l LHHM3Y
WU? Michigan §tate
I University

This is to certify that the
thesis entitled

ADOLESCENT PHYSICAL ACTIVITY AND BREAST
CANCER RISK

presented by

RENEE C. BLOOME

has been accepted towards fulfillment
of the requirements for the

MS. degree in EPIDEMIOLOGY

 

 

fimfiu 1Q. Sigma/L

Majcr Hrofessor’s Signature
‘5 I 10 I “Z 0 0 ‘2
Date

MSU is an affinnative—acfion, equal-opportunity employer

.—._--.-----nu-----.-.-.-.-.-.-n-.-.n.-..~n-.-.-.-.-.—.u...s“.u

an _ I

 

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:lProj/Acc&Pres/ClRC/DateDue,indd

 

ADOLESCENT PHYSICAL ACTIVITY AND BREAST CANCER RISK
By

Renee C. Bloome

A THESIS
Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE
Epidemiology
2008

ABSTRACT
ADOLESCENT PHYSICAL ACTIVITY AND BREAST CANCER RISK
By
Renee C. Bloome

Breast cancer (BC) is a global problem. While inheritance only accounts for a
small portion of BC cases, research speculates that some difference in incidence
can be explained by exposures from lifestyle factors. Adolescence has been
hypothesized to be a critical time when lifestyle factors can effect BC risk. We
aim to assess the relationship between adolescent physical activity (PA) and BC
risk in 422 women from the Polish Women’s Health Study. Logistic regression
models were used to estimate odds ratios (OR) and 95% confidence intervals
(95% Cl). Women with increased adolescent PA (355.7 MET-h/day) had
decreased risk of BC overall (OR=O.56; 95%Cl: 0.34-0.95). Premenopausal
women expending 355.7 MET-h/day, had significantly decreased risk of BC
(OR=0.14; 95%CI: 0.04-0.44), while no effect was observed in postmenopausal
women. The joint effect of adolescent and adult PA on BC Showed decreased
premenopausal BC risk in women with PA 3 55.7 MET-h/day during adolescence

regardless of adult PA level.

To the brave and inspiring members of the Pink Ribbon Crew team

iii

ACKNOWLEDGEMENTS
I would like to thank all of my family and friends for all of their motivation and
support along the way. To my parents, thank you for always supporting me
through the ups and downs of my college career. To Tom, thank you for your
understanding — I know it wasn’t always fun or easy. To Drs. Barbara and Bill
Given, and the rest of the Family Care Studies Team, thank you so much for all
of the wonderful experiences I have gained from working with you and your
endless support on this long journey. To my Classmates, Mary, Pete, & Khalid,
thank you for enhancing my epidemiology experiences through these past two
years. Extraordinary thanks to the members and fellow coaches of the Pink
Ribbon Crew, you ladies are truly amazing and your spirit inspired me when my

mind got stuck — thanks for your never-ending support.

Special thanks to my committee members Dr. Hwan Chung and Dr. Ellen Velie; I
could not have done it without your support both in and out of the classroom.
Also thanks to Jayne Goeddeke for always going the extra step to help us

students through our epidemiology journey.

Finally, to my primary advisor, Dr. Dorothy R. Pathak: words can not express
how thankful I am for everything you have done for me. Without your infinite
wisdom, genuine understanding, and superior nocturnal ability, this work would
not be complete. For this and so many other lessons that you have taught me,

both in epidemiology and in life, you have my sincerest thanks.

iv

TABLE OF CONTENTS

LIST OF TABLES ................................................................................. vii

CHAPTER 1: INTRODUCTION

Organization of this Thesis .............................................................. 1
The Burden of Breast Cancer .......................................................... 1
The Source of the Data ................................................................... 4
Study Specific Aim ........................................................................ 5

Hypothesis ....................................................................................... 5

CHAPTER 2: BACKGROUND

Definition of Breast Cancer ........................................................................ 6
Etiology of Breast Cancer .......................................................................... 6
Breast Cancer Risk Factors ............................................................ 6
Menopausal Status as a Potential Effect
Modifier of Breast Cancer Risk Factors ....................................... 9
Evidence for Importance of Adolescent Physical Activity ............. 10
Hypothesized Biologic Mechanisms ............................................. 13
Evidence on the Importance of Exposures During
Adolescence .............................................................................. 1 5
MET Levels as a Measure of Physical Activity ............................. 17

CHAPTER 3: METHODS

Polish Women’s Health Study ....................................................... 19
Design of Study to be Analyzed .................................................... 19

Case Selection .............................................................................. 19
Control Selection ........................................................................... 20
Sample Size .................................................................................. 21

Main Exposure: Adolescent Physical Activity .......................................... 21
Adolescent Non-Occupational Physical Activity ............................ 21
Adolescent Occupation Physical Activity ....................................... 23

Total Physical Activity .............................................................................. 26
Creation of Adolescent MET-h/d Tertiles ....................................... 26
Creation of Adult MET-h/d Tertiles ................................................. 26

Joint Levels of Adolescent and Adult Physical Activity .................. 26
Other Important Breast Cancer Covariates .............................................. 27
Duration of Stay in the United States ............................................ 27
First-degree Family History of Breast Cancer ............................... 27

Adult BMI(1985-1989) ................................................................... 27

Adolescent Body Size ................................................................... 28

Oral Contraceptive(OC) Use ......................................................... 28
Hormone Replacement Therapy (HRT) Use ................................. 28
Age at Menarche ........................................................................... 28
Age at First Full-term Pregnancy & Parity ..................................... 29
Menopausal Status and Age at Menopause ................................. 29
Adult Total Caloric Intake .............................................................. 30
Adult Alcohol Consumption ........................................................... 30
Age in 1989 ................................................................................... 30
Statistical Analysis ................................................................................... 31

CHAPTER 4: RESULTS

Descriptive Characteristics of Women in the Polish Women’s

Health Study .......................................................................................... 33
Assessment of the Effect of Adolescent Physical Activity on Breast

Cancer Risk .......................................................................................... 33
Assessment of the Joint Effect of Adolescent and Adult Physical

Activity on Breast Cancer Risk ............................................................. 36

CHAPTER 5: CONCLUSIONS & DISCUSSION

Discussion ............................................................................................... 38
Limitations ............................................................................................... 42
Strengths ................................................................................................. 43
Conclusions ............................................................................................. 44
APPENDICES ...................................................................................................... viii

vi

LIST OF TABLES

Table 1: Descriptive Characteristics of Women in the Polish Women’s Health
Study ................................................................................................ 48

Table 2: Age Characteristics of Women in the Polish Women’s Health Study....49

Table 3: Unadjusted & Adjusted (Adj.) Models for Adolescent Physical Activity
(PA) in Total MET-h/day & Breast Cancer Risk ................................................... 50

Table 4: The Adjusted Joint Effect of Adolescent and Adult Physical Activity on
Breast Cancer Risk by Menopausal Status ................................................ 51

vii

APPENDICES
Appendix A: Tables 1 — 4 ................................................................................... 48

Appendix B: Physical Activity Questionnaire from the Polish Women’s
Health Study ...................................................................................................... 53

Appendix C: Adolescent Body Size Figures used in the Polish Women’s
Health Study ..................................................................................................... 58

viii

The Organization of this Thesis

This thesis includes an introduction (Chapter 1), topic background
(Chapter 2), study methods (Chapter 3), study results (Chapter 4), and

conclusions made from this study’s results (Chapter 5).

Chapter 1: Introduction

The Burden of Breast Cancer

In 2004, cancers accounted for 29% of all deaths in the United States and
specifically cancer of the breast is the most incident cancer (age-adjusted) found
in American women (WCRF/AICR 2007). However, the burden of breast cancer
is not only confined to the women of the United States, but effects women
worldwide. The United States has one of the highest incidences of breast cancer
in the world with an overall age-adjusted incidence rate of 126.1 per 100,000
American women according to the Surveillance Epidemiology and End Results
(SEER) database (SEER 2008).

The demographic make-up of the United States is not as homogeneous as
other countries around the globe, thus it is important to note that not all races in
America have an equal incidence rates of breast cancer. In America, white
women have the highest incidence rate (132.5 per 100,000) followed by African
Americans, Hispanics, Asian or Pacific Islanders, and American Indian/Alaskan
natives with incidence rates of 118.3, 89.3, 89.0 and 69.8 per 100,000
respectively (SEER 2008). The number of new breast cancer cases in the

United States will be in excess of 182, 460 in 2008 (A08 2007). Within the last

decade, a woman living in America has seen her overall lifetime risk for breast
cancer jump from one in every ten women (2000) up to one in every eight women
(2007) (A08 2007) .

Increases in incidence rates have been seen across the globe, particularly
in other industrialized countries. In Poland, breast cancer incidence rates in
Polish native women have been observed to be nearly one third of those in the
United States (41.4 per 100,000) (Bebenek 2007). Lifestyle differences, such as
diet or other environmental risk factors, between women living in the United
States and those living in Poland have been suggested to contribute to this
almost three-fold difference in incidence between these two countries (Gronwald
2006 & Dunham 2004).

While the ever-expanding field of genetics has been able to link certain
diseases to specific genes, it has been estimated that the inherited causes of
breast cancer (such as genetic mutations in BRCA1, BRCA2, and p53) account
for a small proportion of breast cancer estimated to be as little as ten percent of
all breast cancers diagnosed each year. This leaves up to 90% of the disease to
potentially attributed either environmental or lifestyle risk factors (Newman 1995
8. [Min 2006). These findings suggest that primary prevention of breast cancer
might be achieved through modifiable factors in a woman’s life (Colditz 2005,
Rockhill 1998, & UCCC 2008).

Studies performed on women that have migrated to the Unites States
provide support to this hypothesis (Nelson 2006, Colditz 2005, McPherson 2000,

& Limer 2004). Thus breast cancer prevention might not require drastic and

expensive newly developed scientific discoveries; if fact, it may be as simple as
modifying a lifestyle factor such as diet or a specific activity. If these hypotheses
are correct, the impact of the reduction of the burden of breast cancer worldwide
would be significant.

A number of modifiable risk factors for breast cancer have been studied
Since the disease became a public concern over a century ago. Several
potentially modifiable risk factors including aspects of diet, physical activity, and
body size have been examined for casual links to breast cancer incidence (Irwin
2006). With the observation that increased occupational physical activity
decreased overall cancer mortality in Australians, British, and Americans in 1922,
physical activity became a highly plausible modifiable risk factor for all cancers
(Kruk 2006). Since then dozens of studies have been published on breast
cancer and physical activity and over 70% of them have observed that women
with the highest physical activity level have a lower risk of breast cancer when
compared to their inactive counterparts (Kruk 2006).

In addition to the identifying risk factors for breast cancer, timing when
exposure occurs also plays an important role in the effect Size of the risk factor.
The younger the age when a woman is exposed to risk factors, such as
irradiation, certain foods, alcohol ingestion, and smoking, the greater her later
risk of breast cancer in adulthood (Colditz 1995, Cho 2003, Colditz 2005, & Linos
2008). This evidence suggests that modification of certain risk factors in
adulthood only may be too late and thus not as effective as if such modification

occurred earlier in life. Therefore, preventative measures may be most effective

in breast cancer risk reduction if they occur between late childhood and early
adulthood in a girl's life when her breast tissue develops. Therefore, additional

research should be focused on this time period (Colditz 2005).

The Source of the Data

This analysis will look at the relationship between physical activity during
adolescent years (specifically between ages 12-13) and breast cancer risk in
Polish migrant women using data from the Polish Women’s Health study. The
Polish Woman’s Health Study was designed to examine if Changes in lifestyle
factors of Polish women who immigrated to the United States, might explain the
increase in beast cancer risk observed in Polish migrant women. The main aim
of the study was to examine the effect of diet and other changes in women who
migrated to the United States and compare it to their counterparts who remained
in Poland. Additionally, information was also collected on other established and
potential risk factors such as reproductive history, physical activity, family history,
and occupational histories.

Two parallel case-control studies were conducted; one of women still
living in Poland and the other of Polish-born women whom have migrated to the
United States before 2000. Data collected from both countries included
questions about women’s lifestyles during 1985-1989 as well as their lifestyles
when the women were aged 12-13 years. The time period of 1985-1989 was

chosen in order to capture the traditional Polish diet before introduction of market

economy after the fall of Communism in 1989. The analysis for this paper will

only include the 422 Polish migrant women residing in the United States.

Study Specific Aim

The specific aim of this study is as follows:
1. To examine the relationship between increased physical activity during
adolescence (defined as ages 12-13 years of age) and later risk of breast cancer
in a United States population of Polish migrant women. Physical activity will be
determined by self-reported, usual total of daily (24-hour) physical activity
(summation of inactive, recreational, household, and occupational).
Hypothesis: We hypothesize that increased adolescent physical activity

will decrease a woman’s risk for breast cancer.

Chapter 2: Background

Definition of Breast Cancer

The American Cancer Society’s definition of breast cancer reads as
follows: “a malignant (cancerous) tumor that starts from cells of the breast (ACS
2007).” There are five stages in the progression of breast cancer: stage 0 (in
situ) — stage 4 (distant metastasis). In situ breast cancer is defined as “cancer
cells that remain within the basement membrane of the elements of the terminal
duct Iobular unit and the draining duct” while invasive breast cancer can be
defined as “a dissemination of cancer cells outside the basement membrane of
the ducts and lobules [of the breast] into the surrounding adjacent tissue [and
distant organsl” (stages one through four) (Sainsbury 2000). For this study only
women diagnosed with histologically- or cytologically-confirmed invasive breast

cancers were included as cases.

Etiology of Breast Cancer

Breast Cancer Risk Factors

Globally, breast cancer is the highest incident cancer in women (Wu
2006). Many risk factors for breast cancer have been established and refined
over the years that this disease has been studied. These risk factors can be
divided into three separate groupings: reproductive or hormonal exposures (such
as ages at menarche, first full-term pregnancy, or menopause), lifestyle or
environmental exposures (such as physical activity, diet, or irradiation), and

genetic predispositions (such as BRCA1 and BRCA2 genes). Based on the

Breast Cancer Detection and Demonstration Project (BCDDP) a Gail Model for
prediction of woman’s 5-year absolute breast cancer risk was developed (Chen
2006). The factors that remained significant predictors of breast cancer risk were
included in the Gail model and were current age, age at menarche, age at first
full-term pregnancy, family history of breast cancer in first degree relatives, and
number of breast biopsies (Gail 1989). Based on this model, high risk women
were defined as those whose five year breast cancer risk was 1.666% or higher
(Gail 1989). However, studies have identified numerous additional risk factors
for breast cancer that are currently not included in the Gail Model.

Recently in the context of developing prediction models for breast cancer
risk over the next five or ten years, Santen et al reviewed and prioritized the
additional risk factors that could be used to improve identification of women at
high risk for breast cancer (Santen 2007). Based on literature review, priority
was scored on a scale from one to five. Risk factors with a priority score of one
to two and a half were considered high priority and low priority was given risk
factors that scored over two and a half. Priority score of 1 was given to the risk
factors in the Gail Model parameters, quantitative breast density, free estradiol
plasma concentrations, total parity (including number of births in addition to never
vs. ever) and age at menopause (Santen 2007). High priority risk factors: score
of 1.5 was given to BMI and weight gain (ages 20-50 years); priority score 2 was
assigned to total estosterone plasma concentrations, paternal and second-
degree family history of breast cancer (including their ages of cancer onset) and

ovarian cancer; priority score 2.5 was given to waist-hip ratio and height (loss

and age final height was attained) (Santen 2007). Low priority risk factors,
scores greater than 2.5, included alcohol use, breast feeding (number of years),
genetic mutation in the BRCA1 and BRCA2 genes (applies to less than ten
percent of the population), bone density, IGF plasma concentrations, use of
hormone replacement therapies (estrogen-based), physical activity, high fat diet,
and oral contraceptive use (Santen 2007). Based on the scoring from the expert
panel, this report indicates the importance of modifiable risk factors, most of the
risk factors listed above, are potentially modifiable with the exception of those
that fall under the hereditary-type exposures. Some of the above identified
modifiable breast cancer risk factors such as BMI, weight gain, and hormone
levels may be effected by level of physical activity.

Some of these modifiable risk factors, both high and low priority, have
been hypothesized to be critical not only in adulthood, but even more importantly
during adolescence (the years of breast development) (Colditz 1995, Berkey
1999, & Kodama 1981). Within the last decade, the focus breast cancer etiology
research has been shifting from exposures in adulthood to exposures that
happen in utero, childhood and adolescent years (De Assis 2006 & Okasha
2003). It has been hypothesized that many of the risk factors that have been
inconsistently associated with breast cancer risk when assessed in adulthood
may eventually play an important role in breast cancer etiology when assessed at

an earlier lifetime period such as adolescence (Hunter 1993 & Colditz 1995).

Menopausal Status as a Potential Effect Modifier of Breast Cancer

Risk Factors

The timing of breast cancer’s onset is often further Characterized by the
woman’s menopausal status (premenopausal and postmenopausal). It has been
suggested that pre- and postmenopausal breast cancers could have different
etiologies and new evidence on modifiable risk factors suggests that pre- and
postmenopausal breast cancer Should not be combined into one overall breast
cancer disease (Newman 1995, Irwin 2006, Colditz 2005, & Kruk 2007‘).
Instead, associations should be examined separately in these two groups or
important information regarding breast cancer risks or preventions may be
missed (Enger 2000). Evidence supporting this theory has been particularly
prominent in studies examining body size and energy expenditures (Reeves
2007). Additional examples of where differing effects on breast cancer risk are
seen by differing menopausal status include overall weight, weight gain, and BMI
(Carmichael 2006 & Shin 2008). For one example, having an increased BMI
prior to menopause (premenopausal) reduces breasts cancer risk; however,
increased BMI after menopause (postmenopausal) increases risk (Cleary 1997 &
McPherson 2000).

However, menopausal status has been inconsistently defined in various
studies and sometimes completely omitted when results are reported. At the
American Institute for Cancer Research along with the World Cancer Research
Fund’s recent conference on all cancer sites, the importance of examining pre-

and postmenopausal breast cancer separately was heavily emphasized

(WCRF/AICR 2007). Given this information and the aim of our proposed study,
it is acceptable to hypothesize that we may observe different effect sizes of

adolescent physical activity in premenopausal verses postmenopausal women.

Evidence for Importance of Adolescent Physical Activity

In addition to the importance of the timing of the outcome (breast cancer),

 

timing of exposures is critical to assess (Linos 2008, Fn'edenreich 2001, Colditz
1995, & De Assis 2006). While both lifetime and adult physical activity have
been studied in regards to breast cancer risk, a new rising topic in this field has
been to examine physical activity specifically during the time period of
adolescence. To date, research on adolescent physical activity and breast
cancer risk has been minimal. However, the number of publications on this topic
began to increase in the past decade: over 70% of the studies on adolescent
physical activity and breast cancer have been published after 1998.

To date 32 individual studies have been published on adolescent physical
activity and breast cancer risk in women. Seven of these studies are cohort
(Mamti 2008, Margo/is 2005, MC Tieman 2003, Wyshak 2000, Rockhill 1998,
Paffenbarger 1987, & Fn'sch 1985) and the remaining twenty-five studies
(Sprague 2007, Kruk” 2007, Kruk b 200 7, Magnusson 2005, Yang 2003, Steindorf
2003, Dom 2003, Matthews 2001, Fn'edenreich 2001, Lee 2001, Adams-
CampbeI12001, Verioop 2000, Moradi 2000, Shoff 2000, Levi 1999, Marcus
1999, Carpenter 1999, Gammon 1998, Chen 1997, Hu 1997, MCTieman 1996,

D'Avanzo 1996, Taioli 1995, Mittendorf1995, & Bemslein 1994) are designed as

10

case-control. In addition, two reviews of this literature have been published
(Lagerros 2004 & Monninkhof 2007). The first review was Lagerros et al based
on 23 studies, 19 case-control and 4 cohorts, concluded that moderate to heavy
adolescent (defined as under 24 years of age) physical activity significantly
decreases a woman’s overall risk for breast cancer by 19% (overall OR=0.81;
95%CI: 0.73-0.89) when compared to the lowest level of physical activity
reported (Lagenos 2004). A Similar effect was seen when they summarized
studies that reported effects by menopausal status: premenopausal estimates
(n=9 studies) - OR=0.85, 95%CI:0.70—1.02; and postmenopausal estimates
(n=11 studies) - OR=0.88, 95%CI:0.78-0.99 (Lagemos 2004). Similarly,
Monninkhof et al that included in their review 18 case-control and 3 cohort
studies, saw an overall 20% to 80% reduction in postmenopausal breast cancer
risk with increasing physical activity (sum of all time periods of activity lumped
together), but were unable to make any specific distinctions about the risk for
premenopausal women (Monninkhof 2007). The two reviews, however, did not
include the same studies of the adolescent physical activity and Monninkhof did
not summarize adolescent physical activity separately from adult physical
activity.

The effect of adolescent physical activity on overall breast cancer risk as
observed from cohort studies varies from a Slightly protective relative risk (RR) to
an effect slightly over 1 (Maruti 2008, Margo/is 2005, MCTieman 2003, Wyshak
2000, Rockhill 1998, Paffenbarger1987, & Fn'sch 1985). Exact summarization of

risk from cohort nor the case-control studies is not possible as no single measure

11

of adolescent physical activity was used. In addition, all adolescent physical
activity for these studies was recalled in later adulthood. Two cohort studies
examined the effect of adolescent physical activity in a population of mostly
premenopausal women (Maruti 2008 & Rockhill 1998). Maruti found a significant
protective effect from increased recreational physical activity (measured in MET-
h/wk) (Maruti 2008) while Rockhill did not observe any effect of increasing
vigorous physical activity per month/year (Rockhill 1998). In addition, two cohort
studies looked at this effect in postmenopausal women (Margo/is 2005 &
MCTieman 2003), however, neither found conclusive results.

Nearly half of the case-control studies observed a mild to significant
decreased risk for breast cancer with increasing adolescent physical activity
(Shoff 2000, Kruk“ 2007, Dom 2003, Matthews 2001 , Lee 2001, Marcus 1999,
Bernstein 1994, Levi 1999, Yang 2003, Verioop 2000, & Mittendorf1995).
However, similar to the cohort studies, few studies used the same measurement
for adolescent physical activity. The fourteen other studies (Sprague 2007,
Carpenter 1999, Steindorf 2003, Fn'edenreich 2001 , Kruk b 200 7, Adams-
Campbell 2001, Moradi 2000, Gammon 1998, Chen 1997, Hu 1997, D’Avanzo
1996, MCTieman 1996, Magnusson 2005, & Taioli 1995) observed a null or
slightly increased breast cancer risk with increasing adolescent physical activity.
The age range for these case-control studies was defined in the time period
between 0 and 20 years of age (when adolescent ages were specified). Three
studies (Gammon 1998, Marcus 1999, & Verloop 2000) examined adolescence

around the same time that we did for our analyses and the results from these

12

studies showed a small non-significant protective effect from increasing
adolescent physical activity.

Only 4 studies (Dom 2002, Fn'edenreich 2001 , Matthews 2001, & Hu
1997) examined the relationship of increased adolescent physical activity on
breast cancer risk for both menopausal strata. Overall, results from stratification
by menopausal status suggested a different effect between both pre- and
postmenopausal women in two (Fn'edenreich 200 & Hu 1997), of the four
studies. One study (Matthews 2001), observed that the joint effect from
exercising as both an adolescent and an adult was statistically significant for all
(OR = 0.47 95%CI: 0.36-0.62), premenopausal (OR = 0.57 95%CI: 0.40-0.82),
and postmenopausal women (OR = 0.36 95%CI: 0.24-0.55). Kruk also examined
a joint effect of ever versus never exercising in adolescence and adulthood with
greatest decrease in risk in Polish women who indicated activity for both periods
(OR = 0.45 95%CI: 0.28-0.73) (Kruka 2007).

Additionally, the World Cancer Research Fund and the American Institute
for Cancer Research recently published a report titled “Food, Nutrition, Physical
Activity, and the Prevention of Cancer: a Global Perspective (WCRF/AICR
2007).” Although not broken out at specific ages, total physical activity was
found to be protective both premenopausal (summary OR = 0.97 95%CI:0.95-
1.00) and postmenopausal (summary OR = 0.97 95%CI:0.95-1.00) breast
cancers (WCRF/AICR 2007).

Hypothesized Biologic Mechanisms

13

While the exact mechanism through which physical activity may act upon
breast cancer is unknown, several mechanisms have been proposed. What is
known is that physical activity has the ability to alter multiple aspects of a
woman’s menstrual cycle. Women with increased physical activity have been
observed to have a later age at onset of menarche (Fn‘sch 1981 & Irwin 2006) as
well as reduced total number of ovulations during her lifetime (Bernstein 1987 &
Friedenreich 2002) especially during strenuous physical activity in adolescence
(Rockhill 1998). In women of reproductive ages, physical activity has also been
observed to reduce the level of Circulating estradiol (Jasienska 2006). These
characteristics involve modifying a woman’s exposure to circulating hormones,
which have been shown to play an important role in invasive breast cancer both
premenopausally and postmenopausally (Chavez-MacGregor 2008). Increasing
exposure to physical activity will decrease the level of Circulating ovarian
hormones a woman is exposed to during her reproductive years (Bernstein 2002)
and the resulting fluctuation may account for the overall protective effect
observed between physical activity and breast cancer risk (Bernstein 2005).

Another hypothesized hormonal-altering mechanism is that physical
activity may alter breast cancer risk through regulating both overall weight and
weight gain (Hofi'man-Goetz 1998 & Fn'edenreich 2001). Physical activity is very
strongly associated with weight through the process of balancing the energy
ingested versus expended, known as energy balance (Friedenreich 2002).
Physical activity may also help to decrease the amount of abdominal fat stored

which has been hypothesized to be at an increased metabolic level and possibly

14

associated with increased risk for the development of some cancers including
cancer of the breast (Friedenreich 2002). Rockhill et al notes that weight and/or
weight gain may be the culprit for the differing effects of adolescent physical
activity seen in pre— and postmenopausal women (Rockhill 1998).

Evidence on the importance of exposures during adolescence.

The human breast is the only organ that is not developed (functional) at
birth (Schedin 1997 & Frazier 2004) . It is not until over a decade of life that this
organ even begins to develop and continues its develop for nearly an additional
decade (Shu 2001). During this developmental period, a woman’s breast is
believed to be very susceptible to insults that might predispose to breast cancer
(Dirx 1999, Pryor 1989, & Frazier 2004). While adult breast cells are still
susceptible to carcinogenic insults, many studies have shown greater
susceptibility of breast tissue during adolescence or Childhood than after onset of
adulthood (De Assis 2006, Land 2003, & Ronckers 2005).

The idea that adolescence could be a susceptible time for insult to breast
tissue was first induced from observing Japanese women who survived the
atomic bomb. The Life-Span Study (LSS) was started on residents of Hiroshima
and Nagasaki who were living in the cities on the date the bombs were dropped
on these cities (Land 1995). While all women exposed to these high radiation
blasts were at an increased risk for breast cancer, women exposed before the
age of 20 were at the highest risk and postmenopausal women exposed were at
the lowest risk (Ronckers 2005). Women who were under the age of 19 at the

time of exposure to radiation from the atomic bombings consistently had the

15

highest percentage of radiation-related breast cancers regardless of radiation
dose (Land 1995). For example, the percentage of radiation-related breast
cancer for women exposed to the highest estimated radiation dose of greater
than 0.5 sieverts when they were aged 0-19 years, 20-39 years, and greater than
40 years of age were 72.4%, 60.9%, and 39.1% respectively (Land 1995). This
dose-response relationship became an important observation for estimating a
window of susceptibility for breast cancer risk.

Colditz and Frazier review how the importance of timing of exposure is of
great importance when dealing with many of the risk factors commonly
associated with breast cancer (Colditz 1995). Ages at onset of menarche, first
live birth, ingestion of certain foods, alcohol consumption, exposure to radiation,
and physical activity are some of the important risk factors that have been
observed to change the breast cancer risk depending on time of exposure (TretIi
1996). Animal exposure to isoflavanones during adolescence have shown
increased differentiation in mammary gland development. This is important to
breast cancer risk as early differentiation of breast tissues has been found to be
protective (Thanos 2006).

Other dietary constituents have been shown to have a significant impact
on breast cancer risk when ingested during adolescence. In rodent models a
significantly reduction in incidences of mammary tumors have been observed
after intake of phytoestrogens during their adolescence (Lamartiniere 1998 &
Messina 2001). Recently, Linos et al, based on the Nurses Health Study II,

observed that increased red meat consumption was positively associated with

16

later breast cancer risk when consumed during high school (Linos 2008).
However, no such association was observed in with consumption in adulthood
suggesting that adolescence is the more critical time for this particular dietary
exposure (Linos 2008). Since both diet and activity level are components of the
overall energy balance which, as stated above, is a hypothesized risk factor for
postmenopausal breast cancer, effects of physical activity may be analogous to
observed effects of adolescent dietary exposures.

MET Levels as a Measure of Physical Activity

Physical activity can be measured in three different ways: “intensity (how
much energy is expended), duration (refers to the minutes or hours of individual’s
activity episode), [and] frequency of muscular contractions (e.g. episodes
performed per day, week, or month) (Kruk 2006).” Survey procedures such as
interviews with participants have all three components of physical activity
measurement on the basis of which a summary index of average level of physical
activity is determined (Kruk 2006). One of these summary measures is MET-h
per day or week (Ainsworth 2000). MET levels, units of metabolic equivalent
calculated “as the ratio of the energy expenditure when performing specific
activities to the resting metabolic rate of the subject” calculated from the
assumed average adult (Forsum 2006). In the literature evaluating the effects
of adolescent physical activity and breast cancer risk, most of the recent studies
have used the measure of MET-h as the measure of physical activity. Per unit
time such as day or week, the higher the calculated MET level represents a

higher level of physical activity. In the literature many of the more recent studies

17

do use MET-h to measure physical activity (Monninkhof 2006). Additional
measures such as hours per day or week, for differing levels of intensity (such as

medium or strenuous) have also been reported.

18

Chapter 3: Methods

Polish Women’s Health Study

The Polish Women’s Health Study was designed to examine the lifestyle
factors that differed between Polish immigrants to United States and Polish
native women in Poland that could account for the increased risk of breast
cancer after migration to United States. Dietary differences were the focus of the
original investigation, however, information on physical activity was collected.

Design of the study to be analyzed - Case-control study

Description of the dataset to be used

The data for this thesis work will be obtained from Dr. Dorothy R. Pathak’s
Polish Women’s Health Study. This case-control study was performed to identify
protective factors in the lifestyles of Polish women. Data on breast cancer risk
factors, including amount and intensity of physical activity, were collected from
two sources of women: those living in one of four cities in Poland (Gliwice,
Katowice, Poznan, and Bialystok) and Polish migrants living in the Detroit or
Chicago area (United States). Women in both countries provided information on
their reproductive risk factors as well as specific exposures (including diet and
physical activity) during specific years of their adulthood (1985—1989) as well as
their adolescent years, defined as the time a woman was 12-13 years of age.
The dataset that will be used to test our hypothesis stated above (Chapter 1) is
the US. component of the study and includes the Polish migrant women residing
in Chicago, Illinois or Detroit, Michigan at the time of the study.

Case Selection

19

Breast cancer cases were identified from Cook County (Chicago) using
the Illinois State Cancer Registry and from Detroit Metropolitan Area using the
Surveillance Epidemiology and End Results (SEER) database at Karmanos
Cancer Institute in Michigan. Cases who had an unknown country of birth were
first screened for race (White), followed by their place of birth.

Inclusion Criteria_: Only incident cases identified between January
13‘, 1994 and December 31"t 2001, histologically—confirmed, ages
of 20 to 79 years, and Polish-born were included in the study.
Exclpsion Criteria: Women diagnosed with cancer of other Sites
(except squamous and basal cell skin cancers) were ineligible for
the study. Also excluded from the study were cases of breast
cancer in situ and women who were non-Polish born.

Control Selection

Control selection was population-based and frequency-matched to the
cases by the two areas of residence and 5-year age group. Controls under 65
years of age were identified through Random Digit Dialing (RDD). The
distribution of Polish-born from the 1990 census was used to calculate sampling
fractions for RDD. Controls between the ages of 65-79 years of age were
identified from Medicare records of the Health Care Financing Administration.
Upon identifying potential controls, women were screened for place of birth and
only those who were Polish-born, had no history of breast cancer, and no
previous other cancers (except squamous and basal cell Skin cancers) were

included in the study.

20

Sample Size

The size of the initial dataset was 443 women, 140 cases and 303
controls. We further restricted our analytical dataset by dropping 6 (2 cases & 4
controls) study participants whose total adolescent physical activity per day was
less then 10 hours including sleep time. Upon review of their questionnaires we
noticed comments by the interviewer that these women were unable to recall
their complete daily physical activity during adolescence. In addition, we defined
adulthood as 18 years or older, thus 15 women (all controls) were also excluded
from analyses because they were under 18 years of age in 1989 (our proxy for
adulthood). The resulting analytic dataset contains 284 controls and 138 cases

for a total of 422 women.

Main Exposure: Adolescent Physical Activity

Women in the study were asked to report their non-occupational and
occupational physical activity during the time when they were aged 12- 13 years
old and during their adulthood years during 1985-89 (see Appendix B). The
process for derivation of total hours spent in a given activity was the same for
both time periods.

Adolescent Non-Occupational Physical Activity

Average number of hours spent doing activities such as: school athletic
participation, other recreational, household, occupational, and inactive/sedentary
actives including hours of sleep per day were reported. Women could report the

number of hours spent participating in the given activity in terms of per day, per

21

week, per month, and in some types of activities such as sports, per year. For the
purpose of these analyses, we converted all answers into hours spent per day.
To accomplish this, all hours of activities reported per week were divided by 7,
those reported per month were divided by 30.4 (an average number of
days/month), and activities reported per year were divided by 365.25. Hours of
each activity were then summed up to calculate the total number of hours each
woman reported per day. To calculate total MET-hs per day, the hours spent on
each type of activity that has its own unique MET value (Ainsworth 2000) were
multiplied by that value and added together.

Since interviewers were instructed not to question respondent’s answers,
the derived number of non-occupational hours/day when combined with
occupational hours of physical activity to create total hours per day, was both
under and over 24 hours. Therefore, the total reported hours per day needed to
be standardized to a 24-hour day for each individual. To calculate the
adjustment fraction, number of hours reported were divided by 24.
Subsequently, the total calculated MET-hs for each individual, were divided by
their same unique standardization fraction. This ensured that hours reported for
each activity were uniformly adjusted by the inverse of the standardization
fraction for each individual.

MET-h from stair climbing were not included in our total daily physical
activity variable for the following reason. Literature states that the average
person takes about only one second per step to climb stairs (Bassett 1997).

When this value is multiplied by the amount of stairs (or flights) climbed and

22

converted into a per day measure, MET-hs calculated from hours spent climbing
stairs is insignificant compared to all other activities. For example, if a woman
were to climb even 6 flights of stairs a day, with an average of 20 stairs per flight,
and a MET value for stair Climbing set to 8.5, her total MET-hs from stairs
contribute less than 0.3 of a MET-h/day. For this reason, we have not included
MET-hld from stair Climbing in our total daily MET-hs for either the adult period or
the adolescent period collected from our subjects.

Similar process was used to calculate non-occupational daily activity for
aduflhood.

Adolescent Occupational Physical Activity

Women reported the number of months or years they were employed
during adolescence and adulthood. For the time when they were 12-13 years
old, the maximum time was reported either as 24 months or 2 years. For
adulthood, 1985-1989, it was either 60 months or 5 years. Women had the
option of reporting their work activities, either in months or in years. For the
adolescent years, women who reported working during the summer, when not in
school, were assigned their work duration to be 3 months/year. Women were
then asked to report their average number of hours they worked in a given week.

Only 42 women reported being employed when they were 12-13 years
old. For the women who reported holding an occupation during their
adolescence, the number of hours worked per week ranged from 3 hours per
week to 84. It is important to note that for some women, the time period during

which they were 12-13 years old overlapped with the years that Europe was

23

involved in World War II. Several of our study participants indicated they had
worked in a concentration and/or labor camp during this time period. This often
resulted in seemly high reporting of hours worked per week as well as percent of
those hours spent in strenuous physical activity.

The typical work and school weeks in Poland prior to 1985 consisted of six
school and work days as opposed to the typical five-day work week found in the
United States. To incorporate this knowledge, we assumed that any women
reporting an adolescent or adult occupation worked this six-day work week thus
the number hours reported working per week was divided by 6 instead of 7 as for
non-occupational activities.

If a woman reported that she worked a certain number of months during
these two time periods, the number of months reported was multiplied by 26.1 to
convert the number of months into the number of work days during that time
period. The value 26.1 was calculated by multiplying the average number of
days in a month, 30.4 by (6/7) to account for the 6 day work week. The result
gave us the average number of work days that each woman worked during either
the 2 year adolescent time period or during the 5 years in adulthood.

If a woman reported that she worked a certain number of years in either
adolescence or adulthood, then the number of years was multiplied by 312
working days per year. The conversion of 312 was calculated by multiplying 52
weeks per year times the number of working days per week, 6. The result gave

us an average number of work days during the specific time period.

24

To calculate the average number of days worked/year, the above
calculated number of days was divided by 2 for adolescence and 5 for adulthood.
Since respondents were reporting the average number of hours worked/week,
that number was divided by 6 to account for the average number of hours
worked/day. To calculate the final number of average hours worked per day
during the given time period (adolescence or adulthood), the number of working
days/year was multiplied by the average hours worked/day and then divided by
365.25.

In the questionnaire, after reporting the number of hours worked per week,
women were asked to indicate either the number of hours, or percent of time,
they spent each week sitting, standing, walking without lifting, walking with lifting
less than 25 pounds, and doing heavy physical work. Each of these occupational
activities carries a different MET value. Sitting was assigned a MET value of 1.5,
standing was coded at 2.3 MET value, walking without lifting 3.0 MET value,
walking with some lifting (less than 11.5 kg or 25le) 4.0 MET value, and heavy
physical work 7.0 MET value (Ainsworth 2000). Percentage of time in each type
of activity was then multiplied by corresponding MET-h value. To obtain the total
MET-h for the occupational daily activity, the occupational daily hours were
weighted by percentage of time spent in that activity times the corresponding
MET-h and summed over all types of activities involved in their work day. The
standardization fraction that was required to adjust each day to 24 hours and
initially calculated by summing the hours of non—occupational and occupational

daily activities (described in the non-occupational physical activity section

25

above), was then applied to the total MET-hs calculated for the occupational

activity.

Total Physical Activity

After both occupational and non-occupational physical activity converted
to MET-hs per day were calculated, all activity for each time period was summed
up creating a variable of total daily physical activity for both the adolescent and
adult periods.

Creation of Adolescent MET-hld Tertiles

Tertiles were created using the total daily MET-hs as reported by controls.
The adolescent tertiles created from the data were as follows: low = 0-<45.9
MET-h/day, medium = 45.9-<55.7 MET-h/day, and high as greater than or equal
to 55.7 MET-h/day.

Creation of Adult MET-hld Tertiles

Adult physical activity tertiles were calculated by a process similar to those
for adolescent physical activity. The adult tertiles were as follows: low = 0 - <
48.8 MET-h/day, medium = 48.8 - <59.6 MET-hlday, and high as greater than or
equal to 59.6 MET-h/day.

Joint Levels of Adolescent and Adult Physical Activities

We created nine categories from the tertiles derived for both the
adolescent and adult physical activity. The nine categories are with adolescent
tertile level followed by the adult in the following notation: low/low, low/med

low/high, med/low, med/med, med/high, high/low, high/med, and high/high.

26

Other Important Breast Cancer Covariates:
Duration of Stay in the United States

During the interview, each woman provided a date when she migrated to
the United States (US) from Poland. From this information, a duration of stay
variable was created to account for potential participant acculturation. Duration
of stay was divided into three categories: women who migrated recently (after
1985), those who migrated before 1985 but have lived in the US. less than 10
years, and those migrated before 1985 but have lived in the US. for more than
10 years since.

First-degree Family History of Breast Cancer

Women that indicated their mother, sister, or daughter had been
diagnosed with breast cancer were considered to have a first-degree family
history of breast cancer. This variable was divided into a binary outcome of
positive or negative for familial history of breast cancer.

Adult BMI(1985-1989)

Women reported their respective height and weights during 1985-1989
and from this their adult BMI was calculated using the formula - weight (kg)/
[height]2 (m) (NHLBI 1998). Once BMI was calculated, women were further
categorized into one of four BMI groups: BMI<18.5 = undenlveight; BMI 18.5-<25
= normal weight; BMI 25-<30 = ovenrveight; and BMI>30 = obese (NHLBI 1998).

Some women could not recall their weight during the time period of 1985-

1989 and to include them in the analysis, their weight had to be extrapolated

27

from answers to Similar questions. In addition to weight during 1985-1989 weight
at age of 18, 30, 40, 50 or 60, were also collected and where appropriate used to
calculate the missing BMIS. For example, if a 25 year-old woman did not report
weight in 1985-1989 but did report weights for ages 18 and 30, her weight was
interpolated between those two values and appropriate BMI was calculated for
1985-1 989.

Adolescent Body Size

Height and weight for the time when participants were aged 12 to 13 years
were not obtained. Instead, women were asked to identify their closest body size
and shape from a series of nine pictures (Koprowski 2001). Women selecting
one of the first two figures were labeled as ‘underweight,’ those selecting one of
the subsequent three figures were considered ‘normal weight,” and women
reporting a figure of six through nine were grouped into an ‘ovenrveight’ category
(see Appendix C).

Oral Contraceptive(OC) Use

This covariate was divided as a binary variable. Women that had ever
used oral contraceptives of any kind were considered ‘users;’ all other women
were considered ‘non-users.’

Hormone Replacement Therapy (HRT) Use

Similar to oral contraceptive users, HRT users were divided into ‘users’
and ‘non-users.’ Women that had ever used any type HRT was placed in the
‘users’ category while all others were labeled as ‘non-users.’

Age at Menarche

28

Women were asked to report the age at which their natural menstruation
began. The median age at menarche for all women was about 14 years. Age at
menarche was then categorized into 4 different groups: less than 13 years, 13-
14, 14-15, and greater than or equal to 15 years. Four women did not report
their age at menarche and were assigned the median value of 14 to avoid
exclusion.

Age at First Full-term Pregnancy & Parity

Women whose pregnancies reached a gestational age of 24 weeks or
greater were considered to have a full-term birth. Women were then divided on
the age when they experienced their first full-term birth. The variable for age at
first full-term pregnancy was divided into four categories for analysis: nulliparous,
under the age of 22 years, 22 to less than 30 years, and 30 years and older.

Menopausal Status and Age at Menopause

Women who did not report natural or surgical menopause were
considered premenopausal (n=171). All other women (n=251) were considered
postmenopausal. For postmenopausal women, the average age at menopause
was 50.5 years. In addition, women who reported stoppage in menstrual cycles,
were older than 50 years of age, but did not report their age at menopause were
assigned an age at menopause equal to 50 years. Women who had surgeries
that removed both of their ovaries were also considered postmenopausal at the
age of their surgery.

Women were further divided into three different categories for age at

menopause: premenopausal, 50 years of age or less, and over 50 years of age.

29

Adult Total Caloric Intake

As part of the questionnaire women filled out a Food Frequency
Questionnaire (FFQ) to capture usual intakes of certain foods during 1985-89.
Calories assigned to an average serving for a specific food were then multiplied
by the frequency of consumption standardized to daily consumption. For a few
foods including breads, eggs, and alcohol, where the number of servings was
also reported, the frequency of consumption was multiplied by number of
servings, subsequently multiplied by calories per serving. Total caloric intake
was then calculated as the sum of the caloric intake from each food-type in the
questionnaire. Tertiles were then formed based on the distribution in controls:
less than 2047 calories/day; between 2047 and less than 2660 calories/day; and
over 2660 calories/day.

Adult Alcohol Consumption

Three questions in the F FQ asked about alcohol consumption. Intakes of
beer, wine, and hard liquor during 1985 -1989 were assessed and subsequently
calculated as weekly number of servings. Total number of weekly servings was
calculated as the sum of the weekly servings of beer, wine, and hard liquor.
Tertiles were formed from the total average weekly alcohol consumption: low =0
— <0.21 serving/week; medium = 0.21 - <1.09 servings/week; and high 3 1.09
servings/week.

Age in 1989

Controls were matched on to the distribution of age at diagnoses of cases

however, case identification ranged from 1994 through 2001. Therefore in 1989,

30

cases were older than controls. This variable was created in order to adjust the
models for the potential age difference between cases and controls during 1985-
1989. This variable (presented in Table 2) was divided into six categories : < 18
years; 18 - < 35 years; 35 - < 45 years; 45 - < 55 years; 55 - < 65 years; 3 65
years. Since we used the years between 1985 through 1989 as a proxy for
adulthood, women in the < 18 years of age category (n = 15 controls) were
excluded from all analyses since they had not started adulthood (defined as 3 18

years) between 1985 and 1989.

Statistical Analysis

Logistic regression always adjusted for matching variable (age and site)
was used to perform all univariate and multivariate modeling to assess the effect
the various risk factors on breast cancer risk.

All potential confounders were chosen based on literature review and
were all included as adjustment variables in the models. In order to make out
results for the effect of adolescent physical activity on breast cancer risk to be
comparable to previous studies, none of the potential confounders were
eliminated based on the change they had in point estimates for our main
exposure. To assess for potential effect modification by all risk factors
considered as potential confounders, interaction terms of these risk factors with
tertiles of adolescent physical activity were included in the models. The
Likelihood Ratio Test comparing the -2logL from the model that did not include

specific interaction, with the model that incorporated potential effect modification,

31

was calculated and assessed for significance at a p<0.05. If effect modification
was observed, analyses of the effect of adolescent physical activity on breast
cancer risk were performed in the separate strata of that variable.

Three types of models were run: first in addition to the main exposure of
adolescent physical activity, included adjustment for the design matching
variables (i.e. age and site); the second model, in addition, included adjustment
for all potential confounders, including age in 1989, except adult physical activity,
and the final model included all variables in Model 2 plus adjustment for adult

physical activity.

32

Chapter 4: Results

All tables from these analyses can be found in Appendix A. Tables 1 and
2 provide a distribution of the demographic and various risk factors for cases and

controls.

Descriptive Characteristics of Women in the Polish Women’s Health Study

Upon viewing Table 1, we were able to establish that relative to controls,
cases had a significantly higher proportion of first degree relatives with breast
cancer (p=0.05); a lower level of physical activity in adulthood (p=<0.01); a higher
proportion of later first full-term pregnancies (p=0.02); and a non-Significant lower
mean age at menarche (p=0.06).

Table 2 provides the distribution of age at diagnosis (reference ages for
controls) and age in 1989 for cases and controls by menopausal status. The age
distributions did not differ significantly between cases in either premenopausal
(p=0. 43) or postmenopausal (p=0.29) women. Over 92% of both cases and
controls did not spend their adolescent years (through age 13) in the United

States even after stratification by menopausal status.

Assessment of the Effect of Adolescent Physical Activity on Breast Cancer
_Ri_sl_<_

Table 3 reports three models for the effect of adolescent physical activity
on breast cancer risk. In the first model, the odds ratios (OR) and 95%

confidence intervals (95% CI) for the effect of adolescent physical activity on

33

breast cancer risk are adjusted only for the matching variables of site and age.
The results Show that women in the highest intensity group during adolescence
(greater than or equal to 55.7 MET-h per day) had a significant 44% reduction of
breast cancer risk (OR = 0.56, 95%CI: 0.36-0.95). Reduced, though not
significantly, risk was also observed for women who fell in the medium level of
physical activity during adolescence (OR = 0.71, 95%CI: 0.43-1.16). Adjusting
for known breast cancer risk factors and potential confounders shows very little
Change in the odds ratio (OR) of women at either medium or high level of
adolescent physical activity, however neither of the estimated odds ratios
reaches statistical significance; medium: OR=0.77, 95%CI: 0.43-1.35 and high:
OR=0.66, 95%CI:0.37-1.19.

Further adjustment for adult physical activity in tertiles does not
substantially change the observed effect medium and high adolescent physical
activity on breast cancer risk in Model 3 (medium: OR=0.76, 95%CI:0.43-1.37;
and high: OR= 0.71, 95%CI: 0.39-1.29). Additionally, we observed that in this
model the highest level of adult physical activity also has a significant protective
effect (OR=O.51, 95%CI:0.27-0.98) (data not shown).

A test for trend of adolescent physical activity was run for each model.
The trend in the first model was statistically significant with OR=0.75; p=0.03.
This implies that for each tertile increase in adolescent MET-h per day, a 24%
reduction in breast cancer risk is observed. The magnitude of the risk reduction
based on linear trend in both of the adjusted models (Model 2 and 3) was in the

same direction as to that from the unadjusted model, OR=0.81, p=0.16 and

34

OR=0.84, p=0.25 Model 2 and Model 3 respectively, however, these estimates
did not reach statistical significance at p=0.05.

Each of the potential confounders (listed in Table 1), were assessed for
potential effect modification in models evaluating the effect of adolescent
physical activity and breast cancer risk. Assessment for potential effect
modification by all of the adjustment risk factors included in Models 2 and 3
revealed a significant effect modification of the adolescent physical activity and
breast cancer risk relationship by menopausal status (LRT, chi square with 4
degrees of freedom = 13.107 p<0.05). Therefore, models assessing the
relationship between adolescent physical activity and breast cancer risk were run
within each strata of menopausal status. For premenopausal women, a
consistent significant reduction across all three models was observed for the
highest tertile of adolescent physical activity with a lesser reduction in risk
observed in women who were in the medium level of adolescent physical activity.
Odds ratios and confidence intervals for premenopausal women with high
adolescent physical activity were: Model 1 (OR=0.23, 95%CI:0.09-0.56), Model 2
(OR=0.14, 95%CI:0.04-0.44), Model 3 (OR=0.12, 95%CI:0.04-0.40). The
respective odds ratios for medium adolescent physical activity were: Model 1
(OR=0.70, 95%CI:0.33-1.47), Model 2 (OR=O.52, 95%CI:0.21-1.32), Model 3
(OR=0.44, 95%CI:0.17—1.18). The tests for linear trend for all three models in
premenopausal women were also statistically significant with the OR for the

linear trend in the final fully-adjusted model being OR=0.35, p<0.001 indicating

35

65% reduction in breast cancer risk per tertile increase in adolescent physical
activity.

For postmenopausal women, no consistent effect of increasing adolescent
physical activity (medium or high compared to low) on breast cancer risk was
observed. However, the inclusion of adjustment for adult physical activity, Model
3 (data not shown), shows an increased protection for postmenopausal breast
cancer risk for both medium and high levels of adult physical activity: medium —

OR = 0.70 95%CI: 0.32-1.55; high — OR = 0.31 95%CI: 0.13-0.77.

Assessment of the Joint Effect of Adolescent and Adult Physical Activity

on Breast Cancer Risk

The differing effects of adolescent and adult physical activities by
menopausal status motivated the evaluation of the joint effect of both adolescent
and adult physical activity on breast cancer risk by menopausal status. The ORs
in Table 4 are from fully-adjusted models within each menopausal strata (see
Table 3 - Model 2). Table 4 shows that high levels of physical activity during both
adolescence and adulthood relative to women with low levels of activity during
both time periods, provide a significant protection in premenopausal women (OR
= 0.11, 95%CI:0.02-0.67) and a non-Significant protection in postmenopausal
women (OR = 0.36, 95%CI:0.08-1.58).

For the combination of medium adolescent physical activity with high adult
physical activity, although point estimates were protective for both

premenopausal and postmenopausal women, statistical significance was

36

reached only in premenopausal women. The ORs were: premenopausal
(OR=0.13, 95%CI:0.02-0.95); postmenopausal (OR=0.27, 95%CI:0.05-1.59).
When the level of adolescent physical activity was low, no significant
protection from increasing high physical activity in adulthood was observed.
Results in Table 4 suggest that the greatest reduction in breast cancer risk is
achieved from the combination of high physical activity during both adolescence
and adulthood. However, high adult physical activity does not reduce breast
cancer risk when the individual had low level of adolescent physical activity.
These observations jointly imply that greatest reduction in overall breast cancer

risk can be achieved by increasing adolescent physical activity.

37

Chapter 5: Conclusions & Discussion

Discussion

This study provides evidence that increased physical activity (both medium
and high relative to low) during the adolescent years (specifically ages 12—13
years) even after adjustment for physical activity during adulthood, provides
Significant reduction in risk for breast cancer in premenopausal women. We
observed no such protective effect among postmenopausal women. However, in
analyzing the joint effect of adolescent and adult physical activity shows that
women who had medium or high adolescent activity as well as high adult activity
had Significant reduction in risk irrespective of menopausal status. Women with
low physical activity in adolescence did not gain significant breast cancer risk
reduction due to increased levels of activity in adulthood. Our findings point to
the importance of increased levels (medium or high) of physical activity during
adolescence in reducing breast cancer risk especially in premenopausal women.

Physical activity throughout life may modify a woman’s lifetime exposure to
circulating hormones therefore altering her breast cancer risk, however this is
only one hypothesis (Bemstein 1987 & Friedenreich 2002). The exact biological
mechanisms of adolescent physical activity on breast cancer risk are still not
completely understood. Further research on physical activity and breast cancer
risk is needed especially for different life periods (including adolescence) as
hormone levels change throughout a woman’s lifetime (Friedenreich 2001).

Findings from our analysis add to the hypothesis that increased adolescent

physical activity decreases risk of breast cancer in women especially for

38

premenopausal women. Few studies, however, have examined the effect of
adolescent physical activity on breast cancer risk across menopausal strata
(Monninkhof 2007). Results similar to our findings in premenopausal women
(decreased breast cancer risk with increased physical activity during
adolescence) have been previously observed in both cohort (Wyshak 2000 &
Maruti 2008) as well as case-control studies (Bemstein 1994, Dom 2002, &
Verloop 2000) on populations comprised of a majority of premenopausal women.
Our results, however, did not corroborate study findings from four other case-
Control studies (Chen 1997, Gammon 1998, Friedenreich 2001, & Steindorf
2003) and one cohort (Rockhill 1998) on populations containing large numbers of
premenopausal women. Measures such as higher number of episodes of
adolescent physical activity as well as increased adolescent MET-h of active
(sum of recreation, occupational, and household) physical activity were seen to
have inconclusive or increased findings on breast cancer risk when compared to
the least active group(s) (Rockhill 1998, Chen 1997, Gammon 1998,
Friedenreich 2001 , & Steindorf 2003). Only one review (Lagenos 2004)
observed a protective effect of increased adolescent physical activity on
premenopausal breast cancer risk.

Although not significant, our findings also suggest a null effect of increasing
adolescent physical activity on postmenopausal breast cancer risk. Three other
case-control studies (MCTieman 1996, Carpenter 1999, Moradi 2000) have also
found null effects of increasing adolescent physical activity on postmenopausal

breast cancer risk. However, six other case-control studies examining the effect

39

of adolescent physical activity, showed a decreasing risk of postmenopausal
breast cancer (Levi 1999, Mittendorf 1995, Taioli 1995, Friedenreich 2001, Dom
2002, & Shoff 2000). But measures of adolescent physical activity have not been
consistent between these studies. Similar to our methods, one previous case-
control study (Friedenreich 2001) examined adolescent physical activity using
MET-h. In that study, a significant protective effect of the highest versus the
lowest levels of adolescent physical activity was observed (OR = 0.73
95%CI:0.54-0.98). Both reviews of adolescent physical activity literature found a
reduced risk for postmenopausal women (Lagerros 2004 & Monninkhof 2007).

In addition, the results from our analyses are comparable to other published
data on Polish women. Although employing the use of hospital-based controls
and having low response rates for both cases (49.2%) and controls (41.4%),
Kruk observed a decreased overall breast cancer risk with increasing physical
activity (measured in MET-h/week/year) at ages 14 to 20 years old: OR: 0.37
95%CI: 0.23-0.60 (Kruk 2007). However, unlike our analyses, these results were
not provided specifically by menopausal strata.

Another component of our analyses, the joint effect of adolescent with adult
physical activity, has also been previously explored. Two studies examining
dichotomous responses (yes/no or ever/never) to participating in both adolescent
and adult physical activities have observed significant reduction in overall breast
cancer risk among women whom were active during both periods when
compared to women inactive in both periods: OR = 0.46, 95%CI:0.28-0.73 (Kruk

2007), OR = 0.47, and 95%CI:0.36-0.62 (Matthews 2001). Three studies

40

(Vertoop 2000, Matthews 2001, & Maruti 2008) further examined this joint effect
by menopausal strata. Similar to the results from our analyses, all three studies
found significant reduction in risk for premenopausal breast cancer when
comparing exercisers in both adolescence and adulthood with women who were
low or non-exercisers in both time periods: OR = 0.68, 95%CI: 0.53-0.87
(Verloop 2000), RR = 0.70 95%CI: 0.53-0.93 (Maruti 2008), & OR = 0.57 95%CI:
0.40-0.82 (Matthews 2001 ). Although our results for postmenopausal breast
cancer risk of high adolescent and adult physical activity (when compared to low
for both) were not significant (OR = 0.36 95%CI: 0.08-1.58; cases = 4; controls
=27) possibly due to small sample size, were Similar to a previous statistically
significant postmenopausal joint effect (OR 0.36 95%CI:0.24-0.55) which was
based on a larger sample size (cases = 43; controls = 95)(Matthews 2001).

However direct comparison of our results to all other studies on adolescent
physical activity and breast cancer risk poses a challenge. Only about 50% of
the previously published literature uses the measure of MET values (Carpenter
1999, Chen 1997, Lee 2001, Gammon 1998, Kruk 2007, Kruk 2007, Mittendorf
1995, Matthews 2001 , Marcus 1999, Friedenreich 2001 , Steindorf 2003, &
Maruti 2008) to assess adolescent physical activity while other use additional
methods such as ever vs. never (Wyshak 2000 & Mctieman 2003), episodic
(Shoff 2000 & Yang 2003) , hourly (MCTieman 1996, Bemstein 1994, Levi 1999,
Moradi 2000, & Dom 2002), energy expenditure (Hu 1997) or comparisons to
peers (Fn'sch 1985, Paffenbarger 1987, 8 Vertoop 2000). Along with the

measure of the adolescent physical activity, physical activity can be further

41

categorized into several different subtypes allowing each individual researcher
focus on one specific type of activity such as recreational, occupational, or
household. To date, our analyses appear to be unique in their measurement of
total daily adolescent physical activity (in MET-hlday) which includes inactive as
well as sleep time in the daily MET-h calculation.

As indicated in the hypothesized biological mechanisms above, energy
balance may play an important role in how adolescent physical activity effects
breast cancer risk. Data on both body size and usual daily dietary consumption
were also captured in this questionnaire and thus allowed us to make

adjustments for them.

Limitations

Limitations in this thesis are Similar to ones seen in other case-control studies
on adolescent physical activity and breast cancer risk. Even cohort studies
examining this association of adolescent physical activity and breast cancer risk
rely on recalled data of adolescent physical activity due to the large lag time
between adolescence and breast cancer occurrence. This limitation was
minimized by using memory prompts (i.e. picture cards and reference dates)
during the data collection for this study.

While two different time periods were captured for this data, one was
determined by a specific age (adolescent years of between ages 12-13) and the
other time period was assessed between stationary calendar years (adulthood

1985-1989) during which ages ranged from 18 to 72 years. As stated previously

42

in the methods section in Chapter 3, women whose period of adolescence
(defined as < 18 years) overlapped with the second time period (1985-89), were
eliminated from analyses to prevent multiple reporting between the two time
periods and to ensure more appropriate captures of adolescent and adulthood
exposures. Although the adult physical activity was captured for different ages
in 1985-1989, the information on all other variables that could be related to
energy expenditure, such as diet and BMI, were also recorded for this same time
period (1985-1989) to make appropriate adjustments in our models.

Finally, the questionnaire’s assessment of physical activity (for both
adolescent and adults) used in this analysis has not yet been validated in a
United States or Polish population; however, it was developed at the National
Institutes of Health specifically for this study. The dietary component, however,

has been validated in a United States population (unpublished data).

Strengths

Physical activity assessment in the Polish Women’s Health Study
questionnaire captured data from two separate time periods: during adolescence
(ages 12-13) and adulthood (years between 1985-1989). We were able to
assess the effect of physical activity during two different time periods
(adolescence and adulthood) as well as the joint effect between physical
activities in the two separate life periods. The questionnaire also captured a wide
range of non-physical activity information including most, if not all, other breast

cancer risk factors. By having access to information on all of these breast cancer

43

risk factors allowed us to account and adjust for them, thus giving us a more
accurate estimation of adolescent physical activity’s true effect on breast cancer
risk.

The Polish Women’s Health Study questionnaire captured all daily activity,
both inactive times (such as sleeping and sitting) as well as active times (such as
recreational and household physical activity). Since women were asked to report
hours for each activity, MET-h/day were able to be assigned for all recreational,
household, inactive, as well as occupational activity. To date this study appears
to be the first to analyze total daily MET-h of adolescent physical activity on
breast cancer risk. In addition, using our unique population, we captured
considerable heterogeneity in adolescent physical activity exposure ranging from
27 MET-h/day — 89 MET-h/day. Some of the high reporting of MET-h/day can be
attributed to the overlapping of World War II with the years some of the women
were aged 12 to 13 years. As previously mentioned several women included in
our dataset reported being in a labor and or concentration camp during their
adolescent years. This information was reported in the adolescent occupational
physical activity portion of the questionnaire. This resulted in some women
reporting extremely high levels of adolescent physical activity (up to about 82

MET-h/day).

Conclusions
In conclusion, these results enhance the current limited findings in the

literature about the effect of adolescent physical activity and breast cancer risk;

44

however, given that the majority of the women were not residing in the United
States during their adolescent years (ages 12-13), it will be imperative to
compare these results with those from the Polish component of the Polish
Women’s Health Study.

Given little evidence for harmful effects of adolescent physical activity on
breast cancer risk and the growing evidence for beneficial effects encourage this
topic to be further explored and findings implemented. This analysis suggests
that there should be more campaigns to increase the level of physical activity in
girls under the age of 21. In addition, to encourage adolescent physical activity,
maintenance of higher levels of physical activity should be of equal importance to

gain the largest benefit from physical activity on breast cancer risk reduction.

45

APPENDICES

46

APPENDIX A

Tables 1 — 4

47

Table 1: Descriptive Characteristics of Women In the Polish Women's Health Study

 

Cases Controls P- Vglue
(n=138) (n=284)

 

Site
Detroit 21% 24%
Chicago 79% 76%
Duration of Stay In the US.
Recent move 44% 45% 0.15
< 10 years 24% 25%
2 10 years 32% 30%
1st Degree Family Relative with Breast Cancer 14% 8% 0.05
Adolescent Physical Activity (MET-h/day)
Low (0 - < 45.9) 43% 33% 0.08
Medium (45.9 - < 55. 7) 30% 33%
High (2 55. 7) 26% 34%
Adult Physical Activity (MET-h/day)
Low (0 - < 48. 8) 43% 33% <0.01
Medium (48.8 - < 59. 6) 38% 34%
High ( 2 59. 6) 19% 33%
BMI 1985-89
Underweight 4% 6% 0.69
Normal 65% 59%
Overweight 24% 26%
Obese 7% 9%
Adolescent Bodysize
Underweight 59% 59% 0. 13
Normal 36% 32%
Overweight 4% 9%
Oral Contraceptive Use 17% 1 2% 0.27
HRT Use 12% 12% 0.59
Age at Menarche 13.5 13.8 0.06
Age at First Full Term Pregnancy/Parity
Nulliparous 12% 9% 0.02
< 22 years 22% 29%
22 - < 30 years 50% 54%
2 30 years 17% 8%
Age at menopause
Premenopausal 46% 39% 0. 77
5 50 years 26% 30%
> 50 years 28% 31%
Adult Total Caloric Intake (calories/day)
< 2047 40% 33% 0.34
2047 - < 2660 33% 34%
2 2660 28% 33%
Adult Alcohol Consumption (servingslweek)
0 - < 0.21 27% 33% 0.62
0.21 - <1.09 38% 34%
2 1.09 35% 33%

 

 

48

Table 2: Age Characteristics of Women in the Polish Women's Health Study

 

Premenopausal Postmenopausal
Cases Controls Cases Controls

(n=62) (n=109) (n=76) (n=175)

 

Age at Diagnosis
20 - < 35 years 10% 15% 0% 0%
35 - < 45 years 42% 39% 0% 0%
45 - < 55 years 47% 44% 20% 18%
55 - < 65 years 2% 3% 36% 25%
65 - < 75 years 0% 0% 36% 38%
2 75 years 0% 0% 9% 19%
p=0.86 p=0.06
Age in 1989*
18 - < 35 years 47% 58% 0% 0%
35 - < 45 years 48% 40% 20% 21%
45 - < 55 years 5% 2% 32% 33%
55 - < 65 years 0% 0% 33% 39%
2 65 years 0% 0% 16% 7%
p=0.43 p=0.29
Adolescence spent in US (%) 2% 3% 7% 2%

 

 

* Excludes 15 controls who were adolesents (<18 years of age) during 1985-89

49

”.088 6.81.8 .883 E .32 N .888 .o8uv-»8~ ~39. 9.9... 0.3.8 »._8 Eu .818 .8v

-8 .988 .8v-8 .88.. 82 c. 8.... .28» SN .28» 2v .96... .88.. 8.9.0. 8...... o... c. a...» .o 88...... ..e8» 08 .28»
omw ._mm:mqoccEc.& $3328.... 8 end .853. can .28.» RVNN .83» «av 8:93:35 .8530... E813. .9... .m mam .co>cc ..c>c.
cm: .53.. ....c>c.. .85. cm: 9.88828 _m..o .aoem .mo.P-_.N.o £9..me _.N.ov-o. 5.58328 .288 >282. .8: .85 .850 8.8....
.o 39...... 2:5. 352. .2: 5.5.8290 .858... 390259.... 3.963 288.03 .330 ....m_czcc>o ._mE.o.. 4.583823 =2m .58
..m.~ fit... it... .23 9.288... .m 8.. ....o8o .8833 2a ....»~ .818 .818 .8v-8 .8v-8 81.3 88 .8. 8.8.3..

 

 

mv.ouq ,8.ch meduq .8. new... .0. 839....
.8. 7R3 2.. .8. 78.3 3.. .8. 78.3 8... 82. .82....
. 8.8.83 8.. .888... 8.. . E. 78.3 8... 8.8 :8 N . 5...:
.8783 m»... .8. 78.3 m»... .878... 8... 8.2 .58 v - 88. E28...
.8. 8.. .8. 8.. .8. 8.. 8.8 6.8 v - 3 30..
E3333»

 

Law—aw OBOE—«00¢

 

 

53.x. Sodvq Edvq .8. can... .6. cams...
.3323 8... .8683 8... 5.98.... 8... .88. 8......
.8583 2... .3333 3... .8............ 8... 8... .58 N . 8.:
.2. 72.3 8... .8. 78.3 8... .8. 78.3 o»... 8.8 .58 v - 88. E28...
.8. 8.. .8. 8.. .8. 8.. 8.8 6.8 v - 3 26..
2.38.23
.umsamOp—OEEm
mmduq o » 6...... 864... .8. use. .6. 839....
.3783 8... .8. 78.3 5... .8683 8... use. .82....
.8. 78.3 E... 6.. 78.3 8... .8583 8... 5.8 .58 N . 8.:
.8783 8... .8783 R... 6.783 E... 8.8 :8 v - 88. E28...
.8. 8.. .8. 8.. .8. 8.. 8.8 3.8 v - 3 3....
“53203..
.6. .NI8 a .c °N|8 md I6 08.... MN... caloac
. E .39.. + N .8828 .oI..I.o_... nrlJl.II|8.8 8 N .28:

 

 

8.... .880 88.... w »8>.-..ms. .98 :. .<3 »..>..o< .8.m»..n. 288.8... .0. 8.8.2 88%.. .8 8.8.85 .8 as:

50

 

.688 68823... .868. 9.2... 2.28 »..8
8a .818 .818 .888 .818 .81... 8... c. 88 .28» on .28» 6V .26... .282. 8.28 8...... 9.. 2 88 .o 5.2.... .28»
of 888» omw 8330:8888. mmamaocmE 8 man .88.... 0mm .88» om-- 888.. wwv 8008053... 880088 E815. .2... .8 8mm .88..
.88. mm: .5... .88.. .88. mm: 82.808200 .80 .893 $0.730 89.3.3 3.010. :0_.0E:m..00 8:008 2.883 .8: .8». 8080
.885 .0 >88... 2E8. 858.. .2... ..Em8»>8>0 8&8: €982,800... 88.60.. 2808.08 .8850 .....m_838>0 ._mE.0: .Em83805. .55
.58 .38 .88. .38. 82 8282.. .m m8 .2280 668.5. a... ..8~ 88-8 .8v-8 88.8 83.8 .818. m8 2... 8.8.2 .

 

 

.8786. 86 R...
8.2.86. 8.. :3
.8678... 66.8 :8.
£26.86. :6 E...
.8686. 86 88
86.866. 86 6:.
I.0.I|2.8 a 8'80.
.IJImI..8~ .. ....

.8. 78.3
.8826.
.8886.

.8686.
.8826.
.8. 78.3

_0 Qamm

“Nd mEm
wmd FNB
nmé 2......

2.6 02%
_Rd 9.6—.
mud 0:2

mm 0080

 

afimm v - Wm... E2085.
.cmomo.00<

«End—.mfi

 

.8883
22.1.83
.3.

.8882...
.8883.
.8.

_0 o\..mm

mm... m to
mm... 5 FR
00... Nah...

N50 En
and c {or
ooé 229.

Md 0080

 

56¢ v - o. 26..

.28 m. 8....

.28 v - 28. 56.8.2

.28 v - 3 26..
82.820.82.88

.28 m. 56....

.28 v - 28. 56.8.2

.28 v - 3 33
88.30.82.021

 

 

 

mafiw _mmamaocms. .3 .83 .8280 889m :0 3.2.0... Regan. .33.. w .cmowm_ou< .0 .0th .50.. $0.833» 9: "v 03a...

51

APPENDIX B

Physical Activity Questionnaire from the Polish Women’s Health
Study

52

Physical Activity Questionnaire from t_he Polish Women’s Heajh
Study

I’ll start with asking questions about gym a_nd sports dun'ng mmchool vegr_s.

H1. When you were 12 and 13 years old, abut how many hours per day or per
week did you spend...(READ FIRST.NEXT ACTIVITY):

a. ...participating in gym classes as part of your school program?

Hours per
a. Day b. Week c. Month d. N/A (circle one)

b. ...participating in sports such as basketball, volleyball, soccer, or swimming as
part of a competitive team in or outside of school? Please include time spent at
practices.
_ Hours per
a. Day b. Week c. Month d. N/A (circle one)

Now, I’d like to ask you about recreational activities amt: transgortation...

H2. When you were 12 and 13 years old, abut how many hours per day or per
week did you spend...(READ FIRST.NEXT ACTIVITY) Then ask: and during
1985- 1989?

a. (SHOW CARD H-1) doing activities while you were sitting or reclining, such
as eating, watching television, reading, playing cards, sewing or knitting, or just
doing nothing? Please include the time you spent sitting in class.

Hours per
a. Day b. Week c. Month d. N/A (circle one)

b. (SHOW CARD H-2) participating in recreational sports, such as softball,
soccer, volleyball, skating, swimming, calisthenics, running or jogging, skiing or
cross-country skiing. Please do not include time spent training for a competitive
sports team or doing aerobics.

Hours per
a. Day b. Week c. Month d. Year e. NIA (circle one)

o. walking to get to places such as school, work or shopping, or for recreation
or exercise, including walking with persons, pets?

Hours per

53

a. Day b. Week c. Month d. NIA (circle one)

(I. bicycling to get to places such as school, work or shopping, or for
recreation. Please include time you spent exercising on a stationary bike.

Hours per
a. Day b. Week 0. Month d. NIA (circle one)

e. dancing or doing aerobic exercise?

Hours per
a. Day b. Week c. Month d. NIA (circle one)

Next, I will ask you to estimate the number of stairs you walked up on an average,
day, week, or month. You can give me you answer as the number of stairs
climbed in a day, week, or month, or as a number of floors climbed in a day,
week, or month. One floor is about 20 steps. Please remember to include stairs
in your home, at school, at work, or other places such as where you shop.

H3. When you were 12 and 13 years old, abut how many hours per day or per
week did you spend...(READ FIRST.NEXT ACTIVITY) Then ask: and during
1985- 1989?

a. Please estimate the number of stairs or floors that you walked up on an
average day, week, or month in you home, at, work, or other places.

Stairs or Floors (circle one) per
a. Day b. Week 0. Month d. N/A (circle one)

Next, I’d like to ask you about outdoor activities related to unpaid garden or farm
M done after regular working hours or another paying job. About outdoor
actives related to jobs that you were paid for and outdoor activities related to
work on your own or family fa_r1n_. I will ask later.

H4. When you were 12 and 13 years old, abut how many hours per day or per
week did you spend...(READ F IRST.NEXT ACTIVITY) Then ask: and during
1985- 1989?

a. (SHOW CARD H-3) moderate and heavy chores such as gardening,
sweeping, raking, mowing, digging, planting, weeding, shoveling, chopping wood,
milking cows, animal pens, feeding and caring for large animals (cow, pigs,
horses, etc)?

54

Hours per
a. Day b. Week c. Month d. Year e. NIA (circle one)

Next, I’d like to ask you about household activities. Again, please do not include
h_ousehold activities that you may have done as art of a lab for which youfi were
aid.
H5. When you were 12 and 13 years old, abut how many hours per day or per
week did you spend...(READ FIRST.NEXT ACTIVITY) Then ask: and during

1 985- 1 989?

a. ...light chores such as cooking, cleaning, washing dishes, making beds,
sweeping or vacuuming, mopping, doing laundry (by machine)m and light
shopping or standing in lines?
__ Hours per
a. Day b. Week c. Month d. NIA (circle one)

D. ...moderate or heavy chores such as heavy cleaning, scrubbing, hand
washing clothes with a wash-board or using an impeller-type machine, making
home repairs, and heavy shopping like caring and lifting groceries?

Hours per
a. Day b. Week c. Month d. N/A (circle one)

c. ...moderate or heavy chores related to children, others, or small pets such as
bathing, dressing , carrying children, and active play with children or pets?

Hours per
a. Day b. Week c. Month d. NIA (circle one)

H6. When you were 12 and 13 years old, abut how many hours per day or per
week did you spend...(READ FIRST.NEXT ACTIVITY) Then ask: and during
1 985- 1989?
a. ...sleeping? Please include naps taken during the day.

Hours per Day
Now I want to ask you just a few questions about your work relate physical
activity. Work includes any part-time or full-time jobs, any self-employment or
work for a family business, jobs in the military or on you own family farm, and any
jobs you may have held during Wortd War II.
H7. ...how many month or years were you employed?

Months or Years (circle one)

55

H8. During those months (NUMBER OF MONTHS OR YEARS FROM H7),
about how many hours per week did you usually work?

Hours per Week
Now I ’d like to know what percentage or how many hours of you time you spent
in each of these five kinds of actives. The total should add up to 100% or the
total number of hours you worked each week.
H9. When you were 12 and 13 years old, abut how many hours per day or per
week did you spend...(READ FIRST.NEXT ACTIVITY) Then ask: and during
1 985- 1 989?

a. ...sitting?
Hours or Percent (circle one)

D. ...standing?
Hours or Percent (circle one)

c. ...walking with no lifting?
_ Hours or Percent (circle one)
d. ...walking with some lifting (less than 11.5 kg or less than 25 lbs)?
_ Hours or Percent (circle one)
e. ...doing heavy physical work?
__ Hours or Percent (circle one)
H10. TOTAL: SHOULD EQUAL 100% OR HOURS PER WEEK FROM H8

Hours or Percent (circle one)

56

APPENDIX C

Adolescent Body Size Figures used in the Polish Women’s
Health Study

57

 

 

:22:

3:0 «923 939%
..pzmommuoas $23.50 >oom
TS. Emu

 

 

 

 

 

136 £62.. page; 5:8 9: s .82. 85 E 35 8m £88.02.

58

BIBILOGRAPHY

Adams-Campbell LL, Rosenberg L, Rao RS, Palmer JR. Strenuous physical
activity and breast cancer risk in African-American women. J Natl Med Assoc.
2001 Jul-Aug;93(7-8):267-75.

Ainsworth BE, Haskell WL, Whitt MC, lnrvin ML, Swartz AM, Strath SJ, O'Brien
WL, Bassett DR Jr, Schmitz KH, Emplaincourt PO, Jacobs DR Jr, Leon AS.
Compendium of physical activities: an update of activity codes and MET
intensities. Med Sci Sports Exerc. 2000 Sep;32(9 Suppl):S498-504.

American Cancer Society (ACS). How Many Women Get Breast Cancer? 2007.
www.cancer.org, accessed 8/21/2008.

American Cancer Society (ACS). What is Breast Cancer? 2007.
www.cancer.org, accessed 8/21/2008.

Bassett DR, Vachon JA, Kirkland AO, Howley ET, Duncan GE, Johnson KR.
Energy cost of stair climbing and descending on the college alumnus
questionnaire. Med Sci Sports Exerc. 1997 Sep;29(9):1250-4.

Bebenek M, Pudelko M, Blaszczyk M. Breast cancer incidence and mortality in
Lower Silesia (Poland) between 1984 and 2003 - trends and perspectives.
CEJMed 2 (2)2007. 208-215.

Berkey CS, Frazier AL, Gardner JD, Colditz GA. Adolescence and breast
carcinoma risk. Cancer. 1999 Jun 1;85(11):2400-9.

Bernstein L, Ross RK, Lobo RA, Hanisch R, Krailo MD, Henderson BE. The
effects of moderate physical activity on menstrual cycle patterns in adolescence:
implications for breast cancer prevention. Br J Cancer. 1987 Jun;55(6): 681-5.

Bernstein L. Epidemiology of endocrine-related risk factors for breast cancer. J
Mammary Gland Biol Neoplasia. 2002 Jan;7(1):3-15.

Bernstein L, Patel AV, Ursin G, Sullivan-Halley J, Press MF, Deapen D, Berlin
JA, Daling JR, McDonald JA, Norman SA, Malone KE, Strom BL, Lift J, Folger
SG, Simon MS, Burkman RT, Marchbanks PA, Weiss LK, Spirtas R. Lifetime
recreational exercise activity and breast cancer risk among black women and
white women. J Natl Cancer Inst. 2005 Nov 16;97(22):1671-9.

Bernstein L, Henderson BE, Hanisch R, Sullivan-Halley J, Ross RK. Physical
exercise and reduced risk of breast cancer in young women. J Natl Cancer Inst.
1994 Sep 21 ;86(18):1403-8.

59

Bondy ML, Vogel VG, Halabi S, Lustbader ED. Identification of women at
increased risk for breast cancer in a population-based screening program.
Cancer Epidemiol Biomarkers Prev. 1992 Jan-Feb;1(2):143-7.

Carmichael AR. Obesity and prognosis of breast cancer. Obes Rev. 2006
Nov;7(4):333-40.

Carpenter CL, Ross RK, Paganini-Hill A, Bernstein L. Lifetime exercise activity
and breast cancer risk among post-menopausal women. Br J Cancer. 1999
Aug;80(11):1852-8.

Chavez-MacGregor M, van Gils CH, van der Schouw YT, Monninkhof E, van
Noord PA, Peeters PH. Lifetime cumulative number of menstrual cycles and
serum sex hormone levels in postmenopausal women. Breast Cancer Res Treat.
2008 Mar;108(1):101-12.

Chen J, Pee D, Ayyagari R, Graubard B, Schairer C, Byrne C, Benichou J, Gail
MH. Projecting absolute invasive breast cancer risk in white women with a

model that includes mammographic density. J Natl Cancer Inst. 2006 Sep
6;98(17):1215-26.

Chen CL, White E, Malone KE, Daling JR. Leisure-time physical activity in
relation to breast cancer among young women (Washington, United States).
Cancer Causes Control. 1997 Jan;8(1):77-84.

Cho E, Spiegelman D, Hunter DJ, Chen WY, Zhang SM, Colditz GA, Willett WC.
Premenopausal intakes of vitamins A, C, and E, folate, and carotenoids, and risk
of breast cancer. Cancer Epidemiol Biomarkers Prev. 2003 Aug;12(8):713-20.

Cleary MP, Maihle NJ. The role of body mass index in the relative risk of
developing premenopausal versus postmenopausal breast cancer. Proc Soc
Exp Biol Med. 1997 Oct;216(1):28-43.

Colditz GA. Epidemiology and prevention of breast cancer. Cancer Epidemiol
Biomarkers Prev. 2005 Apr;14(4):768-72.

Colditz GA, Frazier AL. Models of breast cancer show that risk is set by events of
early life: prevention efforts must shift focus. Cancer Epidemiol Biomarkers Prev.
1995 Jul-Aug;4(5):567-71 .

D'Avanzo B, Nanni 0, La Vecchia C, Franceschi S, Negri E, Giacosa A, Conti E,

Montella M, Talamini R, Decarli A. Physical activity and breast cancer risk.
Cancer Epidemiol Biomarkers Prev. 1996 Mar;5(3):155-60.

6O

De Assis S, Hilakivi-Clarke L. Timing of dietary estrogenic exposures and breast
cancer risk. Ann N Y Acad Sci. 2006 Nov;1089:14-35.

Dirx MJ, van den Brandt PA, Goldbohm RA, Lumey LH. Diet in adolescence and
the risk of breast cancer: results of the Netherlands Cohort Study. Cancer
Causes Control. 1999 Jun;10(3):189-99.

Dorn J, Vena J, Brasure J, Freudenheim J, Graham S. Lifetime physical activity
and breast cancer risk in pre- and postmenopausal women. Med Sci Sports
Exerc. 2003 Feb;35(2):278-85.

Dunham DP, Czysczon A, Chavez N, Piorkowski J, Persky V. Dietary differences
among women of Polish descent by country of birth and duration of residency in
the United States. Ethn Dis. 2004 Spring;14(2):219—26.

Enger SM, Ross RK, Paganini-Hill A, Carpenter CL, Bernstein L. Body size,
physical activity, and breast cancer hormone receptor status: results from two
case-control studies. Cancer Epidemiol Biomarkers Prev. 2000 Jul;9(7):681-7.

Forsum E, Lof M, Schoeller DA. Calculation of energy expenditure in women
using the MET system. Med Sci Sports Exerc. 2006 Aug;38(8):1520-5.

Frazier AL, Li L, Cho E, Willett WC, Colditz GA. Adolescent diet and risk of
breast cancer. Cancer Causes Control. 2004 Feb;15(1):73-82.

Friedenreich CM, Courneya KS, Bryant HE. Influence of physical activity in
different age and life periods on the risk of breast cancer. Epidemiology. 2001
Nov;12(6):604-12.

Friedenreich CM, Orenstein MR. Physical activity and cancer prevention:
etiologic evidence and biological mechanisms. J Nutr. 2002 Nov;132(11
Suppl):3456S-3464S.

Frisch RE, Gotz-Welbergen AV, McArthur JW, Albright T, Witschi J, Bullen B,
Birnholz J, Reed RB, Hermann H. Delayed menarche and amenorrhea of
college athletes in relation to age of onset of training. JAMA. 1981 Oct 2;
246(14):1559-63.

Frisch RE, Wyshak G, Albright NL, Albright TE, Schiff I. Lower prevalence of
non-reproductive system cancers among female former college athletes. Med Sci
Sports Exerc. 1989 Jun;21(3):250-3.

Gail MH, Brlnton LA, Byar DP, Corle DK, Green SB, Schairer C, Mulvihill JJ.
Projecting individualized probabilities of developing breast cancer for white
females who are being examined annually. J Natl Cancer Inst. 1989 Dec
20;81(24):1879-86.

61

Gammon MD, Schoenberg JB, Britton JA, Kelsey JL, Coates RJ, Brogan D,
Potischman N, Swanson CA, Daling JR, Stanford JL, Brinton LA. Recreational
physical activity and breast cancer risk among women under age 45 years. Am J
Epidemiol. 1998 Feb 1;147(3):273-80.

Gronwald J, Byrski T, Huzarski T, Cybulski C, Sun P, Tulman A, Narod SA,
Lubinski J. Influence of selected lifestyle factors on breast and ovarian cancer

risk in BRCA1 mutation carriers from Poland. Breast Cancer Res Treat. 2006
Jan;95(2):105-9.

Hoffman-Goetz L, Apter D, Demark-Wahnefried W, Goran Ml, McTiernan A,
Reichman ME. Possible mechanisms mediating an association between
physical activity and breast cancer. Cancer. 1998 Aug 1;83(3 Suppl):621-8.

Hu YH, Nagata C, Shimizu H, Kaneda N, Kashiki Y. Association of body mass
index, physical activity, and reproductive histories with breast cancer: a case-
control study in Gifu, Japan. Breast Cancer Res Treat. 1997 Mar;43(1):65-72.

Hunter DJ, Willett WC: Diet, body size, and breast cancer. Epidemiol Rev 1993,
15:1 10-132.

Irwin ML. Randomized controlled trials of physical activity and breast cancer
prevention. Exerc Sport Sci Rev. 2006 Oct;34(4):182-93.

Jasienska G, Ziomkiewicz A, Thune I, Lipson SF, Ellison PT. Habitual physical
activity and estradiol levels in women of reproductive age. Eur J Cancer Prev.
2006 Oct;15(5):439-45.

Kodama M, Kodama T. Adolescence, a critical stage for the genesis of female
cancers (review). Anticancer Res. 1981;1(2):93-9.

Koprowski C, Coates RJ, Bernstein L. Ability of young women to recall past body
size and age at menarche. Obes Res. 2001 Aug;9(8):478—85.

Kruk J. Association of Lifestyle and Other Risk Factors with Breast Cancer
According to Menopausal Status: A Case-Control Study in the Region of
WesternPomerania (Poland). Asian Pac J Cancer Prev. 2007 Oct-Dec;8(4):513-
24.

Kruk Ja. Lifetime physical activity and the risk of breast cancer: a case-control
study. Cancer Detect Prev. 2007;31(1):18-28.

Kruk J". Self-reported childhood physical activity and breast cancer in adulthood.
Asian Pac J Cancer Prev. 2007 Jan-Mar;8(1):63-8.

62

Kruk J°. Association of Lifestyle and Other Risk Factors with Breast Cancer
According to Menopausal Status: A Case-Control Study in the Region of
WestemPomerania (Poland). Asian Pac J Cancer Prev. 2007 Oct-Dec;8(4):513-
24.

Kruk J, Aboul—Enein HY. Physical activity in the prevention of cancer. Asian Pac
J Cancer Prev. 2006 Jan-Mar;7(1):11-21.

Lagerros YT, Hsieh SF, Hsieh CC. Physical activity in adolescence and young
adulthood and breast cancer risk: a quantitative review. Eur J Cancer Prev. 2004
Feb;13(1):5-12.

Lamartiniere CA, Murrill WB, Manzolillo PA, Zhang JX, Barnes S, Zhang X, Wei
H, Brown NM. Genistein alters the ontogeny of mammary gland development

and protects against chemically-induced mammary cancer in rats. Proc Soc Exp
Biol Med. 1998 Mar;217(3):358-64.

Land CE. Studies of cancer and radiation dose among atomic bomb survivors.
The example of breast cancer. JAMA. 1995 Aug 2;274(5):402-7.

Land CE, Tokunaga M, Koyama K, Soda M, Preston DL, Nishimori l, Tokuoka S.
Incidence of female breast cancer among atomic bomb survivors, Hiroshima and
Nagasaki, 1950-1990. Radiat Res. 2003 Dec; 1 60(6):707-1 7.

Lee IM, Cook NR, Rexrode KM, Buring JE. Lifetime physical activity and risk of
breast cancer. Br J Cancer. 2001 Sep 28;85(7):962-5.

Levi F, Pasche C, Lucchini F, La Vecchia C. Occupational and leisure time
physical activity and the risk of breast cancer. Eur J Cancer. 1999
May;35(5):775-8.

Limer JL, Speirs V. Phyto-oestrogens and breast cancer chemoprevention.Breast
Cancer Res. 2004;6(3):1 19-27.

Linos E, Willett WC, Cho E, Colditz G, Frazier LA. Red Meat Consumption
during Adolescence among Premenopausal Women and Risk of Breast Cancer.
Cancer Epidemiol Biomarkers Prev. 2008 Jul 31. [Epub ahead of print]

Magnusson CM, Roddam AW, Pike MC, Chilvers C, Crossley B, Hermon C,
McPherson K, Peto J, Vessey M, Beral V. Body fatness and physical activity at
young ages and the risk of breast cancer in premenopausal women. Br J Cancer.
2005 Oct 3;93(7):817-24.

Marcus PM, Newman B, Moonnan PG, Millikan RC, Baird DD, Qaqish B,

Sternfeld B. Physical activity at age 12 and adult breast cancer risk (United
States). Cancer Causes Control. 1999 Aug;10(4):293-302.

63

Margolis KL, Mucci L, Braaten T, Kumle M, Trolle Lagerros Y, Adami HO, Lund
E, Weiderpass E. Physical activity in different periods of life and the risk of breast
cancer: the Norwegian-Swedish Women's Lifestyle and Health cohort study.
Cancer Epidemiol Biomarkers Prev. 2005 Jan;14(1):27-32.

Maruti SS, Willett WC, Feskanich D, Rosner B, Colditz GA. A prospective study
of age-specific physical activity and premenopausal breast cancer. J Natl Cancer
Inst. 2008 May 21 ;100(10):728-37. Epub 2008 May 13.

Matthews CE, Shu XO, Jin F, Dai Q, Hebert JR, Ruan ZX, Gao YT, Zheng W.
Lifetime physical activity and breast cancer risk in the Shanghai Breast Cancer
Study. Br J Cancer. 2001 Apr 6;84(7):994-1001.

McPherson K, Steel CM, Dixon JM. ABC of breast diseases. Breast cancer-
epidemiology, risk factors, and genetics. BMJ. 2000 Sep 9;321(7261):624-8.

McTiernan A, Stanford JL, Weiss NS, Daling JR, Voigt LF. Occurrence of breast
cancer in relation to recreational exercise in women age 50-64 years.
Epidemiology. 1996 Nov;7(6):598-604.

McTiernan A, Kooperberg C, White E, Wilcox S, Coates R, Adams-Campbell LL,
Woods N, Ockene J; Women's Health Initiative Cohort Study. Recreational
physical activity and the risk of breast cancer in postmenopausal women: the
Women's Health Initiative Cohort Study. JAMA. 2003 Sep 10;290(10):1331-6.

Messina MJ, Loprinzi CL. Soy for breast cancer survivors: a critical review of the
literature. J Nutr. 2001 Nov;131(11 Suppl):3095$-1088.

Mittendorf R, Longnecker MP, Newcomb PA, Dietz AT, Greenberg ER, Bogdan
GF, Clapp RW, Willett WC. Strenuous physical activity in young adulthood and
risk of breast cancer (United States). Cancer Causes Control. 1995 Jul;6(4):347-
53.

Monninkhof EM, Elias SG, Vlems FA, van der Tweel l, Schuit AJ, Voskuil DW,
van Leeuwen FE; TF PAC. Physical activity and breast cancer: a systematic
review. Epidemiology. 2007 Jan;18(1):137-57.

Moradi T, Nyrén O, Zack M, Magnusson C, Persson I, Adami HO. Breast cancer
risk and lifetime leisure-time and occupational physical activity (Sweden). Cancer
Causes Control. 2000 Jul;11(6):523-31.

National Institutes of Health NHL , and Blood Institute (NHLBI). Clinical
Guidelines on the Identification, Evaluation, and Treatment of Overweight and
Obesity in Adults: the evidence report. Rockville: US. Department of Health and
Human Services, Public Health Service. 1998: 1-228.

64

 

Nelson NJ. Migrant studies aid the search for factors linked to breast cancer risk.
J Natl Cancer Inst. 2006 Apr 5;98(7):436-8.

Newman B, Moorman PG, Millikan R, Qaqish BF, Geradts J, Aldrich TE, Liu ET.
The Carolina Breast Cancer Study: integrating population-based epidemiology
and molecular biology. Breast Cancer Res Treat. 1995 Jul;35(1):51—60.

Okasha M, McCarron P, Gunnell D, Smith GD. Exposures in childhood,
adolescence and early adulthood and breast cancer risk: a systematic review of
the literature. Breast Cancer Res Treat. 2003 Mar;78(2):223-76.

Pryor M, Slattery ML, Robison LM, Egger M. Adolescent diet and breast cancer
in Utah. Cancer Res. 1989 Apr 15;49(8):2161-7.

Paffenbarger RS Jr, Hyde RT, Wing AL. Physical activity and incidence of cancer
in diverse populations: a preliminary report. Am J Clin Nutr. 1987 Jan;45(1
Suppl):312-7.

Reeves GK, Pirie K, Beral V, Green J, Spencer E, Bull D; Million Women Study
Collaboration. Cancer incidence and mortality in relation to body mass index in
the Million Women Study: cohort study. BMJ. 2007 Dec 1;335(7630):1134.

Rockhill B, Willett WC, Hunter DJ, Manson JE, Hankinson SE, Spiegelman D,
Colditz GA. Physical activity and breast cancer risk in a cohort of young women.
J Natl Cancer Inst. 1998 Aug 5;90(15):1155-60.

Ronckers CM, Erdmann CA, Land CE. Radiation and breast cancer: a review of
current evidence. Breast Cancer Res. 2005;7(1):21-32.

Sainsbury JR, Anderson TJ, Morgan DA. ABC of breast diseases: breast cancer.
BMJ. 2000 Sep 23;321(7263):745-50.

Santen RJ, Boyd NF, Chlebowski RT, Cummings S, Cuzick J, Dowsett M, Easton
D, Forbes JF, Key T, Hankinson SE, Howell A, lngle J; Breast Cancer Prevention
Collaborative Group. Critical assessment of new risk factors for breast cancer:
considerations for development of an improved risk prediction model. Endocr
Relat Cancer. 2007 Jun;14(2):169-87.

Schedin PJ, Byers T. Adolescent diet and the risk of breast cancer in adulthood:
a role for vitamin A? Nutrition. 1997 Oct;13(10):924-5.

Shin A, Matthews CE, Shu XO, Gao YT, Lu W, Gu K, Zheng W. Joint effects of

body size, energy intake, and physical activity on breast cancer risk. Breast
Cancer Res Treat. 2008 Jan 29.

65

 

Shoff SM, Newcomb PA, Trentham-Dietz A, Remington PL, Mittendorf R,
Greenberg ER, Willett WC. Early-life physical activity and postmenopausal breast
cancer: effect of body size and weight change. Cancer Epidemiol Biomarkers
Prev. 2000 Jun;9(6):591-5.

Shu XO, Jin F, Dai Q, Wen W, Potter JD, Kushi LH, Ruan Z, Gao YT, Zheng W.
Soyfood intake during adolescence and subsequent risk of breast cancer among
Chinese women. Cancer Epidemiol Biomarkers Prev. 2001 May;10(5):483-8.

Sprague BL, Trentham-Dietz A, Newcomb PA, Titus-Ernstoff L, Hampton JM,
Egan KM. Lifetime recreational and occupational physical activity and risk of in
situ and invasive breast cancer. Cancer Epidemiol Biomarkers Prev. 2007
Feb;16(2):236-43.

Surveillance Epidemiology and End Results (SEER). SEER Stat Fact Sheet —
Cancer of the Breast. www.seer.cancerppy, accessed 8/21/08.

Steindorf K, Schmidt M, Kropp S, Chang-Claude J. Case-control study of
physical activity and breast cancer risk among premenopausal women in
Germany. Am J Epidemiol. 2003 Jan 15;157(2):121-30.

Taioli E, Barone J, Wynder EL. A case-control study on breast cancer and body
mass. The American Health Foundation-Division of Epidemiology. Eur J Cancer.
1995;31A(5):723-8.

Thanos J, Cotterchio M, Boucher BA, Kreiger N, Thompson LU. Adolescent
dietary phytoestrogen intake and breast cancer risk (Canada). Cancer Causes
Control. 2006 Dec;17(10):1253-61.

Tretli S, Gaard M. Lifestyle changes during adolescence and risk of breast
cancer: an ecological study of the effect of World War II in Nonlvay. Cancer
Causes Control. 1996. 7: 507-512.

University of Colorado Cancer Center (UCCC). What is Breast Cancer?
http://www.uccc.info/conditions/cancerlbreast-cancerlwhat-is-breast-cancer.aspx,
accessed 8/21/2008.

Verloop J, Rookus MA, van der Kooy K, van Leeuwen FE. Physical activity and
breast cancer risk in women aged 20-54 years. J Natl Cancer Inst. 2000 Jan
19;92(2):128-35.

Warri A, Saarinen NM, Makela S, Hilakivi-Clarke L. The role of early life

genistein exposures in modifying breast cancer risk. Br J Cancer. 2008 May
6;98(9):1485-93. Epub 2008 Apr 8.

66

World Cancer Research Fund /American Institute for Cancer Research
(WCRF/AICR). Food, Nutrition, Physical Activity, and the Prevention of Cancer: a
Global Perspective. Washington DC: AICR, 2007.

Wu MH, Chou YC, Yu JC, Yu CP, Wu CC, Chu CM, Yang T, Lai CH, Hsieh CY,
You SL, Chen CJ, Sun CA. Hormonal and body-size factors in relation to breast

cancer risk: a prospective study of 11,889 women in a low-incidence area. Ann
Epidemiol. 2006 Mar;16(3):223-9.

Wyshak G, Frisch RE.Breast cancer among former college athletes compared to
non-athletes: a 15-year follow-up. Br J Cancer. 2000 Feb;82(3):726-30.

Yang D, Bernstein L, Wu AH. Physical activity and breast cancer risk among
Asian-American women in Los Angeles: a case-control study. Cancer. 2003 May
15;97(10):2565-75.

67