Skip to main content

Table 1 The effect of BMI on macrosomia: Different reference categories influence effect estimates and precision.

From: Categorisation of continuous exposure variables revisited. A response to the Hyperglycaemia and Adverse Pregnancy Outcome (HAPO) Study

  

Crude analysis

Adjusted analysis**

        

HAPO model I

HAPO model II

Reference category

BMI <22.6

BMI 22.6-28.4

BMI <22.6

BMI 22.6-28.4

BMI <22.6

BMI 22.6-28.4

BMI (kg/m2)*

n

Beta (SE)

OR (95% CI)

Width of CI

Beta (SE)

OR (95% CI)

Width of CI

OR

OR

OR

OR

<22.6

2974

0

1.00

 

-0.80 (0.10)

0.45 (0.37-0.55)

0.2

1.00

0.45

1.00

0.46

22.6-28.4

11934

0.80 (0.10)

2.23 (1.83, 2.71)

0.9

0

1.00

 

2.24

1.00

2.17

1.00

28.5-32.9

5127

1.27 (0.10)

3.57 (2.91, 4.38)

1.5

0.47 (0.05)

1.60 (1.44-1.78)

0.3

3.62

1.62

3.31

1.52

33.0-37.4

2064

1.47 (0.11)

4.36 (3.49, 5.45)

2.0

0.67 (0.07)

1.96 (1.71-2.25)

0.5

4.43

1.98

3.89

1.79

37.5-41.9

735

1.50 (0.14)

4.47 (3.40, 5.88)

2.5

0.70 (0.11)

2.01 (1.63-2.48)

0.9

4.52

2.02

3.80

1.76

≥42.0

383

1.55 (0.17)

4.71 (3.38, 6.56)

3.2

0.75 (0.14)

2.12 (1.60-2.80)

1.2

4.55

2.03

3.52

1.62

  1. * BMI categories are based on BMI measured at 28 weeks gestation, and the categories correspond to pre-gestational BMI categories as suggested by WHO: "underweight" (<18.5), "normal weight" (18.5-24.9), "overweight" (25.0-29.9) and "obese class I-III" (30.0-34.9, 35.0-39.9 and ≥40, respectively) [4].
  2. ** HAPO Model I: Adjusted for the variables used in estimating 90th birth weight percentiles (gender, gestational age, ethnicity, field centre, and maternal parity), age, height and gestational age at the oral glucose tolerance test, smoking, alcohol use, hospitalisation before delivery and any family history of diabetes. Model II: Model I with additional adjustment for fasting plasma glucose and mean arterial pressure [4].
  3. The table shows results from logistic regression analyses of the effect of BMI on macrosomia, with two different choices of reference category. All results are based on the HAPO paper [4].
Back to article page

BMC Medical Research Methodology

ISSN: 1471-2288

Contact us