775

The Epidemiology of Arm and Hand Swelling in

Premenopausal Breast Cancer Survivors

Electra D. Paskett,1,2 Michelle J. Naughton,3 Thomas P. McCoy,4

L. Douglas Case,4 and Jill M. Abbott1
1
Comprehensive Cancer Center and 2School of Public Health, The Ohio State University, Columbus, Ohio and
Departments of 3Public Health Sciences and 4Biostatistical Sciences, Wake Forest
University School of Medicine, Winston-Salem, North Carolina

Abstract

Background: Breast cancer survivors suffer from lymphe- greater number of lymph nodes removed [hazards ratio
dema of the arm and/or hand. Accurate estimates of the (HR), 1.02; P < 0.01], receiving chemotherapy (HR, 1.76; P =
incidence and prevalence of lymphedema are lacking, as are 0.02), being obese (HR, 1.51 versus normal weight; P = 0.01),
the effects of this condition on overall quality of life. and being married (HR, 1.36; P = 0.05). Factors associated
Methods: Six hundred twenty-two breast cancer survivors with persistent swelling were having more lymph nodes
(age, V45 years at diagnosis) were followed with semian- removed (odds ratio, 1.03; P = 0.01) and being obese (odds
nual questionnaires for 36 months after surgery to deter- ratio, 2.24 versus normal weight; P < 0.01). Women
mine the incidence of lymphedema, prevalence of reporting swelling had significantly lower quality of life
persistent swelling, factors associated with each, and as measured by the functional assessment of cancer
quality of life. therapy-breast total score and the SF-12 physical and mental
Results: Of those contacted and eligible for the study, 93% health subscales (P < 0.01 for each).
agreed to participate. Fifty-four percent reported arm or Conclusions: Lymphedema occurs among a substantial
hand swelling by 36 months after surgery, with 32% proportion of young breast cancer survivors. Weight man-
reporting persistent swelling. Swelling was reported to agement may be a potential intervention for those at greatest
occur in the upper arm (43%), the hand only (34%), and risk of lymphedema to maintain optimal health-related
both arm and hand (22%). Factors associated with an quality of life among survivors. (Cancer Epidemiol Bio-
increased risk of developing swelling included having a markers Prev 2007;16(4):775 – 82)

Introduction
Breast cancer is the most common type of cancer and the Lymphedema following treatment for breast cancer has
second leading cause of cancer mortality among women in the received attention in multiple studies. The overall incidence of
United States (1). It ranks second among cancer deaths in all arm lymphedema can range from 8% to 56% 2 years following
women (1) and first in cancer deaths among women ages 20 to surgery, depending on the extent of axillary surgery and the
59 years (2). Although advancements in cancer treatment and use of radiotherapy (7-15). Most women with lymphedema
emphasis on early detection through mammography screening develop it within the first 12 to 14 months following treatment
have allowed more cancer patients to become survivors, there (16, 17).
has been little change in the number of new cases of invasive Lymphedema can cause limitations in range of motion, pain,
breast cancer in women younger than age 40 years (3). weakness, or stiffness in the affected arm (18, 19). It also results
Survivors face psychological, physical, and emotional chal- in psychological problems, including anxiety, depression,
lenges, all of which affect quality of life (4). sexual dysfunction, social avoidance, and exacerbation of
Lymphedema is a common complication of cancer therapy existing psychiatric illness (20). The effect of arm swelling on
and is characterized by an accumulation of lymphatic fluid in appearance has been suggested to be greater than the effect of
the interstitial tissue that causes swelling, most often in the coping with the initial diagnosis and treatment of breast cancer
arms or legs. Lymphedema can occur anywhere lymph nodes as the swollen arm or hand is a constant reminder of breast
have been surgically removed or lymph flow has been cancer, is a subject of curiosity to others, and may suggest a
disturbed (5). An unwanted consequence of cancer treatment recurrence to the survivor (6). Generally, quality of life is
(4), lymphedema is especially concerning to patients who think compromised for breast cancer patients with lymphedema
they have been cured of their cancer (6). (5, 21-25).
Consistency among studies about prevalence and incidence
rates, risk factors, and prevention and treatment for lymphe-
dema among breast cancer survivors is lacking. Many reasons
for this gap have been proposed. Lymphedema continues to be
Received 3/8/06; revised 1/9/07; accepted 1/25/07. under diagnosed and is not defined or measured in a
Grant support: U.S. Army Medical Research and Materiel Command grants DAMD17-96-1- standardized manner (26-28), thus making estimates of
6292 and DAMD17-01-1-0447.
The costs of publication of this article were defrayed in part by the payment of page charges.
incidence difficult to obtain. Few studies are designed to
This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. follow newly diagnosed women in a longitudinal manner to
Section 1734 solely to indicate this fact. capture the incidence of this condition and determine the
Note: Presented in part at both an oral paper session and a poster session at the Era of Hope
Department of Defense Breast Cancer Research Program Meeting, June 11, 2005, prevalence of the condition over time. Cross-sectional study
Philadelphia, PA and at Congress of Epidemiology 2001, Toronto, Ontario, Canada. designs, most commonly used, only provide a snap shot of the
Requests for reprints: Electra D. Paskett, Ohio State University Comprehensive Cancer Center, prevalence of lymphedema at a single point in time, not in a
A356 Starling-Loving Hall, 320 West 10th Avenue, Columbus, OH 43210.
E-mail: [email protected] longitudinal fashion following diagnosis.
Copyright D 2007 American Association for Cancer Research. Although the etiologic factors for lymphedema have not
doi:10.1158/1055-9965.EPI-06-0168 been studied extensively, some studies have identified several

Cancer Epidemiol Biomarkers Prev 2007;16(4). April 2007

Downloaded from cebp.aacrjournals.org on October 12, 2015. © 2007 American Association for Cancer
Research.
776 Arm and Hand Swelling in Young Breast Cancer Survivors

common factors associated with the development of this tion about patient demographics, their cancer diagnosis and
condition. Extent of axillary dissection, radiation therapy, treatment, patient risk factors for disease and/or lymphedema,
obesity at diagnosis, older age, postoperative fluid formation, and life quality.
and infection in the arm have been reported as related factors Demographics. Age, race/ethnicity, marital status, educa-
(10, 29-31). tional background, income, and employment and insurance
Although breast cancer affects women of all ages, young status were collected from self-report.
women with breast cancer (i.e., those under age 50 years) tend to
have more aggressive breast tumors (32, 33), which necessitates Medical and reproductive history. Information was collected
treatments that may be more toxic than those offered to older about comorbid conditions, family history, and reproductive
women (34-36). Although breast cancer incidence and mortality history, including parity, pelvic surgery, and menstrual cycling.
continue to decline among women younger than 50 years (37), Medical chart review. An extensive medical chart review was
these toxic treatments may cause significant side effects that done on all patients by clinic staff at 1 year after recruitment.
may last a long time. Furthermore, very little has been reported Information was obtained on the date and technique of breast
on lymphedema in this patient population that receives more cancer diagnosis, tumor size, location, grade, hormone
intensive treatment and may suffer from the associated effects of receptor status, number of nodes examined, number of
this side effect for a longer time period, especially during the positive lymph nodes, type of definitive cancer surgery, and
most productive years of life. reconstructive surgery, if any. Chemotherapy information
This prospective study is one of the first in the United States (dates, drugs, and dosages in milligrams) was gathered from
to establish reliable estimates of and factors associated with the medical oncology office records. Likewise, dose per treatment,
incidence and prevalence of arm swelling, to explore which treatment area, and total dosage and duration of treatment
patient characteristics are associated with the incidence and were recorded for women receiving radiation therapy.
persistence of lymphedema, as well as to document the effect Hormonal therapies, such as tamoxifen, were recorded with
of swelling on quality of life among young breast cancer dates, routes of administration, and dosages.
survivors. Thus, this study provides data not found in
previous studies that have used cross-sectional designs. Arm and hand swelling form. Patients were asked if they had
experienced any swelling in their arm or hand since their
surgery (at baseline) or in the last 6 months (at each follow-up
Materials and Methods assessment), location of swelling, and severity. Patients were
also asked to assess the effect of swelling on daily life
Procedures and Participants. Data for this study were taken functions, such as wearing clothing, the completion of routine
from participants recruited to the Menstrual Cycle Mainte- personal, home, and work tasks, exercise, and general use of
nance and Quality of Life After Breast Cancer Treatment the affected hand/arm(s). In addition, participants indicated
Study, a prospective, observational study of patients ages 18 to whether they had sought treatment for the condition, and if
45 years (38). The objectives of this study are to document and yes, what type of treatment they received (31). These questions
identify determinants of menstrual cycle maintenance after have been successfully used in previous studies and research
breast cancer treatment, to examine survivor’s quality of life protocols (4). Other methods to measure arm volume (i.e.,
longitudinally, to track reproductive events among those volume by perometry, water displacement, or arm circumfer-
attempting pregnancy, and ultimately to investigate the effect ence measurements) were not feasible given data collection by
of subsequent pregnancy on survival. Recruitment to this mail. Previous work, however, has indicated moderate
study occurred from January 1998 to December 2005. This correlation between objective measurements of swelling and
article includes 3 years of prospective data from the first 627 self-report of swelling (39).
women who were recruited to this study through July of 2002
to address secondary goals related to lymphedema. Follow-up Personal habits questionnaire. Information about women’s
of participants continues. smoking and alcohol use, height in inches, weight in pounds,
Patients were recruited from clinical centers at the Memorial weight change, and exercise habits were collected. Body mass
Sloan-Kettering Cancer Center in New York City, New York index (BMI) was calculated from height and weight measure-
(449 women); M. D. Anderson Cancer Center in Houston, Texas ments (weight/height2, as kg/m2) and then categorized
(92 women); Presbyterian Hospital in Dallas, Texas (37 women); (referred to as weight status) as normal/underweight (BMI,
and the Wake Forest University Baptist Medical Center in <25 kg/m2), overweight (BMI, 25-29.9 kg/m2), or obese (BMI,
Winston-Salem, North Carolina (49 women). Women were z30 kg/m2).
identified at these clinical centers using tumor/surgical
Functional assessment of cancer therapy-breast. The functional
registries or physician referrals. Inclusion criteria included
assessment of cancer therapy-breast (FACT-B) is a multidi-
female patients ages 18 to 45 years at diagnosis with a stage I, II,
mensional, cancer-specific quality of life measure. This scale
or III invasive breast cancer within the previous 8 months. All
assesses physical well-being, social/family well-being, rela-
patients were required to have regular menstrual cycles at the
tionship with doctor, emotional well-being, functional well-
time of diagnosis. Thus, patients who had a previous hysterec-
being, and concerns specific to breast cancer patients. Scores
tomy, even with intact ovaries, were ineligible for this protocol.
can be calculated for each of the six subscales, as well as a total
Patients were excluded if they had a prior or concurrent history
score composed of all six subscales. Higher scores on this
of any cancer, excluding basal or squamous cell skin carcinoma
measure indicate better levels of functioning (40).
and stage 0 cervical cancer. This study was approved by the
Institutional Review Board of each hospital as well as the U.S. SF-12 health status questionnaire. The SF-12 is a 12-item short
Department of Defense Human Subjects Committee. form of the SF-36 Health Status Questionnaire, a generic
health-related quality of life instrument (41, 42). The SF-12 is
Data Collection and Instruments. Patients completed
composed of two components, a physical health and a mental
questionnaires at baseline and at 6-month intervals thereafter.
health subscale. These subscales are scored with a mean of 50
All follow-up data collection was conducted by mail through
and a SD of 10. Higher scores on these subscales indicate
the study coordinating center at the Wake Forest University
higher levels of functioning.
School of Medicine. Descriptions of the questionnaires perti-
nent to the incidence, prevalence, quality of life and develop- Follow-up questionnaires. Participants completed updates of
ment of lymphedema, and used in the current analyses are their medical and reproductive history every 6 months,
listed below. In brief, these questionnaires provided informa- including their general medical status, cancer recurrences,

Cancer Epidemiol Biomarkers Prev 2007;16(4). April 2007

Downloaded from cebp.aacrjournals.org on October 12, 2015. © 2007 American Association for Cancer
Research.
 Cancer Epidemiology, Biomarkers & Prevention 777

reproductive events, surgical procedures, and any current or Table 1. Demographic, medical, and psychosocial charac-
newly initiated drugs and/or therapy. Patients also completed teristics of sample (N = 622)
the FACT-B, SF-12, follow-up Arm and Hand Swelling, and
Personal Habits forms at 6-month intervals. Updates of Characteristic n (%)
participant demographics, primarily changes in marital status, Age at diagnosis, 38.5 [4.9 (20-45)]
education, income, employment, and insurance status were mean [SD (range)], y
collected every 12 months. Race/ethnicity
White 555 (89)
Analytic Methods. Of the 849 women contacted about the African-American 28 (4)
study, 672 were eligible to participate. Of these eligible women, Hispanic 23 (4)
627 (93%) agreed to participate. The 45 women who refused Asian/Pacific Islander 16 (3)
participation in the study were similar in age to those who Marital status
Single/separated/widowed 156 (25)
participated (median age, 40.1 versus 39.8 years, respectively; Married/marriage-like 466 (75)
P = 0.82, Wilcoxon rank-sum) but different in race (77% White Education
versus 89% White, respectively; P = 0.04, Fisher’s exact). Five of High school graduate or less 54 (9)
the 627 participants completed no follow-up surveys after Some college 158 (25)
surgery and were excluded from all data analyses. Time was College graduate (4 y) or above 409 (66)
calculated from surgery until the date of the survey and divided Annual family income ($)
<50K/y 177 (29)
into 6-month intervals for descriptive purposes. Prevalence of 50-100K/y 232 (38)
swelling during these intervals was calculated by dividing the >100K/y 201 (33)
number of participants who indicated that they had experi- Employment status
enced swelling by the number of participants who filled out a Full-time 339 (55)
survey during that time. Time to first swelling occurrence was Part-time 86 (14)
calculated as the time from surgery until the first occurrence of Full-time homemaker 107 (17)
Other 90 (14)
swelling. Because participants were asked if they had experi- Insurance
enced swelling since surgery (baseline survey) or in the last None 4 (1)
6 months (follow-up surveys), we used the midpoint of the HMO only 155 (29)
interval for the event time when swelling was noted. Group only 298 (56)
The Kaplan-Meier method (43) was used to estimate time to Medicaid only 8 (2)
VA/military only 3 (1)
swelling, and Cox proportional hazards regression was used to Other 48 (9)
determine which covariates were significantly associated with Multiple types 15 (3)
this outcome (44). Age (in years), race (White versus other), No. children
marital status (married versus single), education (high school 0 234 (38)
graduate or less, some college, and college graduate), weight 1 117 (19)
2 182 (29)
status (normal/underweight, overweight, and obese), current 3+ 89 (14)
smoking status (smoker and nonsmoker), weekly exercise Children <8 years of age 225 (36)
(none, walking, mild, moderate, and strenuous), having a child Weight status
<8 years of age, employment status (full-time, part-time, Underweight/normal (BMI, <25 kg/m2) 408 (66)
homemaker, and other), reconstructive surgery, lumpectomy, Overweight (BMI, 25-29.9 kg/m2) 118 (19)
mastectomy, nodal dissection [none, sentinel node dissection Obese (BMI, z30 kg/m2) 96 (15)
Smoking status
(SND) only and axillary node dissection (AND)], number of Never 354 (57)
lymph nodes removed, number of positive lymph nodes, Former 268 (43)
antibiotic use at baseline, radiation therapy, chemotherapy, Current 48 (8)
and tamoxifen use were included as covariates in the model. All Weekly exercise
covariates were considered fixed except for receiving tamoxi- None 113 (18)
Walking only 86 (14)
fen, which was treated as a time varying covariate. Mild 73 (12)
To account for missing visits, the probability of swelling at the Moderate 158 (25)
missing visit was determined using participants with complete Strenuous 191 (31)
data whose patterns matched that of the participant with Type of surgery
missing data. Then, 1,000 samples were taken from the original Lumpectomy only 320 (51)
population, with the time to first swelling for a particular Mastectomy 296 (48)
Type of node dissection
participant sampled with probability p (45). The Kaplan-Meier Sentinel (SND) only 26 (4)
and Cox proportional hazards analyses were then run on each Axillary (AND) 580 (93)
sample, and estimates across all the analyses for the results were Neither 14 (2)
pooled. For time to swelling, the estimates were simply the No. nodes removed, 15.0 [9.3 (0-51)]
means of the monthly Kaplan-Meier estimates. For the Cox median [SD (range)]
No. positive nodes
proportional hazards models, the HR was estimated as the 0 350 (56)
exponential of the average h for a particular covariate, the 95% 1-3 172 (28)
confidence interval was obtained as the exponential of the 4-9 59 (9)
average h F 1.96 times the square root of the average variance of 10+ 41 (7)
the h, and the P value was calculated based on Wald tests (the Antibiotics for infection 19 (3)
average h divided by the square root of the average variance). Chemotherapy (yes) 545 (88)
Tamoxifen (yes) 343 (55)
To assess which demographic and clinical factors were Radiation therapy (yes) 435 (70)
associated with self-reported swelling over time, a longitudinal Quality of life measures,
logistic regression model was fit using the Generalized mean [SD (range)]
Estimating Equations method to account for the multiple FACT-B 105.5 [19.2 (44-146)]
observations per person (46, 47). An autoregressive covariance SF-12M 43.0 [8.2 (14-61)]
SF-12P 44.2 [8.9 (20-65)]
structure was used to model the correlation of the repeated
measurements over time. Time was considered in this model by NOTE: N = 622, at the baseline survey unless otherwise specified.
flooring the time since surgery to the nearest month and Abbreviations: HMO, Health Maintenance Organization; VA, Veterans Affairs.

Cancer Epidemiol Biomarkers Prev 2007;16(4). April 2007

Downloaded from cebp.aacrjournals.org on October 12, 2015. © 2007 American Association for Cancer
Research.
778 Arm and Hand Swelling in Young Breast Cancer Survivors

Table 2. Association between patient characteristics and participants with both baseline and postsurgery follow-up are
time to first swelling (results from weighted resampling, presented in Table 1. The median age of the women when
Cox regression analysis; N = 622) diagnosed with breast cancer was 39 years (range, 20-45 years).
The majority of the participants were non-Hispanic White
Characteristic HR (95% CI) P (89%) and were married or had a live-in partner (75%). Most
Non-White vs White 1.30 (0.89-1.88) 0.17 (62%) women had children, and 36% currently had children
Married vs single 1.36 (1.00-1.85) 0.05 <8 years. Two thirds (66%) had a 4-year college degree or
Education 0.56 higher. Thirty-eight percent of the women had an annual
Some college vs 1.08 (0.68-1.73) 0.74 household income between $50,000 and $100,000, and one
high school graduate and below
College graduate vs 0.93 (0.60-1.46) 0.76 third had a household income above $100,000 annually. Less
high school graduate and below than half (43%) of the participants had a past smoking history
Weight status 0.04 when enrolled in the study; eight percent reported current
Overweight vs normal/underweight*
c
1.24 (0.92-1.69) 0.16 smoking at baseline. Weekly level of exercise varied, with 82%
Obese vs normal/underweight 1.51 (1.09-2.09) 0.01 of the women exercising weekly at some level but only 31%
Age at diagnosis, y 1.00 (0.97-1.03) 1.00 reporting any strenuous exercise. Thirty-four percent of the
Current smoker (Y/N) 0.84 (0.51-1.40) 0.51
Weekly exercise 0.12 women were overweight or obese, as classified by their weight
Walking vs none 1.14 (0.73-1.77) 0.56 status (BMI, z25 kg/m2). The mean (FSD) scores for FACT-B,
Mild vs none 1.49 (0.97-2.30) 0.07 SF-12M, and SF-12P at baseline were 105.5 (19.2), 43.0 (8.2), and
Moderate vs none 1.18 (0.80-1.72) 0.40 44.2 (8.9), respectively.
Strenuous vs none 1.06 (0.73-1.55) 0.76 Over half (51%) of the women had a lumpectomy only, and
Children <8 y old 0.86 (0.64-1.14) 0.29
Employment 0.53
48% had a mastectomy. Sixty percent of the participants
Part-time vs full-time 0.84 (0.58-1.23) 0.38 having mastectomies underwent immediate reconstructive
Full-time homemaker vs full-time 0.91 (0.63-1.31) 0.61 surgery. Ninety-three percent of all women had AND, 4%
Other vs full-time 1.06 (0.76-1.48) 0.74 had SND only, and 2% had neither. Seventy-one percent of the
Reconstructive surgery (Y/N) 0.78 (0.55-1.09) 0.15 women had 10 or more nodes removed (median of 15 nodes
Lumpectomy (Y/N) 0.96 (0.69-1.34) 0.81 removed), and 44% had one or more positive nodes. Eighty-
Mastectomy (Y/N) 1.36 (0.92-2.03) 0.13
Node dissection 0.56 eight percent of the women received chemotherapy, 70%
None vs AND 1.05 (0.38-2.91) 0.93 received radiation therapy, and 55% received tamoxifen
SND only vs AND 0.68 (0.33-1.41) 0.30 sometime after their diagnosis of breast cancer.
No. nodes removed 1.02 (1.01-1.04) <0.01
No. nodes positive 0.99 (0.96-1.01) 0.28 Follow-up Surveys. The 622 participants with postsurgery
Antibiotic use at baseline (Y/N) 1.66 (0.90-3.03) 0.10 follow-up completed up to seven follow-up surveys over the
Radiation therapy (Y/N) 1.23 (0.86-1.77) 0.26 course of the first 3 years after survey. Of the 622 participants,
Chemotherapy (Y/N) 1.76 (1.10-2.82) 0.02 296 (48%) had complete data for all visits. More than three
Tamoxifen (Y/N) 0.81 (0.62-1.05) 0.11 fourths (n = 482; 77%) of the women had >50% data for all
Abbreviations: 95% CI, 95% confidence interval; Y/N, yes/no.
visits over the follow-up period. The effect of missing data was
*BMI, 25-29.9 versus <25 kg/m2. examined by comparing analyses with complete-case data to
cBMI, z30 versus <25 kg/m2. analyses using imputation as described in the analytic
methods. Prevalence and correlates of outcomes were similar
modeled as continuous. Persistent swelling was defined as the and results presented are based on the analyses with
report of two or more swelling episodes within the first 3 years resampling, where applicable.
after surgery. A multivariable logistic regression model was Swelling. Twenty percent of the women reported having
used to determine which demographic and medical factors arm/hand swelling during the first 6 months following
were associated with persistent swelling by 3 years after surgery, 36% by 1 year, and over half (54%) by the 3rd year
surgery. The effect of missing swelling data was examined by following surgery. Figure 1 shows the Kaplan-Meier estimate
looking at patterns of missing data and calculating weights of ever reporting arm/hand swelling over the first 3 years
from observed proportions of persistent swelling from com- following surgery. The median time to swelling was f26
plete case data. Weighted imputation for missing data was done months after surgery. Prevalence of swelling varied from 23%
using resampling as described above for time to first swelling to 29% for any 6-month window following surgery. Women
occurrence. Logistic regression for persistent swelling was
analyzed for each sample, and mean estimates were calculated.
Longitudinal mixed models were fit to explore how
swelling, demographic, and medical characteristics affected
women’s quality of life as measured by the SF-12 and FACT-B.
An autoregressive covariance structure was used to model the
correlation of the repeated measurements over time. In
addition to the covariates described above for the survival
analysis, the Generalized Estimating Equations and quality of
life mixed models also included linear and quadratic terms for
months past surgery. For these models, all covariates were
considered to be fixed in the analyses except for: weight status,
antibiotic use, current smoking status, exercise, and tamoxifen
use. Tamoxifen was modeled a lag1 time-varying covariate. All
the analyses were conducted using SAS version 8.2 (SAS
Institute, Inc., Cary, NC).

Results
Characteristics of the Participants. The baseline demo- Figure 1. Swelling incidence over first 3 y following surgery
graphic, medical, and psychosocial characteristics of the 622 (N = 622).

Cancer Epidemiol Biomarkers Prev 2007;16(4). April 2007

Downloaded from cebp.aacrjournals.org on October 12, 2015. © 2007 American Association for Cancer
Research.
 Cancer Epidemiology, Biomarkers & Prevention 779

Table 3. Factors associated with swelling that occurred 3 years after surgery. Only two variables were found to be
during the 1st 3 y (results from the Generalized Estimating significantly related to persistent swelling: the number of
Equations modeling; N = 622) nodes removed and weight status. For each additional lymph
node removed, the odds of persistent swelling increased by 3%
Characteristic OR (95% CI) P (OR, 1.03; P = 0.01). For women with baseline weight status in
Months past surgery the obese range (i.e., >30 kg/m2) compared with normal/
Linear 1.06 (1.02-1.10) <0.01 underweight weight status (<25 kg/m2), the odds of persistent
Quadratic 0.998 (0.997-0.999) <0.01 swelling were 2.24 times higher (P < 0.01).
Non-White vs White 1.69 (1.03-2.75) 0.04
Married vs single 1.38 (0.95-2.01) 0.09 Quality of Life. Table 5 shows the results for the
Education 0.17 longitudinal mixed modeling exploring the relationships
Some college vs 1.20 (0.66-2.18) 0.55 between arm and/or hand swelling and quality of life,
high school graduate and below
College graduate vs 0.83 (0.47-1.47) 0.53
adjusted for demographic and clinical factors. Women with
high school graduate and below no swelling had significantly higher (better) SF-12M, SF-12P,
Weight status 0.07 and FACT-B scores than women reporting swelling (P value
Overweight vs normal/underweight*
c
1.35 (0.98-1.87) 0.07 <0.01 for each subscale).
Obese vs normal/underweight 1.54 (1.02-2.32) 0.04
Age at diagnosis, y 1.00 (0.96-1.03) 0.83
Current smoker (Y/N) 0.82 (0.51-1.33) 0.42 Discussion
Weekly exercise 0.64
Walking vs none 1.12 (0.85-1.48) 0.43 Lymphedema is an often debilitating consequence of breast
Mild vs none 1.29 (0.90-1.84) 0.17
Moderate vs none 0.97 (0.67-1.41) 0.88
cancer treatment (5, 9, 48-51). The goal of this study was to
Strenuous vs none 1.04 (0.73-1.48) 0.83 determine prospectively the incidence and prevalence of
Children <8 y old 0.79 (0.55-1.13) 0.19 lymphedema in young breast cancer survivors, to assess
Employment 0.99 which factors were associated with reporting lymphedema
Part-time vs full-time 1.05 (0.63-1.75) 0.86 (ever and persistently), and to assess the effect of lymphedema
Full-time homemaker vs full-time 1.00 (0.64-1.57) 0.99 on quality of life. To date, this is the first study in the United
Other vs full-time 0.98 (0.63-1.51) 0.91
Reconstructive surgery (Y/N) 0.68 (0.44-1.04) 0.08 States to address these issues among young breast cancer
Lumpectomy (Y/N) 0.76 (0.49-1.17) 0.21 survivors.
Mastectomy (Y/N) 1.36 (0.82-2.25) 0.23
Node dissection 0.31
None vs AND 0.88 (0.31-2.54) 0.82
SND only vs AND 0.57 (0.26-1.25) 0.16
No. nodes removed 1.04 (1.02-1.06) <0.01 Table 4. Factors associated with persistent swelling during
No. nodes positive 0.97 (0.94-1.01) 0.12 the first 3 y (results from logistic regression using weighted
Antibiotic use (Y/N) 2.41 (1.66-3.48) <0.01
Radiation therapy (Y/N) 1.48 (0.95-2.32) 0.09
resampling; N = 622)
Chemotherapy (Y/N) 1.58 (0.91-2.76) 0.10 Characteristic OR (95% CI) P
Tamoxifen (Y/N) 1.45 (1.06-1.99) 0.02
Non-White vs White 1.52 (0.82-2.80) 0.18
*BMI, 25-29.9 versus <25 kg/m2. Married vs single 1.41 (0.88-2.26) 0.15
cBMI, z30 versus <25 kg/m2. Education 0.22
Some college vs 1.20 (0.57-2.51) 0.64
high school graduate and below
reported swelling in the upper arm only more frequently (43%) College graduate vs 0.81 (0.40-1.65) 0.56
than in the hand only (34%) or in both the arm and hand (22%). high school graduate and below
Seventy percent of the cases of swelling were reported as being Weight status 0.01
Overweight vs normal/underweight* 1.58 (0.97-2.56) 0.07
mild, 25% were moderate, and 5% reported severe swelling. c
Obese vs normal/underweight 2.24 (1.33-3.78) <0.01
Forty-three percent of the swelling cases were accompanied by Age at diagnosis, y 0.99 (0.95-1.03) 0.75
pain in the affected hand and/or arm. Current smoker (Y/N) 0.73 (0.35-1.55) 0.42
Table 2 summarizes the results of the Cox proportional Weekly exercise 0.28
hazards model assessing which factors were associated with Walking vs none 0.77 (0.39-1.54) 0.46
Mild vs none 1.62 (0.82-3.21) 0.17
ever swelling. Number of nodes removed was the most Moderate vs none 1.34 (0.74-2.42) 0.33
significant factor (P = 0.003), with the hazard increasing by Strenuous vs none 1.17 (0.66-2.08) 0.59
2.2% for each additional node removed (f24% increase for 10 Children <8 y old 0.76 (0.48-1.20) 0.24
nodes). Additionally, the hazard of swelling was increased by Employment 0.79
76% for women who received chemotherapy (P = 0.02), by 51% Part-time vs full-time 1.17 (0.65-2.09) 0.60
for those who were obese at diagnosis (relative to normal Full-time homemaker vs full-time 1.14 (0.64-2.04) 0.66
Other vs full-time 1.16 (0.67-1.99) 0.60
weight; P = 0.01), and by 36% for those who were married (P = Reconstructive surgery (Y/N) 0.78 (0.45-1.35) 0.37
0.05). Lumpectomy (Y/N) 1.01 (0.58-1.74) 0.98
Table 3 presents the results of the longitudinal analysis of Mastectomy (Y/N) 1.22 (0.63-2.35) 0.55
factors associated with the prevalence of swelling during the Node dissection 0.45
first 3 years following breast cancer surgery. Non-white race, None vs AND 0.96 (0.21-4.44) 0.96
[odds ratio (OR), 1.69; P = 0.04], a greater number of nodes SND only vs AND 0.51 (0.15-1.66) 0.26
No. nodes removed 1.03 (1.01-1.06) 0.01
removed (OR, 1.04; P < 0.01), tamoxifen use (OR, 1.45; P = No. nodes positive 0.98 (0.94-1.03) 0.44
0.02), and needing antibiotics for arm or hand infection (OR, Antibiotic use at baseline (Y/N) 1.37 (0.47-4.03) 0.56
2.41; P < 0.01) were factors significantly related to the Radiation therapy (Y/N) 1.26 (0.70-2.24) 0.44
prevalence of arm and/or hand swelling over time. Chemotherapy (Y/N) 1.80 (0.94-3.46) 0.08
Logistic regression was used to examine the simultaneous Tamoxifen (Y/N) 1.27 (0.86-1.87) 0.23
effects of demographic and clinical factors on persistent NOTE: Persistent swelling was defined as the report of two or more swelling
swelling during this 3-year period (Table 4). Approximately episodes within the first 3 y after surgery.
32% of the women experienced persistent swelling (i.e., two or *BMI, 25-29.9 versus <25 kg/m2.
more episodes of arm and/or hand swelling) during the first cBMI, z30 versus <25 kg/m2.

Cancer Epidemiol Biomarkers Prev 2007;16(4). April 2007

Downloaded from cebp.aacrjournals.org on October 12, 2015. © 2007 American Association for Cancer
Research.
780 Arm and Hand Swelling in Young Breast Cancer Survivors

Table 5. Adjusted effects of swelling on quality of life: SF- the incidence of lymphedema (8, 54), and reasons for this
12M, SF-12P, and FACT-B (results from longitudinal mixed difference are unclear.
modeling; N = 622) The relationship between weight status (i.e., BMI category)
and swelling is particularly significant because weight status
Scale Estimate SE (95% confidence P was associated with swelling in all models. This finding is
limits)
corroborated in other studies (8, 52, 55) and has many
SF-12M 1.67 0.39 (0.91,2.43) <0.01 implications. Overweight and obesity can be easily identified
Mean scores at 3 y in breast cancer patients and, with some effort, it can be
Not swelling 44.8 0.73 modified after treatment to reduce a woman’s risk of swelling.
Swelling 43.1 0.78
SF-12P 1.29 0.37 (0.56,2.03) <0.01 There is overwhelming evidence that overweight and obesity
Mean scores at 3 y contribute significantly to other health problems, not only
Not swelling 49.8 0.84 among cancer survivors but also among all Americans (56, 57).
Swelling 48.6 0.88 One recent study by Denmark-Wahnefried et al. (58) reported
FACT-B 2.19 0.67 (0.87,3.51) <0.01 that a majority (70%) of breast cancer survivors are overweight
Mean scores at 3 y or obese, putting most survivors at greater risk for cancer
Not swelling 113.8 2.03
Swelling 111.7 2.08 recurrence, cardiovascular disease, diabetes, and overall
poorer quality of life (59-63). Given these indications,
NOTE: Models adjusted for months past surgery (linear and quadratic term), age oncologists should strongly encourage their breast cancer
at diagnosis (in years), race (White versus other), marital status (married versus patients to engage in and routinely practice weight control
single), education (high school graduate or less, some college, college graduate), strategies to minimize their risk for swelling, cancer recur-
weight status category (time varying), current smoking status (time varying),
weekly exercise category (time varying), having a child <8 y of age, employment
rence, and development of other chronic diseases.
status (full-time, part-time, homemaker, and other), reconstructive surgery, Other demographic and medical characteristics that signif-
lumpectomy, mastectomy, nodal dissection (none, SND only, and AND), icantly affected the onset of lymphedema or persistent
number of nodes removed, number of positive nodes, antibiotic use (time swelling were marital status, chemotherapy, race, tamoxifen
varying), radiation therapy, chemotherapy, and tamoxifen use (lag1 time use, and antibiotic use at baseline. The risk of swelling among
varying). married women was 1.36 times higher than the risk of swelling
for unmarried women, although a prior study by Engel et al.
(7) suggested no relationship between swelling and relation-
Lymphedema Incidence. Swelling may occur at any point ship status. It is not known why relationship status would be
following axillary node dissection or radiation therapy, related to risk of swelling. If married women had higher BMIs
beginning immediately after or even delayed by several years. or had more nodes removed, they might be more likely to
Our findings revealed a high cumulative incidence of swelling experience higher incidence of swelling; however, those
among young breast cancer survivors, with more than half relationships were not found in this study. Higher rates of
(54%) reporting ever swelling by 36 months after surgery. Few swelling could be related to the types of activities, in which
studies have assessed the cumulative incidence of swelling in a married women engage (e.g., more routine household chores,
prospective study design; however, other cross-sectional care of children, etc.) compared with other women.
studies have reported swelling rates of 8% to 39% at 18 Women receiving chemotherapy, taking tamoxifen, and
months to 20 years after treatment, respectively (10, 11). receiving antibiotics at baseline were also more likely to report
Previous studies have not been limited to young (V45 years of swelling over time. The risk of swelling was increased by 76%
age) breast cancer survivors, as was true in the present study. for women receiving chemotherapy, which contradicts find-
In addition, the present study used a prospective design with ings from several other studies that found no association
assessment of swelling every 6 months, which enabled between receipt of chemotherapy and swelling, even after
estimation of both the incidence and the prevalence of accounting for axillary node dissection (8, 27, 52, 54). Perhaps,
swelling. the more aggressive treatment offered to younger breast cancer
Using various multivariable regression models, several survivors was related to the increased risk of postoperative
factors were shown to contribute significantly to swelling swelling in this population.
during the 3-year study period in this study. Number of nodes The odds of reporting swelling were greater in non-White
removed and obesity were significantly associated (or border- women compared with White women (OR, 1.69), and this
line associated) with time to first swelling, swelling over time, finding is supported by others (52). In addition, the women in
and persistent swelling. The finding that young breast cancer our study who took tamoxifen were more likely to report
survivors had a greater risk of swelling as more nodes were swelling (OR, 1.45) than those who did not take tamoxifen,
removed during surgery is consistent with other studies. A whereas a previous study reported no association with
German study by Engel et al. (7) found that the odds of tamoxifen use (54). Clearly, this finding warrants further study
swelling among women who had between 10 and 20 nodes as this relationship was based on self-report of swelling.
removed were 2.6 times the odds of swelling among women Similarly, women receiving antibiotics also reported more
who did not have axillary surgery, and the effect was even swelling (OR, 2.41). Using antibiotic use as a proxy for arm
more pronounced for women who had >20 nodes removed. infection, Petrek et al. (64) also reported a significant
Similar findings have been reported by others (52, 53). Our relationship between arm infection and arm swelling.
study found an OR of 1.04, revealing that for every node Several factors in our study did not have a significant effect
removed, the odds of swelling increased by 4%. If 10 nodes on the incidence of lymphedema, including education, type of
were removed, the odds of swelling among young breast surgery, having reconstructive surgery, having radiation, the
cancer survivors would increase to 48%, and if 20 nodes were number of positive nodes, age at diagnosis, smoking, and
removed, the odds of swelling would increase to 119%. exercise frequency. In some cases, other findings support ours
Although this is somewhat lower than that reported in a (7, 8, 52-54), whereas in other cases, they show a different trend
study of similar design (7), it is comparable and similar in (9, 27, 55, 65, 66). For example, there was no relationship
magnitude. Additionally, it is medically plausible that the between receiving radiation and swelling among women in
removal of more nodes contributes to higher risk of swelling our study; however, others have reported receipt of radiation
given that there is greater disruption of lymph flow as more as part of breast cancer treatment to be a risk factor for arm
nodes are removed. Other studies, however, have not found a swelling (9, 65, 66). These differences may, in fact, be the result
significant effect for the number of lymph nodes removed on of the populations studied and the type of study design. The

Cancer Epidemiol Biomarkers Prev 2007;16(4). April 2007

Downloaded from cebp.aacrjournals.org on October 12, 2015. © 2007 American Association for Cancer
Research.
 Cancer Epidemiology, Biomarkers & Prevention 781

present study focuses solely on young breast cancer survivors standing those factors that increase the odds of lymphedema
in a prospective design. incidence and persistent swelling will allow clinicians,
researchers, and educators to more accurately identify those
Lymphedema Prevalence. Persistent swelling has been less
at greatest risk (i.e., those with axillary node dissection and
studied. Our findings provide an estimate of repeated or
who are obese) and to develop programs and practices that
continuous swelling up to 3 years after surgery. Within 6
best meet the needs of breast cancer survivors. For example, a
months of surgery, f20% of young breast cancer survivors
weight management program that promotes weight loss or
reported swelling. By 36 months, 54% of the participants had
prevention of weight gain postoperatively may reduce the
swelled and 32% (59% of those with any swelling) had
incidence of lymphedema among those at greater risk.
persistent swelling. These findings are similar to those
Similarly, providing lymphedema prevention education to
reported by Engel et al., (7) where 38% of the participants
those younger women who undergo more extensive axillary
experienced continuous swelling 5 years from surgery. Factors
node dissection and/or chemotherapy may reduce the risk of
related to persistent swelling were similar to those related to
prevalent swelling or the severity if swelling does develop. In
incident swelling.
so doing, cancer survivors of all ages will ultimately enjoy
Lymphedema and Quality of Life. Young breast cancer better quality of life.
survivors who reported swelling experienced a poorer quality
of life compared with women who did not report swelling, as
evidenced by scores on the FACT-B and the mental and
References
physical scales of the SF-12. The results of numerous other
1. American Cancer Society. Cancer facts and figures. Atlanta: American
studies confirm these findings with more heterogeneous Cancer Society, Inc.; 2006.
populations, thus suggesting that, in this respect, the effect of 2. Jemal A, Murray T, Ward E, et al. Cancer statistics, 2005. CA Cancer J Clin
lymphedema on quality of life is not different for young breast 2005;55:10 – 30.
cancer survivors compared with older breast cancer survivors 3. American Cancer Society. Breast cancer facts and figures 2005 – 2006.
Atlanta: American Cancer Society, Inc.; 2005.
(7, 9-11, 52, 67, 68). 4. Paskett ED, Stark NN. Lymphedema: knowledge, treatment, and impact
Research as has also shown that as arm problems (i.e., among breast cancer survivors. Breast J 1999;5:341.
swelling and limited movement) are treated and subside, 5. Guralnik JM, LaCroix AZ, Abbott RD, et al. Maintaining mobility in late life:
quality of life significantly improves (7). This suggests that demographic characteristics and chronic conditions. Am J Epidemiol 1993;
137:845 – 57.
prompt diagnosis and treatment of lymphedema can help 6. Petrek JA, Pressman PI, Smith RA. Lymphedema: current issues in research
maintain quality of life among survivors. and management. CA Cancer J Clin 2000;50:292 – 307.
7. Engel J, Kerr J, Schlesinger-Raab A, Sauer H, Holzel D. Axilla surgery
Strengths and Limitations. Because of the large sample size severely affects quality of life: results of a 5-year prospective study in breast
and prospective design, our study provides prospective cancer patients. Breast Cancer Res Treat 2003;79:47 – 57.
estimates of lymphedema incidence and prevalence and the 8. Ozaslan C, Kuru B. Lymphedema after treatment of breast cancer. Am J Surg
effect of lymphedema on quality of life among young breast 2004;187:69 – 72.
9. Kwan W, Jackson J, Weir LM, Dingee C, McGregor G, Olivotto IA. Chronic
cancer survivors. To date, relatively few studies have arm morbidity after curative breast cancer treatment: prevalence and impact
prospectively determined lymphedema incidence or examined on quality of life. J Clin Oncol 2002;20:4242 – 8.
persistent swelling; however, our estimates of both incidence 10. Velanovich V, Szymanski W. Quality of life of breast cancer patients with
and prevalence of swelling are consistent with previous lymphedema. Am J Surg 1999;177:184 – 8.
11. Kornblith AB, Herndon JE, Weiss RB, et al. Long-term adjustment of
estimates from cross-sectional studies. Thus, our study extends survivors of early-stage breast carcinoma, 20 years after adjuvant
what is known about the prevalence of lymphedema and its chemotherapy. Cancer 2003;98:679 – 89.
effect as a chronic health condition. 12. Petrek JA, Heelan MC. Incidence of breast carcinoma-related lymphedema.
Several limitations should be noted, however. Young breast Cancer 1998;83:1776 – 81.
13. Swedborg I, Wallgren A. The effect of pre- and post-mastectomy radio-
cancer survivors in this study were mostly White, affluent, and therapy on the degrees of edema, shoulder-joint mobility, and gripping
well-educated women who may be healthier (e.g., only 8% force. Cancer 1981;47:877 – 81.
self-reported smoking) than the average young breast cancer 14. Kissin MW, Quero Della Rovere G, Easton D, Westbury G. Risk of
survivor. Therefore, caution should be used when generalizing lymphedema following the treatment of breast cancer. Br J Surg 1986;73:
580 – 4.
these results to a more representative population. Second, 15. Gerber L, Lampert M, Wood C, et al. Comparison of pain, motion, and
women in this study were recruited through tumor registries edema after modified radical mastectomy vs. local excision with axillary
and physician practices. Consequently, those who chose to dissection and radiation. Breast Cancer Res Treat 1992;21:139 – 45.
participate may be different than those who chose not to 16. Werner RS, McCormick B, Petrek J, et al. Arm edema in conservatively
managed breast cancer: obesity is a major predictive factor. Radiology 1991;
participate. Third, the sample for this study included only 180:177 – 84.
those women who were 45 years of age or younger at 17. Guedes Neto HI. Arm edema after treatment for breast cancer. Lymphology
diagnosis; therefore, caution should again be used when 1997;30:35 – 6.
generalizing these results to older populations. Finally, 18. Sneeuw KC, Aaronson NK, Yarnold JR, et al. Cosmetic and functional
outcomes of breast conserving treatment for early stage breast cancer:
lymphedema was measured through self-report and was not relationship with psychosocial functioning. Radiother Oncol 1992;25:160 – 6.
validated by physical measurement. Other studies (39) have 19. Maunsell E, Brisson J, Deschenes L. Arm problems and psychological
reported moderate correlation between objective measure- distress after surgery for breast cancer. Can J Surg 1993;36:315 – 20.
ments indicating swelling and self-report of swelling 20. Passik S, Newman M, Brennan M, Holland J. Psychiatric consultation for
women undergoing rehabilitation for upper-extremity lymphedema follow-
(c-statistic, 0.919; ref. 39), thus self-report is fairly accurate. ing breast cancer treatment. J Pain Symptom Manage 1993;8:226 – 33.
21. Parker SL, Tong T, Bolden S, Wingo PA. Cancer statistics, 1997. CA Cancer J
Clin 1997;47:5 – 27.
Conclusions 22. Balducci L, Schapira DV, Cox CE, Greenberg HM, Lyman GH. Breast cancer
of the older woman: an annotated review. J Am Geriatr Soc 1991;39:1113 – 23.
In general, our findings further support results of previous 23. Yancik R, Ries LG, Yates JW. Breast cancer in aging women: a population-
based study of contrasts in stage, surgery, and survival. Cancer 1989;63:
studies examining lymphedema incidence and its effect on 976 – 81.
quality of life and extend what is known about lymphedema 24. Kessler LG. The relationship between age and the incidence of breast cancer:
prevalence and incidence using a prospective design with a population and screening program data. Cancer 1992;69:1896 – 903.
large sample size. Lymphedema is clearly a chronic condition, 25. Anderson RT, James MK, Miller ME, Worley AS, Longino CF, Jr. The timing
of change: patterns in transitions in functional status among elderly persons.
which negatively affects breast cancer survivors’ quality of life. J Gerontol B Psychol Sci Soc Sci 1998;53:S17 – 27.
These issues underlie the importance of awareness, preven- 26. Hull MM. Lymphedema in women treated for breast cancer. Semin Oncol
tion, early diagnosis, and treatment of lymphedema. Under- Nurs 2000;16:226 – 37.

Cancer Epidemiol Biomarkers Prev 2007;16(4). April 2007

Downloaded from cebp.aacrjournals.org on October 12, 2015. © 2007 American Association for Cancer
Research.
782 Arm and Hand Swelling in Young Breast Cancer Survivors

27. Geller BM, Vacek PM, O’Brien P, Secker-Walker RH. Factors associated with 48. Tobin MB, Lacy HJ, Meyer L, Mortimer PS. The psychological morbidity of
arm swelling after breast cancer surgery. J Womens Health 2003;12:921 – 30. breast cancer related arm swelling. Cancer 1993;72:3248 – 52.
28. Warmuth M, Bowen G, Prosnitz L, Chu L, Broadwater G, Peterson B. 49. Woo B, Dibble SL, Piper BF, Keating SB, Weiss MC. Differences in fatigue by
Complications of axillary lymph node dissection for carcinoma of the breast: treatment methods in women with breast cancer. Oncol Nurs Forum 1998;
a report based on a patient survey. Cancer 1998;83:1362 – 8. 25:915 – 20.
29. Paskett ED, Naughton M, Robertson J, Petrek J. Lymphedema in young 50. Witte MH, Wutte CL, Mortimer PS, Jamal S. Lymphedema in the
breast cancer survivors. Proc Am Assoc Cancer Res 2001;42:740. developing and developed world: contrasts and prospects. Lymphology
30. Lin PP, Allison DC, Wainstock J, et al. Impact of axillary lymph node 1988;21:242 – 3.
dissection on the therapy of breast cancer patients. J Clin Oncol 1993;11: 51. Kosorok MR, Omenn GS, Diehr P, Koepsell TD, Patrick DL. Restricted
1536 – 44. activity days among older adults. Am J Public Health 1992;82:1263 – 7.
31. Bartelin H, Garavaglia G, Johansson KA, et al. Quality assurance in 52. Beaulac SM, McNair LA, Scott TE, La Morte WW, Kavanah MT.
conservative treatment of early breast cancer: report on a consensus meeting Lymphedema and quality of life in survivors of early-stage breast cancer.
of the EORTC Radiotherapy and Breast Cancer Cooperative groups and the Arch Surg 2002;137:1253 – 7.
EUSOMA (European Society of Mastology). Radiother Oncol 1991;22:323 – 6. 53. Goffman TE, Laronga C, Wilson L, Elkins D. Lymphedema of the arm and
32. Kroman N, Jensen MB, Wohlfahrt J, Mouridsen HT, Andersen PK, Melbye breast in irradiated breast cancer patients: risks in an era of dramatically
M. Factors influencing the effect of age on prognosis in breast cancer: changing axillary surgery. Breast J 2004;10:405 – 11.
population based study. Br Med J 2000;320:474 – 8. 54. Hinrichs CS, Watroba NL, Rezaishiraz H, et al. Lymphedema secondary to
33. Talley LI, Grizzle WE, Waterbor JW, Brown D, Weiss H, Frost AR. Hormone postmastectomy radiation: incidence and risk factors. Ann Surg Oncol 2004;
receptors and proliferation in breast carcinomas of equivalent histologic 11:573 – 80.
grades in pre- and postmenopausal women. Int J Cancer 2002;98:118 – 27. 55. Johansson K, Ohlsson K, Ingvar C, Albertsson M, Ekdahl C. Factors
34. Shapiro CL, Recht A. Late effects of adjuvant therapy for breast cancer. associated with the development of arm lymphedema following breast
J Natl Cancer Inst Monogr 1994;16:101 – 12. cancer treatment: a match pair case-control study. Lymphology 2002;35:
35. Quivey J, Luce JM, Stewart SL, Johnston M, Banks PJ, Bloom JR. Younger 59 – 71.
women with breast cancer: patterns of care in 5 San Francisco Bay Area 56. Hu FB. Overweight and obesity in women: health risks and consequences.
Counties. In: Proceedings of the American Society of Clinical Oncology; J Womens Health 2003;12:163 – 72.
1998; Baltimore (MD); Waverly Press 1998. 57. Aronne LJ. Classification of obesity and assessment of obesity-related health
36. Osteen RT, Cady B, Friedman M, et al. Patterns of care for younger women risks. Obes Res 2002;10:105 – 15S.
with breast cancer. J Natl Cancer Inst Monogr 1994;16:43 – 6. 58. Denmark-Wahnefried W, Aziz NM, Rowland JH, Pinto BM. Riding the crest
37. SEER Cancer statistics review, 1975 – 2002. Bethesda (MD): National Cancer of the teachable moment: promoting long-term health after the diagnosis of
Institute; 2005. cancer. J Clin Oncol 2005;23:5814 – 30.
38. Petrek JA, Naughton MJ, Case LD, et al. Incidence, time course, and 59. Brown JK, Byers T, Doyle C, et al. Nutrition and physical activity during and
determinants of menstrual bleeding after breast cancer treatment: a after cancer treatment: an American Cancer Society guide for informed
prospective study. J Clin Oncol 2006;24:1045 – 51. choices. CA Cancer J Clin 2003;53:268 – 91.
39. Armer JM, Radina ME, Porock D, Culbertson SD. Predicting breast 60. Rock CL, Denmark-Wahnefried W. Nutrition and survival after the
cancer-related lymphedema using self-reported symptoms. Nurs Res 2003; diagnosis of breast cancer: a review of the evidence. J Clin Oncol 2002;20:
52:370 – 9. 3302 – 16.
40. Brady M, Cella D, Mo F. Reliability and validity of the functional assessment 61. Chlebowski RT, Aiello E, McTiernan A. Weight loss in breast cancer patient
of cancer therapy-breast (FACT-B) quality of life instrument. J Clin Oncol management. J Clin Oncol 2002;20:1128 – 43.
1997;15:974 – 86. 62. Nuver J, Smit AJ, Postma A, Sleijfer DT, Gietema JA. The metabolic
41. Ware J, Kosinski M, Keller SD. A 12-item short-form health survey: syndrome in long-term cancer survivors: an important target for secondary
construction of scales and preliminary tests of reliability and validity. Med measures. Cancer Treat Rev 2002;28:195 – 214.
Care 1996;34:220 – 33. 63. Argiles JM, Lopez-Soriano FJ. Insulin and cancer. Int J Oncol 2001;18:683 – 7.
42. Gandek B, Ware J, Aaronson NK, et al. Cross-validation of item selection 64. Petrek JA, Senie RT, Peters M, Rosen PP. Lymphedema in a cohort of breast
and scoring for the SF-12 Health Survey in nine countries: results from the carcinoma survivors 20 years after diagnosis. Cancer 2001;92:1368 – 77.
IQOLA Project, International Quality of Life Assessment. J Clin Epidemiol 65. Schijven MP, Vingerhoets AJJM, Rutten HJT, et al. Comparison of morbidity
1998;51:1171 – 8. between axillary lymph node dissection and sentinel node biopsy. Eur J Surg
43. Kaplan EL, Meier P. Nonparametric estimation from incomplete observa- Oncol 2003;29:341 – 50.
tions. J Am Stat Assoc 1958;53:457 – 81. 66. Deo SVS, Ray S, Rath GK, et al. Prevalence and risk factors for development
44. Cox DR. Regression models and life tables. J R Stat Soc Ser B 1972;34: of lymphedema following breast cancer treatment. Indian J Cancer 2004;41:
187 – 220. 8 – 12.
45. Good PI. Resampling methods. 2nd ed. New York: Springer; 2005. 67. Pereira de Godoy JM, Braile DM, de Fatima Godoy M, Longo O, Jr. Quality
46. Liang KY, Zeger SL. Longitudinal data analysis using general linear models. of life and peripheral lymphedema. Lymphology 2002;35:72 – 5.
Biometrika 1986;73:13 – 22. 68. Pain SJ, Vowler SL, Purushotham AD. Is physical function a more
47. Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous appropriate measure than volume excess in the assessment of breast
outcomes. Biometrics 1986;42:121 – 30. cancer-related lymphoedema (BCRL)? Eur J Cancer 2003;39:2168 – 72.

Cancer Epidemiol Biomarkers Prev 2007;16(4). April 2007

Downloaded from cebp.aacrjournals.org on October 12, 2015. © 2007 American Association for Cancer
Research.
The Epidemiology of Arm and Hand Swelling in
Premenopausal Breast Cancer Survivors
Electra D. Paskett, Michelle J. Naughton, Thomas P. McCoy, et al.

Cancer Epidemiol Biomarkers Prev 2007;16:775-782.

Updated version Access the most recent version of this article at:
http://cebp.aacrjournals.org/content/16/4/775

Cited articles This article cites 63 articles, 10 of which you can access for free at:
http://cebp.aacrjournals.org/content/16/4/775.full.html#ref-list-1

Citing articles This article has been cited by 11 HighWire-hosted articles. Access the articles at:
http://cebp.aacrjournals.org/content/16/4/775.full.html#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].

Downloaded from cebp.aacrjournals.org on October 12, 2015. © 2007 American Association for Cancer
Research.