The nutrient profile of the beverage market: associations with national-level sugar-sweetened beverage taxes and nutrition labelling policies

Abstract

Background A healthy diet across the lifespan is critical to help prevent malnutrition and non-communicable diseases. However, global dietary patterns have shifted in recent decades toward increased intake of less healthful foods and beverages, which are often energy-dense and nutrient-poor. An increasing number of countries have implemented nutrition policies aimed at improving the consumer nutrition environment, such as sugar-sweetened beverage (SSB) taxes, front-of-package nutrition labels (FOPLs), and serving size regulations. While these policies may be associated with reduced sugar consumption, product (re)formulation and/or consumer understanding of nutritional information, several evidence gaps remain on their real-world impact. First, most research on SSB taxes and FOPLs has focused on within-country changes in sales or nutrient content, with less cross-country evidence examining how different policy designs are associated with beverage composition. Second, existing literature on (re)formulation has focused on sugar reduction, with fewer studies examining the presence of low/no-calorie sweeteners (LNCS), which may influence beverage composition without reducing the level of processing. Third, there is limited cross-country evidence comparing serving size formats and amounts.

Research Objectives The overall objective of this thesis was to examine how nutritional composition, sweetener use, and serving sizes of non-alcoholic pre-packaged beverages vary across countries with different sugar-related policies (i.e., SSB taxes, FOPLs, and serving size regulations). Study 1 examined differences in sugar concentrations across countries and beverage categories. Study 2 examined the proportion of beverages containing sweeteners across countries and beverage categories. Study 3 examined differences in serving size formats (‘per 100mL’, ‘per serving’, or both) and voluntary serving sizes on labels selected by brands across countries and beverage categories, when applicable.

Methods Data for the three studies were based on a market scan of non-alcoholic, pre-packaged beverages across six countries: Australia, Canada, Chile, Mexico, the United Kingdom (UK), and the United States (US). Data were collected by web scraping websites of grocery stores, food delivery platforms, and brands between October 2023 and February 2025. Data collection involved two approaches: manual web scraping (i.e., copying data from brand websites) and automatic web scraping (i.e., applying Python 3.13 scripts to grocery store and food delivery websites). Study 1 included 15,640 beverages from Australia (n=1,728), Canada (n=2,643), Chile (n=1,242), Mexico (n=1,588), the UK (n=2,598), and the US (n=5,841). Linear regression examined cross-country differences in sugar concentrations, adjusting for beverage category and retail source. Proportions of beverages within 1.0g/100mL below (or at) versus above country-specific sugar concentration thresholds from SSB tax or sugar FOPL policies were also examined. Study 2 included 15,994 beverages from Australia (n=2,105), Canada (n=2,672), Chile (n=1,229), Mexico (n=1,564), the UK (n=2,587), and the US (n=5,837). Binary regression models examined differences in sweetener presence between countries and beverage categories, adjusting for beverage category and retail source. Study 3 included 26,552 beverages from Australia (n=4,589), Canada (n=4,907), Chile (n=1,816), Mexico (n=3,430), the UK (n=3,554), and the US (n=8,256). Linear regression examined cross-country differences in serving sizes, adjusting for beverage category and retail source. Proportions examined cross-country differences in serving size formats and containers with multiple servings.

Results Study 1: Sugar concentrations differed across countries and by policy design. Adjusted mean sugar concentrations were lowest in countries with SSB tax or sugar FOPL policies triggered by sugar-concentration thresholds (the UK: 4.3g/100mL; Chile: 4.3g/100mL) versus other countries (the US: 5.2g/100mL; Mexico: 5.5g/100mL; Australia: 5.5g/100mL; Canada: 6.0g/100mL; all pairwise contrasts p<0.05, except Chile versus the UK). The proportion of beverages with sugar concentrations that were 1.0g below (or at) the 5.0g/100mL threshold used by the UK tax and Chile FOPL (versus above) was greatest in the UK (21.6-23.7%) and Chile (23.8-31.7%) versus other countries (≤12.0-15.2%). Sugar concentrations were lower in countries with threshold-based sugar-related taxes and FOPLs, with greater proportions just below (or at) thresholds.

Study 2: The percentage of beverages with LNCS was highest in countries with mandatory SSB taxes and/or sugar FOPLs (Chile: 64.7%; UK: 48.8%; Mexico: 45.3%) versus countries without mandatory policies (US: 36.1%; Australia: 24.4%; Canada: 17.5%; p<0.001). Similar patterns were observed for combined presence of LNCS and added sugars (Chile: 31.7%; UK: 27.1%; Mexico: 26.5% vs US: 17.4%; Australia: 8.4%; Canada: 7.5%; p<0.001). Patterns were less consistent for added sugars without LNCS (Chile: 45.2%; UK: 59.2%; Mexico: 58.9% vs US: 56.3%; Australia: 46.8%; Canada: 56.4%; p=0.002). Across all countries, LNCS presence was highest in energy drinks (75.2% vs ≤49.5% in other categories) and added sugars presence was lowest in water drinks (36.7% vs ≥52.9% in other categories).

Study 3: Results: Over 60% of beverages in Australia, Chile, and the UK used both ‘per 100mL’ and ‘per serving’ sizes, while nearly half in Mexico used ‘per 100mL’ only, and virtually all in Canada and the US used ‘per serving’ only. Adjusted mean serving sizes were highest in Canada (327.4mL) and the US (340.1mL) versus other countries (≤304.3mL; all contrasts p<0.001). The proportion of small containers (<355mL) with multiple servings was highest in the UK (7.9%) and Mexico (16.1%) versus other countries (≤5.6%), while proportions among moderate containers >355-600mL were highest in the UK (40.1%), Chile (67.5%), and Mexico (79.8%) versus other countries (≤20.1%).

Contribution/Significance The findings of this thesis add to the growing body of evidence on the complexity of the consumer nutrition environment. Specifically, this thesis highlighted that nutrition policies can shape how beverages are formulated across countries: mandatory policies involving sugar thresholds may incentivize (re)formulation to lower-sugar options and greater LNCS use, suggesting that these approaches do not necessarily improve overall product healthfulness. While these observed patterns are consistent with policy-related differences in product composition, the cross-sectional design does not allow causal conclusions to be drawn, and other factors may have contributed to differences across countries. Additionally, the findings suggest opportunities to standardize the use of ‘per 100mL’ and ‘per serving’ formats, as well as serving size labelling requirements for smaller and larger beverage containers, which may improve the comparability and interpretability of nutrition information across beverages. The branded beverage composition database developed as part of this thesis provides a baseline for ongoing monitoring of beverages across countries, supporting the examination of sugar-related policies aimed at promoting healthier intakes and reducing the burden of diet-related disease worldwide.