Shelf to Health: Does Product Innovation Change National Estimates of Dietary Impacts?

This paper presents a simulation process to augment nutrition surveillance in the United States which incorporates product innovation data. Traditional point-estimates of nutritional quality in a food category are compared to those based on distributions of nutrient compositions using product-level variability seen in the market. Nationally representative consumption patterns provide dietary intakes. Cookies are used as an example food category. Nutrient composition data from Global New Product Database (GNPD) for 5259 cookies launched 2005 to 2012 were matched to dietary intakes from 2005 to 2012 National Health and Nutrition Examination Survey (NHANES) over the 2 y cycles of NHANES for 8284 cookie consumers. Average dietary intakes from traditional NHANES and GNPD-based estimations produced similar mean values for energy, carbohydrates, sugars, total fat, and protein. Saturated fat, fiber and cholesterol contributions using new product compositions were significantly higher than traditional NHANES approaches, estimates of sodium were significantly lower. These differences become pronounced when comparing adult and child consumption patterns and over time. This process also simulated trans fat consumption estimates not traditionally available within NHANES. On average cookies contributed 0.3 g/d (range 0 to 4.1 g/d). Much variability in food composition is seen in the market which is shown to influence estimates of the national diet.

Numerous factors drive changes in the food supply, including health trends, firm strategic choices, and food policy. This evolution presents a challenge for dietary assessments and nutrition monitoring. The public health impact of variability in nutritional composition, subpopulation consumption patterns and market dynamics are particularly difficult to evaluate and are shown to influence estimates of the national diet.