Table 4. Direct and Mediated Effects per 100 000 person-years of Educational Level on Breast Cancer for each Mediator derived from Multinominal Logistic Regression Parameter Estimates and the Additive Hazards Model, Copenhagen, Denmark, 1981-2001.
| Potential mediatorsa | Educational level | Direct effectb (additional cases per 100 000 py) | 95% CI | Mediated effectb (additional cases per 100 000 py) | 95% CI | Mediated proportion | 95% CI |
|---|---|---|---|---|---|---|---|
| BMI | Low -> Medium | 66 | 24, 108 | 0 | -4, 4 | 0% | -7%, 6% |
| Low -> High | 65 | 13, 118 | 1 | -6, 8 | 1% | -24%, 14% | |
| Alcohol consumption | Low -> Medium | 59 | 16, 101 | 10 | 5, 15 | 15% | -1%, 28% |
| Low -> High | 55 | 3, 107 | 19 | 11, 27 | 26% | 14%, 69% | |
| Parity | Low -> Medium | 61 | 19, 103 | 10 | 6, 14 | 14% | 8%, 36% |
| Low -> High | 60 | 9, 111 | 14 | 8, 19 | 19% | 10%, 45% | |
| Age at first birthc | Low -> Medium | 54 | 7, 102 | 7 | 1, 14 | 12% | 0%, 21% |
| Low -> High | 39 | -20, 99 | 19 | 4, 33 | 32% | 10%, 257% | |
| HT used | Low -> Medium | 67 | 24, 110 | 5 | 4, 6 | 7% | 4%, 19% |
| Low -> High | 63 | 8, 119 | 7 | 6, 8 | 10% | 6%, 18% |
Abbreviations: BMI, body mass index; CI, confidence interval; HT, hormone therapy; PY, person-years.
a In categories of: BMI <25 kg/m2, 25-29.9 kg/m2, and 30+ kg/m2; Alcohol <1, 1-7, and >7 drinks/week; Physical activity <2 hours of light physical activity; 2-4 hours of light physical activity, and >4 hours of light activity/> 2 hours of vigorous activity per week; Parity 0, 1, 2, 3+; Age at first birth <25, 25-29, 30+ yrs, and nulliparous. HT use ‘yes’ vs. ‘no’.
b Adjusted for age and study.
c Excluding participants with missing information on age at first birth, N=3475.
d Excluding participants with missing information on HT use, N=2641.