Ultra-Processed Food Consumption and Prostate Cancer Risk: An NHANES Analysis of 17,024 Men

KEY FINDINGS

• Scope: 17,024 men aged 18 and over who participated in NHANES between 2003 and 2023 were assessed in a nationally representative US sample.
• Definition of exposure: Ultra-processed food (UPF) intake was calculated from two 24-hour dietary recalls; participants were divided into quartiles according to the percentage of total energy intake under the NOVA classification (lowest quartile <20%; highest quartile >44%).
• Main result: With the lowest consumption quartile as reference, quartiles 2-4 combined were associated with a 29% higher risk of prostate cancer (RR 1.29; 90% CI 1.03-1.62; P=0.031).
• Dose-response trend: The risk increase was 30% for quartiles 3-4 combined (RR 1.30; 90% CI 1.03-1.65; P=0.035) and 31% for quartile 4 alone - although this comparison did not reach significance (RR 1.31; 90% CI 0.96-1.78; P=0.079).
• After adjustment: On adjustment for age, smoking, race/ethnicity and poverty status, the risk increases were 27% (P=0.045), 31% (P=0.037) and 34% (P=0.072) respectively.
• Critical limitation: The study is cross-sectional and observational; the prostate cancer diagnosis was based on patient report - no causal inference can be drawn.

Background: Why Ultra-Processed Foods?

Ultra-processed foods are products manufactured using refined oils, sugars, starches, salt and numerous industrial additives; they are optimised for taste, texture and shelf life but have low nutrient density. Carbonated drinks, packaged breakfast cereals, ready-to-eat snacks and processed meat products are typical examples of this category. The NOVA classification is a framework that groups foods according to the degree of industrial processing they have undergone rather than the nutrients they contain.

Over the past decade, consistent associations have been reported between ultra-processed food consumption and the incidence of obesity, type 2 diabetes, cardiovascular disease and various malignancies. In prostate cancer specifically, however, the data have to date remained limited and partly contradictory. The present analysis, conducted in a nationally representative sample of the US population, therefore represents a meaningful contribution to a gap in the literature.

Study Design

The study, conducted by researchers at Florida Atlantic University and published in The American Journal of Medicine, used NHANES data from the period 2003-2023.

Key Results

Note on statistical interpretation: Although the point estimate is highest in the comparison confined to quartile 4 (31-34%), the confidence interval includes the value 1.00 and the threshold of significance is not reached. The most likely explanation is the reduction in case numbers when narrowed to a single quartile and the consequent loss of statistical power. In addition, the use of 90% rather than 95% confidence intervals in the study is a methodological choice that should be taken into account when interpreting the results.

Senior author Professor Charles H. Hennekens emphasises that the findings add to the accumulating evidence on the adverse health effects of ultra-processed foods, and that large-scale observational studies and randomized research are needed for the hypothesis to be adequately tested.

Methodological Limitations

• Cross-sectional and observational design: The temporal sequence between exposure and outcome cannot be established with certainty; the possibility of dietary change after diagnosis (reverse causality) cannot be excluded.
• Dietary data are self-reported: Two 24-hour recalls reflect only a limited snapshot of long-term habits.
• The diagnosis is based on patient report: Information on pathological confirmation, stage, grade (ISUP grade) and clinical significance is not available.
• Unmeasured confounders: Family history, frequency of PSA screening, physical inactivity, body mass index and total calorie intake were not included in the model. Screening intensity in particular is a critical candidate confounder.
• The effect size is modest: The relative risk increase is in the range of 27-34%; in observational epidemiology, associations of this magnitude are regarded as sensitive to residual confounding.

Implications for Clinical Practice

1. These data do not change a screening or treatment decision

The present analysis is not of a nature to change PSA screening indications, biopsy thresholds or treatment algorithms. Age, family history, ethnicity and germline genetic predisposition remain the primary determinants in prostate cancer risk assessment.

2. But it strengthens preventive counselling

The urologist is one of the physicians who has contact with a patient more than once during the year. BPH follow-up, active surveillance visits and PSA checks are natural windows of opportunity for lifestyle counselling. The recommendation to reduce ultra-processed food consumption already has strong evidence behind it in terms of cardiovascular risk, metabolic syndrome and general oncological risk.

3. Particularly meaningful in active surveillance patients

Patients followed on active surveillance for low-risk prostate cancer have a desire to take an active role in managing their disease. A constructive and actionable lifestyle recommendation that can be placed alongside the message "no treatment is needed for now" strengthens the patient's sense of control and engagement with the process.

4. Inequality of access must not be overlooked

Reducing ultra-processed food consumption is not merely a matter of individual choice; access to healthy food and its cost are significant structural obstacles. When counselling, it is essential to use non-judgemental language that takes account of the patient's socioeconomic reality.

A practical framework for recommendations: Offering the patient gradual and measurable goals is more effective than an absolute instruction such as "give up processed foods": replacing sugary drinks with water, limiting processed meat consumption on a weekly basis, increasing the proportion of meals prepared at home. Small and sustainable changes achieve higher adherence than short-lived radical restrictions.

Conclusion

This analysis of 17,024 men, nationally representative of the US population, shows that prostate cancer risk is approximately 27-31% higher in men with high ultra-processed food consumption. Because of the cross-sectional design, the self-reported outcome and unmeasured confounders, the findings do not establish a causal relationship and are not on their own capable of changing clinical decisions.

Nevertheless, the signal is biologically plausible, consistent with findings reported for other malignancies in the literature, and points to a recommendation that is harmless to adapt into clinical practice. To establish the role of diet in prostate cancer prevention, prospective cohorts using pathologically confirmed outcomes and, where possible, randomized dietary intervention studies are needed.

References

• Scott H, Willett Y, Dunn J, Dye TD, Hennekens CH. Consumption of Ultra-Processed Foods and Increased Risks of Prostate Cancer in a Random Sample of United States Adults. Am J Med. 2026. doi:10.1016/j.amjmed.2026.07.043
• Clarke H. Higher intake of ultra-processed food associated with increased prostate cancer risk. Urology Times, 12 August 2026.
• Florida Atlantic University. Ultra-Processed Foods Linked to Notably Higher Prostate Cancer Risks. Press release, 12 August 2026.
• EurekAlert! Study links ultra-processed foods to significantly higher prostate cancer risks. 12 August 2026.
• News-Medical.net. Ultra-processed foods linked to higher prostate cancer risk in men. 12 August 2026.