The independent and combined effect of price increase and price labelling on sugar-sweetened beverage vending machine sales in an Australian university: a factorial randomised controlled trial

Abstract Background No studies have examined how price increases on unhealthy drinks in university vending machines impact customer purchasing or vendor financial implications. Vending machines often display prices only upon product selection, hindering consumers’ price evaluation. Objective To assess the independent and combined effects of (i) a 20% price increase on unhealthy drinks, and (ii) prominent labels displaying the price below each product, on the healthiness of customer purchases and revenue in university vending machines. Design Two-year 2 × 2 factorial cluster randomized controlled trial (March 2020 to February 2022). Machines were clustered by proximity. Itemised weekly sales data were collected for 33 weeks pre-trial and 103 weeks during the trial. Six-monthly audits assessed treatment maintenance. Participants/setting Sixty vending machines across four campuses of a university in Victoria, Australia, with 59 machines analyzed. Intervention Machine groups were randomly allocated to four conditions: (i) 20% price increase on unhealthy drinks (‘Price Increase Only’); (ii) prominent labels displaying the price below each product (‘Price Label Only’); (iii) ‘Price Increase + Price Label’; and (iv) no price increase or labels (‘Control’). Main outcome measures The primary outcome was the change from pre-trial mean of volume percentage of unhealthy drinks sold per machine per week. Secondary outcomes included volume percentage of healthier drinks sold per week and revenue. Statistical analyses Sales data were analysed using linear mixed models with machine cluster and machine as random effects; and fixed effects for price, label, price × label, pre-trial mean outcome, teaching period, COVID-19 lockdowns, and season. Results Mean volume percentage of unhealthy drink sales was higher for the ‘Price Label Only’ compared to ‘Control’ condition (6.34 percentage points, 95%CI: 1.85, 10.82), percentage of unhealthy drink units sold, and energy and total sugar content of drinks sold, and lower for volume and unit percentage of unhealthy drink sales. No other significant differences were observed for the primary or secondary outcomes. Conclusions This study in university vending machines found no impact of a 20% unhealthy drink price increase, and an increase in volume sales of unhealthy drinks in response to prominent price labelling only. Trial registration Australian New Zealand Clinical Trials Registry Number: ACTRN12620000206921 (prospectively registered 20 February 2020).