Mobley 2012.

Study characteristics
Methods	Trial name: HEALTHY Study design: cluster‐RCT for implementation outcome Intervention duration: about 3.5 years (autumn 2006 to spring 2009) Length of follow‐up from baseline: about 3.5 years Differences in baseline characteristics: stated as similar but not shown Unit of allocation: schools Unit of analysis: schools
Participants	School type: middle schools Region: 10 school districts located at 7 sites across the US Demographic/socioeconomic characteristics: schools were serving largely minority and lower income populations Inclusion criteria ‐ Student body was representative of the adolescent population at risk for type 2 diabetes, defined as either ≥ 50% minority (African American, Hispanic/Latino, American Indian, or a combination) or > 50% eligible for free or reduced lunch, or both. ‐ Annual student attrition from all causes ≤ 25%. ‐ Expected cohort size at end of study was ≥ 50 per school. ‐ School authorities were willing to accept randomisation of an individual school to intervention or control. If a school was assigned to the intervention programme, this meant that the school must have arranged tasks/requirements needed to comply with the trial protocol. ‐ School permitted to perform grade‐wide collection of height, weight, gender, age and race/ethnicity at baseline. ‐ The school assisted with mass mailings of study materials to students' homes. ‐ The school district possessed or obtained Federal Wide Assurance to conduct research. ‐ Appropriate school authorities agreed to adhere to the protocol. Number of schools allocated: 42 (21 intervention, 21 control) Numbers by trial group n (controls baseline) = 21 n (controls follow‐up) = 21 n (interventions baseline) = 21 n (interventions follow‐up) = 21 Recruitment ‐ Each site recruited 6 middle schools that were randomised into intervention or control. ‐ Students provided parental informed consent. Recruitment rate: not reported
Interventions	Number of experimental conditions: 2 (1 intervention, 1 control) Policies, practices or programmes targeted by the intervention The following nutritional goals were set for the NSLP, SBP and à la carte school food services ‐ Lower the mean fat content of food served in schools. ‐ Serve ≥ 2 servings of fruit or vegetables (or both) per student on NSLP and ≥ 1 serving per student on SBP each day. ‐ Serve all dessert and snack foods with ≤ 200 kcal per single size serving or package, or both. ‐ Eliminate milk > 1% fat, all other added sugar beverages, and 100% fruit juice (100% fruit juice may only be served as ≤ 6 ounces as part of SBP). ‐ Serve ≥ 2 servings of high fibre (≥ 2 g of fibre per serving) grain‐based foods or legumes (or both) per student on NSLP and ≥ 1 serving per student on SBP each day. Implementation strategies EPOC: educational meetings ‐ Provision of staff training (Food Services Manager and staff). ‐ Educational events held during lunchtime in and around the cafeteria. EPOC: educational games ‐ Conducted 'taste tests' of new products and unfamiliar foods, including conducting comparison of available items. EPOC: external funding ‐ Intervention schools received USD3000 per year to defray expenses and potential loss of income and received USD125 for cafeteria enhancements, and to attend training. EPOC: tailored intervention ‐ Research staff worked with food service managers to identify barriers and develop solutions for schools to achieve selected goals. EPOC: educational materials ‐ Curricula, posters, brief messages displayed near serving lines. EPOC: educational outreach ‐ Research staff met weekly with food service managers to observe the food environment and to plan and support goal achievement. EPOC: the use of information and communication technology ‐ Engagement with social marketing experts to generate content and offer guidance on the school social marketing efforts. EPOC: other ‐ Intervention launch and finale; these events were designed to promote global awareness for the programme at participating schools. Family outreach including delivery of newsletters and materials. ‐ Meetings with district level staff and buyers who procure food and with food distributors, to solicit support for change. Theoretical underpinning: social‐ecological models Description of control: control school followed existing school district standards and guidelines.
Outcomes	Outcome relating to the implementation of school policies, practices or programmes 12 scores across the following variables: ‐ Lower than average fat content ‐ Serve 2 servings of fruit and vegetables ‐ Serve all desert and snack foods with < 200 kcal ‐ Eliminate milk > 1% fat ‐ Serve ≥ 2 servings of high fibre Data collection method ‐ Data collected by trained staff not involved in the intervention. ‐ Nutrition data were extracted from food service management source documents maintained by school food service personnel. ‐ Mean food group servings and nutrient amounts served per day were calculated by the trained programme staff using the food service staff records. Validity of measures used: not reported; however, the measures used were objective. Outcome relating to cost: yes Data collection method: by income statements, federal meal records and sales data Validity of measures used: not reported Outcome relating to adverse consequences: adverse impact on scholastic performance. Data collection method: state accountability tests and the total number and passing rates of students taking the test. Grade and school level data were recorded – no individual student data were collected. Validity of measures used: not reported Outcome relating to child diet, PA or weight status: assessed self‐reported dietary intake (energy, macronutrient and grams consumed of selected food groups). Height, waist circumference and BMI. Data collection method Dietary intake: Block Kids Questionnaire Anthropometry: height (Perspective Enterprises PE‐AIM‐101 stadiometer) and weight (SECA Alpha 882 and SECA Large Capacity 634 electronic scales) were measured without shoes. Waist circumference was taken using a tape measure on bare skin measured just above the iliac crest. Validity of measures used Dietary intake:not reported Anthropometry: valid
Notes	Notes: this trial also contained a PA component as part of their policy, practice or programmes implemented; however, the trial was downgraded to a nutrition trial only as it did not report implementation outcomes for PA. Research funding: National Institute of Diabetes and Digestive and Kidney Diseases and the National Institute of Health grant numbers U01‐DK61230, U01‐DK61249, U01‐DK61231, and U01‐DK61223 to the STOPP‐T2D collaborative group. Conflicts of interest: 1 author received consulting fees from McDonald's Global Advisory Committee and another received consulting fees from General Mills and ConAgra Foods. The remaining authors declared no conflict of interest.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Schools were randomly allocated to experimental group. Random sequence generation procedure was not described.
Allocation concealment (selection bias)	Unclear risk	No information provided about allocation concealment and, therefore, it was unclear if allocation was concealed.
Blinding of participants and personnel (performance bias) Implementation outcome	High risk	Outcome group: all Given the nature of the intervention, participants and study personnel were likely to have been aware of study allocation and, therefore, high risk of performance bias. Outcome: child BMI, adverse effects Low: objectively measured.
Blinding of outcome assessment (detection bias) Implementation outcome	Low risk	Outcome group: à la carte, nutrition goals, food group servings, adverse effects. Data were collected at baseline and end of study by trained study staff not involved in implementing the intervention.
Incomplete outcome data (attrition bias) Implementation outcome	Low risk	Outcome group: all outcomes All 42 schools were retained; therefore, risk of attrition bias was low. Loss to follow‐up among students similar (about 28%) for intervention and control groups at 2 years. No mention of ITT analysis in manuscripts.
Selective reporting (reporting bias)	Low risk	Protocol was available and outcomes reported were consistent.
Other bias	Low risk	Appeared free from other bias.
Overall risk of bias assessment	Low risk	Most domains were at low risk of bias.