Sugar does not appear to make you fat in any special way, but it can add calories that do. In a meta-analysis of randomised trials, adults eating freely had slightly higher body weight when they ate more sugar, yet when sugar was swapped for other carbohydrates at the same calories, body weight did not change1.
That distinction, extra calories versus sugar itself, is what the rest of this article unpacks. It covers the nutrition research on body weight only; the other questions people ask about sugar are outside what these trials measured.
First, the words
- Free sugars are sugars added to food by a maker, a cook or you, plus the sugar in honey, syrups and fruit juice1. Sugar inside whole fruit is not part of this definition.
- Calories are the energy in food.
- Carbohydrate means starches and sugars together: bread, rice and pasta as well as table sugar.
- A meta-analysis is a study that pools the results of many studies into one estimate.
Two kinds of trial answer different questions:
- Eating freely (the paper calls it *ad libitum*): people are told to eat more or less sugar, with no limit on how much they eat overall1. This asks what happens in real life.
- Calorie-matched (the paper calls it *isoenergetic*): both diets carry the same calories; only the source changes, sugar or other carbohydrate1. This asks whether sugar has an effect of its own.
In plain terms: one kind of trial lets calories drift, the other holds them still. Comparing them shows whether sugar matters beyond its calories.
What was studied
The main evidence is a 2012 review done for the World Health Organization's sugar guideline1. The authors screened 7,895 trials and 9,445 cohort studies and found 30 trials and 38 cohort studies that qualified. Trials had to last at least two weeks, and trials designed for weight loss were excluded.
| What | Number |
|---|---|
| Trials included | 30 |
| Cohort studies included | 38 |
| Less-sugar trials, eating freely | 5 (10 weeks to 8 months) |
| More-sugar trials, eating freely | 10 (mostly sugary drinks) |
| Calorie-matched trials | 12 (2 weeks to 6 months) |
What the studies found
Eating freely, less sugar went with lower body weight. Across 5 trials in adults, the pooled average difference between groups was −0.80 kg (95% confidence interval −1.21 to −0.39)1. A confidence interval is the range the true value most likely sits in.
Eating freely, more sugar went with higher body weight. Across 10 trials, the pooled average difference was +0.75 kg (0.30 to 1.19). Most of these trials added sugary drinks1. Only two lasted longer than eight weeks, and in those two the difference was larger: +2.73 kg (1.68 to 3.78).
At the same calories, nothing changed. Across 12 calorie-matched trials, swapping sugar for other carbohydrates gave a pooled average difference of 0.04 kg (−0.04 to 0.13)1. The paper calls this "no evidence of difference".
Body weight difference between groups, by kind of sugar trial · kg
The authors' reading is direct: the change in body fatness "results from an alteration in energy balance rather than a physiological or metabolic consequence" of sugars1.
In plain terms: when sugar added calories, body weight rose a little. When sugar replaced other carbohydrate calorie for calorie, it made no difference.
Why sugary drinks come up again and again
Liquid sugar is the clearest case in the data. A 2023 meta-analysis pooled 85 articles on sugary drinks3. In adult trials against a non-caloric control:
- Adding sugary drinks led to a pooled average of 0.83 kg more body weight (0.47 to 1.19).
- Taking them away led to 0.49 kg less (−0.66 to −0.32).
Sugary drinks added or taken away: adult trials · kg
In adult cohort studies, which follow people over years, each extra daily serving was associated with 0.42 kg higher body weight (0.26 to 0.58)3. That is an association, not proof of cause.
Why drinks in particular is not something these trials tested. One common explanation is that calories from drinks may not reduce how much people eat at meals. Read that as a hypothesis, not a finding from these papers.
What about insulin?
A popular claim is that sugar makes you fat through insulin, the hormone that rises after carbohydrate meals and helps store energy. If that were the main route, calorie-matched swaps should show it. They did not1.
A tightly controlled metabolic-ward study tested a related idea. Nineteen adults with obesity lived on a research ward and received two diets with equal calories, one cutting carbohydrate and one cutting fat, each for six days2. Cutting fat led to more body fat loss per day (about 89 g) than cutting carbohydrate (about 53 g), group averages measured in a metabolic chamber. The authors' model predicted the body would narrow such differences over longer periods.
That study is short and small, and it was about carbohydrate in general, not sugar alone. It does not support the idea that carbohydrate has a unique fattening effect at matched calories.
In plain terms: the hormone story is a mechanism; the matched-calorie trials are the test, and they did not find sugar special.
The nuance
- "Sugar is harmless." This review measured body weight only. Other proposed effects of sugar were "beyond the scope of this review"1.
- "A calorie is a calorie, so sugar never matters." In real life, eating freely, more sugar did go with more body weight1. Sugar-rich foods can make it easy to take in more energy.
- "Sweet taste itself drives weight." A 2026 trial found no evidence for that over six months (see Latest research)4.
For how energy balance works in general, see calories in, calories out, and for where sugar sits among macros, how many carbs per day.
The studies at a glance
| Study | Who was studied | Key result | Year |
|---|---|---|---|
| Te Morenga, Mallard, Mann, *BMJ*1 | 30 trials and 38 cohort studies; adult trials used here | Same calories: 0.04 kg difference; eating freely: ±0.75 to 0.80 kg | 2012 |
| Hall et al., *Cell Metab*2 | 19 adults with obesity on a metabolic ward, 6 days per diet | Equal-calorie fat cut: more body fat loss than carbohydrate cut | 2015 |
| Nguyen et al., *Am J Clin Nutr*3 | 85 articles; adult trials in 1,343 people | Adding sugary drinks: +0.83 kg vs non-caloric control | 2023 |
| Čad et al., *Am J Clin Nutr*4 | 180 healthy adults, 6 months | Low, regular or high sweet-taste exposure: no difference in body weight | 2026 |
Limits worth knowing
- Short trials. Most trials in the 2012 review were short; only two of the more-sugar trials lasted over eight weeks1.
- Possible publication bias. Among the eating-freely trials there was a sign of it, though the result held after adjustment1.
- Intake measured with error. Food intake in free-living trials is reported, not observed.
- Body weight only. None of these reviews answers questions about teeth, the liver or blood sugar.
- Abstract only for three papers. We read the 2015, 2023 and 2026 papers as abstracts.
- Funding. The 2012 review was funded by the University of Otago, the Riddet Institute and the World Health Organization1.
Latest research
The newest trial we opened is from 20264. In 180 healthy adults, researchers supplied about half of daily energy needs for six months, with 7%, 35% or 80% of the provided foods and drinks tasting sweet. The sweet sources included sugars, low-calorie sweeteners, fruit and dairy. There were no differences between groups in sweet-taste liking, energy intake or body weight. It tested sweet taste, not sugar calories, so it adds to the picture rather than replacing the calorie-matched trials.
The short version
- When sugar was swapped for other carbohydrates at the same calories, body weight did not change.
- When people ate freely, more sugar went with slightly higher body weight, mostly from sugary drinks.
- The effect ran through calories, not a unique property of sugar, in these trials.
- These studies measured body weight only, not other health questions.
Educational overview only — not medical advice.
