Background Promoting healthy eating in the workplace can enhance staff productivity as employees spend 60% of their waking hours at work. The WHO recommends eating at least 5 portions of fruit and vegetables a day, but the UK falls short of this at only 3 portions a day. Posters and leaflets are routinely used in health promotion but are rarely compared to other interventions such as price subsidisation. In a cafe serving the Nuffield Department of Population Health, University of Oxford, we found that the average number of fruit purchases was 10–15 per day out of 300 transactions. Our aim was to investigate the impact of a price subsidy on fruit consumption within the Department.

Methods We selected two similar sized cafes for the intervention and control arms. The study extended over a 4 week period, with data collection occurring throughout. We initially collected baseline data for both cafes (Week 1), then introduced health promotion posters, leaflets and changing the fruit display (Week 2). Subsequently, we introduced a price subsidy in the intervention cafe only (Week 3), reducing the price of fruit from 45p to 20p for one week. We removed all interventions and monitored both cafes for a washout period (Week 4). Outcomes measured were items of fruit sold per day and total number of transactions in each cafe.

Results Baseline fruit sales were 10–15 pieces per day, accounting for 3–5% of transactions in both cafes (Week 1). Health promotion activities (Week 2) resulted in no significant change to fruit sales. Fruit subsidisation (Week 3) in the intervention cafe increased fruit sales from 4% to 28% which was strongly statistically significant at the 95% level (p < 0.001). Fruit sales also increased in the control cafe, despite no subsidisation, although this was not statistically significant. After removal of all interventions (Week 4), both cafes decreased in daily fruit sales but continued to show levels about the original baseline (intervention cafe p = 0.025, control cafe p = 0.04).

Discussion Health promotion activities had no significant impact on fruit sales. However, the addition of a price subsidisation led to a dramatic increase in sales which was strongly statistically significant. Both cafes demonstrated an ongoing increase in fruit sales after removal of the intervention. Although we cannot identify the distribution or change in individual buying practices, this study demonstrates the potential impact of price subsidisation on fruit sales in a relatively health-aware population.