Skip to main content
. 2000 Aug 12;321(7258):420–424. doi: 10.1136/bmj.321.7258.420

Table.

Relative risk (95% confidence interval) for development of diabetes assuming linear trend across categories

Univariate analysis P value* Multivariate analysis P value*
Maternal age (years) v <20
 20-24 1.25 (1.19 to 1.32) <0.0001 1.25 (1.17 to 1.34) <0.0001
25-29 1.57 (1.39 to 1.77) 1.57 (1.33 to 1.85)
30-34 1.97 (1.65 to 2.35) 1.98 (1.55 to 2.52)
35-39 2.47 (1.95 to 3.13) 2.48 (1.79 to 3.43)
⩾40 3.10 (2.31 to 4.16) 3.11 (2.07 to 4.66)
Paternal age (years) v <20
 20-24 1.18 (1.12 to 1.24) <0.0001 1.09 (1.03 to 1.16) 0.003
25-29 1.39 (1.24 to 1.54) 1.20 (1.05 to 1.37)
30-34 1.63 (1.39 to 1.92) 1.31 (1.08 to 1.59)
35-39 1.92 (1.55 to 2.38) 1.43 (1.10 to 1.86)
40-45 2.27 (1.74 to 2.95) 1.57 (1.13 to 2.17)
Birth order v 1st born
 2nd 0.97 (0.92 to 1.03) 0.37 0.85 (0.79 to 0.90) <0.0001
3rd 0.95 (0.85 to 1.06) 0.72 (0.64 to 0.80)
4th 0.93 (0.78 to 1.09) 0.61 (0.52 to 0.71)
⩾5th 0.90 (0.72 to 1.13) 0.52 (0.42 to 0.63)
Male v female 1.20 (1.08 to 1.33) 0.0007 1.21 (1.08 to 1.35) <0.0001
*

Significance of overall effect of variable in model. 

Relative risk adjusted for three other factors in model.