Sweeteners: they help your weight only when they replace sugar

The strongest data on low-calorie sweeteners says one clear thing. They help with weight when they take the place of sugar, and that is where the help ends. A meta-analysis of 15 randomized trials found body weight fell by 1.8 lb (95% CI 0.95 to 2.6; metric: 0.80 kg, 0.43 to 1.17) and waist by 0.33 in (0.83 cm) with sweetened foods and drinks. The comparison in most of those trials was the sugared version, so the loss reflects calories removed. A larger review of intervention studies, not all of them randomized, split the question in two. Against sugar, sweeteners lowered weight by 2.3 lb (1.06 kg) in 2,267 people. Against water or nothing, in 1,068 people, the difference was 0.22 lb (0.10 kg), with an interval from 1.9 lb (0.87 kg) lower to 2.4 lb (1.07 kg) higher. That is no detectable effect. In 2023 the WHO issued a conditional recommendation against using non-sugar sweeteners to control weight or lower disease risk, for everyone except people who already have diabetes.

The numbers, per 3.5 oz (100 g)

What a real portion looks like

• One packet of sucralose or stevia tabletop sweetener weighs about 1 g in US survey data. — US household measures, FNDDS Evidence E source
  1 individual packet = 1 g for sucralose powder (food code 91107000) and stevia powder (91108000)

What your body absorbs

• Most sucralose is not absorbed: about 85% reaches the colon unchanged, where it meets the gut bacteria. — — Evidence D source
  About 85% of ingested sucralose reaches the colon unmetabolised
• In a crossover RCT on 20 volunteers, a berry-citrus drink sweetened with sucralose gave the highest plasma flavonoid metabolites, then stevia, then sucrose. — 20 volunteers, single acute intake of a maqui-citrus beverage with sucrose, stevia or sucralose (n = 20) Evidence B source
  Peak metabolite concentrations at 30-60 min; ranking sucralose > stevia > sucrose for most metabolites

What it does to your health

• Across 15 RCTs, low-calorie sweeteners, mostly compared with sugared versions, modestly lowered body weight by 1.8 lb (0.80 kg) and waist by 0.33 in (0.83 cm). — 15 randomized controlled trials of low-calorie sweeteners from foods, beverages or tabletop use (n = 15 RCTs) Evidence A source
  Body weight -1.8 lb (95% CI -2.6 to -0.95; metric: -0.80 kg, -1.17 to -0.43), fat mass -2.4 lb (-3.9 to -0.97; metric: -1.10 kg, -1.77 to -0.44), waist -0.33 in (-0.51 to -0.15; metric: -0.83 cm, -1.29 to -0.37)
• Across 7 short RCTs in healthy adults, sucralose did not change gut microbial diversity, fasting glucose or HOMA-IR. — healthy adults, ten RCTs of 8-61 participants each, seven pooled (n = 7 RCTs) Evidence A source
  Shannon SMD -0.00 (95% CI -1.16 to 1.15); fasting glucose SMD -0.04 (-0.63 to 0.56); HOMA-IR SMD -0.13 (-0.36 to 0.10)

Safety, limits and interactions

• Sugar-alcohol sweeteners are partly unabsorbed, and in a dose-dependent way they cause gas, discomfort and a laxative effect, including in people with IBS. — healthy volunteers and patients with irritable bowel syndrome, 79 studies (n = 79 studies) Evidence C source
  Dose-dependent malabsorption and GI symptoms; no pooled number

What people believe that the data does not support

• Low-calorie sweeteners lowered weight by 2.3 lb (1.06 kg) only when they replaced sugar. Against water they made no difference. — 88 intervention studies of at least 1 week; 2267 participants in LCS vs sugar, 1068 in LCS vs water or nothing (n = 2267 and 1068) Evidence A source
  Vs sugar -2.3 lb (95% CI -3.3 to -1.4; metric: -1.06 kg, -1.50 to -0.62); vs water or nothing +0.22 lb (-1.9 to 2.4; metric: 0.10 kg, -0.87 to 1.07)
• WHO advises against using non-sugar sweeteners to control weight or lower disease risk, for everyone except people who already have diabetes. — all people except those with pre-existing diabetes Evidence E source
  Conditional recommendation; review found no long-term benefit in reducing body fat
• Europe refused the label claim that intense sweeteners help keep a healthy body weight or control calorie intake. — — Evidence E source
  Claim non-authorized
• Europe also refused the claim that intense sweeteners have no effect on blood glucose control or insulin. — — Evidence E source
  Claim non-authorized

Who this works differently for

• In people with diabetes, a Cochrane review of RCTs found non-nutritive sweeteners left HbA1c unchanged versus placebo (0%, 360 people). — people with type 1 or type 2 diabetes in RCTs lasting 4 to 10 months (n = 360) Evidence A source
  HbA1c mean difference 0% (95% CI -0.1 to 0.1), very low certainty
• A 2026 meta-analysis in adults with diabetes linked artificial sweeteners to higher insulin and HbA1c, but the link vanished in random-effects models. — adults with diabetes, 22 studies (12 RCTs, 7 cross-sectional, 2 case-control, 1 cohort), 6,969 participants (n = 6969) Evidence C source
  Insulin g = 0.50 (95% CI 0.19 to 0.82), HbA1c g = 0.30 (0.06 to 0.54) in common-effects models only
• In children, non-sugar sweeteners instead of sugar slowed BMI z-score gain by 0.15, with no clear difference in body weight. — children in two studies, n=528 for BMI z score and n=467 for body weight (n = 528) Evidence C source
  BMI z score -0.15 (-0.17 to -0.12), body weight -1.3 lb (-2.9 to 0.31; metric: -0.60 kg, -1.33 to 0.14)

The bottom line

People with diabetes are the group the WHO left out, and the trials in them are thin. A Cochrane review pooled 4 trials with 360 people with type 1 or type 2 diabetes, lasting 4 to 10 months, and found no change in HbA1c against placebo (0%, 95% CI -0.1 to 0.1). The authors rated that very low certainty, because none of their nine trials was at low risk of bias in every domain. A 2026 meta-analysis in 6,969 adults with diabetes linked artificial sweeteners to higher insulin and HbA1c, but only in common-effects models. In random-effects models the link disappeared. That question is still open for a stated reason. The 2026 analysis pooled 12 trials with cross-sectional, case-control and cohort studies, and the results between them varied a lot. Europe refused both label claims tested here: that intense sweeteners help keep a healthy weight, and that they have no effect on blood glucose or insulin. A refusal means the evidence was judged insufficient at assessment. It proves neither side. For the gut worry about sucralose, 7 pooled randomized trials in healthy adults, each with 8 to 61 people and all of them short, found no change in microbial diversity, fasting glucose or insulin resistance. The review authors add, citing older work, that about 85% of sucralose reaches the colon unchanged. In children, two studies with 528 children found that non-sugar sweeteners instead of sugar slowed the rise in BMI z-score by 0.15, while the gap in body weight (1.3 lb / 0.60 kg, in 467 children) was not significant. Those studies mixed randomized and observational designs. The one caution backed by data here is the sugar alcohols. A review of 79 studies found that polyols are partly unabsorbed and cause gas, abdominal discomfort and a laxative effect that grows with the dose, in healthy volunteers and in people with irritable bowel syndrome. It gave no pooled number, and doses differ by polyol. We found no card on sweeteners in pregnancy or alongside medicines. A tabletop packet of sucralose or stevia weighs about 1 g in US survey data, and most of that gram is bulking agent.

Compared with other foods

Sources

1. swtn-r8-13 · reference dataset
   FNDDS 2710263 Sugar substitute, sucralose, powder: 1 individual packet 1 g · FNDDS 2710264 Sugar substitute, stevia, powder: 1 individual packet 1 g, 1 teaspoon 3 g
   USDA FNDDS 2021-2023 · checked 2026-09-28
2. swtn-r8-06 · pharmacokinetic statement
   Sucralose is a widely consumed non-nutritive sweetener, with approximately 85% reaching the colon unmetabolized and directly interacting with the gut microbiota.
   Int J Mol Sci, 2026 · checked 2026-09-28
3. swtn-r8-07 · randomized crossover trial
   In general, sucralose provided the greatest absorption value for most of these metabolites, followed by stevia.
   Nutr Metab Cardiovasc Dis, 2021 · checked 2026-09-28
4. swtn-r8-01 · meta-analysis of RCTs
   In RCTs, LCSs modestly but significantly reduced all outcomes examined, including body weight (-0.80 kg; 95% CI: -1.17, -0.43), body mass index [BMI (in kg/m²): -0.24; 95% CI: -0.41, -0.07], fat mass (-1.10 kg; 95% CI: -1.77, -0.44), and waist circumference (-0.83 cm; 95% CI: -1.29, -0.37).
   Am J Clin Nutr, 2014 · checked 2026-09-28
5. swtn-r8-05 · meta-analysis of RCTs
   Sucralose did not significantly alter alpha diversity (Shannon index: SMD = -0.00; 95% CI -1.16 to 1.15; p = 0.99; I2 = 25.67%), fasting blood glucose (SMD = -0.04; 95% CI = -0.63 to 0.56; p = 0.88; I2 = 69.61%), and HOMA-IR (SMD = -0.13; 95% CI = -0.36 to 0.10; p = 0.212; I2 = 0.00%).
   Int J Mol Sci, 2026 · checked 2026-09-28
6. swtn-r8-08 · systematic review
   Overall, available work has shown that polyol malabsorption generally occurs in a dose-dependent fashion in healthy individuals, and malabsorption increases when polyols are ingested in combination. ... Polyols can induce dose-dependent symptoms of flatulence, abdominal discomfort, and laxative effects when consumed by both healthy volunteers and patients with IBS.
   Adv Nutr, 2017 · checked 2026-09-28
7. swtn-r8-02 · meta-analysis of intervention studies
   Results showed an effect in favour of LCS vs sugar for BW (29 parallel-groups studies, 2267 participants: BW change, -1.06 kg, 95% CI -1.50 to -0.62, I2 = 51%), BMI and EI. ... Overall, results showed no difference in effects of LCS vs water/nothing for BW (11 parallel-groups studies, 1068 participants: BW change, 0.10 kg, 95% CI -0.87 to 1.07, I2 = 82%), BMI and EI
   Int J Obes (Lond), 2021 · checked 2026-09-28
8. swtn-r8-10 · government agency guidance page
   The recommendation applies to all people except individuals with pre-existing diabetes and includes all synthetic and naturally occurring or modified non-nutritive sweeteners that are not classified as sugars found in manufactured foods and beverages, or sold on their own to be added to foods and beverages by consumers.
   WHO, news release 15 May 2023: WHO advises not to use non-sugar sweeteners for weight control · checked 2026-09-28
9. swtn-r8-11 · regulatory register entry
   POL-HC-8016 Table top sweeteners and foods beverages containing intense sweeteners -intense sweetners help to maintain a healthy body weight; -intense sweetners help to control calorie intake. Non-authorised
   EU Register on nutrition and health claims, claim POL-HC-8016, snapshot 2026-09-21 · checked 2026-09-28
10. swtn-r8-12 · regulatory register entry
    POL-HC-8015 Table top sweeteners and foods beverages containing intense sweeteners -intense sweeteners have no effect on carbohydrate metabolism, short or long-term blood glucose control or insulin secretion; -product [x] assists in blood glucose control. Non-authorised
    EU Register on nutrition and health claims, claim POL-HC-8015, snapshot 2026-09-21 · checked 2026-09-28
11. swtn-r8-03 · Cochrane review
    The MD for HbA1c was 0%, 95% CI -0.1 to 0.1; P = 0.99; 4 trials; 360 participants; very low-certainty evidence.
    Cochrane Database Syst Rev, 2020 · checked 2026-09-28
12. swtn-r8-04 · meta-analysis, mixed designs
    Artificial sweeteners were associated with higher insulin (g = 0.50, 95% CI, 0.19-0.82) and higher HbA1c (g = 0.30, 95% CI, 0.06-0.54) in common-effects models; however, these associations were not relevant in random-effects models, and heterogeneity was substantial.
    J Med Life, 2026 · checked 2026-09-28
13. swtn-r8-09 · systematic review and meta-analysis
    In children, a smaller increase in body mass index z score was observed with NSS intake compared with sugar intake (-0.15, -0.17 to -0.12; two, n=528, moderate certainty of evidence), but no significant differences were observed in body weight (-0.60 kg, -1.33 to 0.14; two, n=467, low certainty of evidence)
    BMJ, 2019 · checked 2026-09-28