory analysis determined the distribution of the ratios by food, food category, brand type, and source.

RESULTS

For sodium, 5% of the composites had ratios of laboratory to label values >120% and 14% had ratios ≤80%. Twenty-two percent of private-label brand composites had ratios ≤80%, compared with 12% of national brands. Only 3% of store composites had ratios >120% compared with 11% of restaurant composites. Ratios ≤80% were more prevalent among sit-down restaurants (37%) compared with fast-food restaurants (9%).

CONCLUSIONS

This study shows that a majority of label and laboratory values sampled agree and underdeclaration of label values is limited. However, there is some disagreement. Periodic monitoring of the nutrient content of foods through laboratory analyses establishes validity of the food labels and helps identify foods and food categories where the label and laboratory values do not compare well, and hence may need laboratory analyses to support accuracy of food composition data.