Development of a method for the estimation of non-nutritive sweetener intake from beverages using daily dietary data (SWEET-EST)

Abstract

Background: The long-term health effects of non-nutritive sweetener (NNS) consumption remain uncertain, partially due to limited data on population-level intakes. Most food composition databases lack information on NNSs, and technical approaches to assess NNS content of products may be inaccessible for most researchers. Therefore, effective methods to quantify NNS content within products and estimate consumption are important for public health research. Objective: To develop a method for the estimation of NNS intake from beverages using daily dietary data. Design: SWEET-EST, a novel seven-step method for the estimation of NNS intake from beverages using daily dietary data was developed to be capable of estimating intake of different NNSs, in milligrams per day (mg/d). SWEET-EST aimed to provide flexibility to researchers, allowing adaption to different contexts and populations based on user needs, and data available to them. SWEET-EST was designed to use existing information from (i) food composition databases; (ii) contemporary ingredient lists of beverages; (iii) sweetness intensity ratings of comparable beverages; and (iv) concentration-response models relevant to specific NNSs. The method was applied to a UK paediatric dietary dataset collected via three-day diet diaries from children aged 21-months (n=2,592) with repeat measures at 7 years (n=592). Results: At 21-months, 37.0% of the sample were classified as consumers of any NNS; 46.1% were at 7 years of age. Beverages identified as sources of NNS were ‘fruit-flavoured concentrates’, ‘carbonated soft drinks’ and ‘fruit-juice based drinks’. Intake estimates (in mg/d) were obtained for aspartame, acesulfame-potassium (ace-K), saccharin and sucralose. Conclusions: SWEET-EST supports the estimation of NNS intake from beverages (in mg/d) in large population studies. This novel method opens new opportunities for the quantification of NNS intake from beverages in large population-based samples when research teams do not have access to quantitative information from ingredients lists or resources to perform additional analysis or assays.