A novel score to evaluate nonlinear association between insulin resistance and new-onset diabetes mellitus: A cohort analysis

The rising global prevalence of diabetes mellitus (DM) highlights the need for early risk assessment to delay disease onset. This study aims to evaluate the association between metabolic score for insulin resistance (Mets-IR) and the risk of new-onset DM. Data were obtained from the China Health and Retirement Longitudinal Study database. Participants aged <50 years or with missing data were excluded. Using 2011 as the baseline, the study endpoint was defined as the occurrence of new-onset DM during follow-up until 2018. The association between Mets-IR and new-onset DM risk was assessed using logistic regression models. Subgroup and sensitivity analyses were performed to evaluate robustness. Nonlinear associations were explored through threshold effect analysis and smoothed curve fitting. A total of 5257 people were included. Mets-IR was 1 unit per liter, and the risk of new-onset diabetes increased by 1%. The highest quartile risk was the lowest 5.67 times, showing a dose–response relationship. The threshold is 49.3, and the following risks rise sharply by 11% per unit. Subgroups showed that those under 65 years old, married, rural, nonsmokers, and those without liver and kidney diseases were more associated. The nonlinear association remains after comprehensive adjustment for confounders, suggesting that Mets-IR may provide a valuable additional tool for DM risk stratification and early intervention.