Cancer remains a leading cause of global morbidity and mortality. The World Cancer Research Fund has identified various nutritional factors as potentially causally associated with cancer risk. The Mediterranean diet (MedDiet) and the Mediterranean-DASH Intervention for Neurodegenerative Delay diet (MIND diet) are dietary patterns associated with various health benefits, including potential cancer prevention. MedDiet, characterized by high consumption of fruits, vegetables, fish, and moderate red wine, has shown mixed results regarding cancer risk reduction in previous studies. MIND diet, a modified MedDiet focusing on plant-based foods, has been primarily studied in relation to neurodegenerative diseases, with limited research on its impact on cancer risk. Previous research on the association between dietary adherence and cancer risk has several limitations: inconclusive findings for MedDiet's association with cancer, limited prospective cohort studies on MIND diet's association with cancer, geographically limited studies hindering generalizability, and a lack of understanding regarding the underlying mechanisms. This study utilizes the UK Biobank dataset to overcome these limitations by investigating the prospective associations between MedDiet and MIND diet adherence and overall and 22 specific cancer risks, and further elucidates the mediating roles of metabolites in this association. The study population includes a large, diverse sample, allowing for stronger generalizability. The use of metabolomics allows for the investigation of the biological mechanisms linking diet and cancer risk.

Existing literature on the relationship between Mediterranean diet (MedDiet) and cancer risk presents mixed results. Some prospective cohort studies found no significant association between MedDiet and overall cancer risk, while a recent meta-analysis of numerous studies suggested a reduction in overall cancer risk with greater adherence to MedDiet. Regarding the MIND diet, research on its association with cancer is limited, with few studies, mostly case-control, suggesting inverse associations with specific cancers like breast cancer and glioma. The lack of large-scale prospective cohort studies on MIND diet and cancer limits the generalizability of existing findings. Furthermore, most studies have been geographically limited and haven't fully explored the underlying biological mechanisms through which these diets might affect cancer risk. The potential mediating role of metabolites, which are altered in cancer cells due to changes in their metabolic environment, requires further investigation.

This prospective cohort study used data from the UK Biobank, a large-scale biomedical database. The study included 187,485 participants after exclusions for pre-existing cancer, loss to follow-up, missing dietary data, or missing covariate information. Dietary adherence was measured using the Mediterranean Diet Adherence Screener (MEDAS) and MIND Diet scores. Cox regression models were used to assess the association between diet scores (treated as both continuous and categorical variables) and overall and 22 specific cancer risks. Three models were used, adjusting for different sets of covariates, including demographic and behavioral risk factors. A sequential analysis combining Cox regression, elastic net model (ENM), and gradient boost model (GBM) identified metabolites associated with overall cancer risk from an initial set of 168 NMR metabolites in a separate subset of 202,303 participants. Mediation analyses using the 'mrm' package in R investigated the mediating effects of these metabolites on the association between diet scores and overall cancer risk. Sensitivity analyses were conducted to assess the robustness of the results using different methods of scoring, handling missing data, and adjusting for additional covariates. The study adhered to STROBE and ARRIVE 2.0 guidelines for reporting.

The study found significant negative associations between both MEDAS and MIND scores and overall cancer risk. These associations were consistent across various models adjusting for different covariates. Higher adherence to both diets was associated with a lower risk of overall cancer. Significant protective effects were observed for 13 specific cancers for both diets. The analysis identified 10 metabolites significantly associated with overall cancer risk: total lipids in VLDL, total cholines, omega-3 fatty acids, tyrosine, glucose, citrate, creatinine, albumin, free cholesterol in IDL, and total lipids in large HDL. Mediation analyses showed that several of these metabolites (total choline, omega-3 fatty acids, free cholesterol in IDL, total lipids in large HDL) played a mediating role in the relationship between diet scores and overall cancer risk. Subgroup analyses showed generally consistent associations across different demographic and lifestyle subgroups, although some interactions were observed between diet scores and specific variables (e.g., sex, smoking status, BMI) for some cancer types. Beneficial dietary components associated with lower cancer risk included vegetables, fruits, berries, nuts, whole grains, seafood, legumes, and wine. Conversely, butter, margarine, cheese, red meat, fast/fried food, and sweets were associated with higher cancer risk.

This study provides robust evidence supporting the association between adherence to MedDiet or MIND diet and a reduced risk of overall and several specific cancers. The identification of specific metabolites mediating this relationship contributes significantly to our understanding of the biological mechanisms underlying the protective effects of these diets. The large sample size and multiple sensitivity analyses enhance the reliability and generalizability of these findings. The study's findings align with previous research highlighting the benefits of plant-based diets rich in fruits, vegetables, and healthy fats. The identified metabolites offer potential targets for future cancer prevention and treatment strategies. However, further research is needed to confirm these findings in diverse populations and to elucidate the precise mechanisms through which these metabolites influence cancer development.

This large-scale prospective cohort study provides compelling evidence supporting the protective effects of MedDiet and MIND diets against overall and several specific cancers. The identification of key mediating metabolites highlights the complex interplay between diet, metabolism, and cancer risk. These findings underscore the importance of dietary interventions for cancer prevention and suggest potential therapeutic targets for future research. Future studies should investigate the causal relationships between these diets, metabolites, and cancer outcomes and explore the specific roles of individual dietary components and metabolites in cancer development.

The study is observational, limiting causal inference. Dietary information was self-reported, potentially introducing recall bias. A time lag between dietary assessment and cancer incidence data could introduce bias. The average diet score used might not fully reflect individual dietary patterns. The study lacked information on certain components of the MedDiet (e.g., olive oil, sofrito), which might affect the accuracy of the MEDAS scores. Finally, the large number of multiple comparisons in the metabolomics analysis warrants cautious interpretation of the results.