Journal of Clinical Oncology, Vol 21, Issue 3 (February), 2003: 463-466 2003 American Society for Clinical Oncology

Effect of Upper Extremity Exercise on Secondary Lymphedema in Breast Cancer Patients: A Pilot Study
Donald C. McKenzie, Andrea L. Kalda
From the Division of Sports Medicine and School of Human Kinetics, University of British Columbia, Vancouver, British Columbia, Canada. Address reprint requests to D.C. McKenzie, MD, PhD, The University of British Columbia, Division of Sports Medicine, 3055 Wesbrook Mall, Vancouver, BC, Canada V6T 1Z3; email: kari@interchange.ubc.ca.

ABSTRACT
Purpose: To examine the effect of a progressive upper-body exercise program on lymphedema secondary to breast cancer treatment. Methods: Fourteen breast cancer survivors with unilateral upper extremity lymphedema were randomly assigned to an exercise (n = 7) or control group (n = 7). The exercise group followed a progressive, 8-week upper-body exercise program consisting of resistance training plus aerobic exercise using a Monark Rehab Trainer arm ergometer. Lymphedema was assessed by arm circumference and measurement of arm volume by water displacement. Patients were evaluated on five occasions over the experimental period. The Medical Outcomes Trust Short-Form 36 Survey was used to measure quality of life before and after the intervention. Significance was set at alpha 0.01. Results: No changes were found in arm circumference or arm volume as a result of the exercise program. Three of the quality-of-life domains showed trends toward increases in the exercise group: physical functioning (P = .050), general health (P = .048), and vitality (P = .023). Mental health increased, although not significantly, for all subjects (P = .019). Arm volume measured by water displacement was correlated with calculated arm volume (r = .973, P < .001), although the exercise and control group means were significantly different (P < .001). Conclusions: Participation in an upper-body exercise program caused no changes in arm circumference or arm volume in women with lymphedema after breast cancer, and they may have experienced an increase in quality of life. Additional studies should be done in this area to determine the optimum training program.

INTRODUCTION

LYMPHEDEMA SECONDARY to breast cancer treatment remains a common problem in this patient population. Contributing factors include axillary dissection, radiotherapy to the breast, radiotherapy to the axilla, pathological nodal status, obesity, and tumor stage.1 Estimates of the incidence of lymphedema are varied, but, in the United Kingdom, a large-scale prevalence study found that 28% of women who were treated for breast cancer and were still surviving had lymphedema.2 This is similar to other recent studies, with reported incidence of 25.5%3 and 24%.4 Recent data also indicate that patients have close to a 50% probability of developing lymphedema over a 20-year period after their treatment.5 Many treatment options for lymphedema are available, but none offer a permanent reduction or elimination of arm swelling.6 Upper extremity exercise is rarely mentioned in the rehabilitation of breast cancer survivors and seems to be contraindicated in the clinical lore, especially for women who have developed lymphedema after treatment for this disease. This recommendation cannot be substantiated in the literature. In fact, Harris and Niesen-Vertommen7 followed 20 women living with breast cancer who participated in an upper-body exercise program followed by repetitive training and racing in a dragon boat. Measurements of arm circumference were made at the beginning of the training program, at the start of the racing season, and at the end of the season. None of the women showed a clinically significant difference in arm circumference. The effect of upper-body exercise on patients with lymphedema has not been investigated systematically. This study examined the changes in arm circumference, arm volume, and quality of life in women with secondary lymphedema resulting from breast cancer treatment throughout an 8-week upper-body exercise program.

METHODS
Subjects Fourteen subjects were randomly assigned to either a treatment group (n = 7) or a control group (n = 7). Before testing, informed consent was obtained from each subject and the study was approved by the Clinical Screening Committee for Research and Other Studies Involving Human Subjects of the University of British Columbia. Subjects were eligible for the study if they had undergone breast cancer treatment for stage I or II breast cancer that had been completed more than 6 months before enrolling in the study and had subsequently developed unilateral lymphedema that was greater than 2 cm and less than 8 cm on at least one measurement point8. They were excluded if they had stage III lymphedema, bilateral disease, or if they required medication that might affect upper extremity swelling. Testing Procedures All subjects were tested every 2 weeks for 8 weeks, beginning with the baseline measurement, producing a total of five measurements. Height and weight were measured to 0.1 cm and 0.1 kg, respectively. For the arm circumference measurements, subjects lay prone, arms relaxed by their sides and elbows straight. Both arms were measured at each test date. Circumference was measured every 3 cm beginning at the styloid process of the ulna and continuing 45 cm proximally, as well as at the metacarpals and midhand. The measuring tape was placed around the extremity so that there was no slack but also so there was no indentation in the tissue. The

volume of each arm was calculated from the arm circumference measurements using the following formula: volume = (circumference/2)2h, where circumference equals the mean of the two bounding circumferences, and height (h) = 30 mm (the length of each segment between circumference measurements). The sum of all of the segments from the styloid process of the ulna to 45 cm proximal for each arm equals the total volume of the arm, excluding the hand.9 Upper extremity volume was measured by water displacement. Two volume measurements were taken for each upper extremity; the first to the ulnar styloid process only (hand), the second to 45 cm proximal to the ulnar styloid process (hand, forearm, and arm). The arm was kept straight and was immersed slowly into the water, sliding the fingers straight down the inside wall of the volumeter. Water was collected from the instant the arm was first immersed until the water was dripping less than once per second. Water was displaced into a graduated cylinder, and the volume was read to the nearest 5 mL. The Medical Outcomes Trust 36-Item Short Form Survey (SF-36) general quality-of-life questionnaire was administered at the first and last measurement dates. This scale has been validated and reliably detects quality-of-life deficits in general medical patients. This form measures eight domains: physical functioning, role of physical functioning, bodily pain, general health, vitality, social functioning, role of emotional functioning, and mental health. The completed questionnaires were scored according to the SF-36 Health Survey Manual and Interpretation Guide.10 Exercise Program Experimental subjects completed an 8-week exercise program. A professionally fitted compression sleeve was used by each subject for all exercise sessions. In addition, these sleeves were used daily by all subjects in both the control and exercise groups. A resistance-training program was initiated immediately after the baseline test and continued three times per week for the duration of the experimental period. Stretching exercises were prescribed for each major body part. Resistance training included specific exercises, beginning with a light weight and progressing as tolerated by each subject. The strength exercises prescribed were the seated row, bench press, latissimus dorsi pull down, one arm bent-over rowing, tricep extension, and bicep curl. Two sets of 10 repetitions for each exercise were done for the first week, three sets of 10 were done thereafter. The training sessions consisted of a 5- to 7-minute period of aerobic warmup such as cycling or walking, 5 minutes of stretching, the strength training program, and a cooldown period. After 2 weeks, upper-body aerobic exercise, using an arm cycle ergometer (model 881 Monark Rehab Trainer, Monark Exercise AB, Varberg, Sweden), was added to the program. Subjects exercised under supervision. After a program that began with five 1-minute bouts of cycling at a resistance of 8.3 W, the program progressed to 20 minutes of continuous cycling with a resistance of up to 25 W. This program was used as a guide and was adjusted according to subjects self-report of fatigue or discomfort in their affected arm. Work in kilojoules was calculated for each session for every subject, and this was used to calculate cumulative work done over the course of the program.

Control subjects were given no specific exercise instruction until after they completed the study, at which time they had the option of being taught the exercise program. Statistical Analyses Repeated-measures analysis of variance (ANOVA) was done using SPSS software (SPSS Inc., Chicago, IL). Significance was set at P < .01 for all comparisons to compensate for the number of tests being done on a small sample. Percentage difference in arm circumference, percentage difference in measured arm volume, and percentage difference in calculated arm volume were tested separately. Pearsons product-moment coefficients of correlation were calculated to examine the relationships between measurement techniques, and paired-samples t tests were used to determine whether the means were significantly different. The eight SF-36 domains were also tested using separate repeated-measures ANOVA. The difference between domain scores at time 1 and time 2 was then calculated for those domains that showed a significant difference in the ANOVA, and these difference scores were then correlated with the difference in both measured and calculated volumes (time 2 - time 1). Pearsons product-moment correlation coefficients were determined for the four domains with P .05 from the repeated-measures ANOVA to assess the viability of these results.

RESULTS
Group characteristics at baseline are shown in Table 1. There were no significant differences found between groups at baseline with respect to any of these descriptive variables. It is interesting to note that nine of the 14 subjects would be classified as overweight or obese according to their body mass index value. View this table: In this window In a new window Table 1. Group Characteristics at Baseline Volume Changes There were no significant differences in the percentage change of measured arm volume found by ANOVA. The changes in arm volume as determined by water displacement are shown in Fig 1. Figure 2 illustrates the changes in arm volume calculated from the circumference measurements. There was no significant change in either the control or exercise groups over time.

View larger version (14K):

In this window In a new window Fig 1. Response of arm volume measured by water displacement.

View larger version (13K): In this window In a new window Fig 2. Response of arm volume calculated from arm circumference. Comparison of Volume Measurement Techniques Pearsons product-moment coefficients of correlation were calculated to determine whether the two methods of measuring volume were correlated, and paired-samples t tests were used to determine whether the means were significantly different. The highest correlation was found between measured arm volume of the affected arm, excluding the hand, and calculated arm volume of the affected arm (N = 140, r = .97, P < .001), although the t test showed that the means of the two techniques were significantly different (P < .001). There was also a correlation between measured arm volume of the affected arm, including the hand, and calculated arm volume of the affected arm (N = 140, r = .97, P < .001), but again, the means were found to be significantly different (P < .001). Percentage difference of arm volume calculated from the two measurement techniques were also correlated (N = 70, r = .94, P < .001), although the means were significantly different (P = .001). Quality of Life P values .05 were found for four of the SF-36 domains. Physical functioning (P = .050), general health (P = .048), and vitality (P = .023) increased in the exercise group and decreased in the control group, although the changes were not significantly different between groups across time. Mental health increased over time in all subjects (P = .019), although, again, this was not statistically significant. There was a trend indicating that as percentage difference of calculated volume decreased, the general health domain increased (r = -.53, P = .052). There was a decrease, although it was not statistically apparent, in the bodily pain scores of both groups. The changes over time of the other three domains (role physical, social functioning, and role emotional) were not statistically significant.

DISCUSSION
The primary finding of this pilot study is that this progressive, controlled upper-body exercise program does not significantly affect the volume of the upper extremities in women with lymphedema after breast cancer treatment. This finding is valid independent of the method used to quantify limb volume. The use of this specific exercise program challenges the myth that

strenuous upper extremity exercise may lead to, or worsen, lymphedema. One significant feature is that this study used women who already had developed lymphedema after breast cancer therapy. In fact, it was our hope that exercise would act as a treatment modality rather than a causative factor for this condition. There are several reasons to suggest the use of a gradual, progressive upper-body exercise program in the rehabilitation of lymphedema.11 Lymph is propelled by both passive and active forces. Passive forces are already promoted as treatment for lymphedema: manual lymphatic drainage, massage therapy, sequential pneumatic compression pumping, elastic compression sleeves, and limb elevation. These treatments mimic the passive forces of the body, such as skeletal muscle pumping, respiratory movement, and arterial pulsation. Exercise also stimulates the skeletal muscle to pump venous and lymphatic fluid.12 This type of exercise should also stimulate the contraction of the lymph vessels themselves because these vessels are innervated by the sympathetic nervous system. Regaining control over these internal contractions by resetting the sympathetic drive to these vessels through upper-body exercise may assist in the long-term treatment for lymphedema. The lack of change in arm volume after the exercise intervention may have been the result of small sample size or possibly because the exercise program itself was not long or intense enough to result in statistically significant changes. Changes in arm composition were controlled for by using the subjects normal arm as a control for their affected arm, assuming that changes induced by exercise would be occurring in both arms equally. This method has been used previously.6,13 However, it is impossible with these measurement techniques to know what tissue changes occurred within the upper extremity. Regular exercise is associated with muscle hypertrophy and a loss of adipose tissue. Although arm circumference did not change, we do not know if the level of lymphedema was influenced by the exercise; other tissue changes may have masked any positive effects on lymph accumulation. The data show a trend toward increases in self-reported physical functioning, general health, and vitality with participation in an upper-body aerobic exercise program. Mental health scores increased for both groups, although not significantly, suggesting that this change may be a result of being a participant in a study. The trend toward increases in physical functioning scores in the exercise group supports the theory that upper-body aerobic exercise is beneficial for women with secondary lymphedema after breast cancer treatment. Exercise subjects expressed feeling more confident using their affected arm for activities of daily living, and some mentioned that they had returned to lifting objects or carrying groceries with that arm. This may explain the trend toward increased general health and vitality scores, inasmuch as the exercise subjects were reminded less often of their disease and therefore felt healthier overall. The decrease in physical functioning scores in the control group perhaps represents the normal progress of the disease, with the arm gradually becoming more congested over time. It appears that upper-body aerobic exercise may result in an improvement in physical functioning.

Anecdotally, subjects felt that the exercise program improved the health of their affected arm in ways that were not measurable with our techniques. They reported softening of hardened areas, reduced pain and swelling when using their arm for activities of daily living, and reappearance of hand tendons. Many subjects continued the exercise program independently after they completed the study. Obesity and arm dominance have been shown to confound circumference and volume measurements of the arms.13 They can mask the severity of the lymphedema by either enlarging the normal arm relative to the affected arm, or enlarging both arms and thereby lowering the proportion of the affected arm to the normal arm. This can be problematic for diagnosing lymphedema, and may have affected our results by lowering the percentage ratios of arm circumference and arm volume. However, because the primary question of this study involved change in arm swelling caused by an intervention and not categorization of lymphedema according to severity, this was most likely not a major limitation. Lymphedema can be viewed as primarily a quality-of-life issue for the women whom it affects. Difficulties functioning at work or at home, altered body image, low self-esteem, problems with dress, and a loss of interest in social activities have all been shown to result from living with lymphedema.14 Many of the difficulties performing activities of daily living, such as lifting or carrying, cleaning, and reaching up or down, may be due to a lack of strength and flexibility in the affected arm, as well as a lack of confidence. This 8-week upper-body aerobic exercise program does not affect arm swelling significantly. However, the trend toward improvement in quality of life is an important reason to consider it as therapy for secondary lymphedema resulting from breast cancer treatment.

NOTES
Supported by the Canadian Breast Cancer Research Initiative.

REFERENCES
1. Kissin MW, Rovere GQ, Easton D, et al: Risk of lymphoedema following the treatment of breast cancer. Br J Surg 73:580584, 1986[Medline] 2. Mortimer PS, Bates DO, Brassington HD, et al: The prevalence of arm oedema following treatment for breast cancer. Quart J Med 89:377380, 1996[Abstract/Free Full Text] 3. Tobin MB, Lacey HJ, Meyer L, et al: The psychological morbidity of breast cancer-related arm swelling. Cancer 72:32483252, 1993[CrossRef][Medline] 4. Maunsell E, Brisson J, Deschenes L: Arm problems and psychological distress after surgery for breast cancer. Can J Surg 36:315320, 1993[Medline] 5. Petrek JA, Senie RT, Peters M, et al: Lymphedema in a cohort of breast carcinoma survivors 20 years after diagnosis. Cancer 92:13681377, 2001[CrossRef][Medline]

6. Mirolo BR, Bunce IH, Chapman M, et al: Psychosocial benefits of postmastectomy lymphedema therapy. Cancer Nurs 18:197205, 1995[Medline] 7. Harris SR, Niesen-Vertommen SL: Challenging the myth of exercise-induced lymphedema following breast cancer: A series of case reports. J Surg Oncol 74:9599, 2000[CrossRef] [Medline] 8. Cluzan RV, Alliot F, Ghabboun S, et al: Treatment of secondary lymphedema of the upper limb with CYCLO 3 FORT. Lymphology 29:2935, 1996[Medline] 9. Farncombe M, Daniels G, Cross L: Lymphedema: the seemingly forgotten complication. J Pain Symptom Manage 9:269276, 1994[CrossRef][Medline] 10. Ware JE, Snow KK, Kosinski MK, et al: SF-36 Health Survey Manual and Interpretation Guide. Boston, MA: Nimrod Press, 1993 11. McKenzie DC: Abreast in a boata race against breast cancer. Can Med Assoc J 159:376 378, 1998[Medline] 12. Witte CL, Witte MH: Contrasting patterns of lymphatic and blood circulatory disorders. Lymphology 20:171178, 1987[Medline] 13. Bunce IH, Mirolo BR, Hennessy JM, et al: Post-mastectomy lymphoedema treatment and measurement. Med J Aus 161:125128, 1994 14. Woods M, Tobin M, Mortimer P: The psychosocial morbidity of breast cancer with lymphoedema. Cancer Nurs 18:467471, 1995[Medline] Submitted April 8, 2002; accepted October 7, 2002.