What options do children choose for their school lunch? Time-series and cluster analyses of food choice data from a primary school in England

In England, more than a fifth of 10–11-year-olds live with obesity. Given its detrimental effects on health and well-being, addressing childhood obesity is critical, and understanding how children select foods is crucial to this. This study aimed to investigate children's selection of school lunches. Selection data (>11,000 transactions) for 155 schoolchildren (5–11 years), were collected from a primary school. The school offered three lunch options: meat/fish-based dish (MFD); a vegetarian dish (VEG); and sandwich/jacket potato (SJP). A time-series analysis revealed stable selection at the school population level, with consistent patterns across identical rounds of the menu cycle. There was also no difference in selections for younger and older children, and for pupils entitled to Free School Meals (FSM) and pupils who were not. Cluster analysis revealed four patterns of selection: Cluster 1 (n = 43) with a distinct preference for meat/fish-based dishes (76.8 % MFD, 10.5 % SJP, 12.7 % VEG); Cluster 2 (n = 42) with low selection of vegetarian dishes (50.2 % MFD, 36.8 % SJP, 13.0 % VEG); Cluster 3 (n = 40) with a tendency toward main meals (57.8 % MFD, 12.4 % SJP, 29.8 % VEG); and Cluster 4 (n = 30) with an assorted selection (36.8 % MFD, 27.9 % SJP, 35.3 % VEG). Cluster membership was not associated with age nor FSM entitlement. Findings, such as the stability of children's choices and the relative size of clusters point to the valuable insights afforded by selection data. These offer unique opportunities to examine children's behaviour within a school food environment, including in the longer term and as a means to inform school-based interventions.