Martinez‐Andrade 2014

Methods	Study design: Cluster‐randomised controlled trial Funding: Not reported
Participants	Description: Children aged 2 to 5 years at 4 primary care clinics and their parent N (Randomised): 4 primary care clinics, 306 children Age: Child (mean): Intervention = 40.1 months, Control = 41.1 months Parent (mean): Intervention = 29.3 years, Control = 29.5 years % Female: Child: 47% Parent: not specified SES and ethnicity: Education: no schooling = 0.3%, Primary school = 8.9%, Junior high = 33.7%, High school = 39.3%, Professional school – 12.5%, Postgraduate = 1.7% Inclusion/exclusion criteria: Inclusion criteria: “Participants comprised children aged 2 ‐ <5 years of age whose BMI (calculated as weight in kilograms divided by height in meters squared) was above the median for age and sex (BMI z‐score 0 ‐ 3); who attended one of the participating IMSS clinics during the recruitment period for pediatric care, vaccination, or accompanying a family member; and whose parent or caregiver gave written consent to participate.” Exclusion criteria: “Families were excluded if they planned to move residences or change primary care clinics during the study period; the child had motor limitations (e.g., physical disability or delay); or required a special diet by medical indication.” Recruitment: “The project manager approached the directors of the 6 primary care clinics in Mexico City with the greatest proportion of preschoolers (approximately 5% children <5 years) to request their support for the project.” Recruitment rate: Primary care clinic = 67% (4/6) Child = 10% (306/3095) (using number of participants approached as denominator) Region: Mexico City
Interventions	Number of experimental conditions: 2 Number of participants (analysed): Intervention = 168, control = 138 Description of intervention: Intervention participants received a 6‐week curriculum focused on obesity awareness and prevention. 5 aspects dealt with throughout the 6 sessions: 1) Dietary culture, risk‐benefit practices; 2) The process of feeding acquisition/preparation/service/eating behaviours; 3) Physical activity habits; 4) Importance of weighing/measuring oneself and its meaning; 5) feedback and evaluations Duration: 6 weeks Number of contacts: 6 sessions (2 hrs a session) Setting: Primary care clinics Modality: Face‐to‐face, group sessions Interventionist: Nutritionist, nurse and health educator Integrity: Delivery of intervention: “To ensure fidelity, a small group of study staff (nutritionist, nurse and health educator) administered all intervention sessions and completed all screening, baseline and follow‐up assessments. No quantitative measure of delivery of intervention components” Attendance: “Only 52% (88 of the 168 who agreed to participate) attended ≥ 1 educational session (405 sessions attended in total). The total number of expected attendances at educational sessions was 1008 (168 participants attending 6 sessions each). Thus, compliance in the intervention group was 40% (405/1008) of total expected attendances. However, of the 88 receiving any intervention content, 67% (59/88) attended 5‐6 of the intended 6 workshops” Date of study: March 2012 to April 2013 Description of control: Usual‐care control ‐ received no intervention
Outcomes	Outcome relating to children's fruit and vegetable consumption: Child’s consumption of fruits and vegetables (servings per week), “staff assisted parents in completing a child Food Frequency Questionnaire (FFQ) adapted from the FFQ used to assess dietary intake among 1‐4 year old children in the 2006 Mexican National Nutrition Survey.” 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: 3 and 6 months Length of follow‐up post‐intervention: 1½ and 4½ months Subgroup analyses: None Loss to follow‐up (at 1 ½ and 4 ½ months): Intervention = 41%, 35% Control = 26%, 26% Analysis: Adjusted for clustering Unknown if sample size calculation was performed
Notes	First reported outcome (fruit servings/week) at the longest follow‐up < 12 months (3 months after intervention completion ‐ as 6‐months follow‐up did not report retention values by group) was extracted for inclusion in meta‐analysis The reported estimate which adjusted for clustering assessed change from baseline, we therefore used post‐intervention data and calculated an effective sample size using ICC of 0.016 to enable inclusion in meta‐analysis Sensitivity analysis ‐ primary outcome: Fruit or vegetable intake listed as primary outcome
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	A computer‐generated randomisation list designed by a statistician with no connection to the intervention was used for random allocation to experimental group
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	Child dietary intake (parent‐reported): “Only after informed consent did participants learn of their treatment assignment”. There is no blinding to group allocation of participants at follow‐up described and this is likely to influence performance
Blinding of outcome assessment (detection bias) All outcomes	High risk	Child dietary intake (parent reported): “Only after informed consent did participants learn of their treatment assignment”. There is no blinding to group allocation of participants at follow‐up described and because self‐reported measures were used this is likely to influence detection bias
Incomplete outcome data (attrition bias) All outcomes	High risk	“Non‐participation was greater in the intervention (75 (45%) of 168 participants) than in the usual care (42 (30%) of 138 participants) arm (Figure 1).” Attrition rate was high with >35% of families not completing follow‐up at 3 months. Multiple imputations were performed to address missing data however non‐participation was greater in the intervention than in the usual care condition
Selective reporting (reporting bias)	Low risk	The primary outcomes reported in the paper align with those specified in the trial registration
Other bias	Unclear risk	There were baseline imbalances between the groups, but results were adjusted. Unclear risk of recruitment bias as individuals were recruited to the trial after clusters have been randomised