Table 3

Maximally adjusted pooled women to men ratio of relative risks (RRR) for lung cancer associated with smoking, in subgroup analyses

N studiesFormer versus neverP for interaction*N studiesCurrent versus notP for interaction*
Study region
 Asia401.07 (0.83 to 1.37)460.93 (0.72 to 1.20)
 Non-Asia490.73 (0.58 to 0.93)0.06530.87 (0.66 to 1.14)0.73
Study region
 Asia401.07 (0.83 to 1.37)460.93 (0.72 to 1.20)
 USA50.60 (0.42 to 0.84)60.58 (0.37 to 0.91)
 Europe360.81 (0.60 to 1.10)370.99 (0.63 to 1.57)
 ANZ81.41 (0.65 to 3.04)0.55101.11 (0.97 to 1.28)0.69
Year of study baseline
 1985 or before230.79 (0.56 to 1.12)250.96 (0.66 to 1.40)
 After 1985660.92 (0.73 to 1.16)0.43740.85 (0.68 to 1.06)0.66
Women-to-men smoking prevalence
 >67% lower in women391.16 (0.83 to 1.62)430.96 (0.72 to 1.28)
 33%–67% lower in women190.72 (0.51 to 1.03)200.75 (0.50 to 1.14)
 0%–33% lower in women280.78 (0.59 to 1.03)0.26290.97 (0.59 to 1.58)0.90
Women-to-men lung cancer rate
 ≥50% lower in women800.85 (0.65 to 1.10)830.85 (0.62 to 1.17)
 0%–50% lower in women60.83 (0.61 to 1.13)0.7990.94 (0.68 to 1.31)0.64
Study endpoint
 Fatal lung cancer only570.94 (0.68 to 1.29)670.97 (0.77 to 1.21)
 Fatal and non-fatal lung cancer320.80 (0.64 to 1.01)0.57320.72 (0.48 to 1.06)0.21
Duration of follow-up
 ≤10 years480.90 (0.60 to 1.35)530.91 (0.68 to 1.24)
 >10 years410.85 (0.71 to 1.05)0.92460.87 (0.67 to 1.12)0.83
Study quality
 ≤6 points580.85 (0.64 to 1.13)610.84 (0.53 to 1.12)
 >6 points310.89 (0.66 to 1.20)0.81380.93 (0.72 to 1.20)0.69
  • Random-effects meta-analyses were used for all subgroup analyses and differences between subgroups were examined using meta-regression.

  • *P for interaction assessed using meta-regression.

  • ANZ, Australia and New Zealand.