Dietary inflammatory index predicts cancer mortality in male patients but not female patients: Results from NHANES 1999 to 2014

Abstract

This study explored sex differences between dietary inflammatory index (DII) and cancer prognosis and their mechanisms. We hypothesized that association between dietary inflammatory index and cancer prognosis differs by sex. The study included 2874 adults with cancer from the National Health and Nutrition Examination Survey covering 1999 to 2014. Mortality status was linked to National Death Index mortality data through 31 December 2019. Cox proportional hazards regression models were applied to calculate hazard risk and 95% confidence intervals (Cis) in male patients and female patients. Sex-specific cancer and nonsex-specific cancer subgroup analyses were performed, and the role of C-reactive protein in sex differences was analyzed. The Cancer Genome Atlas pan-cancer transcriptome data were combined to explore the biological mechanisms of the sex differences. Multivariate Cox regression showed higher DII in male patients correlated with increased all-cause mortality (hazard risk highest vs lowest quartile = 1.57 [95% confidence intervals 1.24-1.98]; P for trend <.01), but not in female patients (P = .44). For sex-specific cancers, higher DII potentially correlated with increased mortality in prostate cancer (unadjusted P for trend = .04), but not in breast (P = .83), ovarian (P = .49), or cervical cancers (P = .91). In melanoma and colon cancer, higher DII correlated with increased mortality in male patients but not female patients. Serum C-reactive protein, interleukin-1 binding, interleukin-35 pathway, and programmed cell death protein 1 pathway may contribute to these sex differences. In conclusion, sex differences exist between DII and mortality risk in cancer patients.