Ebbeling 2006.

Methods	Design: RCT Timing: Prospective Allocation to group: Random Number of clusters or sites: N/A Number of individuals: 103 Length of intervention: 25 weeks
Participants	General description of participants: Teenagers consuming at least 360 mL SSB per day living in Boston, Massachusetts, USA Age: Teenagers (an age of 13 to 18 years was an eligibility criterion, and mean age was 16 years) Inclusion criteria: Quote: "We enrolled 103 adolescents (47 males and 56 females), aged 13 to 18 years, who reported consuming at least 1 serving (ie, 360 mL or 12 fl oz) per day of SSB (ie, soft drinks, juice drinks containing <100% juice, punches, lemonades, iced teas, and sports drinks). Each subject lived predominantly in 1 household (ie, no more than 1 weekend every 2 weeks in a secondary household)." Exclusion criteria: Quote: "We excluded those who were currently dieting for the purpose of weight loss or taking prescription medications that might affect body weight. We also did not enroll those who reported smoking at least 1 cigarette in the past week or were diagnosed as having a major medical illness or eating disorder. To decrease the likelihood of enrolling individuals with eating disorders or undernutrition, we excluded those with a BMI below the 25th percentile" Recruitment: "Recruitment and screening of subjects were conducted in collaboration with a local high school that provided mailing lists and space for obtaining measurements. Packets containing an invitation letter and informed consent and assent documents were sent to parents of all students enrolled at the school" Weight status at baseline: Body weight at baseline was not an inclusion criterion, and mean BMI at baseline was 26 kg/m2 in the IG and 25 kg/m2 in the CG. Study authors did a sub‐group analysis according to weight status SSB consumption at baseline: Consumption of at least 360 mL per day of SSB was an inclusion criterion. Mean energy intake from SSB at baseline was 1466 kJ/day in the intervention and 1596 kJ/day in the control group Equity considerations: Participants lived in Boston, USA, included 65% non‐white participants, and 18% participants residing in subsidized housing. Quote: "The results of our pilot study were not materially affected by gender, race or ethnicity, age, household income, household size, physical activity, or television viewing"
Interventions	Intervention: Home‐delivery of non‐caloric beverages (Quote: "The intervention group received weekly home deliveries of noncaloric beverages for 25 weeks. The target number of individual beverage servings (ie, 360 mL or 12 fl oz per referent serving) delivered to each home was based on household size: 4 servings per day for the subject and 2 servings per day for each additional member of the household. This extra allotment was provided to avoid competition between the subject and family members for the beverages. We distributed an order form to each household for selecting beverage preferences from a wide variety of options (eg, bottled water and “diet” beverages including soft drinks, iced teas, lemonades, and punches). The beverage order form listed options in units, based on manufacturer packaging. The units contained bundles of 4 to 6 cans or bottles, with volumes ranging from 300 to 720 mL (10–24 fl oz) per can or bottle. The target number of delivered servings, specified above, was approximately equal to 5 units per week for the subject and 3 units per week for each additional member of the household. A regional supermarket delivery service filled the orders and delivered the beverages, with research staff coordinating and monitoring the process") Behavioural co‐intervention: Dietary counselling focused on beverages ("We instructed subjects to drink the noncaloric beverages delivered to their homes and not to buy or drink SSBs. In addition, we offered advice on how to choose noncaloric beverages when not at home. (...) Written instructions regarding beverage consumption were mailed to subjects at the beginning of the intervention period. We also contacted each household by telephone during the first week of the intervention to speak with the subject and a parent. This telephone contact provided an opportunity to reinforce instructions, answer questions, and address concerns. Thereafter, we contacted each subject by telephone on a monthly basis throughout the intervention period to assess satisfaction with beverage choices and deliveries, discuss beverage consumption, and provide motivational counseling") Control: No intervention (Quote: "We asked subjects in the control group to continue their usual beverage consumption habits throughout the 25‐week intervention period. They received weekly home deliveries of noncaloric beverages for 4 weeks after completion of follow‐up measurements, as a benefit for having participated in the study")
Outcomes	Measures of SSB intake: Energy intake from SSB (kJ/day) as change from baseline, assessed through multiple‐pass dietary recall at baseline and 25 weeks Measures of intake of alternatives to SSB: Non‐caloric beverage intake (ml/day) as change from baseline, assessed through multiple‐pass dietary recall at baseline and 25 weeks Anthropometric measures: Mean BMI as change from baseline, assessed through a standardized diagnostic protocol at baseline and 25 weeks Adverse outcomes: The study does not report how data on adverse outcomes were collected, but states that none were observed Other outcomes: None included in this review
Context and implementation	Setting: Participants' homes Sector: Health care Country: USA Year(s) when implemented: 2003‐2004 Mode of implementation: Pilot trial by researchers Level of implementation: Setting‐based intervention
Declarations	COI: "The authors have indicated they have no financial relationships relevant to this article to disclose" Funding: "This study was supported by grants R01 DK63554 and K01 DK62237 from the National Institute of Diabetes and Digestive Kidney Diseases, the Charles H. Hood Foundation, and grant M01 RR02172 awarded by the National Institutes of Health to support the General Clinical Research Center at Children's Hospital Boston" Trial registration: N/R Protocol availability: Protocol mentioned in the report but not publicly available
Notes	None
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	We contacted the study's corresponding author for additional information on the randomization procedure, which is not described in detail in the study's published report, and received the information that a computer random number generator was used to create the list of assignments. Quote: "To determine the experimental groups, we used a computer random number generator to prepare a list of assignments, mixed in random order. These were concealed until an enrollment list had been obtained from each participating school. We then juxtaposed the participant list with the assignment list to define the groups"
Allocation concealment (selection bias)	Low risk	Quote: "To avoid any bias in the enrollment procedure, personnel conducting recruitment were masked to sequence"
Similarity of baseline outcome measurements (selection bias)	Low risk	Quote: "There were no significant group differences between intervention and control subjects in (...) anthropometrics (weight, height, and BMI). Likewise, the groups did not differ in baseline levels of daily EISSB [energy intake from SSB], noncaloric beverage intake, physical activity, television viewing, or total media time"
Similarity of other baseline characteristics (selection and performance bias)	Low risk	Quote: "There were no significant group differences between intervention and control subjects in demographics (gender, race, ethnicity, age, household income, and household size)"
Incomplete outcome data (attrition bias) All outcomes	Low risk	There was not attrition
Blinding (performance and detection bias) Subjective outcomes	High risk	Dietary intake was self‐reported, and blinding of participants was not possible (interviewers were, however, masked to group assignment)
Blinding (performance and detection bias) Objective outcomes	Low risk	Participants were not blinded, but the outcome (BMI) was objective and assessed based on a standardized clinical protocol
Contamination (performance bias)	Low risk	It is unlikely that the CG received the intervention
Selective reporting (reporting bias)	Low risk	All outcomes mentioned in the methods section are reported; moreover, the study reports all outcomes one would generally expect in such a study. However, no information on study registration is provided
Other bias	Low risk	No other concerns