Table 1.

ORs and 95% CIs for the association of selective nondietary variables with the risk of ovarian cancer, Hawaii and Los Angeles, 1993-1999

VariableCases (n = 558), n (%)Controls (n = 607), n (%)OR* (95% CI)P for trend
Center
    Hawaii200 (35.8)283 (46.6)
    Los Angeles358 (64.2)324 (53.4)
Age (y)
    <45130 (23.3)149 (24.5)
    45-54160 (28.7)171 (28.2)
    55-64117 (21.0)98 (16.1)
    ≥65151 (27.1)189 (31.1)
Ethnicity
    Caucasian258 (46.2)266 (43.8)
    Asian206 (37.2)254 (41.8)
    Other94 (16.6)87 (14.3)
Education (y)
    <13189 (33.9)163 (26.8)1§ (—)
    13-14194 (34.8)214 (35.2)0.75 (0.55-1.02)
    15113 (20.2)150 (24.7)0.63 (0.41-0.96)
    ≥1662 (11.1)80 (13.2)0.59 (0.41-0.83)0.002
No. full term pregnancies
    0159 (28.5)107 (17.6)1§ (—)
    177 (13.8)90 (14.8)0.60 (0.41-0.89)
    2139 (24.9)179 (29.5)0.60 (0.42-0.84)
    ≥3183 (32.8)231 (38.1)0.53 (0.37-0.75)0.002
Oral contraceptive pill use (y)
    0320 (57.3)269 (44.3)1§ (—)
    0.1-1.893 (16.7)99 (16.3)0.74 (0.51-1.08)
    1.9-5.382 (14.7)110 (18.1)0.61 (0.42-0.88)
    ≥5.463 (11.3)129 (21.2)0.36 (0.24-0.54)<0.0001
History of tubal ligation
    No490 (87.8)487 (80.2)1§ (—)
    Yes68 (12.2)120 (19.8)0.70 (0.50-0.99)
Menopausal status
    Premenopausal217 (38.9)256 (42.2)1§ (—)
    Postmenopausal341 (61.1)351 (57.8)1.36 (0.93-1.99)
Drinking status
    Nondrinker354 (63.6)369 (60.8)1§ (—)
    Drinker203 (36.4)238 (39.2)0.88 (0.67-1.66)
    Former drinker104 (18.6)95 (15.6)1.16 (0.82-1.64)
    Current drinker99 (17.7)143 (23.6)0.69 (0.50-0.96)
Smoking status
    Nonsmoker351 (62.9)367 (60.5)1§ (—)
    Smoker207 (37.1)240 (39.5)0.92 (0.72-1.19)
    Former smoker143 (18.6)161 (15.6)0.96 (0.72-1.27)
    Current smoker64 (17.7)79 (23.6)0.86 (0.59-1.25)
Vitamin/mineral supplement use
    None210 (37.6)151 (24.9)
    Multivitamin279 (50.0)387 (63.8)
    Vitamin A or β-carotene68 (12.2)79 (13.0)
    Vitamin C180 (32.2)212 (34.9)
    Vitamin E151 (27.1)193 (31.8)
  • * After adjustment by unconditional multiple logistic regression for age, ethnicity, study center, education, oral contraceptive pill use, parity, and tubal ligation (where appropriate).

  • Based on the likelihood ratio test comparing models with and without a trend variable that was assigned the median for the categories.

  • Cases and controls were frequency matched on these variables.

  • § Reference category.