Impact of Different Front-of-Package Labelling Models on Food Choices Among Residents in Shanghai, China: A Randomized Controlled Trial

Zhuo Sun¹Xiao Chen²Jiangyue Yu¹Wei Lu¹Zehuan Shi¹Kunyi Huang³Liping Shen¹Wenqing Ma¹Shupeng Mai¹Qi Song¹Zhengyuan Wang^1*Jiajie Zang¹

• ¹Shanghai Municipal Center for Disease Control and Prevention (SCDC), Shanghai, China
• ²Shanghai Putuo Center for Disease Control and Prevention, Shanghai, China
• ³School of Public Health, Shanghai University of Traditional Chinese Medcine, Shanghai, China

Background: Increased convenience food consumption poses considerable public health concerns because of its high oil, salt, and sugar content. The World Health Organization endorses front-of-pack labelling (FOPL) as a strategy to enhance consumer choice towards healthier options. Despite their global adoption, the efficacy of various FOPL systems in China's unique market remains unexplored. Here, we evaluated the impact of different FOPL models on the food purchasing behaviours of Shanghai residents.Methods: This randomized controlled trial included 7,346 respondents, who were randomly assigned to one of four FOPL groups: Nutri-Choice (NC), Nutrition Information Panel (NIP), Comprehensive Nutri-Choice (CNC), and Warning Label (WL). An online questionnaire was used to collect the respondents' demographic profiles. The sensory perceptions, nutritional ranking ability, and intention to purchase healthy foods of the respondents were evaluated using a simulated shopping experiment.Results:In total, 88.03% of the respondents supported FOPL use in China. Respondents completed the NC and WL questionnaires more quickly than the CNC or NIP questionnaire, with the NIP questionnaire taking the longest to complete (Z = 24.209, df = 3, p < 0.001).Sensory perception total approval rates were significantly higher for NC (57.01%), CNC (58.04%), and WL (57.77%) than for NIP (43.33%; χ ² = 112.958, df = 3, p < 0.001).Similarly, the total accuracy rate in the nutritional ranking was significantly higher for NC (69.94%) and CNC (72.45%) than for WL (55.44%) and NIP (32.99%; χ²= 737.823, df = 3, p < 0.001). The purchasing intentions were the healthiest for NC (80.09%), followed by CNC (83.10%)-with both significantly outperforming NIP (75.11%) and WL (73.42%; χ ² = 63.360, df = 3, p < 0.001).NC and CNC potentially improve consumers' understanding of packaged food nutritional data, encouraging healthier purchasing decisions. Our results provide a scientific rationale for formulating and implementing an impactful FOPL initiative in China.