Childhood obesity is a national epidemic that disproportionately burdens low income and ethnic minority populations. By preschool, nearly one-third of low income children are already overweight or obese, setting the stage for adverse health outcomes over their life course. Evidence is mounting that individual-focused, single-component interventions fail to produce long-term population-level changes in obesity-related outcomes. Evidence increasingly supports creating entire environments (e.g., schools) in which the healthy choice is the default option.

The investigators will conduct a pragmatic cluster randomized trial to evaluate the effectiveness of a multilevel (organization, teacher, parent, child) intervention to support healthy eating and physical activity in preschools. The study is being conducted in partnership with Child Care Resource Center (CCRC), a non-profit organization dedicated to ensuring that all children receive high quality preschool experiences. The investigators will randomly assign 60 preschool sites (stratified by size), located in underserved areas of Los Angeles, to the intervention or a usual practice control conditions.

The investigators will evaluate the effectiveness of the intervention on child BMI z-scores (primary outcome) and parent-reported child nutrition and physical activity (secondary outcomes). The investigators will systematically examine the implementation process to understand factors that may facilitate or hinder intervention uptake and success. The findings from this work will be critical for informing future dissemination efforts.

This study involves a 2-group design with stratified cluster randomization at the level of the preschool. The investigators will recruit 60 preschools from LA County's most vulnerable communities, stratified by size (large, small). The investigators will randomly select pairs of schools from each stratum and randomly assign one member of each pair to the intervention or control conditions. Given the 10-month instructional year (Sept. 1-June 30), preschools will be recruited in three cohorts over three instructional years (~20 schools per cohort). To control for time trends in outcomes, the randomization protocol will be implemented independently within each cohort. Baseline data collection, conducted in the first six weeks of the school year, will include surveys (site directors, teachers, parents), anthropometric measures (children), preschool policy reviews and direct observation in preschools including child-level eating behavior and physical activity.

The multilevel intervention will be initiated immediately following baseline data collection and will consist of 1) establishment of organizational policies to support healthy nutrition and physical activity, 2) integration of nutrition and physical activity content into curricula 3) supports for teachers and staff to model healthy behavior, and 4) engagement of parents as partners in implementing healthy policies in preschool and at home. Implementation measures will be collected throughout the project period. Follow-up data collection, in the last four-six weeks of the school year, will mirror baseline data collection.

Study Hypotheses:

1. Children at intervention schools will gain less weight relative to height gains compared to children at control schools, resulting in lower mean BMI z-scores at follow-up (effectiveness). Outcomes will be assessed in the overall sample and among the subset of children who are overweight or obese at baseline.
2. Intervention preschools will have greater implementation of healthy nutrition and physical activity policies and practices at follow-up compared to control preschools.

The primary effectiveness outcome will be child age-and sex-adjusted BMI z-score. The investigators will obtain baseline data on 1,630 children, 815 per condition, 60 preschool site directors and approximately 135 teachers. Estimating 80% child retention at follow-up, the investigators anticipate having follow-up data on 1,304 children, 652 per study arm. Data will also be collected from the parents of participating children at baseline and follow-up.

Inclusion Criteria:

Preschool Sites:

• Must be in CCRC network of preschools
• Have 20 or more preschool aged children (2.5-5 years old) in attendance at site
• Children must eat at least one meal on-site per school day
• Must have at least one preschool classroom

Providers (Teachers and Preschool Site Directors):

• Teachers must teach a preschool class at participating preschool
• Be able to provide informed consent

Parents and Children:

• Child must be 2.5-5 years old and enrolled in the participating preschool
• Child primary caregiver must have at least one child 2.5-5 years old that is enrolled in the preschool
• Child primary care giver must be able to read and write in English or Spanish

Exclusion Criteria:

Preschool Sites:

• Have less than 20 preschool aged children (2.5-5 years old) enrolled.
• Do not have a preschool classroom
• Are not in the CCRC network
• Preschools that are based in a family-home or serve special needs children exclusively
• Children do not eat at least one meal on-site per school day

Providers (Teachers and Preschool Site Directors):

-Teachers not currently teaching a preschool class

Parents and Children:

• Child primary care giver unable to read or write in English or Spanish
• Primary care giver who does not have a child currently enrolled in a preschool class

All participants for whom valid height, weight, age, and sex were collected at baseline and/or follow-up. The discrepancies between Baseline Participants and results provided in the Baseline Analysis Population Description vary due to missing and/or invalid data presented across demographic categories.

Over the 4 year study period, we recruited 4 cohorts of participants. The period of time over which adverse event data was collected averaged between 8-10 months or the length of the intervention.

This study implemented a multi-level obesity prevention intervention in preschools that posed no risk for serious adverse events or all cause mortality. We monitored for "other" or non-serious adverse events.

The impact of COVID-19 led to fewer number of participants analyzed in the intervention and control conditions.