Table 2—

Treatment categories: description, incidence of lung cancer, and comparison with nondiabetic control subjects by observation period

Treatment groupPatients n (% total)Age (mean ± SD)Mean years obser-vation*Lung cancer (n)*Incidence (95% CI)Comparison with nondiabetic control subjects [total number of cases, hazard ratio (95% CI)]
Primary end dateGroup end date
No drug treatment13,330 (19.9)64.7 ± 14.72.91112.91 (2.42–3.51)520, 1.22 (0.99–1.52)394, 1.00 (0.77–1.31)
Biguanides only4,852 (7.3)60.4 ± 13.42.4211.77 (1.15–2.71)150, 0.97 (0.61–1.55)86, 1.32 (0.76–2.29)
Insulin only11,628 (17.4)42.9 ± 21.14.2360.72 (0.52–0.99)230, 0.94 (0.66–1.35)90, 1.02 (0.60–1.73)
Biguanides + any other treatment17,727 (26.5)62.2 ± 12.54.7961.16 (0.95–1.41)788, 0.63 (0.51–0.79)222, 1.14 (0.82–1.60)
Oral treatment only (not biguanides)17,400 (26.0)68.5 ± 12.72.91523.02 (2.58–3.55)879, 1.16 (0.97–1.39)457, 1.13 (0.89–1.43)
Insulin + oral non-biguanide1,911 (2.9)60.0 ± 16.64.5131.53 (0.89–2.63)91, 0.76 (0.42–1.38)30, 1.24 (0.53–2.91)
All diabetic patients66,84860.8 ± 17.33.94301.63 (1.48–1.79)2,659, 0.88 (0.79–0.98)1,280, 0.98 (0.84–1.13)
  • *

    * From date of first prescription for those in prescription treatment groups and from date of first diagnosis for others;

  • incidence of lung cancer per 1,000 person-years using primary end date;

  • multivariable Cox regression with the nondiabetic control subjects as the baseline, adjusting for age, sex, and smoking.