Long-term Effectiveness of a Multistrategy Behavioral Intervention to Increase the Nutritional Quality of Primary School Students’ Online Lunch Orders: 18-Month Follow-up of the Click & Crunch Cluster Randomized Controlled Trial

Background
School food services, including cafeterias and canteens, are an ideal setting in which to improve child nutrition. Online canteen ordering systems are increasingly common and provide unique opportunities to deliver choice architecture strategies to nudge users to select healthier items. Despite evidence of short-term effectiveness, there is little evidence regarding the long-term effectiveness of choice architecture interventions, particularly those delivered online.


Objective
This study determined the long-term effectiveness of a multistrategy behavioral intervention (Click & Crunch) embedded within an existing online school lunch-ordering system on the energy, saturated fat, sugar, and sodium content of primary school students’ lunch orders 18 months after baseline.


Methods
This cluster randomized controlled trial (RCT) involved a cohort of 2207 students (aged 5-12 years) from 17 schools in New South Wales, Australia. Schools were randomized to receive either a multistrategy behavioral intervention or the control (usual online ordering only). The intervention strategies ran continuously for 14-16.5 months until the end of follow-up data collection. Trial primary outcomes (ie, mean total energy, saturated fat, sugar and sodium content of student online lunch orders) and secondary outcomes (ie, the proportion of online lunch order items that were categorized as everyday, occasional, and caution) were assessed over an 8-week period at baseline and 18-month follow-up.


Results
In all, 16 schools (94%) participated in the 18-month follow-up. Over time, from baseline to follow-up, relative to control orders, intervention orders had significantly lower energy (–74.1 kJ; 95% CI [–124.7, –23.4]; P=.006) and saturated fat (–0.4 g; 95% CI [–0.7, –0.1]; P=.003) but no significant differences in sugar or sodium content. Relative to control schools, the odds of purchasing everyday items increased significantly (odds ratio [OR] 1.2; 95% CI [1.1, 1.4]; P=.009, corresponding to a +3.8% change) and the odds of purchasing caution items significantly decreased among intervention schools (OR 0.7, 95% CI [0.6, 0.9]; P=.002, corresponding to a –2.6% change). There was no between-group difference over time in canteen revenue.


Conclusions
This is the first study to investigate the sustained effect of a choice architecture intervention delivered via an online canteen ordering systems in schools. The findings suggest that there are intervention effects up to 18-months postbaseline in terms of decreased energy and saturated fat content and changes in the relative proportions of healthy and unhealthy food purchased for student lunches. As such, this intervention approach may hold promise as a population health behavior change strategy within schools and may have implications for the use of online food-ordering systems more generally; however, more research is required.


Trial Registration
Australian New Zealand Clinical Trials Registry ACTRN12618000855224; https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=375075