Table 2.
Model | Description |
---|---|
Base | No anthropometric or MS predictors. |
ARI | ARI as a predictor. |
MS | MS occurrence as a predictor. |
MS score | MS score as a predictor. |
ARI + MS | ARI and MS occurrence as predictors. |
ARI + MS score | ARI and MS score as predictors. |
ARI + MSx score | ARI and MS score (excluding Waist) as predictors. Excluding the WC component from the MS score was hypothesized to be reasonable when considering it alongside ARI since the BMI and ABSI components of ARI already correlate strongly with WC. |
MS components | The 5 MS components as individual predictors. This tested the relative utility of the components of MS, which could suggest refinements of its definition for testing in future work. |
ARI + MS components | ARI and the 5 MS components (Table 1) as individual predictors. |
ARI + MSx components | ARI and the 4 MS components (excluding Waist) as individual predictors. |
Models for predicting mortality that were compared in this study. MS: metabolic syndrome; ARI: anthropometric risk index; WC: waist circumference; BMI: body mass index; ABSI: a body shape index.