Skip to main content

Table 3 Comparison of optimal BMI and 5-year mortality estimates across models

From: Examining the BMI-mortality relationship using fractional polynomials

  MFP Linear-Quadratic Categorical
Never Smokers    
Male    
   Optimal BMI 26.97 [26.41, 27.54] 31.84 [30.34, 33.34] 25-30
   Mortality at Optimum 0.0176 [0.0158, 0.0196] 0.0188 [0.0168, 0.0210] 0.0169 [0.0152, 0.0188]
   Mortality at BMI = 50 0.0548 [0.0428, 0.0699] 0.0544 [0.0383, 0.0766] 0.0355 [0.0251, 0.0499]
Female    
   Optimal BMI 22.34 [20.10, 24.57] 23.04 [18.15, 27.93] 25-30
   Mortality at Optimum 0.0100 [0.0088, 0.0115] 0.0128 [0.0115, 0.0141] 0.0120 [0.0107, 0.0134]
   Mortality at BMI = 50 0.0355 [0.0274, 0.0460] 0.0306 [0.0377, 0.0248] 0.0291 [0.0233, 0.0363]
Ever Smokers    
Male    
   Optimal BMI 26.97 [26.41, 27.54] 31.84 [30.34, 33.34] 25-30
   Mortality at Optimum 0.0306 [0.0283, 0.0331] 0.0329 [0.0301, 0.0358] 0.0299 [0.0273, 0.0326]
   Mortality at BMI = 50 0.0939 [0.0750, 0.1168] 0.0926 [0.0667, 0.1271] 0.0619 [0.0447, 0.0852]
Female    
   Optimal BMI 22.34 [20.10, 24.57] 23.04 [18.15, 27.93] 25-30
   Mortality at Optimum 0.0223 [0.0199, 0.0250] 0.0237 [0.0217, 0.0260] 0.0223 [0.0200, 0.0248]
   Mortality at BMI = 50 0.0765 [0.0610, 0.0955] 0.0559 [0.0459, 0.0680] 0.0534 [0.0433, 0.0657]
  1. 95% confidence intervals are in brackets.