Rejection thresholds for sweetness reduction in a model drink predict dietary sugar intake

Abstract

Links between individual sweetness perception and sugar intake have been previously explored, with highly contrasting results. The current study aims to test individual-based rejection thresholds (RjTs) to ascending and descending sucrose concentrations in a model drink, and to evaluate potential links between these measures and sugar intake behaviour in real life. Sixty-three participants (final N = 43 [f = 29]; 20 ± 5 years) each undertook three sensory testing sessions and completed a 4-day weighed food records. Each individual was tested for their Sweet Liker status, detection thresholds to sucrose, RjTs to 5 descending and 5 ascending sucrose concentrations in an orange-flavoured drink, as well as sugar intake in daily diet and from sweetened beverages. The results demonstrate that RjTs can be estimated an individual level. With each participant completing 4 replications of the RjT task, data from over 88% of participants produced a reasonable individual-based psychometric function. Intriguingly, descending, but not ascending, RjTs were positively correlated with sweet liking (r = 0.43, p < 0.01), and acted as a significant predictor of both overall dietary sugar intakes (R2 = 0.10, F(4,38) = 2.23, p = 0.050) and sugar intakes from sweetened beverages (R2 = 0.21, F(4,38) = 3.73, p = 0.012). While sweet likers (SL) and dislikers (SD) had comparable ascending RjTs (acceptance for sugar addition), they differed substantially in terms of descending RjTs (p = 0.002; acceptance for sugar reduction), with SL rejecting samples having > 47% sugar reduction (47 g/L), whereas SD rejecting samples with > 68% reduction (28 g/L). Overall, these findings suggest that descending RjTs is a useful measure for detecting a sensory-diet relationship, and should be considered to better inform sugar reduction targets and facilitate policies on product reformulation.