Single-point insulin sensitivity estimator and atrial fibrillation: association and incremental discriminative value...

The single-point insulin sensitivity estimator (SPISE), derived from triglycerides, high-density lipoprotein cholesterol, and body mass index, has emerged as an easily calculated marker of insulin sensitivity. Its association with atrial fibrillation (AF) has not been well defined.

This retrospective cross-sectional study included 1,122 hospitalized participants, including 309 without AF, 409 with paroxysmal AF, and 404 with persistent AF. The association between SPISE and AF was examined with logistic regression by modeling SPISE continuously, per standard deviation (SD), and by quartiles. Restricted cubic splines were used to assess the shape of the association. Candidate variables were reduced by LASSO and then entered into univariable and multivariable logistic regression to construct the AF classification model. Model performance was evaluated with discrimination, calibration, decision-curve analysis, repeated 10-fold cross-validation, and bootstrap optimism correction. The incremental discriminative value of SPISE was assessed by changes in area under the receiver operating characteristic curve (AUC), net reclassification improvement (NRI), and integrated discrimination improvement (IDI).

Lower SPISE was associated with AF after multivariable adjustment. In the fully adjusted model, the corresponding odds ratios were 0.68 (0.61-0.76) per unit increase and 0.56 (0.47-0.66) per SD increase. Compared with the lowest quartile, the highest quartile showed lower odds of AF, with an odds ratio of 0.36 (0.23-0.58). The relationship was predominantly linear. The AF classification model achieved an AUC of 0.859 (0.836-0.882), and the mean cross-validated C-statistic was 0.855. Adding SPISE increased AUC from 0.843 to 0.862 and significantly improved both NRI and IDI; however, the absolute AUC gain was small, although NRI and IDI improved statistically.

Lower SPISE was associated with AF after multivariable adjustment and added discriminative information beyond conventional clinical variables in this hospital-based sample. Given the small absolute improvement in AUC, SPISE should be interpreted as an adjunctive marker of AF-related metabolic phenotype rather than a stand-alone component of routine AF assessment. Prospective studies are needed to determine whether SPISE is related to incident AF.