Table 2.

Colorectal cancer risk by smoking status, intensity, duration, and latency period

Male (1,112 case-control pairs)
Female (815 case-control pairs)
n*OR (95% CI)n*OR (95% CI)
Never smoker (reference group)363/2821.00529/4891.00
Smoking status
    Former smoker574/6071.21 (0.97-1.50)185/2131.32 (1.02-1.71)
    Current smoker175/2231.32 (0.99-1.76)101/1131.17 (0.85-1.62)
Intensity (cigarettes per day)
    MalesFemales
        1-101-5175/1590.94 (0.62-1.43)54/641.88 (1.02-3.45)
        11-206-10275/2821.00 (0.67-1.49)81/661.06 (0.58-1.94)
        21-3011-20140/1390.89 (0.56-1.42)90/1101.68 (0.90-3.15)
        >30>20156/2471.51 (0.99-2.29)61/851.73 (0.87-3.43)
    Ptrend§0.0070.33
Duration (y)
    1-1084/741.12 (0.70-1.80)55/531.74 (0.95-3.16)
    11-20130/1130.96 (0.63-1.48)47/441.56 (0.80-3.03)
    21-30153/1771.24 (0.80-1.91)64/651.76 (0.90-3.43)
    31-40168/2041.22 (0.78-1.91)66/711.80 (0.85-3.80)
    >40212/2601.18 (0.71-1.96)54/923.17 (1.40-7.19)
    Ptrend§0.390.06
Time since smoking commenced (y)
    1-3049/471.18 (0.61-2.32)52/511.54 (0.75-3.13)
    31-40119/1451.14 (0.70-1.87)89/931.35 (0.74-2.46)
    41-50284/2921.07 (0.71-1.59)90/1031.23 (0.68-2.24)
    >50297/3441.05 (0.68-1.62)55/791.64 (0.82-3.30)
    Ptrend§0.860.32
  • * Number of controls/number of cases. The analysis of smoking intensity and duration gave rise to missing values for this variable in three male and one female subjects. These case-control pairs were excluded from the analysis.

  • OR and 95% CI are estimated by conditional logistic regression, matched on age and ethnicity, compared with never smokers, and adjusted for family history of colorectal cancer, body mass index 5 y ago (males only), lifetime hours in recreational activities (males only) and egg, total calcium, non-starch polysaccharides from vegetables, methionine (females only), and alcohol (males only) intake. All levels of the smoking characteristics are compared against never smokers. Separate models were run for intensity, duration, and time since smoking commenced.

  • Further adjusted for smoking status.

  • § Ptrend is estimated by fitting the medians of each category as a continuous variable in the multivariate model.

  • Further adjusted for smoking status and duration.