— June 10, 2026 · Sugary Drinks and Liver Cancer: 11 Cohorts, 1.5 Million Adults

Table of Contents

1. At a Glance
2. What Was Done
3. What Was Found, Number by Number
4. The Diet-Drink Result and Aspartame
5. How Sugar Reaches the Liver
6. How It Fits With Earlier Research
7. What It Cannot Show
8. What the Authors and the Commentary Said
9. What Remains Unknown
10. Sources and Primary Documents
11. Key Research Papers
12. Connections
13. Featured Videos

At a Glance

• Paper — Watling CZ, McGlynn KA and 23 co-authors, “Artificially Sweetened and Sugar-Sweetened Beverage Intake and Risk of Liver Cancer,” JAMA Network Open 9(6):e2617754, published online June 10, 2026.
• Who — led by the US National Cancer Institute, with researchers from Yale, Harvard-affiliated hospitals and schools, the American Cancer Society, Boston University, Vanderbilt, the Medical University of Vienna and the International Agency for Research on Cancer (IARC) of the World Health Organization.
• Design — pooled analysis of individual-level data from 11 prospective cohort studies, analysed with the same methods and the same adjustments in each cohort, then combined by meta-analysis.
• Size — 1,518,411 adults (58.2% women, mean age 57.8); median follow-up 17.8 years; 2,811 liver cancers, of which 1,699 were hepatocellular carcinoma (HCC) and 444 intrahepatic cholangiocarcinoma (ICC).
• Main result — per extra sugar-sweetened drink a day: HCC hazard ratio 1.10 (95% CI 1.03–1.18); ICC hazard ratio 1.15 (95% CI 1.00–1.32). No association with liver cancer overall in the main analysis.
• Diet drinks — no association with liver cancer, HCC or ICC after adjustment for body mass index and diabetes.
• Kind of evidence — observational; an association, not a demonstrated cause.

Back to Table of Contents

What Was Done

The 11 cohorts are some of the best-known long-term health studies in the United States and Europe: the NIH-AARP Diet and Health Study, the Agricultural Health Study, the Cancer Prevention Study-II Nutrition Cohort, the European Prospective Investigation into Cancer and Nutrition (EPIC), the Health Professionals Follow-Up Study, the Nurses’ Health Study, the Physicians’ Health Study II, the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial (PLCO), the Southern Community Cohort Study, the Women’s Health Initiative and the Women’s Health Study. Participants enrolled between 1980 and 2009, and follow-up ran to end dates between 2000 and 2019. Four of these cohorts had published on sweetened drinks and liver cancer before; two of them were updated with extra years of follow-up, and the others were re-analysed so that every cohort was handled the same way.

At enrolment, participants filled in a validated food frequency questionnaire. From it the researchers totalled two kinds of drink:

• Artificially sweetened beverages (abbreviated ASB in the paper) — diet cola or soda and other low-calorie sweetened drinks.
• Sugar-sweetened beverages (SSB) — cola, soda, caffeine-free soda and juices, added together. The paper describes the category as including “soda, sugary beverages, and juice.” It does not report a separate result for juice.

One drink was defined as a standard US can, about 355 mL (12 fl oz). Intake was analysed in five bands, from “never or rarely” to more than one a day, and as a continuous measure per extra drink a day. Most cohorts asked about drinks only once; one cohort asked again about four years later, and two asked every four years and used the running average.

Liver cancers were identified through cancer registries or medical-record review and, where the records allowed, split into HCC and ICC. ICC was analysed only in cohorts with more than 10 cases, which left 6 cohorts for that subtype. In each cohort the same factors were adjusted for: age, sex, race and ethnicity, body mass index (BMI), physical activity, smoking, alcohol, regular aspirin or anti-inflammatory drug use, diabetes, coffee intake and total calorie intake. The team also ran the models without BMI and diabetes, because weight gain and diabetes could themselves lie on the path between drinks and cancer.

To check that no published cohort had been missed, the authors also ran a systematic review (registered in PROSPERO as CRD42024518773) through February 12, 2026. It found 4 earlier studies covering 5 cohorts, all already inside the pooled data.

Back to Table of Contents

What Was Found, Number by Number

All hazard ratios below are per one extra drink a day, after full adjustment, unless stated otherwise. A hazard ratio of 1.10 means a 10% higher rate of the cancer in question, relative to people drinking one fewer a day; it is a relative figure, not an individual’s absolute chance.

• Sugar-sweetened drinks and HCC (10 cohorts): hazard ratio 1.10 (95% CI 1.03–1.18).
• Sugar-sweetened drinks and ICC (6 cohorts): hazard ratio 1.15 (95% CI 1.00–1.32). The lower end of the interval sits at 1.00, the point of no difference, so this estimate is the less certain of the two.
• Sugar-sweetened drinks and all liver cancer: not associated in the main analysis. The cohorts disagreed strongly with one another here (a heterogeneity measure, I², of 68.2%).
• The cohort that pulled the overall result toward zero: the Southern Community Cohort Study showed an inverse result. It had the highest rate of diabetes at enrolment (21.4%, nearly twice that of any other cohort) and the highest sugar-sweetened intake (a mean of 1.00 a day). With that cohort removed, sugar-sweetened drinks were associated with all liver cancer: hazard ratio 1.09 (95% CI 1.03–1.15). This cohort did not supply subtype data, so it did not affect the HCC and ICC estimates.
• No difference between subtypes: statistically, the HCC and ICC estimates did not differ from each other.
• Shape of the relationship: the authors found no evidence that the association bent away from a straight line (P > .10). Comparing the top intake band (more than one a day) with “never or rarely” showed no significant association for either kind of drink, which the authors suggest may reflect limited statistical power in the bands.
• Diabetes: results were similar in people with and without diabetes at enrolment.
• Before full adjustment: adjusting only for age, sex and calories, sugar-sweetened drinks had a hazard ratio of 1.06 (95% CI 1.01–1.11) for all liver cancer, and diet drinks 1.12 (1.08–1.16).

The paper itself puts the scale of the problem in context: liver cancer is the third leading cause of cancer death in the world, more than 42,000 new US cases and about 30,000 US deaths were expected in 2025 (an American Cancer Society figure), HCC makes up 75% to 85% of liver cancers, and an estimated 35% of HCC cases are not explained by known risk factors.

Back to Table of Contents

The Diet-Drink Result and Aspartame

Part of the reason for the study was aspartame. In July 2023 an expert working group convened by IARC classified aspartame as Group 2B, “possibly carcinogenic to humans,” citing limited evidence linking artificially sweetened drinks with HCC. Diet-drink intake was used as a stand-in for aspartame intake. According to the new paper, that conclusion rested on three studies covering four cohorts, which disagreed: one found a positive association, one found it only in people with diabetes, and one, a mortality study, found a non-significant increase only among non-smoking men drinking two or more a day.

In the pooled data, diet drinks did show an association before adjustment (hazard ratio 1.12 with only age, sex and calories accounted for; 1.11 without BMI and diabetes). Once BMI and diabetes were added, it disappeared for all liver cancer, for HCC (10 cohorts) and for ICC (6 cohorts; hazard ratio 1.04, 95% CI 0.94–1.14). The authors point out that, in most cohorts, people who drank one or more diet drinks a day were more likely to have diabetes and a higher BMI to begin with. That pattern would produce an apparent link even if the drinks themselves played no part. Results for diet drinks also varied a lot from cohort to cohort (I² 61.0%), mostly because of EPIC, the European cohort, where diet drinks were uncommon; without EPIC the variation was small (I² 12.2%).

The authors conclude: “The overall evidence for an association in humans between ASB and liver cancer, and more specifically aspartame and HCC, is not strongly supported by these analyses.” They also write that whether diet drinks raise liver cancer risk independent of BMI and diabetes “remains unclear.” The paper gives the acceptable daily intake of aspartame as 40 to 50 mg per kilogram of body weight (about 2,800 mg for a 70-kg adult), and a typical diet drink as containing about 130 mg. This study measured drinks, not aspartame itself, and it did not test other sweeteners separately.

Back to Table of Contents

How Sugar Reaches the Liver

The study did not test a mechanism, but its discussion sets out the leading explanations, each from earlier research.

• The indirect route: weight and blood sugar. Sugar-sweetened drinks have been linked to weight gain and type 2 diabetes, and obesity and diabetes are established risk factors for liver cancer. High blood glucose, high insulin and higher total sugar intake have each been associated with HCC.
• The direct route: fructose. The sugar in these drinks, whether table sugar or high-fructose syrup, is made up of glucose and fructose. Unlike glucose, which is used throughout the body, fructose is handled mainly by the liver, without needing insulin and bypassing the first steps of ordinary sugar breakdown. In the paper’s words, fructose “has been shown to promote hepatic lipogenesis and block fatty acid oxidation, similar to ethanol”: the liver turns part of it into new fat (de novo lipogenesis) while burning less of the fat it already holds.
• Fat in the liver. That fat build-up is the defining feature of metabolic dysfunction–associated steatotic liver disease (MASLD), the new name for non-alcoholic fatty liver disease (see the site’s April 13, 2026 News article on the renaming). The paper lists MASLD alongside diabetes and obesity as an established risk factor for liver cancer.
• The gut. Animal studies suggest fructose can encourage bacterial overgrowth and loosen the tight junctions of the gut lining, so that more bacterial endotoxin reaches the liver through the portal vein.

Controlled feeding studies give a sense of how quickly the liver responds. In a 10-week trial, overweight and obese adults drinking fructose-sweetened drinks (25% of their energy) showed a specific rise in fat-making in the liver and in visceral fat, which was not seen with glucose-sweetened drinks. In a 6-month randomised trial of 47 overweight adults, those drinking 1 litre a day of regular cola had liver fat rise 132% to 143% more (relative change) than those given the same volume of milk, an aspartame-sweetened diet cola or water. And in 41 children with obesity and high sugar intake, swapping sugar for starch at the same calories for just 9 days cut median liver fat from 7.2% to 3.8% and fat-making in the liver from 68% to 26% (area under the curve). These trials measured liver fat, not cancer.

Back to Table of Contents

How It Fits With Earlier Research

Several of the cohorts in the pooled analysis had already reported on their own, with mixed results. The new paper brings them together under one method.

• Women’s Health Initiative (JAMA, 2023). In 98,786 postmenopausal women followed for a median 20.9 years, those drinking one or more sugar-sweetened drinks a day, compared with three or fewer a month, had a liver cancer rate of 18.0 versus 10.3 per 100,000 person-years (adjusted hazard ratio 1.85, 95% CI 1.16–2.96) and a death rate from chronic liver disease of 17.7 versus 7.1 (hazard ratio 1.68, 1.03–2.75). Diet drinks were not significantly associated with either. In that paper, sugar-sweetened drinks meant regular soft drinks and fruit drinks, not 100% fruit juice. The pooled study re-analysed this cohort with its own harmonised definitions.
• EPIC (European Journal of Nutrition, 2016). In 477,206 Europeans, combined soft drinks at more than six servings a week were associated with HCC (hazard ratio 1.83, 1.11–3.02). Here the signal came from artificially sweetened drinks (1.06 per serving, 1.03–1.09), not sugar-sweetened ones (1.00, 0.95–1.06). Juice was not associated with HCC, and no associations were seen for bile-duct cancers. This is the cohort that drove the cohort-to-cohort variation for diet drinks in 2026.
• NIH-AARP and PLCO pooled (Cancer Epidemiology, 2022). Among people without diabetes, sugar-sweetened soda was associated with liver cancer in the first part of follow-up (hazard ratio 1.18, 1.03–1.35); among people with diabetes, artificially sweetened drinks were (1.13, 1.02–1.25). The 2026 analysis, with more cohorts and more years, found no such difference by diabetes status.
• Cancer Prevention Study-II (2022). In 934,777 adults followed for cancer death, sugar-sweetened drinks were associated with deaths from colorectal and kidney cancer, and diet drinks with pancreatic cancer deaths.
• Fatty liver (Framingham Heart Study, 2022). In the older Offspring cohort, frequent sugar-sweetened drink consumers had 2.53 times the odds of developing fatty liver over about six years (95% CI 1.36–4.7); no consistent association was seen in the younger generation.

The 2026 result therefore agrees with the earlier American findings for sugar-sweetened drinks and HCC, and it does not reproduce the European diet-drink signal once all cohorts are handled together. It also adds ICC, the bile-duct cancer inside the liver, which few earlier studies could examine.

Back to Table of Contents

What It Cannot Show

The authors list these limits themselves:

• Self-report. Drink intake was self-reported and is “inevitably prone to measurement error.”
• One snapshot. Most cohorts measured drinks once, at the start. Habits change over decades, which the authors expect would generally pull results toward showing no effect. Diabetes, too, was usually recorded only once.
• No data on existing liver disease. The analysis could not adjust for, or split results by, underlying liver disease such as MASLD or cirrhosis.
• Hepatitis. Information on hepatitis B and C infection, two major causes of liver cancer worldwide, was incomplete. In the one cohort with blood tests, drink intake did not differ by infection status.
• Who was studied. The cohorts were mainly White adults in the United States and Europe, where chronic hepatitis is relatively uncommon. The authors write that the findings may apply less well where hepatitis is the main driver of liver cancer.
• Confounding. As in all observational studies, results “may be influenced by residual and unmeasured confounding”: people who drink more sugary drinks may differ in other ways that the adjustments did not fully capture.
• Disagreement between cohorts. Despite identical methods, the cohorts varied, especially for all-liver-cancer results.

Two further points follow from the design. Because the study adjusted for BMI and diabetes, the hazard ratios describe an association beyond those factors, and the authors write that “it remains unclear whether increased consumption of SSB contributes to increased risk of HCC independent of obesity and diabetes.” And because sugar-sweetened drinks were counted together with juices, the paper cannot say whether soda and juice carry the same association.

Back to Table of Contents

What the Authors and the Commentary Said

The authors write: “To our knowledge, this is the largest study to date to assess the association of ASB and SSB intake with liver cancer risk.” Their summary: the study “found that SSB consumption was associated with increased risk of liver cancer subtypes, while the evidence did not support an independent association between ASB intake and liver cancer risk.” They regard the subtype estimates, which the outlying cohort did not affect, as the ones that “likely represent the most robust findings of this analysis.” Their closing sentence states their own reading: “Based on this study and prior research associating SSB with adverse outcomes, individuals may benefit from reducing their intake of SSB.”

The journal published an invited commentary alongside the paper, by D. Clayton-Chubb and A. T. Chan, subtitled “A Hard Look at Soft Drinks.” It states: “While this analysis was well-conducted, observational studies cannot prove causality.” It names two limits: “the data on SSB and ASB consumption are mostly taken from a single time,” and, although the association persisted after adjustment for obesity and type 2 diabetes, “there may be residual confounding or mediation due to metabolic dysregulation given that these disorders incompletely capture metabolic risk.” It ends by calling for “further mechanistic work to identify causal pathways.”

Press coverage followed within days: HealthDay on June 17 and 2 Minute Medicine on June 19, the latter reporting the 10% and 15% figures and the null result for diet drinks in line with the paper. Some headlines described sugary drinks as boosting or raising liver cancer risk, wording that reads as cause; the paper and the commentary both describe an association. The paper also does not identify a “safe” or “unsafe” amount: it found no evidence that the association departed from a straight line, and no significant result when the top intake band was compared with the lowest.

On funding, the work was supported mainly by the US National Institutes of Health intramural research program, with a Canadian government research fellowship, a university professorship and an NIH grant to one of the cohorts; the paper states that the funders had no role in the design, analysis or decision to publish. Two authors reported fees or equity from drug or biotechnology companies for work unrelated to this study.

Back to Table of Contents

What Remains Unknown

• Cause or marker? Whether sugar-sweetened drinks themselves raise liver cancer risk, or mark a wider pattern of diet, weight and metabolism, is not settled by this or any earlier cohort study. The commentary points to mechanistic research as the next step; the feeding trials described above measured liver fat, not cancer.
• Juice versus soda. The pooled category combined both; the separate role of fruit juice, as distinct from whole fruit, was not reported.
• Through the fat in the liver? Without data on MASLD or cirrhosis, the study could not say how much of the association runs through fatty liver disease.
• Other populations. How the findings apply in regions where hepatitis B or C infection is the main driver of liver cancer, a gap the authors name themselves, is unknown.
• Individual sweeteners. The study measured diet drinks as a group and did not separate aspartame from other sweeteners or examine long-term changes in intake.
• Bile-duct cancer. With 444 ICC cases from 6 cohorts and a confidence interval that touches no effect, the ICC finding needs confirmation in more data.

Back to Table of Contents

Sources and Primary Documents

Every figure attributed to the 2026 study above was taken from the paper’s open-access full text on PubMed Central, not from press coverage. The publication date of June 10, 2026 is printed in the paper’s article information and recorded by Crossref; PubMed lists the issue date as June 1.

1. Watling CZ, Zhao L, Zhang X, Deubler E, Gonzalez-Feliciano AG, Graubard BI, et al. (2026). Artificially Sweetened and Sugar-Sweetened Beverage Intake and Risk of Liver Cancer. JAMA Network Open 9(6):e2617754. — doi:10.1001/jamanetworkopen.2026.17754 · PubMed PMID: 42268607
2. Clayton-Chubb D, Chan AT (2026). Sugar-Sweetened and Artificially Sweetened Beverages and Liver Cancer—A Hard Look at Soft Drinks. JAMA Network Open 9(6):e2617763. — doi:10.1001/jamanetworkopen.2026.17763 · PubMed PMID: 42268615

Press coverage: HealthDay, June 17, 2026 (Sugar-Sweetened Beverages Tied to Liver Cancer Risk); 2 Minute Medicine, June 19, 2026 (High intake of sugar-sweetened beverages is associated with increased risk of hepatocellular carcinoma).

Back to Table of Contents

Key Research Papers

Background papers on sweetened drinks, the liver and liver cancer, in the order they appear above. Each entry was checked against PubMed and Crossref before it was written down.

1. Riboli E, Beland FA, Lachenmeier DW, Marques MM, Phillips DH, Schernhammer E, et al. (2023). Carcinogenicity of aspartame, methyleugenol, and isoeugenol. Lancet Oncology 24(8):848-850. — PubMed PMID: 37454664
2. Stanhope KL, Schwarz JM, Keim NL, Griffen SC, Bremer AA, Graham JL, et al. (2009). Consuming fructose-sweetened, not glucose-sweetened, beverages increases visceral adiposity and lipids and decreases insulin sensitivity in overweight/obese humans. Journal of Clinical Investigation 119(5):1322-1334. — PubMed PMID: 19381015
3. Maersk M, Belza A, Stødkilde-Jørgensen H, Ringgaard S, Chabanova E, Thomsen H, et al. (2012). Sucrose-sweetened beverages increase fat storage in the liver, muscle, and visceral fat depot: a 6-mo randomized intervention study. American Journal of Clinical Nutrition 95(2):283-289. — PubMed PMID: 22205311
4. Schwarz JM, Noworolski SM, Erkin-Cakmak A, Korn NJ, Wen MJ, Tai VW, et al. (2017). Effects of Dietary Fructose Restriction on Liver Fat, De Novo Lipogenesis, and Insulin Kinetics in Children With Obesity. Gastroenterology 153(3):743-752. — PubMed PMID: 28579536
5. Jensen T, Abdelmalek MF, Sullivan S, Nadeau KJ, Green M, Roncal C, et al. (2018). Fructose and sugar: A major mediator of non-alcoholic fatty liver disease. Journal of Hepatology 68(5):1063-1075. — PubMed PMID: 29408694
6. Zhao L, Zhang X, Coday M, Garcia DO, Li X, Mossavar-Rahmani Y, et al. (2023). Sugar-Sweetened and Artificially Sweetened Beverages and Risk of Liver Cancer and Chronic Liver Disease Mortality. JAMA 330(6):537-546. — PubMed PMID: 37552302
7. Stepien M, Duarte-Salles T, Fedirko V, Trichopoulou A, Lagiou P, Bamia C, et al. (2016). Consumption of soft drinks and juices and risk of liver and biliary tract cancers in a European cohort. European Journal of Nutrition 55(1):7-20. — PubMed PMID: 25528243
8. Jones GS, Graubard BI, Ramirez Y, Liao LM, Huang WY, Alvarez CS, et al. (2022). Sweetened beverage consumption and risk of liver cancer by diabetes status: A pooled analysis. Cancer Epidemiology 79:102201. — PubMed PMID: 35728406
9. McCullough ML, Hodge RA, Campbell PT, Guinter MA, Patel AV (2022). Sugar- and Artificially-Sweetened Beverages and Cancer Mortality in a Large U.S. Prospective Cohort. Cancer Epidemiology, Biomarkers & Prevention 31(10):1907-1918. — PubMed PMID: 36107009
10. Park WY, Yiannakou I, Petersen JM, Hoffmann U, Ma J, Long MT (2022). Sugar-Sweetened Beverage, Diet Soda, and Nonalcoholic Fatty Liver Disease Over 6 Years: The Framingham Heart Study. Clinical Gastroenterology and Hepatology 20(11):2524-2532.e2. — PubMed PMID: 34752964

PubMed Topic Searches

1. PubMed: Sugar-sweetened beverages and hepatocellular carcinoma
2. PubMed: Artificially sweetened beverages and liver cancer
3. PubMed: Fructose and de novo lipogenesis in the liver
4. PubMed: Sugar-sweetened beverages and fatty liver disease
5. PubMed: Diet and intrahepatic cholangiocarcinoma risk

Back to Table of Contents

Connections

• News Archive
• Liver Cancer — types, risk factors and research
• Hepatocellular Carcinoma — the most common liver cancer, the subtype in the main finding
• Cholangiocarcinoma — bile-duct cancer, including the type inside the liver
• MASLD (Fatty Liver Disease) — the liver condition linked to sugar and to liver cancer
• MASLD: A New Name for Fatty Liver Disease — the April 2026 News article on the renaming
• NAFLD / MASLD and Insulin Resistance — fatty liver as the liver side of insulin resistance
• Added Sugar, Sugary Drinks, and Weight — the research on sugar, fructose and weight gain
• Artificial Sweeteners — the sweeteners used in diet drinks
• Aspartame — the sweetener behind the 2023 IARC classification
• Cirrhosis — liver scarring, a condition the study could not adjust for
• Liver Disease — an overview of conditions of the liver
• Coffee and Liver Disease — coffee, one of the factors the study adjusted for
• Type 2 Diabetes — a liver cancer risk factor tied to sugary drinks
• Obesity — another established liver cancer risk factor

Back to Table of Contents

Featured Videos