Table 3 Regression Models for predicting Insulin Resistance (IR)*.

From: InsuTAG: A novel physiologically relevant predictor for insulin resistance and metabolic syndrome

Surrogate Markers of IR

Model 1

Model 2

Model Coefficients

Model Statistics

β-coefficient

Model Coefficients

p

Model Statistics

β-coefficient

p

adj. R 2

p

Partial R2

Semi-Partial R2

adj. R 2

p

InsuTAG

  

0.739

<0.0001

    

0.738

<0.0001

InsuTAG*

0.637

<0.0001

  

0.613

0.650

0.487

<0.0001

  

Age

—

—

  

0.0002

<0.0001

<0.0001

0.930

  

Gender

—

—

  

−0.022

0.0009

0.0002

0.524

  

WC

—

—

  

0.003

0.0077

0.002

0.058

  

CRP

—

—

  

0.00005

<0.0001

<0.0001

0.988

  

TyG

  

0.183

<0.0001

    

0.303

<0.0001

TyG*

0.559

<0.0001

  

0.342

0.079

0.060

<0.0001

  

Age

—

—

  

−0.003

0.002

0.001

0.406

  

Gender

—

—

  

0.176

0.022

0.016

0.001

  

WC

—

—

  

0.020

0.149

0.120

<0.0001

  

CRP

—

—

  

0.007

0.003

0.002

0.214

  

TG/HDL-C Ratio

  

0.145

<0.0001

    

0.284

<0.0001

TG/HDL-C*

0.380

<0.0001

  

0.217

0.054

0.041

<0.0001

  

Age

—

—

  

−0.003

0.001

0.001

0.426

  

Gender

—

—

  

0.214

0.032

0.024

<0.0001

  

WC

—

—

  

0.021

0.157

0.132

<0.0001

  

CRP

—

—

  

0.006

0.003

0.002

0.260

  
  1. *IR defined by HOMA-IR value, InsuTAG, TyG and TG/HDL-C loge transformed prior to analysis. Model 1: Unadjusted estimates. Model 2: Adjusted for age, gender, waist circumference, CRP. A 1% increase in InsuTAG corresponds with a 0.265% increase in HOMA-IR or a 10% increase in InsuTAG corresponds with an increase to HOMA-IR of 2.57%.
Back to article page