(Photo by WINDCOLORS on Shutterstock)
Adults Assigned Diet Drinks Lost About as Much Weight as Water Drinkers Over Two Years
In A Nutshell
- Among adults who finished a two-year trial, those assigned diet drinks lost 10.6 pounds on average and those assigned water lost 8.2 pounds, a gap too small to rule out chance.
- Both groups still weighed less than at the start after a final year without coaching, though each regained some weight after reaching a low point.
- Only 220 of the 493 adults who started completed the full two years, so the results come from the people who stayed.
- Small differences in cholesterol, blood sugar, and blood pressure favored water, but the authors judged none large enough to matter medically.
Adults in a two-year weight-loss trial who drank diet sodas and other artificially sweetened beverages ended up with about the same results as adults who drank only water. Among participants who finished, those who drank water had lost an average of 8.2 pounds after 104 weeks, while those who drank diet beverages had lost 10.6 pounds, a gap too small to rule out plain chance.
Researchers at the University of Liverpool in England followed 493 adults with overweight or obesity through a structured weight-loss program, then checked in for a full extra year with no coaching at all. Only 220 stayed to the end, and those who did still weighed less than they had at the start of the study.
Diet drinks have long drawn suspicion as a saboteur of weight-loss plans. The World Health Organization advises against using non-sugar sweeteners to control weight, and some observational studies, the kind that spot patterns but cannot prove cause and effect, have linked diet drinks to weight gain. The study’s authors argue that guidance leaned mostly on that type of research rather than on randomized trials like this one. People already gaining weight may simply reach for diet drinks more often, one reason such studies can mislead. Known as SWITCH, the trial provides rare long-term randomized data comparing diet drinks directly with water for weight management, and the results appear in the British Journal of Nutrition.
Nearly 500 Adults Were Assigned Diet Drinks or Only Water
Volunteers were adults ages 18 to 65 living with overweight or obesity, with an average age of about 45 and a typical starting weight near 200 pounds. Roughly 70% were women. All lived within 50 miles of Merseyside, the region that includes Liverpool, and already drank at least three cold beverages a week. Anyone with diabetes, cardiovascular disease, digestive disease, food allergies, or a history of weight-loss surgery was left out, as were current and recent smokers.
Random assignment sent 246 people to water and 247 to diet drinks, and everyone was asked to drink at least two 330-milliliter servings a day, roughly 11 ounces each. The researchers supplied the drinks. The diet drink group chose from 20 branded options, including diet colas, diet fruit drinks, and flavored waters, sweetened mostly with aspartame, acesulfame potassium, and sucralose. The water group gave up all artificially sweetened drinks, even sweeteners stirred into coffee or tea.
Everyone also attended group weight-loss sessions with a nutritionist, weekly for the first 12 weeks and monthly for the next 40. A final 52-week stretch was optional and came with no coaching sessions, though measurements continued monthly. Enrollment ran from 2016 to 2021, so the COVID-19 pandemic disrupted some in-person visits. Researchers tracked weight, waist and hip size, and hunger ratings throughout.
Diet Drinks and Water Led to Similar Weight Loss After Two Years
By the two-year mark, 117 people remained in the water group and 103 in the diet drink group. Among them, about 98% in both groups kept to their daily drink target near the end, according to beverage logs. High dropout means the two-year results come from the people who stayed.
Both groups followed the same arc. Weight came off fastest in the first 12 weeks. Diet drinkers hit their lowest average weight around week 36, about eight months in, and water drinkers around week 44, roughly ten months in. Both groups then slowly regained some weight, though regain was slower with diet drinks. Even so, both stayed clearly below their starting weights at week 104. Re-running the numbers with estimates filled in for dropouts led to the same conclusion about the two groups.
Waists and hips shrank by similar amounts, with waists down roughly 2 inches in both groups. A subset of 48 volunteers who had body scans lost fat and some lean tissue at similar rates in each group. Hunger ratings rose slightly in both groups over two years, though only the diet drink group’s rise was big enough to stand out from chance, and the two groups did not differ from each other. Water drinkers sharply cut their sweetener intake, as expected, and sugar intake dropped in both groups.
A Few Small Health Markers Favored Water, but Not by Amounts That Matter Medically
Total cholesterol crept up slightly in the diet drink group, and a long-range blood sugar measure called HbA1c edged up with diet drinks while holding roughly steady with water. Diastolic blood pressure, the bottom number in a reading, fell more among water drinkers. Cholesterol and blood sugar results came from only 14 to 16 people per group, and the authors judged none of the differences large enough to matter medically.
Diet Drinks Did Not Hinder Weight Loss in a Structured Program
Participants were otherwise healthy adults who received coaching during the first year, so the results speak to weight loss in a program like this one rather than to every health effect of artificial sweeteners.
For adults in a weight-management program who are mainly concerned about shedding pounds, this trial found no evidence that choosing these diet drinks over water sabotaged their results.
Disclaimer: This article is for general informational purposes only and does not constitute medical or dietary advice. Individual results vary, and anyone considering changes to a diet or weight-management plan should consult a qualified health professional.
Paper Notes
Limitations
This trial was conducted at a single site in England and did not collect data on participants’ race or ethnicity, which limits how well the findings apply to other populations. Only 220 of 493 participants completed measurements at week 104. The authors name the COVID-19 pandemic, the two-year length, and the voluntary final phase as potential reasons, noting that the voluntary phase was added by amendment after recruitment targets were not being met. The design was open-label, meaning participants knew which drink they were assigned. During pandemic restrictions, some weights and body measurements were self-reported at home, though the authors report that weight collection location did not affect results. Cholesterol and blood sugar comparisons rested on small subsets of participants.
Funding and Disclosures
Funding for the trial came from the American Beverage Association, which also paid for the beverages and the medical writing support for the manuscript. The authors state the association was not involved in the design, conduct, data analysis, or interpretation of the trial, though a representative was given the opportunity to review the manuscript. Final content decisions were made by the authors, and a third-party organization, Biofortis, audited data at the clinical site. Several authors reported that funding for this study was paid to the University of Liverpool by the American Beverage Association. Outside this study, authors reported grants from the European Commission, funding from Coca-Cola for a PhD studentship, honoraria from the International Sweeteners Association, and consultancy or other fees from companies including Novo Nordisk, AstraZeneca, Boehringer Ingelheim, Dupont, and Mars.
Publication Details
Titled “Effect of non-nutritive sweetened beverages v. water on body weight: long-term results of a randomised controlled trial,” the paper was written by Joanne A. Harrold, Scott Hill, Cristina Radu, Paul Thomas, Paula Thorp, Charlotte A. Hardman, Paul Christiansen, Una Masic, Lauren McGale, Dan J. Cuthbertson, and Jason C.G. Halford. Authors were affiliated with the University of Liverpool, the Ministry of Defence, Edge Hill University, and the University of Leeds in the United Kingdom. Joanne A. Harrold is the corresponding author. It was published in 2026 in the British Journal of Nutrition (DOI: 10.1017/S0007114526108150). The trial is registered with ClinicalTrials.gov under the identifier NCT02591134.
