Table 4.
Select reproductive and demographic factors associated with terminal duct lobular unit counts among postmenopausal women (n = 551)
| Characteristic | 0 THLUs | ≥1 THLUs | Poisson regression | |
|---|---|---|---|---|
| N | N | RR (95% CI)* | P | |
| Age, y | ||||
| <50 | 30 | 50 | 1.00 (reference) | .27† |
| 50–59 | 99 | 157 | 0.82 (0.73 to 0.92) | |
| 60–75 | 100 | 115 | 0.91 (0.79 to 1.05) | |
| Race/ethnicity | ||||
| Non-Hispanic Caucasian | 179 | 226 | 1.00 (reference) | .01‡ |
| African American | 34 | 77 | 0.95 (0.85 to 1.05) | |
| Hispanic | 10 | 11 | 0.97 (0.77 to 1.21) | |
| Other/missing | 6 | 8 | 0.58 (0.41 to 0.82) | |
| Education | ||||
| High school/GED or less | 29 | 61 | 1.00 (reference) | .004‡ |
| Vocational/tech or associates degree | 45 | 59 | 0.83 (0.72 to 0.95) | |
| College degree | 58 | 84 | 1.00 (0.88 to 1.14) | |
| Graduate/professional degree | 74 | 89 | 1.00 (0.88 to 1.14) | |
| Unknown/Missing | 23 | 29 | 0.81 (0.67 to 0.97) | |
| Age at menarche, y | ||||
| ≤12 | 118 | 149 | 1.00 (reference) | .006† |
| 13 | 56 | 92 | 0.82 (0.74 to 0.91) | |
| ≥14 | 55 | 81 | 0.89 (0.8 to 0.99) | |
| Number of live births | ||||
| Nulliparous | 52 | 53 | 0.67 (0.56 to 0.79) | .006ठ|
| 1 | 44 | 45 | 1.00 (reference) | |
| 2 | 75 | 130 | 1.09 (0.96 to 1.25) | |
| ≥3 | 58 | 94 | 1.20 (1.04 to 1.37) | .007†ǁ |
| Breastfeeding | ||||
| Never | 60 | 85 | 1.00 (reference) | <.001‡ |
| Ever | 117 | 184 | 1.24 (1.11 to 1.38) | |
| Bilateral oophorectomy | ||||
| No | 167 | 258 | 1.00 (reference) | .004‡ |
| Yes | 62 | 64 | 1.26 (1.08 to 1.47) | |
| Menopausal hormone therapy | ||||
| Nonuser | 111 | 175 | 1.00 (reference) | <.001‡ |
| Current user | 37 | 48 | 0.93 (0.81 to 1.06) | |
| Former user | 81 | 99 | 0.80 (0.71 to 0.89) | |
| Years since natural menopause | ||||
| <5 | 43 | 79 | 1.00 (reference) | <.001‡ |
| 5–10 | 39 | 52 | 0.90 (0.79 to 1.02) | |
| 11–15 | 30 | 44 | 0.98 (0.84 to 1.16) | |
| >15 | 30 | 27 | 0.70 (0.57 to 0.85) | |
| Gynecologic surgery | 87 | 120 | 0.79 (0.68 to 0.92) | |
| Number of first degree relatives with breast cancer | ||||
| 0 | 171 | 241 | 1.00 (reference) | <.001† |
| 1 | 44 | 61 | 1.26 (1.14 to 1.4) | |
| ≥2 | 14 | 20 | 1.25 (1.06 to 1.48) | |
| Fat percentage, % | ||||
| 0–75 | 15 | 121 | 1.00 (reference) | <.001† |
| 76–100 | 214 | 201 | 0.52 (0.48 to 0.57) | |
* Relative risk and 95% confidence intervals were obtained from zero-inflated Poisson regression models adjusted for age and all factors that showed an association with terminal duct lobular unit (TDLU) counts at P < .05. Poisson model was fit to both 0 and 1+ TDLU counts so that the groups jointly equaled to total subjects, and covariates were selected for inclusion in multivariable models if their trend or Wald test or any category was associated at P < .05. CI = confidence interval; RR = relative risk; TDLU = terminal duct lobular unit.
† Two-sided P value for trend test.
‡ Two-sided P value for Wald type 3 test.
§ Two-sided P value calculated for parous and nulliparous women.
ǁ P value for trend among parous women only.