Grocery Purchase Quality Index‐2016 Scores Are Moderately Correlated with Healthy Eating Index‐2010 Scores in the Food Acquisition and Purchase Survey, 2012–13

Abstract

Objective This study evaluates the Grocery Purchase Quality Index-2016 (GPQI-2016), a new tool developed at the University of Utah for assessing household grocery food purchase quality. We used the Healthy Eating Index-2010 (HEI-2010), a well-validated assessment method, as the reference standard. Instead of using amounts of foods or nutrients, the GPQI-2016 is based on the expenditure shares for the 29 food categories found in the USDA Food Plans. It includes the eight food-based adequacy components (Total Vegetables, Greens and Beans, Total Fruit, Whole Fruit, Whole Grains, Dairy, Total Protein Foods, and Seafood and Plant Proteins) and one food-based moderation component of the HEI-2010 (Refined Grains), plus additional moderation components for Added Sugars and Processed Meats. Components of the HEI-2010 not included in the GPQI-2016 are Fatty Acids, Sodium, and Empty Calories. Methods In 2012–13 the USDA Economic Research Service (USDA-ERS) conducted the National Household Food Acquisition and Purchase Survey (FoodAPS). Household members recorded all foods purchased for a week--both foods purchased at stores for at-home use (“foods at home” (FAH)) and foods obtained and eaten away from home. FoodAPS includes information on prices paid for each food item in a purchasing event or transaction. We mapped the USDA-ERS food group classifications, provided in the survey database, to the 29 food categories used in USDA Food Plan market baskets so that the expenditure shares could be estimated. The 8-digit USDA food codes, provided in the survey database, were used to calculate the HEI-2010. After scoring FAH purchases by participating households (n=4,134), using each assessment method, the weighted Pearson's r coefficient was used to compare the total GPQI-2016 scores with the total HEI-2010 scores. The correlations for the nine food-based components used in both indexes were also estimated. Results The weighted Pearson's correlation coefficient for the Total HEI-2010 and the total GPQI-2016 scores was 0.61 (p < 0.01). The strength of the correlations between the HEI-2010 and the GPQI-2016 components varied (p < 0.01 for all). The strongest associations were found for three of the adequacy components, Whole Grains (0.77) and Total and Whole Fruit (both 0.75); while for the others, associations were more moderate, ranging from 0.42 for Total Protein Foods to 0.58 for Total Vegetables. For Refined Grains, a moderation component, the correlation was 0.45. Conclusions Overall, the association of the GPQI-2016 and the HEI-2010 was moderate, and it varied by component. The process used to map the FoodAPS data to the USDA Food Plan categories could be refined by mapping at the food item level rather than at the food group level and may result in higher correlations in future versions of the GPQI. Support or Funding Information NIH grant T15-LM007124, the Western Regional Nutrition Education and Obesity Prevention Center of Excellence (USDA-NIFA-OP-004574), and the USDA National Institute of Food and Agriculture (2015-09151)