A color-coded beverage label in Singapore was linked to less sugar in household shopping carts and lower-sugar drink formulas, according to a new study that looked at what people bought and what companies sold before and after the policy took effect.
The study, published in the American Journal of Public Health, evaluated Singapore’s Nutri-Grade policy, a front-of-package labeling system for beverages. The before-after study used household scanner data from April 2019 to March 2024 to examine purchases of prepackaged beverages. Researchers also analyzed changes in product nutrient composition after the policy was implemented in December 2022.
Singapore’s Nutri-Grade system grades drinks from A to D based on sugar and saturated fat content. Products graded C or D must display the label, while D-graded beverages face advertising restrictions. The system is designed to make less healthy drinks easier to spot and to encourage manufacturers to reformulate products below key sugar and saturated fat thresholds.
“Nutri-Grade is doing what front-of-pack labels are meant to do: change what people buy and push companies to reduce sugar,” said Soye Shin, principal investigator and assistant professor from Duke-NUS Medical School’s Health Services Research and Population Health Programme at the time of the study.
The researchers found that after the policy was implemented, households bought prepackaged beverages with 18% less sugar, equal to 3.1 fewer grams per day. Purchases of less healthy drinks, those graded C or D, fell by 44%.
The findings suggest labels may work in two ways. They can help consumers compare drinks more quickly, and they can give companies a reason to change product formulas so drinks earn a better grade.
In this study, manufacturers appeared to respond to the policy. Overall reformulation lowered sugar content by 21%. The researchers also found a 6% increase in the use of non-sugar sweeteners across reformulated beverages.
That detail matters. Reducing sugar can be useful from a public health perspective, especially because sugary drinks are a major source of added sugar in many diets. But reformulation does not simply make the nutrition conversation disappear. Companies may replace sugar with other sweeteners, change ingredients or adjust serving patterns in ways that still need to be monitored.
“However, the market adapts simultaneously, with some products improving their grade by replacing sugar with other sweeteners,” Shin said. “This shows why labelling policies may benefit from periodic updates in response to industry practices.”
The study also found that purchases shifted away from Grade C and D beverages and toward Grade A and B beverages. The effect was seen across households, but higher-income households and those with higher nutrition knowledge showed larger reductions.
“We noted that consumers purchased less Grade C and D beverages and more Grade A and B beverages,” said Tan Yan Ning, a co-author of the study, former research assistant at Duke-NUS and now a Duke-NUS MD student.
Tan said the stronger effects among households with more nutrition knowledge suggest labels may work best when paired with broader nutrition education.
“This points to a need to strengthen nutrition literacy alongside policy measures such as Nutri-Grade labelling,” Tan said.
The study is especially useful because it looked at both sides of the market: what households bought and how companies reformulated products. Many studies of front-of-package labels focus only on consumer behavior, but labeling policies can also change the food environment by influencing what manufacturers put on shelves.
Still, the findings have limits. This was not a randomized trial, and the study design cannot prove the Nutri-Grade policy alone caused every change in beverage purchases or product formulas. Other factors, such as changing consumer preferences, pricing, marketing or broader health messaging, could have played a role.
The study also measured purchases, not what people actually drank or their long-term health outcomes. It cannot show whether the policy lowered rates of type 2 diabetes, heart disease, weight gain or other health conditions. And because the study took place in Singapore, the findings may not apply directly to the United States, where food labeling rules, beverage habits, marketing practices and public trust in nutrition policy are different.
Even with those cautions, the study points to a broader lesson: nutrition choices are not shaped by willpower alone. Labels, advertising rules, product design and nutrition literacy all affect what people notice, buy and drink.
“This study shows that mandatory labelling can shift the market at scale — changing what people buy and what companies sell,” said Lok Shee Mei, Duke-NUS' interim vice dean for research.
The researchers thanked the Singapore Health Promotion Board for supporting the data for this study, which received an institutional review board exemption from the National University of Singapore because it used secondary de-identified data.