University of Southern California Dissertations and Theses

DEXA measures of body fat percentage and acute phase proteins among breast cancer survivors

(USC Thesis Other)

DEXA measures of body fat percentage and acute phase proteins among breast cancer survivors

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

DEXA MEASURES OF BODY FAT PERCENTAGE AND ACUTE PHASE PROTEINS
AMONG BREAST CANCER SURVIVORS

by
Anne Dee

A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(APPLIED BIOSTATICS AND EPIDEMIOLOGY)

May 2010

Copyright 2010 Anne Dee
ii

Acknowledgements
This thesis was completed under the direction of Dr. Roberta McKean-Cowdin and edited
by HEAL investigators including Drs. Marian L. Neuhouser, Cornelia Ulrich, Katherine
Baumgartner, Richard N. Baumgartner, Anne McTiernan, Catherine M. Alfano, Rachel Ballard-
Barbash, and Leslie Bernstein.
I would also like to thank Dr. Anna Wu for her suggestions in the preparation of the
manuscript.

iii

Table of Contents
Acknowledgements
List of Tables
List of Figures
Abstract
Chapter 1: Introduction
Chapter 2: Material and Methods
Study Population
Clinical Variables
Questionnaire Variables

Chapter 3: Statistical Analyses
Chapter 4: Results
Chapter 5: Discussion
Bibliography
iii
iv
v
vi
1
3
3
4
5
7
9
21
26

iv

List of Tables
Table 1: Descriptive characteristics of 138 female breast cancer survivors in HEAL
Study.
Table 2a: Adjusted geometric means and 95% CI of CRP stratified by obesity status
at follow-up (30 months post-diagnosis)(n=134).

Table 2b: Adjusted geometric means and 95% CI of SAA stratified by obesity status
at follow-up (30 months post-diagnosis)(n=134).

Table 3a: Linear regression models for ln(CRP) on body fat percentage for study
population of breast cancer survivors from HEAL (n=134).

Table 3b: Regression models for ln(SAA) on body fat percentage for study
population of breast cancer survivors from HEAL (n=134).

9

11

12

16

17

v

List of Figures
Figure 1: Scatterplot and predicted regression line of ln(CRP) on centered
body fat percentage in 134 breast cancer survivors HEAL study
and the predicted values of ln(CRP) based on body fat
percentage while controlling for age, caloric intake (kcal),
weight difference(kg), and NSAID use (Correlation (95% CI):
0.52 (0.39-0.63) Adj. r
2
= .33).
Figure 2: Scatterplot and predicted regression lines of ln(SAA) on centered
body fat percentage of 138 breast cancer survivors from HEAL
study and the predicted values of ln(SAA) based on body fat
percentage while controlling for age, history of arthritis, caloric
intake (kcal), weight difference(kg), and NSAID use
(Correlation (95% CI): 0.42 (0.27-0.54) Adj. r
2
=.20).

Figure 3: ln(CRP) and ln(SAA) on BMI or Percent Body fat.
14

15

19

vi

Abstract
Objective: To examine the association between body fat percentage as measured by DEXA and 2
systemic inflammatory markers, C-reactive protein (CRP) and Serum amyloid A protein (SAA)
in breast cancer survivors and its modulation by lifestyle factors.

Design: Cross-sectional measures from a longitudinal study of breast cancer survivors.
Subjects: 138 non-Hispanic White and Hispanic breast cancer survivors participating in the
Washington and New Mexico components of the Health, Eating, Activity, and Lifestyle (HEAL)
Study.

Measures: DEXA measures of body fat percentage and circulating levels of 2 acute phase
proteins (CRP, SAA) approximately 30 months after breast cancer diagnosis.

Results: Circulating levels of CRP and SAA were associated with increased adiposity as
measured by DEXA. Breast cancer patients with body fat ≥35% had significantly higher CRP
and SAA levels compared to non-obese patients. The data suggest that geometric mean levels of
CRP are lower among obese women taking non-steroidal anti-inflammatory drugs (NSAIDs)
compared to obese women not taking them and higher among women eating ≥ 2000 kcal/day,
although neither of these associations was statistically significant. Physical activity and other
lifestyle factors were not significant modifiers of the associations of CRP and SAA with
percentage body fat.

Conclusion: Women with body fat ≥35% had higher mean levels of systemic inflammatory
markers than women with lower body fat. Modest changes in lifestyle (e.g. taking NSAIDs or
vii

decreasing caloric intake ) may help to reduce circulating inflammatory markers, which may
beneficially impact survival.

1

Chapter 1: Introduction
C-reactive protein (CRP) and serum amyloid A (SAA) are nonspecific acute-phase
proteins that increase during systemic inflammation.
1
The high levels of these proteins among
the obese may indicate a low-grade chronic inflammatory condition, which could result from the
expansion of blood vessels and other supporting structures necessary for growth of adipose
tissue
2
. Obese individuals have been shown to have a higher circulating levels of pro-
inflammatory cytokines (e.g. TNF-α and IL-6) and acute-phase proteins (including CRP and
SAA)
3
. The pro-inflammatory cytokine IL-6 has a dramatic impact on the secretion of acute-
phase proteins by the liver and may result in a 10 to 100 fold increase in circulating CRP and
SAA
4
. The inflammatory process is considered critical to both the development and progression
of cancer
5
. Elevated CRP concentration has been associated with the development and
progression of a variety of cancers including breast cancer.
6
Elevated CRP and SAA have been
associated with lower breast cancer survival and with stage of disease
7,8,10
.
Previous studies evaluating the relationship between adiposity and CRP and SAA levels
have used anthropometric measures of obesity including BMI, waist circumference, and
bioelectrical impedance
2, 9-11
. In this study, we use a measure of body fat obtained from Dual
Caloric Intake X-ray Absorptiometer (DEXA) which may be a more valid and reliable estimate of
total and discrete depots of fat . DEXA includes measures of body mineral mass, lean mass, and
fat mass which is especially important when measuring body composition of postmenopausal
women who may have less bone and muscle mass than younger women
12,13
.
In this analysis, we investigated the relationship between body fat percentage and
systemic inflammatory markers among a cohort of Hispanic and non-Hispanic White breast
cancer survivors enrolled in the HEAL (Health, Eating, Activity and Lifestyle) Study and the
possible modifying effects of lifestyle changes including exercise, weight loss and use of non-
steroidal anti-inflammatory drugs (NSAIDs), such as aspirin. While the biological mechanism is
2

not known, elevated concentrations of post-diagnostic serological CRP and SAA and high post-
diagnostic body mass index have been associated with poor prognosis in breast cancer patients
14
.
Whether the association between obesity and CRP/SAA can be modified by lifestyle changes is
not known.

3

Chapter 2: Material and Methods
Study Population
The data for this analysis were collected for the HEAL Study, a population-based
prospective cohort study that includes women diagnosed with in-situ to stage IIIa breast
cancer from 1996 through 1999. Baseline data was collected within the first year after
diagnosis, on average 7.5 months post diagnosis. The HEAL study included 1,183
women, 18 years of age or older, who were identified through the Surveillance,
Epidemiology, and End Results

(SEER) registries in New Mexico, Los Angeles County,
California and western Washington. Of these, 615 women were recruited from New
Mexico, 202 from Washington and 366 from Los Angeles. The study was designed to
evaluate the independent roles of sex-hormones, diet, weight, physical activity, genetics,
and other factors on post-diagnostic breast cancer prognosis and survival. Details of
study design and recruitment procedures have been described previously
9, 15, 16
.
The current analysis was restricted to HEAL participants who completed the 24
month follow-up questionnaire, had body composition measured by DEXA, and
CRP/SSA measured. Of the 1,183 HEAL participants at baseline, 944 women completed
the follow-up in-person interview 24 months after the baseline survey (approximately 30
months after breast cancer diagnosis). Reasons why women did not complete the follow-
up interview included: death (n=44), refusal to continue participation (n=104), spouse
refusal to permit further contact (n=1), unable to contact (n=17), unable to locate (n=55),
moved from area (n=16), and too ill (n=2). Funding was available to complete a second
DEXA measure on 20 participants at the Washington study site; the first 20 women who
agreed to complete the follow-up examination were measured. In New Mexico, funding
4

was available to complete DEXA on 135 women at the 24 month follow-up examination;
the 135 women were randomly selected from women who completed DEXA at baseline.
Of the 155 women with DEXA at the 24 month follow-up visit, 140 women from the
New Mexico and Washington study sites also had data available on circulating levels of
CRP and SAA. The final analysis was restricted to 138 women (18 from Washington and
120 from New Mexico) who were self-identified as non-Hispanic White (n=108) or
Hispanic (n=30).
Clinical Variables
Trained staff at the respective study centers obtained waist and hip
circumferences (in centimeters) at baseline and at the follow-up visit. Waist
circumference was measured just above the superior margin of the iliac crests. Hip
circumference was measured at the maximal posterior

projection of the buttocks.
A measure of percent body fat was obtained using DEXA; this measure of body fat was
collected for the New Mexico and western Washington study centers at baseline and at the
follow-up examinations, but not at USC. DEXA uses a scan of the whole body using two low
dose x-rays to read bone and soft tissue mass simultaneously. DEXA can be used to estimate
percent body fat by using several measures of body composition collected at the time of the scan
including: height, weight, fat free mass, skeletal muscle mass, skeletal size, bone mineral mass
and body fat distribution. Measurements were taken from participants in New Mexico using the
Lunar model DPX (GE Medical Systems,

Milwaukee, WI) and in Washington using the Hologic
model QDR 1500 (Hologic Inc,

Waltham, MA). For each participant, total percent body fat was
estimated using body composition software which provides proportion of fat and lean tissue in
each pixel based on the assumed constant attenuation of pure fat and of bone mineral-free lean
5

tissue. Furthermore, we calculated body mass index (BMI) as kilograms per meters squared
using height and weight measured at the same time as the DEXA was performed.
Levels of circulating CRP and SAA were measured by latex-enhanced nephelometry
using highly sensitive assays on the Behring Nephelometer II analyzer (Dade Behring
Diagnostics, Deerfield, IL) at the University of Washington. Tests were completed using fasting
blood samples collected at the follow-up interviews at 30 months after diagnosis. Each sample
was processed within 3 hours of collection and stored at -70° to -80°C until analysis. Interassay
coefficients of variation were 5% to 9% for CRP and 4% to 8% for SAA. The lowest detectable
value for CRP is 0.2 mg/L, the lowest for SAA is 0.7mg/L. The control materials that were
included with assay batches for quality control purposes came from Bio-Rad Laboratories
(Hercules, CA).
Questionnaire Variables
In-person interviews (New Mexico) and self-administered questionnaire forms
(Washington) provided information on demographics, dietary intake, menopausal status,
smoking status, disease history (arthritis, chronic lung disease, diabetes, heart attack,
heart failure, hypertension, other cancers) and current use of any over the counter or
prescription NSAIDs at the 30-month follow-up survey. Cancer treatment history,
including history of radiation, chemotherapy and tamoxifen use, was obtained through
medical record review, participants’ SEER reports, or the questionnaire. Weight gain is
defined as increase of more than 5% in body weight from the baseline to the follow-up
exam, weight loss is defined as decrease of more than 5% in body weight since
baseline
17
.
Physical activity was assessed using the Modifiable Activity Questionnaire
6

developed by Kriska and colleagues
18
. The type, duration and frequency of activities
performed in the past year were evaluated. The sports/recreation and household activity
section of the questionnaire addressed 29 popular activities, such as fast walking,
moderate/slow walking, jogging, aerobic, tennis, household cleaning, and yard work.
Hours of activity per week for each activity type were calculated by multiplying the
frequency each activity by the duration. Activities were further classified by intensity --
light (<3 METs), moderate (3-6 METs) or vigorous ( >6 METs) -- based on the values
required to lose 3% of body fat given by Compendium of Physical Activities
19
.
Metabolic equivalent tasks (METs) were calculated as the ratio of the associated
metabolic rate for a specific activity divided by the resting metabolic rate (RMR).
Individual physical activity was evaluated using the sum of the MET values times hours
per week of moderate and vigorous activities.
Caloric intake was calculated using data from the Women’s Health Initiative
food frequency questionnaire
20
and the nutrient database from the University of
Minnesota’s Nutrition Coordinating Center’s Nutrition Data Systems for Research (NDS-
R, version 2005). The questionnaire produced nutrient estimates similar to 4-day and 24-
hour dietary recalls but is not validated against independent and unbiased measures that
take into account selective recall in the post-menopausal population. Women at the New
Mexico study site were asked their usual dietary intake in the preceding year, while
women at the Washington site were asked about their intake in the preceding month.

7

Chapter 3: Statistical Analyses
Geometric means and 95% confidence intervals (CI) were calculated for
logarithmically transformed values of CRP and SAA by obesity status as defined by
DEXA measurements. Values of CRP and SAA were transformed due to the skewed
distribution of the data. Participants were classified as obese if their body fat percentage
was equal to or greater than 35%
21
. Models for ln(CRP) were adjusted for age
(continuous), race/ethnicity(non-Hispanic White or Hispanic), caloric intake(kCal/day
continuous), weight change from the baseline (approximately 6 months post-diagnosis) to
the 30-month post-diagnosis interview (kg continuous), and current NSAID use at the 30-
month post-diagnosis interview. Models for ln(SAA) were further adjusted for history of
arthritis. Additionally, stratified analyses were performed to evaluate the potential
modifying effects of medical history, individual characteristics and lifestyle including
NSAID use at 30-months, chemotherapy history, menopausal status at 30-month,
arthritis, hypertension, smoking, alcohol intake, caloric intake, weight change, physical
activity, and change in physical activity. Physical activity levels are based on
recommended activity for weight maintenance in METS
22
. Smoking history at the 30-
month post-diagnosis interview was classified as never, past, or current smoker. For the
stratified analyses, we collapsed levels of caloric intake (<1200kcal, 1200-1500kcal,
>1500kcal) into two levels (<1500kcal versus ≥1500kcal) since the geometric mean CRP
and SAA levels for <1200kcal and 1200-1500kcal were almost identical. Smoking
history was further defined as never smoked, smoked ≤6 months ago, and smoked >6
months ago as we expected time since cessation of smoking to influence CRP or SAA
levels at blood draw.

8

Regression models were used to assess the associations between percent body fat
and continuous values of CRP and SAA. All final models were adjusted for age at
baseline, caloric intake (kCal continuous), weight change between baseline and follow-up
(kg continuous), and current NSAID use (yes/no). Adjustment for other potential
confounders was evaluated, including physical activity, effects of smoking and alcohol
consumption and chronic conditions including arthritis, hypertension, chronic lung
disease, diabetes, heart attack, heart failure, hypertension, and other cancers. These
variables were not included in the final model as they did not substantially affect our
results with the exception of arthritis in the model for SAA . Stratum-specific estimates
were calculated to evaluate potential effect modification by the same medical history,
individual characteristics and lifestyle factors described for the analysis of geometric
means.
Regression analysis to compare measures of BMI to DEXA were completed.
Correlations of continuous BMI values and percent body fat values measured by DEXA
with ln(CRP) and ln(SAA) were examined. Statistical analyses were performed using
SAS version 9.2 (SAS Institute Inc, North Carolina, USA).

9

Chapter 4: Results
Table 1 describes the characteristics of the HEAL study population by obesity status.
Participants who were obese were more likely to be older (p< .0001) post-menopausal (p<.001),
and had larger average hip circumference (p<.001) and larger average waist circumference
(p<.001) than women who were not obese. However, the proportion of obese and non-obese
survivors who gained 5% or more of their bodyweight over the two year follow-up period was
similar (26.5% vs. 25.0%). A higher proportion of non-obese cases experienced weight loss of
5% of their bodyweight or more over the same time period, but this difference was not
statistically significant.
Obesity status was not associated with demographic or lifestyle factors (Table 1)
including alcohol consumption (p=0.14), smoking history (p=0.49), or education level (p=0.23).
When we examined the frequency of six other self-reported medical conditions commonly
associated with chronic inflammation (see methods), only history of arthritis was significantly
associated with obesity (p<.001). Neither tamoxifen nor chemotherapy was associated with
obesity status (p=0.56, and p=0.68, respectively). However, obese women were more likely to
report current use of any prescription or over the counter NSAIDs at the 30-month post-diagnosis
interview (p=.03).
Table 1. Descriptive characteristics of 138 female breast cancer survivors in HEAL Study.
Not Obese Obese
1

(n=40) (n=98)
Characteristics at 30 month post-diagnosis interview Mean (SD) Mean (SD) p-value
Age at interview 53.9(9.8) 61.0(10.7)
<.0001

Weight (kg)
2
58.9(7.1) 74.8(12.3) <.0001
BMI (kg/m
2
) 21.9(2.5) 28.4(4.3) <.0001
Percentage of body fat 29.8(5.0) 42.6(4.7) <.0001
CRP (mg/L) 1.2(1.5) 3.4(3.7) <.0001
SAA (mg/L) 4.2(2.2) 9.4(17.7) 0.006
Hip Circumference (cm) 96.0(5.4) 109.6(9.3) <.0001
Waist (cm) 74.2(7.3) 92.5(11.23) <.0001

1
Obese ≥35% Body Fat, not obese < 35% Body Fat
21
.
2
Weight at baseline is taken at the first DEXA measurement. Weight at follow-up is taken at the second DEXA
measurement.
10

Table 1, Continued.
Not Obese Obese
1

(n=40) (n=98)
Characteristics at 30 month post-diagnosis interview Mean (SD) Mean (SD) p-value
n (%) n (%)
Physical Activity (MET hrs/week) 44.0(45.0) 28.5(24.8) 0.046

Change in Physical Activity since Baseline
\

6.5(4.6) -5.1(3.2) 0.04
(approximately 6 months post-diagnosis, MET-hrs/week)
Caloric intake (kcal)
2

<1500 kcal 27(67.50) 56(57.14) 0.26
≥ 1500 kcal 13(32.50) 42(42.86)
Weight Difference (%) 0.06

Loss 7(17.5) 5(5.1)
No Change
3
23(57.5) 67(68.4)
Gain 10(25) 26(26.5)

Postmenopausal (yes, %)
4
21(53.9) 83(90.2) <.0001
History of medical conditions (yes, %)
Arthritis 5(12.5) 50(51.0) <.0001
Chronic lung disease, emphysema or bronchitis 3(7.5) 11(11.2) 0.51
Diabetes or high blood sugar 3(7.5) 10(10.2) 0.62
Heart attack or myocardial infarction 0(0.0) 4(4.1) 0.20
Heart failure or congestive heart failure 0(0.0) 3(3.1) 0.26
Hypertension
5
6(15.0) 32(32.7) 0.08

Past and current medication use (yes, %)
Tamoxifen
6
15(37.5) 42(42.9) 0.56
NSAID
7
9(22.5) 41(41.8) 0.03
Chemotherapy
7
7(18.4) 21(21.4) 0.68

Smoking status (%) 0.53
Smoked >6 months ago 18(45.0) 37(37.8)
Smoked ≤6 months ago 7(17.5) 14(14.3)
Never 15(37.5) 47(48.0)
Alcohol consumption (%)
8
0.14
None 2(5.0) 12(12.2)

<10gm 28(70.0) 73(74.5)
≥10gm 10(25.0) 13(13.3)

Highest level of education 0.23
Some high school 1(2.50) 4(4.1)
High school grad 8(20.0) 15(15.3)
Some college/tech 9(22.5) 37(37.8)
College grad 13(32.5) 17(17.4)
Grad school 9(22.5) 25(25.5)

Geometric mean CRP levels were significantly higher among obese than non-obese
women (p<.001)(Table 2a). There patterns were statistically significant primarily for women
who did not have other medical conditions, did not use NSAIDS or had received chemotherapy,

1
Obese ≥35% Body Fat, not obese < 35% Body Fat
21
.
2
1 obese and 3 non-obese had no caloric intake data available.
3
Within 5% change in body weight since baseline weight taken during DEXA measurements
17
.
4
I non obese and 6 obese excluded due to unknown menopausal status
5
1 obese answered "Don't know" and was excluded.
6
Current use of tamoxifen or over the counter or prescription non-steroidal anti-inflammatory drug (NSAID).
7
2 non-obese refused to answer and were excluded.
8
10gm is assumed to be 1 glass of alcohol
23
.
11

and were inactive and non-smokers. Most of these groups were larger in size, suggesting the
power to observe associations if they existed. Furthermore, we also observed a dose-response
relationship between CRP and tertile of percent body fat in all women (i.e. not obese and obese)
combined (p<.001, data not shown). The significant positive trend persisted after adjustment for
of NSAIDs use, level of caloric intake, and history of arthritis (tables not shown). Women with
≥1500 kcal/day intake had significantly higher mean levels of CRP than women who consumed
less than 1500 kcal/day regardless of obesity status (unadjusted pairwise comparison p=.001
among the non-obese, p=.004 among the obese).
Table 2a. Adjusted geometric means and 95% CI of CRP stratified by obesity
status at follow-up (30 months post-diagnosis)(n=134
1
).
Stratifying variable Not Obese Obese
2

(N=40) (N=98)
CRP Mean
(95%CI)
3

CRP Mean (95%CI)
3
p-value
4

Overall 0.85(0.61-1.20)
a
2.01(1.64-2.46)
b
.0086

Post Menopausal
No (n= 18, 8) 0.66(0.38-1.14)
a
2.40(1.14-5.08)
b

Yes (n=20, 81) 0.86(0.54-1.35)
a,b
2.02(1.60-2.57)
b
<.0001

History of arthritis
No (n=34,45) 0.81(0.57-1.17)
a
1.89(1.41-2.52)
b

Yes (n=5,50) 0.91(0.38-2.18)
a,b
2.19(1.65-2.90)
b
<.0001
History of hypertension
5

No (n=33,63) 0.77(0.53-1.12)
a
2.04(1.60-2.60)
b

Yes (n=6,31) 1.16(0.52-2.56)
a,b
2.12(1.45-3.10)
b
<.0001

Chemotherapy
No (n=30,76) 0.80(0.54-1.17)
a
2.19(1.73-2.77)
b

Yes (n=9,19 ) 0.86(0.45-1.65)
a,b
1.59(1.02-2.48)
a,b
<.0001
NSAID use
6

No (n= 31,55) 0.84(0.58-1.21)
a
2.31(1.78-3.00)
b

Yes(n=8,40) 0.99(0.49-1.97)
a,b
1.64(1.20-2.23)
a,b
<.0001

1
4 women lacked caloric data and were excluded from most of the models.
2
Obese ≥35% Body Fat, not obese < 35% Body Fat
21
.
3
Adjusted for age, ethnicity, caloric intake(kcal), NSAID (yes/no), and weight difference(kg).
4
Analysis of covariance was used to compare logarithmically transformed geometric means by obesity status. For each
variable, means with different letter(a-c) are statistically different from one another using Scheffe multiple comparison
procedure (p<.05). P-value shown is from the ANCOVA test of significance between groups.
5
5 excluded because 5 lacked caloric data, and 1 answered "Don't know" to history of hypertension.
6
Current use of over-the-counter or prescription non-steroidal ant-inflammatory drugs (NSAID). Adjusted for age,
caloric intake (kg) weight difference (kg).
12

Table 2a, Continued.

Stratifying variable Not Obese Obese
1

(N=40) (N=98)
CRP Mean
(95%CI)
2

CRP Mean (95%CI)
3
p-value
3

Caloric Intake (kcal)
4

<1500 kcal (n=26,53) 0.78(0.51-1.18)
a
1.82(1.38-2.40)
b

≥ 1500 kcal (n=13,42) 0.91(0.52-1.60)
a
2.42(1.78-3.29

)
b
<.0001
Weight Change
5

Loss (n=7,5) 0.75(0.36-1.57)
a,b
1.96(0.81-4.78)
a,b

No Change (n=22,66 ) 0.70(0.46-1.07)
a
1.93(1.51-2.47)
b

Gain (n=10,24) 1.12(0.59-2.12)
a,b
2.50(1.69-3.69)
b
<.0001

Physical Activity
6

<13 MET hr/wk (n=7,21) 1.09(0.52-2.25)
a.b
2.26(1.47-3.47)
b

13-26 MET hr/wk (n=7,32 ) 0.76(0.35-1.62)
a,b
2.06(1.46-2.90)
b

> 26 MET hr/wk (n=25,42 ) 0.78(0.51-1.19)
a
1.93(1.41-2.64)
a,b
<.0001
Physical Activity Change
7

Less (n=16,45) 0.86(0.52-1.40)
a
2.32(1.73-3.11)
b

Same (n=4,13) 0.94(0.34-2.57)
a,b
1.63(0.95-2.80)
a,b

More (n=19,37) 0.81(0.51-1.29)
a
1.84(1.32-2.57)
a,b
<.0001

Smoking
Never (n=14, 45) 1.22(0.71-2.10)
a
2.20(1.64-2.94)
a,b

Smoked >6 months ago
(n=18,36)
0.63(0.39-1.00)
b
1.89(1.36-2.61)
a

Smoked ≤6 months ago (n=7,14) 0.82(0.39-1.69)
a,b
1.89(1.14-3.13)
a,b
<.0001

The results for SAA were similar in pattern to those for CRP (Table 2b) for most factors
with the exception of hypertension, chemotherapy, calorie intake, physical activity and change in
physical activity. However, among obese women, mean SAA levels were significantly lower
among obese women taking NSAIDs compared to women not taking them.

1
Obese ≥35% Body Fat, not obese < 35% Body Fat
21
.
2
Adjusted for age, ethnicity, caloric intake(kcal), NSAID (yes/no), and weight difference(kg).
3
Analysis of covariance was used to compare logarithmically transformed geometric means by obesity status. For each
variable, means with different letter(a-c) are statistically different from one another using Scheffe multiple comparison
procedure (p<.05). P-value shown is from the ANCOVA test of significance between groups.
4
Adjusted for age, ethnicity, weight difference, NSAID (yes/no).
5
Adjusted for age, ethnicity, caloric intake(kcal), NSAID (yes/no). Weight gain is >5% increase in bodyweight since
baseline and weight loss is >5% decrease since baseline.
6
Based on recommended MET values
22
.
7
More is >5% difference in MET values since initial interview. Less is >5% difference in MET values since initial
interview
13

Table 2b. Adjusted geometric means and 95% CI of SAA stratified by
obesity status at follow-up (30 months post-diagnosis)(n=134
1
).
Stratifying variable Not Obese Obese
2

(N=40) (N=98)
SAA Mean (95%CI)
3
SAA Mean
(95%CI)
3

p-value
4

Overall 4.21(3.34-5.30)
a
6.21(5.42-7.11)
b
.0083

Post Menopausal
No (n= 18, 8) 4.28(2.96-6.20)
a
7.21(4.40-11.80)
b

Yes (n=20, 81) 4.00(2.97-5.39)
a,b
6.26(5.34-7.35)
a,b
<.0001

History of arthritis
No (n=34,45) 3.70(2.93-4.67)
a
5.37(4.46-6.47)
a,b

Yes (n=5,50) 4.06(2.31-7.16)
a,b
7.75(6.46-9.31)
b
<.0001
History of hypertension
8

No (n=33,63) 4.42(3.44-5.67)
a
6.27(5.35-7.35)
a

Yes (n=6,31) 2.94(1.76-4.92)
a
6.06(4.73-7.76)
a
<.0001

Chemotherapy
No (n=30,76) 4.21(3.25-5.44)
a
6.39(5.48-7.45)
a

Yes (n=9,19 ) 3.77(2.46-5.77)
a
5.84(4.34-7.85)
a
<.0001
NSAID use
9

No (n= 31,55) 4.29(3.36-5.48)
a
7.24(6.13-8.56)
b

Yes(n=8,40) 4.59(2.93-7.19)
a,b
4.87(3.95-6.00)
a
<.0001

Caloric Intake (kcal)
10

<1500 kcal (n=26,53) 4.94(3.79-6.44)
a
6.22(5.24-7.39)
a

≥ 1500 kcal (n=13,42) 2.99(2.10-4.25)
b
6.38(5.26-7.76)
a
<.0001
Weight Change
11

Loss (n=7,5) 3.11(1.94-4.98)
a
5.27(2.97-9.35)
a,b

No Change (n=22,66 ) 4.28(3.23-5.67)
a,b,
5.88(4.99-6.92)
a,b

Gain (n=10,24) 4.51(2.97-6.86)
a,b
7.81(6.07-10.04)
b
<.0001

Physical Activity
12

<13 MET hr/wk (n=7,21) 5.50(3.44-8.80)
a
5.39(4.08-7.14)
a

13-26 MET hr/wk (n=7,32 ) 4.23(2.59-6.91)
a
6.75(5.40-8.43)
a

> 26 MET hr/wk (n=25,42 ) 3.72(2.82-4.89)
a
6.44(5.26-7.87)
a
<.0001
Physical Activity Change
13

Less (n=16,45) 4.23(3.07-5.83)
a
5.68(4.70-6.87)
a

Same (n=4,13) 3.93(2.05-7.52)
a
5.73(4.05-8.10)
a

More (n=19,37) 3.89(2.86-5.28)
a
7.46(6.00-9.28)
a
<.0001

1
4 women lacked caloric data and were excluded from most of the models.
2
Obese ≥35% Body Fat, not obese < 35% Body Fat
21
.
3
Adjusted for age, ethnicity, caloric intake(kcal), NSAID (yes/no), weight difference(kg), and arthritis (yes/no).
4
Analysis of covariance was used to compare logarithmically transformed geometric means by obesity status. For each
variable, means with different letter(a-c) are statistically different from one another using Scheffe multiple comparison
procedure (p<.05). P-value shown is from the ANCOVA test of significance between groups.
8
5 excluded because 5 lacked caloric data, and 1 answered "Don't know" to history of hypertension.
9
Current use of over-the-counter or prescription non-steroidal ant-inflammatory drugs (NSAID). Adjusted for age,
caloric intake (kg) weight difference (kg).
10
Adjusted for age, ethnicity, weight difference, NSAID (yes/no).
11
Adjusted for age, ethnicity, caloric intake(kcal), NSAID (yes/no). Weight gain is >5% increase in bodyweight since
baseline and weight loss is >5% decrease since baseline.
12
Based on recommended MET values
22
.
13
More is >5% difference in MET values since initial interview. Less is >5% difference in MET values since initial
interview
14

Table 2a, Continued.

Stratifying variable Not Obese Obese
1

(N=40) (N=98)
SAA Mean
(95%CI)
2

SAA Mean (95%CI)
3
p-value
3

Smoking
Never (n=14, 45) 4.68(3.31-6.60)
a
6.27(5.21-7.56)
a

Smoked >6 months ago
(n=18,36)
3.99(2.93-5.43)
a
7.26(5.90-8.93)
a

Smoked ≤6 months ago (n=7,14) 3.52(2.21-5.61)
a
4.24(3.05-5.91)
a
<.0001

Figure 1 presents the plot for ln(CRP) versus percent body fat centered on the mean. We
observed an approximate linear relationship between ln(CRP) and percent body fat, both in an
unadjusted model and after adjusting for age, caloric intake, weight difference and NSAID use.
Our linear regression model indicates that the estimated ln(CRP) increases by .07 mg/L for each
1% increase in body fat beyond the mean. At the mean percent body fat of 38.9%, the estimated
ln(CRP) based on the adjusted model is 0.45 mg/L. The slopes (β’s) for the unadjusted and
adjusted models were similar.

1
Obese ≥35% Body Fat, not obese < 35% Body Fat
21
.
2
Adjusted for age, ethnicity, caloric intake(kcal), NSAID (yes/no), and weight difference(kg).
3
Analysis of covariance was used to compare logarithmically transformed geometric means by obesity status. For each
variable, means with different letter(a-c) are statistically different from one another using Scheffe multiple comparison
procedure (p<.05). P-value shown is from the ANCOVA test of significance between groups.
15

Figure 1. Scatterplot and predicted regression line of ln(CRP) on centered body fat
percentage in 134 breast cancer survivors HEAL study and the predicted values of ln(CRP)
based on body fat percentage while controlling for age, caloric intake (kcal), weight
difference(kg), and NSAID use (Correlation (95% CI): 0.52 (0.39-0.63) Adj. r
2
= .33).

16

Figure 2 presents the plot for ln(SAA) versus percent body fat centered on the mean; we
found that for every 1% increase in body fat above the mean, estimated ln(SAA) increased 0.033
mg/L. At the mean body fat percentage of 38.9%, the estimated ln(SAA) is 1.72 mg/L.
Figure 2. Scatterplot and predicted regression lines of ln(SAA) on centered body fat
percentage of 138 breast cancer survivors from HEAL study and the predicted values of
ln(SAA) based on body fat percentage while controlling for age, history of arthritis, caloric
intake (kcal), weight difference(kg), and NSAID use (Correlation (95% CI): 0.42 (0.27-0.54)
Adj. r
2
=.20).

Table 3a describes the β coefficients and 95% confidence intervals (CIs) for the
regression models describing the relationship between percent body fat and ln(CRP). The β
coefficients represent the change in ln(CRP) per 1% increase in body fat. The values for the
overall model and those stratified by select lifestyle factors and other medical conditions (or other
health factors) are shown. No significant interaction was found between body fat percentage and
any lifestyle or health-related factor. When we stratified by recent NSAID use, we found a greater
increase in ln(CRP) for each 1% increase in body fat for women who took NSAIDs compared
17

with women who did not. Additionally, we found that for women with a recent history of
hypertension, the effect size is smaller compared with women reporting no hypertension.
Table 3a. Linear regression models for ln(CRP) on body fat percentage for study
population of breast cancer survivors from HEAL (n=134)
1
.
lnCRP

%BodyFat

Full Model
2
β SE 95%CI p-value
0.068 0.012 (0.04,0.09) <.0001
p-value
for Stratified Models interactio
n Chemotherapy (n=28) 0.072 0.026 (0.02,0.13) 0.0106 0.39
No Chemotherapy (n=106) 0.065 0.014 (0.04,0.09) <.0001

NSAID use (n=48)
3
0.037 0.027 (-0.02,0.09) 0.1766 0.13
No NSAID use (n=86) 0.076 0.013 (0.05,0.10) <.0001

History of arthritis (n=55) 0.057 0.024 (0.01,0.11) 0.0238 0.80
No History of arthritis (n=83) 0.069 0.015 (0.04,0.10) <.0001

History of hypertension
(n=37)
4

0.024 0.029 (-0.04,0.08) 0.4181 0.16
No History of hypertension
(n=96)
0.073 0.014 (0.05,0.10) <.0001
0.073 0.014 (0.05,0.10) <.0001
Physical Activity
5
0.46
<13 MET hr/wk (n=28) 0.060 0.029 (0.00,0.12) 0.047
13-26 MET hr/wk (n=39) 0.063 0.026 (0.01,0.12) 0.0194
> 26 MET hr/wk (n=67) 0.061 0.018 (0.02,0.10) 0.0015

Physical Activity Change
6
0.21
Less (n=61) 0.081 0.017 (0.05,0.12) <.0001
Same (n=17) 0.067 0.046 (-0.04,0.17) 0.1775
More (n=56) 0.054 0.019 (0.02,0.09) 0.0062

Weight Change
7
0.53
Loss (n=12) 0.128 0.229 (-0.43,0.69) 0.5969
Same (n=88) 0.079 0.013 (0.05,0.10) <.0001
Gain (n=34) 0.048 0.026 (-0.01,0.10) 0.0789

The results for ln(SAA) (see Table 3b) are generally similar to those described for
ln(CRP) (Table 3a), although the effect size differences are smaller between the strata. One
exception is for women with a history of hypertension in which the effect size was higher rather
than smaller as found for CRP.

1
4 women lacked caloric data and were excluded from most of the models.
2
Adjusted for age, ethnicity, caloric intake(kcal), NSAID (yes/no), and weight difference(kg).
3
Current use of over-the-counter or prescription non-steroidal ant-inflammatory drugs (NSAID). Adjusted for age,
ethnicity, caloric intake (kg) weight difference (kg).
4
5 excluded because 5 lacked caloric data, and 1 answered "Don't know" to history of hypertension.
5
Based on recommended MET values
22
.
6
More is >5% difference in MET values since initial interview. Less is >5% difference in MET values since initial
interview
7
Adjusted for age, caloric intake(kcal), NSAID (yes/no). Weight gain is >5% increase in bodyweight since baseline
and weight loss is >5% decrease since baseline.
18

Table 3b. Regression models for ln(SAA) on body fat percentage for study population of
breast cancer survivors from HEAL (n=134)
1
.

lnSAA

%BodyFat

Full Model
2
β SE 95%CI p-value
0.032 0.008 (0.02,0.05) 0.0002
p-value for
Stratified Models interaction
Chemotherapy (n=28) 0.055 0.017 (0.02,0.09) 0.0034 0.22
No Chemotherapy (n=106) 0.022 0.010 (0.00,0.04) 0.0255

NSAID use (n=48)
3
-0.005 0.023 (-0.05,0.04) 0.8253 0.16
No NSAID use (n=86) 0.032 0.008 (0.02,0.05) 0.0002

History of arthritis (n=55)
4
0.037 0.018 (0.00,0.07) 0.0469 0.45
No History of arthritis (n=83) 0.020 0.009 (0.00,0.04) 0.0308

History of hypertension
(n=37)
5

0.048 0.021 (0.00,0.09) 0.0318 0.43
No History of hypertension
(n=96)
0.024 0.010 (0.00,0.04) 0.0183

Physical Activity
6
0.46
<13 MET hr/wk (n=28) 0.028 0.017 (-0.01,0.06) 0.1179
13-26 MET hr/wk (n=39) 0.030 0.025 (-0.02,0.08) 0.2315
> 26 MET hr/wk (n=67) 0.024 0.011 (0.00,0.05) 0.0292

Physical Activity Change
7
0.30
Less (n=61) 0.019 0.015 (-0.01,0.05) 0.211
Same (n=17) 0.015 0.041 (-0.08,0.11) 0.7241
More (n=56) 0.034 0.010 (0.01,0.05) 0.0013

Weight Change
8
0.53
Loss (n=12) -0.029 0.065 (-0.19,0.13) 0.6724
Same (n=88) 0.034 0.008 (0.02,0.05) <.0001
Gain (n=34) -0.009 0.023 (-0.06,0.04) 0.6894

Figure 3 shows the plots for the distribution of acute phase proteins (ln CRP, ln SAA) by
continuous measures of DEXA (% body fat) or BMI. There is a noticeable difference in the
scatter pattern for body fat measured by BMI or DEXA and ln(CRP) or ln(SAA). We note that
there is less clustering of data points for DEXA measures than for BMI, reflecting a stronger

1
4 women lacked caloric data and were excluded from most of the models.
2
Adjusted for age, ethnicity, caloric intake(kcal), NSAID (yes/no), arthritis, and weight difference(kg).
3
Current use of over-the-counter or prescription non-steroidal ant-inflammatory drugs (NSAID). Adjusted for age,
ethnicity, caloric intake (kg) weight difference (kg).
4
Adjusted for age, ethnicity, caloric intake(kcal), NSAID (yes/no), and weight difference(kg).
5
5 excluded because 5 lacked caloric data, and 1 answered "Don't know" to history of hypertension.
6
Based on recommended MET values
22
.
7
More is >5% difference in MET values since initial interview. Less is >5% difference in MET values since initial
interview
8
Adjusted for age, caloric intake(kcal), NSAID (yes/no). Weight gain is >5% increase in bodyweight since baseline
and weight loss is >5% decrease since baseline.
19

linear relationship for DEXA and the two acute phase proteins. The measurements were taken
concurrently. The correlation coefficients for DEXA were 0.52 (0.39-0.63) for ln(CRP) and .42
(0.27-0.54) for ln(SAA) while the correlation coefficients for BMI were 0.47 (0.34-0.60) for
ln(CRP) and .38 (0.22-0.42) for ln(SAA). However, there is no statistically significant difference
between the slope of BMI versus the slope for percent body fat for either ln(SAA) or ln(CRP).
20

Figure 3. ln(CRP) and ln(SAA) on BMI or Percent Body fat.

21

Figure 3, Continued

22

Chapter 5: Discussion
In this study, we found a significant association between acute phase proteins (CRP or
SAA) and adiposity as measured by DEXA which reaffirms the association between fat
accumulation and inflammation. The group classified as obese at 35% body fat or greater had
consistently higher CRP levels compared to non-obese women regardless of lifestyle factors. Our
analysis indicates that percent body fat is a strong predictor of CRP and SAA levels. We found
some suggestive modifying factors for these associations. While we did not find a significant
interaction between obesity and lifestyle factors with acute phase proteins, our data suggest some
modifiable associations including recent NSAID use, caloric intake and smoking. For example,
when we modeled the relationship between percent body fat and ln(CRP), we found the
association was diminished among NSAID users compared to non-NSAID users. We also found
that high caloric intake (>1500 kcal/day) was associated with higher levels of acute phase
proteins regardless of percent body fat (obesity).
However, while suggestive, the potential modifying effects of these factors did not reach
statistical significance for CRP though it did for SAAperhaps because small numbers of women
in some of the strata limited power. The association between acute phase proteins and adiposity is
consistent with earlier reports based on anthropometric measures. Increased adiposity has
previously been reported as significantly correlated with elevated levels of CRP and SAA among
breast cancer patients
11, 23
. A previous study of HEAL participants (n=741) using anthropometric
measures (BMI, waist circumference) found significant associations between BMI and
inflammatory proteins
8
. Furthermore, measures of inflammation in breast cancer patients have
been reported as associated with existing medical conditions, cancer and tumor metastasis
24
.
Obesity itself may induce a low-grade chronic inflammatory condition; the accumulation of
adipocytes may trigger a biological response that leads to increased levels of pro-inflammatory
23

cytokines including IL-6. These cytokines may then cause a greater production of acute phase
proteins.
25

DEXA may provide a more accurate measure of body fat for studies of acute phase
proteins than other anthropometric measures because DEXA can distinguish between bone,
muscle and fat mass
26
. The accuracy of BMI as a measure of percent body fat may vary by age,
gender, race or ethnicity and physical conditioning. It is an overall measure of body mass and
does not provide measures of body composition. . Thus, there is no absolute cross-tabulation of
DEXA definition of fat by standard BMI. For example, the accuracy of BMI as a measure of
adiposity in postmenopausal women may be reduced due to age-related changes in bone and
muscle mass
27
. There may also be important differences in accuracy by race/ethnicity. Studies
have previously found that Hispanic women, on average, have a higher % body fat than women
of other ethnicities, even after adjustment for differences in age and socioeconomic status
28, 29
.
One study found that Hispanic women have a higher percentage of fat in the trunk and abdominal
region, which is positively associated with elevated CRP
28
. Body fat determined through clinical
measures of bio-impedance also may be less reliable than DEXA, as the measure tends to
underestimate adiposity
12, 13
. Additionally, while BMI, waist-hip ratio, and waist circumference
have been used as indicators of body fatness, these measurements were found to be more closely
correlated with each other than with body fatness measured by DEXA based on data from a
nationally representative US population sample (National Health and Nutrition Examination
Survey)
13
.
In this study, the association of percent body fat (using a precise measure such as DEXA)
with levels of acute phase proteins suggests that circulating inflammatory markers are associated
with greater adiposity and not just greater weight. In the literature, weight loss and physical
activity have inconsistently been associated with circulating acute phase proteins. Some studies
suggest that physical activity is strongly correlated with levels of acute phase proteins, while
24

other studies have not supported this association
30
. In a study by Bastard et al., weight loss was
associated with a significant reduction in IL-6 levels in both plasma and adipose; the authors also
found a non-statistically significant reduction in CRP. Bochud et al. suggested that the lack of
statistical significance with CRP in the Bastard study may be due to the short duration of weight
loss and the small amount of weight lost.
23
In our study, the relationship between percent body
fat and CRP was similar among people who lost or gained weight.
When we looked at the direct relationship between some of our lifestyle factors and CRP
and SAA levels we found several significant associations. We found women who consumed at
least 1500 kcal/day had higher CRP levels than those who consumed less than 1500 kcal/day.
Low caloric intake has been associated with lower CRP, however weight loss due to a negative
caloric intake balance and fat mobilization has also associated with elevated CRP
31
. Previous
studies of caloric restriction (where subjects were limited to reduced caloric intake compared to a
typical Western diet) found that caloric restriction is associated with lower levels of inflammatory
proteins, lower body fat, and longer survival
32
. Animal and human studies in caloric restriction
have shown that caloric restriction has anti-inflammatory effects associated with longer survival
and may be more effective than exercise alone. One proposed mechanism suggests that reduction
in caloric intake reduces hyperinsulimia and insulin-like growth factor (IGF) levels - which are
known to promote angiogenesis and inhibit apoptosis
33
.
In our HEAL sample, we found acute inflammatory proteins did not differ across levels
of physical activity. It has been suggested that the type of exercise plays a role in reducing CRP
levels. One study suggests that aerobic exercise may be more effective than flexibility or
resistance exercise for CRP reduction
34
.
Studies of some diseases such as arthritis have compared sensitivity of SAA to CRP as
predictors of degree of inflammation and disease activity and have found SAA to be more
25

sensitive
35-37
. While increases in CRP and SAA levels during inflammation are correlated , the
magnitude of change may vary depending on disease
38
.
Though weight change is typically associated with the loss of or accumulation of fat, the
relationship between weight and adiposity is highly variable depending on the individual.
Predisposition to obesity (BMI ≥30) is not necessarily a predisposition to fat accumulation.
Genetics, age, and lifestyle factors including diet composition, existing medical conditions, and
activity level interact and influence adiposity levels. In our analyses, we observed a slight
elevation in inflammatory markers associated with weight loss. Nevertheless, the effects of
weight loss combined with lifestyle change, in particular, dietary changes, have been shown to
reduce the production of inflammatory cytokines
32
.
A strength of this study is the use of DEXA measurements for estimates of adiposity,
which may give a more accurate and valid assessment of body fatness, than BMI or percent body
fat from bio-impendence. Our sample includes non-Hispanic white and Hispanic breast cancer
survivors. Differences in percent body fat and CRP or SAA levels by race/ethnicity may be
expected due to previous studies that have indicated fat-patterning differs by race and ethnicity.
Specifically, higher measures of central adiposity have been found in Hispanic women. Since it
had been shown by Pierce et al.
14
that higher CRP and SAA levels adversely affect survival
among breast cancer survivors, identifying a more accurate measure of body fat (a predictor of
CRP and SAA) may be useful for accurately identifying those women who could improve
prognosis by decreasing body fat.
Our study has several limitations. The primary limitation of this study is that despite this
being the largest study to date with such comprehensive measures of body composition, BMI and
inflammation, it remains a relatively small study. Therefore, stratified analysis to examine the
effect of various clinical and lifestyle factors must be interpreted with caution as power is limited
for these stratified analyses. In specific, this small sample size limited our ability to explore
26

whether there was a specific threshold for which increases in calorie intake led to increases in
inflammation. In addition were were not able to explore changes in caloric intake over the 2
year interval, because some women (n=43) were missing baseline caloric data. Therefore, we do
not think that the current analysis provides specific information about the precise level of
reduction in calorie intake that may be beneficial.
This analysis suggests that circulating CRP and SAA are positively associated with
percent body fat. Beyond a reduction in adiposity, changes in other lifestyle factors, including a
reduction of caloric intake or use of NSAIDs, may help to decrease levels of acute-phase
inflammatory proteins. However, the potential benefits or risks of NSAIDs or caloric restriction
will require further exploration in a larger study. Assessment of the association between use of
NSAIDs and survival time may be worthwhile as lower levels of CRP have been linked with
longer survival time and NSAIDs have been suggested as an adjuvant treatment for breast cancer
39
. Our analysis confirms the finding that adiposity is strongly associated with circulating levels of
CRP/SAA.

27

Bibliography

[26] Obesity: preventing and managing the global epidemic. Report of a WHO
consultation. World Health Organ Tech Rep Ser 2000;894:i-xii, 1-253.

[22] Physical Activities Guidelines Advisory Committee. Physical Activity
Guidelines Advisory Committee Report, 2008. Washington (DC): US
Department of Health and Human Services; 2008.

[19] Ainsworth B HW, Whitt M, et al. Compendium of physical activities: An
update of activity codes and MET intensities. 2000;(suppl 9)(32):6.

[8] Albuquerque KV, Price MR, Badley RA, Jonrup I, Pearson D, Blamey RW,
et al. Pre-treatment serum levels of tumour markers in metastatic breast
cancer: a prospective assessment of their role in predicting response to
therapy and survival. Eur J Surg Oncol 1995;21(5):504-9.

[23] Bochud M, Marquant F, Marques-Vidal PM, Vollenweider P, Beckmann JS,
Mooser V, et al. Association between C-reactive protein and adiposity in
women. J Clin Endocrinol Metab 2009.

[28] Casas YG, Schiller BC, DeSouza CA, Seals DR. Total and regional body
composition across age in healthy Hispanic and white women of similar
socioeconomic status. Am J Clin Nutr 2001;73(1):13-8.

[5] Coussens LM, Werb Z. Inflammation and cancer. Nature
2002;420(6917):860-7.

[3] Das UN. Is obesity an inflammatory condition? Nutrition 2001;17(11-
12):953-66.

[29] Fernandez JR, Heo M, Heymsfield SB, Pierson RN, Jr., Pi-Sunyer FX,
Wang ZM, et al. Is percentage body fat differentially related to body mass
index in Hispanic Americans, African Americans, and European
Americans? Am J Clin Nutr 2003;77(1):71-5.

[24] Fingleton B. Matrix metalloproteinases: roles in cancer and metastasis.
Front Biosci 2006;11:479-91.

[13] Flegal KM, Shepherd JA, Looker AC, Graubard BI, Borrud LG, Ogden CL,
et al. Comparisons of percentage body fat, body mass index, waist
circumference, and waist-stature ratio in adults. Am J Clin Nutr
2009;89(2):500-8.

[32] Fontana L. Neuroendocrine factors in the regulation of inflammation:
excessive adiposity and calorie restriction. Exp Gerontol 2009;44(1-2):41-5.

28

[25] Fried SK, Bunkin DA, Greenberg AS. Omental and subcutaneous adipose
tissues of obese subjects release interleukin-6: depot difference and
regulation by glucocorticoid. J Clin Endocrinol Metab 1998;83(3):847-50.

[27] Going S, Williams D, Lohman T. Aging and body composition: biological
changes and methodological issues. Exerc Sport Sci Rev 1995;23:411-58.

[4] Heinrich PC, Castell JV, Andus T. Interleukin-6 and the acute phase
response. Biochem J 1990;265(3):621-36.
[36] Hilliquin P. Biological markers in inflammatory rheumatic diseases. Cell
Mol Biol (Noisy-le-grand) 1995;41(8):993-1006.

[1] Hirschfield GM, Pepys MB. C-reactive protein and cardiovascular disease:
new insights from an old molecule. QJM 2003;96(11):793-807.

[15] Irwin ML, Crumley D, McTiernan A, Bernstein L, Baumgartner R,
Gilliland FD, et al. Physical activity levels before and after a diagnosis of
breast carcinoma: the Health, Eating, Activity, and Lifestyle (HEAL) study.
Cancer 2003;97(7):1746-57.

[16] Irwin ML, McTiernan A, Bernstein L, Gilliland FD, Baumgartner R,
Baumgartner K, et al. Physical activity levels among breast cancer
survivors. Med Sci Sports Exerc 2004;36(9):1484-91.

[34] Kohut ML, McCann DA, Russell DW, Konopka DN, Cunnick JE, Franke
WD, et al. Aerobic exercise, but not flexibility/resistance exercise, reduces
serum IL-18, CRP, and IL-6 independent of beta-blockers, BMI, and
psychosocial factors in older adults. Brain Behav Immun 2006;20(3):201-9.

[18] Kriska AM CC. Introduction to the collection of physical activity
questionnaires in a collection of physical activity questionnaires for health-
related research. In: Caspersen Ka, editor. Med Sci Sports Exerc: Centers
for Disease Control and Prevention.; 1997.

[38] Malle E, De Beer FC. Human serum amyloid A (SAA) protein: a prominent
acute-phase reactant for clinical practice. Eur J Clin Invest 1996;26(6):427-
35.

[6] Marsik C, Kazemi-Shirazi L, Schickbauer T, Winkler S, Joukhadar C,
Wagner OF, et al. C-reactive protein and all-cause mortality in a large
hospital-based cohort. Clin Chem 2008;54(2):343-9.

[9] McTiernan A, Rajan KB, Tworoger SS, Irwin M, Bernstein L, Baumgartner
R, et al. Adiposity and sex hormones in postmenopausal breast cancer
survivors. J Clin Oncol 2003;21(10):1961-6.

[12] Neovius M, Hemmingsson E, Freyschuss B, Udden J. Bioelectrical
impedance underestimates total and truncal fatness in abdominally obese
women. Obesity (Silver Spring) 2006;14(10):1731-8.
29

[10] Nguyen XM, Lane J, Smith BR, Nguyen NT. Changes in inflammatory
biomarkers across weight classes in a representative US population: a link
between obesity and inflammation. J Gastrointest Surg 2009;13(7):1205-12.

[7] O'Hanlon DM, Lynch J, Cormican M, Given HF. The acute phase response
in breast carcinoma. Anticancer Res 2002;22(2B):1289-93.

[37] O'Hara R, Murphy EP, Whitehead AS, FitzGerald O, Bresnihan B. Acute-
phase serum amyloid A production by rheumatoid arthritis synovial tissue.
Arthritis Res 2000;2(2):142-4.

[20] Patterson RE, Kristal AR, Tinker LF, Carter RA, Bolton MP, Agurs-Collins
T. Measurement characteristics of the Women's Health Initiative food
frequency questionnaire. Ann Epidemiol 1999;9(3):178-87.

[11] Pierce BL, Neuhouser ML, Wener MH, Bernstein L, Baumgartner RN,
Ballard-Barbash R, et al. Correlates of circulating C-reactive protein and
serum amyloid A concentrations in breast cancer survivors. Breast Cancer
Res Treat 2009;114(1):155-67.

[14] Pierce BL, Ballard-Barbash R, Bernstein L, Baumgartner RN, Neuhouser
ML, Wener MH, et al. Elevated Biomarkers of Inflammation Are
Associated With Reduced Survival Among Breast Cancer Patients. J Clin
Oncol 2009.

[31] Salas-Salvado J, Bullo M, Garcia-Lorda P, Figueredo R, Del Castillo D,
Bonada A, et al. Subcutaneous adipose tissue cytokine production is not
responsible for the restoration of systemic inflammation markers during
weight loss. Int J Obes (Lond) 2006;30(12):1714-20.

[35] Sipe JD. Acute-phase proteins in osteoarthritis. Semin Arthritis Rheum
1995;25(2):75-86.

[17] Stevens J, Truesdale KP, McClain JE, Cai J. The definition of weight
maintenance. Int J Obes 2005;30(3):391-399.

[30] Tchernof A, Nolan A, Sites CK, Ades PA, Poehlman ET. Weight loss
reduces C-reactive protein levels in obese postmenopausal women.
Circulation 2002;105(5):564-9.

[39] Takkouche B, Regueira-Mendez C, Etminan M. Breast cancer and use of
nonsteroidal anti-inflammatory drugs: a meta-analysis. J Natl Cancer Inst
2008;100(20):1439-47.

30

[2] Visser M, Bouter LM, McQuillan GM, Wener MH, Harris TB. Elevated C-
reactive protein levels in overweight and obese adults. JAMA
1999;282(22):2131-5.

[21] Wilmore J. Excercise, obesity and weight control. . Washington DC:
President's Council on Physical Fitness and Sports Research Digest 1994;
Series 1(No. 6).

Abstract (if available)

Abstract Objective: To examine the association between body fat percentage as measured by DEXA and 2 systemic inflammatory markers, C-reactive protein (CRP) and Serum amyloid A protein (SAA) in breast cancer survivors and its modulation by lifestyle factors.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Arm lymphedema in a multi-ethnic cohort of female breast cancer survivors

PDF

Factors that influence mammographic density: role of estrogen metabolism genes, biomarkers of inflammation, and lifestyle

PDF

The role of heritability and genetic variation in cancer and cancer survival

PDF

The ADRB3 TRP64ARG variant and obesity in African American breast cancer cases

PDF

Genes and hormonal factors involved in the development or recurrence of breast cancer

Asset Metadata

Creator Dee, Anne (author)

Core Title DEXA measures of body fat percentage and acute phase proteins among breast cancer survivors

School Keck School of Medicine

Degree Master of Science

Degree Program Applied Biometry

Publication Date 05/04/2010

Defense Date 04/19/2010

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag body fat percentage,cancer,DEXA,OAI-PMH Harvest,obesity

Place Name California (states), Los Angeles (counties), New Mexico (states), Washington (states)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Azen, Stanley Paul (committee chair), McKean-Cowdin, Roberta (committee chair), Wu, Anna (committee member)

Creator Email adee@usc.edu,needane@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m2989

Unique identifier UC1329710

Identifier etd-Dee-3663 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-308397 (legacy record id),usctheses-m2989 (legacy record id)

Legacy Identifier etd-Dee-3663.pdf

Dmrecord 308397

Document Type Thesis

Rights Dee, Anne

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu