This article was downloaded by: [Dalhousie University]

On: 30 December 2014, At: 23:41

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer
House, 37-41 Mortimer Street, London W1T 3JH, UK

Behavioral Medicine
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/vbmd20

A Comparison of Physical Activity of Posttreatment

Breast Cancer Survivors and Noncancer Controls
Chris M. Blanchard PhD , Vilma Cokkinides PhD , Kerry S. Courneya PhD , Eric J. Nehl
MS , Kevin Stein PhD & Frank Baker PhD
Published online: 25 Mar 2010.

To cite this article: Chris M. Blanchard PhD , Vilma Cokkinides PhD , Kerry S. Courneya PhD , Eric J. Nehl MS , Kevin
Stein PhD & Frank Baker PhD (2003) A Comparison of Physical Activity of Posttreatment Breast Cancer Survivors and
Noncancer Controls, Behavioral Medicine, 28:4, 140-149, DOI: 10.1080/08964280309596052

To link to this article: http://dx.doi.org/10.1080/08964280309596052

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”)
contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors
make no representations or warranties whatsoever as to the accuracy, completeness, or suitability
for any purpose of the Content. Any opinions and views expressed in this publication are the opinions
and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of
the Content should not be relied upon and should be independently verified with primary sources of
information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands,
costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or
indirectly in connection with, in relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or
systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in
any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
 A Comparison of Physical Activity
of Posttreatment Breast Cancer Survivors
and Noncancer Controls
Chris M. Blanchard, PhD; Vilma Cokkinides, PhD; Kerry S. Courneya, PhD;
Eric J. Nehl, MS; Kevin Stein, PhD; Frank Baker, PhD

The authors obtained data from 335 breast cancer survivors and 6,880 non-
Downloaded by [Dalhousie University] at 23:41 30 December 2014

cancer controls. They proposed ( I ) to determine whether; after treatment, the

survivors were meeting the Center for Disease Control and PreventiodAmeri-
can College of Sports Medicine recommendationsfor physical activity and were
similar to the controls in physical activity and ( 2 ) to compare the modes of
activity of the 2 groups in frequency, midsession, and sessions/wk. Adjusted
logistic regression analyses revealed that the breast cancer survivors engaged
in as much moderate, vigorous, and combined physical activity as the non-
cancer controls did. However; chi-square analyses showed that survivors
engaged in more yard work than the controls did, whereas independent-sample
t tests showed that the frequency and the total midwk of stretching were sign$-
icantly higher in breast cancer survivors compared with noncancer controls.
Findingsfrom the study suggest that breast cancer survivors engage in as much
physical activity as controls do, but that the groups differ in specific activities.
Index Terms: breast cancer survivors, noncancer controls, physical activity,
treatment

Despite the positive physical, functional, and psychologi-
call4 benefits that posttreatment breast cancer survivors
(BCS) report when they engage in physical activity, uncer-
tainty remains about how physically active these survivors
are. One partial explanation for this uncertainty is the lack
of consistency in the operationalization of their physical
activity in previous studies. For example, Young-McCaugh-
an and Sexton' found that 41% of BCS did not engage in
Dr Blanchard is an assistant professor with the Faculty of Health
Sciences in the School of Human Kinetics, the University of
Ottawa, Ontario, Canada. Mr Nehl, Dr Stein, and Dr Baker are
with the Behavioral Research Center; American Cancer Society,
Atlanta, Georgia, where Dr Cokkinides is with Risk Factor Sur-
veillance in Epidemiology. Dr Courneya is with the Faculty of
Physical Education, University of Alberta, Edmonton, Canada.
physical activity at least 2 times/wk. Leddy* reported that
38% engaged in physical activity occasionally or not at all,
whereas McBride and associates9 found that 44% did not
engage in physical activity on a weekly basis. Finally, Cour-
neya and Friedenreich'" reported that survivors engaged in
moderate/vigorous physical activity an average of 2.68
times/wk.Thus, inconsistency in measures of physical activ-
ity across studies has made it difticult to compare findings
and ascertain how physically active survivors are.
One way to overcome this limitation is to measure physi-
cal activity on the basis of recommendations suggested by the
Centers For Disease Control and PreventiotdAmerican Col-
lege of Sports Medicine (CDC/ACSM). That is, to accumu-
late 30 or more minutes of moderately intense physical activ-
ity on 5 or more d/wk or vigorously intense physical activity

140

3 times/wk for at least 20 minhession." In using these rec-
ommendations, researchers can compare activity across stud-
ies and can discern how physically active posttreatment BCS
are compared with the national averages for healthy women.
Pinto and associates6 used the CDC/ACSM recommen-
dations and found that only 20% of the BCS were meeting
the standards, a level that was below the national average of
25.5% for women.I2 Although that study was a significant
improvement over previous survivor studies in terms of
measuring physical activity, it was similar to previous stud-
ies in that the participants knew they were being recruited
into a study of physical activity and cancer. The transparent
purpose of that study may have caused a selection bias that
attracted participants who were more active physically. If
Downloaded by [Dalhousie University] at 23:41 30 December 2014
this was known, it is highly likely that sedentary adult can-
cer survivors were underrepresented in the sample; if
included, they might have lowered the percentage (ie, 20%)
of survivors who met the CDC/ASCM recommendations
and limited the generalizability of the findings.
In addition, the sample in that study (as with the other
survivor studies) was opportunistic. The physical activity
findings from this opportunistic and potentially biased sam-
ple might not be indicative of physical activity in BCS in
the population at large." Hence, additional research is
needed that (1) does not use physical activity as the focal
point in an attempt to recruit posttreatment survivors who
are not interested in physical activity and (2) uses nonop-
portunistic samples (eg, randomly selected population-
based participants) of survivors after treatment. Doing so
should increase the potential generalizability of the findings
to survivors in the general population.
Another limitation in the preceding studies is the absence
of a noncancer control group. Indeed, Nelson4 suggested the
importance of researchers' inclusion of such a group when
examining health behaviors such as physical activity among
posttreatment breast cancer survivors. The experience of
having had breast cancer has been said to lead to permanent
changes in a woman's view of the world,4 changes that
might make her views different from the views of women
who have never experienced breast cancer. Such disparate
views could also have a different influence on the women's
health behaviors.
Nelson's study: which compared physical activity levels
in after-treatment BCS with age-matched healthy controls,
reported that 74% of the survivors reported participating in
some form of physical activity, compared with 85% of age-
matched healthy controls. These data supported previous
s t ~ d i e s ' ~and
. ' ~ suggested that the cancer experience may
have had a negative effect on the survivors' physical activi-
ty after treatment. However, Nelson's study was limited by
Vol 28, Winter 2003
BLANCHARD ET AL
the use of a potentially biased and small ( n = 5 5 ) oppor-
tunistic sample, as were the previous physical activity and
cancer studies. Therefore, replication is warranted before
one can draw any firm conclusions.
Activities that BCS engage in after they complete treat-
ment and whether these activities are similar to those of con-
trols is another limitation in past research that has received
little attention. Nelson4 found that survivors' most commonly
used activities were walking (60%), stretching (1 5%), sta-
tionary cycling (lo%), aerobics (5%), and swimming (5%).
Their age-matched healthy controls frequently used walking
(37%), aerobics (30%), biking (1 l%), jogging (4%), and
swimming (4%). Although these percentages suggest that the
activities survivors and healthy women engaged in were quite
similar, it appears that nearly twice as many survivors as
healthy controls were walkers. On the other hand, nearly
twice as many healthy controls as survivors engaged in aero-
bics. Although each group's activities appear to be similar,
the way they use them might still be different. It would be
useful to measure the frequencylwk and duration of each
activity in each group, and we believe that future research
should attempt to shed light on these issues.
Given that the number of after-treatment breast cancer
survivors is steadily increasing15 and the number of studies
describing their physical activity is so limited, further
research to reflect physical activity in the general popula-
tion more accurately is needed. In particular, previous limi-
tations, such as biased and opportunistic samples, lack or
neglect of information about the styles of activities, and
inconsistency in operationalizations of physical activity
should be considered. Therefore, we used data from the
1998 National Health Interview Survey (NHIS) in the pre-
sent study. The NHIS provided health-related information
from an unbiased, nationally representative civilian sample
(note that the focal point of the survey was not physical
activity) to examine physical activity among women with
self-reported diagnoses of breast cancer and noncancer con-
trols (referred to throughout as controls).
Our first goal was to determine whether the proportion of
after-treatment beast cancer survivors who met the CDC/
ACSM recommendations for moderate, vigorous, and com-
bined physical activity was similar to those of the controls.
Our second purpose was to compare (1) modes of activities
in each group and (2) the frequency, minlsession, and total
minlwk for each group.
METHOD
Data Source
The NHIS, a multipurpose health survey conducted by
the CDC National Center for Health Statistics, is the princi-
141
 BREAST CANCER SURVIVORS
pal source of information on the health of noninstitutional-
ized US civilian households. It has been conducted continu-
ously since 1957 and data are released annually. The popu-
lation-based NHIS provides national data on the incidence
of illness, accidental injuries, chronic conditions and
impairments, extent of disability, the use of healthcare ser-
vices, and prevalence of selected health-related behaviors
used in monitoring the Healthy People 2010 objectives.16
The US Office of Management and Budget has approved the
study and participants provided informed consent when the
survey instrument was
The NHIS uses a complex, stratified, multistage sampling
design in which the primary sampling units are individual
counties or contiguous groups of counties. The interviewed
Downloaded by [Dalhousie University] at 23:41 30 December 2014
adult sample for 1998 consisted of 32,440 randomly sam-
pled persons aged 18 years and older (18,238 women and
14,202 men) from 38,209 households or 38,773 families.
The response rate for the adult sample was 83.8%, and the
final response rate for the adult sample components was
73.9%. The latter was calculated as the product of the
response rates of the adult sample and of the family sample.
Participants
The NHIS used in-person interviews to obtain demo-
graphic characteristics and health-related information on
everyone living in a surveyed household. Of the 18,238
women who participated in the study, we included 17,233 in
the analytic sample on the basis of complete valid data on
health status (chronic diseases) and physical activity. We
classified the women by their self-reported breast cancer
diagnosis, resulting in 2 groups- 335 women who had had
breast cancer (eg, BCS) and 16,880 who had not had a
breast cancer diagnosis (ie, controls). The NHIS made no
attempt to validate diagnoses, although the NHIS data on
chronic diseases (including major cancers) is considered
reliable. It is widely used in public-health monitoring of US
adults’ health status and health-related behaviors. We used
self-reported information on leisure-time physical activity
of women from the core and adult prevention sections of the
NHIS questionnaire to examine and compare patterns of
leisure physical activities between survivors and controls.
Measures
Demographics
Information included self-reported information on age,
race, education, income, marital status, smoking, height,
weight, obesity (body mass index [BMI] greater than 30
kg/m2 is defined as obese; BMI less than 30 kg/m2is defined
as nonobese), and time since cancer diagnosis (survivors
only). We also obtained information concerning the pres-
142
ence (or absence) of comorbidities known to influence
physical activity for all participants (ie, hypertension, coro-
nary heart disease, angina, myocardial infarction, stroke,
emphysemdasthma, diabetes, and chronic bronchitis). We
weighted the comorbidities equally and summed them to
obtain an overall comorbidity score.
vigorous Physical Activity
We used the following 2 questions from the NHIS survey
to assess vigorous physical activity: “How often do you do
vigorous activities for at least 10 minutes that cause heavy
sweating or large increases in breathing or heart rate?’ and
“About how long do you do these vigorous activities each
time?’ Because the possible time frames for the frequency
question ranged from daily to weekly to monthly to yearly,
we rescaled the question so that all frequencies were on a
weekly basis. We wanted to determine whether participants
in the sample met CDC/ACSM” recommendations for vig-
orous physical activity.
Light/Moderate Physical Activity
To make this assessment, we used the following 2 ques-
tions from the NHIS survey: “How often do you do light or
moderate activities for at least 10 minutes that cause only
light sweating or a slight to moderate increase in breathing
or heart rate?’ “About how long do you do these light or
moderate activities each time?’ As we had done with the
vigorous frequency question, we used the weekly time
frame to rescale the moderate frequency question.
Exercise Modes
We took the frequency and duration of 9 modes of phys-
ical activity from the NHIS survey. More specifically, the
questions asked participants, “How many times in the past
2 weeks did you (1) walk for exercise? (2) do gardening or
yard work? (3) do stretching exercises? (4) do weight lifting
or other exercises to increase muscle strength? (5) jog or
run? (6) do aerobics or aerobic dancing? (7) ride a bicycle
or exercise bike? (8) stair climb for exercise? and (9) swim
for exercise?’ For each activity, we also asked the partici-
pants “On the average, about how many minutes did you.. .”
followed by the specific activity? To be consistent with the
vigorous and moderate activity questions, we rescaled all
activities so that all frequencies were on a weekly basis and
we multiplied the frequency and intensity of each activity to
obtain the total midwk for each activity.
Analysis
The data collected for the NHIS was obtained through a
complex sample design involving stratification, clustering,
Behavioral Medicine
 BLANCHARD ET AL

and multistate sampling, with the sample chosen so that

each person in the covered population had a known non- TABLE 1. Demographic Characteristics of the Breast
Cancer Survivors and Noncancer Controls
zero probability of being selected. These selection probabil-
ities, along with adjustments for nonresponse and poststrat-
ification are reflected in the sample weights. Because of the Noncancer
Cancer control
complex NHIS sampling plan and the use of adjusted sam-
(n = 353) ( n = 16,880)
pling weights, the direct application of standard analysis
Characteristic M SE M SE
methods for estimation and hypothesis testing might yield
misleading results. Therefore, we used SUDAAN a statisti-
cal software package that uses a series of linearization meth- Time since diagnosis (y) 10.I 0.6
Age (Y)* 66.3 0.8 44.5 0.2
ods to compute the appropriate variance estimates on the
basis of the complex sample design.”
% SE % SE
Purpose 1
Downloaded by [Dalhousie University] at 23:41 30 December 2014

Race* *
We conducted multivariate logistic regression analyses in White 90.2 1.7 81.4 0.4
SUDAAN to determine whether the proportion of after- Other 9.8 1.7 18.6 0.4
treatment BCS who were meeting the CDC/ACSM recom- Education**
mendations for moderate, vigorous, and combined physical 5 High school 58.9 3.1 49.5 0.5
activities was similar to the proportions of the controls, after 2 High school 41.1 3.1 50.5 0.5
Income ($/y)
controlling for age, obesity, race, education, income, smok- < $20,000 30.3 2.6 25.1 0.5
ing, and comorbidity. (Nearly two thirds of the BCS had 2 $20,000 69.7 2.6 74.9 0.5
survived breast cancer for 5 years or longer. Because the Marital status**
mean time since diagnosis was 10 years, we examined this Marriedcommon-law 57.7 2.7 62.0 0.5
group [controlling for demographic variables] to determine Separateddivorced 11.6 1.7 11.3 0.3
Widowed 26.7 2.5 9.7 0.3
whether the interval between diagnosis and age at comple- Single 4.0 1.1 17.0 0.4
tion of the survey was related in any way to meeting the Smoking**
CDC/ACSM physical activity recommendations. The ad- Nonsmoker 58.0 2.9 60.0 0.5
justed odds ratio [ R = 0.66; 95% CI = 0.35-1.251 suggested Former smoker 28.8 2.6 18.1 0.4
that the time since diagnosis did not influence survivors’ Current smoker 13.2 2.2 21.9 0.4
Obesity (ns)
physical activity levels.) Nonobese 79.8 2.5 80.1 0.4
Obese 20.2 2.5 19.9 0.4
Purpose 2 Body mass index 28.5 13.7 24.1 3.1
To determine whether the percentages of posttreatment Comorbidit y
survivors who engaged in the various activities was similar to ( M number) 1.8 0.1 1.5 0.01
Comorbidity**
the percentages of the noncancer controls, we performed 1 comorbidity 51.0 3.9 67.8 0.7
weighted chi-square analyses. In addition, we performed 2 comorbidity 28.6 3.4 19.6 0.6
independent sample i tests to learn whether the total frequen- 2 3 comorbidity 20.4 3.2 12.6 0.5
cies/wk, duratiodsession, and total midwk of each activity
were similar in the 2 groups. Unfortunately, because of the Note. ns = nonsignificant test; SE = standard error.
small sample of survivors in each activity and the lack of *Independent sample r tests significant at p < .01; **x2 significant
atp<.01.
power to detect group differences for a specific activity, we
could not control for potential confounders as we had earlier.
the survivors were significantly older (age M = 66.3) than
RESULTS their control group counterparts (age M = 44.5). Further-
more, our chi-square analyses showed that the survivors
Survivor Characteristics
were more likely to be White (90.2%), less educated
For details of the demographic profiles of the BCS and (58.9%), widowed (26.7%), former smokers (28.8%), and
controls, see Table 1. The survivors, with an average time to have had 3 or more comorbidities (20.4%) than were the
since diagnosis of 10.1 years, would be considered long- controls (81.4% White; 49.5% less than high school educa-
term survivors. An independent sample r test showed that tion; 9.7% widowed; 18.1% former smokers; 12.6% at least

Vol 28, Winter 2003 143

 BREAST CANCER SURVIVORS

TABLE 2. Unadjusted and Age-Adjusted Percentages and Standard Errors (SE)

for Breast Cancer Survivors (BCS) and Noncancer Controls for LighUModerate,
Vigorous, and Total Activity

BCS Noncancer controls

Unadjusted Adjusted Unadjusted Adjusted
% % % %
Activity SE SE SE SE
I
Lighvmoderate 11.1 12.2 12.0 11.9
7.4-14.8 4.0-20.7 11.4-12.6 I 1.3-1 2.5
Vigorous 14.7 13.0 17.6 16.7
10.5-18.8 7.7-1 8.4 16.8-18.3 16.0-1 7.3
Total 23.4 22.8 24.9 24.1
Downloaded by [Dalhousie University] at 23:41 30 December 2014

18.2-28.5 13.G32.5 24.1-25.6 23.3-24.8

3 or more comorbidities). Both groups, however, were sim-
ilarly distributed in income and obesity.
Purpose 1: Were Physical Activities Similar
in BCS and Controls?
The data in Table 2 show the unadjusted and adjusted
weighted percentages and standard errors for lighdmoderate,
vigorous, and combined physical activities. The age- adjust-
ed percentages of survivors and controls who met the rec-
ommended standards of physical activity were quite similar.
We found it interesting that unadjusted odds ratios (OR)
from Table 3 suggested that survivors were 9% less likely to
meet the lighdmoderate physical activity levels than the con-
trols, 19% less likely to meet the vigorous physical activity
recommendation, and 8% less likely to meet either or both of
the physical activity recommendations. When we controlled
for demographic variables, we found that survivors were
42% more likely than controls (95% CI = 0.97-1 20) to meet
the vigorous physical activity recommendations and 28%
more likely than controls (95% CI = 0.91-1.80) to meet the
combined physical activity recommendations.
Purpose 2a: Were Modes of Activity Similar
in Survivors and Controls?
The percentages of survivors and controls who engaged
in the various modes of activity are shown in Table 4. Our
chi-square analyses showed that a greater percentage of
controls than survivors engaged in running (controls’ run-
ning = 8.9%, survivors’ running = 3.9%) and aerobics (con-
trols’ aerobics = 11 .O%, survivors’ aerobics = 5.8%). By
contrast, BCS engaged in more yard work (37.4%) than
controls (27.6%) did. We found no group differences for the
remaining physical activities. Finally, it appeared that the 5
activities most frequently engaged in (ie, walking, yard
work, stretching, biking, and weight training) were similar
for both groups.
Purpose 2b: Were Mode Frequency and Duration
Similar in BCS and Controls?
Means and standard deviations of frequency, mitdsession,
and total mitdwk of engaging in each physical activity, by
group, are also shown in Table 4. Independent sample f tests
showed that the frequency and total min/wk of stretching
was significantly higher in survivors (frequency M = 4.4;
total minutes M = 55.2) compared with the controls (fre-
quency M = 3.4; total minutes M = 42.9). However, we
found no differences for the 8 remaining physical activities.
COMMENT
Previous research that examined physical activity in
breast cancer survivors has been limited because of biased
and opportunistic samples, lack of controls, neglect of
attention to modes of activity, and inconsistency in opera-
tionalizing physical activity. To overcome these limitations,
we first analyzed data from the 1998 NHIS. In doing so, we
compared the proportions of after-treatment BCS who met
the CDC/ACSM recommendations for moderate, vigorous,
and combined physical activity with noncancer controls.
Second, we compared the kinds of activities that each group
engaged in and the frequency, min/session, and total
midwk for each activity.
Purpose 1
Previous researchers showed that 15% to 44% of post-
treatment BCS engaged in some form of physical activity in
the 5 years after their cancer d i a g n o s i ~ . ~ -Pinto
’ - ~ - ~and
~ asso-

144 Behavioral Medicine

 Downloaded by [Dalhousie University] at 23:41 30 December 2014

TABLE 3. Unadjusted and Adjusted Prevalence Odds Ratio for Physical Activity Pattern

Lighl/mderdIe activities Vigorous activities Either or both

5 times f o r t 30 midwk Wald 3 times for t 20 midwk Wald ACSM recommendation Wald
Unadj Adj adj Unadj Adj adj Unadj Adj adj
Covariates OR OR 959CI P OR OR 95%CI P OR OR 95’7rCI P

Group ,8303 .074 .161

BCS 0.91 0.95 0.62-1.46 0.8 I 1.42 0.97-2.10 0.92 I .28 0.91-1.80
Control 1 .oo I .oo 1 .oo 1 .oo I .oo 1 .oo
Age (Y) .o009 .m .m
1849 1.46 I . I9 0.97- I .47 5.62 4.36 3.40-5.60 2.8 1 2.2 1 1.8&2.64
50-59 I .44 1.25 0.99-1 .57 3.83 3.19 2.384.26 2. I8 I .83 1 .SO-2.24
6M9 I .77 1.64 1.28-2. I I 2.57 2.47 1.863.28 I .98 I .86 1.52-2.29
70- 85 I .oo 1 .oo I .oo I .oo 1 .oo I .oo
Race .ooOl .oooo .m
Non-White 0.70 0.72 0.614.85 0.63 0.66 0.564.78 0.66 0.68 0.614.77
White 1 .oo I .oo I .oo I .oo 1 .oo 1 .oo
Education .m .m .m
5 High school 0.63 0.7 I 0.624.80 0.35 0.47 0.424.52 0.43 0.53 0.484).58
> High school 1 .oo I .oo 1 .oo 1 .oo 1 .oo 1 .oo
Income level ($) .m .m .m
< 2O,OOo/y 0.60 0.7 I 0.624.83 0.39 0.65 0.584.74 0.49 0.69 0.624.77
t 20,OOo/y I .oo 1 .oo I .oo I .oo I .oo 1 .oo
Obesity ,0024 .m .m
Obese 0.73 0.78 0.674.92 0.46 0.51 0.44-0.59 0.55 0.60 0.53-0.68
Non obe se 1 .oo I .oo I .oo I .oo I .oo 1 .oo
Smoking .4n I .m .m
Current 1.03 1.04 0.88-1.23 0.8 I 0.80 0.714.91 0.85 0.85 0.76496
Former 1.19 1.11 0.95-1.30 1.15 1.22 1.05-1.40 1.18 I .20 1.06-1.35
Nonsmoker I .oo 1 .oo 1 .oo 1 .oo 1 .oo I .oo
t 1 comorbidily .5 18 .oo8 ,024
Yes 0.86 0.95 0.83-1.10 0.54 0.84 0.74-0.95 0.64 0.88 0.794.98
No I .oo 1 .oo 1 .oo 1 .oo 1.oo I .oo

Nofe.BCS = breast cancer surv~vors:OR =odds ratio: CI = confidence interval: ACSM =American College of Sports Medicine

b
m
-1
 Downloaded by [Dalhousie University] at 23:41 30 December 2014

TABLE 4. Activity Comparisons Between Breast Cancer Survivors and Healthy Controls: Frequency of Participating in the Activity, Means and Standard Errors for
the Frequency and Duration of Activities Used by Breast Cancer Survivors and Noncancer Controls

Breast cancer (n = 264) Healthy control (n = 1,472)

Yes Yes
% sE Frequency/wk Midsession Total midwk % SE Frequency/wk Midsession Total midwk
Activity “I M SE M SE M SE “I M SE M SE M SE

Walking 53.4 3.5 3.5 0.2 32.9 1.4 118.8 9.4 53.4 0.5 3.2 0.03 35.9 0.4 116.9 2.4
ns [I481 [7710]
Yard work 37.4 3.5 1.8 0.2 74.6 6.9 134.0 16.2 27.6 0.5 1.8 0.03 72.5 1.3 137.3 4.4
p . x 2 = .007 [971 [3829]
Stretching 32.3 3.5 4.4 0.4 15.1 1.6 55.2 5.1 30.4 0.5 3.4 0.04 13.0 0.2 42.9 1.0
ns [831 [4330]
Weight training 10.1 2.0 2.8 0.4 21.9 2.9 52.8 7.0 13.2 0.3 2.6 0.05 26.4 0.6 68.4 2.2
ns [281 [I8161
Running 3.9 1.2 2.4 0.4 48.2 23.5 88.4 34.7 8.9 0.3 2.3 0.05 28.3 1.0 72.5 3.7
p. x 2 = .oO01 PI [1228]
Aerobics 5.8 1.5 1.7 0.2 48.0 4.0 86.6 14.9 11.0 0.3 2.3 0.05 39.2 0.8 86.9 3.0
p, x 2 = ,001 ~ 7 1 [1510]
Biking 14.3 2.4 2.6 0.4 27.0 5.7 66.0 19.0 12.2 0.3 2.0 0.05 30.4 0.8 57.2 2.0
ns [361 [ 17421
Stair climbing 6.5 1.7 2.9 0.5 11.0 2.6 36.3 10.0 8.4 0.3 3.5 0.2 15.6 0.5 43.9 2.2
ns 1171 [ 12201
Swimming 6.3 1.7 1.9 0.3 29.8 5.7 51.3 10.7 5.8 0.2 1.9 0.09 41.7 1.5 78.7 4.0
ns [151 17981

Note. ns = no significant differences in x2 analyses for group differences, cancer survivors vs healthy controls on the percentage of participants involved in each activity unless otherwise indicated; 2
independent-sample t tests only showed a significant group difference on frequencylwk stretching and total midwk stretching.

ciates6 showed that 20% of this group met the CDC/ACSM
recommendations in a biased and opportunistic sample.
Surprisingly, the (23%) of survivors in the present study
who met the recommendations was greater than the per-
centages in Pinto and associates' studies. These conflicting
findings probably resulted from our moderate intensity def-
inition's having included light activities, which may have
inflated the percentages of survivors who met the moderate
activity guideline. Nonetheless, both studies demonstrated
that survivors were slightly below the national average (ie,
25.5%) for healthy women.I2
When we compared the BCS to the noncancer controls,
we found that survivors were engaging in as much moder-
ate physical activity as the noncancer controls were. In fact,
Downloaded by [Dalhousie University] at 23:41 30 December 2014
controlling for important covariates, we discovered that the
survivors engaged in more vigorous and combined physical
activity than the controls did (although the difference was
not statistically significant). This finding is very encourag-
ing because it suggests that long-term survivors may be
incorporating physical activity into their daily lives as
much, and in some cases more, than the controls are. Fur-
thermore, our survivor analyses showed that time since
diagnosis (ie, S 5 y versus > 5 y) did not influence whether
the survivors met the CDC/ACSM physical activity recom-
mendations, which is consistent with findings in previous
research.'0.20Thus, short- and long-term posttreatment BCS
appeared to be engaging in as much physical activity as the
noncancer controls were.
Although these findings are promising, readers should
note 2 important limitations. First, previous research1°.20
showed that time since diagnosis was not related to the fre-
quency of survivors' physical activity in the 5 years since
diagnosis. We found that physical activity was similar in
survivors after less than 5 years and greater than or equal to
survivors after 5 years. These are cross-sectional findings,
therefore future studies should incorporate long-term fol-
low-ups in survivors that use prospective designs to allow
for comparisons at different time points along the cancer
continuum. In doing so, researchers might be able to iden-
tify fluctuations in physical activity as survivors progress
farther from their diagnosis (ie, they will be able to exam-
ine survivors' changes over time).
Second, although long-term BCS may engage in similar
(or more) moderate, vigorous, and combined physical
activity compared with short-term survivors and controls,
approximately 77% of these survivors still do not meet the
CDC/ASCM" physical activity recommendations. Conse-
quently, this might suggest that it is particularly important
to motivate long-term survivors to engage in physical
activity to alleviate their increased risks of developing sec-
Vol 28, Winter 2003
BLANCHARD ET AL
ondary tumors, cardiovascular disease, osteoporosis, and
Purpose 2
Our second purpose was to compare BCS' and controls'
weekly activities. We found that the top 5 activities of both
survivors and controls were similar (eg, walking, yard
work, stretching. weight training, and biking) and were the
same 5 top-ranked activities as the CDC24reported for US
residents generally. Similarly, Nelson4 reported that approx-
imately 4.5 years after diagnosis and treatment, BCS indi-
cated that walking was their dominant physical activity
(approximately 53% of both groups walked weekly).
However, the top 5 activities for BCS in Nelson's study
differed from the favored activities we found in our investi-
gation. More specifically, our findings in the present study
replaced Nelson's aerobic and swimming activities with
weight training and yard work (walking, stretching, and
biking were the similar activities). Differences between the
2 studies may be a result of the sampling methodology: the
present study used a representative sample of US adult
women, whereas Nelson used a convenience sample. They
could also be attributable to the fact that the time since diag-
nosis was different in the 2 studies. It may also be that as
survivors move farther from their cancer diagnosis, they
change activities. This, of course, is only speculation. Fur-
ther research comparing BCS' modes of physical activities
at different times along the cancer continuum is warranted
before one can draw any firm conclusions.
Another novel finding of the study we have described in
this article was that a significantly larger percentage of BCS
engaged in yard work (37.4%) compared with controls
(27.6%). Furthermore, although similar percentages of sur-
vivors and controls used stretching as an activity, the BCS
engaged in this activity significantly more often and for
longer perioddwk than the controls did. Therefore,
although walking is the dominant activity of both groups,
yard work and stretching are activities favored more by sur-
vivors than by controls. This could be because the survivors
were significantly older than the controls. Unfortunately,
our analyses did not allow us to control for potential con-
founders such as age because of the small numbers of par-
ticipants in each activity. Nonetheless, our current study is
novel in making direct comparisons among physical activi-
ties in survivors and controls, which suggests that the can-
cer experience may influence the choice, frequency, and
duration of activities. Future studies should obtain larger
numbers of survivors who engage in each activity to get a
more accurate assessment and comparisons for covariates
that could potentially confound the findings.
I47
 BREAST CANCER SURVIVORS
Limitations and Future Directions
Despite the important descriptive findings of our present
study, some limitations must be considered. First, our study
used self-reports, which are typically inferior to objective
indicators, to assess physical activity. Second, we used a
cross-sectional design; therefore, we assumed that the one-
time measure of leisure activity is a characteristic pattern
and does not account for sudden changes because of unan-
ticipated health-related physical limitations or other life cir-
cumstances. Future investigations should examine physical
activity prospectively and include objective indicators to
allow the researchers to determine whether the types of
physical activity vary over time. A final limitation is the lack
of cancer-stage assessment. Given that previous research'0.20
Downloaded by [Dalhousie University] at 23:41 30 December 2014
with BCS has shown that cancer stage does not influence the
frequency of engaging in physical activity, we do not believe
that this was a grave concern. Nonetheless, these studies of
survivors were conducted within 5 years of diagnosis. Fur-
ther studies in longer term survivors are therefore warranted.
Conclusion
We performed secondary analyses on a US-population-
based sample taken from the 1998 NHIS in an attempt to
overcome previous limitations in the literature on physical
activity and breast cancer. We obtained a less biased, nonop-
portunistic sample of BCS that included a noncancer control
group. Survivors, we found, were engaging in as much mod-
erate, vigorous, and combined physical activity as controls
were. However, a significantly greater percentage of BCS
were engaging in yard work and stretching significantly
more often and for longer periods of time than the controls
were. We suggest that future studies (1) incorporate objec-
tive physical activity assessments, (2) use prospective
designs, and (3) sample sufficient numbers of BCS at vari-
ous times along the cancer continuum for a more accurate
and reliable assessment of survivors' and controls' physical
activity patterns. Nonetheless, our findings are promising
because they suggest that longer term survivors may be
incorporating as much physical activity in their daily lives
as controls are.
NOTE
For comments and further information, please address correspon-
dence to Chris M. Blanchard, PhD, 125 University Street, Ottawa,
Ontario, Canada KIN 6N5 (e-mail: chblanch@uottawa .ca).
REFERENCES
1. Neiman DC, Cook VD, Henson DA, et al. Moderate exercise
training and natural killer cell cytotoxic activity in breast can-
cer patients. Int J Sports Med. 1995;16:334-337.
148
2. Schulz KG, Szlovak C, Schulz H, et al. Implementation and
evaluation of an ambulatory exercise therapy based rehabilita-
tion program for breast cancer patients. Psychother Psycho-
som Med Psychol. 1998;48(9/10):45-58.
3. Segar MS, Katch VL, Roth RS, et al. The effect of aerobic
exercise on self-esteem and depressive and anxiety symptoms
among breast cancer survivors. Oncol Nurs Forum. 1998:
259(1):107-113.
4. Nelson JP. Perceived health, self-esteem, health habits, and
perceived benefits and barriers to exercise in women who have
had and who have not experienced stage I breast cancer. Oncol
Nurs Forum. 1991:18:1191-211.
5. Baldwin MK, Coumeya KS. Exercise and self-esteem in
breast cancer survivors: An application of the exercise and
self-esteem model. J Sport Exerc Psychol. 1997;19: 347-359.
6. Pinto BM, Maruyama NC, Engebretson TO, Thebarge RW. Par-
ticipation in exercise, mood, and coping in survivors of early
stage breast cancer. J Psychosoc Oncol. 1998;16(2):45- 58.
7. Young-McCaughan S, Sexton DL. A retrospective investiga-
tion of the relationship between aerobic exercise and quality of
life in women with breast cancer. Oncol Nurs Forum. 1991;
8~751-757.
8. Leddy SK. Incentives and barriers to exercise in women with
a history of breast cancer. Oncol Nurs Forum. 1997;24(10):
1715-1 723.
9. McBride CM, Clipp E, Peterson B, Lipkus I, Demark-Wahne-
fried W. Psychological impact of diagnosis and risk reduction
among cancer survivors. Psychooncology. 2000;9:418427.
10. Courneya KS, Friedenreich CM. Relationship between exer-
cise during treatment and current quality of life among sur-
vivors of breast cancer. J Psychosoc Oncol. 1997;15:35-57.
1 1. Pate RR, Pratt M, Blair SN, et al. Physical activity and public
health: A recommendation from the Centers for Disease Con-
trol and Prevention and the American College of Sports Med-
icine. JAMA. 1995;273:402407.
12. National Center for Chronic Disease Prevention and Health
Promotion. Percent of U S . adults with the recommended level
of physical activity by state and gender http://apps.nccd.
cdc.gov/dnpaltarecsex.htm.Accessed June 28, 2002.
13. Lohr S. Sampling: Design and Analysis. Toronto, Ontario;
Duxbury Press; 1999.
14. Rhodes RE, Courneya, KS, Bobick TM. Personality and exer-
cise participation across the breast cancer experience. Psy-
chooncology. 2001 ;10:38&388.
15. Cancer Facts and Figures. Atlanta, Georgia: American Cancer
Society; 2002.
16.Healthy People 2010. Available from www.healthypeople.gov.
Accessed 29 August, 2003.
17. National Health Interview Survey: Research for the 1995-
2004 Redesign. Vital Health Stat 2.1999;No 126.
18. Design and Estimation for the National Health Interview Sur-
vey, 1995-2004. Vital Health Stat 2. 2000;Nol30.
19. Shah BV, Barnacle BG, Bier GS. SUDAAN User's Manual:
Release 8.0. Research Triangle Park, NC: Research Triangle
Institute: 1998.
Behavioral Medicine
 BLANCHARD ET AL

20. Blanchard C, Courneya KS, Rodgers W, Murnaghan D. Deter-
minants of exercise intention and behavior in breast and
prostate cancer survivors: An of the theory Of
planned behavior. Cancer Nurs. 2002;25(2):88-95. Bines J,
Gradishar WJ. Primary care issues for the breast cancer sur-
vivor. Compr Ther 1997;23(9):605-611.
21. Bines J,Gradishar W. Primary care issues for the breast cncer
survivor. Cornpr TIM 1887;23:605-61 I .
22. Brown BW, Brauner C, Minnotte MC. Noncancer deaths in
white adult cancer patients. J Natl Cancer Inst. 1993;85(12):
979-987.
23. Talamini R, Franceschi S, Favero A, et al. Selected medical
conditions and risk of breast cancer. Br J Cancer 1997;
75(11):699-1703.
24. Participation in physical activities: Adults aged I X years und
older. CDC National Center for Chronic Disease Prevention
and Health Promotion. Available at http:l/apps.nccd.cdc.gov/
dnpa/pasports.htm.Assessed June 28, 2002.

Announcing the
Academy of Behavioral Medicine Research
Neal E. Miller New Investigator Award
Downloaded by [Dalhousie University] at 23:41 30 December 2014

The annual Neal E. Miller New Investigator Award is presented for work imagina-
tively conceived and carefully conducted before the recipient’s appointment as an
assistant professor or an equivalent rank. Miller, the first president of the academy,
was a pioneer in the application of learning theory to behavioral therapies and the
use of chemical and electrical stimulation to analyze the brain’s mechanisms of
behavior, homeostasis, and reinforcement.
“Be bold in what you try; cautious in what you claim,” Miller said. This award is
meant to encourage young scientists to pursue research in that spirit. The award, to
be presented to MD or PhD recipients, will consist of a plaque, a cash grant, and an
opportunity to attend the annual meeting of the academy as its guest. The recipient
will ordinarily be invited to present his or her work in a special academy lecture.
The following subject matter will be considered:

The interaction between behavior and biological mechanisms in homeostasis,

the maintenance of health, the pathophysiology of disease, or the susceptibility to
illness.
Development or evaluation of behaviorally based therapeutic interventions for
the primary prevention, secondary prevention, or treatment of disease or injury,
Basic theoretical or empirical studies in any scientific discipline or mathemat-
ics with implications for behavioral medicine research.

The submitted work must be in the form of a published article, chapter, book, or
a manuscript that has been accepted for publication.

All materials, including a current curriculum vitae, must be sent to the sec-
retary of the academy by February 1,2004.

Thomas W. Kamarck, PhD

Professor of Psychology
University of Pittsburgh
2 10 South Bouquet Street
4403 Sennott Square
Pittsburgh, PA 15260
Phone: 4 12-624-2080. FAX: 41 2-624-2205

Vol 28, Winter 2003 149