Sugary Foods and Breast Cancer Risk: What a Korean Study Found

Women who ate certain sugary foods more frequently had a higher rate of breast cancer in a large Korean study. Fruit, despite being the largest source of dietary sugar in the group, did not show the same association.

The findings draw attention to the foods that supply sugar, rather than sweetness alone. They do not prove that eating ice cream or sweetened cereal causes breast cancer.

That distinction matters when interpreting a headline about everyday foods. A study can identify a pattern worth investigating without establishing that a particular ingredient caused the cases observed. Here, the central question is whether the source and nutritional context of sugar deserve more attention alongside the total amount consumed.

What did the researchers study?

Published on October 31, 2026, in the European Journal of Nutrition, the research used data from the Korean Health Examinees-Gem (HEXA-G) cohort. The final analysis included 71,264 women aged 40–69 at enrollment, followed for an average of 9.2 years. Researchers identified 713 breast cancer cases through cancer-registry linkage.

Participants reported their usual diets using a food-frequency questionnaire. Researchers examined seven food groups, including sugary cereals, ice cream, cakes and cookies, sweets, carbonated drinks, sweetened traditional Korean tea, and a combined spreads category. They adjusted their analysis for factors including body mass index, alcohol use, exercise, reproductive history and family history of breast cancer. Read the original study by Shin and colleagues.

The women entered the underlying cohort between 2004 and 2013, and cancer records were followed through December 31, 2018. Although the paper was published in 2026, its findings therefore describe dietary patterns and outcomes recorded years earlier. They are not a direct assessment of every cereal, dessert or eating pattern available today.

The combined spreads category included jam, honey, butter and margarine. This was a questionnaire grouping, not a claim that butter and margarine are themselves high-sugar foods. Keeping that detail visible helps avoid turning a research classification into an inaccurate shopping rule.

Adjustment for other risk factors was important because eating habits occur alongside many other differences between people. However, an adjusted comparison is still not the same as a trial in which researchers assign diets and otherwise keep conditions comparable.

What does the reported 28% increase mean?

Women consuming more than two servings of the combined high-sugar food groups per week had a 28% higher estimated breast cancer hazard than those consuming less than one serving per month. This is a combined intake measure, not two servings of every food listed.

The adjusted hazard ratio was 1.28, with a 95% confidence interval of 1.02–1.61. This describes a relative comparison between groups over follow-up. It does not mean that 28% of women who eat sugary foods will develop breast cancer, or that their probability rises by 28 percentage points. Study results.

The confidence interval also matters. Its width reflects uncertainty around the estimate, and the lower end is only slightly above 1.00, the value representing no difference in hazard. Presenting only the headline percentage can make the result appear more precise than it is.

Across the full analysis, 713 diagnoses among 71,264 women amount to approximately 1% of participants. That is a simple overall proportion, not either food-intake group’s adjusted risk and not a prediction of an individual woman’s future risk. It cannot be used as a baseline to calculate a personal increase from the 28% figure.

Food-specific analyses also reported higher estimates for ice cream and sugary cereals:

FoodComparison reported in the paperRelative estimate
Ice creamMore than once weekly versus no consumption26% higher hazard
Sugary cerealMore than once every two months versus no consumption34% higher hazard

These categories describe comparisons within this study. They should not be turned into universal limits defining how much of a food is safe. The ice-cream estimate was borderline, with a confidence interval of 1.00–1.58; sugary-cereal intake was relatively low in this population. Full findings and discussion.

For example, the cereal finding does not establish that eating a sweetened cereal once every two months triggers a biological change. It identifies a boundary used to compare reported consumption in this particular population. Nor can the ice-cream percentage establish that sugar, saturated fat or another associated habit was responsible for the difference.

Why was fruit different?

Average total sugar intake was about 44 grams per day, with fruit contributing roughly 43%. Neither fruit consumption nor sugar obtained from fruit showed a statistically significant association with breast cancer. Total dietary sugar also showed no significant association.

The authors suggested that fiber, vitamins and other components of fruit might help explain the different pattern. They also discussed possible insulin-related pathways for sugary foods. These are possible explanations, not mechanisms demonstrated by this study. Researchers’ interpretation.

The findings do not establish that fruit prevents breast cancer, or that sugar from fruit can be consumed without limit. They show why foods with different nutritional profiles should not automatically be treated as equivalent.

There is also a difference between failing to detect an association and proving that no association could ever exist. The fruit result is informative within the conditions of this analysis; it should not become a promise about every person, quantity or form of fruit consumption.

How might insulin and food composition be involved?

The researchers discussed insulin resistance and insulin-like growth factor-1, or IGF-1, as possible pathways connecting dietary patterns with breast cancer. IGF-1 participates in cell growth and proliferation. The proposed explanation is that changes in insulin-related signaling might help connect some eating patterns to cancer development.

They also noted that sugary cereals can combine refined grains with added sugar, while ice cream combines added sugar with saturated fat. These features offer possible explanations to investigate. They do not establish that an individual product changes hormone signaling in a way that causes breast cancer.

Similarly, the authors referred to earlier research on sweet-food intake and mammographic breast density. That discussion supplies background for a hypothesis; it does not mean the present cohort demonstrated that sugary foods increased breast density and subsequently caused cancer. Discussion in the original paper.

This distinction between a plausible pathway and a demonstrated pathway is especially important in nutrition reporting. A biological explanation can make an association easier to understand, but it cannot remove the uncertainty inherent in the study design.

What did the age and alcohol analyses show?

Among women younger than 50 at enrollment, the highest combined intake category had a 52% higher estimated hazard than the reference category. An association was also observed among women with a history of alcohol consumption. These findings suggest questions for further research, rather than separate dietary rules for each group.

In particular, finding a statistically significant result in one age group and not in another does not automatically prove the groups respond differently. The researchers’ age-interaction test was not significant. The paper therefore does not establish that sugary foods have a stronger association in younger women or that older women are unaffected.

Age at enrollment should not be treated as a substitute for menopausal status throughout follow-up. A woman could enter the study before menopause and experience it later. The alcohol finding likewise does not prove that sugar and alcohol acted together through a specific hormone pathway. Subgroup analyses and interpretation.

What are the important limitations?

This was an observational study: researchers tracked existing habits rather than assigning diets. Statistical adjustment cannot eliminate every alternative explanation. Diet was assessed once at baseline, and the study did not separately measure added and naturally occurring sugars.

That single diet assessment cannot fully capture changes over nearly a decade. Someone who reported frequent desserts at enrollment might later eat them less often, or the reverse. Self-reported frequency and portion size also introduce uncertainty.

The absence of separate added-sugar measurements matters for the main interpretation. Comparing food groups is useful, but it does not isolate sugar from the other ingredients in those foods or the broader habits of the people eating them. The findings cannot establish which component explains the association.

The researchers also lacked detailed information about cancer subtype and stage. Consequently, the analysis cannot tell readers whether particular biological forms of breast cancer account for the findings. Results from this cohort of Korean women should also be tested in other populations before being treated as universally applicable. Methods and limitations.

How does this fit into breast cancer prevention?

The practical response is to consider overall eating habits without treating an individual dessert as a diagnosis or a cause. Established prevention advice remains broader: the World Cancer Research Fund recommends regular physical activity, avoiding excess weight gain, and limiting alcohol. No lifestyle choice can guarantee prevention. WCRF’s breast cancer prevention guidance.

This research also should not be interpreted as evidence that eliminating sugar treats cancer. The U.S. National Cancer Institute states that stopping sugar intake has not been shown to make cancer shrink or disappear. NCI’s explanation of sugar and cancer.

For readers considering their own habits, the useful question is how frequently sweetened foods become the default part of meals and snacks, rather than whether one serving has caused harm. Reviewing that pattern can sit alongside established health priorities without turning a single study into a restrictive diet.

The fruit findings also give no basis for cutting fruit out solely because it contains sugar. At the same time, this study cannot prescribe an exact number of desserts that will prevent breast cancer. Its contribution is a more focused research question: how much of the observed association relates to sugar sources, the rest of the food, or the dietary pattern around it?

Reporting note: Original English summary and explanation based on the supplied Korean draft and the published research. Sources checked November 6, 2026. The paper’s abstract gives the starting figure of 92,314 women; its methods and results identify 71,264 in the final analysis after exclusions. Research attribution: Shin, W.-K., Lee, H., Huang, D., et al. (2026), European Journal of Nutrition, 65, 291, doi:10.1007/s00394-026-04134-0, licensed under CC BY 4.0. Findings have been paraphrased and contextualized; no original figures are reproduced. Thumbnail: AI-generated editorial illustration.

ViewithU

ViewithU Team