Swaps are often used to encourage healthier food choices, but there is little evidence of their effectiveness. The current study assessed the impact of offering swaps on groceries purchased within a bespoke online supermarket; specifically the objective was to measure the impact on energy density (ED) of food purchases following the offer of lower ED alternatives (a) at point of selection or at checkout, and (b) with or without explicit consent to receive swap promptsMethodParticipants were asked to complete a 12-item shopping task within an online shopping platform, developed for studying food purchasing. 1610 adults were randomly assigned to a no swap control condition or to one of four interventions: consented swaps at selection; consented swaps at checkout; imposed swaps at selection; or imposed swaps at checkout. Each swap presented two lower ED options from the same category as the participant’s chosen food. Swap acceptance rate and purchased food ED were the primary outcomes.
Of the mean 12.36 (SD 1.26) foods purchased, intervention participants were offered a mean of 4.1 (SD 1.68) swaps, with the potential to reduce the ED of purchased food (effect (95 % CI): −83 kJ/100 g (−110 – -56), p = <0.0001). A median of one swap (IQR 0 to 2) was accepted, not significantly reducing the purchased food ED (effect (95 % CI): −24 kJ/100 g (4 – -52), p = 0.094). More swaps were accepted when offered at selection than at checkout (OR (95 % CI) = 1.224 (1.11 – 1.35), p < 0.0001), but no differences were seen with consent. Purchased food ED was unaffected by point of swap or consent, but reduced with number of swaps accepted (effect per swap (95 % CI) = −24 kJ/100 g (−35 – -14), p < 0.0001).
Within category swaps did not reduce the ED of food purchases reflecting the observation that the use of swaps within an on-line shopping platform offered small potential gains in ED and a minority was accepted.