Dietary inflammatory index and incident prostate cancer

Abstract

The Dietary Inflammatory Index (DII) quantifies the inflammatory potential of diet. Evidence linking DII to prostate cancer remains limited and inconsistent. We evaluated the prospective association between DII and incident prostate cancer in a large population-based cohort. We included 30,972 men who were free of prostate cancer at baseline. The DII was derived from 27 dietary components assessed by the Oxford WebQ. Incident prostate cancer was identified from linked cancer-registry records (ICD-10 C61). Cox regression was the primary analysis, with age included in all adjusted models; logistic regression was secondary. Lag, additional-covariate, and age-stratified Cox models assessed sensitivity. Mediation and Mean Adequacy Ratio (MAR) analyses were exploratory. During a median follow-up of 13.8 years (IQR 13.5–14.2; 406,058 person-years), 1,546 participants developed incident prostate cancer. The unadjusted inverse association (HR per 1-unit DII increase 0.951, 95% CI 0.927–0.977) was attenuated after adjustment for age and other covariates (fully adjusted HR 0.986, 95% CI 0.956–1.018; P = 0.389; Q4 vs Q1 HR 0.928, 95% CI 0.788–1.093; P for trend = 0.426). The secondary logistic analysis was likewise null (OR per 1-unit increase 0.982, 95% CI 0.950–1.015; P = 0.275). Age-adjusted lag analyses and exploratory CRP/NLR mediation analyses did not materially alter this interpretation. DII was inversely correlated with the revised 18-nutrient MAR (Pearson r = −0.694; Spearman ρ = −0.756), but neither index was associated with prostate cancer in the age-adjusted joint model. An apparent inverse DII–prostate cancer association in unadjusted analyses was explained largely by age differences across DII categories. After adjustment for age and other covariates, neither logistic nor time-to-event analyses supported an independent association. These findings underscore the importance of rigorous age control in prostate cancer epidemiology and should not be interpreted causally.