Abstract:Objective To explore the relationship between homeostasis model assessment of insulin resistance index (HOMA-IR) and non-alcoholic fatty liver disease (NAFLD). Methods The registered participants aged 18 years or older in the National Health and Nutrition Examination Survey (NHANES) database in the United States from 2017 to 2020 were selected, and those who lacked liver transient elastography data, HOMA-IR index, and those with significant alcohol consumption, hepatitis B, and hepatitis C were excluded. The research subjects were classified into 4 groups (Q1-Q4 group) according to the HOMA-IR quartiles. Age, gender, race, marital status, education, family income-to-poverty ratio (PIR), body mass index (BMI), smoking, alcohol consumption, diabetes mellitus (DM), hypertension, hyperlipidemia, and HOMA-IR level were compared among four groups. The risk association between HOMA-IR and NAFLD was explored through descriptive analysis, multivariable logistic regression, restricted cubic spline regression analysis, subgroup and interaction analysis. Results A total of 3 370 cases were enrolled, including 843 cases in Q1 group, 842 cases in Q2 group, 841 cases in Q3 group, and 844 cases in Q4 group. After adjusting for age, gender, race, education, marital status, PIR, BMI, smoking, alcohol intake, DM, hypertension, and hyperlipidemia, compared with HOMA-IR Q1 group, the HOMA-IR Q4 group had a significantly increased risk of NAFLD (OR=7.055, 95%CI: 5.136-9.772, P＜0.01). There was a positive dose-response relationship between HOMA-IR and the risk of NAFLD (P＜0.01), and the relationship was nonlinear (P_non-linear=0.002). HOMA-IR had significant interactions with gender, PIR, BMI, and smoking, with statistical significances (P_interaction＜0.05). Conclusion The HOMA-IR level is significantly associated with the risk of NAFLD.