Butte 2007.

Methods	Study design: prospective cohort study. Analyses for cohorts: analyses conducted on subsample of 798 children who gained weight after 1 year. Predictors of weight gain were individually examined using GEE. To account for correlated data within families, a family identification number was used as the cluster variable. Preliminary graphical analysis indicated that weight gain increased non‐linearly with age; thus, a quadratic term was needed. To address potential confounding between BMI status and predictors of weight gain, GEE analyses were repeated and adjusted for BMI status, age, age squared, sex and Tanner stage. How were missing data handled? Lost to follow‐up at 1 year: 14.6% (151/1030) (reasons not stated). Number of study contacts: 3 (2 baseline visits, at 1 year' follow‐up). Period of follow‐up (total period of observation): 1 year. Periods of recruitment: November 2000 to August 2004. Sample size justification adequately described? No. Sampling method: convenience sample. Recruitment conducted through local TV and radio stations and community outreach efforts. Each family was selected from an overweight proband aged 4‐19 years using bivariate ascertainment scheme (i.e. overweight ≥ 95th percentile for BMI and ≥ 85th percentile for FM). In addition, families were required to have ≥ 3 children aged 4‐19 years. Study objective: to test putative sociodemographic, metabolic and behavioural predictors of weight gain: familial characteristics, birth information, child acculturation, dietary intake, eating behaviour, physical activity, energy expenditure and fasting blood biochemistries, while controlling for sex, age and sexual maturation. Study population: children aged 4‐19 years in Hispanic community.
Participants	Baseline characteristics (reported for 1 overall group) Age (mean in years): overall (n = 1030) 10.95 (SD 4.55); boys (n = 510) 11.2 (SD 4.52); girls (n = 520) 10.7 (SD 4.56). Sex (% girl): 50.5% girls. Ethnicity: Hispanic. Education: paternal education (years): mean (SD): 8.8 (4.4); maternal education: 9.6(4.1). Fathers with 8 years of education or less (%): 42%; some high school/high school graduate: 38%; some college/college graduate: 19%. Mothers with 8 years of education or less (%): 35%; some high school/high school graduate: 43%; some college/college graduate: 22%. Income (%): < USD 19,999: 22%; USD 20, 000 to USD 29,999: 34%; USD 30,000 to USD 39,999: 25%; > USD 40,000: 19%. Pubertal stage (Tanner stage, %): stage I: 51%; stage II: 14%; stage III: 16%; stage IV: 12%; stage V: 7%. Parental BMI (kg/m2): father: 30.7 (SD 4.7); mother: 33.8 (SD 8.2). Child total energy (kJ): overall (n = 1030) 8388 (SD 2877); boys (n = 510) 9138 (SD 3054); girls (n = 520) 7653 (SD 2481); P < 0.05. Child total fat (%TE): overall (n = 1030) 33.9 (SD 6.81); boys (n = 510) 33.9 (SD 6.77); girls (n = 520) 33.9 (6.84). Child total protein (%TE): overall (n = 1030) 14.1 (SD 2.28); boys (n = 510) 14.3 (SD 2.26); girls (n = 520) 13.9 (SD 2.28). Child total CHO (%TE): overall (n = 1030) 53.2 (SD 6.81); boys 53 (SD 6.77); girls 53.4 (SD 6.84). Child physical activity (physical activity count x 10‐4/d): sedentary physical activity (%): overall (n = 1030) 37.8 (SD 13.6); boys (n = 510) 38 (SD 13.55); girls (n = 520) 37.5 (SD 13.7); light physical activity (%): overall 52.8 (SD 11.4); boys 51.7 (SD 11.3); girls 53.9 (SD 11.4); P < 0.05, boys vs girls; moderate physical activity (%): overall 9.2 (SD 6.9); boys 10 (SD 6.8); girls 8.4 (SD 6.8); P < 0.05, boys vs girls; vigorous physical activity (%): overall 0.3 (SD 0.6); boys 0.4 (SD 0.7); girls 0.2 (SD 0.5). Child physical inactivity or screen time or both (hours/day): overall (n=1030) 3 (SD 1.6). Child CVD risk (excluding fatness): SBP (mmHg): overall (n = 1030) 107.93 (SD 10.5); boys (n = 510) 110.2 (SD 11.29); girls (n = 520)105.7 (SD 9.12); DBP (mmHg): overall 50.95 (SD 6.81); boys 51.1 (SD 6.77); girls 50.8 (SD 6.84); total cholesterol (mmol/L): overall 4.45 (SD 0.91); boys 4.47 (0.9); girls 4.43 (0.91); HDL‐C (mmol/L): overall 1.21 (SD 0.23); boys 1.21 (SD 0.23); girls 1.21 (0.23); TGs (mmol/L): overall 1.58 (0.58); boys 1.22 (SD 0.45); girls 1.93 (SD 0.46). Child body fatness:BMI (kg/m2): overall (n = 1030) 25.09 (SD 8.04); boys (n = 510) 25.9 (SD 9.03); girls 24.3 (SD 6.84); P < 0.05, boys vs girls; weight (kg): overall (n = 1030) 54.41 (SD 27.6); boys (n = 510) 58.6 (SD 29.36); girls (n = 520) 50.3 (SD 25.08); P < 0.05, boys vs girls; FM (kg): overall (n = 1030) 18.95 (SD 12.51); boys (n = 510) 19 (SD 13.55); girls (n = 520) 18.9 (SD 11.4); truncal FM (kg): overall (n = 1030) 8.3 (SD 17.04); boys (n = 510) 8.1 (SD 18.07); girls (n = 520) 8.5 (SD 15.96); WC (cm): overall (n = 1030) 71.06 (SD 17.07); boys (n = 510) 79 (SD 18.07); girls (n = 520) 72.1 (SD 15.81); P < 0.05, boys vs girls; hip circumference (cm): overall (n = 1030) 87.74 (SD 20.46); boys (n = 510) 89 (SD 20.32); girls 86.5 (SD 20.52); P < 0.05, boys vs girls. Included criteria: Hispanic families with ≥ 3 children aged 4‐19 year and ≥ 1 overweight child aged 4‐19 year (overweight was defined as BMI ≥ 95th percentile and FM > 85th percentile). Excluded criteria: NR. Brief description of participants: Hispanic children aged 4‐19 years in the Viva la Familia Study enrolling families with ≥ 1 overweight child. Total number completed in cohort study: 879 (analyses conducted on 798 children). Total number enrolled in cohort study: 1030.
Interventions	Description of exposure for cohorts Time span: 1 year. Dietary assessment method used: 24‐hour dietary recalls. Frequency: 2 multiple‐pass 24‐hour dietary recalls performed on 2 random occasions (2‐4 weeks apart) at baseline. See Table 9; Table 10; Table 11; Table 12; Table 13; Table 14; Table 15; Table 16; Table 17; Table 18 for details of total fat intake exposure per outcome.
Outcomes	Weight Weight gain (kg per year).
Identification	Sponsorship source: National Institutes of Health (NIH), US Department of Agriculture. Country: USA. Setting: Hispanic communities, Houston, TX. Comments: Viva la Familia Study. Author's name: Nancy F Butte. Institution: US Department of Agriculture, Agricultural Research Service Children’s Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine, Houston, TX, USA. Email: nbutte@bcm.edu. Declaration of interests: yes. "None of the authors had a financial conflict of interest in relation to this study." Study ID: Butte 2007. Type of record: journal article.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Were adequate outcome data for cohorts available? All outcomes	High risk	Attrition at 1 year: 14.6% (151/1030).
Was there matching of less‐exposed and more‐exposed participants for prognostic factors associated with outcome or were relevant statistical adjustments done? All outcomes	Low risk	The model using dietary fat intake to predict weight gain did not adjust for parental BMI, physical activity, family income or parental education. However, there was no association between physical activity, family income and parental education and weight gain after adjustment for gender, age, pubertal stage and baseline BMI of the child.
Did the exposures between groups differ in components other than only total fat? All outcomes	Low risk
Can we be confident in the assessment of outcomes? All outcomes	Unclear risk	Insufficient description of outcome measurement methods.
Can we be confident in the assessment of exposure? All outcomes	High risk	Dietary intake only assessed once, at baseline.
Can we be confident in the assessment of presence or absence of prognostic factors? All outcomes	High risk	Single assessment of physical activity performed. Pubertal stage self‐reported. Unclear whether parental BMI was self‐reported or measured.
Was selection of less‐exposed and more‐exposed groups from the same population? All outcomes	Low risk	All children were participants of the Viva la Familia Study.