sugar and artificial sweeteners both increase liver cancer

 a recent study found that drinking either artificially sweetened or sugar-sweetened beverages was associated with increased risk of liver cancer: see artificial sweeteners and sugar inc liver cancer JAMA2026 in dropbox, or doi:10.1001/jamanetworkopen.2026.17754

Details:

-- analysis of 11 prospective cohort studies assessed the relationship between the intake of artificially sweetened beverages (ASBs) and sugar-sweetened beverages (SSBs) with the subsequent development of liver cancer in adults

    -- 10 US cohorts and 1 European cohort:

        -- National Institutes of Health-AARP (NIH-AARP)

        -- Agricultural Health Study (AHS)

        -- Cancer Prevention Study-II Nutrition Cohort (CPS-II)

        -- European Prospective Investigation Into Cancer and Nutrition (EPIC)

        -- Health Professionals Follow-Up Study (HPFS)

        -- Nurses’ Health Study (NHS)

        -- Physicians’ Health Study-II (PHS-II)

        -- Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial (PLCO)

        -- Southern Community Cohort Study (SCCS)

        -- Women’s Health Initiative (WHI)

        -- Women’s Health Study (WHS)

-- the researchers excluded participants in these studies who had prevalent cancer (except nonmelanoma skin cancer) prior to enrollment, or those who had implausible total energy intake (<800 or >4200 kcal/d for males and <600 or >3500 kcal/d for females), or those that did not have dietary information on ASB (artificially sweetened beverages) or SSB (sugar-sweetened beverages) consumption

-- participants were further assessed via linkage to state cancer registries or follow-up surveys for incident liver cancer.

    -- analysis included individual-level data among these 11 cohorts that included long-term follow-up for cancer incidence

-- a total of 1,518,411 participants were included, with mean [SD] age of 57.8 [10.1] years; 883,832 females  [58.2%] were in this analysis. 

-- the results were based on the self-reported intake of ASB and SSB assessed at baseline, including intake of cola, soda, caffeine-free soda, and juices, using validated food frequency questionnaires and analyzed per 1-beverage/day increment, between 1980 and 2009 and followed up from 2000 to 2019

--statistical analysis:

    -- In the minimally adjusted models, researchers used the time in the study as the underlying time variable adjusted for age (years, continuous), sex (in cohorts with both sexes), and total energy intake (kilocalories per day, continuous).

    -- In the multivariable analyses, researchers additionally adjusted for factors selected a priori that are known to be associated with liver cancer, including race and ethnicity (White and racial and ethnic minority groups [those identifying as groups included American Indian or Alaska Native, Asian, Black, Hispanic, and Pacific Islander]), smoking status (never, former, or current), body mass index (BMI <25.0, 25.0-29.9, 30.0-34.9, and >35.0), physical activity (cohort-specific quintiles by sex), alcohol intake (0, 0.1-4.9, 5.0-14.9, 15.0-29.9, or >30.0 g/d), regular use of aspirin or nonsteroidal anti-inflammatory drugs (yes or no), diabetes history (yes or no), and coffee intake (non-consumers, <1 cup/d, 1-2 cups/d, or >2 cups/d). Race and ethnicity were based on self-report by choosing options included on each survey.

        -- Race and ethnicity were dichotomized as White and racial and ethnic minority groups due to the large number of individuals identifying as White. in particular, race and ethnicity were used in this study in order to adjust for potential confounding differences in dietary intake and liver cancer risk.

-- main outcomes: Incident liver cancer overall and by their subtypes of hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC), adjusting for potential confounders, including demographics and lifestyle factors, and weighted mean meta-analysis of estimates.

-- the median (IQR) duration of follow-up was 11.4 (10.8-27.9) to 31.4 (27.6-31.5) years, with data analysis between September 2024 and August 2025.

Results:

-- 1,518,411 participants

-- 2,811 incident liver cancer cases were identified, including 1,699 HCC and 444 ICC cases 

Associations Between ASB and SSB Intake and Liver Cancer Risk

-- Minimally adjusted analysis:

    -- ASB intake was positively associated with 12% increased liver cancer risk per 1-beverage/day intake, HR 1.12 (1.08-1.16)

    -- SSB intake was positively associated with 6% increased liver cancer risk per 1-beverage/day, HR 1.06 (1.01-1.11)

-- multivariable adjusted model analysis:

    -- ASB intake was positively associated with an 11% increased risk of liver cancer per 1-beverage/day intake, in analysis that did not adjust for BMI or diabetes, HR 1.11 (1.07-1.15)

    -- SSB intake overall was no longer associated with liver cancer, though there was evidence of significant heterogeneity between studies; when the SCCS (Southern Community Cohort Study) estimate was deleted from the meta-analysis (a common approach in large meta-analyses, as done here, to sequentially removed one of the studies to see if any of them was an outlier that changed the results), finding that there was a positive association between SSB and liver cancer in the remaining studies, with a 9% increased liver cancer risk per 1-beverage/day, HR 1.09 (1.03-1.15); P = .03

-- categories of ASB and SSB consumption, comparing participants with the highest category of intake (>1 drink/day) with those who never or rarely drank ASBs or SSBs:

    -- ASB intake had a 6% nonsignificant increase in liver cancer risk: HR 1.06 (0.94-1.20)

    -- SSB intake had a liver cancer risk was also a nonsignificant 6% increase in liver cancer risk: HR 1.06 (0.90-1.24)

        -- there was significant evidence of heterogeneity between studies from the European EPIC study, since there was no observed heterogeneity when this cohort was removed

-- assessment of the specific liver cancer risks, hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC), after multivariable adjustment (HCC was assessed in 10 cohorts, and ICC was assessed in 6 cohorts):

    -- HCC: ASB intake per 1-beverage/day increase was not associated with liver cancer risk; but SSB was associated with a 10% increased risk per 1 drink/day, with HR 1.10 (1.03-1.18)

    -- ICC: ASB intake per 1-beverage/day was not associated with ICC risk with HR per 1-beverage/day intake having a nonsignificant 4% increased risk, HR 1.04 (0.94-1.14), but a positive 15% increased risk per 1-beverage/day per intake of SSB, HR 1.15 (1.00-1.32)

Subgroup Analyses

-- analyses stratified by diabetes status at recruitment found that associations between ASB and SSB consumption with liver cancer risk remained similar for participants

Commentary:

-- liver cancer is the 3^rd leading cause of cancer mortality globally and is predicted to increase by about 55% by 2040

-- in the US, the American Cancer Society anticipates about 42,240 new cases of liver cancer in 2025 with 30,090 deaths, about 2.1% of all new cancer diagnoses: https://www.cancer.org/content/dam/cancer-org/research/cancer-facts-and-statistics/annual-cancer-facts-and-figures/2025/2025-cancer-facts-and-figures-acs.pdf

-- the predominant liver cancer is hepatocellular carcinoma (HCC), 75% to 85% of liver cancers, though the risk factors for it may be different in different countries: chronic hepatitis B and C, alcohol consumption, smoking, aflatoxin-contaminated foods, MASLD (Metabolic Dysfunction-Associated Steatotic Liver Disease), diabetes, and obesity

    -- 35% of HCC  are not explained by these known risk factors.

-- a 2023 report by the International Agency for Research on Cancer (IARC) used the intake of ASBs as a proxy for aspartame, then classifying aspartame as a group 2B carcinogen (ie possibly carcinogenic to humans), though some studies have not found aspartame to be related to cancers

    -- the ASB used most commonly has been aspartame since 1983

    -- of potential concern is that a Mendelian randomization study did find that aspartame may be a liver carcinogen via the icaspase-1 protein which may promote liver inflammation, a cancer precursor

    -- the current acceptable daily intake of aspartame is 40-50 mg/kg of body weight (eg 2800 mg for a 70kg person)

-- 3 prior studies assessed the relationship between ASBs/SSBs and liver cancer: one large US study found an association between ASBs and SSBs and cancer mortality that included liver cancer (sugar and artificial drinks inc cancers CancerEpidBiomarkers2022 in dropbox, or doi: 10.1158/1055-9965.EPI-22-0392);  another study found the association with liver cancer only in those participants with diabetes; and the third found a non-statistically significant association with liver cancer only among nonsmoking men who consumed 2 or more ASBs/day, with no significant association with SSBs

--Artificially sweetened beverages (ASBs) and sugar-sweetened beverages (SSBs) are widely consumed worldwide and have been linked to metabolic disorders, including obesity and type 2 diabetes, which are established risk factors associated with liver cancer

-- this study was done to examine the associations between ASB and SSB consumption and the risk of incident liver cancer overall and by subtype, including hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC)

-- SSBs have documented associations with bad health outcomes, largely from being associated with weight gain and with more type 2 diabetes globally; and glucose intolerance/diabetes are important liver cancer risk factors; in addition, SSBs are associated with increased risk of MASLD (ie NAFLD of yore) and with more fatty liver accumulation with more sugar intake

    -- of note, not all sugars are created equally: fructose (most abundantly in foods having high-fructose corn syrup, brought to you from Japan chemists in 1957) which has the rather sneaky habits of being insulin-independent, bypassing hepatic glycolysis, promoting hepatic lipogenesis (and adverse lipid effects), and blocking fatty acid oxidation (similar to ethanol). Animal models have found that fructose is associated with bacterial overgrowth and disruption of the gut's tight junctions, thereby increasing gut permeability and endotoxin exposure from gut contents that normally would be excreted

    -- and sucrose, a combination of glucose and fructose in this disaccharide that is table sugar, has a glucose component that increases fructose absorption (fructose is passively absorbed via the glucose transporter 5 (GLUT5), though the glucose is absorbed by passive transport across the small intestine by SGLT1 but also actively and passively in the kidneys' proximal tubules by SGLT1 and SGLT2: https://www.ncbi.nlm.nih.gov/books/NBK576428/

-- the current study found that "In this pooled analysis that combined data from 11 prospective studies assessing the association of ASB and SSB consumption with liver cancer risk, increased SSB beverage intake was associated with increased risk of HCC and ICC. In contrast, after adjustment for confounders, there was no association between ASB intake and total liver cancer, HCC, or ICC risk. Associations remained largely the same when stratifying by diabetes status at recruitment" 

-- "However, in models not adjusted for diabetes or BMI, we observed positive associations between ASB consumption and liver cancer risk, but after adjustment for these confounders was made, outcomes were attenuated and there were no longer associations."

    -- but see the last of the many study limitations below that really challenges these conclusions of this study... 

--prior blogs of interest:

    -- a large French population-based nutrition study (NutriNet-Sante) found that artificial sweeteners (especially aspartame and acesulfame-K) were associated with increased cancer risk: https://gmodestmedblogs.blogspot.com/2022/05/artificial-sweeteners-and-cancer-risk.html

    -- another study found that ultraprocessed foods were associated with an increase of early onset colorectal cancers, where SSBs and ASBs specifically were associated with a 25% increased risk: https://gmodestmedblogs.blogspot.com/2026/01/more-cancers-are-being-diagnosed-at.html

    -- a broadly based European study found that either SSBs plus ASBs were associated with 17% increased risk of all-cause mortality, 52% increase in circulatory disorder deaths by ASBs, 59% increase in deaths from digestive diseases by SSBs, but no significant increase in cancers overall, but a significant 25% increase in colorectal cancer deaths, and a 59% increase in Parkinson's disease mortality: https://gmodestmedblogs.blogspot.com/2019/10/soft-drinks-increase-mortality.html

Limitations:

-- systematic reviews and meta-analyses inherently have significant problems, especially in a report like this one where 11 studies are included: 

    -- the included studies all had different inclusion and exclusion criteria (ie, the different populations of patients with different comorbidities, medications, lifestyles...)

    -- these studies represented different types of populations: some were country-wide, some were limited to certain areas of the country, one was from Europe...

    -- the dietary questionnaires used in the studies varied. the dietary information is all self-reported and may well have inaccuracies (likely having SSB/ASB intake underreported, since many people may well think that these are unhealthy and report lower numbers, similar to underreporting of smoking or alcohol consumption)

    -- some studies assessed cancer mortality vs cancer incidence

    -- some studies evaluated subcohorts of larger studies where there was better cancer incidence information; but evaluating subcohorts can diminish statistical validity

    -- the studies had different follow-up times, ranging from 11.4 to 31.4 years, again potentially leading to different results since the latency to develop liver cancer in some cases may take longer than the 11.4 years to be reported and thereby confounding the results

    -- the current study did not take into account other potential causes of liver cancer. Was the association of increased liver cancer risk found in this study actually from hepatitis B or C, and not from the amount of SSBs/ASBs consumed??

    -- most importantly: the enrollment in the studies was over a large range of times (from 1980 to 2009) and ended over a large range of times (2000-2019) is problematic: this large time range of enrollment and study cessation may represent significantly different times where there were major variations in the types of foods consumed, the types of artificial sweeteners used, and the types/intensities of comorbidities and their therapies that could affect results. for example, were individuals on statins???? [very likely not in the earlier studies, very likely included in the later ones], and statins decrease chronic inflammation (a precursor of cancer) and have been found to decrease breast cancer mortality, for example: https://gmodestmedblogs.blogspot.com/2026/09/statin-use-after-breast-cancer-dec.html)

        -- very important information was not included (and may not have been available from the studies), such as for diabetes which was presented in a binary fashion (yes or no) and not whether the A1c was 6.8% or 13.8% that might have influenced the results, or what meds were used (since there is some concern that some of the older ones may be associated with cancer, including the finding that liver cancer may be increased with first- and second-generation sulfonylureas, as well as insulin, meglitinides, thiazolidinediones, though GLP-1s may reduce several cancers (breast and colon): https://pmc.ncbi.nlm.nih.gov/articles/PMC4046473/.

        -- there may have been significant changes in the effectiveness of detection and treatment for early liver cancer from the earlier to the later studies, which could affect the liver cancer mortality statistics cited.

        -- the information was ascertained only at the onset of the studies, and this information may well have changed over time, especially since many of the studies were long-term; for example, alcohol consumption, a clear hepatotoxin, may well have changed in individuals, perhaps because of changes in their medical conditions or social pressures. so, conceivably during a midrange study above with about 20 years of followup, people may have changed their alcohol intake dramatically after about 1 or 2 years and alcohol may not have been a significant liver cancer risk though they would have been stated to have potentially very considerable alcohol intake. or they may have other comorbidities, or be on different medications.....

        -- they state that "There was no evidence of effect modification by diabetes status". but, as noted, there are differences in outcomes by quality of diabetes control, meds used, diabetes comorbidities: diabetes is not just "present" or "absent". and importantly "prediabetes", a documented liver cancer risk factor, was not included in this study, and can affect liver cancer mortality ("prediabetes" is associated with hepatic steatosis, a liver cancer precursor in 35.1%-47.8% of cases: https://pubmed.ncbi.nlm.nih.gov/32801011/)

            -- and given the widespread differences in the time periods of these studies, there are likely very different degrees of prediabetes and diabetes in the later periods, since the overall incidence of diabetes has increased so much by the time of the later studies, and diabetes is such an important risk factor for liver cancer

            -- also, it is a bit strange to control for diabetes and BMI in their multivariable adjustment model: ASBs/SSBs are associated with the development of diabetes/increased BMI. so there is an obvious link that these ASB/SSB fortified drinks lead to diabetes/increased BMI which leads directly to increased liver cancer risk. so, this association is obscured by controlling for diabetes/BMI. And they actually showed that by finding that their analysis that did not adjust for BMI or diabetes made their results change from nonsignficant to significant. so, their overall conclusions are in fact contradicted by this analysis and there was a real, statistically valid conclusion that ASB/SSB are bad actors associated with liver cancer!!!

-- so, the aggregation of these diverse studies into a single outcome is fraught with potential unreconciled biases!!

    -- the SSB and ASB assessments were limited to beverage consumption, which does not include the broad range of sugar- and artificially-sweetened foods (and the frequent use of high fructose sweeteners in lots of foods as noted above; this is important given that fructose is metabolized differently in the body than the other sweeteners, along with very different adverse effects on the body vs other sweeteners)

    -- consumption of these sweeteners was by self-report, which could be swayed/minimized by many external factors, such as public perception of appropriate levels of consumption that might well lead to the minimization of self-reported ASB/SSB consumption), and the level of reporting accuracy may have changed over the time course of the included studies

    -- see the above limitations section for the many problems with the results of this study

-- that all being said, it is clear from other studies (as well as an appropriate interpretation of this study) that sugar- and artificially sweetened beverages (and other sweetened consumables) are in fact associated with many bad medical conditions, including liver cancer and others leading to premature deaths

    -- we as clinicians/nurses/dieticians/others who interact medically with patients should routinely question about the consumption of ASBs/SSBs, specifically from beverages (the most common consumables that are sweetened) as well as other "no sugar" or "diet" consumables with ASBs that are still associated with weight gain, diabetes, etc.

    -- the advertising for these consumables is flagrantly misleading ("diet sodas" that actually lead to increased weight and diabetes), and many individuals are unaware of their potentially severe adverse medical conditions...

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

to get access to the dropbox: https://www.dropbox.com/scl/fo/pcf4vymtpo77f49wd2ljr/ANyJqB40CuBd-DMEr67XDoI?rlkey=5dsalmuhf8ekd0svsgvn93xm1&dl=0

or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them