Skip to content
HabitSeek
Back to results

Cutting sugar-sweetened drinks

Reducing or replacing sugar-sweetened beverages such as soda, sweetened juice drinks, sweetened iced tea and energy drinks with non-caloric alternatives.

At a glance

  • Evidence: Trials that reduce or replace sugar-sweetened beverages (SSBs) with non-caloric drinks produce small but consistently favourable changes in body weight and adiposity.Body weight / adiposity change (randomized trials and prospective cohorts) · Moderate certainty · Randomized intervention
  • What’s uncertain: Many included trials are small (the substitution network meta-analysis pooled 1733 adults across 17 RCTs) and short (2 weeks to a few months), and outcome certainty was graded moderate at best.
  • First step: Editorial starting point, not a validated prescription: pick the one sugar-sweetened drink you have most often and replace it with a non-caloric alternative.
  • Before you try it: Replacing sugar-sweetened drinks with drinks that are still sweetened and caloric, such as sweetened fruit juice drinks, does not follow the trial evidence. All precautions
Read the full takeaway

Drink substitutions

The evidence supports cutting sugar-sweetened drinks and replacing them with non-caloric alternatives. Pooled randomized trial evidence supports low- and no-calorie sweetened drink substitution, which reduced body weight by about 1 kg over the moderate term, and trials that subtracted sugar-sweetened drinks showed weight loss while trials that added them showed weight gain.

Water and cohort evidence

Water substitution, however, was null in the pooled analysis, so this habit should not be sold as proven weight benefit from swapping to water. Prospective cohorts consistently link higher intake to incident type 2 diabetes and to higher all-cause mortality, with the diabetes association only partly explained by body weight.

Limits

The evidence stops short of certainty: mortality evidence is graded low quality, absolute effects are small, and key syntheses carry food-industry competing interests including the Calorie Control Council, International Sweeteners Association, International Life Sciences Institute, National Honey Board and Canadian Sugar Institute.

A glass of cola
NutritionLong-term healthEnergyUnder 10 min · Easy · Free
Reviewed Sep 23, 2026
Save

Scientific evidence

What the research shows

Outcome assessments based on research, including its uncertainties and limitations.

Body weight / adiposity change (randomized trials and prospective cohorts)

Finding: Benefit supportedCertainty: Moderatestudies found a benefit from doing this

Evidence basis: randomised trials

Trials that reduce or replace sugar-sweetened beverages (SSBs) with non-caloric drinks produce small but consistently favourable changes in body weight and adiposity.

Population: Adults and children in prospective cohort studies and randomized controlled trials; adults in the substitution network meta-analysis had overweight or obesity and were at risk for or had diabetes.

Research details

Summary

Trials that reduce or replace sugar-sweetened beverages (SSBs) with non-caloric drinks produce small but consistently favourable changes in body weight and adiposity. In a 2023 systematic review and meta-analysis, randomization to subtraction of SSBs produced weight loss (MD -0.49 kg; 95% CI -0.66 to -0.32) and addition of SSBs produced weight gain (MD 0.83 kg; 95% CI 0.47 to 1.19) versus control, while SSB-reduction trials in children showed less BMI gain (MD -0.21 kg/m2; 95% CI -0.40 to -0.01). A network meta-analysis of RCTs found that substituting low- and no-calorie sweetened beverages for SSBs reduced body weight by 1.06 kg, but substituting water for SSBs was not associated with any outcome in the pooled analysis. Effects are small in absolute terms and cohort data show only modest dose-response associations.

Why this certainty

Rated Moderate: the direction of effect is consistent across randomized and cohort designs with statistically significant pooled estimates, and the narrative is corroborated by an independent umbrella review that graded the SSB body-weight association as moderate quality. It is not Strong because absolute effects are small, water substitution alone was null in pooled RCT analysis, industry-linked competing interests are pervasive, and GRADE certainty was reported as moderate or low rather than high.

Limitations

Many included trials are small (the substitution network meta-analysis pooled 1733 adults across 17 RCTs) and short (2 weeks to a few months), and outcome certainty was graded moderate at best. Cohort estimates carry residual confounding and self-reported SSB intake. Several authors of the key syntheses report competing interests from food-industry-linked bodies including the Calorie Control Council, International Sweetener Association, International Life Sciences Institute, National Honey Board, Canadian Sugar Institute and Institute for the Advancement of Food and Nutrition Sciences.

Effect estimate

Adults: subtraction of SSBs vs control MD -0.49 kg (95% CI -0.66 to -0.32); addition of SSBs vs control MD 0.83 kg (95% CI 0.47 to 1.19); children: SSB-reduction RCTs MD -0.21 kg/m2 (95% CI -0.40 to -0.01); LNCSB substituted for SSB MD -1.06 kg (95% CI -1.71 to -0.41)

Sources for this outcome:

Incident type 2 diabetes (prospective cohorts)

Finding: Benefit supportedCertainty: Moderatestudies found a benefit from doing this

Evidence basis: observational studies (association, not proof of cause)

Multiple independent dose-response meta-analyses of prospective cohorts find that higher habitual sugar-sweetened drink intake is associated with higher type 2 diabetes incidence, only partly explained by adiposity.

Population: Adults without diabetes at baseline from prospective cohort studies conducted in Europe, the United States, Asia, Australia and Latin America; large occupational and population cohorts including health professionals and nurses.

Research details

Summary

Multiple independent dose-response meta-analyses of prospective cohorts find that higher habitual SSB consumption is associated with higher incidence of type 2 diabetes, and that the association is only partly explained by adiposity. Imamura and colleagues pooled 17 cohorts (38,253 cases/10,126,754 person years) and reported an 18% higher incidence per one serving/day before adjustment for adiposity and 13% after adjustment. A 2025 dose-response meta-analysis of 18 cohorts reported RR 1.25 (95% CI 1.17 to 1.35) per additional daily serving, and an earlier pooled analysis found a 26% higher risk in the highest versus lowest intake category.

Why this certainty

Rated Moderate: at least four independent meta-analyses of prospective cohorts agree on direction and statistical significance, a dose-response gradient is present, the association survives adjustment for adiposity, and a separate umbrella review of dietary factors and incident type 2 diabetes graded the SSB association as high quality of evidence (one serving/day, HR 1.26, 95% CI 1.11 to 1.43; Neuenschwander 2019). It is not Strong because the evidence is entirely observational, heterogeneity is high, and no randomized trial has tested type 2 diabetes incidence as an endpoint.

Limitations

All estimates are observational, with self-reported beverage intake and residual confounding by overall diet, weight trajectory and socioeconomic factors. Heterogeneity was substantial (I2 79-89% for SSB in some analyses). Effect estimates are per serving/day and cannot establish causality. The 2025 meta-analysis reported conflicts of interest from one author with the International Life Sciences Institute Europe and other food-sector bodies.

Effect estimate

Per 1 serving/day SSB: RR 1.18 (95% CI 1.09 to 1.28; I2 = 89%) before adiposity adjustment and RR 1.13 (95% CI 1.06 to 1.21; I2 = 79%) after; per additional serving/day: RR 1.25 (95% CI 1.17 to 1.35); highest vs lowest quantile: RR 1.26 (95% CI 1.12 to 1.41)

Sources for this outcome:

All-cause mortality (long-term clinical outcome, prospective cohorts)

Finding: Benefit supportedCertainty: Limitedstudies found a benefit from doing this

Evidence basis: observational studies (association, not proof of cause)

Prospective cohort evidence links higher long-term SSB intake to higher all-cause mortality, with a graded dose-response.

Population: Middle-aged and older US adults in the Nurses' Health Study and Health Professionals Follow-up Study, plus adults in prospective cohorts included in dose-response meta-analyses.

Research details

Summary

Prospective cohort evidence links higher long-term SSB intake to higher all-cause mortality, with a graded dose-response. In the Nurses' Health Study and Health Professionals Follow-up Study (118,363 participants; 36,436 deaths over 3,415,564 person-years), pooled hazard ratios across increasing intake categories below once a month, one to four a month, two to six a week, one to under two a day and two or more a day were 1.00, 1.01, 1.06, 1.14 and 1.21, and each additional serving a day was associated with 7% higher mortality. A dose-response meta-analysis of 39 articles found a 4% higher all-cause mortality risk per 250 mL/day of SSB, and an umbrella review graded the SSB all-cause mortality association as low quality.

Why this certainty

Rated Limited: the direction and dose-response are consistent and the cohorts are large with long follow-up and repeated dietary measures, but all evidence is observational, effect sizes are small, attenuation after adjustment for intermediates indicates that much of the association runs through diagnosed intermediate disease, and the umbrella review explicitly graded the evidence as low quality. No trial has tested mortality as an endpoint.

Limitations

Observational design with self-reported intake; the association attenuates after adjustment for intermediate conditions that were diagnosed during follow-up - incident hypertension, hypercholesterolaemia, type 2 diabetes, coronary heart disease and stroke - from HR 1.07 to 1.05 per serving/day, which is a mediation structure rather than a clean independent effect, and reverse causation cannot be excluded. The same paper reports a sex interaction for total mortality (p = 0.02) with stronger associations in the women's cohort, and explicitly declines to call it biological: "Whether there is a biological basis for this difference is not clear." Cohorts are predominantly white US health professionals, limiting generalisability. The umbrella review graded this association as low quality.

Effect estimate

Pooled HR per 1 serving/day increment in SSB: 1.07 (95% CI 1.05 to 1.09); highest (>=2/day) vs lowest (<1/month) category pooled HR 1.21 (95% CI 1.13 to 1.28; p-trend <0.0001); per 250 mL/day increment RR 1.04 (95% CI 1.01 to 1.07; I2 = 58.0%)

Sources for this outcome:

What people say

Who discusses this?

Public statements, with positions specific to each source and topic. These views do not determine the scientific evidence assessment.

What the research shows ↑

No sourced public statements added for this habit yet. This does not mean no one discusses it.

Practical steps and precautions

How to try it

  1. Editorial starting point, not a validated prescription: pick the one sugar-sweetened drink you have most often and replace it with a non-caloric alternative. In trials, this substitution was tested over roughly 2 weeks to several months and produced small average weight changes of about 1 kg at most.
  2. Choose the substitute deliberately. Pooled trial evidence supports low- and no-calorie sweetened drinks as the replacement with a measurable weight signal; pooled trials of water as the replacement showed no association with any outcome, which is an absence of evidence rather than evidence of harm.
  3. Editorial framing, not a study-tested threshold: treat it as an every-day drink swap rather than a one-off. Cohort dose-response analyses report risk rising per additional daily serving, so the pattern of habitual intake is what the long-term evidence tracks.
  4. If fizzy drinks are the main source, keep a non-caloric option cold and visible at home so the default is the replacement rather than the sugar-sweetened version.

Precautions

  • Replacing sugar-sweetened drinks with drinks that are still sweetened and caloric, such as sweetened fruit juice drinks, does not follow the trial evidence. Sugar-sweetened fruit juice was itself associated with higher incident type 2 diabetes in pooled cohort analysis (RR 1.28; 95% CI 1.04 to 1.59), while 100% fruit juice was not (Xi 2014, doi 10.1371/journal.pone.0093471).
  • Long-term safety of high intakes of low- and no-calorie sweetened drinks is not settled. In the substitution network meta-analysis the certainty of evidence for substituting these drinks for sugar-sweetened drinks was moderate, and substituting them for water was associated with higher glycated haemoglobin (MD 0.21%, 95% CI 0.02 to 0.40) and lower systolic blood pressure (MD -2.63 mmHg, 95% CI -4.71 to -0.55), so they are not a demonstrated upgrade over water.
  • Most included randomized trials were short (2 weeks to a few months) and involved adults with overweight or obesity who were at risk for or had diabetes, so the size of any weight effect in other groups is not established.
  • Do not present this as a treatment for diagnosed type 2 diabetes. The cohort evidence concerns incident diabetes in people without diabetes at baseline, and no randomized trial tested diabetes incidence as an endpoint.

References

Sugar-sweetened beverage consumption and weight gain in children and adults: a systematic review and meta-analysis of prospective cohort studies and randomized controlled trials

Nguyen M, Jarvis SE, Tinajero MG, Yu J, Chiavaroli L, Mejia SB, Khan TA, Tobias DK, Willett WC, Hu FB, Hanley AJ, Birken CS, Sievenpiper JL, Malik VS · The American Journal of Clinical Nutrition · 2023

Systematic review and meta-analysis of prospective cohort studies and randomized controlled trialsVerified from abstract
Open original source

Association of Low- and No-Calorie Sweetened Beverages as a Replacement for Sugar-Sweetened Beverages With Body Weight and Cardiometabolic Risk: A Systematic Review and Meta-analysis

McGlynn ND, Khan TA, Wang L, Zhang R, Chiavaroli L, Au-Yeung F, Lee JJ, Noronha JC, Comelli EM, Blanco Mejia S, Ahmed A, Malik VS, Hill JO, Leiter LA, Agarwal A, Jeppesen PB, Rahelic D, Kahleova H, Salas-Salvado J, Kendall CWC, Sievenpiper JL · JAMA Network Open · 2022

Systematic review and network meta-analysis of randomized clinical trialsVerified from abstract
Open original source

Dietary sugar consumption and health: umbrella review

Huang Y, Chen Z, Chen B, Li J, Yuan X, Li J, Wang W, Dai T, Chen H, Wang Y, Wang R, Wang P, Guo J, Dong Q, Liu C, Wei Q, Cao D, Liu L · BMJ · 2023

Umbrella review of existing meta-analyses (RCTs, cohort, case-control and cross-sectional studies)Verified from abstract
Open original source

Consumption of sugar sweetened beverages, artificially sweetened beverages, and fruit juice and incidence of type 2 diabetes: systematic review, meta-analysis, and estimation of population attributable fraction

Imamura F, O'Connor L, Ye Z, Mursu J, Hayashino Y, Bhupathiraju SN, Forouhi NG · BMJ · 2015

Systematic review and meta-analysis of prospective cohort studies with population attributable fraction estimationVerified from abstract
Open original source

Dietary Sugar Intake and Incident Type 2 Diabetes Risk: A Systematic Review and Dose-Response Meta-Analysis of Prospective Cohort Studies

Della Corte KA, Bosler T, McClure C, Buyken AE, LeCheminant JD, Schwingshackl L, Della Corte D · Advances in Nutrition · 2025

Systematic review and dose-response meta-analysis of prospective cohort studiesVerified from abstract
Open original source

Sugar-sweetened beverages and risk of metabolic syndrome and type 2 diabetes: a meta-analysis

Malik VS, Popkin BM, Bray GA, Despres JP, Willett WC, Hu FB · Diabetes Care · 2010

Meta-analysis of prospective cohort studiesVerified from abstract
Open original source

Sugar and artificially sweetened beverages and risk of obesity, type 2 diabetes mellitus, hypertension, and all-cause mortality: a dose-response meta-analysis of prospective cohort studies

Qin P, Li Q, Zhao Y, Chen Q, Sun X, Liu Y, Li H, Wang T, Chen X, Zhou Q, Guo C, Zhang D, Tian G, Liu D, Qie R, Han M, Huang S, Wu X, Li Y, Feng Y, Yang X, Hu F, Hu D, Zhang M · European Journal of Epidemiology · 2020

Dose-response meta-analysis of prospective cohort studiesVerified from abstract
Open original source

Intake of fruit juice and incidence of type 2 diabetes: a systematic review and meta-analysis.

Xi B, Li S, Liu Z, Tian H, Yin X, Huai P, Tang W, Zhou D, Steffen LM · PloS one · 2014

Systematic review and meta-analysis of prospective cohort studiesVerified from full text
Open original source

Role of diet in type 2 diabetes incidence: umbrella review of meta-analyses of prospective observational studies.

Neuenschwander M, Ballon A, Weber KS, Norat T, Aune D, Schwingshackl L, Schlesinger S · BMJ (Clinical research ed.) · 2019

Umbrella review of meta-analyses of prospective observational studiesVerified from abstract
Open original source

Long-term Consumption of Sugar-Sweetened and Artificially Sweetened Beverages and Risk of Mortality in US adults

Malik VS, Li Y, Pan A, De Koning L, Schernhammer E, Willett WC, Hu FB · Circulation · 2019

Prospective cohort study of two cohorts (Nurses' Health Study and Health Professionals Follow-up Study) with repeated dietary measures, analysed separately by age-stratified Cox proportional hazards regression and then pooled by combining the two cohorts' estimates. The paper does not state whether the pooling was fixed- or random-effects, so this record does not either. Verified in this pass against two retrieved documents of the same article: the publisher's abstract as indexed by Europe PMC, and the NIH accepted author manuscript held at PMC6488380 (NIHMS1520746). The publisher's version of record answered HTTP 403 and was not inspected, and Supplemental Tables 1-4 could not be retrieved, so no claim here rests on them. Facts 1 to 6 are quotations from the indexed abstract. Facts 7 to 10 are quotations from the manuscript's results and discussion that are not in the abstract: the per-serving SSB estimate, the intermediate-conditions sensitivity analysis, the sex-interaction sentence with the authors' own caveat that a biological basis is not clear, and the residual-confounding statement. A verifier that compares stored facts against the abstract will report this record because of those four facts; that is expected and is why they are labelled here rather than removed.Verified from full text
Open original source

Community

Discussion

Have you tried this? What helped or made it difficult?

Reader experiences are separate from the scientific evidence. Explore all discussions

No discussions yet.

Be the first to ask a question or share your experience.

Sign in to vote or join the discussion.

Related habits

Explore other habits in a shared category. Each has its own evidence and precautions.

Editorial history and review method

Revision 10 · Approved for publication · Codex (publication approval explicitly authorized by site owner; category migration)

Prepared and checked with AI. Publication approval does not imply independent expert review. Request a correction.

Applied the audit corrections for sugary-drink-reduction. (1) Precaution 1's quantitative claim about sugar-sweetened fruit juice and diabetes had no source on this entry; xi-2014-juice-diabetes already existed in the library but was attached only to fruit-over-juice, so it is now cited from ssb-incident-type-2-diabetes and named in the precaution. (2) The certainty rationale for ssb-incident-type-2-diabetes credited 'an umbrella review' with grading the SSB-type 2 diabetes evidence as high quality, which is not what the cited umbrella review says: huang-2023 grades all cohort-derived evidence as low or very low. The high-quality grading belongs to Neuenschwander 2019, which is now retrieved, stored as neuenschwander-2019-diet-t2d-umbrella, cited from the same outcome, and named with its estimate (one serving/day, HR 1.26, 95% CI 1.11 to 1.43). (3) Completed the stored headline truncation on ssb-incident-type-2-diabetes, which ended mid-word at 'only partly explain…'. (4) Replaced the card summary, which was a one-line endorsement, with a version carrying the magnitude, the certainty, the null water comparison and the low quality of the mortality evidence. (5) Added the other half of the McGlynn comparison to precaution 2: substituting low- and no-calorie sweetened drinks for water was associated with higher glycated haemoglobin but also with lower systolic blood pressure (MD -2.63 mmHg, 95% CI -4.71 to -0.55), so the entry no longer reports only the unfavourable half. Second reviewer must re-check: (a) whether the pervasive food-industry competing interests disclosed in the limitations warrant a reader-facing disclosure, which the audit raised as a judgement call and which was not applied; (b) that no downstream copy implies proven weight benefit from swapping to water; (c) that the Della Corte estimate is quoted for the intended exposure. NOTE ON REVISION HISTORY: the revision-3 record for this habit was written by a separate, concurrent truncated-string pass that had already completed the same ssb-incident-type-2-diabetes headline this pass rewrites. This pass read revision 3, so nothing was lost, but these changeNotes replace that record - a second reviewer should not expect a separate revision-3 note. This pass re-read Malik 2019 against the full manuscript and found the stored facts for it were wrong in four places, which means this outcome's numbers were wrong too: the paper reports 3,415,564 person-years (not 2,923,338), a five-category SSB scheme with pooled HRs 1.06 / 1.14 / 1.21 (not a four-category scheme reaching 1.30), and 1.07 (95% CI 1.05-1.09) per additional serving/day (not 1.16, 95% CI 1.10-1.22). The 1.05 estimate is the intermediate-conditions sensitivity analysis and adds incident hypertension, hypercholesterolaemia, type 2 diabetes, CHD and stroke; it does not include BMI, which the stored text wrongly listed. The p = 0.02 is a sex interaction for total mortality only - the CVD-mortality sex interaction is p = 0.70 and the ASB one is p = 0.01 - and the authors state that a biological basis for it is not clear, so it must not be presented as an established biological sex difference. The malik-2019 source record now stores eight corrected facts with a verification note; Supplemental Tables 1-4 remain unretrieved and nothing in this entry depends on them. Publication release 2026-10-04: site owner authorized a bounded AI audit and release of this exact revision. See docs/release-audit-2026-10-04.md for source-check findings and limits. No independent health-professional or peer review is claimed. Category migration 2026-10-04: removed the separate goals field; categories: Nutrition, Long-term health → Nutrition, Long-term health, Energy. Research assessments, sources, guidance, precautions, and evidence review dates are unchanged. This revision awaits publication approval; no new evidence review is claimed. Publication approval 2026-10-04: site owner explicitly accepted all current category-migration revisions for deployment and requested that remaining limitations be retained in docs/content-review-backlog.md. Approval applies to this exact content revision. No new evidence review, independent expert review, or clinical validation is claimed; existing precautions and disclosures remain in effect.