Dose-Response Associations of Metabolic Score for Insulin Resistance Index with Nonalcoholic Fatty Liver Disease among a Nonobese Chinese Population: Retrospective Evidence from a Population-Based Cohort Study

PURPOSE: This study is aimed at investigating the association between the metabolic score for insulin resistance (METS-IR) index and nonalcoholic fatty liver disease (NAFLD) in the nonobese population and its predictive value. METHODS: 10730 nonobese subjects were selected from longitudinal cohort research conducted from January 2010 to December 2014. Cox proportional hazards models were employed to assess the relationship between METS-IR and new-onset NAFLD. Generalized additive models were used to identify nonlinear relationships. In addition, we performed subgroup analyses and interaction tests. The time-dependent receiver operating curve (ROC) and area under the ROC (AUC) were utilized to measure the discriminatory ability of METS-IR for new-onset NAFLD. Beyond clinical risk factors, the incremental predictive value of METS-IR was appraised using integrated discrimination improvement (IDI), C-index, and net reclassification index (NRI). RESULTS: Over a median period of 804.50 days of follow-up, 1859 (17.33%) participants had a new onset of NAFLD. After adjusting for confounders, the HR for new-onset NAFLD in the Q4 group was 6.40 compared with the Q1 group. When METS-IR was considered a continuous variable, the risk of NAFLD increased by 34% for every 1 SD increase in METS-IR. The smoothing curve shows the dose-response relationship between METS-IR and the presence of new-onset NAFLD. Using a two-piecewise linear regression model, we derived a METS-IR inflection point of 36. HRs were 1.31 on the left side of the inflection point and 1.04 on the right side of the inflection point (log-likelihood ratio test, P < 0.001). Subgroup analyses and interaction tests revealed an interaction between gender and SBP in the association between METS-IR and new-onset NAFLD. In the subgroup analysis of gender and SBP, we observed a higher risk of new-onset NAFLD in men and in those with abnormal SBP levels. We evaluated the ability of METS-IR to identify new-onset NAFLD at different time points. The AUCs at 1, 2, 3, and 4 years were 0.784, 0.756, 0.758, and 0.752, respectively, which represent good discrimination of new-onset NAFLD. The addition of METS-IR greatly improved the reclassification and differentiation of clinical risk factors, with an NRI of 0.276 and an IDI of 0.068. In addition, the addition of METS-IR increased the C-index from 0.719 to 0.771. CONCLUSION: In a nonobese Chinese population, elevated METS-IR was independently associated with an enhanced risk of NAFLD development and a dose-response relationship existed. In addition, METS-IR might be a reliable indicator for screening individuals at risk for early NAFLD, especially in nonobese populations.