← Longevity & Supplement Guides
Diet Drinks vs Water: What a Two-Year Trial Really Settled
This week the verdict went round the health pages: diet drinks work as well as water for weight loss. It arrived like a ceasefire in one of nutrition’s longest arguments. One side has spent a decade calling sweeteners a metabolic trap. The other keeps a fridge full of zero-sugar cans and feels quietly vindicated.
The study behind it is unusually good for this field. It is randomised, it ran for two years, and it was published in a peer-reviewed journal. But the most useful thing in it is not the comparison everyone quoted. It is the line on the chart that both camps ignored.
The trial, in one paragraph
Researchers at the Universities of Liverpool and Leeds recruited 493 adults with a BMI between 27 and 35 who drank cold drinks regularly. Half were asked to drink at least two 330 ml servings of water a day and to give up diet drinks. The other half were asked to drink at least two servings of sweetened, zero-calorie drinks a day, and could drink water as well. Everyone joined the same behavioural weight programme: weekly group sessions for twelve weeks, then monthly sessions until the end of the first year. In the second year, they were on their own.
After two years, the diet-drink group was 4.8 kg lighter than at the start and the water group 3.7 kg lighter. The authors tested whether the two results were equivalent, and they were. At one year, in an earlier report, the diet-drink group had been about 1.4 kg ahead — a gap the same team called statistically, but not clinically, significant.
The two-year trial
Average change in body weight (kg) from the start of the programme
The drink was never the main event
Look at the shape of the lines rather than the gap between them. Almost all of the weight came off in the first twelve weeks, when people met every week. It largely stayed off while they met monthly. Once the meetings stopped, both lines started to drift back up.
The difference between the two drinks is about a kilogram. The difference between having support and not having it is visible in both groups. The trial was designed to compare drinks, so it cannot prove that the meetings did the work. But the curve is hard to read any other way.
This is the part I find most useful, and it has nothing to do with sweeteners. We tend to argue about ingredients because ingredients are easy to argue about. Structure — a weekly check-in, a plan someone else expects you to follow — is harder to sell and harder to put on a label. It is also what moved the numbers here.
Why the WHO said the opposite
In 2023 the World Health Organization advised people not to use non-sugar sweeteners for weight control. If you remember that headline, this week’s one looks like a contradiction. It is less of one than it seems.
The WHO recommendation was explicitly conditional, and the evidence behind it was rated low certainty. Much of the worry came from long observational studies, where people who drink a lot of diet drinks tend to weigh more over time. The trouble is direction. People who are already gaining weight are exactly the people most likely to switch to diet drinks. The can may be a sign of the problem rather than its cause.
Randomised trials avoid that trap, because the drink is assigned, not chosen. Most of them are short, though, which is why a two-year trial matters. It does not overturn the WHO’s caution about long-term use in general. It does answer one narrow question well: for adults in a weight programme, switching water for diet drinks did not undo the programme.
Who paid, and who left
Two things belong next to the result, and both are in the paper. The trial was funded by the American Beverage Association, and several authors report links to the sweetener and soft-drink industry. The authors say the funder had no role in the analysis. Randomisation protects the comparison itself, but funding tends to show up in what gets asked and how it is framed. This trial asked whether diet drinks are as good as water, not whether either one is better than changing the habit altogether.
The second is that more than half of the people who started did not finish both years. Of 493, 220 were weighed at week 104. The researchers used two methods to account for the missing data, and neither changed the conclusion. But dropout on that scale always leaves a question: the people who stayed may be the people for whom the plan was easiest.
What the drink replaces matters more than what it is
Every comparison in this debate has a hidden second half: diet drinks instead of what? The answer changes the result more than the sweetener does.
Same can, three comparisons
What the evidence says depends on what the diet drink replaces
Instead of a sugary drink. A 2021 meta-analysis of sustained trials found about 1 kg less body weight with sweetened drinks than with sugar-sweetened ones, and more the more sugar was replaced.
Instead of water. The two-year trial above: equivalent weight change inside a structured programme, with no difference in reported hunger.
Instead of changing the habit. No trial here asks whether people do better when the sweet drink simply stops being a daily ritual. That is the comparison most of us actually face.
Read that way, the two camps are both partly right. If a can of diet cola is standing in for a sugary one, it is a reasonable swap. If it is standing in for water, this trial suggests the cost to your weight is small at most. Neither side gets to call it a health drink.
The sweet-tooth question, honestly
One worry about sweeteners is that they keep a craving for sweetness alive. The trial offers a small, interesting hint. The water group used noticeably less sweetener over time — which is no surprise, since they had been asked to stop. But hunger rose slightly in both groups, and it did not differ between them.
That does not settle the craving argument. It suggests the effect, if it exists, is not big enough to show up on the scale over two years. For most people the real question is personal and easy to test: does a diet drink at 4 pm end the craving, or start a search for something to eat with it?
What to try this week
Pick the one daily drink you reach for without thinking. Write down what it replaced when the habit began — water, a sugary drink, a coffee, a break. That tells you more about whether it helps you than any trial can.
If it replaced something sugary, keep it without guilt. If you are curious whether it is doing anything for you, swap it for sparkling water for two weeks and notice two things: your afternoon hunger, and your waist measured at the same point each Sunday. Our guide to why waist size often says more than BMI explains how. For the rest of what you drink, hydration and electrolytes covers the basics.
And if the trial’s curve looks familiar, it should. Our post on losing weight versus keeping it off tells the same story from the training side: what you keep doing after the programme ends counts for more than the programme. If you use the Agen Band, judge any change by the trend across weeks, not by a single morning.
The bottom line
The two-year trial found that, inside a structured weight programme, diet drinks and water did about equally well. That is a useful result, and a better-designed one than most of the evidence the sweetener debate runs on. It came with industry funding and heavy dropout, and both belong in the same sentence as the headline.
But the loudest part of the chart is not the gap between two drinks. It is the slow climb that starts when the weekly support stops. Use numbers to correct the fantasy, not to replace the experience. The fantasy was that the right can would do the work. The experience, again, is that the habits around the can did it.
