Background: There is an increasing awareness that diet-related inflammation may have an impact on the stroke. Herein, our goal was to decipher the association of dietary inflammatory index (DII) with stroke in the US general population.
Methods: We collected the cross-sectional data of 44,019 participants of the National Health and Nutrition Examination Survey (NHANES) 1999-2018. The association of DII with stroke was estimated using weighted multivariate logistic regression, with its nonlinearity being examined by restricted cubic spline (RCS) regression. The least absolute shrinkage and selection operator (LASSO) regression was applied for identifying key stroke-related dietary factors, which was then included in the establishment of a risk prediction nomogram model, with the receiver operating characteristic (ROC) curve being built to evaluate its discriminatory power for stroke.
Results: After confounder adjustment, the adjusted odds ratios (ORs) with 95% confidence intervals (CIs) for stroke across higher DII quartiles were 1.19 (0.94-1.54), 1.46 (1.16-1.84), and 1.87 (1.53-2.29) compared to the lowest quartile, respectively. The RCS curve showed a nonlinear and positive association between DII and stroke. The nomogram model based on key dietary factors identified by LASSO regression displayed a considerable predicative value for stroke, with an area under the curve (AUC) of 79.8% (78.2-80.1%).
Conclusions: Our study determined a nonlinear and positive association between DII and stroke in the US general population. Given the intrinsic limitations of cross-sectional study design, it is necessary to conduct more research to ensure the causality of such association.
Keywords: Cross-sectional study; DII; LASSO; NHANES; Stroke.
© 2024. The Author(s).