Exploring the association between 5 different alternative indicators of insulin resistance and the risk of multiple cardiovascular and metabolic diseases: A cross-sectional NHANES study from 2005 to 2018

There is currently an absence of research exploring the correlation between insulin resistance (IR) surrogates and the risk of cardiometabolic multimorbidity (CMM). This study sheds light on the link between different IR surrogates to CMM risk and seeks to identify the optimal surrogate index for IR. Using the National Health and Nutrition Examination Survey 2005 to 2018 data, we applied logistic regression, the Boruta algorithm, trend tests, restricted cubic spline analysis, subgroup analysis, Brier scores, and receiver operating characteristic curve analysis to assess the relationship between CMM risk and IR markers including the triglyceride-glucose index (TyG index), the triglyceride-glucose-body mass index (TyG-BMI index), the metabolic score for insulin resistance (METS-IR), the triglyceride to high-density lipoprotein cholesterol ratio (TG/HDL-C ratio), and homeostasis model assessment of insulin resistance (HOMA-IR). The study included 15,537 participants, of whom 2881 developed CMM. Increased levels of IR markers were significantly associated with a higher CMM risk. Trend analysis showed a dose-response association (P for trend < .05). Restricted cubic spline analysis indicated a J-shaped nonlinear relationship for TyG index and ln[HOMA-IR] with CMM (P for overall < .001, P for nonlinear < .05). Conversely, ln[TyG-BMI], ln[METS-IR], and ln[TG/HDL-C] exhibited linear relationships with CMM (P for overall < .001, P for nonlinear > .05). HOMA-IR showed the highest area under the curve (0.699) for predicting CMM risk. As TyG, TyG-BMI, METS-IR, TG/HDL-C, and HOMA-IR levels increase, the CMM risk increases. Among these, HOMA-IR demonstrates a J-shaped nonlinear relationship with CMM risk and the best predictive performance.