Obarzanek 2001 (RCT).

Methods	Study design: RCT. Study grouping: parallel group. Allocation ratio in RCTs: 1:1. Analyses methods for RCTs: ITT; end values reported. Description of randomisation: "Computer‐generated randomisation assignments were provided by the coordinating centre to produce within each clinical center approximately equal numbers of participants assigned to the intervention and usual care groups balanced by age and sex;" central allocation; NR who enrolled participants. How were missing data handled? "It was assumed that missing data in both groups would have come from the same distribution as observed data in the usual care group, so missing year 3 LDL‐C data were estimated by drawing values from the usual care group distribution;" "Analyses of secondary outcomes using no imputation for missing values used ANCOVA models for continuous outcomes and Wilcoxon tests for ordered categorical outcomes. Baseline level and sex were included as covariates." Number of study contacts: 8. Period of follow‐up (from when duration of active intervention period ended): approximately 3 years. Period of recruitment: 2.5 years. Sample size justification adequately described? yes: "The sample size of 300 in each treatment group was based on estimates of intervention efficacy. The primary outcomes will be tested at a two‐sided significance level of u=0.05. To test the primary efficacy hypothesis with 90% power, the sample size needed per group is given by n = 2 (1.96 + 1.28)*var/A2, where A is the difference between the average changes in the treatment and control groups, and var is the variance of A. Variance estimates were derived from Bogalusa Heart Study data, using 8‐ to I0‐year‐old children with LDL‐C levels in the 75 to 98th percentile, and calculating baseline and 36‐month follow‐up variances as well as the correlation at these two times." Sampling method: mass mailing used to recruit children from schools, a health maintenance organization and paediatric practices; > 47,000 children were prescreened for potential eligibility; n = 5122 seen for screening 1; n = 1637 for screening 2; n = 752 for baseline visit. Study objective: to assess efficacy and safety of lowering dietary intake of total fat, saturated fat and cholesterol to decrease LDL cholesterol levels in children. Study population: prepubescent boys and girls with primary elevated serum LDL cholesterol levels.
Participants	Baseline characteristics (reported for 2 groups and overall group) Lower fat intake (≤ 30%TE) Age (mean in years): 9.5 (SD 0.74). Sex: 46.4% girls. Ethnicity: white 86.5%; black 7.5%; other 6%. Education: NR. Income: 15.1% had household income USD 20,000. Pubertal stage: Tanner stage I (prepubertal). Parental BMI: NR. Child total energy (kJ): 7364 (SD 1832). Child total fat: 33.4 (SD 5.5). Child saturated fat: 12.5 (SD 2.7). Child total protein: 14.8 (SD 2.8). Child total CHO: 53.0 (SD 6.7). Child physical activity: NR. Child physical inactivity or screen time or both: NR. Child CVD risk (excluding fatness): SBP (mmHg): 97.31 (SD 9.1); DBP (mmHg): 61.97 (SD 9.54); total cholesterol (mmol/L): 5.17 (SD 0.38); LDL‐C (mmol/L): 3.38 (SD 0.31); HDL‐C (mmol/L): 1.48 (SD 0.28); TGs (mmol/L): 0.9 (SD 0.33). Child body fatness: weight (kg): 32.7 (SD 6.8); BMI (kg/m2): 17.5 (SD 2.3); triceps skinfold (mm): 11.97 (SD 4.54); subscapular skinfold (mm): 8.02 (SD 4.41); supra‐iliac skinfold (mm): 9.45 (SD 5.8). Child height (cm): 136.2 (SD 6.8). Usual or modified fat intake Age (mean in years): 9.5 (SD 0.70). Sex: 44.4. Ethnicity: white 86.6%; black 9.4%; other 4%. Education: NR. Income: 5.9% had household income USD 20,000. Pubertal stage: Tanner stage I (prepubertal). Parental BMI: NR. Child total energy (kJ): 7229 (SD 1841). Child total fat: 34.0 (SD 4.9). Child saturated fat: 12.7 (SD 2.5). Child total protein: 14.6 (SD 2.7). Child total CHO: 52.8 (SD 6.2). Child physical activity: NR. Child physical inactivity or screen time or both: NR. Child CVD risk (excluding fatness): SBP (mmHg): 97.55 (SD 9.4); DBP (mmHg): 61.67 (SD 10.23); total cholesterol (mmol/L): 5.17 (SD 0.38); LDL‐C (mmol/L): 3.38 (SD 0.3); HDL‐C (mmol/L): 1.47 (SD 0.29); TGs (mmol/L): 0.92 (SD 0.32). Child body fatness: weight (kg): 33.1 (SD 6.9); BMI (kg/m2): 17.6 (SD 2.4); triceps skinfold (mm): 12.6 (SD 5.26); subscapular skinfold (mm): 8.59 (SD 4.73); supra‐iliac skinfold (mm): 10.1 (SD 6.04). Child height (cm): 136.5 (SD 7.0). Overall Age: NR. Sex: NR. Ethnicity: NR. Education: NR. Income: P = 0.002. Pubertal stage: All children were prepubertal at enrolment. Parental BMI: NR. Child total energy (kJ): P > 0.05. Child total fat: P > 0.05. Child total protein: P > 0.05. Child total CHO: P > 0.05. Child physical activity: NR. Child physical inactivity or screen time or both: NR. Child CVD risk (excluding fatness): BP P > 0.05; blood lipids P > 0.05. Child body fatness: weight: P > 0.05; BMI: P > 0.05; skinfolds: P > 0.05. Included criteria: boys aged 8 years 7 months to 10 years 10 months and girls aged 7 years 10 months to 10 years 1 month, with primary elevated serum LDL‐C levels (defined as mean of 2 fasting values between 80th and 98th age‐ and sex‐specific percentiles), with no evidence of pubertal development (Tanner stage I) and normal psychosocial and cognitive development. Excluded criteria: major illness; medications that might affect blood lipids or growth (or both); weight‐for‐height < 5th or > 90th percentile, or height < 5th percentile for sex‐ and race‐specific growth curves; any household member on a LF or "cholesterol‐lowering" diet; and parental factors such as prior heart disease, extreme obesity or excessive intake of alcohol, which are potential barriers to dietary adherence by the child. Children with serum levels of TGs > 200 mg/dL or of HDL‐C < 30 mg/dL. Pretreatment: NR. Brief description of participants: prepubertal boys (approximately n = 362) and girls (approximately n = 301) aged 7‐11 years with LDL‐C levels ≥ 80th and < 98th percentiles for age and sex percentiles of the Lipid Research Clinics population. Total number completed in RCT: last visit for BMI (> 5 years): intervention group n = 293; control group n = 283. Total number randomised: total n = 663; intervention group n = 334; control group n = 329.
Interventions	Intervention characteristics Lower fat intake (≤ 30%TE) Energy prescription: NR. Total fat prescription: 28%TE. SFA, PUFA, MUFA prescription: SFA 8%TE; PUFA 9%TE; MUFA 11%TE. Total protein prescription: 14%TE. Total CHO prescription: 58%TE. Other diet prescription details: cholesterol 75 mg/1000 kcal, not to exceed 150 mg/day; encourage water‐soluble fibre; each family given child and adult DISC "guidebooks" that outlined each session including activities and recipes. Participants provided with DISC recipe book and DISC "dictionary," which described grams of SFAs and a "GO or WHOA" score to help identify more appropriate and less appropriate foods. Method number of dietary assessments: 3 non‐consecutive 24‐hour dietary recalls at baseline (using standardised protocol, which included 2‐dimensional food models for portion) size estimates. Dieticians interviewed child and if necessary obtained additional information from parent. 1st recall collected at baseline, and 2 more collected by telephone with child within 2 weeks; thereafter 3 non‐consecutive 24‐hour recalls at 1, 3 and 5 years, and the last visit. Data from 3 recalls were averaged. For intervention group only, at least 3 × 24‐hour recalls collected quarterly to monitor dietary adherence for 3 years. Other components prescribed: group and individual sessions with multidisciplinary team to support behaviour change. Duration of intervention: 4 years. Implementation: in 1st 6 months, 6 weekly and then 5 biweekly group sessions led by nutritionists and behaviourists, and 2 individual visits held with nutritionist. Over 2nd 6 months, 4 group and 2 individual sessions held. During 2nd and 3rd years, group and individual maintenance sessions held 4‐6 times/year, with monthly telephone contacts between group sessions. During 4th year of follow‐up, 2 group events plus 2 individual visits conducted with additional telephone contacts as appropriate. Usual or modified fat intake Energy prescription: NR. Total fat prescription: prescription NR, heart healthy guidelines available to public provided. SFA, PUFA, MUFA prescription: prescription NR, heart healthy guidelines available to public provided. Total protein prescription: prescription NR, heart healthy guidelines available to public provided. Total CHO prescription: prescription NR, heart healthy guidelines available to public provided. Other diet prescription details: families provided with AHA publications "Dietary Guidelines for Americans" and "How to Make Your Heart Last a Lifetime." Method number of dietary assessments: 3 × 24‐hour recalls every year for 7 years. Other components prescribed: NR. Duration of intervention: once at baseline. Implementation: at trial entry, parents or guardians informed that their children's blood cholesterol level was high. No specific recommendations to see their physician given. Subsequent contacts limited to data collection visits. 3‐year lipid results provided for them to share with their physician. In addition, cases exceeding cut‐off points for clinical monitoring, which included LDL‐C, height and ferritin, reviewed to assess whether physician referral warranted based on NCEP guidelines for drug treatment and clinical judgement.
Outcomes	Weight Weight (kg) (MD at 1 and 3 years adjusted for baseline value and sex). BMI BMI (kg/m2) (MD at 1, 5 and 7 years adjusted for baseline value, sex and age at last visit (for last visit only)). Total cholesterol Total cholesterol (mmol/L) (MD at 1, 5 and 7 years adjusted for baseline value, sex and age at last visit (for last visit only)). LDL‐C LDL‐C (mmol/L) (MD at 1, 5 and 7 years adjusted for baseline value, sex and age at last visit (for last visit only)). HDL‐C HDL‐C (mmol/L) (MD at 1, 5 and 7 years adjusted for baseline value, sex and age at last visit (for last visit only)). TGs TGs (mmol/L) (MD at 1, 5 and 7 years adjusted for baseline value, sex and age at last visit (for last visit only)). SBP SBP (mmHg) (adjusted for baseline BP and sex). DBP DBP (mmHg) (adjusted for baseline BP and sex). Height Height (cm). Energy intake Energy intake (kJ) (MD at 1 and 3 years adjusted for baseline value and sex. Energy intake was 98 kcal/day (411 kJ/day) lower in the intervention than usual care group at 1 (P = 0.01) year and 148 kcal/day (619 kJ/day) lower at 3 years (P = 0.001), and not different at subsequent time points). Fat intake Fat intake (%TE) (MD at 1, 5 and 7 years adjusted for baseline value, sex and age at last visit (for last visit only)). Saturated Fat intake Saturated fat intake (%TE) (MD at 1, 5 and 7 years adjusted for baseline value, sex and age at last visit (last visit only)). Protein intake Protein intake (%TE) (MD at 1 and 3 years adjusted for baseline value and sex). CHO intake CHO intake (%TE) (MD at 1 and 3 years adjusted for baseline value and sex).
Identification	Sponsorship source: NHLBI. Country: USA. Setting: 6 clinical centres. Comments: NA. Author's name: Eva Obarzanek. Institution: DISC Coordinating Center, Maryland Medical Research Institute, Baltimore, MD, USA. Email: obarzane@nhlbi.nih.gov. Declaration of interests: no. Study ID: DISC 2001. Type of record: journal article.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	In DISC 1995, "computer‐generated randomisation assignments were provided by the coordinating center to produce within each clinical center approximately equal number of participants assigned to the intervention and usual care groups balanced by age and sex." Baseline characteristics similar between groups.
Allocation concealment (selection bias)	Low risk	In DISC 1993 authors stated, "eligible children were allocated randomly to intervention and usual‐care groups by the coordinating centre..." thus it appeared that there was a central allocation centre and recruitment at the clinical centres could not have been manipulated.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	In DISC 1993, "though it was not possible to have a double blind trial due to the nature of dietary intervention, a single blind was maintained by using data collectors unaware of group assignment." Participants not blinded. However, lack of double blinding was not likely to influence the outcomes.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Outcome assessors blinded to group assignment.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Numbers lost to follow‐up: at 3 years: intervention group 14/334 (4.2%) and control group 26/329 (7.9%) (no reasons). At 7 years: intervention group 39/334 (11.7%) and control group 44/329 (13.4%) (no reasons). No differences in age, height, weight, BMI, total and saturated fat intake, serum LDL‐C or serum ferritin, and in distributions of sex, household income and education in those attending final visit vs dropouts. Missing the last visit was not related to treatment assignment. Primary outcomes analysed using ITT, imputation process described; secondary outcomes analysed using per protocol analyses.
Selective reporting (reporting bias)	Low risk	Protocol not available, but paper with study design and baseline characteristics available and all the study's prespecified outcomes were reported in the results section.
Other bias	Unclear risk	Intervention diet focused only on fat intake changes and encouraged water‐soluble fibre, and control diet AHA publications "Dietary Guidelines for Americans" and "How to Make Your Heart Last a Lifetime" but no detailed nutrition composition detail provided.