HCSF 1 2007.

Study characteristics
Methods	Study design: RCT Study recruitment: not reported, follow‐up 2 years Published protocol/trial: protocol published and trial not registered
Participants	Description: American Indian children aged 2–5 years (interventions n = 67, control n = 83) Exclusion criteria: children with presence of major physical or behavioural conditions that would preclude participation. Indigenous population: Bad River Band of Lake Superior Chippewa Indians, the Lac du Flambeau Band of Lake Superior Chippewa Indians, the Menominee Nation, and the Oneida Nation Setting: USA, reservation Place of delivery: Indian Health Service and Head Start sites Principle health condition: obesity Age of child (years), mean: intervention 4.0 (SD 0.9), control 4.0 (SD 0.9) Gender of child: intervention 34 (50.7%) girls, control 36 (43.4%) girls First child in family: none reported Family unit: none reported Primary carer of child: mother: intervention 58 (86.6%), control 70 (84.3%); father: intervention 2 (3.0%), control 1 (1.2%); grandparent/other: intervention 7 (6.0%), control 12 (14.5%) Socioeconomic status: none reported Employment of primary carer: none reported Education of primary carer: high school or less: intervention 15 (19.7%), control 16 (19.3%); some college: intervention 24 (35.8%), control 30 (36.1%); completed college and beyond: intervention 16 (23.9%), control 22 (26.5%); unknown: intervention 12 (17.9%), control 15 (18.1%)
Interventions	Intervention Intervention name: Mentored Health Child, Strong Families (Mentored HCSF) Intervention aim: HCSF aims to change behaviours through increased knowledge of healthy lifestyles, enhanced parenting and increased self‐efficacy. Theory used to develop intervention: HCSF is based on social cognitive and family systems theories and seeks to change behaviours at the family level. Consumer and community involvement: tool kit lessons and activities were developed by the University of Wisconsin–Madison and Great Lakes Inter‐Tribal Council research team and University of Wisconsin Extension specialists. Community members and tribal leaders were integral throughout the conceptualisation and planning of the intervention. HCSF's Supportive Communities component, which worked with 3 tribal communities to develop community advisory boards was aimed at assessing and eliminating environmental barriers to health. Overall grouping: education Fees, reimbursement, or incentives: gift cards to local merchants and lesson‐specific incentives (non‐monetary) were provided. Individual amounts varied but averaged USD 175/person. Procedures: year 1: initial contact with the family was made by telephone, and mentors were encouraged to share information to create a friendly and supportive relationship. The first in‐person lesson was designed to create dialogue between the mentor and the family and to begin building a supportive rapport. During each visit, mentors reviewed the lesson with the primary carer and child, led discussions and activities to help the carer and child learn about the topic, considered behaviour change related to the topic, and assisted the family in setting goals to attempt behaviour change. Ideally, mentor‐led discussions assisted the primary carer and child in progressing along the continuum of motivation towards actual behaviour change, while helping the primary carer build skills and confidence in his or her own ability to adopt healthier lifestyle choices. During year 1, intervention families were also invited with their extended family to 3 mentor‐led group sessions. Year 2: intervention families participated in monthly group meetings and continued to receive a monthly newsletter with parenting tips/recipes/local programme notices to help in sustaining behaviour changes implemented in year 1. Monthly group meetings focused on topics such as basic nutrition concepts (sugar, fats, appropriate serving sizes, food choice variety) and ideas for physical activities. Monthly newsletters were disseminated for the 2 years Materials: lesson‐specific incentives included an HCSF calendar to track goals and progress, cooking utensils, and physical activity items such as balls, frisbees, pedometers, exercise videos, etc. Mode of delivery: year 1: individual face‐to‐face home visits, year 2 group session When and how often was the intervention delivered? year 1: 12 toolkit lessons, 3 group lessons; year 2: 12 monthly group meetings Who delivered the intervention? mentors: tribal members or other people connected to the tribe. This included parents, grandparents, and respected community members who were able to deliver the intervention. Was there any training provided to the people who delivered the intervention? mentors were trained extensively by the University of Wisconsin Extension staff, tribal well‐ness staff (including nurses, diabetes educators, and dieticians), knowledgeable tribal elders, and HCSF research staff. Additional training was provided on child development, nutrition, and physical activity. Mentors received a full protocol manual, yearly training, and refresher sessions. The University of Wisconsin–Madison and Great Lakes Inter‐Tribal Council project co‐ordinators worked with mentors, discussing issues with in‐home visiting and families' lack of progress, and assessed mentor progress. Was the study modified or adapted? if at any time the home visits were unable to be scheduled or completed for participants, the intervention materials were provided by mail. This was not outlined in the protocol. Was the fidelity assessed? not reported Comparison Mailed group This was an education intervention. In year 1, the control families received the same 12 lessons by mail + the monthly newsletter. In year 2, the control families received only the monthly newsletter. Fees, reimbursement, or incentives: gift cards to local merchants and lesson‐specific incentives (non‐monetary) were provided. Individual amounts varied but averaged USD 175/person.
Outcomes	Overall health and well‐being: BMI z‐score Child physical health and development: BMI z‐score Family enhancing lifestyle or behaviour outcomes: 12‐item Short Form Survey (SF‐12) Physical Health Parent/carer psychological health: SF‐12 Mental Health Time points: 12 months
Funding source and conflicts of interest	Funding: Wisconsin Partnership Program Community–Academic Partnership Fund and the NIH (grant number U01 HL 087381). Author EJT was supported through an NIH T32 training grant to the University of Wisconsin Department of Nutritional Sciences (grant number 5T32DK007665). The funders had no role in the design, analysis, or writing of the article. Conflict of interest: none reported
Notes	Other outcomes recorded but not used within the review Child: waist circumference; fruit and vegetable servings/day; soda/sweetened drink and candy servings/day; hours physical activity/day; hours television viewing time/day; accelerometry Mother: waist circumference; height and weight; BMI; glucose tolerance; blood lipid profile; C‐reactive protein level; urine microalbumin level; creatinine level; nutrition and physical activity behaviours; fruit and vegetable servings/day; soda/sweetened drink and sweets (candy) servings/day; hours of physical activity/day; hours television viewing time/day; accelerometry; health behavioural efficacy
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	It is unclear how sequence generation was completed; however, there appeared to be a sufficient process of randomisation used. Quote: "Within each stratum, half of the families were randomly assigned to the intervention condition and half to the control condition. Furthermore, within each stratum, a blocked randomization strategy was used to ensure that there was an equal number of families in the intervention and control groups."
Allocation concealment (selection bias)	Unclear risk	Randomisation occurred after enrolment; no information about allocation concealment. Quote: "Randomization at the family level was done after obtaining consent from and completing baseline measurements on participating families."
Blinding of participants and personnel (performance bias) All outcomes	High risk	Participants and people delivering the intervention were not blinded to the intervention.
Blinding of outcome assessment (detection bias) All outcomes	High risk	The same people who delivered the intervention also collected the data.
Incomplete outcome data (attrition bias) All outcomes	High risk	Intention‐to‐treat analysis completed. High level of missing data with more data missing in the intervention (83%) than in the control (67%) group; intention‐to‐treat analysis. Missing imputation was completed but did not account for the high attrition rate in the intervention group.
Selective reporting (reporting bias)	Low risk	Protocol available. Data provided as per definition.
Other bias	High risk	Changes in participant allocation after randomisation Participants were moved after randomisation. This could have influenced outcomes in favour of the intervention. No sensitivity analysis was completed. No other obvious sources of bias. Quote: "After randomization, participants who were unable to be scheduled for their initial mentoring visit within two months were moved to the mailed toolkit group, resulting in a higher number of participants in this group (eight families were transferred before any intervention was administered, resulting in eighty‐three in the mailed only group instead of the seventy‐five expected after randomization)."