Yoga for Women with Metastatic Breast Cancer: Results from a Pilot Study
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Vol. 33 No. 3 March 2007 Journal of Pain and Symptom Management 331
Clinical Note
Yoga for Women with Metastatic Breast
Cancer: Results from a Pilot Study
James W. Carson, PhD, Kimberly M. Carson, MPH, Laura S. Porter, PhD,
Francis J. Keefe, PhD, Heather Shaw, MD, and Julie M. Miller, MA
Departments of Psychiatry and Behavioral Sciences (J.W.C., K.M.C., L.S.P., F.J.K., J.M.M.)
and Medicine (H.S.), Duke University Medical Center, Durham, North Carolina, USA
Abstract
Metastatic breast cancer (MBC) remains a terminal illness for which major treatment
advances are slow to appear, and hence it is crucial that effective palliative interventions be
developed to reduce the cancer-related symptoms of women with this condition during the
remaining years of their lives. This pilot/feasibility study examined a novel, yoga-based
palliative intervention, the Yoga of Awareness Program, in a sample of women with MBC.
The eight-week protocol included gentle yoga postures, breathing exercises, meditation,
didactic presentations, and group interchange. Outcome was assessed using daily measures
of pain, fatigue, distress, invigoration, acceptance, and relaxation during two
preintervention weeks and the final two weeks of the intervention. Thirteen women completed
the intervention (mean age ¼ 59; mean time since diagnosis ¼ 7 years; two African
American, 11 Caucasian). During the study, four participants had cancer recurrences, and
the physical condition of several others deteriorated noticeably. Despite low statistical power,
pre-to-post multilevel outcomes analyses showed significant increases in invigoration and
acceptance. Lagged analyses of length of home yoga practice (controlling for individual mean
practice time and outcome levels on the lagged days) showed that on the day after a day
during which women practiced more, they experienced significantly lower levels of pain and
fatigue, and higher levels of invigoration, acceptance, and relaxation. These findings
support the need for further investigation of the effects of the Yoga of Awareness Program in
women with MBC. J Pain Symptom Manage 2007;33;331e341. Ó 2007 U.S. Cancer
Pain Relief Committee. Published by Elsevier Inc. All rights reserved.
Key Words
Metastatic breast cancer, pain, fatigue, yoga, meditation
This work was supported by funding from the
Introduction
North Carolina Academic Alliance for Integrative Metastatic breast cancer (MBC) constitutes
Medicine. a serious life threat for women, with an average
Address reprint requests to: James W. Carson, PhD, Pain prognosis of 18e24 months to live.1 Pain, fa-
Prevention & Treatment Research Program, Depart- tigue, and emotional distress are often re-
ment of Psychiatry, Duke University Medical Center, ported as the most debilitating symptoms of
Box 90399, Durham, NC 27708, USA. E-mail:
jim.carson@duke.edu MBC patients.2e5 Although pharmacological
interventions may help these women to some
Accepted for publication: August 11, 2006. degree, these symptoms rarely resolve.4,6,7
Ó 2007 U.S. Cancer Pain Relief Committee 0885-3924/07/$esee front matter
Published by Elsevier Inc. All rights reserved. doi:10.1016/j.jpainsymman.2006.08.009332 Carson et al. Vol. 33 No. 3 March 2007 Given that MBC is a terminal illness for which in MBC. Second, historically in the practice major treatment advances are slow to appear, it of yoga, a fundamental emphasis is placed on is crucial that effective adjunctive palliative in- accepting one’s moment-to-moment experi- terventions be developed to reduce the cancer- ences, whatever they may be, without forcing related symptoms of these women during the the body beyond its comfortable limits. This remaining years of their lives. is because struggles to control the body, or to Few palliative behavioral interventions have control one’s physical sensations, thoughts, been tested with MBC patients, with mixed re- or emotions often only exacerbate problems. sults. A recent review of cognitive-behavioral Such struggles also detract from focusing on group therapy8e11 and supportive-expressive the personally fulfilling activities at hand. group therapy12e17 trials concluded that al- The important role of a healthy sense of accep- though some evidence exists for short-term tance in decreasing distress in the face of benefits from these approaches (e.g., better unpleasant symptomology has received in- mood9,15 and less increase in pain14,16), such creased research and clinical attention,31 and changes are not maintained even for a few may be especially important for individuals months.18 More recently, results from an exer- dealing with a life-threatening illness.32 Third, cise intervention tailored for MBC demon- studies have demonstrated that yoga produces strated short-term effects in terms of slower the relaxation response characterized by de- rates of deterioration in well-being and creased sympathetic and increased parasympa- fatigue.19 thetic activity.33e38 The relaxation response is There is thus a clear need to test new behav- likely to improve symptoms of pain, fatigue, ioral approaches to ameliorating cancer-re- and distress.39,40 lated symptoms in women with MBC.18 Yoga In summary, yoga appears to promote at least is a popular approach to health maintenance, three therapeutic processesdinvigoration, ac- which holds promise for benefiting women ceptance, and relaxationdwhich are likely to with MBC. As a mind/body discipline originat- have a favorable effect on MBC patients. The ing in India, yoga has been practiced for its purposes of this pilot study were to determine proposed physical, mental, and spiritual bene- the feasibility of a novel, yoga-based eight-week fits for thousands of years.20 Yoga is one of group intervention in women with MBC and a variety of complementary and alternative to examine the intervention’s effects on pain, medicine adjunctive approaches that cancer fatigue, distress, invigoration, acceptance, and patients of all cultural backgrounds have been relaxation. We did not specify a priori hypothe- seeking out in dramatically larger numbers in ses regarding improvements in these outcomes recent years.21e24 Yoga is now offered at several because of (a) the preliminary, uncontrolled de- major treatment centers (e.g., M.D. Anderson, sign of the study and (b) the fact that significant Memorial Sloan-Kettering, UCLA/Jonsson) via findings from intervention studies with this their complementary therapy services, and re- population have often consisted of less deterio- cently yoga has been the focus of several ration in the intervention vs. control condi- cancer studies.25e29 Results from two random- tions, rather than actual improvement in ized trials of yoga for cancer patients have symptoms.12,14,16,17,19 The intervention we been published. Cohen et al. reported im- useddthe Yoga of Awareness Programdwas provement in sleep disturbance in a trial with specifically designed for MBC patients, and in- lymphoma patients,27 and Culos-Reed et al. cluded gentle yoga postures, regulated breath- demonstrated improvements in mood, quality ing, guided meditations, brief didactic of life, and stress in breast cancer survivors.28 presentations, and group discussions. However, thus far no study has investigated Compared to previous MBC intervention the use of yoga with MBC patients. The ratio- studies, this study makes a unique contribution nale for applying a yoga-based intervention in terms of the method of analyzing treatment in this population rests on several premises. outcome. Data on symptoms (pain, fatigue, First, research suggests that yoga can produce distress) and therapeutic processes (invigora- an ‘‘invigorating effect on mental and physical tion, acceptance, relaxation) were collected energy’’ that improves physical fitness and in the form of prospective daily diaries. Rela- counteracts fatigue,30 a problematic symptom tive to traditional survey questionnaires,
Vol. 33 No. 3 March 2007 Yoga for Women with Metastatic Breast Cancer 333
diaries offer improved accuracy, reduced recall et al.9 ¼ 26%, Goodwin et al.14 ¼ 32%, Spiegel
bias, and increased recall of symptoms and re- and Bloom ¼ 31%16).
lated events, especially regarding the temporal This left 13 women who completed the inter-
sequencing of events.41e45 To analyze the diary vention and provided pre- and postmeasures.
data, we used multilevel random effects The mean age of this sample was 59 years
models.46 These analyses are preferable to (range ¼ 44e75). The average time since diag-
data analytic strategies that aggregate daily as- nosis was seven years (range ¼ 0e26). Seven
sessments (e.g., ordinary regression models) were concurrently receiving chemotherapy
because they permit 1) tracking of symptom-re- treatments, and six were not; four had new can-
lated processes as they occur in naturalistic set- cer reoccurrences while participating in the
tings, 2) accounting for two levels of sampling, study. Six patients previously practiced yoga or
that is, both within-person variation and be- meditation to some extent (of which two had
tween-person variation, 3) controlling for auto- practiced during the previous year). Eleven pa-
correlation, that is, the serial dependency that tients were Caucasian and two were African
results from successive daily assessments, and American. Eleven were married, one was wid-
4) handling missing data that often occur in owed, and one was divorced. Fifteen percent
intensive longitudinal data collection.46e49 had graduated from high school, 54% had
attended college, and 31% had attended
graduate school.
Methods
Participants and Setting Procedure
Volunteers for this study included 21 adult The protocol for this study was approved by
women with MBC referred by oncologists at the Duke Institutional Review Board. Prior to
the Duke University Medical Center breast on- the study, informed consent was obtained
cology unit and affiliate sites. Sample illustra- from all participants. All women were asked to
tions of the postures used during the yoga provide basic demographic information and in-
sessions were provided to referring physicians formation relevant to their breast cancer treat-
and to potential participants to give them a bet- ment history. Participants continued to receive
ter sense of what the intervention would involve. the standard care provided by their health
Patients were excluded if they had less than six- care providers. To control for a potentially im-
month life expectancy, if changes had taken portant medication-related confound, after
place in their use of any antidepressants during completing the intervention, patients were
the previous three months, if they had received asked about any changes in antidepressant
treatment for serious psychiatric disorders (e.g., use; no changes were reported. After complet-
schizophrenia) in the previous six months, if ing the yoga program and postintervention
they were currently engaged in intensive yoga measures, women were invited to participate
practice (>three days per week), if driving in a focus group to give qualitative feedback
time to attend weekly sessions was excessive regarding the program.
(>one hour as a rule of thumb), or if they
were not English speaking. Of the women who Data Collection
volunteered for the study, three withdrew before
beginning the intervention (2 ¼ scheduling dif- Daily Symptom and Process Measures. A brief
ficulties, 1 ¼ custody-case court appearances). daily-diary measurement strategy was chosen
Of the 18 remaining women, four withdrew because of the demonstrated reliability and
shortly after beginning the intervention validity of this method in health-related fields,
(1 ¼ scheduling difficulties, 1 ¼ traveling dis- and the fact that diaries allow subjective events,
tance, 1 ¼ relocated to another state, 1 ¼ health such as sensations and feelings, to be reported
status deterioration), and one woman did not unobtrusively and with little introspection in
complete post diaries because of mental status individuals’ natural settings.45 Data from dia-
deterioration (dementia-like). Attrition subse- ries also permit a careful analysis of day-to-
quent to beginning the intervention was thus day interrelated happenings for each person
28%, which is similar to other intervention stud- and provide increased statistical power when
ies with MBC patients (e.g., Edelman analyzing a small clinical sample.48334 Carson et al. Vol. 33 No. 3 March 2007
Before going to bed each evening, patients qualitative feedback regarding the Yoga of
completed a daily-diary log, in which they re- Awareness Program. Ten of the 13 participants
corded their levels of pain, fatigue, distress, in- attended. Open-ended questions were used to
vigoration, acceptance, and relaxation during explore women’s experiences regarding the
two preintervention weeks (pre), and during various yoga methods used, the length and
the last two weeks of the intervention period number of sessions, and the overall program.
(post). All daily variables were indicated by Women were invited to offer suggestions for
marking 100-mm visual analogue scales how the program could be better tailored to
(VAS), in which higher scores reflected greater their needs.
amounts (e.g., for pain, the item read ‘‘Please
indicate the average level of pain you had to- Focus Group Anonymous Questionnaire. Focus
day,’’ with anchors set as ‘‘No Pain’’ and group participants were also asked to rate the
‘‘Pain as Bad as it Can Be’’; for acceptance, success of the program on a five-item anony-
the item read ‘‘Please indicate how easy it was mous questionnaire. The questionnaire asked
for you today to accept and let be the ways patients to rate, on 10-point Likert-type scales,
you have been affected by your medical condi- how helpful the program was overall, how suc-
tion,’’ with anchors set as ‘‘No Acceptance’’ cessful the program was in helping cope with
and ‘‘Complete Acceptance’’). Similar VAS feelings of distress, how successful the program
measures are extensively used in clinical set- was in helping manage pain, how successful
tings to measure subjective phenomena, and the program was in helping manage fatigue,
have been shown to be valid, reliable, rapid, and whether they would recommend the pro-
and sensitive in measuring such variables as gram to a friend who has similar concerns.
global affect, pain, and fatigue.50 The post di-
ary also asked participants to indicate how Yoga of Awareness Program
many minutes were spent in completing the The intervention consisted of eight weekly
day’s yoga practice assignment. group sessions (four to five patients per group)
An important requirement for the effective conducted at the Duke Pain Prevention and
use of diary measures is that participants receive Treatment Research Program offices. The
adequate training in the completion of dia- groups were jointly led by a certified yoga
ries.45 Participants in this study were trained in teacher (registered with the national Yoga Alli-
how to complete the diary by a research assistant ance) who holds a master’s degree in health be-
who helped them complete sample diaries. Pa- havior and education (KMC) and a clinical
tients also were called during the first week of health psychologist (JWC). Both intervention
each recording period to inquire about any dif- leaders had received comprehensive training
ficulties and to answer any questions. Another in traditional schools of yoga and had extensive
key requirement of most daily-diary studies is experience in teaching yoga and meditation
that each day’s responses be recorded by the techniques to medical patients and the general
end of that day.51 To bolster this expectation, public. On average, participants attended seven
we asked participants to mail each day’s diary of the eight sessions (range, five to eight). To
back to us the following morning in pread- ensure consistency in delivering the interven-
dressed stamped envelopes. To further facilitate tion, a manual was developed to delineate the
motivation and compliance, patients were paid program and provide detailed session guide-
$0.25 for each completed daily diary and a $1.50 lines to be followed by the intervention leaders.
bonus for each week of complete recording. All sessions were videotaped and reviewed in
The diary completion rate was 89% (324 of weekly treatment team meetings.
364 potentially reportable days across 13 partic- Yoga of Awareness is an innovative behavioral
ipants; range, 71e100%). On average, patients intervention specifically designed and tailored
completed 12 of the 14 diaries at both pre and to address patients’ pain, fatigue, and emo-
post (SD ¼ 2). tional distress. The intervention is based in the
ancient Indian discipline of yoga (meaning
Focus Group Feedback. After all participants ‘‘yoking’’ or ‘‘union’’). During recent years, as
had completed the intervention, the women the physical exercises of yoga have become pop-
were invited to a focus group meeting to give ularized in Western countries, the term yoga inVol. 33 No. 3 March 2007 Yoga for Women with Metastatic Breast Cancer 335
common usage has largely become synonymous based on data from study completers
with this single aspect of the fuller discipline. only.41,53 A series of regression and Chi-square
Yoga in actuality comprises a wide variety of analyses comparing study completers and
methods and approaches.27,52 The Yoga of those who withdrew (either prior to or soon af-
Awareness intervention is a comprehensive ter beginning the intervention) indicated no
yoga program that systematically integrates significant differences in demographic charac-
a broad spectrum of traditional yogic tech- teristics. However, a significant effect was
niques and tenets. Each 120-minute session in- found for mean baseline scores on fatigue
cluded gentle physical stretching postures (F[1, 19] ¼ 8.26, P < 0.01). Those who with-
(asanas, e.g., seated forward folds, supine lateral drew were likely to have lower scores on fatigue
twists) complemented by breathing exercises (M ¼ 46.92 for completers vs. 26.15 for non-
( pranayama, e.g., extended exhalation, breath- completers). There were also findings ap-
ing into sensation), meditation techniques proaching significance for noncompleters to
(dhyana, e.g., awareness of breath, awareness have higher baseline scores on relaxation
of awareness itself), study of pertinent topics (M ¼ 45.21 for completers vs. 63.16 for non-
(swadhyaya, e.g., themes such as the value of completers, F[1, 19] ¼ 3.89, P ¼ 0.06) and in-
watching oneself in one’s daily life with the in- vigoration (M ¼ 39.99 for completers vs.
tention to understand rather than to judge 54.58 for noncompleters, F[1, 19] ¼ 3.19,
and of maintaining one’s poise even amid the P ¼ 0.09) and lower scores on pain
tumult of ever-changing challenges), and group (M ¼ 33.92 for completers vs. 21.53 for non-
discussions (satsang, e.g., discussion of experi- completers, F[1, 19] ¼ 2.65, P ¼ 0.12). Note
ences of practicing yoga at home, changes in that statistical power for these tests was low
cancer-related symptoms during the week). Pa- due to the small sample size.
tients were supplied with a yoga mat, a blanket,
CDs/audiotapes, and illustrated handbooks to Approach to Multilevel Data Analyses
guide them in home practice. Participants The analyses for this study are based on a set
were encouraged to spend at least 10 minutes of recently developed statistical procedures
a day practicing yoga strategies on their own, called multilevel modeling.54 Multilevel mod-
and applications of yoga to daily living were as- eling is an advanced methodology for integrat-
signed each week (e.g., acceptance during inter- ing data from multiple levels of sampling, such
vals of pain). as this study’s two levels (within-person and be-
During sessions, the intervention leaders tween-persons). Multilevel models are particu-
emphasized the importance of gentle posture larly advantageous in analyzing data sets with
practice when one’s body is challenged by many repeated measures, such as daily-diary re-
chronic illness, and instructions were modified cords.48,53 By preserving the rich detail in each
to appropriately address individual patient individual’s full data set, multilevel models al-
needs (e.g., backache). Although the yoga pos- low for a sensitive independent determination
tures used in this study presented no more risk of day-to-day interrelated happenings for each
of permanent injury than is associated with ev- patient, as well as aggregation of individual es-
eryday activities such as climbing stairs or timates for reliable results for the average pa-
kneeling down to pick something up, nonethe- tient. The SAS Proc Mixed procedure55
less, as a precaution, a physician assistant or produced parameters in the form
nurse was present during all sessions to ad- of unstandardized maximum likelihood
dress any medical concerns that may have estimates (b coefficients). These are partial
arisen. However, their medical services were correlations, adjusted for between-person dif-
never needed. [Please contact the first author ferences, which serve as effect size estimates
for a full description of the yoga postures of magnitude and direction of changes in de-
used in the study.] pendent variables associated with changes in
independent variables.56 Multilevel models al-
low for strict control for potential confounds,
Results such as serial autocorrelation in measurements.
Because of the preliminary nature of this For all analyses, we allowed intercepts to vary
study, analyses of outcome measures were randomly, thus allowing us to generalize the336 Carson et al. Vol. 33 No. 3 March 2007
findings to the population of persons from were predictive of same-day levels of diary vari-
which the sample was taken and the population ables. These models controlled for individuals’
of observations from which their daily reports mean levels of yoga practice, and practice rates
were drawn.46 [For a more complete descrip- were person-centered to control for potentially
tion of the multilevel equations reported spurious within-person associations.56
herein, please contact the first author.] Table 2 presents the results of these tests.
Same-day tests indicated greater yoga practice
Treatment Effects on Daily Outcomes was significantly associated with decreased
To examine treatment effects, models tested pain, increased invigoration, and increased ac-
whether patients’ intercept levels for daily ceptance. Trends were also present for greater
pain, fatigue, distress, invigoration, accep- yoga practice to be associated with decreased
tance, and relaxation changed significantly fatigue (P ¼ 0.07) and increased relaxation
across time from the pre to the post recording (P ¼ 0.07).
periods.46,53 Table 1 shows the outcomes for
the effect of treatment (Time). Despite the Length of Yoga Practice and Lagged
small sample size, the results demonstrated sig- Day Outcomes
nificant improvements in daily invigoration The preceding same-day analyses did not ad-
and acceptance, along with trends for improve- dress the question of temporal precedence
ment in pain and relaxation. and hence cannot be used to make causal in-
ferences. Yoga practice potentially could have
influenced levels of pain, invigoration, and ac-
Length of Yoga Practice
ceptance, or the reverse could be true. There-
and Same Day Outcomes
fore, to clarify whether increases in yoga
On 71% (SD ¼ 25%) of post diaries (col-
practice preceded and may have had a causa-
lected during the last 14 days of the interven-
tive influence on day-to-day fluctuations in out-
tion), patients reported spending some time
come variables, tests were conducted for lags
formally practicing yoga techniques. On aver-
of one and two days’ practice. In lagged
age, they reported practicing for 21 minutes
models, along with controlling for individuals’
per day (SD ¼ 11). Analyses examined whether
mean levels of practice, the lagged day’s level
the number of minutes spent in yoga practice
of the dependent variable was also included
Table 1 as a within-person control variable (e.g., the
Multilevel Random Effects Estimates present day’s pain when predicting next day’s
for Baseline Intercept and Effect pain). As before, practice rates were person-
of Treatment (Time) centered to control for any spurious within-
Predictor b t P person associations.
Daily pain The results for next-day lagged associations
Baseline intercept 34.62 8.38Vol. 33 No. 3 March 2007 Yoga for Women with Metastatic Breast Cancer 337
Table 2
Multilevel Random Effects Estimates for Associations of Length of Yoga Practice with Day-to-Day Outcomes
Same Day Outcomes Next Day Outcomes
Variable b t P b t P
a
Daily pain 0.15 2.71338 Carson et al. Vol. 33 No. 3 March 2007 potential benefits in this vulnerable popula- of completers and noncompleters, a trial of tion of women with limited life expectancy. cognitive-behavioral group therapy, Edelman Because of the absence of a control group, it et al. found the opposite: the noncompleters is difficult to directly compare the magnitude tended to be patients with greater symptom of our findings with those of other palliative in- levels (e.g., higher anxiety, lower vigor).9 tervention studies in MBC patients. As noted The Yoga of Awareness intervention we above, however, significant findings with this used, which comprehensively integrates population have often consisted of less in- a wide spectrum of ancient yoga techni- crease in symptoms in the intervention vs. quesdpostures, breathing exercises, medita- control conditions rather than actual improve- tion, study of guiding tenets, and group ments.12,14,16,17,19 So, while the absence of discussionsdstands in strong contrast to previ- a control group does not allow us to test for ous palliative interventions with MBC patients, differential decline in symptoms, the fact that such as cognitive-behavioral group therapy the treatment-related coefficients for daily and supportive-expressive group therapy.8e17 pain, fatigue, invigoration, acceptance, and re- This intervention also contrasts with some laxation (with the sole exception of daily yoga trainings that have been used with cancer distress) were in the direction of true improve- patients, which emphasize only one compo- ments has important implications for the nent of the yoga system, that is, the posture ex- potential efficacy of the intervention. ercises that have become widely popularized in Beyond examining the effects of the yoga- Western countries (e.g., Culos-Reed et al.28). based intervention, this study marks the first Notably, however, only a minority of the partic- time multilevel modeling has been applied to ipants in our study had ever practiced yoga or daily-diary outcomes in MBC patients. Impor- meditation before, and among this minority, tantly, the present multilevel treatment results only a few had more than minimal exposure were obtained by first calculating independent to these methods. Despite this unfamiliarity, estimates for each participant and then aggre- our findings clearly demonstrated the feasibil- gating them to derive reliable results for the av- ity of conducting future studies of the Yoga of erage patientdthus avoiding the problem of Awareness Program with MBC patients. It is overlooking the impact of individual differ- important to note that four of the 13 women ences, as in standard regression approaches. in our sample experienced new cancer recur- Moreover, the advantages offered by this statis- rences during the study, and the physical tical approach were particularly well suited for condition of several others was noticeably dete- the analysis of real-time processes that have riorating. Nonetheless, no adverse events were strong causal implications. The tangible im- related to yoga participation. It is likely that pact of yoga practice was highlighted by the the gentle approach to yoga postures we finding that greater practice on a given day used, including careful monitoring of teaching was associated with improvements not only methods in weekly treatment team meetings, on the same day, but the next day as well. Fu- was vital to making the yoga program safe for ture studies could profit from using daily these women. In addition, attendance at yoga data collection to examine more refined hy- sessions during the study was good (an average potheses about changes in patients’ symptoms of seven of eight sessions), as was adherence to (e.g., would yoga lead to same-day or next-day yoga practice (an average of 21 minutes per fatigue becoming more resilient to the nega- day). Further confirmation of the feasibility tive impact of a chemotherapy session60). of offering this program to MBC patients Attrition in this study was similar to previous came from the focus group meeting. Along palliative trials for MBC.9,14,16,19 Analyses with giving the program high ratings on the showed that completers tended to be patients anonymous questionnaire, participants’ com- who were experiencing higher levels of fatigue ments confirmed the suitability of the yoga in- and other outcomesdpresumably because tervention’s content and the number and these women were more highly motivated by length of yoga sessions. The only drawback pa- the possibility of ameliorating their symptoms. tients noted was that the program did not lead In contrast, in the only previous psychosocial to an ongoing opportunity to practice yoga in intervention study to report on a comparison the company of other women with MBC.
Vol. 33 No. 3 March 2007 Yoga for Women with Metastatic Breast Cancer 339
Future studies should consider ways to create conference statement: symptom management in
a structure for continuing specialized yoga cancer: pain, depression, and fatigue. J Natl Cancer
classes past the study termination. Finally, pa- Inst 2003;95:1110e1117.
tients’ anecdotal reports revealed that the 5. Rowland JH, Massie MJ, Holland JC. Breast can-
yoga training appeared to be very useful in cer. In: Holland JC, Breitbart W, Jacobsen PB,
Lederberg MS, eds. Psycho-oncology. New York:
helping them better adjust to the challenges
Oxford University Press, 1998: 380e401.
of living with metastatic disease, including
‘‘keeping their balance even when the waves 6. Foley KM, Arbit E, DeVita VT, Hellman S,
Rosenberg SA. Management of cancer pain. In:
get rough.’’ DeVita VT, Rosenberg SA, Hellman S, eds. Princi-
Several important limitations of our study ples and practice of oncology, 3rd ed. Philadelphia:
should be noted. The generalizability of these J.B. Lippincott, 1989: 2064e2087.
preliminary findings is restricted by the very 7. Osoba D, Slamon DJ, Burchmore M, et al. Ef-
small sample, the absence of a control group, fects on quality of life of combined trastuzumab
and the lack of follow-up data. To clearly estab- and chemotherapy in women with metastatic breast
lish the efficacy of the Yoga of Awareness Pro- cancer. J Clin Oncol 2002;20:3106e3113.
gram, a controlled (e.g., support group 8. Cunningham AJ, Edmonds CV, Jenkins GP, et al.
control), well-powered trial including follow- A randomized controlled trial of the effects of
up evaluations is needed. Further methodo- group psychological therapy on survival in women
with metastatic breast cancer. Psychooncology
logical improvements for such a study could 1998;7:508e517.
include supplementation of self-report data
9. Edelman S, Bell DR, Kidman AD. A group cog-
by other types of measures (e.g., immune re- nitive behaviour therapy programme with metastatic
sponse,61,62 physical fitness28), formal analyses breast cancer patients. Psychooncology 1999;8:
of treatment integrity,63 and analyses of predic- 295e305.
tors of treatment outcome (e.g., is yoga more 10. Edelman S, Lemon J, Bell DR, Kidman AD. Ef-
helpful for patients with certain characteris- fects of group CBT on the survival time of patients
tics). Future studies can also seek to determine with metastatic breast cancer. Psychooncology
minimum amounts of effective yoga practice, 1999;8:474e481.
and given the dose/response effect observed, 11. Edmonds CV, Lockwood GA, Cunningham AJ.
how adherence can be bolstered in those Psychological response to long-term group therapy:
who practice less. a randomized trial with metastatic breast cancer
patients. Psychooncology 1999;8:74e91.
In conclusion, the findings of this pilot study
provide promising preliminary support for the 12. Bordeleau L, Szalai JP, Ennis M, et al. Quality of
life in a randomized trial of group psychosocial
palliative health benefits of yoga in MBC pa-
support in metastatic breast cancer: overall effects
tients. The improvements that have been of the intervention and an exploration of missing
documented, and the potential impact these data. J Clin Oncol 2003;21:1944e1951.
improvements may have on patients’ adjust- 13. Classen C, Butler LD, Koopman C, et al. Sup-
ment during the remaining years of their lives, portive-expressive group therapy and distress in pa-
are important enough to warrant further study. tients with metastatic breast cancer: a randomized
clinical intervention trial. Arch Gen Psychiatry
2001;58:494e501.
References 14. Goodwin PJ, Leszcz M, Ennis M, et al. The effect
1. Leonard RC, Rodger A, Dixon JM, Dixon JM. of group psychosocial support on survival in meta-
Metastatic breast cancer. In: Dixon M, ed. ABC of static breast cancer. N Engl J Med 2001;345:
breast diseases, 2nd ed. London: BMJ Books, 2000: 1719e1726.
65e71. 15. Spiegel D, Bloom JR, Yalom I. Group support
2. Aranda S, Schofield P, Weih L, et al. Mapping for patients with metastatic cancer. Arch Gen Psychi-
the quality of life and unmet needs of urban women atry 1981;38:527e533.
with metastatic breast cancer. Eur J Cancer Care 16. Spiegel D, Bloom JR. Group therapy and hypno-
2005;14:211e222. sis reduce metastatic breast carcinoma pain. Psycho-
3. Cella D. Factors influencing quality of life in som Med 1983;45:333e339.
cancer patients: anemia and fatigue. Semin Oncol 17. Spiegel D, Bloom JR, Kraemer HC, Gottheil E.
1998;25:43e46. Effect of psychosocial treatment on survival of pa-
4. Patrick DL, Ferketich SL, Frame PS, et al. tients with metastatic breast cancer. Lancet 1989;2:
National Institutes of Health State-of-the-science 888e891.340 Carson et al. Vol. 33 No. 3 March 2007
18. Edwards AG, Hailey S, Maxwell M. Psychological with alpha wave activation. Percept Mot Skills 2000;
interventions for women with metastatic breast can- 90:1027e1032.
cer. Cochrane Database Syst Rev 2004;CD004253.
34. Patel C, North WR. Randomised controlled trial
19. Headley JA, Ownby KK, John LD. The effect of of yoga and bio-feedback in management of hyper-
seated exercise on fatigue and quality of life in tension. Lancet 1975;2:93e95.
women with advanced breast cancer. Oncol Nurs
Forum 2004;31:977e983. 35. Raghuraj R, Telles S. Effect of a yoga-based and
forced uninostril breathing on the autonomic ner-
20. Telles S, Naveen KV. Yoga for rehabilitation: an vous system. Percept Mot Skills 2003;96:79e80.
overview. Indian J Med Sci 1997;51:123e127.
36. Schell FJ, Allolio B, Schonecke OW. Physiologi-
21. Alferi SM, Antoni MH, Ironson G, Kilbourn cal and psychological effects of Hatha-Yoga exercise
KM, Carver CS. Factors predicting the use of com- in healthy women. Int J Psychosom 1994;41:46e52.
plementary therapies in a multi-ethnic sample of
early-stage breast cancer patients. J Am Med 37. Telles S, Reddy SK, Nagendra HR. Oxygen con-
Womens Assoc. 2001;56:120e123. sumption and respiration following two yoga relaxa-
tion techniques. Appl Psychophysiol Biofeedback
22. Eisenberg DM, Davis RB, Ettner SL, et al. 2000;25:221e227.
Trends in alternative medicine use in the United
States, 1990e1997: results of a follow-up national 38. Telles S, Joshi M, Dash M, et al. An evaluation of
survey. JAMA 1998;280:1569e1575. the ability to voluntarily reduce the heart rate after
a month of yoga practice. Integr Physiol Behav Sci
23. Lee MM, Lin SS, Wrensch MR, Adler SR,
2004;39:119e125.
Eisenberg D. Alternative therapies used by women
with breast cancer in four ethnic populations. 39. Lehrer PM, Woolfolk RL. Principles and prac-
J Natl Cancer Inst 2000;92:42e47. tice of stress management, 2nd ed. New York: Guil-
24. Richardson MA, Sanders T, Palmer JL, ford Press, 1993.
Greisinger A, Singletary SE. Complementary/alter- 40. Nespor K. Pain management and yoga. Int J
native medicine use in a comprehensive cancer cen- Psychosom 1991;38:76e81.
ter and the implications for oncology. [see
comment]. J Clin Oncol 2000;18:2505e2514. 41. Carson JW, Keefe FJ, Lynch TR, et al. Loving--
kindness meditation for chronic low back pain: re-
25. Bower JE, Woolery A, Sternlieb B, Garet D. Yoga sults from a pilot trial. J Holist Nurs 2005;23:1e18.
for cancer patients and survivors. Cancer Control
2005;12:165e171. 42. Carson JW, Keefe FJ, Affleck G, et al. A compar-
ison of conventional pain coping skills training and
26. Brawley LR, Culos-Reed SN, Angove J, pain coping skills training with a maintenance train-
Hoffman-Goetz L. Understanding the barriers to ing component: a daily diary analysis of short- and
physical activity for cancer patients: review and rec- long-term treatment effects. J Pain 2006;7:615e625.
ommendations. J Psychosoc Oncol 2002;20:1e21.
43. Stone AA, Kessler RC, Haythornthwaite JA. Mea-
27. Cohen L, Warneke C, Fouladi RT, suring daily events and experiences: decisions for
Rodriguez MA, Chaoul-Reich A. Psychological ad- the researcher. J Pers 1991;59:575e607.
justment and sleep quality in a randomized con-
trolled trial of the effects of a Tibetan yoga 44. Verbrugge LM. Health diaries. Med Care 1980;
intervention in patients with lymphoma. Cancer 18:73e95.
2004;100:2253e2260.
45. Bolger N, Davis A, Rafaeli E. Diary methods:
28. Culos-Reed SN, Carlson LE, Daroux LM, capturing life as it is lived. Annu Rev Psychol 2003;
Hately-Aldous S. A pilot study of yoga for breast can- 54:579e616.
cer survivors: physical and psychological benefits.
Psycho-Oncol 2006;15:891e897. 46. Affleck G, Zautra A, Tennen H, Armeli S. Multi-
level daily process designs for consulting and clini-
29. Joseph CD. Psychological supportive therapy for cal psychology: a preface for the perplexed.
cancer patients. Indian J Cancer 1983;20:268e270. J Consult Clin Psychol 1999;67:746e754.
30. Wood C. Mood change and perceptions of vital- 47. Jaccard J, Wan CK. Statistical analysis of tempo-
ity: a comparison of the effects of relaxation, visual- ral data with many observations: issues for behav-
ization and yoga. J R Soc Med 1993;86:254e258. ioral medicine data. Ann Behav Med 1993;15:
31. McCracken LM, Carson JW, Eccleston C, 41e50.
Keefe FJ. Acceptance and change in the context of 48. Schwartz JE, Stone AA. Strategies for analyzing
chronic pain. Pain 2004;109:4e7. ecological momentary assessment data. Health Psy-
32. Clark D. Between hope and acceptance: the chol 1998;17:6e16.
medicalisation of dying. BMJ 2002;324:905e907.
49. West SG, Hepworth JT. Statistical issues in the
33. Kamei T, Toriumi Y, Kimura H, et al. Decrease study of temporal data: daily experiences. J Pers
in serum cortisol during yoga exercise is correlated 1991;59:609e662.Vol. 33 No. 3 March 2007 Yoga for Women with Metastatic Breast Cancer 341
50. Cella DF, Perry SW. Reliability and concurrent 58. Oken BS, Kishiyama S, Zajdel D, et al. Random-
validity of three visual-analogue mood scales. Psy- ized controlled trial of yoga and exercise in multiple
chol Rep 1986;59:827e833. sclerosis. Neurology 2004;62:2058e2064.
51. Affleck G, Urrows S, Tennen H, Higgins P. Daily 59. Williams KA, Petronis J, Smith D, et al. Effect of
coping with pain from rheumatoid arthritis: pat- Iyengar yoga therapy for chronic low back pain.
terns and correlates. Pain 1992;51:221e229. Pain 2005;115:107e117.
52. Feuerstein G. The yoga tradition: Its history, lit- 60. Irvine D, Vincent L, Graydon JE, Bubela N,
erature, philosophy and practice. Boston: Sham- Thompson L. The prevalence and correlates of fa-
bala, 2001. tigue in patients receiving treatment with chemo-
53. Carson JW, Carson KM, Gil KM, Baucom DH. therapy and radiotherapy. Cancer Nurs 1994;17:
Mindfulness-based relationship enhancement. Be- 367e378.
hav Ther 2004;35:471e494.
61. Giese-Davis J, Sephton SE, Abercrombie HC,
54. Kreft I, de Leeuw J. Introducing multilevel mod- Duran RE, Spiegel D. Repression and high anxiety
eling. London: Sage, 1998. are associated with aberrant diurnal cortisol
55. SAS Institute. SAS/STAT software: Changes and rhythms in women with metastatic breast cancer.
enhancements through release 6.12. Cary, NC: SAS Health Psychol 2004;23:645e650.
Institute, 1996. 62. van der Pompe G, Antoni M, Visser A,
56. Bryk AS, Raudenbush SW. Application of hierar- Garssen B. Adjustment to breast cancer: the psycho-
chical linear models to assessing change. Psychol biological effects of psychosocial interventions. Pa-
Bull 1987;101:147e158. tient Educ Couns 1996;28:209e219.
57. Garfinkel MS, Schumacher HR Jr, Husain A, 63. Waltz J, Addis ME, Koerner K, Jacobson NS.
Levy M, Reshetar RA. Evaluation of a yoga based Testing the integrity of a psychotherapy protocol: as-
regimen for treatment of osteoarthritis of the sessment of adherence and competence. J Consult
hands. J Rheumatol 1994;21:2341e2343. Clin Psychol 1993;61:620e630.You can also read
