Diet drinks and water produced similar long-term weight-loss results in adults taking part in a two-year behavioral weight-management program, according to a randomized trial published in the British Journal of Nutrition. Among 493 adults with overweight or obesity, those assigned to non-nutritive sweetened beverages maintained an average 4.8 kg weight loss at 104 weeks, compared with 3.7 kg in the water group, a between-group difference that was not statistically significant.
SWITCH followed participants through weight loss and a year without routine support
The trial enrolled adults ages 18 to 65 living near Liverpool, England, with a body mass index (BMI) between 27 and 35 kg/m². All participants regularly consumed cold beverages before entering the study.
Participants were randomized to one of two beverage strategies. One group was assigned at least two 330 ml servings per day of non-nutritive sweetened beverages, including diet colas, diet carbonated drinks, diet fruit drinks and flavored waters. The other group was assigned the same minimum amount of water and was asked to avoid non-nutritive sweetened beverages.
Both groups entered the same behavioral weight-management program. They attended weekly sessions during a 12-week weight-loss phase, followed by monthly sessions during a 40-week assisted maintenance phase. After the first year, routine nutrition-support meetings stopped, and participants entered a voluntary 52-week unassisted follow-up period.
That final year is what makes the new analysis different from the trial’s earlier reports. The investigators were asking whether any difference between water and diet drinks would persist once structured support had largely ended.
After two years, the two beverage strategies were equivalent for weight change
Of the 493 participants who started the assigned intervention, 220 had body-weight measurements at both baseline and week 104: 117 in the water group and 103 in the non-nutritive sweetened beverage group.
Among those completers, average weight loss from baseline was 3.7 kg in the water group and 4.8 kg in the diet-drink group. The 1.1 kg difference between groups was not statistically significant, and the trial’s prespecified analysis concluded that the two strategies were equivalent for long-term weight change.
Sensitivity analyses that accounted for missing data reached the same general conclusion. Assignment to diet drinks rather than water was not a significant predictor of body weight at week 104 after baseline weight and other factors were considered.
The result is notably different from the trial’s one-year analysis. At 52 weeks, participants who completed follow-up had maintained a mean weight loss of 7.5 kg with non-nutritive sweetened beverages versus 6.1 kg with water. That difference was statistically significant, although the authors said it did not reach clinical significance. By week 104, some weight had been regained in both groups and the between-group difference had narrowed.
The study does not show that diet drinks cause weight loss on their own
The trial was not a test of simply adding diet soda to an otherwise unchanged diet. Everyone participated in an intensive behavioral weight-management program during the first year, including dietary guidance and increasing physical activity.
That matters when interpreting the result. The study supports the idea that people trying to lose weight did not do worse when they used non-nutritive sweetened beverages instead of water. It does not show that diet drinks independently produce a 4.8 kg weight loss.
The same distinction comes up with other approaches to weight management. Medications such as GLP-1 drugs have direct pharmacologic effects on appetite and metabolism, while beverage substitution is one component of overall dietary intake. The size and mechanism of the effect are not comparable.
Participants drank at least two assigned beverages per day
The intervention was more structured than simply telling people to choose whatever zero-calorie drink they preferred. Investigators provided the study beverages, and participants were asked to consume at least two 330 ml servings each day.
The non-nutritive sweetened beverage options contained no more than 5 kcal per 8 oz and included 20 branded choices. Participants assigned to water were asked to avoid all non-nutritive sweetened beverages, including sweeteners added to hot drinks. Both groups were counseled to reduce sugar intake, and participants in the diet-drink group could also drink water.
Reported adherence remained high among participants still in the study at week 104, with more than 98% in each group meeting the two-servings-per-day target according to beverage logs.
Cardiometabolic markers did not consistently favor either group
Both groups showed improvements from baseline in several measures over the two-year study, but the differences between beverage assignments were inconsistent.
At week 104, changes in waist and hip circumference did not differ significantly between groups. The investigators reported between-group differences in total cholesterol, glycated hemoglobin (HbA1c) and diastolic blood pressure, but not in most other measured biomarkers. Total cholesterol moved less favorably in the non-nutritive sweetened beverage group, while HbA1c and diastolic blood pressure changed more favorably in the water group.
The authors cautioned that biomarker sample sizes were smaller than the weight-analysis sample for some measures. These secondary outcomes therefore do not provide a strong basis for claiming that either beverage strategy was metabolically superior.
For readers interested in broader diet guidance, the key distinction is between replacing sugar-sweetened beverages and treating diet drinks as a health food. The 2025–2030 Dietary Guidelines emphasize overall dietary pattern rather than a single beverage choice.
The dropout rate limits how confidently the two-year result applies to everyone
Less than half of the 493 participants who started the intervention had a body-weight measurement at week 104. That degree of attrition is one of the study’s main limitations.
The researchers used multiple approaches to test whether missing data changed the conclusion, including multiple imputation and last-observation-carried-forward analyses, and those analyses also found no significant difference between beverage groups. Even so, people who stay in a two-year weight-management trial may differ from those who leave.
The trial was also open-label, so participants knew whether they were assigned water or diet drinks. Roughly three-quarters of participants were already regular users of non-nutritive sweetened beverages before the study, which may limit how directly the findings apply to people who rarely or never consume them.
The study was funded by the beverage industry
The American Beverage Association funded the SWITCH trial and also funded medical-writing assistance for the manuscript. The paper states that the association was not involved in the study’s design, conduct, data analysis or interpretation, although an association representative was given an opportunity to review the manuscript before publication. Final decisions about the content remained with the authors.
Several authors reported study funding from the American Beverage Association, and some disclosed additional relationships with food, beverage or pharmaceutical organizations.
Industry funding does not by itself invalidate a randomized trial, but it is relevant context when interpreting findings about commercially important products. Replication by independently funded groups would strengthen confidence in the long-term result.
The result is reassurance about substitution, not evidence that diet drinks are healthier than water
The most defensible conclusion from SWITCH is narrow: within a structured weight-management program, using non-nutritive sweetened beverages did not produce worse long-term weight outcomes than using water.
That is useful because diet drinks are often discussed as though replacing them with water must improve weight loss. This randomized trial did not support that assumption over two years.
It also does not settle broader questions about long-term non-nutritive sweetener exposure, cardiovascular outcomes, diabetes risk, the gut microbiome or mortality. A two-year weight-management trial was not designed to answer those questions.
For someone replacing a sugar-sweetened soda, either water or a non-nutritive sweetened alternative can substantially reduce beverage calories. Which one is preferable may depend on taste, adherence and the rest of the diet rather than an expectation that one choice will produce meaningfully greater weight loss.
References
- Harrold JA, Hill S, Radu C, et al. Effect of non-nutritive sweetened beverages v. water on body weight: long-term results of a randomised controlled trial. Br J Nutr. 2026. DOI: 10.1017/S0007114526108150.
- Harrold JA, Hill S, Radu C, et al. Non-nutritive sweetened beverages versus water after a 52-week weight management programme: a randomised controlled trial. Int J Obes. 2024;48:83-93. DOI: 10.1038/s41366-023-01393-3.
- ClinicalTrials.gov. Effects of Non-Nutritive Sweetened Beverages Versus Water on Weight Management. NCT02591134. Clinical trial record.