American Association for Cancer Research

Table 2

Association of total testosterone with age, ethnicity, BMI, and waist circumference; regression coefficients (β)^a from Generalized Estimating Equations

Variable	Model 1^b			Model 2^b			Model 3^b
Variable		β (ng/ml)	P	β (ng/ml)	P	β (ng/ml)	P
Visit age, per yr	−0.0579	0.0013	−0.0521	0.0024	−0.0433	0.012
Black race	0.0764	0.44	0.2059	0.028	0.0493	0.62
Year 2 BMI, per 1 kg/m²			−0.1091	<0.0001	0.0119	0.65
Δ BMI, per 1 kg/m²			−0.1766	<0.0001	−0.1087	<0.0001
Year 2 waist circumference, per 1 cm					−0.0544	<0.0001
Δ waist circumference, per 1 cm					−0.0278	0.0006

Variable	Model 1^b			Model 2^b			Model 3^b
Variable		β (ng/ml)	P	β (ng/ml)	P	β (ng/ml)	P
Visit age, per yr	−0.0579	0.0013	−0.0521	0.0024	−0.0433	0.012
Black race	0.0764	0.44	0.2059	0.028	0.0493	0.62
Year 2 BMI, per 1 kg/m²			−0.1091	<0.0001	0.0119	0.65
Δ BMI, per 1 kg/m²			−0.1766	<0.0001	−0.1087	<0.0001
Year 2 waist circumference, per 1 cm					−0.0544	<0.0001
Δ waist circumference, per 1 cm					−0.0278	0.0006

Mean change in total testosterone corresponding to the indicated difference in the independent variable.

Each of the models also included terms for secular changes in hormone levels between the Years 2 and 7, and the Years 2 and 10 examinations.

View Large

This Feature Is Available To Subscribers Only