A cross-sectional exploration of the dietary inflammation index association with cardiovascular disease in gout: application of machine learning algorithms

Qiang Zhang^1*Xue-bing Lyu²Chang-quan Liu²Wei-zhen Zhang²Yu-guang Wang³Wei-zhe Deng^1*Xuan-hua Yu^2*

• ¹The 962nd Hospital of the PLA, Department of Rheumatology and Chinese Medicine, Harbin, Jilin Province, China
• ²Department of Rheumatology, People's Hospital Affiliated to Fujian University of Traditional Chinese Medicine, Fuzhou, Fujian Province, China
• ³962 Hospital of Chinese People's Liberation Army, Harbin, China

Objective: Gout is a condition strongly associated with dietary patterns and elevated risk of cardiovascular disease (CVD) in affected individuals. Given the potential influence of dietary diversity on inflammatory responses, this study aimed to explore the association between the dietary inflammatory index (DII) and CVD prevalence in gout patients. Methods: Data from gout patients in NHANES 2007-2018 were extracted for analysis. Correlation matrices were employed to examine the relationships among 28 dietary inflammation indices. Machine learning algorithms were utilized to identify key features for constructing a covariate subset for the final model, and Random Forest SHAP interpretations were applied to assess variable risk factors. The relationship between DII and CVD risk in gout patients was assessed using multi-model logistic regression. RCS were applied to evaluate the risk trend and to assess model discrimination, predictive probability, and clinical benefit using ROC, calibration curves, and DCA, respectively. Subgroup analysis was evaluated the heterogeneity in CVD across different populations. Results: 1,437 gout patients met inclusion criteria were included in the study, with mean age of 60.84 years, consisting of 435 females (31.23%) and 1,002 males (68.77%), and an overall CVD prevalence of 32.92%. DII was linearly associated with CVD risk (P for overall = 0.002; P for nonlinear = 0.810). In the final model, DII was positively associated with CVD risk, showing 118% increased risk in Q4 compared to Q1 (OR: 2.18, 95%CI: 1.52-3.13, P < 0.001). The constructed model exhibited stability performance (AUC = 0.750, 95%CI: 0.722-0.775). Segmented subgroup analysis indicated that gout patients with high DII (> 1.934) had a increased risk of CVD (OR: 1.33, 95%CI: 0.06-1.65, P = 0.012), while those younger than 60 years had higher risk (OR: 2.19, 95%CI: 1.36-3.54, P = 0.001). Conclusion: Higher DII was associated with increased prevalence of CVD in gout patients. Dietary modification may serve as an effective strategy for preventing disease progression and reducing CVD risk. Our findings support the clinical development of dietary and nutritional guidance programs.