A New Method to Monitor the Nutritional Quality of Packaged Foods in the Global Food Supply in Order to Provide Feasible Targets for Reformulation

Abstract

Nutrient profiling systems, initially designed to promote healthy food choices at the point of sale, can also provide the scientific basis for innovation and product reformulation by the food industry. This work presents a new profiling system to help define feasible nutrient targets for reformulation of packaged foods. The focus is on five key nutrients for which the World Health Organisation (WHO) has set population-level goals: sugar, saturated fat, sodium, fiber, and protein. The methodology uses Mintel’s Global New Products Database of packaged foods to (1) identify nutrients relevant to each food category (2) sort products into sub-categories defined by a unique nutritional signature, and (3) develop standards for “best of class” products. For instance, if targeted to be amongst the best 15% of the global food supply, pizza must have less than 4.0 g/100 g saturated fat, less than 520 mg/100 g total sodium and more than 9.8 g/100 g protein. Fiber and sugar are not identified as relevant nutrients for the pizza category and no targets are provided.

Monitoring improvements in the nutritional quality of new packaged foods launched between 2016 and 2020

“…The present model relies on three algorithms to analyze the distributions of five public health sensitive nutrients—total sugars, total sodium, and saturated fat along with protein and fiber ( 7 )—that were recently developed and discussed in detail ( 6 ). The first algorithm identified which of the five considered nutrients were relevant for each category; a nutrient was considered category-relevant if more than half of the products within this category had a declared nutrient content with a value that exceeded the “low in” or the “source of” limit.…”

“…For the full 3-year period 2018–2020, the Mintel Global New Products Database lists 397,958 products that can be split into 288 nutritional subcategories. In a previous study, we used 350,994 products for 263 subcategories, but the 3 years period was truncated to 30 months instead of 36 ( 6 ).…”

“…Breakfast cereals or pizza can be of variable nutrient density. To overcome this problem, we have developed a new methodology ( 6 ) to disaggregate marketing categories into product subgroups or “classes” that were more nutritionally homogeneous. That procedure was used to set reformulation targets for new products that were comparable to those that in 2016-18 were judged to be “best of class” ( 6 ).…”

Algorithm-based mapping of products in a branded Canadian food and beverage database to their equivalents in Health Canada’s Canadian Nutrient File

“…In recent decades, food marketing and retail databases have been revisited as a largely untapped source of low-bias, high-quality data for researching trends in consumer health and nutrition ( 1 ). Such databases can also be utilized for the study of front-of-pack labeling, marketing and advertising to children, and the implementation and surveillance of dietary guidelines ( 2 ). For food and beverage manufacturing industries, these data on the nutrient content of their products and associated sales can guide healthy eating initiatives and even product reformulations ( 3 ).…”