Associations Between Physical Activity and Quality of Life in Cancer Patients Receiving Palliative Care: A Pilot Survey

Transcription

1 Vol. 38 No. 5 November 2009 Journal of Pain and Symptom Management 785 Clinical Note Associations Between Physical Activity and Quality of Life in Cancer Patients Receiving Palliative Care: A Pilot Survey Sonya S. Lowe, MD, MSc, Sharon M. Watanabe, MD, Vickie E. Baracos, PhD, and Kerry S. Courneya, PhD Department of Symptom Control and Palliative Care (S.S.L.), Cross Cancer Institute; and Division of Palliative Care Medicine (S.M.W., V.E.B.), Department of Oncology (K.S.C.), University of Alberta, Edmonton, Alberta, Canada Abstract The primary aim of this study was to examine the association between physical activity and quality of life (QoL) in cancer patients receiving palliative care. Fifty advanced cancer patients aged 18 years or older with clinician-estimated life expectancy of 3e12 months and Palliative Performance Status Scale scores greater than 30% were recruited from an outpatient palliative care clinic and palliative home care. Participants completed a cross-sectional survey by means of face-to-face interview assessing self-reported QoL (McGill Quality of Life Questionnaire [MQOL]), self-reported physical function (Late-Life Function and Disability Instrument), symptoms (Edmonton Symptom Assessment System), and physical activity behavior. Seventy-six percent (38 of 50) of the participants were deceased at the time of data analysis, with a median survival of 104 days from time of survey to time of death. Walking was the most common reported physical activity. Analyses of variance indicated that participants who reported walking more than 30 minutes per day also reported higher existential subscores (0.8 [95% CI, 0.0e1.5]; P ¼ 0.045), support subscores (0.7 [95% CI, 0.1e1.4]; P ¼ 0.027), and total scores (0.5 [95% CI, 0.0e0.9]; P ¼ 0.046) on the MQOL. There were no significant differences for self-reported physical function or symptoms. Our findings show a significant positive association between physical activity and QoL scores in this sample of patients with advanced cancer. A pilot intervention trial testing the causal effects of physical activity on QoL in cancer patients receiving palliative care is warranted. J Pain Symptom Manage 2009;38:785e796. Ó 2009 U.S. Cancer Pain Relief Committee. Published by Elsevier Inc. All rights reserved. Key Words Physical activity, quality of life, palliative care, cancer This study was funded by Operating Grant MOP from the Canadian Institutes of Health Research. Sonya Lowe is supported by a full-time Clinical Fellowship Award from the Alberta Heritage Foundation for Medical Research. Kerry Courneya is supported by the Canada Research Chairs Program as well as a Research Team Grant from the National Cancer Institute of Canada, with funds Ó 2009 U.S. Cancer Pain Relief Committee Published by Elsevier Inc. All rights reserved. from the Canadian Cancer Society and the Sociobehavioral Cancer Research Network. Address correspondence to: Sonya S. Lowe, MD, MSc, Department of Symptom Control and Palliative Care, Room 2001, Cross Cancer Institute, University Avenue, Edmonton, Alberta, T6G 1Z2, Canada. sonyalow@cancerboard.ab.ca Accepted for publication: April 1, /09/$esee front matter doi: /j.jpainsymman

2 786 Lowe et al. Vol. 38 No. 5 November 2009 Introduction Cancer-related fatigue and loss of physical function have been reported by cancer patients as being among their most distressing symptoms. 1 In combination with progressive muscle weakness and cachexia, the inability to perform activities of daily living independently adds to the distress level and disease burden for cancer patients. Associations between progressive debility and poorer social and psychological well-being have redirected emphasis toward examining quality of life (QoL) interventions that maintain a patient s mobility for as long as possible. 2 Recent meta-analyses have shown that physical activity can improve several aspects of physical and psychological well-being that contribute to QoL in early-stage cancer patients, 3 but few studies have focused on cancer patients at the palliative phase of the cancer spectrum. In our recent systematic review, there is preliminary evidence to show that selected cancer patients who are receiving palliative care are able to complete physical activity interventions, and that at least some of these patients report improvement in supportive care outcomes post intervention. 4 Of these preliminary studies, Oldervoll et al. 5 conducted an uncontrolled pilot trial of a six-week group exercise intervention in 34 advanced cancer patients with a clinicianestimated life expectancy of 3e12 months. Despite showing improvements in objective physical functioning, from the patientreported emotional functioning and physical fatigue, the authors concluded that future research should be directed at earlier physical exercise habits. to see if this has significance for whether palliative patients want to participate in an exercise intervention. 6 The primary objective of this study was to examine the associations between patient-reported physical activity, QoL, physical functioning, and symptoms in cancer patients with advanced illness. We hypothesized that these patients would be participating in minimal levels of physical activity, and that patients who were participating in physical activity would report better QoL, physical functioning, and less severe symptoms. Methods Setting and Participants The study was conducted at the Department of Symptom Control and Palliative Care, Cross Cancer Institute (CCI) and the Capital Health Palliative Home Care Program in Edmonton, Canada. Ethical approval for the study was received from the Health Research Ethics Board of the University of Alberta and the Research Ethics Committee of the Alberta Cancer Board. All participants were diagnosed with progressive, incurable, and locally recurrent or metastatic cancer. Eligibility criteria also included 1) 18 years of age or older; 2) ability to understand, provide written informed consent in, and speak English; 3) cognitive ability to participate (defined as a normal Folstein s Mini-Mental State score for patient s age and education level 7 ); and 4) clinician-estimated life expectancy of 3e12 months. Participants were ineligible if they presented with 1) any absolute contraindications to physical activity; 8 or 2) a Palliative Performance Status Scale score of 30% or less. 9 Eligible participants were required to read and sign a consent form, which detailed the right to withdraw, confidentiality, and the risks and benefits of participating in the study. Design and Recruitment The study was a cross-sectional survey using an interviewer-administered questionnaire. Potential participants were recruited from the Department of Symptom Control and Palliative Care, CCI (n ¼ 26) or from the Capital Health Palliative Home Care Program (n ¼ 24) from November 2006 to May In the Capital Health Palliative Home Care Program and the CCI outpatient radiotherapy units and lung clinics, ethical approval was obtained whereby every patient, irrespective of eligibility, was made aware of all research studies by means of a distributed copy of a research participation document. Patients thereafter had the option to indicate interest in a research study and to give their permission to be contacted by the study coordinator or to indicate that they were not interested in research. Patients were only assessed for study eligibility after they had granted permission to be contacted by the study coordinator. From the Multidisciplinary Pain and Symptom Outpatient

3 Vol. 38 No. 5 November 2009 Physical Activity, QoL, and Palliative Care 787 Clinic and the Department of Symptom Control and Palliative Care inpatient consultation service at CCI, potential participants were identified from consecutive referrals. The Capital Health Palliative Home Care Program, the Department of Symptom Control and Palliative Care at the CCI, and the outpatient radiotherapy units and lung clinics at the CCI constitute a heterogeneous population and cover a wide spectrum of palliative care services. Between April 2006 and March 2007, for example, the median survival in palliative home care was 64 days, and more than 50% of patients were bedridden or spent most of the day in their bed or chair (H. Quan, personal communication). On the other hand, most of the outpatients seen by the Department of Pain and Symptom Control, CCI, are more ambulatory and less cognitively impaired (B. Fisher, personal communication). We anticipated that many of these patients would be unable or unwilling to participate in physical activity because of the morbidity of their disease; therefore, our inclusion criteria were selected to purposefully target a subset of cancer patients with advanced disease who expressed an interest in and ability for physical activity. Survey Instruments QoL was assessed by the MQOL. 10 The 16-item MQOL covers five domains, including physical symptoms; physical, psychological, and existential well-being; support; and global QoL. The MQOL has been found to be comprehensive, widely tested, and valid across end-of-life populations. 11 Patient-reported symptoms were assessed by the Edmonton Symptom Assessment System (ESAS). 12 The ESAS covers nine items, including physical, psychological, and well-being subscales, and also has been widely tested and validated in palliative care populations. 13 Both MQOL and ESAS items were modified for this study to incorporate open-ended questions with respect to impact on physical function and activity. Physical activity was assessed by four questions modified from concepts and short items drawn from the Physical Activity Scale for the Elderly (PASE), which requires participants to recall their most common physical activities, including frequency, intensity, and duration, performed over the past week. 14 For the purposes of this study, physical activity was defined as any bodily movement produced by the skeletal muscles that results in a substantial increase in energy expenditure over resting levels; 15 this definition was explained to the participant before beginning the questionnaire. The PASE was developed for assessment of community-dwelling older adults and has been widely used and validated in various clinical populations, including end-stage renal patients; 16,17 given the symptom burden of palliative cancer patients, the PASE was selected for its sensitivity in assessing activity in frail populations. Physical functioning was assessed by an abbreviated version of the Late-Life Function and Disability Instrument (LLFDI). 18 The LLFDI comprises both a function component, which examines lower and upper extremity function, and a disability component, which examines the limitation in performing both instrumental and basic activities of daily living. For the purposes of this study, only the function component of the abbreviated LLFDI was used. The LLFDI has been widely used and validated in elderly populations. 19 Medical and demographic information was collected using self-report measures and by means of medical chart review. This information consisted of demographic variables, including age, sex, height, weight, marital status, education, income, employment status and ethnicity, and medical variables, including months since diagnosis, type and duration of adjuvant treatment, current medications, smoking and alcohol status, medical comorbidities, current Palliative Performance Status Scale score, and date of death. Statistical Analysis As with our previous studies examining associations between exercise and QoL in earlystage cancer patients, 20e22 the sample size calculation was based on Cohen s guideline for effect sizes of 0.20 ¼ small, 0.50 ¼ medium, and 0.80 ¼ large. When comparing two means with 25 participants per group, we can detect a standardized effect size of 0.80 in the various outcomes (QoL, physical activity) with a power of 80% and a two-tailed alpha value of less than 0.10; hence, our accrual goal was 50 participants in total.

4 788 Lowe et al. Vol. 38 No. 5 November 2009 Given that there are no established physical activity-level recommendations for advanced cancer patients, 23 participants were divided into two categories based on a roughly median split on their most common self-reported physical activity over the past week (<30 vs. $30 minutes of walking per day). Participants also were divided into two categories based on their self-reported total physical activity over the past week (<60 vs. $60 minutes of physical activity per day). Data were analyzed using SPSS version 15.0 software (SPSS Inc., Evanston, IL). Differences in patient-reported QoL, physical function, and symptoms between participants in the two physical activity categories were tested using independent t-tests. Effect sizes (d ) were computed by dividing the mean difference between groups by the pooled standard deviation. 24 To examine potential confounding variables, analyses were repeated using analysis of covariance (ANCOVA) to adjust for the following demographic and medical variables: age (<60 vs. $60 years), gender (male vs. female), marital status (single vs. partnered), number of metastatic sites (<2 metastatic sites vs. $2 metastatic sites), and current chemotherapy (yes vs. no). To test for potential moderators of the associations between physical activity and total MQOL scores, total LLFDI scores, ESAS pain scores, and ESAS fatigue scores, analyses of variance were performed with age (<60 vs. $60 years), sex (male vs. female), body mass index (BMI) (normal/underweight vs. overweight/ obese), Palliative Performance Status Scale score (<60% vs. $60%), total number of comorbidities (<2 vs. $2), cancer diagnosis (lung cancer vs. other), and survival from time of interview to time of death (<90 vs. $90 days) as potential moderators. With the same covariates as the main-effect analyses, the moderator analyses were repeated using ANCOVAs to control for potential confounding variables. Probabilities of less than 0.10 and Cohen s effect size d $ 0.35 (small-medium) were interpreted as potentially meaningful. Results Sample Characteristics Recruitment began in November 2006 at the CCI and in January 2007 at the Capital Health Palliative Home Care Program and ended May 2007 at both sites. Figure 1 shows the flow of participants through the study. Forty-seven of 244 (19%) screened home care patients consented to being contacted by the study coordinator, and 23 of 47 (49%) patients were eligible for and recruited to the study. One home care patient was recruited through the mail-out recruitment letter of invitation. Eight of 119 (7%) screened outpatient radiotherapy unit patients consented to be contacted by the study coordinator through the recruitment handout, with one of eight (13%) patients eligible for and recruited to the study. Two of 62 (3%) screened outpatient lung clinic patients consented to be contacted by the study coordinator through the recruitment handout, with one patient eligible for and recruited to the study. The estimated accrual rate from the Department of Symptom Control and Palliative Care was 24 of 92 potentially eligible participants (26%). Of all patients who contacted the study coordinator, the most common reason for declining to participate was severe fatigue. Demographic and Medical Characteristics As of July 2008, 38 of 50 (76%) study participants were deceased, with a median survival of 104 days from the date of conducting the survey to the date of death. Of those participants who were deceased, the maximum duration from time of survey to time of death was 356 days, whereas the minimum duration was 23 days. When combining the deceased participants into the total sample, 16 of 50 (32%) participants had a survival of less than 90 days, and 34 of 50 (68%) participants had a survival of 90 days or more at the time of survey. To examine the representativeness of the sample, participants recruited from the CCI (n ¼ 26) were compared with those recruited from the Capital Health Palliative Home Care Program (n ¼ 24) across demographic and medical variables. The only significant difference was observed for sex (P < 0.05), with participants recruited from the CCI being more likely to be male than those recruited from the Capital Health Palliative Home Care Program (55.6% vs. 21.7%). The demographic and medical characteristics of

5 Vol. 38 No. 5 November 2009 Physical Activity, QoL, and Palliative Care consecutive home care patients 119 consecutive outpatient radiotherapy unit patients 62 consecutive outpatient lung clinic patients 47 consented to being contacted by study coordinator 8 patients consented to being contacted by study coordinator 2 patients consented to being contacted by study coordinator Home care patients excluded: 8 declined due to severe fatigue 5 declined due to feeling overwhelmed 3 deceased at time of initial contact 3 language barrier 2 cognitive impairment 2 unable to contact 1 patient contacted study coordinator via mail-out recruitment letter Radiotherapy patients excluded: 3 declined due to severe fatigue 2 did not meet prognostic criteria 2 did not have advanced cancer Lung clinic patient excluded: 1 did not meet prognostic criteria 24 home care patients completed study 1 outpatient radiotherapy unit patient completed study 1 outpatient lung clinic patient completed study 50 palliative cancer patients recruited 26 from Cross Cancer Institute 24 from Regional Palliative Home Care 24 Department of Symptom Control and Palliative Care patients completed study 30 patients referred from Department of Symptom Control and Palliative Care Multidisciplinary Pain and Symptom Outpatient Clinic inpatient consultation service Department of Symptom Control and Palliative Care patients excluded: 4 declined due to severe fatigue 2 declined due to feeling overwhelmed Fig. 1. Flow of participants through the study. participants are presented in Tables 1 and 2, respectively. Physical Activity Behavior, Quality of Life, Physical Functioning, and Symptoms Table 3 presents descriptive data for physical activity behavior, QoL, physical function, and symptoms of all participants (n ¼ 50). Fifty percent of the participants reported walking 30 minutes or more per day over the past week (n ¼ 25), and 50% reported that they engaged in physical activity for 60 minutes or more per day over the past week (n ¼ 25). Associations Between Physical Activity Behavior and Patient-Reported Quality of Life Data for the primary outcomes of patientreported QoL are presented in Tables 4 and 5. There was an overall trend favoring the high walking group and high physical activity group over the low groups for patientreported QoL. Participants who reported walking 30 minutes or more per day over the past week also reported higher MQOL existential subscores (mean between-group difference ¼ 0.8; 95% confidence interval [CI] ¼ 0.0e1.5; P ¼ 0.045), higher MQOL support subscores (mean between-group difference ¼ 0.7; 95% CI ¼ 0.1e1.4; P ¼ 0.027), and higher MQOL total scores (mean between-group difference ¼ 0.5; 95% CI ¼ 0.0e0.9; P ¼ 0.046). Cohen s effect size d for these differences ranged from 0.58 to These differences in QoL did not change substantially after adjusting for the aforementioned covariates. Similarly, participants who reported physical activity of 60 minutes or more per day over the past week also reported higher MQOL existential subscores (mean between-group difference ¼ 1.1; 95% CI ¼ 0.4e1.8; P ¼ 0.002); Cohen s effect size d for this difference was This difference in QoL remained after adjusting for the aforementioned potential covariates (mean adjusted between-group difference ¼ 1.0; 95% CI ¼ 0.2e1.8; P ¼ 0.011). Interestingly, participants who reported participating in physical activity for 60 minutes

6 790 Lowe et al. Vol. 38 No. 5 November 2009 Table 1 Demographic Characteristics of Participants (n ¼ 50) Demographic Variables n (%) Age, years <60 21 (42) $60 29 (58) Mean SD Sex Male 20 (40) Female 30 (60) Marital status Married/common law 21 (42) Other 29 (58) Education Completed Grade 12 or higher 25 (50) Lower than Grade (50) Annual family income <$40,000 8 (16) $$40, (84) Employment status Employed full/part time 4 (8) Not employed 46 (92) SD ¼ standard deviation. or more per day over the past week also reported lower MQOL psychological subscores (mean between-group difference ¼ 1.3; 95% CI ¼ 2.5 to 0.1; P ¼ 0.039), and this difference was only slightly attenuated in adjusted analysis (mean adjusted between-group difference ¼ 1.2; 95% CI ¼ 2.6 to 0.2; P ¼ 0.079). Associations Between Physical Activity Behavior and Patient-Reported Physical Functioning Data for the secondary outcomes of patientreported physical functioning are presented in Tables 4 and 5. On comparison across categories of walking and total physical activity over the past week, there was an overall pattern favoring the higher activity groups over their lower counterparts for patient-reported physical functioning. Although not statistically significant, there was a trend toward better functioning in participants who reported walking 30 minutes or more per day over the past week for the LLFDI advanced lower extremity function subscores (mean between-group difference ¼ 2.1; 95% CI ¼ 5.1 to 0.9; P ¼ 0.l72) and lower LLFDI total scores (mean between-group difference ¼ 3.3; 95% CI ¼ 9.1 to 2.5; P ¼ 0.261). Cohen s effect size d for these differences were 0.39 and 0.32, respectively. These differences in patient-reported physical functioning did not change after adjusting for the aforementioned covariates. Although not statistically significant, participants who reported engaging in physical activity for 60 minutes or more per day over the past week also reported better LLFDI advanced lower extremity function subscores (mean betweengroup difference ¼ 2.5; 95% CI ¼ 5.5 to 0.5; P ¼ 0.096) and better LLFDI total scores (mean between-group difference ¼ 1.8; 95% CI ¼ 7.7 to 4.0; P ¼ 0.530). Adjusting for the aforementioned potential covariates did not substantially alter these results. Associations Between Physical Activity Behavior and Patient-Reported Symptoms Data for the secondary outcomes of patient-reported symptoms are presented in Tables 4 and 5. There is an overall trend favoring the higher walking and physical activity groups over their lower counterparts for better patient-reported symptoms. Although not statistically significant, there was a trend toward association between participants who reported walking 30 minutes or more per day over the past week and lower fatigue scores (mean between-group difference ¼ 0.7; 95% CI ¼ 2.1 to 0.6; P ¼ 0.273) on the ESAS; Cohen s effect size d for this difference was These differences in patientreported symptoms did not change after adjusting for the covariates. With respect to the total physical activity categories, there was more variability apparent in associations between total physical activity behavior over the past week and patient-reported symptoms, although none achieved statistical significance. Adjusting for the aforementioned potential covariates did not substantially alter these results. Moderator Analysis of Associations Between Physical Activity Behavior, Patient-Reported Quality of Life, Physical Functioning, and Symptoms None of the aforementioned variables moderated the association between physical activity behavior and total MQOL scores, total LLFDI scores, ESAS pain scores, and ESAS fatigue scores. Covarying did not substantially alter the results of our moderator analyses.

7 Vol. 38 No. 5 November 2009 Physical Activity, QoL, and Palliative Care 791 Table 2 Medical Characteristics of Participants (n ¼ 50) Medical Variables n (%) Body Mass Index Underweight 8 (16) Healthy 17 (34) Overweight 17 (34) Obese 8 (16) Mean SD Number of comorbidities <2 29 (58) $2 21 (42) Most common comorbidities Hypertension 17 (34) Arthritis 12 (24) Dyslipidemia 11 (22) COPD 6 (12) Smoking status Never smoked 20 (40) Ex-smoker 22 (44) Current smoker 8 (16) CAGE 0 48 (96) 1 2 (4) Cancer diagnosis Lung 15 (30) Genitourinary 11 (22) Breast 8 (16) Gastrointestinal 8 (16) Hematological 4 (8) Head and neck 2 (4) Other 2 (4) PPS level, % 40 2 (4) 50 2 (4) (50) (38) 80 2 (4) Sites of metastases Bone 22 (44) Lung 22 (44) Liver 11 (22) Brain 5 (10) Other 18 (36) Number of metastastic sites #1 23 (46) $2 27 (54) Treatment received Surgery Never 23 (46) Completed 27 (54) Chemotherapy Never 15 (30) Completed 18 (36) Current 17 (34) Radiation Never 19 (38) Completed 28 (56) Current 3 (6) Recurrence Never 34 (68) $1 16 (32) SD ¼ standard deviation; CAGE ¼ screen for alcohol abuse, rated from 0 to 4 for a maximum total score of 4; PPS ¼ Palliative Performance Scale; COPD ¼ chronic obstructive pulmonary disease. Table 3 Descriptive Statistics for Physical Activity Behavior, Quality of Life, Physical Function, and Symptoms (n ¼ 50) Variable Mean SD Physical activity over the past week, minutes Walking Housework Stair climbing Miscellaneous Total physical activity McGill Quality of Life Questionnaire Physical symptom subscale (0e10) Physical well-being subscale (0e10) Psychological subscale (0e10) Existential subscale (0e10) Support subscale (0e10) Total score (0e10) Late-Life Function and Disability Instrument Basic lower extremity functioning subscale (5e25) Advanced lower extremity functioning subscale (5e25) Upper extremity functioning subscale (5e25) Total function score (15e75) Edmonton Symptom Assessment System Pain Fatigue Nausea Depression Anxiety Drowsiness Appetite Feeling of well-being Shortness of breath SD ¼ standard deviation Discussion This study reports the associations between physical activity and patient-reported QoL, physical functioning, and symptoms in a sample of advanced cancer patients receiving palliative care. Our results showed moderate to large associations between participants who reported walking 30 minutes or more per day over the past week and MQOL existential, support, and total scores. Similarly, there was a large association between participants who reported engaging in physical activity for 60 minutes or more per day over the past week and MQOL existential scores. Overall, there appears to be an advantage in QoL for those palliative cancer patients who reported greater levels of walking and total physical activity over the past week. As hypothesized, the amount of physical activity reported by this sample of cancer

8 792 Lowe et al. Vol. 38 No. 5 November 2009 Table 4 Quality of Life, Physical Functioning, and Symptom Differences Between Participants Reporting Walking 30 Minutes or More Per Day (n ¼ 25) and Walking Less Than 30 Minutes Per Day (n ¼ 25) Over the Past Week Variable Walking <30 Minutes/Day (Mean SD) Walking $30 Minutes/Day (Mean SD) Between-Group Difference Mean (95% CI) F Effect Size (d ) McGill Quality of Life Questionnaire Physical symptom (0e10) ( 0.5 to 0.9) Physical well-being (0e10) ( 0.3 to 1.6) Psychological (0e10) ( 1.4 to 1.1) Existential (0e10) (0.0e1.5) Support (0e10) (0.1e1.4) Total score (0e10) (0.0e0.9) Late-Life Function and Disability Instrument Basic lower extremity function (0e25) ( 2.2 to 2.0) Advanced lower extremity function (0e25) ( 5.1 to 0.9) Upper extremity function (0e25) ( 3.4 to 1.2) Total function score (0e75) ( 9.1 to 2.5) Edmonton Symptom Assessment System Pain (0e10) ( 1.0 to 1.8) Fatigue (0e10) ( 2.1 to 0.6) Nausea (0e10) ( 2.4 to 0.0) Depression (0e10) ( 2.1 to 0.5) Anxiety (0e10) ( 1.4 to 1.4) Drowsiness (0e10) ( 2.2 to 1.0) Appetite (0e10) ( 2.5 to 0.6) Feeling of well-being (0e10) ( 2.0 to 0.3) Shortness of breath (0e10) ( 2.4 to 0.8) SD ¼ standard deviation. Cohen s effect size d $ 0.35, P < 0.05 highlighted in boldface. McGill Quality of Life Questionnaire: higher scores ¼ better quality of life; Late-Life Function and Disability Instrument: lower scores ¼ better functioning; Edmonton Symptom Assessment System: lower scores ¼ lower symptom severity. P patients was low. None of our participants reported engaging in physical activities that would meet the Godin Leisure-Time Exercise Questionnaire criteria for strenuous exercise (i.e., running, aerobic classes, cross-country skiing, vigorous swimming, or vigorous bicycling) or that of even moderate intensity (i.e., brisk walking). 25 Most of our participants reported independent activities of daily living, such as housework and climbing house stairs, as their common physical activities over the past week. The median time spent walking was 225 minutes in total over the past week by means of patient self-report. Given the tendency of patients to overestimate physical activity levels on self-report, 26 it is clear that the amount of physical activity undertaken by the participants in our sample is very low, which is not unexpected given the progressive fatigue, cachexia, and debility that patients encounter at the end stages of cancer. Also consistent with our hypotheses, we found that physical activity was positively associated with QoL in advanced cancer patients. In their study of the MQOL s test-retest reliability and responsiveness in advanced cancer patients, Cohen and Mount 27 determined that producing a difference of 0.7 points in MQOL support score was similar to changing a bad day into an average day, as perceived by the patient. This difference corresponds to our mean difference of 0.7 points for MQOL support score when comparing patientreported walking categories. Our mean difference of 1.1 points for MQOL existential score between patient-reported total physical activity categories is greater than the 0.9 points that Cohen and Mount 27 deemed was similar to changing an average day into a good day for the MQOL existential subscale. In a pilot study of a group exercise intervention in 34 mixed cancer patients with an estimated life expectancy of 3e12 months, Oldervoll et al. 6 reported a statistically significant improvement in both the emotional and social functioning subscores of the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-C30). In our sample, the MQOL existential subscore had a statistically significant

9 Vol. 38 No. 5 November 2009 Physical Activity, QoL, and Palliative Care 793 Table 5 Quality of Life, Physical Functioning, and Symptom Differences Between Participants Reporting Total Physical Activity of 60 Minutes or More Per Day (n ¼ 25) and Total Physical Activity Less Than 60 Minutes Per Day (n ¼ 25) Over the Past Week Variable Physical Activity <60 Minutes/Day (Mean SD) Physical Activity $60 Minutes/Day (Mean SD) Between-Group Difference Mean (95% CI) F Effect Size (d ) McGill Quality of Life Questionnaire Physical symptom (0e10) ( 0.5 to 0.9) Physical well-being (0e10) ( 0.5 to 1.4) Psychological (0e10) ( 2.5 to 0.1) Existential (0e10) (0.4e1.8) Support (0e10) ( 0.5 to 0.9) Total score (0e10) ( 0.3 to 0.6) Late-Life Function and Disability Instrument Basic lower extremity function (0e25) ( 2.9 to 1.3) Advanced lower extremity ( 5.5 to 0.5) function (0e25) Upper extremity function (0e25) ( 0.8 to 3.7) Total function score (0e75) ( 7.7 to 4.0) Edmonton Symptom Assessment System Pain (0e10) ( 1.9 to 0.9) Fatigue (0e10) ( 1.2 to 1.4) Nausea (0e10) ( 1.5 to 1.0) Depression (0e10) ( 1.9 to 0.6) Anxiety (0e10) ( 0.1 to 2.6) Drowsiness (0e10) ( 1.6 to 1.6) Appetite (0e10) ( 2.2 to 1.0) Feeling of well-being (0e10) ( 1.4 to 0.9) Shortness of breath (0e10) ( 1.6 to 1.7) SD ¼ standard deviation. Cohen s effect size d $ 0.35, P < 0.05 highlighted in boldface. McGill Quality of Life Questionnaire: higher scores ¼ better quality of life; Late-Life Function and Disability Instrument: lower scores ¼ better functioning; Edmonton Symptom Assessment System: lower scores ¼ lower symptom severity. P positive association with both patient-reported walking and total physical activity over the past week. The MQOL existential subscore measures the patient s concerns about their personal meaning in life and has been established as a fundamental determinant of QoL in cancer patients with advanced illness. 11 In two studies of 420 mixed cancer survivors and 167 breast cancer survivors, Jim and Andersen 28 reported that the association between impairments in physical functioning and distress were mediated by the individual s perceived meaning in life. This evidence linking existential concerns and physical functioning may explain the strong positive association between patient-reported physical activity and the MQOL existential subscore in our participants, as they may perceive being physically active with delaying or slowing functional impairments. Interestingly, there was a statistically significant association between higher patientreported total physical activity over the past week and worsened MQOL psychological subscores. In a study of 170 mixed cancer survivors before chemotherapy, Pinquart et al. 29 reported that decreased functional status was associated with decreased EORTC QLQ-C30 QoL scores in patients with low and average levels of depression but not in patients with mild to moderate depression; this suggests that impairments in physical functioning may contribute less to overall QoL when there are more significant psychological issues present. Given that the MQOL psychological subscore is a composite measure of the patient s depression, anxiety, and sadness, this phenomenon may explain the negative association between patient-reported physical activity and psychological well-being in our participants. Oldervoll et al. 6 reported no change in EORTQ QLQ-C30 physical functioning scores postintervention in their pilot study of 34 advanced cancer patients. In our sample, there was an overall pattern favoring the higher activity groups over their lower counterparts for physical functioning; however, this did not

10 794 Lowe et al. Vol. 38 No. 5 November 2009 reach statistical significance. The abbreviated LLFDI instrument may not have had sufficient sensitivity to discriminate functional differences within an advanced cancer population experiencing inevitable physical decline. In a recent systematic review, Jordhoy et al. 30 showed that physical functioning is a neglected dimension in palliative care QoL measures, and that there is little consensus as to how physical functioning should be assessed or what components of physical functioning should be elicited in cancer patients. Clearly, future research is required into standardized tools of physical functioning that are validated for this population. Oldervoll et al. 6 reported statistically significant improvement in dyspnea, with a trend of improvement in EORTC QLQ-C30 fatigue, pain, and nausea scores in 34 advanced cancer patients post exercise intervention. In our sample, variability in patient-reported symptoms between physical activity groups may be explained by our use of the ESAS, which assesses each symptom using a single-item visual analogue scale. 12 Multidimensional symptoms, such as cancer-related fatigue, may require the use of multidimensional assessment tools to elucidate potential associations. 31 Future studies may benefit from the selection of a multidimensional instrument for more in-depth exploration of the patientreported symptom as the primary outcome. Neither age, gender, BMI, Palliative Performance Scale level, total number of comorbidities, cancer diagnosis, nor survival moderated the association between physical activity behavior and total MQOL scores, total LLFDI scores, ESAS pain scores, and ESAS fatigue scores. Given our small sample size, it was not feasible to establish the dose effect of physical activity behavior or relationships with specific symptom clusters without compromising statistical power and validity. Larger studies are required to confirm these results and delineate potential interactions between demographic and medical variables with physical activity behavior and QoL, physical functioning, and symptoms in this cancer population. Considering the variable morbidity of patients within our heterogeneous referral population, we anticipated that only a select subset of patients would express interest in physical activity. The distributed copy of research participation document was provided to consecutive patients, irrespective of study eligibility, in the Capital Health Palliative Home Care Program and the CCI outpatient radiotherapy units and lung clinics. Given that the patient had to grant permission to be contacted by the study coordinator, demographic and medical data for nonconsenting patients was not available for comparison with our sample; there may be differences between the two groups that may have influenced our findings and may reduce generalizability. Because of the observational nature of this study, one cannot infer a cause-effect relationship between physical activity and QoL in advanced cancer patients or if other variables are responsible for this association. Despite adjusting for multiple potential covariates, there may be other factors involved that were not measured in our sample. Finally, we relied on selfreported physical activity, which has not been validated in this type of cancer population. Our survey items did not discriminate between the ability to walk or the choice to walk, and the distinction between the two may have different implications for overall QoL. Future studies combining both objective measurements and patient-reported assessments of physical activity levels should be conducted in the cancer population with advanced illness. Nonetheless, this study has key implications for future research. Our results show a positive association between patient-reported physical activity and QoL in advanced cancer patients. In particular, there is a great association between higher physical activity levels and better existential well-being, which is a fundamental determinant of QoL in these cancer patients. 32 Further exploration of a global QoL index that incorporates physical activity may be of clinical utility for this population. Given the benefits of physical activity interventions in early-stage cancer populations, these results lend impetus to initiate pilot physical activity intervention trials in cancer patients with more advanced disease. References 1. Stromgren AS, Sjogren P, Goldschmidt D, et al. Symptom priority and course of symptomatology in specialized palliative care. J Pain Symptom Manage 2006;31:199e206.

11 Vol. 38 No. 5 November 2009 Physical Activity, QoL, and Palliative Care Simmonds MJ. Physical function in patients with cancer: psychometric characteristics and clinical usefulness of a physical performance test battery. J Pain Symptom Manage 2002;24:404e Conn VS, Hafdahl AR, Porock DC, McDaniel R, Nielsen PJ. A meta-analysis of exercise interventions among people treated for cancer. Support Care Cancer 2006;12:699e Lowe SS, Watanabe SM, Courneya KS. Physical activity as a supportive care intervention in palliative cancer patients: a systematic review. J Support Oncol 2009;7:27e Oldervoll LM, Loge JH, Paltiel H, et al. The effect of a physical exercise program in palliative care: a phase II study. J Pain Symptom Manage 2006;31:421e Oldervoll LM, Loge JH, Paltiel H, et al. Are palliative cancer patients willing and able to participate in a physical exercise program? Palliat Support Care 2005;3:281e Mungas D, Marshall SC, Weldon M, Haan M, Reed BR. Age and education correction of Mini-- Mental State Examination for English and Spanish-- speaking elderly. Neurology 1996;46:700e Whaley MH, Brubaker PH, Otto RM. Pre-exercise evaluations. In: Whaley MH, Brubaker PH, Otto RM, eds. American College of Sports Medicine s guidelines for exercise testing and prescription, 7th ed. Baltimore: Lippincott Williams & Wilkins, 2006: 39e Anderson F, Downing GM, Hill J, Casorso L, Lerch N. Palliative performance scale (PPS): a new tool. J Palliat Care 1996;12:5e Cohen SR, Mount BM, Strobel MG, Bui F. The McGill Quality of Life Questionnaire: a measure of quality of life appropriate for people with advanced disease. A preliminary study of validity and acceptability. Palliat Med 1995;9:207e Cohen SR, Mount BM, Bruera E, et al. Validity of the McGill Quality of Life Questionnaire in the palliative care setting: a multi-centre Canadian study demonstrating the importance of the existential domain. Palliat Med 1997;11:3e Bruera E, Kuehn N, Miller MJ, Selmser P, Macmillan K. The Edmonton Symptom Assessment System (ESAS): a simple method for the assessment of palliative care patients. J Palliat Care 1991;7:6e Nekolaichuk C, Watanabe S, Beaumont C. The Edmonton Symptom Assessment System: a 15-year retrospective review of validation studies (1991e2006). Palliat Med 2008;22:111e Washburn RA, Smith KW, Jette AM, Janney CA. The Physical Activity Scale for the Elderly (PASE): development and evaluation. J Clin Epidemiol 1993;46:153e Bouchard C, Shephard RJ. Physical activity, fitness, and health: the model and key concepts. In: Bouchard C, Shephard RJ, Stephens T, eds. Physical activity, fitness and healthdinternational proceedings and consensus statement, 1st ed. Champaign, IL: Human Kinetics Publishers, 1994: 77e Martin KA, Rejeski WJ, Miller ME, et al. Validation of the PASE in older adults with knee pain and physical disability. Med Sci Sports Exerc 1999;31:627e Johansen KL, Painter P, Kent-Braun JA, et al. Validation of questionnaires to estimate physical activity and functioning in end-stage renal disease. Kidney Int 2001;60:1586e McAuley E, Konopack JF, Motl RW, Rosengren K, Morris KS. Measuring disability and function in older women: psychometric properties of the Late-Life Function and Disability Instrument. J Gerontol A Biol Sci Med Sci 2005;60:901e Sayers SP, Jette AM, Haley SM, et al. Validation of the Late-Life Function and Disability Instrument. J Am Geriatr Soc 2004;52:1554e Courneya KS, Karvinen KH, Campbell KL, et al. Associations among exercise, body weight, and quality of life in a population-based sample of endometrial cancer survivors. Gynecol Oncol 2005;97:422e Vallance JK, Courneya KS, Jones LW, Reiman T. Differences in quality of life between non-hodgkin s lymphoma survivors meeting and not meeting public health exercise guidelines. Psychooncology 2005; 14:979e Stevinson C, Faught W, Steed H, et al. Associations between physical activity and quality of life in ovarian cancer survivors. Gynecol Oncol 2007;106: 244e Doyle C, Kushi LH, Byers T, et al. Nutrition and physical activity during and after cancer treatment: an American Cancer Society guide for informed choices. CA Cancer J Clin 2006;56:323e Cohen J. Statistical power analysis for the behavioral sciences, 2nd ed. Hillsdale, NY: Lawrence Erlbaum Associates, Godin G, Shephard RJ. A simple method to assess exercise behavior in the community. Can J Appl Sport Sci 1985;10:141e Sallis JF, Saelens BE. Assessment of physical activity by self-report: status, limitations and future directions. Res Q Exerc Sport 2000;71(2 Suppl):S1eS Cohen SR, Mount BM. Living with cancer: good days and bad daysdwhat produces them? Cancer 2000;89:1854e Jim HS, Andersen BL. Meaning in life mediates the relationship between social and physical functioning and distress in cancer survivors. Br J Health Psychol 2007;12(Pt 3):363e Pinquart M, Koch A, Eberhardt B. Associations of functional status and depressive symptoms with health-related quality of life in cancer patients. Qual Life Res 2006;15:1565e Jordhoy MS, Inger Ringdal G, Helbostad JL, et al. Assessing physical functioning: a systematic

12 796 Lowe et al. Vol. 38 No. 5 November 2009 review of quality of life measures developed for use in palliative care. Palliat Med 2007;21:673e Munch TN, Stromgren AS, Pedersen L, et al. Multidimensional measurement of fatigue in advanced cancer patients in palliative care: an application of the multidimensional fatigue inventory. J Pain Symptom Manage 2006;31:533e Cohen SR, Leis A. What determines the quality of life of terminally ill cancer patients from their own perspective? J Palliat Care 2002;18:48e58.