Diet drinks and weight loss: what trials show when they are compared with water and with sugary drinks

A two-year randomized trial has found that adults trying to lose weight did about as well drinking diet soft drinks as drinking water. The result, from the University of Liverpool, was publicized in early October 2026 under the headline that diet drinks “work as well as water.” It lands in a field where the World Health Organization has advised people not to use sweeteners for weight control at all.

Both positions rest on evidence, and they are less contradictory than they look. The answer to whether diet drinks help depends heavily on what they are being compared with.

What diet drinks are, and what the argument is about

Diet drinks replace sugar with non-nutritive sweeteners, compounds such as aspartame, sucralose and acesulfame potassium that taste sweet and supply few or no calories. The case for them is arithmetic. In one trial described below, an 8-ounce serving of the sugary drink supplied 104 calories and its sugar-free twin almost none.

The case against is that the body may not cooperate. Critics have argued that sweetness without calories could sharpen appetite, sustain a taste for sweet food or disturb metabolism, so that the saved calories are eaten back or worse. Some long-term studies that track people’s habits have found diet-drink users gaining more weight, not less.

A team at the University of Toronto summarized where official advice stood in 2022, with major dietary guidelines recommending the use of water, and not diet drinks, in place of sugary drinks. The question they and the Liverpool group set out to answer was whether that preference for water is supported by trials.

What the two-year trial of diet drinks and water found

The Liverpool study, called SWITCH, recruited adults with a body mass index of 27 to 35, a range that runs from overweight into obesity, who regularly consumed cold beverages. In all, 493 were randomly assigned, 246 to water and 247 to diet drinks. Everyone joined the same group weight-management course, with weekly meetings for 12 weeks and monthly meetings for a further 40.

It was designed as a parallel-group, open-label, controlled equivalence trial. Parallel-group means each person stayed in one group throughout. An equivalence trial asks whether two options perform the same within a preset margin, instead of asking whether one is better. Open-label means participants knew which drink they had been given, which is unavoidable when one of them is water.

At the end of the first year, both groups had lost weight. Participants consuming water maintained a weight loss of 6.1 kg over 52 weeks versus 7.5 kg with diet drinks. That 1.4 kg gap was large enough that the two could not be declared equivalent, though the authors judged it too small to matter clinically.

The second year was the harder test. Support was withdrawn, and the trial was followed by a voluntary, unassisted 1-year extension. Some weight came back in both groups. At 104 weeks the water group was 3.7 kg below its starting weight and the diet-drink group 4.8 kg below. Statistically the two were not different, and this time the equivalence test was met.

The authors’ own summary is modest. Their results, they wrote, suggest people can lose weight and maintain it for 2 years when drinking either kind of beverage and taking part in a weight-management course.

Other measurements moved together. According to the university, researchers found no clinically meaningful differences in measures such as blood pressure and cholesterol. Sugar intake fell in both groups, with a slightly larger reduction among participants drinking the diet beverages.

Jason Halford, who led the work with Jo Harrold, drew a broad conclusion. “This study provides the long-term clinical evidence that has been missing from international discussions,” Halford said.

Two facts about the trial need to sit beside that claim. The study was funded by the American Beverage Association, a trade association for drinks makers. The university states that the study team operated independently from the funder and retained full control over the analysis and reporting of the results. And most participants did not finish. The final 52-week unassisted phase was completed by 220 participants, fewer than half of those who started.

What pooled trials show about diet drinks, water and sugary drinks

One trial is rarely decisive. The Toronto team’s 2022 analysis gathered every randomized trial lasting at least two weeks that compared diet drinks, sugary drinks and water. It found 17 trials with 24 trial comparisons, involving 1733 adults with overweight or obesity.

Their method combines direct and indirect comparisons, so that three options can be ranked against one another even where no single trial tested all of them. It looked at three substitutions. Some trials tested more than one, which is why the rows below add up to more than 17.

Substitution tested	Trials	Effect on body weight	Certainty
Diet drinks in place of sugary drinks	12 trials, 601 participants	1.06 kg lower (range 0.41 to 1.71)	Moderate
Water in place of sugary drinks	3 trials, 429 participants	No association found	Low
Diet drinks in place of water	9 trials, 974 participants	No association found	Low

The first row is the clearest result. Replacing sugary drinks with diet drinks was associated with reduced body weight, along with a lower body mass index and less body fat. One kilogram is a small effect, and the trials were mostly short.

The second row is a surprise at first sight, since water also has no calories. Only three trials tested it, and the reviewers rated that evidence as low certainty.

The third row matches the Liverpool trial. Compared with each other, diet drinks and water showed no difference in weight, on evidence rated low certainty. The authors concluded that diet drinks used as intended were associated with small improvements in body weight and cardiometabolic risk factors without evidence of harm, pointing in the same direction as replacing sugary drinks with water. Cardiometabolic risk factors are measures such as blood pressure, blood fats and blood sugar.

Funding was mixed in these trials as well. Eight were paid for by public or nonprofit agencies, four by industry and five by a combination.

What a trial in children found about sugar-free drinks

The most carefully controlled test was done in children. Dutch researchers gave 641 primarily normal-weight children, aged about 5 to 11, one small can a day for 18 months. Half received a sugar-sweetened drink and half an artificially sweetened one that looked and tasted alike, so neither children nor parents knew which they had.

Weight increased by 6.35 kg in the sugar-free group as compared with 7.37 kg in the sugar group. Growing children gain weight, so both numbers are positive, and the sugar-free group gained about a kilogram less. A quarter of the children had stopped drinking their assigned beverage by the end.

That trial answers a narrow question well. Taking roughly 100 calories of sugar a day out of a child’s diet, without the child knowing, slowed weight gain. It does not show that sweeteners themselves do anything useful.

Why the WHO advises against sweeteners for weight control

In May 2023 the World Health Organization had come down against sweeteners. It released a new guideline on non-sugar sweeteners (NSS), which recommends against the use of NSS to control body weight or reduce the risk of noncommunicable diseases. “Replacing free sugars with NSS does not help with weight control in the long term,” said Francesco Branca, the agency’s director for nutrition and food safety. Noncommunicable diseases are long-term conditions such as heart disease and diabetes, and free sugars are those added to food and drink or present in honey, syrups and juice.

The guideline drew on a systematic review published in The BMJ in 2019. Of 56 individual studies that provided data for this review, 35 were observational studies. Its trial evidence was thin. No evidence of any effect of NSSs was seen on overweight or obese adults or children actively trying to lose weight, and the reviewers warned that most studies had few participants, were of short duration, and their methodological and reporting quality was limited.

The WHO’s stronger worry came from the observational side. Its review suggested that there may be potential undesirable effects from long-term use of NSS, such as an increased risk of type 2 diabetes, cardiovascular diseases, and mortality in adults.

Here the agency was candid about weakness. People who choose diet drinks often do so because they are already overweight or have been told their blood sugar is high, and those same people are more likely to develop diabetes whatever they drink. The WHO acknowledged that the link might be confounded by baseline characteristics of study participants, and for that reason the recommendation has been assessed as conditional. A conditional recommendation is the weaker of the WHO’s two grades.

The two bodies of evidence therefore answer different questions. Trials lasting up to two years show no weight penalty from diet drinks and a small advantage over sugar. Observational studies lasting decades show worse health among users and cannot establish why.

The Liverpool researchers hope to shift the guidance. “These latest findings show that concerns about sweeteners disrupting appetite or causing weight gain are not supported when tested in a rigorous, long-term randomized trial,” Harrold said.

Limits of the evidence on diet drinks and weight loss

The new trial’s dropout is its main weakness. With 220 of 493 participants completing two years, the people who stayed may differ from those who left, and a comparison between the remaining groups is less secure than the original random assignment.

Its funding does not invalidate the result and does call for independent replication. The earlier pooled analysis, which includes trials funded by public agencies, shows the same pattern, which lends it support.

Every participant was in a weight-management course. The trial says nothing about people who drink diet soda without trying to lose weight, which describes most consumers.

Two years is long for a diet trial and short for disease. The trials summarized here did not run long enough to count heart attacks or new cases of diabetes, so those questions still rest on observational data.

The trials also treat sweeteners as a group. Individual compounds differ chemically, and the Liverpool drinks contained aspartame, acesulfame potassium or sucralose.

Whether a diet drink is a sensible swap for a particular person depends on what it replaces and on their health, including diabetes, which is assessed by their doctor.

In trials lasting up to two years diet drinks have matched water for weight and modestly beaten sugary drinks, on evidence that includes an industry-funded trial with heavy dropout and that cannot speak to disease risk over decades.