Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Determinants of mammographic density in African-American, non-Hispanic white and Hispanic white women before and after the diagnosis with breast cancer
(USC Thesis Other)
Determinants of mammographic density in African-American, non-Hispanic white and Hispanic white women before and after the diagnosis with breast cancer
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
DETERMINANTS OF MAMMOGRAPHIC DENSITY IN AFRICAN-
AMERICAN, NON-HISPANIC WHITE AND HISPANIC WHITE WOMEN
BEFORE AND AFTER THE DIAGNOSIS WITH BREAST CANCER.
Copyright 2005
by
Ashish Mark Ingle
A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(APPLIED BIOSTATISTICS/EPIDEMIOLOGY)
December 2005
Ashish Mark Ingle
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 1435084
INFORMATION TO USERS
The quality of this reproduction is dependent upon the quality of the copy
submitted. Broken or indistinct print, colored or poor quality illustrations and
photographs, print bleed-through, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
®
UMI
UMI Microform 1435084
Copyright 2006 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Acknowledgments
I would like to express my utmost gratitude to Drs. Frank Gilliland, Sue
Ingles and W James Gauderman for their guidance and support in preparing this
Master’s Thesis.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTENTS
Acknowledgements...............................................................................................................ii
List of tables.......................................................................................................................... iv
A bstract..................................................................................................................................vi
Introduction.............................................................................................................................1
Methods...................................................................................................................................4
Study setting and patients:...............................................................................................4
Recruitment and Eligibility:............................................................................................ 4
Data Collection:................................................................................................................ 6
M ammograms:..............................................................................................................6
Menopausal status.........................................................................................................8
Physical measurements:...............................................................................................9
Physical Activity:........................................................................................................10
Stage of Disease and Treatment:...............................................................................11
Oral contraceptive use and alcohol:......................................................................... 12
Other variables:........................................................................................................... 12
Statistical Analysis:.....................................................................................................12
Results................................................................................................................................... 14
Discussion............................................................................................................................ 52
Bibliography.........................................................................................................................57
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
iv
LIST OF TABLES
Table 1. Reasons patients were excluded from the HEAL study.................................... 6
Table 2. Demographics of Heal, Eating, Activity and lifestyle study participants with
mammographic data, by race.............................................................................................. 15
Table 3. Univariate mean pre-diagnostic percent mammographic density of African-
American breast cancer women stratified on menopausal status..................................20
Table 4. Adjusted least-squared mean pre-diagnostic percent mammographic density
of African-American breast cancer women stratified on menopausal status...............22
Table 5. Univariate mean post-diagnostic percent mammographic density of African-
American breast cancer women stratified on menopausal status..................................24
Table 6. Adjusted least-squared mean post-diagnostic percent mammographic
density of African-American breast cancer women stratified on menopausal status. 27
Table 7. Univariate mean pre-diagnostic percent mammographic density of Non-
Hispanic Whites breast cancer women stratified on menopausal status...................... 30
Table 8. Adjusted least-squared mean pre-diagnostic percent mammographic density
of Non-Hispanic Whites breast cancer women stratified on menopausal status.........32
Table 9. Univariate mean post-diagnostic percent mammographic density of Non-
Hispanic Whites breast cancer women stratified on menopausal status...................... 36
Table 10. Adjusted least-squared mean post-diagnostic percent mammographic
density of Non-Hispanic Whites breast cancer women stratified on menopausal
status...................................................................................................................................... 39
Table 11. Univariate mean pre-diagnostic percent mammographic density of
Hispanic-Whites breast cancer women stratified on menopausal status......................42
Table 12. Adjustedf least-squared mean pre-diagnostic percent mammographic
density of Hispanic-Whites breast cancer women stratified on menopausal status...44
Table 13. Univariate mean post-diagnostic percent mammographic density of
Hispanic Whites breast cancer women stratified on menopausal status...................... 46
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 14. Adjustedf least-squared mean post-diagnostic percent mammographic
density of Hispanic Whites breast cancer women stratified on menopausal status. ..49
Table 15. Regression models of pre-diagnostic and post-diagnostic percent
mammographic density across all races............................................................................52
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vi
ABSTRACT
Mammographic density has been associated with an increase in breast cancer
risk. This study concentrates on the understanding of the predictors of breast tissue
density before and after cancer diagnosis.
In this study, information on demographics, reproductive, mammographic,
and other prognostic factors were collected at three study centers for African-
American, Non-Hispanic White, and Hispanic White women with breast cancer.
Body mass index (BMI), age, and menopausal status were common
predictors of breast density across race. Age and BMI were inversely associated with
percent breast density, and mammographic density was statistically significantly
higher in pre-menopausal than post-menopausal women (p<0.05). These associations
were consistent for both pre-diagnostic and post-diagnostic breast tissue density.
Data from this ethnically diverse population suggest that age, menopausal
status, and BMI are important predictors for mammographic density.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
INTRODUCTION
Mammographic density has been recognized to have a strong association
with breast cancer. Women with dense breast have 4 to 5 folds increase in risk for
developing breast cancer (Brisson and others 1982; Boyd and others 1995; Oza and
others 1993; Boyd and others 1998; Byrne and others 1995). Light (radiolucent)
areas in a mammogram represent adipose tissue whereas the dark (radioopaque)
areas represent fibro-glandular tissue. The ratio of radioopaque area to total breast
area in a mammogram is one of the ways mammographic density is defined.
Although many studies have observed association of mammographic density
and breast cancer risk factors, these associations vary across racial groups. In Aluet,
Indian and Eskimo Alaskan native women, Roubidooux et al (2003) showed an
association of mammographic density with age at mammogram, parity, hormone
replacement therapy, hysterectomy, history of breast biopsy and race. Other studies
(Byrne and others 1995; Maskarinec and others 2001) found an association of
mammographic density with body mass index (BMI), age, menopausal status, parity,
and estrogen therapy in Japanese, Chinese, Caucasians and Filipino women. Body
mass index has been found in many studies to be negatively associated with percent
mammographic density (Lam and others 2000; Sala and others 1999 Maskarinec
2001). Gram et al (1999) found a weak inverse association of high-risk
mammographic pattern with moderate physical activity in Norwegian women, where
as Vachon and others (2000) failed to see any association between physical activity
and mammographic density in a population of Caucasian women. It has been shown
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
in earlier studies (Greendale and others 1999; Laya and others 1995; McNicholas and
others 1994; Lundstrom and others 1999; Persson and others 1997; Stems and others
2000; Stomper 1990) that there is an increase in mammographic density in
postmenopausal women receiving hormone replacement therapy (HRT). Vachon et
al (2000) found that high intake of alcohol is associated with increased
mammographic density in both pre-menopausal and post-menopausal women, where
as smoking is inversely associated with percent mammographic density in only pre
menopausal women. Gapstur et al (2003) also found an inverse association between
smoking and mammographic density in pre-menopausal Hispanic women but failed
to observe any significant association between smoking and breast density in post
menopausal women. Roubidoux et al (2003) did not find any association between
smoking and breast density in Alaska native women. All these factors may be related
to mammographic density through their effects on hormone levels.
Because mammographic density has a demonstrated strong relationship with
breast cancer and also with its risk factors, it may be a useful intermediate marker for
breast cancer development and prognosis. Thus, there is a need to further investigate
the determinants of mammographic density in women with breast cancer to better
understand its potential as an intermediate marker for prognosis.
Although the role of mammographic density in breast cancer etiology has
been studied in some populations, its role as a prognostic marker has not been
extensively investigated. After diagnosis there are many factors, such as, treatment
with radiation therapy, chemotherapy, Tamoxifen and changes in lifestyle, which
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
may have an effect on mammographic density. Irwin et al (2003) showed that there
is a change in the amount of physical activity after diagnosis of breast cancer as
compared to before diagnosis. These lifestyle changes may also affect
mammographic density. Examining the association between mammographic density
and prognostic variables by racial groups may shed light on factors that explain
racial variations in prognosis.
Most of the studies examining the determinants of mammographic density
have been done on Non-Hispanic White women. To our knowledge, there is only one
study that investigated the association of mammographic density with breast cancer
risk factors in Hispanic Whites (Gapstur and other 2003) and none in African-
Americans. African-Americans and Hispanic Whites have poorer prognosis as
compared to Non Hispanic Whites. In the present study we examine the associations
of mammographic density with pre- and post- diagnostic risk factors in Non-
Hispanic White, Hispanic White and African-American women to investigate
whether mammographic density can help us understand the difference in prognosis
across race.
Pre- and post-diagnostic mammograms were collected from the participants
of the Health, Eating, Activity, and Lifestyle (HEAL) Study to investigate the above.
The HEAL study is a multi-center population-based cohort of African-American,
Non-Hispanic White and Hispanic White women with breast cancer at University of
New Mexico (UNM), University of Southern California (USC) and Fred Hutchinson
Cancer Research Center (FHCRC), Washington.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
METHODS
Study setting and patients:
The HEAL study is a collaborative multi-center cohort study of modifiable
prognostic factors in women with breast cancer in New Mexico, Southern California,
and Western Washington. The HEAL study aims to determine whether body-weight,
physical activity, diet and treatment affected breast cancer prognosis. In addition,
pre- and post-diagnostic mammographic images were collected to investigate the
determinants of mammographic density in women diagnosed with breast cancer.
Recruitment and Eligibility:
The HEAL study included women with invasive and in situ breast cancer
who met the site-specific eligibility criterion. In New Mexico, subjects were eligible
if they were diagnosed with invasive or in situ breast cancer at ages 18 years or
above, were able to participate in an interview one year after the diagnosis, were
diagnosed between July 1996 and March 1999, and if they lived in Bernalillo, Santa
Fe, Sandoval, Valencia, or Taos counties in New Mexico at diagnosis.
In Washington, subjects were eligible for the study if they were diagnosed
with invasive or in situ breast cancer between the ages of 35 and65 years, diagnosed
between September 1997 and September 1998, were able for interview one year after
diagnosis, were not active in another protocol, and lived in King, Pierce or
Snohomish counties in Washington at diagnosis.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
Subjects for HEAL study at the University of Southern California (USC)
were eligible if they were African-Americans with invasive or in situ breast cancer
diagnosed between the ages of 35-64; diagnosed between May 1995 and May 1998;
able to participate in interview after diagnosis and lived in Los Angeles County at
diagnosis. These cases had participated in one of two population based case-control
studies at USC, the Contraceptive and Reproduction Experiences (CARE) study with
invasive breast cancer and a sister study, the INSITU study of women with in situ
breast cancer diagnosed during the same period.
There were 1036 eligible subjects in Washington, and 1037 eligible subjects
in New Mexico. Eight hundred and thirty four subjects in Washington, and 383
subjects in New Mexico were excluded from the study due to the following reasons:
1) doctor refusal (n=59), 2) patient refusal (n=437), 3) inability to locate physician or
subject (n=443) or 4) participating in other studies (n=278). There were 367 African-
American subjects from the CARE and INSITU studies at USC who met the
eligibility criteria for the HEAL study. Two hundred and two subjects from
Washington, 654 subjects from New Mexico and 367 subjects from Los Angeles
participated in the HEAL study. All subjects provided an informed consent. All the
participating centers had approvals from their respective Institutional Review
Boards.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
Table 1. Reasons patients were excluded from the HEAL study.
W ashington and USC
N ew M exico (n)
(n)
Eligible 2073 367
Reason excluded
Physician refusal 59
Patient refusal 437
Inability to locate physician o r patient 443
Participating in other studies 278
H EAL Study patients included 856 367
Data Collection:
Data were collected at approximately 6 and 30 months after diagnosis. In-
person interviews were conducted at Los Angeles and New Mexico at 6 months and
30 months post diagnosis, whereas at Washington, subjects were mailed the study
questionnaire, which they brought to their clinic visit at 6 and 30 months after
diagnosis.
Mammograms:
For the participants who signed medical records and mammogram release
forms, pre-diagnostic and post diagnostic cranio-caudal (cc) views of the breast
contralateral to the affected breast were collected and scanned. For some subjects
only one of these mammograms was available. Any pre-diagnostic images 4 months
to 3 years prior to diagnosis and 6 months to 3 years post diagnosis were scanned.
At USC and UNM, cranio-caudal views were then scanned using a
Cobrascan CX-812M Large Format 12-bit Xray Digitizer (Radiographic Digital
Imaging, Torrance, CA). ScanWizard 3.09 plug-in program to Adobe Photoshop 5.0
was used with 256 shades of gray, scaling of 164, 150 dpi of resolution and gamma
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7
of 1.7. The images were scanned and saved at 300% brightness-contrast exposure
(BCE). If the mammograms were dark, BCE was changed to 500% and if they were
light, BCE was changed to 100 %. These films were then saved in PICT format.
At FHCRC, flatbed Epson Expression 836XL (Epson America, Long Beach,
CA) with a transparency scan head was used to scan the mammograms. SilverFast
plug-in program to Adobe Photoshop 5.5 was used with 12-bit grayscale and 200 dpi
of resolution. The images were cropped to an image size of 736 x 970 pixels (width x
height). These images were then saved in both TIF and BMP format.
These digitized images from all the centers were then converted into BMP
format using Adobe Photoshop and sent to FHCRC, Washington in batches of
approximately 50 images for determining the percent density of the images. One pre
diagnostic and one post-diagnostic image per subject were placed in a batch in a
random order to reduce variability between readings. Priority for selecting images
for reading was based on the time from diagnosis. Mammograms done one to two
years prior to or post diagnosis were given first priority, mammograms done over
two years and up to three years prior to or post diagnosis were given second priority.
Pre-diagnostic mammograms that were collected 4 to 11 months prior to diagnosis
were given third priority, and fourth priority was given to the mammogram done at
the time of diagnosis. For post-diagnostic mammograms, third priority was given to
mammograms done 6 to 11 months post diagnosis and fourth priority was given to
mammograms done more than three years post diagnosis. In the instance that only
pre-diagnostic or post-diagnostic images were available, it was put in the batch with
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
out a corresponding post-diagnostic or pre-diagnostic image. For quality control over
time, two images were put in all batches but with different labels. Within batch
quality control was assured by randomly selecting 10 percent of the total subjects
and assigning them evenly across the batches, so that at least 2 subjects were in the
same batch. Their images were then replicated in the respective batch with a new
label.
Images were read using Cumulus 108 developed by University of Toronto
(Boyd et al., 1995), which uses a gray-scale threshold technique. To determine the
density, the technician first outlined the back edge of the breast excluding any
pectoral muscle. Two sliding scales based on pixel brightness were then used to
determine the total area and the dense area (brightest) by adjusting the scale to select
the pixel value that separated the edge of the breast from the background of the film
and by adjusting the scale to select the pixel value representing brightest area within
breast respectively. The software then calculated percent mammographic density by
dividing dense area by total area and multiplying it by 100.
Menopausal status
Menopausal history was obtained during both 6-month and 30-month post
diagnostic follow-up interviews. Menopausal history was obtained during both 6-
month and 30-month post diagnostic follow-up interviews. An algorithm accounting
for age, menstruation in the past year, hysterectomy status, any surgery to ovaries,
hormone replacement therapy, Tamoxifen and hormone data was used to determine
menopausal status of the women
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
Subjects were assigned post-menopausal status if they were over 54 years of
age and not menstruating in last year; if they were over 54 years of age with
hysterectomy and without at least one ovary; if they were ages 54 or less and without
at least one ovary and not menstruation in last year; or if they were ages 54 or less
with no hysterectomy and had at least one ovary with no menstmation in last year.
An unknown menopausal status was assigned to subjects if they were ages 54 or less
with hysterectomy and had at least one ovary; if they were ages 55 or over with no
hysterectomy, used HRT for at least a year and menstruating; or if they were ages 55
and over, menstruating, used HRT for less than a year and had at least one ovary
with no hysterectomy.
In case hysterectomy and menstrual status data was not available, hormone
and Tamoxifen data was used to assign menopausal status. Subjects not on
Tamoxifen were assigned a post-menopausal status if follicle-stimulating hormone
(FSH) was more than 20 mRJ/mL, and Estrone less than 30 pg/ml or Estrodiol less
than 20 pg/mL. Subjects not on Tamoxifen were assigned a pre-menopausal status if
FSH was less than or equal to 20 mRJ/mL, and Estrone was more than or equal to 30
pg/ml or Estrodiol was more than or equal to 20 pg/ml.
Physical measurements:
Physical measurements were taken at baseline in clinics in Washington and
New Mexico. In Los Angeles, physical measurements were conducted at 30-month
post diagnosis. Trained staff conducted measurements in a defined manner. Height
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
and weight were measured to nearest 0.1 cm and 0.1 kg, respectively. Weights at one
year prior to diagnosis, five year prior to diagnosis, at diagnosis and at different ages
were collected as a part of the interview. Body mass index was classified as normal
2 2
if BMI was less than 25 kg/m , pre-obese if BMI was between 25-30 kg/m and
obese if BMI was more than 30 kg/m2.
Physical Activity:
Six months and 30 months post-diagnostic physical activity data were
collected at all the sites using standardized interview-administered questionnaires
based on the Modifiable Activity Questionnaire developed by Kriska and colleagues
(1997). Questions were asked about the physical activities in which the respondent
may have participated. These activities included strenuous activities such as running,
aerobics, swimming, jogging; leisure sedentary activities such as watching TV,
sleeping, napping; sitting and non-sitting activities at the workplace and household
activities. Information was recorded as number of months over the year doing the
activity, number of times per month, and hours each time doing the activity a year
prior to diagnosis and a year prior to 30-month follow up. The questionnaire also
included information on changes in the physical activities after diagnosis with breast
cancer. Six-month post-diagnostic physical activity information in Los Angeles was
collected by questions on activities done over lifetime. For each activity, start and
stop ages, years, months per year, and hours per week were recorded. Activity in the
year prior to diagnosis was constructed using the lifetime history of physical activity.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
The number of hours per week of the physical activities was calculated at
both baseline and 24 month follow-up using frequency and duration of each activity
from participant interviews. According to Compendium o f physical activities by
Ainsworth et al (2000), activities were classified into light intensity, moderate
intensity and vigorous intensity exercise. Metabolic equivalent task (MET) hours,
which is defined as the ratio of the associated metabolic rates for a specific activity
divided by the resting metabolic rate, was used to classify the activities into light
intensity physical activity with METs less than 3, moderate intensity physical
activity with METs between 3 to 6 and vigorous intensity physical activity with
METs greater than 6 MET hours per week were then calculated by multiplying
respective METs of the activities to the hours they were done in a week. We also
calculated and used total MET hours spent per week doing light intensity, moderate
intensity and vigorous intensity exercise for our analysis.
Stage o f Disease and Treatment:
Both at 6-month and 30-month post-diagnosis follow up, information on
stage of disease, tumor characteristics and treatment was collected and verified from
multiple sources, namely, local Surveillance Epidemiology and End Results (SEER)
registries, information extracted from the hospital charts, and medical records.
Medical records were obtained from the physicians’ offices for the participants who
had signed a medical records release form. Medical records were then used to update
the information on chemotherapy, radiation therapy and hormone therapy agents and
the duration of each therapy. In addition to the above sources, 6-month and 30-month
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
post-diagnostic follow up interviews were also used to supplement the information
on chemotherapy, radiation therapy and hormone therapy. During the 30-month
follow-up interview the information on medication use was also gathered from the
pill bottles during the in-person interview.
Oral contraceptive use and alcohol:
At the 6-month post-diagnostic interview, subjects were asked about their
history of oral contraceptive use. For analysis, oral contraceptive use was categorized
as ever/never. History of alcohol consumption was also collected at both the 6-month
and the 30-month post-diagnostic interview and classified as ever/never.
Other variables:
Information on reproductive history, family history of breast cancer, smoking
history, hormone replacement therapy, medical history and selected demographics
were collected using the self-administered questionnaires in Washington and
interview-administered questionnaires in New Mexico and Los Angeles at baseline
and 24-month follow up.
Statistical Analysis:
The primary goal of our analysis was to determine whether the association of
the risk factors of breast cancer with percent mammographic density varies with
race. All statistical analysis was stratified on race and menopausal status as race and
menopausal status are strong determinants of mammographic density. Separate
analyses were performed on pre-diagnostic and post-diagnostic mammographic
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
density. We conducted descriptive analysis examining mean percent mammographic
density across levels of covariates.
Selection of covariates to obtain least square means was based on prior
studies or if any statistical significance was observed in the univariate analyses. For
Non-Hispanic Whites final models included adjustment for age at the image, parity,
age at first full term birth, age at diagnosis, body mass index, oral contraceptive use,
hormone replacement therapy and alcohol use whereas for Hispanic-Whites models
included adjustment for age at image, parity, body mass index, age at first full term
birth, alcohol use, age at menarche and oral contraceptive use. Multivariate analysis
for African-Americans was adjusted for age at image, body mass index, parity, age at
first full term birth, alcohol use, hormone replacement therapy, age at menarche, oral
contraceptive use and stage of cancer.
One-way analysis of variance and one-way analysis of covariance were used
to calculate the means and least-square means of percent mammographic density and
other physiological and demographic characteristics for the univariate and
multivariate analysis respectively. If the dependent variable was categorical chi-
square was used to observe statistically significant difference. Statistical Analysis
Software (SAS) version 8.1 was used for all the analysis. All statistical analyses were
two-sided and were performed at significance level of 0.05.
General linear model was also used to obtain the predictive models for pre
diagnostic and post-diagnostic percent mammographic density. Backward stepwise
selection method was used to obtain models with race as dummy variables using
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
0.15 as the significance level to enter and stay in the model. All two-way interactions
were also checked for statistical significance.
RESULTS
As shown in table 2, the three racial groups, Non-Hispanic Whites, Hispanic
Whites and African-Americans, differed in many demographic and physiologic
characteristics. African-Americans were statistically significantly heavier and
younger as compared to the other two racial groups. African-Americans were also
significantly younger at their first full term birth. Hispanic Whites were statistically-
significantly more likely t o have a first-degree family history of breast cancer as
compared to African-Americans and Non-Hispanic Whites. Non-Hispanic Whites
had higher incidence of alcohol use and hormonal replacement therapy as compared
to the other racial groups. Twenty-five subjects were excluded from analyses as they
belonged to American Indian (6), Asian/Pacific (14) and others racial groups (5).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 2 Demographics o f Heal, Eating, Activity and lifestyle study participants with mammographic data, by race.
Characteristics Non-Hispanic Whites Hispanic Whites African-Americans
Pre
diagnosis
N (%)
Post
diagnosis
N (%)
Pre
diagnosis
N (%)
Post
diagnosis
N (%)
Pre
diagnosis
N (%)
Post
diagnosis
N (%)
Pre
diagnosis
P value3
Post
diagnosis
P-valueb
N 454 431 93 92 103 172
Age at 24-month
interview' (yrs) 60.2 ±11.2 60.3 ±11.1 57.9 ±11.0 57.3 ±11.0 54.8 ± 6.9 53.8 ±7.6 <0.05 <0.05
Age at
mammogram'
(yrs)
57.1 ±11.2 60.3 ±11.3 54.8 ± 10.9 57.3 ± 10.9 51.3 ±6.9 53.6 ±7.6 <0.05 <0.05
Age at
Diagnosis' (yrs)
57.9 ±11.2 54.9 ±11.1 51.2 ±7.6 >0.05 >0.05
Age at first birth'
(yrs)
24.5 ± 5.2 24.4 ± 5.0 22.9 ±5.2 22.9 ±5.6 22.0 ±5.5 20.9 ±4.6 <0.05 <0.05
Age at Menarche
(yrs)
<12
12
13
>13
106 (23.3)
133 (29.3)
107 (23.6)
107 (23.6)
103 (23.9)
121 (28.1)
107 (24.8)
100 (23.2)
21 (22.6)
23 (24.7)
25 (26.9)
24 (25.8)
24 (26.1)
22 (23.9)
25 (27.2)
21 (22.8)
23 (22.3)
35 (33.9)
29 (28.2)
16 (15.5)
43 (25.0)
54 (31.4)
45 (26.2)
30 (17.4)
>0.05 >0.05
BMI' 25.9 ± 5.3 26.1 ±5.4 27.1 ±5.7 27.3 ± 5.6 31.1 ±7.1 30.9 ±7.6 <0.05 <0.05
Postmenopausal 311 (68.5) 339 (78.7) 60(64.5) 66 (71.7) 55 (53.4) 131 (76.2) >0.05 >0.05
Table continued..
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 2. Demographics o f Heal, Eating, Activity and lifestyle study participants by race.
Characteristics Non-Hispanic Whites Hispanic Whites African-Americans Pre Post
diagnosis
P-valueb
Pre-
diagnosis
Post
diagnosis
Pre
diagnosis
Post
diagnosis
Pre
diagnosis
Post
diagnosis
diagnosis
P value3
Parity
Nulliparous 80 (17.6) 68 (15.8) 15 (16.1) 15 (16.3) 8 ( 7.8) 18 (10.5) <0.05 <0.05
1 72 (15.9) 67 (15.5) 13 (13.9) 12 (13.0) 24 (23.3) 35 (20.3)
2 146 (32.2) 142 (32.9) 22 (23.7) 20 (21.7) 30 (29.1) 46 (26.7)
3 84 (18.5) 86 (19.9) 15 (16.1) 15 (16.3) 20 (19.4) 32 (18.6)
> 4 71 (15.6) 68 (15.8) 28 (30.1) 30 (32.6) 21 (20.4) 41 (23.8)
First degree
family history 116(25.6) 101 (23.4) 28(30.1) 25 (27.2) 13 (12.6) 20 (11.6) <0.05 <0.05
Ever Smoker 232 (51.1) 213 (49.4) 45 (48.4) 43 (46.7) 53 (51.5) 94 (54.7) >0.05 >0.05
Ever Alcohol 258 (56.8) 323 (74.9) 50 (53.8) 63 (68.5) 42 (40.8) 68 (39.5) <0.05 <0.05
Ever Oral
contraceptive 298 (65.6) 284 (65.9) 60 (64.5) 61 (66.3) 78 (75.7) 133 (77.3) >0.05 <0.05
Ever HRT1 269 (59.3) 254 (58.9) 50 (53.8) 51(55.4) 39 (37.9) 50 (29.1) <0.05 <0.05
Tamoxifen given
Yes 160(37.1) 27 (29.3) 69 (40.1) >0.05
Chemotherapy
given
Yes 109(25.3) 27 (29.3) 87 (50.6) <0.05
Radiation therapy given
Yes 280(64.9) 61 (66.3) 81 (47.1) <0.05
Table continued..
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 2. Demographics o f Heal, Eating, Activity and lifestyle study participants by race.
Characteristics Non-Hispanic Whites Hispanic Whites African-Americans Pre Post
diagnosis
P-valueb
Pre
diagnosis
Post
diagnosis
Pre
diagnosis
Post
diagnosis
Pre
diagnosis
Post
diagnosis
diagnosis
P value3
Stage
In Situ 111 (24.5) 100 (23.2) 20 (21.5) 19 (20.7) 23 (22.3) 33 (19.2) >0.05 >0.05
Localized 263 (57.9) 257 (59.6) 55 (59.1) 54 (58.7) 52 (50.5) 75 (43.6)
Regional 77 (16.9) 72 (16.7) 18 (19.4) 19 (20.7) 27 (26.2) 64 (37.2)
Distant 2 (0.4) 2 (0.5)
ERe Positive 254 (58.9) 48 (52.2) 88 (51.2) >0.05
PRf Positive 209 (48.5) 35 (38.0) 66 (38.4) >0.05
a ANOVA across race for pre-diagnostic characteristics
b ANOVA across race for pre-diagnostic characteristics
c Mean + S.D.
d Hormone Replacement therapy,
e Estrogen receptor status,
f Progesterone receptor status.
< 1
18
Of the 368 eligible African-American subjects, 106 were not interviewed due
to inability to locate, decision not to participate in the study, or death. Two hundred
and forty seven women out of the remaining 262 women gave the consent to collect
the mammograms. We obtained at least one image for 202 subjects. The reasons for
failure to obtain the images were closure of facility, no record of mammograms at
any of the facilities listed by the subject, and discarding of older images. We
obtained both pre- and post-diagnosis images for 129 subjects, only post-diagnostic
images for 54 and only pre-diagnosis images for 19 subjects.
Eighty-eight and 159 African-American subjects had interview and
mammographic density data for pre-diagnostic and post-diagnostic period,
respectively. Tables 3 and 4 present pre-diagnostic analyses and tables 5 and 6
present post-diagnostic univariate and multivariate analyses.
BMI was an important indicator of breast density in our sample population.
In the pre-diagnostic period, we found a statistically significant (p<0.05) inverse
association of BMI in pre-menopausal women (Table 4) that remained strongly
associated with percent mammographic density after adjustment for age at
mammogram, alcohol consumption, parity and other covariates. Post-diagnostic
percent mammographic density showed a statistically significant decrease with
increase in BMI in both pre-menopausal and post-menopausal women (Tables 5 and
6). We also observed a decrease in percent-mammographic density with increase in
MET hours per week of total physical activity among pre menopausal women in
adjusted analyses of post diagnostic mammographic density.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
Pre-menopausal African-American subjects also showed a statistically
significant (p<0.05) decrease in post diagnostic percent-mammographic density
when chemotherapy was given as a component of treatment regimen. There was a
marginal statistical significance (p=0.05) of chemotherapy with percent
mammographic density in post-menopausal subjects.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
Table 3. Univariate mean pre-diagnostic percent mammographic density o f African-Am erican breast
cancer women stratified on m enopausal status.
Prem enopausal W om en Postm enopausal W om en
N M ean SD b P-value N M ean SD b P-value
Age at Image
< 4 7 22 26.3 16.9 0.1 4 18.6 11.5 0.1
47.1-52 9 36.3 18.2 6 20.8 13.9
52.1-57 2 10.1 1.1 23 20.3 12.1
57.1-63 - - 22 12.7 11.4
M enarche Age
< 12 9 28.9 9.3 0.9 13 17.6 10.3 0.6
12 11 27.1 18.5 18 15.9 13.5
13 7 30.9 22.7 17 19.8 13.9
> 13 6 25.0 23.1 7 13.2 6.9
Parity
0 4 25.3 11.5 0.4 4 20.2 6.9 0.8
1 9 34.3 17.2 11 14.7 14.3
2 10 25.3 18.1 12 18.4 9.5
3 7 20.5 13.8 10 14.9 11.6
> 4 3 39.3 30.9 18 18.5 14.3
Age at first birth
< 19 8 27.4 19.4 0.9 20 17.1 10.8 0.9
19.1-21 4 28.4 25.3 13 18.2 13.4
21.1-25 6 29.7 22.6 10 14.4 11.9
25.1-41 11 28.5 15.6 8 17.5 17.5
Nulliparous 4 25.3 11.5 4 20.2 6.9
Fam ily historyd
No 29 28.6 17.8 0.6 47 16.9 12.7 0.5
Yes 4 24.1 19.2 6 20.4 10.3
B M Ia
< 2 4 .9 5 48.4 15.7 0.01 9 17.4 16.9 0.2
25-29.9 9 21.4 14.1 17 21.2 12.4
> 2 9 .9 18 25.9 16.9 29 14.7 10.1
1 m issing
Alcohol
Never 15 29.7 18.7 0.6 38 16.7 11.9 0.6
Ever 16 26.7 17.2 17 18.3 13.3
Oral contraceptive
Never 3 16.3 18.7 0.2 20 14.7 8.5 0.2
Ever 30 29.2 17.5 35 18.6 13.8
HRT°
Never 33 - - - 23 13.6 10.2 0.06
Ever 0
- -
32 19.8 13.0
Table continued..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
Table 3. Univariate m ean pre-diagnostic percent m ammographic density o f A frican-A m erican breast
cancer women stratified on menopausal status.
Prem enopausal W om en Postm enopausal W omen
N M ean SD b P-value N M ean SD b P-value
Smoke
Never 15 33.3 19.2 0.1 25 17.4 12.3 0.8
Ever 18 23.7 15.5 30 16.9 12.4
Light-intensity PA e
0
M ET hrs/week)
28 29.6 18.5 0.2 31 19.3 12.4 0.1
0.63-18.82 5 19.5 9.5 22 13.6 11.7
M oderate-intensity P
0
\ e (M E T hrs/week)
30 26.4 17.2 0.05 40 17.9 12.1 0.4
2.67-4.58 1 61.4 - 3 12.4 8.5
4.59-8.81 - 4 14.8 15.5
8.82-16.08 - 4 8.6 3.3
16.08-29 - 4 24.3 16.4
Vigorous-intensity Pj
0
1 m issing
A = (M E T hrs/week)
30 28.1 18.5 0.9 51 16.4 11.9 0.06
0.024-37.5 3 27.1 4.7 3 30.1 14.1
Total Physical Activi
0
;y (M E T h rs/w eek /
24 27.9 18.6 0.9 25 19.2 12.5 0.1
1-8.72 4 27.1 24.5 13 14.6 11.7
8.72-29 3 27.2 4.6 13 11.4 9.1
______________________3 m issing____________________________________ 4 m issing___________________
a Body M ass Index (kg/m '),
b Standard deviation,
c Horm one replacem ent therapy,
d First degree fam ily history,
e Physical Activity.
f. Total physical activity= light intensity + m oderate intensity + vigorous intensity physical activity.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
Table 4. Adjusted least-squared mean pre-diagnostic percent m am m ographic density o f African-
American breast cancer women stratified on m enopausal status.
Prem enopausal subjects Postm enopausal subjects
N Lsm eans P-value N Lsm eans P-value
Age at Im age (yrs)
< 4 7 22 26.8 0.9 4 20.1 0.3
47.1-52 9 35.2 6 21.3
52.1-57 2 10.9 23 21.1
57.1-63 - 22 11.4
M enarche Age (y/rs)
< 12 9 30.8 0.7 13 17.7 0.9
12 11 30.7 18 16.4
13 7 28.4 17 18.7
> 13 6 19.3 7 14.6
Parity
0 4 22.9 0.6 4 17.8 0.5
1 9 34.7 11 13.4
2 10 25.3 12 17.9
3 7 18.7 10 15.7
> 4 3 33.7 18 20.6
Age at first birth (yrs)
< 19 8 31.0 0.8 20 17.1 0.8
19.1-21 4 20.9 13 15.5
21.1-25 6 32.5 10 16.2
25.1-41 11 24.9 8 21.2
Nulliparous 4 22.9 4 17.8
B M P
< 2 4 .9 5 50.9 < 0.001 9 19.6 0.4
25-29.9 9 18.4 17 19.5
> 2 9 .9 18 26.7 29 15.0
1 m issing
Fam ily
historyd
No 29 28.3 0.9 47 16.7 0.3
Yes 4 27.0 6 21.9
Alcohol
Never 15 28.5 0.9 38 16.4 0.5
Ever 16 27.8 17 18.8
Oral contracepth/e
Never 3 17.9 0.4 20 15.6 0.5
Ever 30 29.2 35 18.1
HRT0
Never 33 - - 23 13.6 0.07
Ever 0 - 32 19.7
Table continued..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
Table 4. Adjusted least-squared mean pre-diagnostic percent m am m ographic density of African-
A m erican breast cancer women stratified on menopausal status.
Prem enopausal subjects Postm enopausal Subjects
N Lsm eans P-value N L sm eans P-value
Sm oke
N ever 15 34.9 0.07 25 16.6 0.8
Ever 18 22.9 30 17.6
L ight-intensity P Ae (M E T hrs/week)
0 28 29.8 0.3 31 18.9 0.2
0.63-18.82 5 19.5 22 14.4
M oderate-intensity PA e (M E T hrs/week)
0 30 26.5 0.1 40 17.6 0.2
2.67-4.58 1 61.1 3 10.9
4.59-8.81 - 4 11.5
8.82-16.08 - 4 11.4
16.08-29 - 4 28.6
1 m issing
V igorous-intensity PA e (M E T hrs/week)
0 30 29.3 0.4 51 16.6 0.3
0.024-37.5 3 17.5 3 26.1
Total Physical A ctivity (M E T hrs/w eek)h
0 24 29.4 0.8 25 18.6 0.2
1-8.72 4 23.8 13 15.5
8.72-29 3 22.2 13 11.7
____________________3 m issing____________________________________4 m issing________________________________
a B ody M ass Index (kg/m -),
c H orm one replacem ent therapy,
d First degree fam ily history,
e Physical Activity.
f A djusted for age at the image, body mass index, alcohol use, horm one replacem ent therapy, parity and age at
first full term birth.
g. A djusted for age at the image, body m ass index, alcohol use, parity and age at first full term birth.
h. Total physical activity= light intensity + m oderate intensity + vigorous intensity physical activity.
Adjusted analyses on post-menopausal African-American subjects also
showed a statistically significant (p<0.05) decrease in post-diagnostic percent
mammographic density in subjects who had alcoholic beverages in their lifespan as
compared to the ones who never had alcoholic beverages.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
Table 5. Univariate mean post-diagnostic percent m am m ographic density of A frican-Am erican breast
cancer women stratified on menopausal status.
Prem enopausal W omen Postm enopausal wom en
N M ean SD b P-value N M ean SDb P-value
Age at Im age (yrs)
38-40 6 12.9 11.1 0.2 - - -
40-47 17 23.1 12.1 18 14.9 10.8 0.3
47.1-53 5 28.1 23.8 28 16.9 11.9
53.1-60 - 44 13.9 12.3
60.1-67 - 41 11.5 9.5
Age at Diagnosis (yrs)
36-40 12 16.5 11.4 0.2 4 13.5 7.3 0.2
40-45 11 24.8 12.7 15 17.9 12.1
45.1-51 5 28.1 23.8 27 16.1 11.7
51.1-57.5 - - - 42 14.3 12.7
57.51-64 - - - 43 10.9 8.8
M enarche Age (i/rs)
< 12 7 15.9 7.6 0.2 34 14.6 9.9 0.5
12 7 16.7 14.2 41 11.8 9.8
13 10 29.6 14.7 33 15.6 13.0
> 13 4 21.8 21.5 23 14.5 12.7
Parity
0 4 23.9 17.9 0.5 14 15.0 12.9 0.06
1 5 17.8 13.7 26 13.9 10.1
2 9 17.2 12.5 32 18.2 12.4
3 7 24.8 17.7 23 9.7 10.1
> 4 3 32.9 13.7 36 12.5 10.1
Age at first birth (yrs)
12-18 9 18.9 17.0 0.6 42 12.1 10.1 0.2
18.1-20 4 32.9 3.9 23 11.4 10.8
20.1-23 4 19.0 18.3 25 14.4 12.0
23.1-38 7 19.7 12.6 27 17.9 11.2
N ulliparous 4 23.9 17.9 14 15.0 12.9
BM Ia
< 2 4 .9 7 33.1 12.8 0.02 27 18.0 10.9 0.008
25-29.9 8 24.1 17.5 43 15.7 12.1
> 2 9 .9 13 14.4 9.8 60 10.7 10.0
Smoke
Never 13 17.2 11.8 0.1 58 14.2 12.2 0.8
Ever 15 25.9 16.3 73 13.7 10.5
Table continued..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
Table 5. U nivariate mean post-diagnostic percent m am m ographic density o f African-American breast
cancer women stratified on menopausal status.
Prem enopausal wom en Postm enopausal W omen
N M ean SD b P-value N M ean SD b P-value
Oral C ontraceptive
N ever 3 28.4 9.3 0.4 35 12.9 11.0 0.5
E ver 25 21.0 15.3 96 14.3 11.3
Estrogen recepto r
Positive 15 19.0 13.7 0.5 68 14.7 11.8 0.3
Negative 6 24.5 20.4 35 12.4 10.5
7 m issing 28
m issing
Progesterone receptor
Positive 14 19.9 12.7 0.8 50 14.9 12.3 0.5
N egative 6 21.4 19.4 29 13.1 10.7
8 m issing 52
m issing
HRT0
N ever 27 20.8 14.1 0.06 85 13.8 11.3 0.9
Ever 1 49.5 - 46 14.2 11.3
Fam ily historyd
No 25 22.5 15.1 0.3 111 13.3 10.9 0.07
Yes 2 10.4 12.4 17 18.7 13.2
Tam oxifen use
No 19 22.2 14.2 0.8 75 13.9 11.5 0.9
Yes 9 20.9 16.9 54 13.8 11.1
2 missing
Chem otherapy
No 15 26.6 16.0 0.06 66 12.5 11.4 0.1
Yes 13 16.3 11.6 65 15.4 10.9
Radiation T herapy
No 13 22.0 13.2 0.9 71 14.9 11.9 0.3
Yes 15 21.7 16.6 60 12.7 10.3
Stage
In Situ 6 25.4 16.7 0.5 25 14.6 11.8 0.4
Localized 13 23.7 17.4 56 12.2 9.7
Regional 9 16.8 8.4 49 14.9 11.4
D istant - 1 m issing
Light-intensity PAe (M ET hrs/week)
0-26.4 6 25.2 10.2 0.06 46 15.7 10.9 0.4
26.41-53.4 10 28.7 18.5 40 13.8 11.2
53.4-174 12 14.4 10.1 41 12.4 11.6
4 m issing
Table continued..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
T able 5. Univariate mean post-diagnostic percent m am m ographic density of African-Am erican breast
cancer women stratified on menopausal status.
Prem enopausal wom en Postm enopausal W omen
N Mean SD P-value N M ean SD b P-value
M oderate and vigorous intensity PA eh (M E T
hrs/w eek)
0-6.1 7 22.9 18.3 0.1 44 11.9 9.3 0.1
6.2-17.3 10 28.9 14.6 39 14.1 10.9
17.3-55 10
1 m issing
14.9 11.2 39
9 m issing
17.2 13.3
V igorous-intensity PA e (M ET hrs/week)
0 13 26.9 16.2 0.5 93 13.9 10.7 0.2
0.13-1.33 5 21.3 11.6 8 9.6 10.7
1.34-3.25 6 16.9 16.0 8 9.1 6.6
3.26-10.38 1 8.4 - 9 17.1 15.8
10.39-23.6 3 14.7 7.9 9 20.7 14.2
Total Physical A ctivity1 1 (M ET hrs/week)
0-39.4 8 32.0 15.4 0.05 42 13.5 10.3 0.8
39.4-75.1 7 20.9 15.5 43 14.2 11.3
75.1-208.5 12
1 m issing
15.7 11.8 38
8 m issing
14.9 12.4
a Body M ass Index (kg/m -),
b Standard deviation,
c H orm one replacem ent therapy,
d First degree fam ily history,
e Physical Activity.
f. Total physical activity= light intensity + m oderate intensity + vigorous intensity physical activity.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
Table 6. Adjusted least-squared mean post-diagnostic percent m am m ographic density o f African-
American breast cancer women stratified on menopausal status.____________________________________
Prem enopausal w om en3 Postm enopausal wom en1 5
N Lsm ean P-value N Lsm ean P-value
Age at Im age (yrs)
38-40 6 -
40-47 17 20.4 0.2 18 4.9 0.3
47.1-53 5 28.2 28 12.5
53.1-60 - 44 14.6
60.1-67 - 41 17.0
Age at Diagnosis (yrs)
36-40 12 17.7 0.4 4 20.7 0.2
40-45 11 26.4 15 26.0
45.1-51 5 22.9 27 19.9
51.1-57.5 - 42 13.5
57.51-64 - 43 5.1
M enarche A ge (jtrs)
< 12 7 21.5 0.6 34 14.1 0.9
12 7 18.2 41 12.9
13 10 26.0 33 13.9
> 13 4 18.4 23 13.5
Parity
0 4 24.7 0.4 14 11.3 0.06
1 5 16.5 26 12.8
2 9 21.2 32 18.4
3 7 27.9 23 10.3
. > 4
3 25.4 36 13.3
Age at first birth (yrs)
12-18 9 19.2 0.7 42 13.1 0.4
18.1-20 4 23.8 23 10.9
20.1-23 4 19.6 25 14.5
23.1-38 7 26.8 27 16.0
Nulliparous 4 24.7 14 11.3
BM IC
< 2 4 .9 7 31.8 0.02 27 19.0 <0.001
25-29.9 8 23.6 43 15.4
> 2 9 .9 13 15.4 60 9.9
Smoke
Never 13 19.8 0.5 58 13.4 0.8
Ever 15 23.6 73 13.8
Alcohol
Never 16 22.1 0.9 79 15.3 0.02
Ever 12 21.5 52 11.0
Oral contracepti\'e
Never 3 25.6 0.5 35 13.1 0.8
Ever 25 21.4 96 13.8
Table continued..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
Table 6. Adjusted least-squared mean post-diagnostic percent mammographic density of African-American breast
cancer women stratified on menopausal status.
Prem enopausal w om en1 Postm enopausal w om enb
N Lsm ean P-value N Lsm ean P-value
Estrogen receptor
Positive 15 23.1 0.2 68 13.9 0.5
Negative 6
7 m issing
14.3 35
28 m issing
12.6
Progesterone receptor
Positive 14 20.7 0.9 50 13.2 0.6
Negative 6
8 m issing
19.6 29
52 m issing
14.5
HRTd
N ever 27 21.2 0.2 85 13.1 0.5
E ver 1 38.1 46 14.6
Fam ily history6
No 25 21.5 0.8 111 13.3 0.3
Yes 2 23.4 17 16.1
Tam oxifen use
No 19 23.2 0.4 75 12.9 0.5
Yes 9 18.9 54
2 m issing
14.3
Chem otherapy
No 15 26.8 0.02 66 11.2 0.05
Yes 13 16.1 65 16.0
Radiation Therap
y
No 13 22.5 0.8 71 14.4 0.4
Yes 15 21.3 60 12.7
Stage
In Situ 6 16.9 0.6 25 15.5 0.5
Localized 13 22.4 56 12.7
Regional 9 24.2 49 13.7
Distant - - 1 m issing
Light-intensity PA (M E T hrs/w eek)
0-26.4 6 28.4 0.06 46 14.9 0.6
26.41-53.4 10 27.1 40 12.6
53.4-174 12 14.1 41 13.3
M oderate and vigorous intensity PA* (M E T hrs/week)
0-6.1 7 26.1 0.07 44 11.7 0.2
6.2-17.3 10 28.6 39 14.4
17.3-55 10
1 m issing
12.2 39
9 m issing
16.1
Total Physical A ctivity (M ET hrs/week)
0-39.4 8 33.1 0.03 42 12.9 0.8
39.4-75.1 7 20.6 43 13.9
75.1-208.5 12
1 m issing
15.1 38
8 m issing
14.6
a Body mass index, age at image, age at m enarche, parity, age at first full term birth, oral contraceptive,
chem otherapy and light intensity physical activity.
b Body mass index, age at image, age at m enarche, parity, age at first full term birth, oral contraceptive, age at
diagnosis, alcohol use and stage,
c Body M ass Index (kg/m 2),
d Horm one replacem ent therapy,
e First degree fam ily history,
f Physical Activity.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
Of the 671 eligible Non-Hispanic Whites, interview data and mammographic
data were available for 483 subjects. Failure to acquire the data was due to non
participation of the subjects in follow-up interview, inability to locate, death, lack of
consents by the subjects to obtain mammograms from facilities, closure of facility, or
no record of mammograms at any facility listed. We obtained both pre- and post
diagnostic mammograms for 367 subjects. We were able to obtain only post
diagnostic mammograms for 44 subjects and for other 72 subjects we got only pre
diagnostic mammograms.
In Non-Hispanic White women, BMI was an important determinant of
mammographic density. Pre-diagnostic and post-diagnostic mammographic density
had a statistically significant (p< 0.05) inverse association with BMI in both pre- and
post-menopausal women (Tables 7 and 8). This statistically significant decrease in
percent mammographic density with increase in BMI remained significant in
multivariate analyses after adjustment for age at image, hormone replacement
therapy, parity and other covariates (Tables 9 and 10).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
Table 7. Univariate mean pre-diagnostic percent m am m ographic density o f Non-Hispanic W hites
breast cancer wom en stratified on menopausal status.
Prem enopausal women Postm enopausal wom en
N M ean S.dT6 P-value N M ean S.D. b P-value
A ge at Im age (yrs)
< 4 3 44 38.2 18.2 0.04
- . .
43.1-49 54 29.3 18.9 16 31.7 21.2 0.003
49.1-56 30 29.7 17.4 86 21.0 16.1
56.1-64 101 19.7 13.8
64-86 108 17.2 13.0
M enarche Age ()/rs)
< 12 29 26.3 17.6 0.2 74 18.2 12.9 0.2
12 40 33.3 19.9 88 19.3 14.4
13 31 34.9 17.6 73 18.9 14.7
> 13 28 35.1 18.8 75 23.0 17.5
1 m issing
Parity
0 31 31.8 20.8 0.5 50 22.3 15.5 0.3
1 25 34.2 21.1 43 20.5 14.9
2 55 34.0 16.7 84 20.7 15.3
3 15 24.7 16.4 68 19.3 15.2
> 4 2 37.1 28.6 65 16.4 12.9
1 m issing
Age at first birth (yrs)
< 2 0 16 28.5 15.6 0.2 72 21.3 15.7 0.4
21-22 12 37.4 24.9 51 16.5 13.1
22.1-25 15 26.5 15.0 59 20.4 14.0
25.1-29 25 29.1 17.9 51 18.0 14.6
> 2 9 29 39.3 16.3 28 19.9 17.4
N ulliparous 31 31.8 20.8 50 22.3 15.5
Fam ily historyd
No 92 34.5 18.5 0.1 221 20.2 15.6 0.6
Yes 33 28.4 18.5 81 19.1 13.0
3 m issing 9 m issing
B M P
< 2 0 15 53.4 16.3 <0.001 25 32.9 18.6 < 0.001
20.1-24.9 56 35.7 16.9 120 20.6 13.7
25-29.9 35 29.2 14.5 97 19.9 14.8
> 2 9 .9 20 12.5 10.3 66 13.3 12.5
2 m issing
Sm oke
Never 75 30.5 18.7 0.2 141 18.9 13.8 0.3
E ver 53 35.3 18.5 170 20.6 15.8
Table continued..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
Table 7. U nivariate mean pre-diagnostic percent m am m ographic density o f Non-Hispanic W hites
breast cancer women stratified on menopausal status.
Prem enopausal women Postm enopausa women
N Mean S .D .b P-value N M ean S .D .b P-value
Alcohol
Never 42 30.6 19.2 0.4 147 18.1 13.9 0.04
Ever 86 33.4 18.5 161
3 m issing
21.6 15.7
Oral contraceptix
Never
it
18 30.6 16.2 0.7 135 18.3 13.6 0.1
Ever 109 32.5 19.0 176 21.0 15.9
HRT°
1 m issing
Never 101 32.8 18.6 0.9 75 16.8 13.5 0.04
Ever 26 32.1 19.4 234 20.9 15.3
Light-intensity P
0-13.8
1 m issing
Ae (M ET hrs/week)
55 34.2 19.8 0.4 93 19.9 14.4 0.3
13.9-35 46 33.4 18.4 123 18.3 14.7
35.1-120 25 28.4 16.6 93 21.7 15.6
M oderate-intens
0-10.4
2 m issing
ty PA e (M ET hrs/week)
32 26.4 18.8 0.07
2 m issing
111 20.2 15.5 0.2
10.41-23.5 43 33.2 19.3 104 17.7 13.8
23.51-110 52 35.9 17.6 94 21.8 15.5
Vigorous-intensi
0
1 m issing
ty PA e (M ET hrs/week)
50 26.3 17.7 0.007
2 m issing
206 19.2 15.0 0.7
0.14-2.2 20 31.1 17.3 27 22.8 13.5
2.21-6.75 13 33.0 16.7 29 19.1 15.3
6.76-15.75 18 43.2 21.7 28 20.2 16.6
15.76-73.45 27 37.6 16.7 19 22.7 15.0
Total Physical A
0-33.2
:tivityf (M ET hrs/week)
41 28.8 18.8 0.2
2 m issing
104 20.4 15.7 0.5
33.2-60.42 37 33.5 19.3 107 18.4 13.5
60.42-199.5 48
2 missing
35.5 18.0 96
4 m issing
20.5 15.8
a Body M ass Index (kg/m -),
b Standard deviation,
c H orm one replacem ent therapy,
d First degree fam ily history,
e Physical Activity.
f. Total physical activity= light intensity + m oderate intensity + vigorous intensity exercise.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
T able 8. Adjusted least-squared mean pre-diagnostic percent mammographic density of N on-Hispanic
W hites breast cancer women stratified on menopausal status.
Prem enopausal w om en1 Postm enopausal w om en8
N Lsm ean P-value N Lsm ean P-value
Age at Im age (yrs)
< 4 3 44 39.8 0.6 - -
43.1-49 54 36.6 16 31.6 < 0.001
49.1-56 30 35.7 86 26.3
56.1-64 101 22.8
64-86 108 17.1
M enarche Age (5
< 12
/rs) |
29 28.9 0.4 74 18.8 0.2
12 40 35.8 88 18.4
13 31 31.8 73 19.4
> 13 28 33.0 75 22.9
Parity
0 31 36.2 0.5
1 m issing
50 24.9 0.5
1 25 38.5 43 21.7
2 55 37.7 84 21.2
3 15 30.8 68 22.7
> 4 2 50.5 65 18.2
A ge at first birth
< 2 0
(yrs) |
16 36.8 0.2
1 m issing
72 22.3 0.9
21-22 12 47.7 51 19.9
22.1-25 15 28.6 59 21.3
25.1-29 25 37.0 51 20.0
> 2 9 29 38.3 28 19.7
Nulliparous 31 36.2 50 24.9
Fam ily historyd
No 92 ^ 34.4 0.2 221 20.2 0.5
Y es 33
3 m issing
29.9 81
9 m issing
18.9
B M Ia
< 2 0 15 54.8 <0.001 25 32.9 <0.001
20.1-24.9 56 40.9 120 21.8
25-29.9 35 30.5 97 20.2
> 2 9 .9 20 16.7 66 14.6
Sm oke
2 m issing
N ever 75 31.4 0.3 141 19.3 0.5
E ver 53 34.6 170 20.3
Alcohol
N ever 42 32.8 0.9 147 19.3 0.5
Ever 86 32.7 161
3 m issing
20.3
Table continued..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
Table 8. Adjusted least-squared mean pre-diagnostic percent mammographic density o f Non-Hispanic
W hites breast cancer women stratified on m enopausal status.
Prem enopausal women* Postm enopausal women®
N Lsm ean P-value N Lsm ean P-value
Oral contraceptive
N ever 18 29.4 0.4 135 20.1 0.8
E ver 109 32.9 176 19.6
HRT°
N ever 101 33.1 0.6 75 18.4 0.3
Ever 26 31.3 234 20.2
1 m issing
Light-intensity PA e g (M ET hrs/week)
0-13.8 55 39.7 0.3 93 19.2 0.1
13.9-35 46 37.2 123 20.4
35.1-120 25 31.6 93 23.9
2 m issing
M oderate-intensity PA eg (M ET hrs/week) |
0-10.4 32 38.3 0.8 111 20.2 0.2
10.41-23.5 43 35.9 104 19.8
23.51-110 52 38.5 94 23.4
1 m issing 2 m issing
Vigorous-intensity PA e (M E T hrs/week) |
0 50 28.7 0.02 206 19.9 0.9
0.14-2.2 20 28.9 27 21.2
2.21-6.75 13 31.4 29 17.7
6.76-15.75 18 42.1 28 19.4
15.76-73.45 27 36.7 19 20.8
1 m issing
Total Physical Activity (M ET hrs/week) |
0-33.2 41 38.9 0.6 104 20.3 0.4
33.2-60.42 37 34.6 107 20.1
60.42-199.5 48 38.4 96 22.9
2 m issing 4 m issing
a Body M ass Index (k g /n r),
c H orm one replacem ent therapy,
d First degree fam ily history,
e Physical Activity.
f. Adjusted for age at the image, body mass index, alcohol use, horm one replacem ent therapy, parity, age at first
full term birth, light intensity physical activity and vigorous intensity physical activity.
g. Adjusted for age at the image, body mass index, alcohol use, horm one replacem ent therapy, parity and age at
first full term birth.
h. Total physical activity= light intensity + m oderate intensity + vigorous intensity exercise.
An inverse association of pre-diagnostic percent mammographic density was
also observed with the age of subject at the time of mammogram. There was a
statistically significant (p<0.05) decrease in the pre-diagnostic percent breast density
with increase in age of the woman at the time of mammogram in post-menopausal
Non-Hispanic White women in both univariate and multivariate analyses (tables 9
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
and 10). In pre-menopausal women, a statistically significant (p<0.05) inverse
association of pre-diagnostic percent breast density was seen with age at the time of
mammogram. We also observed a statistically significant (p<0.05) decrease in post
diagnostic percent density with increase in age at mammogram in pre-menopausal
women and a marginally significant (p=0.07) association in post-menopausal
women. Non-Hispanic White women also demonstrated a statistically significant
(p<0.05) decrease in post-diagnostic percent breast density with increase in the age at
diagnosis.
Post-menopausal Non-Hispanic White women who ever consumed alcoholic
beverages in their lifetime also had a statistically significant increase in percent
breast density as compared to the subjects who never consumed alcoholic beverages.
This association maintained its statistical significance in multivariate analysis for
post-diagnostic percent mammographic density (p<0.05).
Post-menopausal Non-Hispanic White women also had a statistically
significant increase in percent breast density in subjects who had alcoholic beverages
in their lifetime as compared to the ones who never had alcoholic beverages. This
association maintained its statistical significance in multivariate analysis for post
diagnostic percent mammographic density (p<0.05).
We also observed a statistically significant (p<0.05) increase in
mammographic density with increase in MET hours/week of vigorous intensity
physical activity in both pre- and post-menopausal women for the pre-diagnostic
period and only for the post-diagnostic period in pre-menopausal women. Pre
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
diagnostic mammographic density was also significantly (p<0.05) associated with
moderate intensity physical activity in post-menopausal women.
Post-menopausal Non-Hispanic White women also showed a statistically
significant (p<0.05) increase in pre-diagnostic breast density in women who were
receiving hormone replacement therapy and women who had a first-degree family
history of breast cancer as compared to women who had never received hormone
replacement therapy and who had no first-degree family history of breast cancer
respectively. A marginally significant (p=0.05) decrease in percent breast density
was also observed in pre-menopausal women who were given Tamoxifen as a part of
their treatment regimen.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
Table 9. Univariate mean post-diagnostic percent mam m ographic density o f Non-Hispanic W hites
breast cancer women stratified on menopausal status.
Prem enopausal women Postm enopausal wom en
N M ean S.D .b P-value N M ean S.D.b P-value
Age at Im age (yrs)
< 4 5 26 33.7 17.1 0.01 5 21.3 11.7 0.07
45.1-58 46 23.1 16.5 121 17.9 14.9
58.1-67 - - - 111 15.1 12.2
67.1-89 - - - 102 13.7 11.9
Age at Diagnosis (yrs)
35-45 40 32.1 18.3 0.01 11 16.0 11.5 0.006
45-50 25 21.9 13.9 36 23.2 16.9
50.1-56 7 15.8 14.3 87 16.1 12.2
56.1-65 105 14.9 13.3
65.1-87 100 13.7 12.1
M enarche Age (\/rs)
< 12 19 18.2 14.8 0.08 79 13.7 11.4 0.2
12 26 30.5 16.8 88 15.4 12.6
13 15 29.0 17.3 87 15.7 13.7
> 13 12 30.3 19.4 84 18.3 15.0
1 m issing
Parity
0 16 29.6 18.3 0.7 51 16.9 15.2 0.04
1 11 32.5 14.3 50 17.3 14.3
2 32 25.1 18.0 100 18.0 13.5
3 11 23.4 17.5 73 14.2 12.1
> 4 2 25.2 22.3 65 12.1 10.8
Age at first birth (yrs)
< 2 0 10 23.9 13.7 0.5 73 13.9 11.9 0.6
21-22 6 38.1 22.8 54 14.9 12.3
22.1-25 9 20.6 20.4 65 15.8 11.9
25.1-29 13 25.3 20.0 62 16.2 14.2
> 2 9 18 26.8 12.3 34 18.7 14.9
Nulliparous 16 29.6 18.3 51 16.9 15.2
Fam ily historyd
No 53 26.2 14.5 0.6 248 15.9 13.5 0.9
Yes 19 28.9 23.9 77 15.8 12.4
Table continued..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
Table 9. U nivariate mean post-diagnostic percent m am m ographic density o f Non-Hispanic W hites
breast cancer women stratified on menopausal status.
Premenopausal wom en Postm enopausal women
N M ean S.D .b P-value N M ean c n
b
cr
P-value
B M Ia
< 2 4 .9 43 34.1 16.5 < 0.001 155 18.8 14.1 <0.001
25-29.9 16 21.7 13.1 106 15.6 12.5
> 2 9 .9 11 7.8 7.2 72 9.7 10.4
Alcohol
Never 9 27.1 20.2 0.9 93 11.5 9.5 <0.001
E ver 63 26.9 17.1 246 17.4 14.1
Oral contraceptiv e
Never 10 24.7 12.1 0.6 134 14.8 12.1 0.3
Ever 62 27.3 18.1 204 16.3 13.9
1 m issing
Estrogen receptor
Positive 28 27.5 17.9 0.4 215 16.1 13.0 0.5
Negative 6 21.0 15.5 39 14.6 14.9
38 m issing 85 m issing
Progesterone receptor
Positive 23 26.0 18.9 0.8 175 16.3 13.6 0.5
Negative 7 23.6 15.7 75 14.9 12.9
42 m issing 89 m issing
HRT°
N ever 59 - - - 108 15.4 13.3 0.7
Ever 0 - - 230 16.0 13.3
13 m issing
Smoke
Never 39 26.9 19.7 0.9 169 15.2 12.2 0.4
Ever 33 26.9 14.4 169 16.4 14.3
Tam oxifen use
No 58 28.8 17.6 0.05 192 15.8 12.9 0.9
Yes 9 16.5 15.0 141 15.8
5 m issing 6 m issing
C hem otherapy
No 61 26.8 17.0 0.9 246 15.7 13.1 0.8
Yes 10 27.7 20.9 93 16.1 13.5
1 m issing
R adiation T herapy
No 27 28.7 19.5 0.5 118 15.8 11.7 0.9
Yes 44 25.8 16.3 221 15.8 14.1
1 m issing
Stage
In Situ 33 25.7 17.2 0.6 60 15.9 13.4 0.9
Localized 31 26.8 16.6 216 15.6 13.8
Regional 8 32.4 21.8 61 16.2 11.5
Distant - 2 16.6 12.7
Table continued..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
Table 9. U nivariate m ean post-diagnostic percent m am m ographic density o f Non-Hispanic W hites
breast cancer women stratified on m enopausal status.
Prem enopausal women Postm enopausal wom en
N M ean S.D .b P-value N M ean S.D.b P-value
L ight-intensity P/
0-20.4
\ e (M E T hrs/week)
24 26.8 17.7 0.9 112 16.5 14.2 0.5
20.41-38.4 28 25.8 16.6 108 16.2 12.6
38.4-123 18
2 m issing
26.1 15.9 115
4 m issing
14.8 12.8
M oderate-intensi
0-10.4
y PA e (M E T hrs/week)
18 22.4 16.9 0.09 118 13.7 13.5 0.03
10.41-25.9 27 24.3 17.3 108 18.2 14.2
25.91-110 26
1 m issing
32.8 17.1 109
4 m issing
15.8 11.6
Vigorous-intensit
0
y PA e (M E T hrs/week)
27 19.3 15.5 0.02 210 14.7 13.3 0.04
0.02-1.193 9 25.9 12.9 32 16.1 12.9
1.194-4.38 11 36.0 15.4 31 14.1 9.3
4.39-12.25 14 27.7 17.6 30 19.3 13.8
12.26-83.4 10 37.2 20.7 33 21.5 14.9
Total Physical Ac
0-39.7
1 m issing
:tivityh (M E T hrs/week)
23 24.3 17.3 0.3
3 m issing
112 14.7 12.9 0.5
39.8-67.2 20 22.9 12.6 117 16.6 15.1
67.3-238.1 27
2 m issing
30.3 17.3 105
5 m issing
16.1 11.2
a Body M ass Index (k g /n r),
b Standard deviation,
c H orm one replacem ent therapy,
d First degree fam ily history,
e Physical Activity.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
Table 10. Adjusted least-squared mean post-diagnostic percent m ammographic density o f Non-
Hispanic W hites breast cancer women stratified on menopausal status.
Prem enopausal w om en1 Postm enopausal women*
N Lsm ean P-value N Lsm ean P-value
Age at Im age (yr
< 4 5
s)
26 27.3 0.8 5 23.3 0.4
45.1-58 46 25.2 121 16.9
58.1-67 - 111 15.9
67.1-89 - 102 12.5
Age at Diagnosis
35-45
(yrs)
40 30.3 0.2 11 12.9 0.2
45-50 25 23.6 36 21.6
50.1-56 7 22.1 87 15.1
56.1-65 105 16.1
65.1-87 100 14.9
M enarche Age (y
< 12
/rs)
19 22.5 0.5 79 15.1 0.3
12 26 28.1 88 14.5
13 15 30.7 87 15.6
> 13 12 28.2 84 18.2
Parity
0 16 30.3 0.8
1 m issing
51 15.1 0.2
1 11 29.3 50 16.3
2 32 25.1 100 17.8
3 11 26.1 73 15.0
> 4 2 33.1 65 13.4
Age at first birth
< 2 0
(yrs)
10 26.6 0.9 73 16.1 0.9
21-22 6 33.7 54 15.9
22.1-25 9 26.8 65 15.5
25.1-29 13 28.1 62 15.2
> 2 9 18 24.5 34 16.9
N ulliparous 16 30.3 51 15.1
Fam ily historyd
No 53 24.8 0.05 248 16.1 0.9
Yes 19 33.0 77 15.9
BM Ia
< 2 4 .9 43 33.8 < 0.001 155 18.4 <0.001
25-29.9 16 20.6 106 15.7
> 2 9 .9 11 10.3 72 10.8
Alcohol
Never 9 30.2 0.5 93 13.2 0.02
Ever 63 26.6 246 16.9
T a b le c o n tin u e d ..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
Table 10. Adjusted least-squared mean post-diagnostic percent m am m ographic density o f Non-
H ispanic W hites breast cancer women stratified on menopausal status.
Prem enopausal women* Postm enopausal women*
N Lsm ean P-value N Lsm ean P-value
Oral contraceptive
Never 10 24.6 0.6 134 16.9 0.4
E ver 62 27.5 204 15.2
1 m issing
E strogen receptor
Positive 28 27.1 0.5 215 16.3 0.4
N egative 6 22.8 39 14.4
38 m issing 85 m issing
Progesterone receptor
Positive 23 25.1 0.9 175 16.7 0.3
N egative 7 26.1 75 14.6
42 m issing 89 m issing
H RT0
N ever 59 - - 108 15.0 0.4
E ver 0 - 230 16.3
13 missing
Sm oke
N ever 39 27.5 0.8 169 15.4 0.4
E ver 33 26.6 169 16.5
T am oxifen use
No 58 28.9 0.07 192 15.8 0.8
Yes 9 17.9 141 16.1
5 m issing 6 m issing
C hem otherapy
No 61 27.1 0.9 246 16.3 0.4
Yes 10 27.2 93 14.7
1 m issing
R adiation Therapy
No 27 29.1 0.4 118 16.2 0.8
Yes 44 25.8 221 15.7
1 m issing
Stage
In Situ 33 27.1 0.7 60 15.9 0.9
Localized 31 26.2 216 15.8
Regional 8 30.8 61 16.2
Distant - 2 13.7
Light-intensity ViA* 8 (M ET hrs/week)
0-20.4 24 27.0 0.9 112 15.9 0.7
20.41-38.4 28 25.5 108 16.5
38.4-123 18 26.9 115 15.1
2 m issing 4 m issing
Table continued..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
Table 10. Adjusted least-squared mean post-diagnostic percent m am m ographic density of Non-
Hispanic W hites breast cancer women stratified on menopausal status.
Prem enopausal wom en Postm enopausal wom en
N Lsm ean P-value N Lsm ean P-value
M oderate-intensity PA eg (M ET hrs/week)
0-10.4 18 26.0 0.4 118 15.3 0.4
10.41-25.9 27 24.9 108 17.2
25.91-110 26 30.1 109 14.9
1 m issing 4 m issing
V igorous-intensity PA e (M ET hrs/week)
0 27 23.1 0.3 210 16.1 0.2
0.02-1.193 9 30.3 32 13.6
1.194-4.38 11 30.5 31 11.8
4.39-12.25 14 24.6 30 17.5
12.26-83.4 10 34.1 33 18.9
1 m issing 3 m issing
Total Physical Ac;tivityh (M ET hrs/w eek)
0-39.7 23 24.8 0.6 112 15.2 0.8
39.8-67.2 20 25.5 117 16.2
67.3-238.1 27 28.4 105 15.9
2 m issing 5 m issing
a Body M ass Index (k g /n rj.
c Horm one replacem ent therapy,
d First degree fam ily history,
e Physical Activity.
f Age at image, age at diagnosis, parity, body m ass index, alcohol use, m oderate intensity physical activity,
vigorous intensity physical activity, light intensity physical activity, and oral contraceptive use.
g Age at image, age at diagnosis, parity, body m ass index, alcohol use and oral contraceptive use.
h Total Physical activity= M oderate intensity PA+ light intensity PA + vigorous intensity PA
Of 159 eligible Hispanic Whites, we had complete interview data and
mammographic data for 99 subjects. Failure to obtain the data was because of refusal
to participate in follow-up interview, inability to locate, death, no consent from the
subject to request mammograms from the facility, or no records of mammograms at
any facility listed. We obtained both pre- and post-diagnostic images for 73 subjects,
11 had only post-diagnostic images and 15 had only-pre-diagnostic images.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
Table 11. Univariate mean pre-diagnostic percent m am m ographic density o f Hispanic-W hites breast
cancer women stratified on m enopausal status.
N M ean S.D .b P-value N M ean S.D .b P-value
Age at Im age (yr s)
< 46 20 29.2 16.6 0.05 3 33.6 8.1 0.08
46.1-55 8 42.4 11.9 17 23.8 16.4
55.1-64 19 15.9 11.6
64.1-78 21 15.8 14.9
M enarche Age (y/rs)
< 12 5 30.6 18.9 0.58 14 18.3 11.4 0.14
12 8 38.2 17.6 13 26.9 15.5
13 7 26.7 16.9 18 17.7 17.3
13 < 8 34.8 14.0 15 14.4 11.6
Parity
0 9 35.7 23.6 0.8 6 39.0 12.2 0.003
1 6 31.6 13.8 7 20.3 15.1
2 8 34.1 13.7 12 21.4 14.5
3 3 33.7 3.6 12 14.9 10.3
4 < 2 19.2 3.7 23 14.2 13.2
Age at first birth (yrs)
<19 4 22.7 10.5 0.53 17 12.9 11.8 0.001
19.1-21 4 33.4 6.2 13 14.2 11.9
21.1-25 4 26.5 19.3 12 20.1 15.2
25.1-41 7 38.8 7.2 12 21.7 13.6
N ulliparous 9 35.7 23.6 6 39.0 12.2
BM T
< 2 4 .9 14 39.5 17.5 0.09 21 23.8 17.3 0.11
25-29.9 7 25.4 13.6 21 18.2 13.6
29.9 < 6 26.3 12.8 18 14.2 11.2
1 m issing
Fam ily historyd
No 22 30.9 13.4 0.33 36 20.9 14.6 0.39
Yes 5 39.0 27.1 21 17.5 14.9
1 m issing 3 m issing
Alcohol
Never 10 29.8 13.9 0.5 29 17.5 14.9 0.45
Ever 17 34.7 18.2 31 20.4 14.5
1 m issing
Oral contraceptive
Never 2 41.1 19.7 0.47 30 16.2 14.1 0.14
Ever 26 32.4 16.4 30 21.8 14.9
HRT°
Never 22 - - - 17 16.7 14.9 0.44
Ever 0 - - 42 19.9 14.8
6 m issing
T a b le c o n tin u e d ..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
Table 11. U nivariate m ean pre-diagnostic percent m am m ographic density o f Hispanic-W hites breast
cancer women stratified on menopausal status.
Prem enopausal women Postm enopausa wom en
N M ean S.D .b P-value N M ean S.D .b P-value
Smoke
Never 13 26.9 14.1 0.07 31 19.8 13.4 0.66
Ever 15 38.3 16.8 29 18.1 16.1
Light-intensity P Ae (M E T hrs/week)
0-12.92 9 30.8 12.9 0.8 21 14.8 11.9 0.3
12.93-35 15 34.5 15.0 20 22.1 14.7
35.1-93 3 37.4 33.9 8 18.7 15.6
1 m issing 1 m issing
M oderate and vijjorous intensity PAe (M E T hrs/week)
0-14.1 9 32.4 10.9 0.9 19 15.7 14.7 0.3
14.2-29.7 8 34.2 22.7 22 23.0 14.1
29.8-75 10 31.8 16.9 19 17.6 14.9
1 m issing
Total Physical Activity* (M E T hrs/week)
0-37.6 9 30.4 12.4 0.9 20 16.5 12.7 0.5
37.7-67.3 9 34.4 13.4 20 19.1 13.1
67.4-168 9 33.4 23.5 19 22.1 18.1
1 m issing
a Body M ass Index (k g /n r),
b Standard deviation,
c Horm one replacem ent therapy,
d First degree fam ily history,
e Physical Activity.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
Table 12. A djusted1 least-squared mean pre-diagnostic percent m am m ographic density of Hispanic-
W hites breast cancer women stratified on menopausal status.
Prem enopausal women Postm enopausal women
N Lsm ean P-value N Lsm ean P-value
Age at Im age (yrs)
< 46 20 27.9 0.02 3 25.1 0.3
46.1-55 8 44.9 17 24.2
55.1-64 19 16.0
64.1-78 21 16.5
M enarche Age ()/rs)
< 12 5 34.5 0.5 14 20.8 0.6
12 8 39.8 13 22.0
13 7 26.6 18 18.7
13 < 8 31.4 15 15.2
Parity
0 9 36.9 0.4 6 29.6 0.5
1 6 31.5 7 16.0
2 8 38.6 12 23.8
3 3 34.0 12 16.7
4 < 2 17.9 23 19.5
Age at first birth (yrs)
< 19 4 33.2 0.09 17 15.3 0.2
19.1-21 4 46.0 13 17.6
21.1-25 4 16.1 12 25.4
25.1-41 7 34.5 12 21.9
N ulliparous 9 36.9 6 29.6
BM Ia
< 2 4 .9 14 39.5 0.1 21 24.0 0.06
25-29.9 7 26.5 21 18.1
29.9 < 6 24.9 18 14.0
1 m issing
Fam ily historyd
No 22 29.7 0.09 36 20.6 0.5
Yes 5 43.3 21 18.2
1 m issing 3 m issing
Alcohol
Never 10 32.9 0.9 29 19.4 0.9
Ever 17 32.9 31 18.7
1 m issing
Oral contraceptive
Never 2 39.8 0.5 30 17.6 0.5
Ever 26 32.3 30 20.3
HRT°
Never 22 - - 17 16.9 0.4
Ever 0 - 42 19.9
6 m issing
Smoke
Never 13 27.6 0.2 31 18.8 0.9
Ever 15 37.8 29 19.2
Table continued..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
Table 12. A djustedf least-squared mean pre-diagnostic percent m am m ographic density o f Hispanic-
W hites breast cancer women stratified on menopausal status.
Prem enopausal women Postm enopausal wom en
N Lsm ean P-value N Lsm ean P-value
Light-intensity 1 ’Ae (M E T hrs/week)
0-12.92 9 31.2 0.9 21 14.9 0.2
12.93-35 15 34.7 20 21.6
35.1-93 3 33.6 8 18.9
1 m issing 1 m issing
M oderate and vigorous intensity PAe (M E T hrs/week)
0-14.1 9 30.1 0.9 19 15.6 0.4
14.2-29.7 8 33.4 22 21.9
29.8-75 10 34.0 19 18.8
1 m issing
Total Physical Activity® (M ET hrs/week)
0-37.6 9 26.6 0.4 20 15.1 0.3
37.7-67.3 9 37.7 20 19.9
67.4-168 9 32.9 19 22.5
1 m issing
a Body M ass Index (k g /n r),
b Standard deviation,
c H orm one replacem ent therapy,
d First degree fam ily history,
e Physical Activity.
f Adjusted for age at the image, body mass index, alcohol use, horm one replacem ent therapy, parity and age at
first full term birth.
g. Total physical activity = light intensity + m oderate intensity + vigorous intensity physical activity.
In Hispanic-White women, univariate and adjusted analyses showed a
statistically significant increase in pre-diagnostic percent breast density with increase
in age at the time of mammogram in pre-menopausal women (tables 11 and 12).
Post-menopausal Hispanic-White women also showed a statistically significant
(p<0.05) increase in pre-diagnostic percent breast density with increase in the age at
first full term birth and increase in number of births.
We also observed a marginally significant (p=0.09) increase in percent breast
density in pre-menopausal women who smoked tobacco as compared to women who
never smoked tobacco in their lifetime for the pre-diagnosis period.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
After adjusting for age at mammogram, parity and other covariates, we
observed a statistically significant (p<0.05) decrease in percent breast density with
increase in BMI and increase in percent breast density when the estrogen receptors
were negative.
Table 13 Univariate mean post-diagnostic percent m am m ographic density o f Hispanic W hites breast
cancer women stratified on menopausal status.
Prem enopausal wom en Postm enopausal women
N M ean S.D .b P-value N M ean S.D.b P-value
Age at Im age (yr
35-50
s)
13 22.7 13.2 0.2 9 18.9 11.9 0.3
50.1-55 5 34.2 20.1 13 16.2 14.7
55.1-65 - 22 13.7 10.2
65.1-82 - 22 10.8 8.9
Age at D iagnosis
32-48
(yrs)
15 ^ 22.9 12.9 0.06 9 18.9 11.9 0.2
48.1-53 3 41.1 22.7 13 16.2 14.7
53.1-63.5 - 21 14.1 10.2
63.5-80 - 23 10.6 8.8
M enarche Age (j
< 12
trs)
5 ^ 31.9 24.2 0.77 17 14.4 10.1 0.4
12 6 24.6 8.2 11 19.0 12.6
13 3 19.7 17.0 21 11.8 12.0
13 < 4 25.0 14.4 17 12.9 9.9
Parity
0 7 29.4 20.9 0.7 8 21.6 15.6 0.3
1 3 19.9 5.5 7 13.1 13.1
2 5 29.8 14.9 12 14.3 11.9
3 2 13.7 4.5 12 14.3 7.9
4 < 1 24.3 - 27 11.6 9.8
Age at first birth
15-19
(yrs)
2 18.8
1
7.8 0.8 20 11.7 8.9 0.2
19.1-21 5 22.1 9.1 12 14.9 12.8
21.1-25 1 14.5 - 12 10.9 6.8
25.1-41 3 32.7 17.8 14 14.5 12.1
N ulliparous
BM Ia
7 29.4 20.9 8 21.6 15.6
< 2 4 .9 8 29.5 19.4 0.6 21 21.1 12.5 0.001
25-29.9 6 25.1 13.1 26 10.3 7.5
29.9 < 4 19.9 12.4 19 11.1 10.6
Smoke
N ever 9 26.3 16.4 0.9 35 12.1 9.8 0.2
Ever 9 25.5 15.9 31 16.0 12.4
Table continued..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
Table 13. Univariate mean post-diagnostic percent m am m ographic density of Hispanic W hites breast
cancer women stratified on menopausal status.
Prem enopausal wom en Postm enopausal women
N Mean S.D .b P-value N Mean S.D .b P-value
A lcohol
N ever 1 37.9 - 0.4 27 11.2 9.1 0.09
E ver 17 25.2 15.9 39 15.9 12.1
Oral contraceptiv e
|
N ever 2 20.3 18.2 0.6 28 9.9 7.7 0.01
E ver 16 26.6 15.9 38 16.9 12.4
E strogen receptor
|
Positive 8 18.2 11.6 0.7 38 11.4 8.7 0.006
N egative 3 20.9 5.7 7 23.0 14.3
7 m issing 21 m issing
Progesterone receptor |
Positive 6 14.5 10.8 0.1 27 13.2 10.5 0.8
Negative 5 24.3 6.7 16 14.3 10.9
7 m issing 23 m issing
HRT°
N ever 12 - - - 23 17.0 14.8 0.1
E ver 0 - - 42 12.6 8.3
6 missing 1 missing
Fam ily historyd
No 17 24.3 14.6 0.08 39 15.6 10.9 0.09
Yes 1 52.9 - 22 10.6 10.9
Tam oxifen use
No 14 28.6 16.5 0.3 44 15.2 12.1 0.2
Yes 3 17.5 9.2 22 11.4 8.8
1 missing
C hem otherapy
No 12 27.9 18.4 0.5 49 13.2 10.0 0.4
Yes 6 21.9 8.2 17 16.1 14.1
Radiation Therapy |
No 6 35.2 20.9 0.08 21 13.7 10.7 0.9
Y es 12 21.3 10.7 45 14.1 11.5
Stage
In Situ 4 37.8 22.9 0.2 13 16.3 11.9 0.5
Localized 11 21.0 13.2 39 12.7 11.3
Regional 3 27.9 4.4 14 15.3 10.3
D istant - -
Light-intensity PA e (M ET hrs/week) |
0 - - - - - -
3.12-17.5 6 20.8 20.4 0.8 18 12.8 11.8 0.8
17.51-32.1 2 28.8 8.2 16 15.1 7.4
32.11-57.6 3 32.7 17.8 16 12.3 11.9
57.61-139.8 5 27.3 16.5 15 15.6 13.7
2 m issing 1 m issing
Table continued..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
Table 13. Univariate mean post-diagnostic percent m am m ographic density of H ispanic W hites breast
cancer women stratified on menopausal status.
Prem enopausal wom en Postm enopausal wom en
N M ean S.D .b P-value N M ean S.D .b P-value
M oderate-intensity PAe (M E T hrs/w eek) |
0 - - 2 4.4 2.5 0.5
0.47-10.5 6 28.1 16.1 0.6 15 13.8 10.5
10.51-20.0 3 15.7 2.1 14 10.4 13.7
20.1-36.67 2 19.4 21.5 18 14.5 9.7
36.68-68.3 7 30.3 17.6 14 15.8 9.4
V igorous-intensity PA e (M E T hrs/week)
0 5 14.4 9.8 0.3 34 12.9 10.0 0.8
0.007-0.88 4 33.7 19.9 7 11.6 8.9
0.89-2.44 3 36.8 16.6 7 15.5 10.8
2.45-7.82 2 34.9 13.5 8 10.7 7.5
7.83-19.5 2 23.9 15.1 8 15.8 14.9
a Body M ass Index (k g /n r),
b Standard deviation,
c H orm one replacem ent therapy,
d First degree fam ily history,
e Physical Activity.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
Table 14. Adjustedf least-squared m ean post-diagnostic percent m am m ographic density o f Hispanic
W hites breast cancer women stratified on m enopausal status.
Prem enopausal wom en Postm enopausal wom en
N Lsm ean P-value N Lsm ean P-value
A ge at Image (yr
35-50
s)
13 21.9 0.1 9 17.4 0.8
50.1-55 5 36.2 13 12.7
55.1-65 - 22 13.8
65.1-82 - 22 13.4
A ge at Diagnosis
32-48
(yrs)
15 25.2 0.8 9 12.6 0.9
48.1-53 3 29.5 13 10.4
53.1-63.5 - 21 14.0
63.5-80 - 23 16.4
M enarche Age ('
< 12
,TS)
5 34.9 0.5 17 16.9 0.4
12 6 29.3 11 16.1
13 3 7.1 21 12.0
13 < 4 23.7 17 11.9
Parity
0 7 19.4 0.1 8 16.7 0.9
1 3 11.9 7 13.7
2 5 44.0 12 12.3
3 2 28.7 12 15.4
4 < 1 69.5 27 14.2
Age at first birth
15-19
(yrs)
2 14.5 0.5 20 12.0 0.6
19.1-21 5 32.3 12 16.9
21.1-25 1 30.4 12 13.4
25.1-41 3 47.8 14 15.8
N ulliparous 7 19.4 8 16.7
B M Ia
< 2 4 .9 8 28.3 0.8 21 20.8 0.002
25-29.9 6 25.1 26 10.5
29.9 < 4 22.2 19 11.1
Sm oke
Never 9 24.4 0.7 35 12.7 0.3
Ever 9 27.4 31 15.4
Alcohol
Never 1 35.5 0.6 27 13.3 0.7
Ever 17 25.3 39 14.4
Oral contraceptb
Never
/e
2 18.1 0.6 28 11.2 0.1
E ver 16 26.9 38 15.9
Table continued..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
Table 14. A djustedf least-squared mean post-diagnostic percent m am m ographic density of Hispanic
W hites breast cancer women stratified on menopausal status.
Prem enopausal women Postm enopausal wom en
N Lsm ean P-value N Lsm ean P-value
Estrogen receptor
Positive 8 15.7 0.7 38 11.7 0.02
Negative 3 27.4 7 21.6
7 m issing 21 m issing
Progesterone receptor
Positive 6 8.7 0.07 27 12.7 0.4
Negative 5 31.2 16 15.1
7 m issing 23 m issing
HRT°
Never 12 - - 23 17.1 0.1
E ver 0 - 42 12.5
6 m issing 1 m issing
Fam ily historyd
No 17 24.6 0.3 39 15.1 0.2
Yes 1 47.6 22 11.6
Tam oxifen use
No 14 29.3 0.2 44 15.2 0.2
Yes 3 14.3 22 11.5
1 m issing
C hem otherapy
No 12 22.1 0.5 49 14.4 0.6
Yes 6 33.5 17 12.8
R adiation Therapy
No 6 30.2 0.5 21 14.5 0.8
Yes 12 23.7 45 13.7
Stage
In Situ 4 24.9 0.3 13 17.9 0.3
Localized 11 22.7 39 12.7
Regional 3 39.1 14 13.8
Distant - -
L ight-intensity PA e (M ET hrs/w eek)
0-19.4 7 17.3 0.6 20 12.8 0.6
19.5-41.1 2 26.3 24 15.1
41.2-108 6 34.1 20 12.2
1 m issing 2 m issing
M oderate and vigorous intensity PAe (M ET hrs/week)
0-14.2 7 28.9 0.2 25 14.1 0.2
14.2-74.0 3 7.6 22 11.3
74.0-180 8 30.2 19 16.8
Total Physical A ctivity8 (M E T hrs/w eek)
0-49.2 7 15.8 0.1 27 12.9 0.8
49.3-74.0 6 31.9 19 15.2
74.1-180 5
----.. 'V ....
32.9 20 14.1
b Standard deviation,
c Horm one replacem ent therapy,
d First degree fam ily history,
e Physical Activity.
f Adjusted for body m ass index, age at image, age o f m enarche, parity, age at first full term birth and oral
contraceptive use.
g. Total physical activity= Light +m oderate + vigorous intensity physical activities.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
We also observed a statistically significant (p<0.05) increase in the post
diagnostic percent breast density of post-menopausal women who had taken oral
contraceptive in their lifetime as compared to postmenopausal women who had never
taken oral contraceptive in their lifetime (table 13). A marginally significant (p=0.06)
increase in post-diagnostic percent breast density was observed with increase in age
at diagnosis in pre-menopausal Hispanic-White women.
Multivariate linear regression models were done on pre and post diagnostic
percent mammographic density with race as dummy variable. There were no
differences in the direction and estimates of the regression coefficients for the
relationship of risk factors and percent density for pre-diagnostic and post-diagnostic
time periods (Tablel5). In our sample population, after adjusting for significant
variables, we also observed that Non-Hispanic White women had slightly higher
percent density as compared to Hispanic-White and African-American women.
Premenopausal women were also shown to have higher percent density as compared
to post-menopausal women (p<0.05). We also observed statistically significant
inverse association between BMI and mammographic density, and the magnitude of
the association varied by menopausal status (p for interaction < 0.05). Partial
Spearman correlation coefficient (r) of pre-diagnostic mammographic density with
BMI, adjusting for race, age and study site was stronger in pre-menopausal women
(r=-0.53) compared with post-menopausal women (r=-0.30). For post-diagnostic
mammographic density, the partial correlation coefficients were similar in direction
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
and magnitude in pre- and post-menopausal women (r=-0.57 and -0.37,
respectively).
Table 15. Regression models of pre-diagnostic and post-diagnostic percent m ammographic density
across all races.
Param eters Pre-D iagnostic Post-D iagnostic
Estim ate S.E. * Partial
R2
P Estim ate S.E. Partial
R2
P
Intercept 134.5 10.7 109.2 10.4
Race
Black 11.6 14.1 0.002 0.4 7.5 12.0 0.001 0.5
Hispanic -2.4 1.7 0.003 0.2 -3.0 1.5 0.01 0.04
Study Site
W ashington -8.7 1.5 0.06 < 0.001 -6.6 1.3 0.04 <0.001
N ew M exico -13.9 14.1 0.003 0.3 - 10.1 12.0 0.001 0.4
Age at m am m ogram (yrs) -0.5 0.1 0.08 <0.001 -0.4 0.1 0.06 <0.001
BM I (kg/m 2) -2.8 0.4 0.05 <0.001 -2.2 0.4 0.05 <0.001
Postm enopausal -30.4 6.2 0.02 <0.001 -25.1 5.7 0.03 <0.001
B M I*M enopausal status 1.0 0.2 0.02 <0.001 0.8 0.2 0.02 <0.001
Total R 2 0.33 0.30
S.E. =Standard Error
Baseline: prem enopausal, N on-H ispanic W hite w om en at U SC study site.
DISCUSSION
In this Health, Eating, Activity and Lifestyle (HEAL) multi-racial and multi
center cohort of African-American, Non-Hispanic White and Hispanic White women
with breast cancer, we found that there are different prognostic factors that affect
percent mammographic density in the three racial groups. These factors differ by
menopausal status and also by time relative to diagnosis.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
In this study, we found that percent mammographic density decreased with
increase in BMI, age, family history and alcohol consumption in Non-Hispanic
White women which is consistent with the findings of other studies (Maskarinec and
others 2001, Byrne and others 1995, Valcohn and others 2000). Our study also
showed that HRT and family history were only associated with pre-diagnostic
mammographic density of postmenopausal women. This finding is consistent with
the findings of Valcohn and others (2000). The plausible reason for not finding this
association in the post-diagnosis period could be because of Tamoxifen, a non
steroidal agent that has anti-estrogen affect on breast, as a part of treatment for breast
cancer that could be nullifying the affect of hormone replacement therapy in post
diagnosis period. We also found a significant association of vigorous and moderate
intensity physical activity with pre-diagnostic mammographic density and with post
diagnostic mammographic density in premenopausal Non-Hispanic White women.
There have been four studies (Ursin and others 2003, del Carmen and others,
2003, El-Bastawissi and others 2001 and Chen and others 2004) that have
investigated mammographic density in African-American women. These studies
were done to inspect the associations of mammographic density with breast cancer
risk and to investigate the difference in the mammographic densities between
different racial groups. Our study is the first study that looks at the determinants of
mammographic density in African American women. We found that BMI and total
physical activity in MET hrs/week were significant predictors of mammographic
density in premenopausal women. Post diagnosis, chemotherapy was also
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
significantly associated with a decrease in mammographic density in both
premenopausal and postmenopausal women and alcohol intake in only
postmenopausal women.
There has been only one study that has looked at the association of breast
cancer risk factors with mammographic density in Hispanic-White women (Gapstur
and others 2003). In their study, they found a positive association of post
menopausal percent-mammographic density with HRT and inverse relationship with
age. They also found an inverse association of BMI with percent mammographic
density in pre-menopausal women. In the multivariate analysis, they found an
inverse association with BMI and a marginally significant association of smoking
with pre-menopausal percent density; inverse association of age and BMI, and
positive association with HRT with post-menopausal percent density.
Even though, we found similar associations of HRT and smoking with
percent mammographic density in our study, we failed to observe statistical
significance in both univariate and multivariate analyses. This could be because of
possible misclassification bias that might exist in the variable measuring exposure to
HRT. In our study, we asked if a subject was ever or never exposed to HRT, as a
result of which, subjects who might have taken HRT for considerable short period of
time would also be classified as ‘ever’ subgroup. This would lead to biasing the
association towards null. The same is true for the variable measuring smoking status
in our study.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
There are many limitations in this study that should to be understood before
interpreting the results. The participants in our study were selected from a cancer
registry. Therefore, due to selection bias we might observe a stronger association of
breast cancer risk factors with mammographic density as compared to other studies
where the cohorts were not breast cancer cases. Another limitation in our study was
the small sample size of African-American and Hispanic women, which was further
reduced after stratification on menopausal status, time relative to diagnosis and sub
groups of the risk factors. The analysis for this study involved multiple comparisons
consisting of multiple statistical tests, which increases our probability of observing a
statistically significant relationship by mere chance. The HEAL study is a multi
center cohort study but the possible biases that could have resulted because of that
were controlled for by having only one person read the mammographic images and
calculating percent density for them and using similar interviewing protocol to
collect data from the participants.
This study is one of the few cohort studies of a population-based cohort of
breast cancer subjects. This is also one of the first cohort studies that followed
African-American women with breast cancer. This study gives us the opportunity to
compare the associations of different breast cancer prognostic factors with
mammographic density (surrogate for sex-steroid hormones) across African-
American, Non-Hispanic White and Hispanic White women as it is one of the first
studies that conducted race specific analysis on the associations of the breast cancer
prognostic factors and mammographic density. Our study suggests that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
mammographic density may be modified by altering risk and prognostic factors.
Future researches should be done to investigate if association of mammographic
density with risk and prognostic factors causal in nature and also if the alteration in
mammographic densities by controlling its determinants can lead to change in breast
cancer risk.
In conclusion, our data suggests that prognostic factors for breast cancer may
affect mammographic density differently in the different racial groups. This affect of
the prognostic factors also varies by the menopausal status and time period from
diagnosis.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
BIBLIOGRAPHY
Ainsworth BE, Haskell WL, Whitt MC, Irwin ML, Swartz AM, et al. 2000.
Compendium of physical activities: an update of activity codes and MET
intensities. Med Sci Sports Exerc 32: S498-504.
Boyd NF, Byng JW, Jong RA, Fishell EK, Little LE, et al. 1995. Quantitative
classification of mammographic densities and breast cancer risk: results from
the Canadian National Breast Screening Study. J Natl Cancer Inst 87: 670-5.
Boyd NF, Lockwood GA, Byng JW, Tritchler DL, Yaffe MJ. 1998. Mammographic
densities and breast cancer risk. Cancer Epidemiol Biomarkers Prev 7: 1133-
44.
Brisson J, Merletti F, Sadowsky NL, Twaddle JA, Morrison AS, Cole P. 1982.
Mammographic features of the breast and breast cancer risk. Am J Epidemiol
115:428-37.
Byrne C, Schairer C, Wolfe J, ParekhN, Salane M, et al. 1995. Mammographic
features and breast cancer risk: effects with time, age, and menopause status.
J Natl Cancer Inst 87: 1622-9.
Chen Z, Wu AH, Gauderman WJ, Bernstein L, Ma H, et al. 2004. Does
mammographic density reflect ethnic differences in breast cancer incidence
rates? Am J Epidemiol 159: 140-7.
del Carmen MG, Hughes KS, Halpem E, Rafferty E, Kopans D, et al. 2003. Racial
differences in mammographic breast density. Cancer 98: 590-6.
El-Bastawissi AY, White E, Mandelson MT, Taplin S. 2001. Variation in
mammographic breast density by race. Ann Epidemiol 11: 257-63.
Gapstur SM, Lopez P, Colangelo LA, Wolfman J, Van Horn L, Hendrick RE. 2003.
Associations of breast cancer risk factors with breast density in Hispanic
women. Cancer Epidemiol Biomarkers Prev 12: 1074-80.
Gram IT, Funkhouser E, Tabar L. 1999. Moderate physical activity in relation to
mammographic patterns. Cancer Epidemiol Biomarkers Prev 8: 117-22.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
Greendale GA, Reboussin BA, Sie A, Singh HR, Olson LK, et al. 1999. Effects of
estrogen and estrogen-progestin on mammographic parenchymal density.
Postmenopausal Estrogen/Progestin Interventions (PEPI) Investigators. Ann
Intern Med 130: 262-9.
Hankinson SE, Willett WC, Manson JE, Hunter DJ, Colditz GA, et al. 1995.
Alcohol, height, and adiposity in relation to estrogen and prolactin levels in
postmenopausal women. J Natl Cancer Inst 87: 1297-302.
Irwin ML, Crumley D, McTieman A, Bernstein L, Baumgartner R, et al. 2003.
Physical activity levels before and after a diagnosis of breast carcinoma: the
Health, Eating, Activity, and Lifestyle (HEAL) study. Cancer 97: 1746-57.
Kriska AM, Caspersen CJ. A collection of Physical Activity Questionnaires for
Health-Related Research. Med Sci Sports Exerc 1997; 29(6Suppl.):Sl-S205.
Lam PB, Vacek PM, Geller BM, Muss HB. 2000. The association of increased
weight, body mass index, and tissue density with the risk of breast carcinoma
in Vermont. Cancer 89: 369-75.
Laya MB, Gallagher JC, Schreiman JS, Larson EB, Watson P, Weinstein L. 1995.
Effect o f postmenopausal hormonal replacement therapy on mammographic
density and parenchymal pattern. Radiology 196: 433-7.
Lundstrom E, Wilczek B, von Palffy Z, Soderqvist G, von Schoultz B. 1999.
Mammographic breast density during hormone replacement therapy:
differences according to treatment. Am J Obstet Gynecol 181: 348-52.
Maskarinec G, Meng L, Ursin G. 2001. Ethnic differences in mammographic
densities. Int J Epidemiol 30: 959-65.
McNicholas MM, Heneghan JP, Milner MH, Tunney T, Hourihane JB, MacErlaine
DP. 1994. Pain and increased mammographic density in women receiving
hormone replacement therapy: a prospective study. AJR Am J Roentgenol
163: 311-5.
Oza AM, Boyd NF. 1993. Mammographic parenchymal patterns: a marker of breast
cancer risk. Epidemiol Rev 15: 196-208.
Persson I, Thurfjell E, Holmberg L. 1997. Effect of estrogen and estrogen-progestin
replacement regimens on mammographic breast parenchymal density. J Clin
Oncol 15: 3201-7.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
Roubidoux MA, Kaur JS, Griffith KA, Stillwater B, Novotny P, Sloan J. 2003.
Relationship of mammographic parenchymal patterns to breast cancer risk
factors and smoking in Alaska Native women. Cancer Epidemiol Biomarkers
Prev 12: 1081-6.
Sala E, Warren R, McCann J, Duffy S, Luben R, Day N. 1999. High-risk
mammographic parenchymal patterns and anthropometric measures: a case-
control study. Br J Cancer 81: 1257-61.
Stems EE, Zee B. 2000. Mammographic density changes in perimenopausal and
postmenopausal women: is effect of hormone replacement therapy
predictable? Breast Cancer Res Treat 59: 125-32.
Stamper PC, Van Voorhis BJ, Ravnikar VA, Meyer JE. 1990. Mammographic
changes associated with postmenopausal hormone replacement therapy: a
longitudinal study. Radiology 174: 487-90.
Ursin G, Ma H, Wu AH, Bernstein L, Salane M, et al. 2003. Mammographic density
and breast cancer in three ethnic groups. Cancer Epidemiol Biomarkers Prev
12: 332-8.
Vachon CM, Kuni CC, Anderson K, Anderson VE, Sellers TA. 2000. Association of
mammographically defined percent breast density with epidemiologic risk
factors for breast cancer (United States). Cancer Causes Control 11: 653-62.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
beta3-adrenergic receptor gene Trp64Arg polymorphism and obesity-related characteristics among African American women with breast cancer: An analysis of USC HEAL Study
PDF
Family history, hormone replacement therapy and breast cancer risk on Hispanic and non-Hispanic women, The New Mexico Women's Health Study
PDF
Association between body mass and benign prostatic hyperplasia in Hispanics: Role of steroid 5-alpha reductase type 2 (SRD5A2) gene
PDF
BRCA1 mutations and polymorphisms in African American women with a family history of breast cancer identified through high throughput sequencing
PDF
A descriptive analysis of medication use by asthmatics in the Children's Health Study, 1993
PDF
Evaluation of the accuracy and reliability of self-reported breast, cervical, and ovarian cancer incidence in a large population-based cohort of native California twins
PDF
Dietary fiber intake and atherosclerosis progression: The Los Angeles Atherosclerosis Study
PDF
A pilot survey of medical abortion knowledge and practices among obstetrician/gynecologists and family practitioners in Los Angeles County
PDF
Comparisons of metabolic factors among gestational diabetes mellitus probands, siblings and cousins
PDF
Extent, prevalence and progression of coronary calcium in four ethnic groups
PDF
The association between recreational physical activity and mammographic density
PDF
Development and evaluation of standardized stroke outcome measures in a population of stroke patients in rural China
PDF
A linear model for measurement errors in oligonucleotide microarray experiment
PDF
A case-control study of passive smoking and bladder cancer risk in Los Angeles
PDF
Polymorphisms in genes involved in steroid hormone metabolism and mammographic density changes in women randomized to menopausal estrogen and progesterone therapy
PDF
Cigarettes and alcohol in relation to colorectal cancer within the Singapore Chinese Health Study
PDF
Recreational physical activity and risk of breast cancer: The California Teachers Study
PDF
Androgens and breast cancer
PDF
Association between latchkey status and smoking behavior in middle school children
PDF
Effect of hormone therapy on the progression of carotid-artery atherosclerosis in postmenopausal women with and without established coronary artery disease
Asset Metadata
Creator
Ingle, Ashish Mark
(author)
Core Title
Determinants of mammographic density in African-American, non-Hispanic white and Hispanic white women before and after the diagnosis with breast cancer
School
Graduate School
Degree
Master of Science
Degree Program
Applied Biostatistics and Epidemiology
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
biology, biostatistics,health sciences, public health,OAI-PMH Harvest
Language
English
Contributor
Digitized by ProQuest
(provenance)
Advisor
Gilliland, Frank (
committee chair
), Gauderman, William James (
committee member
), Ingles, Sue Ann (
committee member
)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c16-48612
Unique identifier
UC11326969
Identifier
1435084.pdf (filename),usctheses-c16-48612 (legacy record id)
Legacy Identifier
1435084.pdf
Dmrecord
48612
Document Type
Thesis
Rights
Ingle, Ashish Mark
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA
Tags
biology, biostatistics
health sciences, public health