Smith 2017

Methods	Study design: Cluster‐randomised controlled trial Funding: Not reported
Participants	Description: Low socio‐economic children aged 3‐5 years attending Head Start preschools in Marion County, Ohio N (Randomised): 4 Head Start centres, 240 children Age: “All clusters combined had a total of 80 (38.3%) three year old children, 116 (55.5%) four year old children, and 13 (6.2%) five year old children in the study sample.” % Female: Access‐only cluster = 54%, access + education = 45%, control = 55% SES and ethnicity: Low socio‐economic “There were 9 (4.3%) Hispanic children, 152 (72.7%) white children, 36 (17.2%) multi‐racial, and 12 (5.7%) black children in the study sample” Inclusion/exclusion criteria: No explicit inclusion criteria Exclusion criteria: “Children or parents were excluded if a medical issue prohibited them from participating in the study. Children who were unable to eat solid foods were asked not to participate in this study. Children with chronic diseases, such as diabetes, were excluded from the study, as children with chronic diseases are known to have reduced carotenoid concentrations” Recruitment: “Purposive sampling was the method chosen for this study”. Parents were approached about consenting to the study at various meetings or when parents were dropping off or picking up their children. Recruitment rate: 83% (240/290) Region: Marion County, Ohio
Interventions	Number of experimental conditions: 3 Number of participants (analysed): Access only: 61 Access + education: 82 Control: 66 Description of intervention: Access only: “received the take home weekly fruits and vegetables, without the educational intervention.” Access + education: “received weekly take home fruits and vegetables, education for the children, and supplemental materials, such as newsletters and recipes, for the families about the produce being provided.” The Supplemental Nutrition Assistance Program Education (SNAP‐Ed) was provided each week. “The Harvest for Healthy Kids curriculum was used and each week the focus was on a high carotenoid fruit or vegetable. Storybooks, activities such as making pumpkin pudding in a bag, and tastings were the foundation of the class sessions.” Duration: 8 weeks Number of contacts: 8 Setting: Preschool + home Modality: Access only: provision of fruit and vegetable Access + education: multiple (provision of fruit and vegetable, face‐to‐face education, written materials) Interventionist: Access only: unclear Access + education: Supplemental Nutrition Assistance Program Education (SNAP‐Ed) programme staff member delivered education Integrity: No information provided Date of study: October‐December 2016 Description of control: “the control group did not receive either the produce or education during the eight weeks.” “The group received education following the study.”
Outcomes	Outcome relating to children's fruit and vegetable consumption: Consumption of fruit and vegetable consumption measured by carotenoid levels in the skin Outcome relating to absolute costs/cost‐effectiveness of interventions: Not reported Outcome relating to reported adverse events: Not reported Length of follow‐up from baseline: 8 weeks Length of follow‐up post‐intervention: Immediately Subgroup analyses: None Loss to follow‐up: Access only: 18% Access + education: 10% Control: 12% Analysis: Unclear if adjusted for clustering Sample size calculations performed
Notes	We pooled the access + education intervention arm compared to the no‐intervention control group in meta‐analysis of multicomponent interventions. We described the access‐only intervention compared to the no‐intervention control group narratively.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	“Site clusters were randomly assigned to one of the treatment or control groups”. Randomly allocated to experimental group but the random sequence generation procedure is not described.
Allocation concealment (selection bias)	Unclear risk	There is no information provided about allocation concealment and therefore it is unclear if allocation was concealed.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Participants and study team were not blinded. Parent self‐reported survey on fruit and vegetable consumption and therefore at high risk of performance bias.
Blinding of outcome assessment (detection bias) All outcomes	High risk	Parents were not blinded which may have affected how they responded to the survey. Parent self‐reported survey on fruit and vegetable consumption and therefore at high risk of detection bias.
Incomplete outcome data (attrition bias) All outcomes	Low risk	There were 31/240 withdrawn (27 from intervention, 4 from control).
Selective reporting (reporting bias)	Unclear risk	No trial protocol is available
Other bias	Unclear risk	There appears to be baseline imbalance between groups with differences between groups on child age and race. Analysis does not appear to account for clustering.