Table 3.

Risk estimates from case-control and family studies

VariablesCase-control study (ncase=325, ncontrol=327), OR* (95% CI)Family study
Familial cases to family controls (ncase=502, ncontrol=1,944)
Familial cases to population controls (ncase=502, ncontrol=327)
OR (95% CI)OR (95% CI)
Duration of smoking (y)
    Never1 (reference)1 (reference)1 (reference)
    <251.20 (0.80-2.00)0.88 (0.63-1.22)0.74 (0.49-1.11)
    ≥251.70 (1.09-2.74)0.75 (0.53-1.07)0.60 (0.39-0.93)
Duration of betel nut use (y)
    Never1 (reference)1 (reference)1 (reference)
    <201.08 (0.57-2.07)0.68 (0.42-1.10)1.22 (0.65-2.30)
    ≥201.37 (0.61-3.10)1.36 (0.81-2.27)1.83 (1.08-4.67)
Cumulative wood exposure
    None1 (reference)1 (reference)1 (reference)
    <251.20 (0.58-2.50)0.49 (0.20-1.19)0.32 (0.13-0.77)
    ≥252.40 (1.20-5.10)1.90 (0.88-4.13)1.74 (0.69-4.39)
Cumulative formaldehyde exposure
    None1 (reference)1 (reference)1 (reference)
    <251.30 (0.70-2.40)1.03 (0.60-1.76)1.30 (0.70-2.39)
    ≥251.50 (0.88-2.70)1.31 (0.87-1.97)4.29 (2.45-7.51)
Guangdong salted fish consumption (age <10)
    Never1 (reference)1 (reference)1 (reference)
    Ever1.40 (0.47-4.50)0.92 (0.54-1.57)4.73 (1.75-12.78)
RSA§
    c1-c11 (reference)1 (reference)1 (reference)
    c1-c20.83 (0.60-1.20)1.44 (0.82-2.54)1.34 (0.84-2.14)
    c2-c22.60 (1.20-5.74)0.68 (0.16-2.86)1.05 (0.27-4.12)
  • * ORs obtained from the case-control study for duration of smoking before diagnosis, cumulative wood and formaldehyde exposure, Guangdong salted fish consumption before age 10, and RSA allele were previously published (9, 18, 20, 21). OR for the duration of betel nut use before diagnosis was calculated using data collected in the same case-control study using unconditional logistic regression. ORs for smoking, betel nut use, Guangdong salted fish consumption, and RSA allele carrier were adjusted for age and sex. ORs for wood and formaldehyde exposure were adjusted for age, sex, education, and ethnicity.

  • All ORs were calculated using conditional logistic regression and were adjusted for age and sex; Guangdong salted fish consumption was presented as ever/never to compare with that of the case-control study.

  • All ORs were calculated using unconditional logistic regression and were adjusted for age and sex.

  • § To compare with the published result from the case-control study, RSA c1-c2 and c2-c2 genotype were analyzed separately; however, because only three cases in the family study had c2-c2, the risk estimate in this category from the conditional logistic regression might not be accurate.