Table 3.

Predictive factors for the incidence model of breast cancer are shown

NameMarginal distribution (source)Meta-analysis estimated relative hazard and CIMeta-analysis reference (if available and applicable)
Race-age (y)California censusaSee Supplementary Table S1 for matrix of risk estimates.N/A
Age at first birth (years, all among parous women)NHANESaCategoryRR(140)
<20 y1.00
20–241.04 (0.91–1.18)
25–291.17 (1.02–1.34)
30–341.30 (1.11–1.51)
35+1.40 (1.15–1.70)
Age at menarche (y)NHANESaAgeRR(139)
<111.19 (1.13–1.25)
111.09 (1.06–1.12)
121.07 (1.05–1.09)
131.00 (0.98–1.02)
140.98 (0.96–1.00)
150.92 (0.89–0.95)
≥160.82 (0.79–0.85)
Age at menopause (y)NHANESaAgeRR(139)
<400.67 (0.62–0.73)
40–440.73 (0.70–0.77)
45–490.86 (0.84–0.89)
50–541.00 (0.98–1.02)
≥551.12 (1.07–1.17)
Alcohol use (g/day)NHANESag/dRR (SE)(141)
01.00 (0.015)
<51.01 (0.020)
5–141.01 (0.023)
15–241.19 (0.048)
25–341.22 (0.056)
35–441.18 (0.093)
≥451.49 (0.110)
OR increase in relative risk per 10 g/day: 7.1% (1.3%)
BRCAb0.1%5.0 (4–6) estimated
Race/ethnicityCalifornia censusN/A
HTNHANESaCurrent combined vs. never use: 2.14 (2.04–2.24)(142)
Current estrogen-only vs. never use: 1.32 (1.24–1.39)
ObesityNHANESaRisk ratio per 5 U of BMI for postmenopausal breast(143)
1.12 (1.08–1.16)
Tobacco useNHANESaRelative risk of ever vs. never smokers, for those who went through natural menopause at:(141)
Age <45 y
RR = 1.11 SE = 0.15
Age 45–49
RR = 0.98 SE = 0.08
Age ≥50 years
RR = 1.12 SE = 0.06
  • aRace/ethnic group–specific marginal distributions were derived from NHANES data; see Table 2 for details.

  • bBRCA was assumed to have a population frequency of 0.1% in all race/ethnic groups independent of all other risk factors.