Appannah 2015.

Methods	Study design: prospective cohort study. Analysis methods for cohorts: prospective associations between DP z‐scores and cardiometabolic risk factors at 14 and 17 years of age analysed using GEE with an exchangeable correlation structure. Beta coefficients resulting from the regression models for these biomarkers were back‐transformed for interpretation. Logarithmic transformation was applied to insulin, HOMA and TG measurements as they were not normally distributed. How were missing data handled? Out of 2337 adolescents eligible at 14 years, 1857 (79.5%) responded to FFQs and 1286 (55%) attended physical assessments. Number of study contacts: 2 (at 14 and 17 years). Period of follow‐up (total period of observation): 3 years. Periods of recruitment: 1989‐1991 (mothers of participants were recruited). Sample size justification adequately described? No. Sampling method: convenience sample. Present analysis uses data collected at 14 (n = 1857) and 17 (n = 1709) years' follow‐up from Raine cohort study. Original cohort comprised 2900 pregnant women recruited into a trial at King Edward Memorial Hospital (Perth, Western Australia) from 1989 to 1991. At 14 years, 2337 adolescents were eligible for follow‐up. Study objective: to examine associations between an "energy‐dense, high‐fat and low fibre" DP and cardiometabolic risk factors, and the tracking of this DP in adolescence. Study population: Australian adolescents aged 14‐17 years.
Participants	Baseline characteristics (reported for 1 overall group) Age (age eligible for inclusion in years): 14. Sex: 49% girls. Ethnicity: majority white. Education: maternal education (n = 767), overall; > 10 years (66%); = 10 years (34%). Income: family income (n = 776), %, USD 30,000 (16%); > USD 30,000‐USD 50,000 (21%); > USD 50,000‐USD 70,000 (20%); > USD 70,000‐USD 104,000 (23%); > USD 104,000 (20%). Pubertal stage: NR. Parental BMI: NR. Child total energy (kJ): overall (n = 785) 9667 (SD 2950); girls (n = 382) 8882 (SD 2815); boys (n = 403) 10,412 (SD 2882). Child total fat (g): overall (n = 785) 91 (SD 32); girls (n = 382), 84 (SD 31); boys (n = 403) 98 (SD 32). Child total protein (g): overall (n = 785) 96 (SD 29); girls (n = 382) 88 (SD 27); boys (n = 403) 103 (SD 29). Child total CHO (g): overall (n = 785) 279 (SD 89); girls (n = 382) 256 (SD 84); boys (n = 403) 301 (SD 89). Child physical activity: physical fitness (PWC‐170): overall (n = 1334) 111.1 (SD 30.1); girls (n = 640) 96.8 (SD 19.4); boys (n = 694) 124.3 (SD 32.2). Child physical inactivity or screen time or both: NR. Child CVD risk (excluding fatness): HDL‐C (mmol/L): overall (n = 1376), 1.39 (SD 0.32); girls (n = 664) 1.43 (SD 0.32); boys (n = 712) 1.35 (SD 0.31); LDL‐C (mmol/L): overall 2.32 (SD 0.63); girls 2.38 (SD 0.61); boys 2.26 (SD 0.64); TGs (mmol/L), median (IQR): girls 0.95 (1.45); boys 0.88 (1.55); glucose (mmol/L): overall 4.63 (SD 0.68); girls 4.59 (SD 0.6); boys 4.66 (SD 0.74); HOMA‐IR, median (IQR): girls 2.32 (1.67); boys 2.14 (1.86); smoker, (%) overall (n = 1582) 1.5%; girls (n = 772) 2.1%; boys (n = 810) 1%. Child body fatness: BMI (kg/m2): median (IQR): overall (n = 1605) 21.3 (4.1); girls (n = 780), 21.5 (4.1); boys (n = 825), 21.1 (4.1); BMI‐for‐age z‐score overall: ‐0 (SD 1.0); girls 0.06 (SD 1.0); boys ‐0.06 (SD 1.0); WC (cm): median (IQR): overall (n = 1580) 75.5 (10.9); girls (n = 766) 74.6 (10.1); boys (n = 814) 76.3 (11.5); WC z‐score: overall ‐0 (SD 1.01); girls ‐0.08 (SD 0.9); boys 0.07 (SD 1.1). Included criteria: adolescents who participated in the Raine cohort study and had complete dietary and cardiometabolic data at 14 and 17 years. Excluded criteria: NR. Brief description of the participants: adolescents aged 14‐17 years participating in Raine cohort study. Total number completed in cohort study: 1709 (1009 completed FFQ). Total number enrolled in cohort study: 2337.
Interventions	Description of exposure for cohorts Time span: 3 years. Dietary assessment method used: semi‐quantitative FFQ. Frequency: single FFQ at 14 (baseline) and 17 years. Components of dietary exposure: an "energy‐dense, high fat, low fibre" dietary pattern (DP)" defined as high intakes of processed meat, chocolate and confectionary, low‐fibre bread, crisps and savoury snacks, fried and roasted potatoes; high intake of these foods increased the DP z‐score calculated for each participant using reduced rank regression. 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	BMI BMI‐for‐age z‐score. Overweight/obese, odds. WC WC z‐score. WC ≥ 80 cm, odds. LDL‐C LDL‐C (mmol/L). HDL‐C HDL‐C (mmol/L). TGs TGs (%).
Identification	Sponsorship source: Medical Research Council (grant number U105960389) and research grants from the National Health and Medical Research Council of Australia (ID#1022134 (2012‐2014)) and the National Heart Foundation of Australia and Beyond Blue Cardiovascular Disease (grant number G 08P 4036) and Depression Strategic Research Program. Country: Australia. Setting: community in Perth. Comments: Western Australian Pregnancy (Raine) Cohort Study. Author's name: G Appannah. Institution: Department of Nutrition and Dietetics, Faculty of Medicine and Health Sciences, Universiti Putra Malaysia, Serdang, Malaysia; Medical Research Council Human Nutrition Research, Cambridge, UK. Email: Gina.Ambrosini@uwa.edu.au Declaration of interests: yes. "Authors have no conflicts of interest to declare." Study ID: Appannah 2015. 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	High lost to follow‐up rate (35‐40% at 14 and 17 years). Authors did not report any comparative analyses between participants lost to follow‐up and participants who completed study.
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	Study included mainly white participants, upper income families, stratified for gender. Adjusted for age, dietary misreporting, physical fitness, smoking and BMI‐for‐age z‐score. Not adjusted for parental BMI.
Did the exposures between groups differ in components other than only total fat? All outcomes	Unclear risk	NR.
Can we be confident in the assessment of outcomes? All outcomes	Low risk	Standard methods performed for measurement of weight, height, WC and fasting blood samples.
Can we be confident in the assessment of exposure? All outcomes	Low risk	Repeated assessment using a validated semi‐quantitative FFQ.
Can we be confident in the assessment of presence or absence of prognostic factors? All outcomes	Low risk	Physical fitness assessed at each session, using validated test (PWC‐170) which was correlated with self‐reported physical activity.
Was selection of less‐exposed and more‐exposed groups from the same population? All outcomes	Low risk	Mothers of participants selected for 1 cohort.