What works to improve school lunch nutritional quality – legislation or self-audit?

OBJECTIVE: Sweden updated its legislation on universal free school meals in 2011 and nutrition was explicitly mentioned. The current study (i) describes cross-sectional changes in school lunch nutritional quality during the following eight years and (ii) examines if repeated self-auditing, using a fully automated, online tool (School Food Sweden), based on the implementation strategy of audit and feedback, was associated with improvements. DESIGN: Both repeated cross-sectional and longitudinal design. Factors associated with meeting nutritional criteria were examined using variance weighted least squares regression and logistic regression. SETTING: Sweden. PARTICIPANTS: Primary schools who self-selected to audit meal quality between March 2012 and July 2019. RESULTS: Almost half of all (ca 4800) primary schools signed up to use the tool and 1500 audited nutritional quality at least once. Repeated cross-sectional analyses showed the proportion meeting the nutritional criteria increased significantly between 2012/13 (11 %) and 2018/19 (34 %). Longitudinally, each additional audit completed increased the odds of meeting the nutritional criteria by 1·30 (CI 1·20, 1·41), controlling for region and time elapsed since the legislative change. In 774 schools with repeat audits, both number of audits and frequency of accessing feedback predicted meeting the nutritional criteria (OR 2·02, CI 1·23, 3·31), even after adjusting for time since the legislative change and days elapsed since previous audit. CONCLUSIONS: Both legislation and self-audit with automatic feedback appear effective in helping schools to improve school meal quality. Self-audit with feedback may be an effective complement to legislation, or a promising alternative in settings where regulation is not an option.