Abstract
PURPOSE: The purpose of the study was to examine the relationship between proinflammatory diet and prostate cancer risk. METHODS: Energy-adjusted Dietary Inflammatory Index (E-DII) scores were computed among 40,161 participants in the California Men's Health Study. Over 9.7 ± 3.8 years of follow-up, 2707 incident prostate cancer cases were diagnosed and categorized as low-, intermediate-, or high-risk, based on disease grade and stage. Accelerated failure-time models assessed time to diagnosis of prostate cancer. Cox proportional hazard models estimated hazard ratios (HR) and 95% confidence intervals (95% CI). Nonlinear effects of E-DII were modeled as third-order polynomials. RESULTS: Time to prostate cancer diagnosis did not differ by E-DII quartile. The HR for high-risk prostate cancer increased in the third E-DII quartile (HR(Q3 vs. Q1) = 1.36; 95% CI: 1.04-1.76), but not in the fourth (HR(Q4 vs. Q1) = 0.99; 95% CI: 0.74-1.32, P(trend) = .74), suggesting a nonlinear dose-response. HR curves for prostate cancer increased exponentially above an E-DII threshold of ≈+3.0. HR curves for high-risk prostate cancer had a much steeper incline above an E-DII threshold of ≈+2.5. Curves were higher among Blacks and Whites relative to other races and among overweight or obese men. No relationship was observed between E-DII scores and intermediate- or low-risk disease. CONCLUSIONS: Relationships between proinflammatory diet and prostate cancer risk may be nonlinear, with an increased risk above an E-DII threshold of ≈+2.5.