In this guide
What participants actually drank (and what the water group gave up)
Who paid for it, and how to read industry-funded nutrition research
Picture it: you're three weeks into another serious attempt at losing weight. In the office fridge there's a cold can of zero-sugar cola, and next to it the water bottle you filled at 8 a.m. One feels virtuous. The other feels like a loophole. Which one actually helps?
For years nobody could answer that with confidence. Observational studies kept finding that people who drink diet soda tend to weigh more, which scared plenty of people off. Short clinical trials said diet drinks were fine, but most of them ended before the hard part of weight management even begins, which is keeping the weight off. The argument went in circles, and official advice drifted toward the cautious side.
Then, at the end of September, the University of Liverpool announced the results of SWITCH, the longest randomised comparison of diet drinks against water that has been run. The paper itself appeared in the British Journal of Nutrition in August, but the university's announcement and a wave of coverage in the first days of October are what put it in front of everyone this week. We read the paper summaries, the trial's earlier 52-week report, the funding statements, and the comments from dietitians who weren't involved. Here is what holds up, what doesn't, and what to do with it.
The short version: after 104 weeks, adults assigned to drink diet drinks lost an average of 4.8 kg and adults assigned to water lost 3.7 kg. The 1.1 kg gap was not statistically significant, so the researchers called the two drinks equivalent for long-term weight change. But only 220 of 493 participants reached the two-year assessment, the study was funded by the American Beverage Association, and it compared diet drinks with water, not with sugary drinks. All three details change how much weight the headline can carry.
Why This Trial Matters So Much Right Now
In 2023 the World Health Organization issued a conditional recommendation against using non-sugar sweeteners to control body weight. The reasoning was mostly about uncertainty: short trials showed a small benefit, while longer observational studies linked sweetener use with higher risks of weight gain and chronic disease. "Conditional" is the important word. It signals that WHO's panel knew the evidence was thin and wanted better long-term trials.
SWITCH is one of those trials. It was registered in 2015 (ClinicalTrials.gov NCT02591134), began enrolling in 2016 and finished data collection in early 2024, so it has been running through nearly the whole era of the diet-drink debate. That long timeline is a genuine strength. It is also why people on every side of the argument are now scrambling to claim it.
So let's slow down and read it the way a sceptical dietitian would: design first, then numbers, then who paid, then what it does and doesn't show.
The SWITCH Trial at a Glance
| Item | What we know |
|---|---|
| Full name | SWITCH: effects of non-nutritive sweetened beverages on appetite during active weight loss |
| Led by | Prof. Joanne Harrold (University of Liverpool) with Prof. Jason Halford (University of Leeds) |
| Published | British Journal of Nutrition, volume 136, issue 4, pages 482–493 (August 2026) |
| Participants | 493 adults aged 18–65 with BMI 27–35, living within about 50 miles of Merseyside; roughly 70% women; all already drank at least three cold drinks a week |
| Design | Randomised, parallel-group, open-label (people knew which drink they were assigned) |
| Intervention | At least two 330 ml servings a day of either water or diet drinks sweetened with aspartame, acesulfame-K and/or sucralose |
| Programme | Everyone joined a structured behavioural weight-management programme: 12 weeks of active weight loss, 40 weeks of assisted maintenance, then a voluntary 52-week unassisted extension (104 weeks total) |
| Two-year result | −4.8 kg (diet drinks) vs −3.7 kg (water); difference not statistically significant |
| Completed week 104 | 220 of 493 (117 water, 103 diet-drink) |
| Funding | American Beverage Association (also paid for the drinks and for medical-writing support); authors state the funder did not design, run, analyse or interpret the trial |
What Participants Actually Drank (and What the Water Group Gave Up)
The diet-drink group picked from about 20 commercially available products: diet colas, other diet fizzy drinks, low-calorie fruit drinks and flavoured waters, still or sparkling. The sweeteners were aspartame, acesulfame potassium and sucralose, with intakes below the European Food Safety Authority's safety thresholds. That is a narrower slice of the sweetener world than people sometimes assume. Stevia-based products, saccharin and sugar alcohols weren't the focus.
The water group had a harder assignment than it sounds. They weren't just told to drink water. They were asked to cut out artificially sweetened drinks entirely, including sweeteners stirred into tea and coffee. In other words, the comparison was "diet drinks allowed" against "sweeteners avoided", which is closer to how a real person would experience the choice.
Everyone also received the same counselling, goal-setting and support that comes with a behavioural programme. This matters for interpretation. The drink was one ingredient in a package, not a stand-alone treatment.
The Results, Checkpoint by Checkpoint
| Checkpoint | What the trial reported |
|---|---|
| Week 12 (active weight loss ends) | Weight loss was equivalent between diet-drink and water groups |
| Week 52 (assisted maintenance ends) | The 52-week analysis judged the groups non-equivalent, with significantly greater weight loss maintained in the diet-drink group |
| Week 104 (after a year without assistance) | −4.8 kg vs −3.7 kg; gap of 1.1 kg not significant; groups judged equivalent |
| Body composition | Both groups reduced waist and hip measurements and body fat, with no meaningful difference between them |
| Appetite and sugar intake | No significant between-group differences in hunger at the end; sugar intake fell in both groups, slightly more with diet drinks |
In everyday units, that two-year gap is roughly 10.6 pounds versus 8.2 pounds. Both groups stayed meaningfully lighter than where they started, even through the final year when the programme's support had stopped. For a field where regain is the norm, that last point may be the quietest and most encouraging finding in the whole paper.
The 52-Week Twist Most Headlines Skipped
Here's a detail that deserves more attention. The team had already reported the 52-week result separately, and at that point the diet-drink group had kept weight off to a statistically greater extent than the water group. The two arms were not "equivalent" at the one-year mark. By two years the difference had shrunk to 1.1 kg and was no longer significant.
There are two honest readings of that pattern. One says diet drinks gave a small early edge that faded as people drifted away from the programme. The other says the later gap narrowed because of who dropped out, or simply because small differences wobble over time. The data can't separate those explanations, which is why we'd treat the safest conclusion as the modest one: over two years, diet drinks did no harm to weight loss in this group, and may have helped a little early on.
An earlier trial from the University of Colorado and Temple University, led by John Peters, reached a similar direction over 52 weeks, finding diet drinks at least as good as water. So SWITCH is not a lone result. It is part of a small, consistent group of trials, which is also worth noting is mostly industry-funded.
The Dropout Problem: 220 of 493
Of 493 people randomised, only 220 completed the 104-week assessment: 117 in the water group and 103 in the diet-drink group. That's fewer than half. Long weight-management trials always lose people, and the final year here was voluntary, so heavy attrition isn't shocking. But it does limit how firmly the two-year numbers can be trusted.
Why? Because people who stay in a weight-loss study to the end are rarely a random slice of those who started. They're often the ones doing better, or the ones who find the programme easier to live with. If the type of person who drops out differs between the two drinks, that can nudge the comparison either way. Reputable trials use statistical methods to handle missing data, and the full paper will say which approach SWITCH used and how sensitive the conclusion is to it. It's the first thing we'd check before leaning hard on the 1.1 kg figure.
Who Paid for It, and How to Read Industry-Funded Nutrition Research
The American Beverage Association funded SWITCH. It also paid for the drinks and for a medical writer working under the authors' direction. The authors and the university state that the funder had no role in design, conduct, analysis or interpretation, though a representative was given the chance to review the manuscript. The trial was pre-registered, the university's team held the analysis, and the paper went through peer review. Those are the right safeguards.
Does that settle it? Not entirely, and it's fair to say so. A well-known 2007 analysis in PLoS Medicine by Lesser and colleagues found that nutrition studies funded entirely by industry were several times more likely to reach conclusions favourable to the sponsor than studies without industry funding. That doesn't mean any particular trial is wrong. It means funding is a reason to read the methods more carefully and to look for replication from independent teams.
A fair stance is neither "ignore it because the beverage industry paid" nor "believe it because the university says it was independent". Treat SWITCH as a decent-quality, pre-registered trial whose conclusions are plausible and consistent with earlier work, and note that it still needs confirmation from a team with no financial link to the product.
What SWITCH Cannot Tell You
- It doesn't show that diet drinks cause weight loss. Everyone was in a structured programme. The comparison was diet drink versus water inside that programme, not diet drink versus nothing.
- It doesn't compare diet drinks with sugary drinks. The big benefit of a diet drink comes from replacing a sugary one, as a dietitian outside the study pointed out, and that wasn't the question being tested here.
- It doesn't speak for people outside BMI 27–35. People with higher BMIs, with diabetes, or with very different diets weren't the target group.
- It doesn't cover every sweetener. Only aspartame, acesulfame-K and sucralose were used.
- It doesn't tell you about long-term gut, dental or cardiovascular outcomes. Weight was the main question, not every possible health effect of sweeteners.
- It was open-label. People knew what they were drinking, so expectation effects can't be ruled out, though they'd be hard to blind in a drink trial anyway.
Why Observational Studies and the WHO Still Sound the Alarm
If diet drinks are fine in trials, why do large population studies keep finding that heavy diet-drink drinkers weigh more? The most likely explanation is reverse causation. Think about who reaches for diet drinks: often people who are already gaining weight, already dieting, or already managing blood sugar. The drink looks like the cause, but it may be a marker of the problem someone is trying to fix. It's the same logic by which people who use inhalers appear to have more asthma. Randomised trials like SWITCH are designed to cut through exactly that tangle.
That said, the WHO's caution wasn't irrational. Its 2023 guideline reflected genuine uncertainty about long-term effects, and it applied broadly, with an exception for people who already have diabetes. SWITCH adds evidence on one outcome, body weight. It doesn't erase uncertainty about everything else.
Aspartame deserves a short note too, because the 2023 headlines confused many people. In July 2023 the WHO's cancer agency (IARC) classed aspartame as "possibly carcinogenic to humans" (Group 2B), a category that describes how strong the evidence is for hazard, not how large the risk is at ordinary intakes. On the same day, the joint FAO/WHO expert committee on food additives kept the acceptable daily intake at 40 mg per kg of body weight. For a 70 kg adult, that works out to somewhere around 9–14 cans of diet soft drink a day, depending on the sweetener content per can. The two-can-a-day pattern in SWITCH sits far below that line.
The Calorie Math of Swapping Sugary Drinks
The sturdiest part of the diet-drink argument has nothing to do with metabolism. It's arithmetic. If a drink has 140 calories and the replacement has about zero, you've removed 140 calories from your day without needing to feel hungry for it. Over a year those small removals add up.
| Daily swap | Calories removed per day | Per year | Theoretical fat-loss equivalent* |
|---|---|---|---|
| One 330 ml can of regular cola → zero-sugar or water | ≈140 kcal | ≈51,000 kcal | ≈6–7 kg |
| One 250 ml glass of orange juice → water | ≈110 kcal | ≈40,000 kcal | ≈5 kg |
| One 500 ml bottle of regular cola → zero-sugar or water | ≈210 kcal | ≈77,000 kcal | ≈10 kg |
*Using the old rule of thumb of roughly 7,700 kcal per kilogram of body fat. Real bodies compensate: appetite, activity and metabolism all adjust, so actual results are usually much smaller than this ceiling. The point isn't to promise a number. It's to show why swapping liquid calories is one of the cheapest changes available, and why the diet-versus-water question matters less than the sugar-versus-no-sugar question. We go through the common offenders in our guide to daily calorie needs, and you can see how much headroom a swap creates with the calorie calculator.
A Practical Decision Guide: Diet Drink or Water?
| If this is you… | What the evidence suggests |
|---|---|
| You drink two or more sugary sodas, juices or sweet teas a day | Replacing them with either water or diet drinks removes the most calories. SWITCH supports diet drinks as a reasonable bridge. |
| You already drink mostly water and like it | No weight-related reason to start diet drinks. Keep what's working. |
| Sweet tastes trigger cravings or overeating for you | An average across 493 people doesn't override your own pattern. Try sparkling water with citrus for three to four weeks and notice what happens to cravings. |
| You have phenylketonuria (PKU) | Avoid aspartame-containing products. Check labels for phenylalanine warnings. |
| You have enamel erosion or sensitive teeth | Diet drinks are still acidic. Drink them with meals rather than sipping all day, and rinse with water afterwards. |
| You rely on cola for caffeine | Count the caffeine separately and make sure you drink enough plain water too. Our water intake calculator helps set a daily target. |
| You're managing a medical condition or pregnancy | Follow your clinician's advice. Population guidance may not fit individual situations. |
Our own view is simple: treat diet drinks as a useful bridge away from sugary ones, not as a health food and not as a villain. Water remains the best default, and a mixed approach is fine for most people. To see how drink changes interact with your overall deficit, try the calorie deficit calculator. If you feel you're already eating "right" but progress has stalled, hidden intake is often the culprit, which we explain in why most people are eating the wrong number of calories.
The Bottom Line
SWITCH is the strongest long-term randomised evidence we have that diet drinks don't sabotage weight loss when they replace sugar inside a structured programme. It deserves to move the conversation forward. It also deserves honest labels: industry-funded, open-label, heavy attrition, narrow sweetener set, and no comparison with sugary drinks. Read it as "reassuring, not final".
If you're trading a sugary drink for a diet one, the trial says you're unlikely to be harming your progress. If you'd rather just drink water, nothing in the trial says you're missing out. And if you're hunting for the next big lever in your weight-loss plan, this isn't it. The levers that matter most are still total calories, protein, movement, sleep and consistency.
Frequently Asked Questions
Does diet soda stall weight loss?
In the SWITCH trial it didn't. Adults assigned to diet drinks lost 4.8 kg on average over two years versus 3.7 kg for water, and the difference wasn't statistically significant. That held inside a structured weight-management programme, so it doesn't prove diet soda would help or be neutral in every situation. But it counters the idea that sweeteners reliably undermine weight loss.
Is diet soda better than water for weight loss?
Not in any way that trial data can prove over two years. The groups were judged equivalent at week 104. At week 52 the diet-drink group had kept more weight off, but the gap narrowed by the end. Water has other advantages (no acidity, no sweeteners, no caffeine unless you add it), so the practical answer is that either can work and the real win is replacing sugary drinks.
Do artificial sweeteners make you hungrier?
SWITCH found no significant between-group difference in hunger or in sugar intake at the end of the trial, which is relevant because appetite compensation is one of the most common concerns. Individual experience varies, though, and the trial reports group averages, not everyone's response.
How does SWITCH fit with the WHO's 2023 advice against sweeteners for weight control?
The WHO's recommendation was conditional and reflected uncertainty, partly because long randomised trials were scarce. SWITCH adds long-term randomised data on body weight specifically. It doesn't address other outcomes the WHO considered, such as chronic disease risk, so it nudges the picture rather than overturning the guideline.
Is aspartame safe at the amounts in diet drinks?
The joint FAO/WHO committee in 2023 kept the acceptable daily intake at 40 mg per kg of body weight, which for a 70 kg adult corresponds to roughly 9–14 cans of diet soft drink a day. IARC's "possibly carcinogenic" label (Group 2B) reflects limited evidence of hazard, not a measure of risk at normal intakes. People with PKU must avoid aspartame.
Why did the researchers compare diet drinks with water instead of sugary drinks?
Because the unresolved question was whether sweeteners themselves interfere with weight loss compared with a neutral drink. Replacing sugary drinks with diet ones is already well supported on calorie grounds. Water is the cleanest control for testing whether the sweeteners add any penalty.
Should I worry that the study was funded by the beverage industry?
It's reasonable to read the methods more critically, since earlier research found industry-funded nutrition studies tend to favour sponsors. The trial was pre-registered, the university states it controlled the analysis, and similar results came from an earlier trial, but independent replication would strengthen confidence.
Does sparkling water count the same as still water?
For calories and hydration, yes. If it contains no added sweeteners or sugar, plain sparkling water fits the water side of this comparison. Flavoured versions vary, so check the label for sweeteners and acids.
What should I drink if I want to cut sugar but dislike both water and diet soda?
Try unsweetened tea or coffee, sparkling water with fruit slices, or diluted fruit squash made much weaker than usual. Gradually lowering sweetness over a few weeks often shifts taste preferences more than people expect.
References & Sources
- Harrold, J. A., Halford, J. C. G. et al. (2026). Effect of non-nutritive sweetened beverages v. water on body weight: long-term results of a randomised controlled trial. British Journal of Nutrition, 136(4), 482–493. doi:10.1017/S0007114526108150.
- Harrold, J. A. et al. SWITCH trial 52-week report: non-nutritive sweetened beverages versus water after a 52-week weight management programme (randomised controlled trial). ClinicalTrials.gov NCT02591134.
- University of Liverpool (2026). Press release on the SWITCH trial; ScienceDaily (2 October 2026), "Diet drinks work as well as water for weight loss, major trial finds."
- Peters, J. C. et al. (2016). The effects of water and non-nutritive sweetened beverages on weight loss and weight maintenance: a randomized clinical trial. Obesity, 24(2), 297–304.
- Tate, D. F. et al. (2012). Replacing caloric beverages with water or diet beverages for weight loss in adults: main results of the CHOICE randomized clinical trial. American Journal of Clinical Nutrition, 95(3), 555–563.
- World Health Organization (2023). Use of non-sugar sweeteners: WHO guideline.
- Riboli, E. et al. (2023). Carcinogenicity of aspartame, methyleugenol, and isoeugenol. The Lancet Oncology, 24(8), 848–850; JECFA (2023) evaluation of aspartame.
- Lesser, L. I. et al. (2007). Relationship between funding source and conclusion among nutrition-related scientific articles. PLoS Medicine, 4(1), e5.
- MacrosMeasure (2026). Free Calorie Calculator.
- MacrosMeasure (2026). Free Calorie Deficit Calculator.
Calculate your own numbers — free.
18 science-backed calculators covering everything in this article and more.