TABLE 5.

Randomized trials of SSBs and ASBs and body weight in youth¹

				Results
First author (ref)	Sample	Design	Duration	Intervention	Dietary intake	Body weight	Comments
Sichieri (51)	n = 1140	47 classrooms randomized to intervention or control	1 y	Education to promote replacing SSBs with water through 10 1-h sessions	SSBs decreased for intervention vs. control: −56 mL/d	No overall effect on BMI change or percentage of overweight or obese	Despite a 4-fold decrease in SSBs for intervention vs. control, both groups had increased BMI over time
	Age: 9–12 y					Decrease in BMI only for overweight intervention girls
	Sex: F, M
James (52)	n = 644	29 classrooms randomized to intervention or control	1 y; a 3-y follow-up on 424 students was also conducted (69)	Four 1-h education sessions over 1 y, or control	Carbonated beverages decreased for intervention (0.6 servings/3 d), no difference in SSBs	No effect on BMI change	Low overall intakes of SSBs as well as a number of important methodologic limitations (56)
	Age: 7–11 y					Percentage of overweight less for intervention vs. control
	Sex: F, M					No difference in overweight at 3-y follow-up
Ebbeling (53)	n = 103	Two-group parallel	6 mo	Water or ASBs delivered to homes, or control condition	82% reduction in SSB intake for intervention, no change in control (P < 0.0001)	No effect on BMI change; significant intervention effect on weight loss for those in the highest third of BMI	It is biologically plausible that the intervention would be most efficacious in those with the highest baseline BMI, perhaps due to gene × environment interactions
	Age: 13–18 y
	Sex: F, M
Ebbeling (54)	n = 224	Two-group parallel	2 y	Water or ASBs delivered to homes, or control condition	Reduction in SSB intake for intervention vs. control = 0.4 servings/d (P < 0.01)	No effect on BMI change between groups at 2 y	No effect on percentage body fat; the effect in Hispanic adolescents was also observed in a post hoc analysis of the earlier study by this group (53)
	Age: 15 ± 0.7					Significant effect on BMI in the interim 1-y analysis only for Hispanic children
	Sex: F, M
de Ruyter (55)	n = 641	Two-group parallel	18 mo	Water or ASBs delivered to schools, or control condition	Urinary sucralose evidence for adherence to the ASB intervention; 477 children who completed the study consumed 83% of the assigned beverages	No effect on BMI change in the total sample; effect only seen after excluding those who did not adhere	The children were normal weight and younger than in the other studies
	Age: 4–12 y
	Sex: F, M

¹ASB, artificially sweetened beverage; ref, reference; SSB, sugar-sweetened beverage.