Related Experiment Videos
Dietary Inflammatory Index and risk of nonalcoholic fatty liver disease: nonlinear associations, metabolic mediation,
Tongyan Chen1, Guanghui Zheng1, Huiyuan Tan2
1Laboratory Diagnosis Center, Beijing Tiantan Hospital, Capital Medical University, Beijing, China.
Background:
Nonalcoholic fatty liver disease (NAFLD) is a common metabolic disorder characterized by chronic low-grade inflammation. The Dietary Inflammatory Index (DII) reflects the inflammatory potential of habitual diet and has been associated with several cardiometabolic diseases. However, the relationship between DII and NAFLD is not fully clarified. In particular, it remains unclear which metabolic factors mediate this association, and whether findings based largely on NHANES data can be confirmed in independent local populations.
Methods:
This cross-sectional study included adults from NHANES and an independently collected local cohort. Multivariable logistic regression was used to assess the association between DII and NAFLD. Restricted cubic spline models examined dose-response patterns. Subgroup analyses evaluated the consistency of the findings. Mediation analyses quantified the indirect effects of metabolic factors in the DII-NAFLD association.
Results:
In the NHANES analysis, higher DII scores were independently associated with an increased risk of NAFLD after adjustment for demographic, socioeconomic, and clinical covariates. Restricted cubic spline analysis indicated a nonlinear association, with NAFLD risk rising more steeply at higher DII levels. Mediation analyses demonstrated that triglycerides, fasting plasma glucose, and insulin partially mediated the DII-NAFLD relationship, although a significant direct effect of DII persisted. Consistent findings were observed in the independent local cohort, where higher dietary inflammatory scores were also associated with greater NAFLD risk.
Conclusion:
DII was positively associated with NAFLD and showed a nonlinear dose-response pattern. This association was partially explained by glucose and lipid metabolic disturbances, and machine learning analysis identified DII as an important predictor of NAFLD. Consistent findings in the local cohort supported the robustness of this association. However, causal relationships cannot be established due to the cross-sectional design.
