Americans are consuming less sugar but more nonnutritive sweeteners

July 29, 2020

A new study in the Journal of the Academy of Nutrition and Dietetics found that between 2002 and 2018, United States household purchases of foods and beverages containing caloric sweeteners (CS, i.e., sugar) declined. Purchases of products containing both caloric sweeteners and non-nutritive sweeteners (NNS, i.e., sugar substitutes) increased, however.

NNS include aspartame, saccharin, rebaudioside A (reb-A) and sucralose/Splenda, which provide sweetness to products without the calories of sugar or high fructose corn syrup.

Dr. Barry Popkin

“With excessive sugar consumption linked to chronic cardiometabolic diseases, sugar reduction has become an important public health strategy. This has resulted in greater innovation by the food industry and increased use of NNS in our food supply,” said lead investigator Barry Popkin, PhD, W.R. Kenan Jr. Distinguished Professor of nutrition at the UNC Gillings Global School of Public Health, and a researcher with the Carolina Population Center.

The study examined how the prevalence and volume of foods with common NNS in the U.S. packaged food supply had changed between 2002 and 2018. Co-investigator Shu Wen Ng, PhD, Distinguished Scholar in Public Health Nutrition at the Gillings School, said the researchers found a decline in market prevalence of products containing aspartame and saccharin, but an increase in products with sucralose/Splenda (up from 38.7% to 71%) and reb-A (increased from 0.1% to 25.9%). Beverages accounted for most of the products purchased that contained NNS alone or in combination with CS. Compared to households without children, households with children buy more beverages and food products that contain NNS. While this aligns with public health objectives related to decreasing sugar intake, it also raises concerns about exposure to NNS.

Additionally, the study showed that non-Hispanic white consumers purchased almost double the volume of products containing NNS compared to Hispanic and non-Hispanic Black consumers. However, non-Hispanic Black households showed a 42% increase in the proportion of households purchasing beverages containing both CS and NNS between 2002 and 2018, indicating that purchasing behavior may be changing for this group.

The researchers used a nationally representative dataset on household purchases at the barcode level (the Nielsen Homescan) to review purchases in 2002 and 2018. These data were linked with Nutrition Facts Panel data and ingredient information using commercial nutrition databases that are updated regularly to capture product reformulations. Researchers performed keyword searches on ingredient lists to classify products containing various types of NNS, and then determined the volume of products containing NNS and/or caloric sweeteners that each household puschased per capita per day in 2002 and 2018.

“There is a need to be able to track our exposure to specific types of sweeteners in order to properly understand their health implications,” noted Elizabeth Dunford, PhD, a postdoctoral fellow in nutrition with the Gillings School and the Carolina Population Center. “The change to the food supply shown in our study documents reinforces the need to develop and maintain data systems to monitor what companies are putting in their foods. This work complements emerging clinical evidence about the different cardiometabolic and health effects of each NNS type.”

Dr. Shu Wen Ng

“Further improvements to the Nutrition Facts Panel, including the amounts of NNS when present in products, would allow for monitoring our exposure to these additives so we can better assess their potential harms or benefits to health,” said Ng.

Previous observational studies have linked NNS consumption to increased body weight, Type 2 diabetes and other adverse cardiometabolic outcomes — but other studies have found the opposite effect. Results from randomized controlled trials and meta-analyses have not demonstrated any relationship between NNS and increased consumption of sweet foods, and it remains unclear whether the inconsistency of these findings is due to studies typically categorizing all NNS together, rather than examining differences in the effect of specific NNS types.

Funding for this work comes primarily from Arnold Ventures with additional support from the Carolina Population Center NIH grant P2C HD050924.

Contact the UNC Gillings School of Global Public Health communications team at sphcomm@unc.edu.