Alaimo 2015.

Study characteristics
Methods	Trial name: School Nutrition Advances Kids (SNAK) Study design: non‐randomised Intervention duration: 2 overlapping cohorts: 1 year 9 month per cohort, 2 years 9 months both cohorts Cohort 1: 2007–2009 Cohort 2: 2008–2010 Length of follow‐up from baseline: follow‐up occurred in the next school year. Students in cohort 1 completed the baseline survey between November 2007 and March 2008 and the follow‐up survey between October and December 2008. Cohort 2 students completed baseline surveys between November 2008 and February 2009 and the follow‐up survey between October 2009 and January 2010. Differences in baseline characteristics: no differences among intervention groups with regard to school characteristics at baseline. There were some differences at baseline with regard to student dietary intake (all subsequent analyses adjusted for baseline dietary values). Table 2 reported difference at baseline in school characteristics. Some differences were apparent in school location and kitchen type but no P values were reported. Unit of allocation: school Unit of analysis: school
Participants	School type: middle schools (seventh and eighth grades) Region: Michigan, USA Demographic/socioeconomic characteristics: low‐income middle schools (mean percentage of students eligible for free/reduced‐price meals for schools in the SNAK project was 68% (range 50–98%), which was similar to all Michigan low‐income middle schools in 2007 (72%; 514 schools). Inclusion/exclusion criteria Inclusion ‐ Having ≥ 50% of students eligible for free or reduced‐price meals and having seventh and eighth grades within the same building (for follow‐up purposes) ‐ Having ≥ 50% of students eligible for free or reduced‐price meals ‐ Having seventh and eighth grades within the same building Number of schools allocated Schools:75 (54 intervention, 21 control) Students:1777 seventh graders (completed baseline) 65 (3 intervention groups, 1 control group stated, but result reported according to 2 groups: intervention and control). Numbers by trial group: study contained 1 control and 3 intervention groups. Intervention groups consisted of 3 different programmes (or programme combinations) 1. the HSAT, 2. the HSAT plus the SNAK Team and 3. the HSAT plus the Michigan State Board of Education (MSBE) nutrition policy. The sample sizes by trial group follow: n (controls baseline) = 21 n (controls follow‐up) = 20 completed some aspect of the project. 17 completed FFQ n (interventions baseline) = 54 n (HSAT baseline) = 24 n (HSAT + SNAK baseline) = 5 n (HSAT + MSBE POLICY baseline) = 25 n (interventions follow‐up) = 45 completed some aspect of the project. 38 completed the FFQ n (HSAT follow‐up) = 18 completing some aspect and 16 with FFQ n (HSAT + SNAK follow‐up) = 5 completing some aspect and 4 with FFQ n (HSAT + MSBE POLICY follow‐up) = 22 completing some aspect and 18 with FFQ Recruitment Schools:recruited through an application for small grant funding with award values ranging from USD2000–4600 with recruitment methods included direct mailings, e‐mails and telephone calls to eligible schools, as well as a posting on the Michigan Team Nutrition website. Students:written parental consent and student assent Recruitment rate Schools: denominator unknown Students: 20.6%
Interventions	Number of experimental conditions: 4 (3 intervention, 1 control, but reported according to 2 groups: intervention and control) Policies, practices or programmes targeted by the intervention The 3 policies, practices or programmes implemented in this study were the HSAT programme, the SNAK programme and the MSBE nutrition policy. ‐ Schools completed the HSAT concerning healthy eating and nutrition, and developed an action plan to improve school‐nutrition practices. ‐ Seventh‐grade student teams were formed in SNAK wherein the students implemented nutrition education and marketing. ‐ The MSBE is a nutrition policy recommending that schools offer and promote healthy foods and beverages in all competitive venues. Implementation strategies EPOC: tailored interventions ‐ The HSAT consisted of a tailored intervention (online assessment and action planning process). EPOC: educational outreach visits ‐ Provision of a facilitator (1 time to complete HSAT action plan). ‐ SNAK intervention group also received: facilitator/student meetings to assess student nutrition environment and policies. EPOC: external funding ‐ Incentives (USD1000 to implement nutrition education or implement aspects of their action plan). ‐ SNAK intervention group also received: incentives: USD1000 for students to implement student nutrition action plan. ‐ Curriculum developed for facilitator/student meetings. ‐ MSBE nutrition policy intervention schools were also received: USD1500 to compensate for any loss to food service revenue. EPOC: local consensus processes ‐ Co‐ordinated School Health Team EPOC: clinical practice guidelines ‐ Implementation of policy (2003 MSBE Healthy Food and Beverage Policy) in cafeteria à la carte lines (during second year) EPOC: educational materials ‐ Guidance documents and assistance were provided to schools, food service staff (MSBE group). Theoretical underpinning: not reported. Description of control: control group schools participated only in data collection during the study period and were offered the HSAT intervention after the last data collection point.
Outcomes	Outcome relating to the implementation of school policies, practices or programmes ‐ Mean Nutrition Policy Change Score (range 0–6) ‐ Mean Nutrition Education or Practice Change Score (or both) (range 0–14) Data collection method: survey: the Middle‐School SEPS was completed either online or by paper (took approximately 30 minutes to complete). There were 2 versions of the survey: 1 for administrators/principals and 1 for food service directors/kitchen managers. Validity of measures used: not reported/self‐report methods. The authors commented that the tool was trialled to establish face and content validity; however, the tool was not subjected to rigorous validity testing. Outcome relating to cost: not reported Outcome relating to adverse consequences: not reported Outcome relating to child diet, PA or weight status: student‐level dietary intake Data collection method: the Block Kids Food Frequency Questionnaire 2004 (ages 8–17 years) at baseline and follow‐up Validity of measures used: not reported
Notes	Research funding: Robert Wood Johnson Foundation's Healthy Eating Research Program, the Michigan Department of Community Health, and the USDA Supplemental Nutrition Assistance Program–Nutrition Education, supported by the Michigan Department of Human Services under contract numbers ADMIN‐07‐99010, ADMIN‐08‐99010, and ADMIN 09‐99010. Conflicts of interest: authors reported no competing financial interests.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Non‐randomised. Schools were allocated based on preference and were not randomly assigned to group resulting in high risk of selection bias.
Allocation concealment (selection bias)	High risk	Non‐randomised trial and no indication that allocation was concealed resulting in high risk of selection bias.
Blinding of participants and personnel (performance bias) Implementation outcome	High risk	Outcome group: all No blinding or incomplete blinding, and the outcome was likely to be influenced by lack of blinding.
Blinding of outcome assessment (detection bias) Implementation outcome	Unclear risk	Outcome group: number of school‐initiated nutrition policy and practice changes. No blinding of outcome assessment, and the outcome measurement is likely to be influenced by lack of blinding; self‐report considered high risk. Outcome group: mean nutrition policy change/nutrition education or practice change (SEPS). No blinding of outcome assessment, and the outcome measurement is likely to be influenced by lack of blinding; self‐report considered high risk. Outcome group: cafeteria à la carte and vending offerings. Self‐reported data from food service directors/other food service personnel and school administrators/principals. No blinding of outcome assessment, and the outcome measurement was likely to be influenced by lack of blinding; self‐report considered high risk. Outcome group: process data. Unclear if personnel conducting analysis of various process data sources blinded.
Incomplete outcome data (attrition bias) Implementation outcome	Low risk	Outcome group: practice and policy. Missing outcome data balanced in numbers across intervention groups, with similar reasons for missing data across groups. Outcome group: changes in à la carte/vending during lunch. 18% missing/unreliable data overall; however, relatively balanced across groups. Outcome group: number of school‐initiated nutrition practice changes; and number of school‐initiated nutrition policy changes. 5% missing data overall (3 schools), relatively balanced across groups but 2 schools missing from HSAT‐only group.
Selective reporting (reporting bias)	Unclear risk	No study protocol, therefore, it was unclear if there was selective outcome reporting.
Other bias	Low risk	Appeared free from other bias.
Potential confounding	Unclear risk	Insufficient information to determine the risk of potential confounders.
Overall risk of bias assessment	High risk	Most domains were at unclear or high risk of bias.