Skip to main content

Table 2 Logistic regression coefficient estimates and standard errors in parentheses for the final adjustment model including smoking status, fractionally transformed BMI1 and age2 and interactions identified as significant

From: Examining the BMI-mortality relationship using fractional polynomials

  Male Female
BMI(p1) 24.260 20.328
  (1.938) (5.186)
BMI(p2) -49.284 -18.490
  (4.307) (3.766)
Age(q1) 0.077  
  (0.001)  
Age(q1)*BMI(p1)   0.244
   (0.165)
Age(q1)*BMI(p2)   -0.115
   (0.125)
Age(q1) | Ever Smoked = 0   0.082
   (0.002)
Age(q1) | Ever Smoked = 1   0.075
   (0.002)
Ever Smoked 0.581 0.832
  (0.050) (0.084)
Constant -4.366 -4.712
  (0.054) (0.075)
Log Likelihood -9503.563 -10179.990
  1. 1 For the male sample, B M I ( p 1 ) = B M I 10 - 2 - 0.137, B M I ( p 2 ) = B M I 10 - 2 * ln B M I 10 - 0.376 and for the female sample, B M I ( p 1 ) = B M I 10 - 2 - 0.142, B M I ( p 2 ) = B M I 10 - 1 - 0.376.
  2. 2 For the male sample, Age ( q 1 ) = Age 10 2 - 20.692 and for the female sample, Age ( q 1 ) = Age 10 2 - 22.216.