Magarey 2001.

Methods	Study design: prospective cohort study. Analyses methods for cohort: generalised linear estimating equations evaluated longitudinal relationship between body fatness and macronutrient intake. Regression analysis assessed whether body fatness at a particular age was predicted by intake at any of the previous ages. How were missing data handled? Considerable attrition occurred from 500 selected at birth to 198 at 2 years and 130 at 11 years. Information on participants lost before 8 years not available, but sociodemographic status of children remaining in cohort at 8 years was upwardly skewed compared to original cohort due to cohort attrition. Therefore, new recruitment (n = 113) done at age 11 years with age‐matched and socioeconomic balanced to the cohort (Magarey and Boulton 1994). Number of study contacts: 7 (at 2, 4, 6, 8, 11, 13 and 15 years of age). Period of follow‐up (total period of observation): 13 years. Periods of recruitment: November 1975 to June 1976. Sample size justification adequately described? No. Sampling method: 500 infants randomly selected by birth order from healthy term infants born at Queen Victoria Hospital, Adelaide, South Australia between November 1975 and June 1976. Core sample of approximately 150 children was retained in a longitudinal study of growth and nutrition from birth to 15 years of age. A further 113 children recruited for the 11‐year assessment from an age‐matched cross‐sectional sample of 715 children who had taken part in a family heart disease risk factor precursor study when they were 8 years of age. Study objective: to investigate the longitudinal relationship between macronutrient intake and adiposity at ages 2‐15 years. Study population: healthy born children aged 2‐15 years in Adelaide, South Australia.
Participants	Baseline characteristics (reported for 1 overall group) Age (range eligible for inclusion in years): 2‐15. Sex: 42.3% girls. Ethnicity: NR. Education: NR. Income: NR. Pubertal stage: 12‐16%, prepubertal girls (aged 2‐8 years); 17‐22%, adolescent girls (aged 11‐15 years). Parental BMI: NR. Child total energy (kJ): overall 4860.1 (SD 949.15); boys 5030 (SD 880); girls 4630 (SD 990), <P0.05. Child total fat: grams/day: overall 50.4 g/day (SD 12.9); boys 52.3 g/day (SD 12.2); girls 47.9 g/day (SD 13.4); <P0.05; overall 38.3%TE (SD 9.8); boys 38.4%TE (SD 5.8); girls 38.1%TE (SD 13.4). Child total protein: overall 39.8 g/day (SD 9.9); boys 41 g/day (SD 9.2); girls 38.3 g/day (SD 10.6); P > 0.05; overall 14%TE (SD 2.4); boys 13.9%TE (SD 2.3); girls 14.1%TE (SD 2.4). Child total CHO: overall 144.9 g/day (SD 34.5); boys 150 g/day (SD 34); girls 138 g/day (SD 34); P0.05; overall 47.8%TE (SD 7.4); boys 47.7%TE (SD 7.4); girls 47.9%TE (SD 7.4). Child physical activity: NR. Child physical inactivity or screen time or both: NR. Child CVD risk (excluding fatness) (n = 129): total cholesterol (mmol/L): overall 4.19 (SD 0.77); boys 4.17 (SD 0.82); girls 4.22 (SD 0.71); LDL‐C (mmol/L): overall 2.13 (SD 0.73); boys 2.06 (SD 0.75); girls 2.21 (SD 0.69); HDL‐C (mmol/L): overall 1.32 (SD 0.5); boys 1.39 (SD 0.61); girls 1.23 (SD 0.28); TG (mmol/L) overall 1.73 (SD 0.9); boys 1.81 (SD 0.93); girls 1.62 (SD 0.86). Child body fatness, weight (kg): overall 12.75 (SD 1.63); boys 13.0 (SD 1.8); girls 12.4 (SD 1.3); P ≤ 0.05. Child body fatness: BMI (kg/m2): overall 16.67 (SD 1.59); boys 16.8 (SD 1.7); girls 16.5 (SD 1.4); P > 0.05; BMI‐SDS: overall 0.07 (SD 1.26); boys 0.22 (SD 1.32); girls ‐0.14 (SD 1.14); P > 0.05. Child body fatness: triceps skinfold (mm): overall 10.1 (SD 2.3); boys 10.0 (SD 2.1); girls 10.2 (SD 2.5); P > 0.05; TC‐SDS: overall ‐0.42 (SD 0.85); boys ‐0.35 (SD 0.81); girls ‐0.51 (SD 0.90); P > 0.05. Child body fatness: subscapular skinfold (mm): overall 7.5 (SD 1.8); boys 7.2 (SD 1.6); girls 7.9 (SD 1.9); P ≤ 0.05; SS‐SDS: overall 0.22 (SD 0.85); boys 0.19 (SD 0.76); girls 0.26 (SD 0.95); P > 0.05. Included criteria: children who participated in the Adelaide Nutrition Study aged 2‐15 years with available follow‐up data. Excluded criteria: NR. Brief description of participants: children who participated in the Adelaide Nutrition Study aged 2‐15 years with 12‐16% of the boys being overweight, 12‐16% of prepubertal girls (aged 2‐8 years) and 17‐22% of adolescent girls (aged 11‐15 years). Total number completed in cohort study: 218 (at 15 years). Total number enrolled in cohort study: 500 (at birth) + 113 (at 11 years).
Interventions	Description of exposure for cohort Time span: 13 years. Dietary assessments used and frequency: single 3‐day weighed food record at ages 2, 4 and 6 years, and 1 single 4‐day weighed food record at 9, 11 and 13 years. From 11 years, children encouraged to take increasing responsibility for completing the food record. 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 (kg). BMI BMI‐for‐age z‐score. Skinfold thickness Sum of 4 skinfolds (triceps, biceps, subscapular and supra‐iliac) (mm). Triceps z‐score. Subscapular z‐score. Height Height (cm).
Identification	Sponsorship source: National Heart Foundation of Australia, Adelaide Children's Hospital Research Foundation and the National Health and Medical Research Council of Australia. Country: Australia. Setting: community in Adelaide. Comments: Adelaide Nutrition Study (birth cohort). Author's name: AM Magarey. Institution: Department of Public Health, The Flinders University of South Australia. Email: NR. Declaration of interests: no. Study ID: Magarey 2001. 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 attrition (71.4% over 8 years). No information available on children lost to study between 2 and 8 years. Attrition at 11 years: 74%. Since the children who returned had an upwardly skewed sociodemographic profile, another 115 children were recruited from an age‐matched cross‐sectional sample.
Was there matching of less‐exposed and more‐exposed participants for prognostic factors associated with outcome or were relevant statistical adjustments done? All outcomes	High risk	No matching reported. Ethnicity, SES, physical activity and pubertal stage not adjusted for in regression analyses.
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	Anthropometric measurements done using standard methods by 1 observer.
Can we be confident in the assessment of exposure? All outcomes	Low risk	Repeated weighed 3‐day DRs completed by parents and children throughout study.
Can we be confident in the assessment of presence or absence of prognostic factors? All outcomes	Unclear risk	Parental anthropometric data were investigator‐measured once when children were 8‐9 years old. Method not described.
Was selection of less‐exposed and more‐exposed groups from the same population? All outcomes	Low risk	It is likely the 2 groups were from the same population although the original sample were selected from a single hospital (Victoria, Adelaide, Australia) and the additional sample from the same birth cohorts were purposively selected to balance demographic characteristics of the cohorts.