Gliomas are the most common primary neoplasm of the central nervous system, accounting for greater than 50% of these diagnoses.1 Despite significant advancements in treatment, concomitant improvements in survival have been poor.2, 3 Currently, the ‘standard’ management includes surgical resection followed by 6 weeks of conventional fractioned radiotherapy (RT), with or without adjuvant chemotherapy. Despite these aggressive approaches, the median survival for patients with brain tumors ranges from 1 year to 5 years from the time of the initial diagnosis.1–3
Patients presenting with primary gliomas may have a broad array of debilitating neurologic symptoms.4 Typically, these symptoms arise from mass effect because of tumor growth and the associated vasogenic edema. The mass effect can be reduced by surgical resection and corticosteroid medication, most commonly dexamethasone. Corticosteroids are currently the only effective medical treatment for the long-term control of intracranial edema associated with gliomas.5, 6 Unfortunately, these pharmacologic therapies are associated with many deleterious side effects, including peripheral edema, Cushing syndrome, fatigue, hyperglycemia, thromboembolism, and, in particular, severe skeletal muscle catabolism. Collectively, these side-effects which can have a profound impact on a patient's quality of life (QOL) and physical functioning/performance status.7–10 Performance status has been consistently demonstrated to be a strong independent predictor of morbidity and mortality in patients with neurologic malignancies.11–13 Clearly, supportive interventions that attenuate therapy-induced functional declines may have substantial clinical benefits for this population.
One supportive intervention that may complement existing conventional therapies and address a multitude of therapy-induced toxicities in patients with brain tumors is physical exercise. In recent years, increased attention has focused on exercise as a rehabilitative intervention for cancer survivors both during and after the cessation of cancer therapy.14, 15 The majority of studies have reported statistically significant findings across a wide range of psychosocial (e.g., depression, anxiety, symptoms, etc.) and physiologic (e.g., muscle strength, immune and metabolic profiles, body composition, etc.) endpoints, culminating in clinically meaningful improvements in the patient's functional capacity16–19 and overall QOL.16, 20–22 However, the vast majority of studies to date have focused on breast cancer or bone marrow transplant patients, with fewer studies focusing on patients with colorectal cancer, prostate cancer, multiple myeloma, or non-Hodgkin lymphoma. To our knowledge, no study to date has examined the potential role of exercise in patients diagnosed with primary brain cancer. Given the limited evidence, we conducted a pilot study to examine the patterns of exercise across the cancer trajectory in a cohort of patients with brain tumors. The primary objective was to examine the percentage of participants meeting national exercise prescription guidelines and examine differences in exercise behavior across the cancer trajectory (i.e., prior to diagnosis, during treatment, and after the completion of treatment). A second objective was to explore differences in exercise behavior across the cancer trajectory based on select medical (e.g., disease grade and adjuvant therapy) and demographic (e.g., age and gender) variables.
MATERIALS AND METHODS
Participants and Procedures
The study was conducted at the Preston Robert Tisch Brain Tumor Center (BTC) at Duke University Medical Center (DUMC) in Durham, North Carolina. Potential participants consisted of all brain tumor patients seen at the BTC between August 2003 and February 2005. Additional inclusion criteria were 1) histologically confirmed primary brain cancer, 2) the patient being of legal age, 3) approval of the patient's primary treating oncologist, and 4) the patient's ability to understand and provide written informed consent in English. Institutional review board approval was obtained before commencement of the study. Using a cross-sectional design, potential participants were identified and screened through the BTC cancer database at DUMC and oncologist approval to contact patients was pursued. After approval, each potential participant was sent a questionnaire package that contained a detailed cover letter; 2 copies of a consent form; a questionnaire; and a stamped, self-addressed return envelope. Survey methods known to increase response rates were used, including multiple reminders (e.g., telephone calls), stamped return envelopes, personalized cover letters, colored paper, assurances of confidentiality, a small incentive of 3 commemorative postage stamps, and BTC sponsorship.23 The study was conducted between May and October of 2005.
Exercise behavior was assessed by the Leisure Score Index (LSI) of the Godin Leisure-Time Exercise Questionnaire.24, 25 The LSI contains 3 questions that assess the average frequency of mild (minimal effort, no perspiration), moderate (not exhausting, light perspiration), and strenuous intensity (heart beats rapidly, sweating) exercise during free time in a typical week. In the current study, participants reported their average weekly exercise for the 3 cancer-related time periods (before their brain tumor diagnosis, during active treatment, and after the completion of treatment). We also asked for the average duration of exercise within each exercise intensity. Separate scores were calculated for total exercise minutes and strenuous, moderate, and mild intensity exercise minutes, as well as the percentage of participants meeting current national exercise guidelines (i.e., at least 150 minutes of moderate intensity or higher exercise per week).26 Finally, we also assessed the percentage of participants reporting no exercise behavior across the 3 cancer-related time periods.
Demographic and Medical Information
Demographic and medical information were collected using self-report measures and via medical chart review. Self-report data included marital status, educational level attained, current employment status, and smoking history. Medical data were abstracted from the most recent chart notes and included age, weight, height, body mass index (BMI) (calculated as weight divided by height squared, [kg/m2]), months since diagnosis, Karnofsky performance status (KPS), tumor grade, brain tumor subtype, prior/current treatment (i.e., surgery, RT, or chemotherapy), and current treatment status (on active therapy, off therapy with active disease, or off therapy without active disease). We also calculated comorbidity status using the Charlson index, a comorbidity index commonly used in health research.27
Descriptive analyses are presented for the demographic/medical characteristics, the patterns of exercise behavior across the cancer trajectory, and the percentage of participants meeting national exercise prescription guidelines. Differences in exercise behavior (total exercise minutes; strenuous, moderate, and mild exercise minutes; percentage of participants meeting national guidelines; and those reporting no exercise behavior) across the cancer trajectory (i.e., prior to diagnosis, during active treatment, and after the completion of treatment) were analyzed using repeated measures multivariate analysis of variance (RM-MANOVA) followed by repeated measures analysis of variance (RM-ANOVA) and pairwise comparisons when appropriate. We used 1-way ANOVA to examine changes in exercise behavior across the cancer trajectory by age, gender, BMI, disease grade, treatment status, and KPS. Finally, we also repeated analyses using analysis of covariance (ANCOVA) with age, gender, marital status, employment status, smoking history, KPS, comorbidity score, months since diagnosis, and treatment status as covariates. Data are presented as the mean (± the standard deviation) with 95% confidence intervals (95% CIs). All statistical tests were 2-sided (α = .05). Data were analyzed using SPSS software (version 12.0; SPSS Inc., Chicago, IL).
Approximately 771 patients were seen at the BTC at DUMC between 2003 and 2005. Of these patients, 560 met inclusion criteria (211 patients were deceased) and 3 oncologists representing these patients were contacted. Approval was granted to contact all the patients. Each of the 560 potential participants were sent a questionnaire package. A total of 177 questionnaires were returned unopened (167 patients were deceased and 10 of the envelopes were marked “return to sender”), leaving a sample of 383 participants. We received 106 completed questionnaires, for a response rate of 28% (106 of 383 patients).
Details of the demographic and medical characteristics of the participants are presented in Table 1. In summary, the mean age of the participants was 44.8 ± 12.0 years;, 50.9% were female, 76.5% were married, 82.1% had completed university/college, 38.7% were currently employed, and 40.6% had smoked >100 cigarettes in their lifetime. Clinical treatment characteristics indicated that 54.7% of the patients were overweight or obese (with a BMI of >25 kg/m2), 75.5% had Grade 3/intravenous disease, and the majority were diagnosed with either glioblastoma multiforme (38.7%) or anaplastic astrocytoma (43.4%) brain tumor subtypes. Approximately 47% of patients were currently receiving active treatment whereas approximately 100%, 80.2%, and 75.5%, respectively, had received prior surgery, chemotherapy, and RT. The mean KPS and Charlson index score of the patients was 89% (range, 50-100%) and 2.3 (range, 2–4), respectively. Finally, the average time since diagnosis was 28 months (range, 6-178 mos).
Table 1. Characteristics of the Patients, Tumors, and Primary Treatments (n = 106)*
|Age in y|| |
| <45||51 (48.1)|
| >45||49 (46.2)|
| Missing data||6 (5.7)|
|Mean age in y||44.8 ± 12.0|
| Male||52 (49.1)|
| Female||54 (50.9)|
| Missing data||0 (0)|
|Marital status|| |
| Married/common law||81 (76.5)|
| Divorced/separated||6 (5.6)|
| Widowed||1 (1.0)|
| Never married||16 (15.1)|
| Missing data||2 (1.9)|
| Some high school||3 (2.9)|
| Completed high school||15 (14.2)|
| Completed university/college||87 (82.1)|
| Missing data||1 (1.0)|
|Employment status|| |
| Retired||12 (11.3)|
| Disability||41 (38.7)|
| Employed full/part-time||41 (38.7)|
| Temporarily unemployed||10 (9.4)|
| Missing data||2 (1.9)|
|Smoked >100 lifetime cigarettes|| |
| Yes||43 (40.6)|
| No||62 (58.5)|
| Missing data||1 (1.0)|
|Current smoker|| |
| Yes||12 (11.3)|
| No||91 (85.8)|
| Missing data||1 (1.0)|
|Medical information|| |
|Mean (SD) weight in kg||78.9 ± 17.9|
|BMI, no. (%)|| |
| <25 kg/m2||46 (43.4)|
| 25–30 kg/m2||39 (36.8)|
| >30 kg/m2||19 (17.9)|
| Missing data||2 (1.9)|
|Mean time since diagnosis in mo||28.4 ± 28.6|
| Range in mo||6–178|
|Tumor grade|| |
| I||7 (6.6)|
| II||17 (16.0)|
| III||36 (34.0)|
| IV||44 (41.5)|
| Missing data||2 (1.9)|
|Tumor subtype|| |
| Glioblastoma multiforme||41 (38.7)|
| Anaplastic astrocytoma||46 (43.4)|
| Astrocytoma/glioma||12 (11.3)|
| Pilocytic astrocytoma||4 (3.8)|
| Cerebellar medulloblastoma||1 (1.0)|
| Meningioma||1 (1.0)|
| Oligodendroglioma||1 (1.0)|
| Missing data||0 (0)|
|Treatment status|| |
| Receiving active therapy||50 (47.2)|
| Off therapy||55 (51.9)|
| Missing data||1 (1.0)|
|Surgery for primary tumor|| |
| Yes||105 (99.0%)|
| No||0 (0)|
| Missing data||1 (1.0)|
|Type of surgery|| |
| Total macroscopic resection||77 (72.6)|
| Subtotal resection||28 (26.4)|
| Missing data||1 (1.0)|
|Treatment with chemotherapy|| |
| Yes||85 (80.2)|
| No||21 (19.8)|
| Missing data||0 (0)|
|Treatment with radiotherapy|| |
| Yes||80 (75.5)|
| No||26 (24.5)|
| Missing data||0 (0)|
|Mean (SD) Karnofsky performance score, %||88.8 ± 8.7|
| Range in %||50–100|
|Mean (SD) Charlson comorbidity score||2.3 ± 0.6|
Exercise Behavior across the Cancer Trajectory
Table 2 shows the descriptive statistics for exercise behavior across the cancer trajectory and between cancer-related time period comparisons. Analyses indicated that 42%, 38%, and 41% of participants, respectively, met national exercise prescription guidelines before their brain tumor diagnosis, during active treatments, and after the completion of treatment, whereas 41%, 41%, and 31% of participants, respectively, reported no exercise behavior over the same cancer-related time periods. The overall RM-MANOVA for all exercise behavior outcomes was not found to be significant (Wilks lambda of .867; F(10, 96) = 1.474 [P = .161]). Although exercise behavior was found to be lower during active and off-treatment periods compared with prior to diagnosis, exploratory RM-ANOVAs indicated no significant changes across the cancer-specific time periods for any exercise behavior outcome except moderate minutes between prediagnosis and off treatment (mean difference of -20.6 mins/wk; 95% CI, -38.0 mins/wk to -3.2 mins/wk) and the percentage of patients reporting no exercise between active and off-treatment periods (mean difference of 10%; 95% CI, 2-19% [P = .021]). Pairwise comparisons indicated that participants reported significantly more moderate intensity weekly exercise minutes while they were off treatment compared with prior to diagnosis, whereas significantly fewer participants reported no physical exercise while off treatment compared with during therapy.
Table 2. Descriptive Statistics for Participants' Exercise Patterns across the Cancer Trajectory
|Total minutes||160.8 ± 224.9||154.5 ± 222.2||177.0 ± 226.5||−6.3 [−62.6 to 49.9]||.824||16.1 [−42.7 to 75.0]||.588||22.5 [−12.1 to 57.2]||.202|
|Strenuous minutes||88.5 ± 134.6||67.9 ± 120.6||72.6 ± 109.4||−20.6 [−54.9 to 13.7]||.236||−15.9 [−49.8 to 8.1]||.355||4.7 [−13.3 to 22.7]||.604|
|Moderate minutes||27.5 ± 77.1||36.1 ± 72.8||48.1 ± 82.5||8.7 [−9.0 to 6.3]||.332||20.6 [−3.2 to −38.0]||.020||11.9 [3.5 to −27.4]||.129|
|Mild minutes||44.9 ± 88.2||50.5 ± 88.2||56.3 ± 86.1||5.6 [−16.2 to 27.5]||.611||11.4 [−10.1 to 32.9]||.294||5.8 [−8.9 to 20.5]||.434|
|% meeting ACSM guidelines†||42 ± 50||38 ± 49||41 ± 49||−4 [−17 to 9]||.558||−1.0 [−15 to −13]||.895||3.0 [−6 to 12]||.534|
|% reporting no exercise||41 ± 49||41 ± 50||31 ± 47||0 [12 to −12]||.992||10 [−3 to 22]||.141||−10 [−8 to 18]||.032|
Changes in Exercise Behavior by Select Demographic and Medical Variables
We also explored differences in exercise behavior across the cancer trajectory based on gender (52 males and 54 females), age (<45 yrs [51 patients] and ≥45 yrs [49 patients]), BMI (<25 kg/m2 [46 patients] and ≥25 kg/m2 [58 patients]), disease grade (Grade I/II [24 patients] and Grade III/intravenous [80 patients]), treatment status (on active treatment [50 patients] and off treatment [55 patients]), and KPS (<90% [24 patients] and ≥90% [75 patients]). Given the lack of differences noted in exercise behavior between patients receiving active treatment and those off treatment in our prior analyses, potential differences in exercise behavior between these time periods by the selected demographic and medical variables were not examined.
We found statistically significant gender by time interactions for total minutes (mean difference of -109.0 mins/wk; 95% CI, -220.1 mins/wk to 2.2 mins/wk [P = .053]), moderate minutes (mean difference of -42.5 mins/wk; 95% CI, -77.1 mins/wk to -8.0 mins/wk [P = .016]), and mild minutes (mean difference of -33.6 mins/wk; 95% CI, -86.7 mins/wk to -0.6 mins/wk [P = .047]) from prior to diagnosis to active treatment. There were no other significant interactions noted with regard to exercise behavior outcome (Table 3). Overall, univariate analyses indicated that the significant time by gender interactions were driven primarily by a decrease in exercise behavior by male participants and an increase in exercise behavior by female participants across the cancer trajectory (Table 3).
Table 3. Descriptive Statistics and Between-Group Comparisons for Participants' Exercise Patterns across the Cancer Trajectory by Gender (52 Males and 54 Females)
|Total minutes|| || || || || || || || |
|Male||172.7 ± 231.3||110.9 ± 182.8||165.1 ± 216.8||−61.8|| || ||−7.6|| |
|Female||149.4 ± 220.2||196.6 ± 249.0||188.4 ± 236.9||47.1||−109.0 [−220.1 to 2.2]||.053||39.0||−46.6 [−164.6 to 71.4]||.436|
|Strenuous minutes|| || || || || || || || || |
|Male||81.4 ± 111.1||49.2 ± 85.5||62.2 ± 88.0||−32.2|| || ||−19.2|| |
|Female||95.4 ± 154.6||85.9 ± 145.2||82.6 ± 126.6||−9.4||−22.8 [−91.5 to 6.0]||.513||−12.7||−6.5 [−74.8 to 61.7]||.849|
|Moderate minutes|| || || || || || || || |
|Male||39.2 ± 94.7||26.2 ± 62.5||46.5 ± 84.2||−13.0|| || ||7.3|| |
|Female||16.1 ± 53.7||45.6 ± 81.0||49.5 ± 81.5||29.5||−42.5 [−77.1 to −8.0]||.016||33.4||−26.1 [−60.7 to 8.4]||.137|
|Mild minutes|| || || || || || || || |
|Male||52.0 ± 94.2||35.4 ± 76.9||56.3 ± 82.2||−16.6|| || ||4.3|| |
|Female||38.0 ± 82.3||65.0 ± 96.4||56.2 ± 90.4||27.0||−33.6 [−86.7 to −0.6]||.047||18.2||−13.9 [−57.0 to 29.2]||.523|
|% meeting ACSM‡|| || || || || || || || |
|Male||44 ± 50||31 ± 47||38 ± 49||13|| || ||−6|| |
|Female||39 ± 49||44 ± 50||42 ± 50||−5||−18 [−44 to 6]||.139||3||−9 [−38 to 19]||.511|
|% reporting no exercise|| || || || || || || || |
|Male||42 ± 50||47 ± 50||31 ± 46||−5|| || ||11|| |
|Female||39 ± 49||35 ± 48||31 ± 47||4||−9 [−16 to 31]||.532||8||3 [−21 to 29]||.747|
With regard to the influence of age on exercise behavior across the cancer trajectory, we found significant age by time interactions for total minutes (mean difference of -134.9 mins/wk; 95% CI, -244.6 mins/wk to -25.0 mins/wk [P = 0.017]) and strenuous minutes (mean difference of -85.5 mins/wk; 95% CI, -154.2 mins/wk to -14.6 mins/wk [P = .021]), whereas moderate minutes (mean difference of -30.9 mins/wk; 95% CI, -65.5 mins/wk to 3.7 mins/wk [P = .079]) and the percentage of participants meeting national exercise prescription guidelines (mean difference of -24%; 95% CI, -49% to 2% [P = .067]) approached statistical significance (Table 4) from prediagnosis to active treatment. With regard to prediagnosis to off-treatment comparisons, we found statistically significant age by time interactions for strenuous minutes (mean difference of -78.0 mins/wk; 95% CI, -147.4 mins/wk to -8.6 mins/wk [P = .028]) and percentage meeting national exercise prescription guidelines (mean difference of -34%; 95% CI, -62% to -5 [P = .020]), whereas the percentage of participants reporting no exercise (mean difference of -22%; 95% CI, -48% to 3% [P = .067]) approached statistical significance. Subsequent univariate analyses indicated that participants age < 45 years substantially decreased exercise behavior whereas older participants (age ≥45 yrs) substantially increased exercise behavior over the cancer trajectory (Table 4).
Table 4. Descriptive Statistics and Between-Group Comparisons for Participants' Exercise Patterns across the Cancer Trajectory by Age (45 Patients Were Age < 45 Yrs and 49 Patients Were Age ≥ 45 Yrs)
|Total minutes|| || || || || || || || || |
|< 45 y||218.3 ± 257.2||135.9 ± 184.7||174.1 ± 248.8||−82.6|| || ||−44.2|| || |
|≥45 y||110.0 ± 179.4||162.3 ± 238.3||164.2 ± 182.2||52.3||−134.9 [−244.6 to −25.0]||.017||54.2||−98.4 [−216.3 to 19.5]||.101|
|Strenuous minutes|| || || || || || || || || |
|<45 y||119.7 ± 166.4||52.9 ± 92.1||60.1 ± 98.3||−66.8|| || ||−59.6|| || |
|≥45 y||6+1.7 ± 91.0||79.4 ± 143.3||80.1 ± 117.8||18.7||−85.5 [−154.2 to −14.6]||.018||18.4||−78.0 [−147.4 to −8.6]||.028|
|Moderate inutes|| || || || || || || || || |
|<45 y||37.9 ± 96.9||30.2 ± 60.8||54.8 ± 93.4||−7.7|| || ||16.9|| || |
|≥45 y||17.1 ± 51.8||40.3 ± 77.0||39.3 ± 63.2||23.2||−30.9 [−65.5 to 3.7]||.079||22.1||−5.2 [40.1 to 29.6]||.764|
|Mild minutes|| || || || || || || || || |
|<45 y||60.7 ± 105.7||52.7 ± 76.9||59.2 ± 93.8||−7.9|| || ||−1.5|| || |
|≥45 y||31.1 ± 67.5||42.7 ± 96.2||44.8 ± 72.1||11.5||−19.4 [−63.6 to 24.6]||.383||13.7||−15.2 [−58.4 to 28.1]||.489|
|% meeting ACSM†|| || || || || || || || || |
|<45 y||57 ± 50||39 ± 49||35 ± 48||−18|| || ||−22|| || |
|≥45 y||29 ± 46||35 ± 48||40 ± 49||6||−24 [−49 to 2]||.067||12||−34 [−62 to −5]||.020|
|% reporting no exercise|| || || || || || || || || |
|<45 yrs||29 ± 46||38 ± 49||33 ± 48||−8|| || ||−4|| || |
|≥45 yrs||49 ± 50||45 ± 50||31 ± 47||4||−12 [−36 to 12]||.325||18||−22 [−48 to 3]||.089|
With regard to the influence of BMI, disease grade, treatment status, and KPS on exercise behavior across the cancer trajectory, we found that overweight/obese patients, patients with Grade III/intravenous disease, patients who were not currently undergoing treatment, and patients with a lower KPS generally reported a greater decrease in exercise behavior across the cancer trajectory. However, none of these differences reached statistical significance for any exercise outcome (data not shown). Finally, repetition of all analyses using ANCOVAs was not found to substantially alter any unadjusted analysis.
In light of the promising evidence of exercise as a supportive therapy for cancer patients, the current study sought to examine and describe the exercise patterns of brain tumor patients using an institution-based, cross-sectional survey. To our knowledge, this is the first report to examine this question and adds to a growing body of literature investigating the potential role of exercise as a supportive therapeutic intervention for patients with cancer.
An unexpected finding in the current study was the relatively high number of participants who reported meeting Centers of Disease Control/American College of Sports Medicine national exercise prescription guidelines of achieving at least 150 minutes of weekly strenuous/moderate exercise minutes.26 Specifically, 42%, 38%, and 41% of participants, respectively, reported meeting these guidelines prior to diagnosis, during active treatment, and after the completion of treatment. Although the number of participants meeting these guidelines is slightly lower than that reported for the general population (45%),28 these numbers are substantially higher than those reported for other cancer populations across similar cancer-related time periods. For example, in a series of studies examining the exercise patterns of patients with multiple myeloma,29 breast cancer,30, 31 colorectal cancer,32 non-Hodgkin lymphoma,33 and endometrial cancer,34 researchers have reported that 5% to 16% and 20% to 31% of participants, respectively, met national exercise guidelines during active treatment and off-treatment periods.
Although the reasons for these contrasting findings are not clear, differences in the medical treatments and demographic profiles of brain tumor patients in comparison with the other cancer populations sampled may partially explain this incongruence. The mean age of the participants in the current study was 44 years, compared with ages 60 years to 65 years in our prior investigations.29, 33, 34 Given that advancing age is associated with lower levels of exercise, the higher percentage of brain tumor patients meeting national exercise guidelines in the current study is not surprising. In addition, conventional chemotherapeutic agents for brain cancer, such as temozolomide, are well tolerated compared with the toxicity profiles of the cytoxic therapies commonly used to treat the aforementioned malignancies (e.g., doxorubicin, paclitaxel, and platinum-containing regimens). As such, brain tumor patients may simply be more capable of exercising during adjuvant therapy.
In concordance with previous investigations in cancer patients, participants in the current study reported lower levels of exercise during active treatments than before their diagnosis of brain cancer. In addition, participants also reported higher levels of exercise during off-treatment periods than during adjuvant therapy, a finding that also is consistent with prior reports. Given the aggressive nature of adjuvant therapy in any cancer population, decreases in exercise behavior during this time are not surprising. However, although the trend in exercise behavior patterns appears to be similar across different cancer populations, the magnitude and changes in the intensity at which exercise is performed is distinct. Specifically, prior surveys in patients with multiple myeloma,29 endometrial cancer,34 and non-Hodgkin lymphoma33 have reported total exercise minute decreases of approximately 100 minutes per week from before diagnosis to during adjuvant therapy compared with only 6 minutes per week noted in the current study over the same time period. Interestingly, although prior investigations have reported that there is a substantial decrease in participation in all exercise intensities (i.e., strenuous, moderate, and mild) for cancer patients, brain tumor patients appear to have decreased participation exclusively in strenuous intensity exercise (Table 2). However, this decrease is partially compensated by a small increase in moderate intensity exercise, resulting in a comparably higher proportion of patients still achieving national exercise prescription guideline requirements (i.e., at least 150 mins/wk of moderate-to-strenuous intensity exercise).
Numerous studies have demonstrated that undergoing cancer treatment has a significantly negative impact on a patient's exercise levels.29, 30, 33 However, these levels tend to recover during off-treatment periods but still remain significantly lower than before the diagnosis. As previously stated, adjuvant therapy does appear to negatively impact exercise levels in patients with brain tumors. This population appears to be unique because brain tumor patients were able to fully recover their exercise habits to prediagnosis levels. In fact, total, moderate, and mild exercise minutes and the percentage of patients meeting national exercise prescription guidelines remained essentially unchanged between the prediagnosis and off-treatment periods, whereas the percentage of participants reporting no physical exercise decreased 10% (from 41% to 31%). In other words, participants actually reported more exercise after the completion of adjuvant therapy than before their cancer diagnosis. The one caveat is that participation in strenuous intensity exercise failed to recover fully to prediagnosis levels.
A secondary purpose of the current study was to explore differences in exercise behavior across the cancer trajectory based on several medical and demographic variables. Accordingly, although the exercise patterns of patients with brain tumors appear to be similar across the selected medical characteristics, there did appear to be statistically significant different patterns based on the patient's age and gender. Specifically, older (age ≥45 yrs) and female participants generally increased their exercise levels during active and off-treatment periods compared with before their diagnosis, whereas younger and male participants tended to decrease their exercise levels over these periods (Tables 3 and 4). It is not clear at the current time why these differences exist. Prior studies generally have reported a lack of association between medical/demographic variables and exercise behavior patterns in cancer patients.29, 33, 34 Clearly, larger studies are required to explore the potential interaction between demographic/medical variables and exercise behavior patterns further in this population.
From an applied perspective, the results of the current investigation may have several important implications for exercise professionals who are designing exercise rehabilitative programs and exercise scientists who are conducting clinical exercise trials in cancer patients. First, the results of the current study suggest that a relatively high percentage of brain tumor patients may be capable of exercising at the currently recommended exercise guidelines both during and after the cessation of adjuvant therapy. Second, despite the lack of evidence, brain tumor patients appear to represent a ‘proactive’ group of patients who are motivated to engage in exercise behavior practices after their diagnosis. Third, although a relatively high percentage of patients adhered to recommended exercise guidelines, 60% still were not meeting these minimum requirements and greater than 30% reported no exercise behavior. Collectively, these findings suggest that future investigations examining the potential efficacy of exercise training interventions on clinically meaningful outcomes in this population are appropriate and timely. Finally, male and younger patients appear to be at the greatest risk of reducing their exercise levels across the brain cancer trajectory. As such, different behavioral and clinical approaches may be required to encourage patients within these subgroups to maintain their exercise behavior practices during the cancer experience.
Although the current study has several strengths, including a well-validated measure of exercise behavior and the abstraction of accurate medical chart data, there are several limitations that need to be considered when interpreting the results and planning future research. One obvious limitation is that selection biases are likely to exist because of the transparent purpose of the study. In addition, the relatively low response rate (28%) and the exclusion of patients with a KPS of < 70 may limit the generalizability of our findings. It is not uncommon to have such biases when conducting exercise surveys. Consequently, patients with brain tumors who were more interested in exercise and had less advanced disease were probably more likely to participate in the study. Nevertheless, the response rate noted in the current study is respectable given that a high percentage of brain tumor patients diagnosed in 2003 were likely deceased when this study was conducted. Moreover, our response rate also compares favorably with previous cancer patient surveys conducted in this population,35 and our mailed survey represented approximately 72% of brain tumor patients seen at the BTC at DUMC between 2003 and 2005. Second, the reliance on self-report rather than objective measures of exercise behavior (e.g., pedometers) may be imprecise. Future studies should strive to use objective measures to verify exercise levels. Finally, the current study used a retrospective observational design, which provides the weakest evidence in terms of causality. Future research that applies prospective experimental designs is warranted.
To our knowledge, the current study is the first to investigate the potential role of exercise in an institution-based sample of brain tumor patients. Overall, brain tumor patients have unique exercise patterns across the cancer trajectory that may be modified by select demographic variables. Moreover, a comparably high percentage of patients are exercising at recommended levels both during active treatment and after treatment has ended. Finally, this patient population appears motivated to engage in exercise behavior practices shortly after their diagnosis. Information gained from this preliminary study will provide critical information for future studies examining the potential clinical value of exercise training interventions in this population.