Does body mass index mediate the association between dietary patterns and colorectal neoplasia among individuals at above-average risk for colorectal cancer?

BACKGROUND & AIMS: Modifiable factors, including dietary patterns and body mass index (BMI) are associated with the risk of colorectal neoplasia in the general population. However, evidence is limited on whether the association between dietary patterns and colorectal neoplasia is mediated through BMI among individuals at above-average risk for colorectal cancer due to a family or personal history of colorectal neoplasia. Therefore, this study aimed to investigate the association between dietary patterns and colorectal neoplasia and assess the mediating role of BMI on their associations. METHODS: Dietary data was collected using the Australian Eating Survey (2023-2024) from 1917 participants undergoing regular surveillance colonoscopy. Linked clinical data assessed presence or absence of neoplasia at the closest colonoscopy. Dietary patterns were analysed using principal component analyses and separated into tertiles (where tertile I indicated the lowest intake to that specific pattern). The association between dietary patterns and colorectal neoplasia, its advanced form and histological polyp subtypes (adenoma or sessile serrated lesions) was analysed using multivariable logistic regression, and the mediating role of BMI on this association was analysed using a mediation model. Adjusted odds ratios (AOR) with 95% confidence interval (CI) were reported. RESULTS: Two dietary patterns were identified: dietary pattern I was characterised by higher intake of vegetables, fruits, and seafood, while dietary pattern II was characterised by higher intake of fast foods, red/processed meat and sweet snacks. Only tertile II level of intake in dietary pattern I was statistically associated with lower odds of adenoma (AOR = 0.75; 95%CI 0.58-0.97) and advanced adenoma (AOR = 0.74; 95%CI 0.58-0.94) compared with tertile I. This association was not mediated by BMI (P = 0.802), but the direct effect (p < 0.001) was statistically significant. CONCLUSION: Dietary patterns are associated with incidence of adenomas, therefore implementing dietary interventions irrespective of BMI status is advisable for minimising adenoma risk.