385-S
REPRODUCTIVE FACTORS AND BREAST CANCER RISK
AMONG OLDER WOMEN. *S Shantakumar, M B Terry, S L
Teitelbaum, J A Britton, A I Neugut, and M D Gammon (UNC-
Chapel Hill, Chapel Hill, NC 27599)
Reproductive factors have been shown to affect pre- and postmenopausal
breast cancer risk differently, but whether there are additional age-specific
differences among menopausal women as they age has not been clarified.
We analyzed data from a large population-based case-control study that
included 1508 breast cancer cases and 1556 controls, aged 20–98 years,
who completed an in-home interviewer-administered questionnaire. The
following subgroups were created to examine if the associations between
reproductive factors and breast cancer risk varied by age- and menopausal-
status: premenopausal (n ¼ 968), postmenopausal <65 years (n ¼ 1045),
postmenopausal 65 years (n ¼ 958). Among postmenopausal women
65 years, ever having breastfed decreased risk (OR ¼ 0.69, 95% CI ¼
0.48, 0.99), and a strong dose-response relationship was observed for longer
durations of breastfeeding (p trend ¼ 0.02), with the most pronounced
protective effect observed for 14 months of breastfeeding (OR ¼ 0.40,
95% CI ¼ 0.19,0.81). Older age at last birth increased risk in this group
(OR ¼ 2.14, 95% CI ¼ 1.04,4.43 for 35 years vs. <28 years), while late
age at menarche and surgical menopause decreased risk. Odds ratios for
multiparity and older age at first birth (AFB) were close to the null. Factors
associated with breast cancer risk among premenopausal women and post-
menopausal women <65 years were multiparity, which decreased risk, and
older AFB, which increased risk. Our findings suggest that the effects of
many well-known reproductive factors for breast cancer attenuate with
older age. Moreover, breastfeeding, one of the few modifiable risk factors,
was an important factor in decreasing breast cancer risk among older parous
postmenopausal women.
386
OPTICAL REFLECTANCE SPECTROSCOPY FOR
PROSPECTIVE STUDIES ON BREAST CANCER RISK IN
ADOLESCENT GIRLS. *S Dick and L Lilge (Ontario Cancer
Institute, Toronto, ON M5G 2M9)
Prospective studies on diet and lifestyle effects on future breast cancer risk
in young girls typically rely on establishing correlations between risk fac-
tors and serum sex and growth hormones. Quantitative physical information
on developing breast tissue is limited to breast density derived from DEXA
at a cost of ionizing radiation exposure. Near-infrared Optical Reflectance
Spectroscopy (ORS) uses light to access structural and metabolic bulk
tissue properties, providing quantitative physical and functional informa-
tion. Here we want to evaluate the ability of ORS to detect spectral differ-
ences between different Tanner stages of breast development indicative of
changes in chromophore concentration and distribution. Measurements
were obtained from 7 adolescent girls ages 10–14 years. Spectrally resolved
reflected light from 2 radial distances, 1.5 and 3.0cm, providing different
sample depth, is detected. Principle Component Analysis (PCA) was used
to reduce the spectral data and generate individual scores, representing
physiologic and morphologic tissue information. Four principle compo-
nents were identified accounting for 99.92% of the spectral variance. Radial
distances and hence tissue depth, had an impact on ORS parameters t1, t2
and t4 (all at p < 0.05, 2-tail independent t-test). Pearson correlation anal-
ysis indicated a positive correlation between t3 (r(110) ¼0.462, p <
0.001) and increasing breast Tanner stages; t4 (r(110) ¼ 0.240, p <
0.011) showed a reverse relationship. r values were strengthened when
adjusted for menarche status and radial distances. ORS may provide an
alternative tool to DEXA to identify morphologic and physiologic changes
to the breast tissue resulting from an intervention during the adolescent
years.
387
RISK OF IN SITU BREAST CANCER AND
MAMMOGRAPHIC DENSITY: THE MULTIETHNIC
COHORT. *J Gill, G Maskarinec, I Pagano, L N Kolonel
(Cancer Research Center of Hawaii, Honolulu, HI, 96813)
Background: It is well established that women with high mammographic
density are at greater risk of breast cancer than women with low breast
density. However, little research has been done on mammographic density
and breast carcinoma in situ (BCIS), which is thought to be a precursor
lesion to some invasive breast cancers. Methods: We conducted a nested
case-control study within the Multiethnic Cohort and compared the mam-
mographic densities of 482 invasive breast cancer cases and 125 BCIS cases
to 667 cancer free controls. A reader blinded to disease status performed
computer-assisted density assessment. For women with more than one
mammogram, mean density values were computed. Polytomous logistic
regression models were used to compute adjusted odds ratios (ORs) and
95% confidence intervals (CIs) for two measurements of mammographic
density: percent density and dense area. Results: Mammographic density
was significantly associated with invasive breast cancer and BCIS. For the
highest category of percent breast density ( 50%) as compared to the
lowest (< 10%), the OR was 3.58, 95% CI (2.26–5.66) for invasive breast
cancer and 3.06, 95% CI (1.50–6.24) for BCIS. Similarly, for the highest
category of dense area ( 45 cm
2
) as compared to the lowest (< 15 cm
2
),
the OR was 2.92, 95% CI (2.01–4.25) for invasive breast cancer and 2.86,
95% CI (1.57–5.20) for BCIS. Trend tests were significant for invasive
breast cancer (P
trend
<0.0001) and BCIS (P
trend
< 0.001) for both percent
density and dense area. Conclusion: The similar strengths of association for
mammographic density with BCIS and invasive breast cancer suggest
a common etiology for both cancers.
388-S
TIME TRENDS IN BREAST CANCER IN THE U.S. SEER
SYSTEM 1975–2002. *J Y Song, Y Y Han, and D L Davis
(University of Pittsburgh, Department of Epidemiology,
Pittsburgh, PA15260)
This paper assesses trends in recent time, period, and birth cohorts for breast
cancers in women. We used data from the SEER 9 Registry to perform age-
period-cohort analyses of breast cancer incidence rates among black and
white women in the U.S. from 1975 to 2002. We also used the SEER*Stat
software to fit Joinpoint models to summarize time trends. For 20–49 year
olds, black and white women had similar incidence rates; however, whites
had higher incidence rate than blacks in women aged 50–84. For whites,
breast cancer shows steeper slopes in more recent years, with no tapering
off at menopause. For blacks, similar patterns are evident, but incidence is
about 30% less at the older age group. Estimates of the long-term increase
in breast cancer rates over time are given by the average annual change:
1.1% for whites and 1.0% for blacks. These simple linear time trends alone
do not adequately capture the temporal patterns in black women. Breast
cancer incidences in white and black women deviated from linearity due to
age, time period, and cohort effect. Regarding generational risk, the chances
of getting breast cancer in new generations are about 30% higher than
previous generations. Mammographic screening only became widespread
after 1994, and it was used more by whites in their fifties, than by those in
their sixties and older. This may account for patterns in whites after 1994,
but cannot explain patterns in blacks. Additional research investigating the
differences in ethnos is needed to determine whether preventable factors
account for some of these patterns and to anticipate demands for health care
delivery.
2006 Congress of Epidemiology Abstracts S97
Am J Epidemiol 2006;163(Suppl):S1–S258 * ¼ Presenter; S ¼ The work was completed while the presenter was a student
Downloaded from https://academic.oup.com/aje/article/163/suppl_11/S97/115160 by guest on 20 August 2022