Slightly, in the trials that tested it, and not by a meaningful amount. In the longest trial, adults in a weight programme who drank diet drinks kept off 1.4 kg more over a year than those who drank water, a gap the authors called statistically but "not clinically significant"3. Both options are close to zero calories; in pooled trials, the larger gap is between diet drinks and sugar-sweetened drinks5.
Diet soda here means drinks sweetened with non-nutritive sweeteners, sweeteners that add sweetness with almost no calories. Below is what the trials compared, why the results lean one way, and why they are weaker than they look.
What was studied
Two randomised nutrition trials, one reported twice, put diet drinks and water head to head inside a behavioural weight programme, the kind with group sessions and weekly weigh-ins. Randomised means people were assigned to a drink by chance.
| Trial | Who | Drinks | Length |
|---|---|---|---|
| Peters 2014 / 2016, US | 303 adults with overweight or obesity who regularly drank diet drinks | 710 ml a day of diet drinks, or water (the water group stopped diet drinks) | 12 weeks, then 1 year |
| Harrold 2024, England | 493 adults, BMI 27 to 35, who regularly drank cold drinks | at least two 330 ml servings a day of diet drinks, or water (the water group gave up sweeteners altogether) | 52 weeks |
In both, everyone got the same weight programme. Only the drink differed. Both were designed as equivalence trials: they set out to test whether the two drinks give the same result, within a pre-set margin.
In plain terms: same programme, same support, one variable. Diet drinks or water.
What the studies found
The 2024 English trial. At 52 weeks, among the 262 people who finished, weight loss averaged 6.1 kg with water and 7.5 kg with diet drinks3. The difference, 1.4 kg (90% confidence interval 0.2 to 2.6 kg), meant the two were not equivalent. A confidence interval is the range the true difference most likely sits in.
But the authors had set 1.5 kg in advance as the smallest difference that would matter clinically, and 1.4 kg fell short. Their conclusion: weight loss was maintained "to a statistically greater extent" with diet drinks, but the difference "was not clinically significant"3.
Only 53% of the people randomised reached week 52. When the authors filled in the missing weights by two standard methods, the gap shrank to 0.8 kg and 0.6 kg, and neither was significant3.
The earlier US trial. After 12 weeks of the programme, the diet-drinks group had lost 5.95 kg against 4.09 kg with water, and reported less hunger1. At one year, the gap was wider: 6.21 kg kept off with diet drinks against 2.45 kg with water, counting dropouts as having no change2.
Weight change with water vs diet drinks in head-to-head trials · kg
Water Diet drinks
In plain terms: both trials lean toward diet drinks. The longer, larger trial found a gap under the size its own authors called meaningful, and the gap faded once dropouts were counted.
Why the US result was bigger
The detail that matters is who was asked to change. In the US trial, everyone was already a regular diet-drink user2. The diet-drinks group kept their habit; the water group had to give it up. The authors list this as a limit: results may differ in people who do not usually drink diet drinks.
The English trial mixed habits: about a quarter of participants had rarely drunk diet drinks before, and the water group was asked to stop sweeteners in hot drinks too3. Its gap was smaller. One reading, ours rather than either paper's, is that some of the US gap reflects the cost of giving up a familiar drink rather than a property of the drink.
Who paid for the trials
Both trials were funded by the same US soft-drinks industry trade association32. Both papers state that the funder was not involved in the design, conduct or interpretation, and both used a third party to audit the data. Funding does not make a result wrong. It is a reason to weight independent pooled evidence alongside it.
What pooled evidence says
A 2025 network meta-analysis pooled 78 randomised trials (4,168 adults) comparing drinks5. Low- or no-calorie sweetened drinks came out 0.79 kg lower than water (95% credible interval −1.35 to −0.18). But in the trials at low risk of bias, and in trials that also cut calories, there was no significant effect. The authors rated the certainty of most comparisons as low.
A 2024 meta-analysis of water trials found the opposite lean in a small subgroup: in two trials where water replaced diet drinks, the water groups ended 1.82 kg heavier4. The authors themselves suggest this may be chance, given only two trials, one of which dominated.
Diet drinks vs water in pooled analyses · kg
Both analyses also point against sugar. Swapping water for sugar-sweetened drinks leaned toward a small loss (−0.81 kg, borderline and not significant)4, and diet drinks beat sugar-sweetened drinks by about 1.08 kg5. Replacing drinks with calories may matter more than which zero-calorie drink replaces them. Calories in, calories out explains why.
In plain terms: diet drinks and water are close. Diet drinks are clearly ahead of drinks with sugar; for water, the edge over sugary drinks was borderline.
What this does not answer
The trials measured body weight over a year inside a structured programme. They did not test health outcomes, long-term use or people outside a programme, and this article makes no claim about those. For individual questions about sweeteners and health, a doctor is the right person to ask.
The studies at a glance
| Study | Who was studied | Key result | Year |
|---|---|---|---|
| Peters et al., *Obesity*1 | 303 adults with overweight or obesity, regular diet-drink users; 12 weeks | Diet drinks −5.95 kg vs water −4.09 kg | 2014 |
| Peters et al., *Obesity*2 | Same 303 adults, 1 year | Diet drinks −6.21 kg vs water −2.45 kg kept off | 2016 |
| Harrold et al., *Int J Obes*3 | 493 adults, BMI 27 to 35, England; 52 weeks | Diet drinks −7.5 kg vs water −6.1 kg; not clinically significant; no significant difference with dropouts imputed | 2024 |
| Chen et al., *Nutrients*4 | 8 RCTs of water, people with overweight or obesity | Water replacing diet drinks: +1.82 kg (2 trials); replacing sugary drinks: −0.81 kg | 2024 |
| Shahinfar et al., *Int J Obes*5 | 78 RCTs, 4,168 adults, network meta-analysis | Low-calorie sweetened drinks −0.79 kg vs water; not significant in low-risk-of-bias trials | 2025 |
| Li et al., *J Endocrinol Invest*6 | 19 RCTs of sweeteners | −0.79 kg overall, mainly vs sugar; not significant beyond 18 weeks | 2026 |
Limits worth knowing
- Dropouts. Only 53% finished the 52-week trial, and imputing the missing data shrank the difference to non-significance3; 73% finished the US trial2.
- Habit. The US trial studied only regular diet-drink users, so the water group was also a "give up your usual drink" group.
- Industry funding for both head-to-head trials, with independent data audits.
- Inside a programme. Everyone had structured weight-management support; results may differ without it.
- Group averages. Individuals varied widely around every number here.
- Abstract only for the 2014, 2025 and 2026 papers.
Latest research
The newest paper we opened is a 2026 meta-analysis of 19 randomised trials of sweeteners6. Replacing caloric sugars with sweeteners lowered weight by 0.79 kg (95% CI −1.55 to −0.04) overall, and the effect was significant only against caloric sugars and in trials shorter than 18 weeks; beyond 18 weeks it was minimal, and in participants with obesity it was not significant. It did not test sweeteners against water specifically. We read its abstract only. Read with the 2024 trial, the current picture is that diet drinks and water end within a clinically small margin over a year, and diet drinks do better than sugar.
The short version
- Head to head, diet drinks edged out water by 1.4 kg over a year in the largest trial, below the authors' own threshold for a meaningful difference.
- With dropouts accounted for, that gap was no longer significant.
- An earlier trial in regular diet-drink users found a bigger gap; both trials were industry-funded.
- In pooled trials, the larger difference is between diet drinks and sugar-sweetened drinks; water's edge over sugary drinks was borderline.
Educational overview only — not medical advice.
