Table 4.

Stepwise linear regression model indicating predictors of HOMA-IR index in male (A) and female (B) study participants

(A) Male (n = 195)
Model	Predictor	R2	Unstandardized B coefficient	Standard error	p-value
1	log FM (Kg)	0.318	−1.641	0.211	< 0.001***
			1.426	0.152	< 0.001***
2	log FM (Kg) log VAI	0.358	−1.373	0.220	< 0.001***
			1.238	0.157	< 0.001***
			0.229	0.067	0.001**
Dependent variable: log HOMA-IR
Model 1 excluded variables: log AGEs (kU/day), log BMI (Kg/mq), log VAI Model 2 excluded variables: log AGEs (kU/day), log BMI (Kg/mq)

(B) Female (n = 239)
Model	Predictor	R2	Unstandardized B coefficient	Standard Error	p-value
1		0.213	−2.263	0.326	< 0.001***
	log BMI		1.780	0.230	< 0.001***
2		0.307	−1.825	0.317	< 0.001***
	log BMI		1.459	0.224	< 0.001***
	log VAI		0.327	0.060	< 0.001***
3		0.343	−2.735	0.407	< 0.001***
	log BMI		1.468	0.219	< 0.001***
	log VAI		0.340	0.058	< 0.001***
	log AGEs		0.223	0.065	0.001***
Dependent variable: log HOMA-IR
Model 1 excluded variables: log AGEs (kU/day), log FM (Kg), log VAI Model 2 excluded variables: log AGEs (kU/day), log FM (Kg) Model 3 excluded variables: log FM (Kg)

***p-value ≤ 0.001

BMI, body mass index; FM, fat mass; VAI, visceral adiposity index; AGEs, advanced glycation end products