Creating A Comprehensive Diet Database
When studying health outcomes, many scholars rely on information about people’s diets. Historically, however, research has often assumed that all plant-based diets are equal. For example, healthy and less healthy plant-based foods (e.g., refined grains and sweets) may be grouped together in diet studies, even though they are qualitatively different. Moreover, methods used to study food intake often don’t account for products or meals with more than one ingredient.
In Australia, the Australian Dietary Guidelines (ADG) database is used to estimate people’s consumption of major food groups (e.g., whole grains and vegetables), with the most recent data from 2011-2012. However, several food and drink items are not classified in the database because they are considered discretionary, rather than part of the core food groups. Nevertheless, many of these unclassified foods are widely consumed plant and animal foods high in saturated fat, salt, and sugar. In addition, some ingredients found in products with more than one ingredient are excluded.
The authors of this study argue that a database with the plant and animal content of all foods — regardless of their health value or how they’re incorporated into a dish — is needed to accurately assess a diet. They aimed to update the national Australian database by determining the plant and animal content of all whole foods, beverages, multi-ingredient products, and mixed dishes.
The first step was to define the plant- and animal-based food groups. They reviewed each of the 5,740 food and beverage items in the ADG database, in addition to data from three other sources. Items were placed into existing food groups based on ingredients or into a new category. For example, they created new “tea and coffee” and “animal fats” categories. Items that did not warrant the creation of a new group (e.g., protein powders and nutritional supplements) remained unclassified. They ended up with 23 food groups.
The second step was to calculate the plant and animal content per 100 grams of each item in the database. Their goal was to standardize the data so people can easily calculate how many servings of each ingredient they’ve consumed based on how much they’ve eaten. They used food labels, recipes, or estimates based on similar products. Finally, to test their database, they analyzed the diet of 20 participants from a national survey using both the original ADG and their expanded version.
Out of the products listed in the expanded database, around 84% were plant-based and 66% were animal-based. Intakes of food groups such as vegetables, nuts and seeds, and refined grains were higher in the updated database. In terms of distribution, plant- and animal-based ingredients were widely spread across the food groups. For example, 97% of foods with animal fats were found in food groups outside of “fats and oils.” The authors were surprised to see that fruits, nuts, and seeds were found in more discretionary products.
The results support the need for a comprehensive database that considers more animal and plant-based ingredients in a variety of forms. The data also show the versatility of plant and animal ingredients. For example, the authors note that people consume vegetables in various ways, including juices, salads, sandwiches, stir-fries, pizzas, and pies. Some of these forms are healthier and more essential than others.
This approach can be applied to the development of other food databases, but it may require modifications based on national and cultural differences as well as changes in the food supply over time. Armed with more precise data, future research can better investigate plant-based diets and their health outcomes. This type of database can also improve diet quality by providing targeted interventions — for example, providing food recommendations to those transitioning to a plant-based diet.