Australian Physiotherapy Journal

Journal of Physiotherapy 60 (2014) 144–150 Journal of PHYSIOTHERAPY journal homepage: www.elsevier.com/locate/jphys Research Several submaximal exercise tests are reliable, valid and acceptable in people with chronic pain, fibromyalgia or chronic fatigue: a systematic review Julia Ratter a, Lorenz Radlinger b, Cees Lucas c a Hospital Rivierenland Tiel, The Netherlands; b Applied Research and Development Physiotherapy, Health Division, Bern University of Applied Sciences, Switzerland; c Department of Clinical Epidemiology, Biostatistics and Bioinformatics, Medical Faculty, University of Amsterdam, Academic Medical Centre, Amsterdam, The Netherlands KEYWORDS ABSTRACT Chronic pain Question: Are submaximal and maximal exercise tests reliable, valid and acceptable in people with Fatigue syndrome (chronic) chronic pain, fibromyalgia and fatigue disorders? Design: Systematic review of studies of the Fibromyalgia psychometric properties of exercise tests. Participants: People older than 18 years with chronic pain, Exercise test fibromyalgia and chronic fatigue disorders. Intervention: Studies of the measurement properties of tests Psychometrics of physical capacity in people with chronic pain, fibromyalgia or chronic fatigue disorders were included. Review (publication type) Outcome measures: Studies were required to report: reliability coefficients (intraclass correlation coefficient, alpha reliability coefficient, limits of agreements and Bland-Altman plots); validity coefficients (intraclass correlation coefficient, Spearman’s correlation, Kendal T coefficient, Pearson’s correlation); or dropout rates. Results: Fourteen studies were eligible: none had low risk of bias, 10 had unclear risk of bias and four had high risk of bias. The included studies evaluated: A˚ strand test; modified A˚ strand test; Lean body mass-based A˚ strand test; submaximal bicycle ergometer test following another protocol other than A˚ strand test; 2-km walk test; 5-minute, 6-minute and 10-minute walk tests; shuttle walk test; and modified symptom-limited Bruce treadmill test. None of the studies assessed maximal exercise tests. Where they had been tested, reliability and validity were generally high. Dropout rates were generally acceptable. The 2-km walk test was not recommended in fibromyalgia. Conclusion: Moderate evidence was found for reliability, validity and acceptability of submaximal exercise tests in patients with chronic pain, fibromyalgia or chronic fatigue. There is no evidence about maximal exercise tests in patients with chronic pain, fibromyalgia and chronic fatigue. [Ratter J, Radlinger L, Lucas C (2014) Several submaximal exercise tests are reliable, valid and acceptable in people with chronic pain, fibromyalgia or chronic fatigue: a systematic review. Journal of Physiotherapy 60: 144–150] ß 2014 Published by Elsevier B.V. on behalf of Australian Physiotherapy Association. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Introduction A challenging diagnostic dilemma with regard to the above diagnoses is overlap of symptoms. Chronic widespread pain, the Functional disorders are illnesses in which there is no obvious pathology or anatomical change in an organ, and there is a cardinal symptom of fibromyalgia, is prevalent and co-occurs with presumed dysfunction of an organ or system. Chronic pain, numerous symptom-based conditions such as chronic fatigue fibromyalgia and chronic fatigue disorders are often-mentioned syndrome, joint pain and psychiatric disorders.5 Estimates of the diagnoses belonging to functional disorders.1 Chronic pain is number of patients with fibromyalgia who meet the criteria for defined as pain that has lasted longer than 3 to 6 months,2 although chronic fatigue disorders range from 30 to 70%.4 Fibromyalgia some use 12 months as the threshold.3 A popular alternative syndrome and chronic fatigue syndrome are similar in many ways definition of chronic pain, involving no arbitrarily fixed durations is ‘pain that extends beyond the expected period of healing’.2 – both conditions lack an accepted disease model that can explain Fibromyalgia is a chronic functional illness that presents with signs and symptoms in terms of specific pathophysiological widespread musculoskeletal pain, including above and below the abnormalities.6 waist, as well as the right and left sides of the body, and the physical finding of 11 of 18 tender points. These simple criteria In Europe, 19% of adults experience chronic pain of moderate to provide 85% specificity and sensitivity in differentiating patients with fibromyalgia from those with other rheumatic diseases.4 severe intensity with serious negative implications for their social Chronic fatigue is defined as persistent or relapsing fatigue lasting and working lives.7 Fatigue is also a common symptom in the more than 6 months, with more than four of the following symptoms: impaired memory, sore throat, tender cervical or community, affecting from 0.007 to 2.8% in the general adult axillary lymph nodes, muscle pain, multifocal joint pain, new population and from 0.006 to 3.0% in primary care.8 Fibromyalgia headaches, unrefreshing sleep, and post-exertion malaise.4 syndrome affects 2 to 4% of the general population, and over 5% of patients in general medical practice.9 Recent studies have confirmed previous evidence of the enormous indirect socioeconomic costs of chronic pain, fibromy- algia and chronic fatigue disorders. The overall financial costs of chronic pain to society are comparable with the costs of cancer or http://dx.doi.org/10.1016/j.jphys.2014.06.011 1836-9553/ß 2014 Published by Elsevier B.V. on behalf of Australian Physiotherapy Association. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/3.0/).

Research 145 cardiovascular diseases.10,11 Chronic pain is also associated with to select the best instrument. Therefore, the research question of this systematic review was: many secondary stressors such as sleep disruption, unemployment and interpersonal tensions.12 Chronic fatigue syndrome is char- In people with chronic pain, fibromyalgia and fatigue disorders, are maximal and submaximal physical capacity tests reliable, valid acterised by profound disabling fatigue lasting at least 6 months and acceptable? and accompanied by numerous symptoms such as pain, sleep Method difficulties and cognitive impairment.13 Chronic pain, fibromyalgia Identification and selection of trials and chronic fatigue also have personal economic, psychological and social consequences for the affected individuals.12,14,15 One in A sensitive search was performed in PubMed, Embase, PEDro and the Cochrane library in October 2012. The search strategy was three people with pain or fatigue disorders is unable or less able to developed by a medical librarian specialist. The detailed strategy maintain an independent lifestyle11 and 50 to 66% of people for PubMed is presented in Appendix 1 (see eAddenda). Eligible studies could use any study design that reported on one or more suffering from chronic pain are less able or unable to exercise, measurement properties of physical capacity tests in adults with chronic pain, chronic fatigue disorders or fibromyalgia. Data were enjoy normal sleep, perform household chores, attend social extracted for reliability coefficients, validity coefficients and activities, drive a car, walk or have sexual relations.16 dropout rates. Studies published in any language and in any year were eligible for inclusion. Although key risk factors have been identified, the incidence of Records retrieved by the search were assessed for eligibility by chronic pain, fibromyalgia and fatigue disorders has been two reviewers (JR, LR) working independently, initially based on increasing, rendering their management a persistent challenge.14 titles and abstracts, with potentially eligible articles being assessed in full-text to confirm eligibility. Discrepancies were reviewed and Fear avoidance models emphasise psychological distress, pain- consensus was achieved by discussion. Reasons for exclusion were given for each reference and are documented in Figure 1. For each related anxiety, anxiety sensitivity, fear of illness/injury, fear of re- included study, the exercise tests assessed were tabulated along with the psychometric tests performed and their results. injury and catastrophising in the development and maintenance of disabling chronic pain.17 International and national guidelines Assessment of characteristics of trials recommend graded activity and graded exposure in the treatment Quality of chronic disorders.15,18–21 The COSMIN 4-point rating scale (excellent, good, fair, poor) The validity of self-reported assessment of pain and physical was used to evaluate elements of the methodological quality disability is controversial. The level of pain reported by people with rug[i(Fe]GI$DT_1) chronic pain is not always related to their reports of their physical Records identified through Records identified database searches (n = 3496) through other disability. Nevertheless, pain, fear of pain and its consequences are • PubMed (n = 2046) sources (n = 0) subjective experiences and are difficult to assess.22 Observational • Embase (n = 1051) • Cochrane (n = 257) measures may be useful to corroborate subjective reports when • PEDro (n = 142) evaluating each person’s capability.23,24 Ideally, evaluation of Records after duplicates removed (n = 2637) physical function in people with chronic pain and chronic fatigue Records screened (n = 2637) disorders should rely on a combination of clinical assessment of Records excluded after screening impairments, behavioural observation of physical function, and titles and abstracts (n = 2563) self-report.25 Despite this, there is limited evidence about the Potentially relevant articles retrieved acceptability, reliability and validity of submaximal and maximal in full text for further evaluation of eligibility (n = 74) exercise tests measuring physical fitness and capacity in this group Articles excluded after evaluation of people. To assess aerobic capacity, maximal testing with of full text (n = 60) calorimetry is considered to be the gold standard.26,27 However, • ineligible assessments (n = 32) • no psychometric data (n = 28) outcomes of this measurement are strongly influenced by motivation, fear and pain.26 Furthermore, outcomes are invalid Studies included in the review (n = 14) when fear and pain expectation rather than aerobic capacity limit Figure 1. Flow of studies through the review. performance.28 In one study, over 90% of the variance in performance among disabled individuals with chronic musculo- skeletal pain was predicted by psychosocial factors like self-efficacy, perceived emotional and physical functioning, pain intensity and pain cognition.29 Several studies of people with chronic pain have identified discrepancies between self-report of physical activity and actual level of physical activity. Poorer achievement on physical performance testing by people with low back pain has been linked to fear of injury during movement, depression, cognitive factors, pain expectations, pain increase during testing, disability status and the presence of a solicitous spouse.23 The conventional A˚ strand bicycle test and maximal exercise capacity tests tend to be unacceptable in people with a very poor aerobic capacity30 and the validity is low in those with chronic low back pain.27 Also, physical assessments used to detect the degree of disability in other disease states have major limitations when applied to people with fibromyalgia and chronic fatigue syn- drome.31 In the last decade, many submaximal tests have been developed as an alternative to maximal exercise testing.28 The most commonly used test in people with chronic low back pain is the submaximal A˚ strand bicycle test. Its test-retest reliability seems to be good in people with chronic low back pain.32 However, submaximal testing tends to underestimate or overestimate maximal oxygen consump- tion (VO2max) in 15% of healthy subjects.33 Nevertheless, due to pain, fatigue and fear of worsening their symptoms, people with chronic pain, fibromyalgia and fatigue disorders are often unable to perform the submaximal A˚ strand bicycle test.34,35 Guidance for clinicians in this area is needed because the variety in attributes of the available instruments makes it difficult

146 Ratter et al: Psychometrics of exercise tests in chronic pain patients (measurement error, reliability, criterion validity) of the studies. A un-blinded assessment, uncertainty about other measurement methodological quality score for each relevant element was errors and absence of gold standards. obtained by taking the lowest rating of any item for that element (‘worse score counts’).36 Two authors (JR, LR) independently Participants assessed the risk of bias in included studies, with consensus Sample sizes in the included studies ranged from 24 to 683. The achieved by discussion. mean age of all participants was 45 years, with mean age in the Participants individual studies ranging from 34 to 82 years. Age, diagnosis and Studies involving adults (ie, aged 18 years or older) with chronic number of participants in individual studies are presented in Table 1. pain, fibromyalgia or chronic fatigue disorders were eligible. Exercise tests Exercise tests The exercise tests listed above were all assessed by one study Studies were required to have assessed the psychometric each, except for the conventional A˚ strand test (three studies), the properties of any of the following submaximal exercise tests to be 5-minute walk test (three studies), and a submaximal bicycle eligible: A˚ strand test; modified A˚ strand test; Lean body mass- ergometer test following a protocol other than the A˚ strand test based A˚ strand test; submaximal bicycle ergometer test following a (three studies). protocol other than the A˚ strand test; 2-km walk test; shuttle walk test; modified symptom-limited Bruce treadmill test; and walking Are maximal and submaximal physical capacity tests in people distance over 5, 6 or 10 minutes. with chronic pain, fibromyalgia and fatigue disorders reliable, valid and acceptable? Psychometric outcomes Data were extracted, where available, for the following No data regarding maximal exercise tests in the population of interest were identified. The data extracted from studies of reliability coefficients: intra class correlation (ICC), alpha reliability submaximal tests are presented in Table 1. The psychometric coefficient, limits of agreements, and Bland-Altman plots. Data properties of each submaximal test are summarised descriptively, were also extracted for the validity coefficients: ICC, Spearman’s below. correlation, Kendal T coefficient, and Pearson’s correlation. Dropout rates were also recorded. A˚ strand test Four studies evaluated the reliability, concurrent validity and Data analysis dropout rates of the A˚ strand test, the modified A˚ strand test or the The following data were extracted from each eligible study and Lean body mass-based A˚ strand test. Based on 19 participants, tabulated: study design, participants (sample size, age, diagnosis), Hodselmans et al reported the test-retest reliability of the Lean aim, exercise test, psychometric outcomes and methodological body mass-based A˚ strand test as an ICC of 0.91 (95% CI 0.76 to quality. Data for individual studies were reported quantitatively 0.97), which changed to 0.96 (95% CI 0.91 to 0.99) when one outlier and the evidence was also summarised qualitatively. No meta- was excluded.30 The limits of agreement for the Lean body mass- analyses were performed because of heterogeneity among the based A˚ strand test were 32.0 and 32.8% including the outlier, and study designs used, heterogeneity of the psychometric properties 13.8 and 16.9% excluding the outlier. Assessing the conventional evaluated and incomplete reporting of the data. A˚ strand test in 31 participants, Keller et al showed a test-retest reliability ICC of 0.96 and a critical difference of 21%.32 Based on The evidence was graded, based on the number of studies, their these studies, test-retest reliability seems to be excellent. methodological quality, and the consistency of the available evidence into five categories: strong (consistent findings in two or Smeets and van Soest evaluated the concurrent validity of the more high-quality studies); moderate (consistent findings in one A˚ strand test with a modified A˚ strand test in 31 participants with high-quality and one low-quality study, or in two or more low- musculoskeletal pain disorder.35 They reported an intraclass quality studies); limited (only one study); conflicting (inconsistent coefficient of 0.79 between the two tests. The limits of agreement findings); and no evidence (no studies). The authors considered for VO2max were 15.9% from the mean difference, which equated findings to be consistent if at least 75% of the available studies to 8.5 ml/kg of lean body mass per minute in VO2max. Viitanen reported the same conclusion37. evaluated the concurrent validity of the A˚ strand test with a modified A˚ strand test and a 2-km walk test in 69 participants.39 Results The ICC was 0.20 (95% CI –0.29 to 0.50) at entry of the study and 0.47 (95% CI 0.15 to 0.67) after 3 months. In addition, Spearman’s Flow of trials through the review rank correlation between these two tests was low: r = 0.37 (p < 0.01) at entry and r = 0.34 (p < 0.01) after 3 months. These The search yielded 3496 records, which amounted to 2637 tests showed low and non-significant correlations with the visual potentially relevant articles after removal of duplicates. After initial analog scale for pain, with r-values ranging from 0.11 to –0.19 for screening, 74 of these articles were obtained in full text for further the A˚ strand test and 0.09 to –0.22 for the 2-km walk test. assessment. The final selection included 14 studies involving 1275 participants. The selection procedure and the reasons for exclusion Smeets and van Soest described a slight underestimation of are presented in Figure 1. Inter-rater agreement about the eligibility VO2max with the modified A˚ strand test,35 with VO2max outcomes of studies was assessed by using an unweighted Kappa. Unweighted an average of 9.96% higher when the conventional A˚ strand test was Kappa for the selection of abstracts was k = 0.91, unweighted Kappa used (95% CI 6.4 to 13.5%) in the pain group. The dropout rates of for the selection of full texts was k = 0.74; this is considered to be the A˚ strand test and the modified A˚ strand test were moderate at 5 excellent inter-rater agreement. to 16%.30,35 The 2-km walk test was not recommended for subjects with chronic pain syndrome, for example fibromyalgia, due to underestimation of exercise capacity.38 Characteristics of the included trials Submaximal bicycle ergometer aerobic capacity test not following A˚ strand protocol Quality The reviewers rated the included studies as being ‘fair’ or ‘poor’ Three of the 14 studies assessed reliability (test-retest reliability) and acceptability (dropout rate) of other submaximal on the COSMIN checklist, as presented in Appendix 2 bicycle ergometer tests. Protocols of these exercise tests are (see eAddenda). Common methodological shortcomings were available from the authors. Test-retest reliability was good in the

Research 147 Table 1 Summary of included studies (n = 14). Study Participants Exercise tests Results Quality Design n = 31 A˚ strand test and modified Validity: ICC = 0.79, against 33 healthy matched Reliability: fair Smeets35 A˚ strand test, repeated at 7 d controls. Criterion validity: fair Cross-over Musculoskeletal pain Critical difference: On B-A plots, the limits of randomised disorders agreement = 15.9% (8.5 ml/kg LBM minÀ1). Slight trial underestimation of VO2max with the modified test Age = 48 yr (SD 8) by 9.96% (95% CI 6.4 to 13.5). Dropout rate: A˚ strand test = 5/31 (16%), modified A˚ strand test: 5/31 (16%). Hodselmans30 n = 20 LBM-based A˚ strand test, Usability: modified A˚ strand = acceptable Reliability: poor repeated at 2 wk Observational Non-specific chronic LBP Test-retest reliability: ICC = 0.91 (95% CI 0.76 to 0.97), study with median duration of 68 or excluding outlier 0.96 (95% CI 0.91 to 0.99). mth (range 8 to 180) Maximal bicycle test, once for Validity: ICC = 0.94 (95% CI 0.85 to 0.98), or excluding validity outlier 0.98 (955 CI 0.95 to 0.99). Age = 34 yr (SD 9) Critical difference: On B-A plots, limits of agreement = 32.0 to 32.8%, or excluding outlier 13.8 to 16.9%. Viitanen38 n = 69 A˚ strand test and 2-km walk Dropout rate: 1/20 (5%) Reliability: fair Fibromyalgia test, repeated at 2 mth Criterion validity: poor Non-randomised Age = 48 yr (SD 7) Validity: ICC = 0.20 (95% CI –0.29 to 0.50) at baseline, intervention 0.47 (95% CI 0.15 to 0.67) at 3 mth. study Correlation between tests: r = 0.37 (p < 0.01) at baseline, r = 0.34 (p < 0.01) at 3 mth. Reliability: alpha coefficients = 0.54 at baseline, 0.47 at 3 mth. Kendal-T coefficients 0.32 and 0.41 for ordinal correlation of the tests. Criterion validity: Results did not correlate with pain VAS, Spearman’s r = –0.19 to 0.11 for A˚ strand test, –0.22 to 0.09 for 2-km walk. Keller32 n = 31 A˚ strand bicycle test, repeated Usability: 2-km walk test not recommended for Reliability: fair Observational twice at 5–10 d intervals subjects with chronic pain syndrome, eg, Unclear study LBP fibromyalgia. Reliability: fair Protas41 Age = 34 yr (range 29 to 43) Test-retest reliability: ICC = 0.96. Non-randomised n = 683 intervention Critical difference: On B-A plots, limits of study Chronic work-related spinal agreement = 21% for LBP. disorders van Santen40 Submaximal bicycle test, Dropout rate before rehabilitation: 20/179 (11%) assessed before and after cervical, 165/504 (33%) lumbar. rehabilitation Dropout rate after rehabilitation: 0/683 (0%). Age = 41 yr (SD 9) (cervical); Bicycle test, repeated at 7 d Reliability: ICC = 0.86. 40 yr (SD 10) (lumbar) n = 37 Randomised Fibromyalgia Bicycle test, repeated at 5 d Reliability: ICC = 0.70. Reliability: fair trial Age = 42 yr (range 20 to 58) n = 30 van Santen39 Randomised Fibromyalgia 5-min walk test, repeated at 5 Test-retest reliability: ICC = 0.89 (95% CI 0.81 to 0.93). Reliability: fair trial to 9 d Age = 45 yr (range 26 to 60) Critical difference: On B-A plots, limits of Smeets42 n = 53 agreement = 20%. Observational Non-specific chronic LBP study without task experience (n = 30) or with task experience (n = 23) Task experience did not significantly influence test- retest differences. Age = 43 yr (SD 9) Usability: 5-min walk test = acceptable

148 Ratter et al: Psychometrics of exercise tests in chronic pain patients Table 1 (Continued ) Participants Exercise tests Results Quality n = 147 Study 10-min walk test, measured Inter-rater reliability: ICC = 0.994 (n = 24). Reliability: poor Design using 2 raters, and repeated Test-retest reliability: Pearson’s r = 0.944 (n = 30). Criterion validity: fair Harding43 at 12 wk Non-randomised Chronic pain Criterion validity: Spearman’s r = 0.985, indicating intervention Age = 50 yr (range 20 to 84) strong correlation with 5-min walk test. study Test-retest reliability: ICC = 0.87. Simmonds45 n = 44 5-min walk test, repeated at Reliability: fair LBP 2 wk Observational Age 43 yr (range 21 to 63) study Criterion validity: Pearson’s r = 0.617 for correlation Criterion validity: poor between 5-min walk and 50-ft walk. Correlations with 7 other physical performance tests were weak. Sato44 n = 82 6-min walk test, repeated at Usefulness = no specialised equipment required, Reliability: fair Chronic pain 2 wk acceptable to all subjects. Criterion validity: poor Observational Age = 82 yr (SD 8) Inter-rater reliability: ICC = 1.000. study Intra-rater reliability: ICC = 0.979. Mannerkorpi46 n = 69 6-min walk test Validity: Spearman’s r = 0.652, indicating significant Criterion validity: fair association with the Functional Independence Measure. Dropout rate: 1/69 (1%). Observational Women with fibromyalgia 6-min walk test showed a fair relationship with the study Age = 45 yr (SD 8) Physical Function domains of the SF-36 and the FIQ, and a moderate-to-good relationship with the ASES function scale. Taylor47 Reliability: Shuttle walk test, repeated Correlations between the performance-based tests Reliability: fair n = 44 within 1 wk and the activity limitations tended to be higher than Observational Chronic LBP those between performance and pain. study Age = 48 yr (SD 14) Reliability: ICC = 0.99. Critical difference: On B-A plots, mean difference = 2.5 m, with limits of agreement of 52 m and –47 m. Responsiveness: Shuttle walk test, repeated at Responsiveness: Effect size = 1.42 for fitness group, n = 111 2 and 25 mth 0.23 for control group, and 0.94 for orthopaedic clinic Age = 35 (SD 9), 40 (SD 9) group. and 46 (SD 13) yr Wittink25 Modified symptom-limited Usability: simple, quick. Criterion validity: fair n = 63 Bruce-treadmill test Observational Validity: Spearman’s r = 0.43, indicating a strong study Chronic LBP correlation with the Physical Function domain of the SF-36. Age = 40 yr (SD 8) ASES = American Shoulder and Elbow Surgeons, B-A = Bland-Altman, FIQ = Fibromyalgia Impact Questionnaire, ICC = Intraclass correlation coefficient, LBM = lean body mass, LBP = low back pain, SF-36 = Short-Form Health Survey, VO2max = maximal aerobic capacity, Wmax = maximal work capacity. studies by van Santen et al, 39,40 with ICCs of 0.70 to 0.86. The r = 0.985.43 Criterion validity of the walk tests was assessed against dropout rates of 0 to 33% among the various tests were considered the 50-foot walk, the Functional Independence Measures (FIM) acceptable.41 scale, various performance-based tests, the Short-Form Health Survey (SF-36), the Fibromyalgia Impact Questionnaire (FIQ), and Walk tests the American Shoulder and Elbow Surgeons (ASES) Function Five studies evaluated the reliability, criterion validity and questionnaire. Simmonds et al45 reported a moderate correlation of the 5-minute walk test with the 50-foot walk, r = 0.617. Sato acceptability of walk tests. Smeets et al42 assessed test-retest et al44 reported a significant correlation of the 6-minute walk test reliability, reporting an ICC of 0.89 (95% CI 0.81 to 0.93). Harding with the Functional Independence Measures scale (r = 0.652, et al43 reported a Pearson’s r of 0.944. Task experience did not p < 0.01), which was used to evaluate activities of daily living. significantly influence test-retest differences.42 Inter-rater reli- Mannerkorpi et al46 correlated the 6-minute walk test against ability was reported as ICCs of 0.994 by Harding et al43 and 1.000 various performance-based tests (chair rising test, hand grip by Sato et al.44 Intra-rater reliability was reported as an ICC of strength, endurance shoulder muscles, abduction, hand to neck, 0.979 by Sato et al44 and day-to-day reliability as an ICC of 0.87 by hand to scapula) but the criterion validity was fair to moderate, Simmonds et al.45 The critical difference was 20%.42 Therefore, with r-values ranging from –0.46 to 0.63. Criterion validity was reliability of the 5-minute, 6-minute or 10-minute walk tests is established between the 6-minute walk tests and two subscales of good to excellent. The 5-minute walk test is considered useful.42,45 the Fibromyalgia Impact Questionnaire: the physical function scale No specialised equipment is required and walk tests appear to be (r = –0.48, p < 0.001) and the pain scale (r = –0.39, p < 0.01). In the acceptable for people with chronic low back pain.45 same study,46 the 6-minute walk test also correlated with the Short-Form Health Survey (SF-36) physical function scale (r = 0.49, Criterion validity was established between the 5-minute and 10-minute walk tests with a high Spearman’s rank correlation of

Research 149 p < 0.001), the SF-36 bodily pain scale (r = 0.38, p < 0.01), the In four of the 14 studies,38–40,42 assessment of the psychometric American Shoulder and Elbow Surgeons function scale (r = 0.68, properties of the submaximal tests was not the primary purpose of p < 0.001) and pain scale (r = 0.53, p < 0.001). In summary, the 6- the study. Data of measurement properties were sparse and the minute walk test showed a fair relationship with the SF-36 methodological shortcomings of the psychometric measurements physical function scale and the Fibromyalgia Impact Questionnaire could have led to bias. physical function scale, and a moderate-to-good relationship with the American Shoulder and Elbow Surgeons function scale.46 Five out of 14 studies investigated test batteries of physical Concurrent validity with the performance-based tests and the performance tasks.42–46 Submaximal exercise tests such as the 5- other quality of life scales was low to moderate. The performance- minute, 6-minute or 10-minute walk tests were merely one item of based measures correlated more strongly with activity limitations the test battery. This could have generated an unclear risk of bias than with pain.46 The dropout rate of 1% was low.46 and could cause underestimation or overestimation of the effect measure because participants had to do the test battery Shuttle walk test completely, and not just one exercise test. Taylor et al47 reported test-retest reliability with an ICC of 0.99, Some uncertainties arose about the reliability and criterion a mean difference of 2.5 m, and upper and lower limits of validity of the conventional A˚ strand test.27,30,34 Good test-retest agreement of –47 and 52 m. They concluded that the shuttle walk reliability (ICC 0.96) was reported in people with chronic low back test is a reliable and responsive test and is simple to administer. pain32 and moderate concurrent validity with the modified A˚ strand test (ICC 0.79) in people with musculoskeletal pain Modified symptom-limited Bruce treadmill test disorders.35 However, the ICC is strongly influenced by the Wittink et al25 assessed the concurrent validity of the modified variation between subjects32 and the low number of participants in the included studies, which may have resulted in a spuriously treadmill test with the SF-36 scale and found a moderate high estimate of reliability. relationship (Spearman’s r = 0.43) in 63 people with chronic low back pain. Despite good reliability and moderate criterion validity, all the studies showed low levels of perceived exertion. The low levels of Discussion perceived exertion may be more likely to be due to fear avoidance than physical deconditioning. This systematic review identified 14 eligible studies about measurement properties of physical capacity tests in people with The gold standard for exercise testing is maximal calorimetry, chronic pain, fibromyalgia and chronic fatigue disorders. Exhaus- with detailed assessment of lactate, VO2max, blood pressure and tive assessment of methodological quality showed some potential electrocardiographic data. However, these detailed assessments bias due to lack of blinding, doubt over whether the measurement are not available to many physiotherapists. Measuring people’s was independent, and no gold standard. This may have allowed subjective perception with standardised assessment (such as overestimation of some of the psychometric properties reported. rating of perceived exertion), monitoring heart rate, and perform- Although the demographic features and disease severity of the ing submaximal exercise tests seem to be the most applicable participants were comparable among the studies, a meta-analysis methods in daily practice. All of the submaximal exercises could not be performed due to heterogeneity among the study identified in this review are useful, feasible, and applicable to designs used, heterogeneity of the psychometric properties the population of interest. At most, one session of 20 to 30 minutes evaluated, and incomplete reporting of the data. Therefore, is necessary for a submaximal test, although a treadmill or a cycle psychometric data from individual studies were reported quanti- ergometer are also needed for some of the tests. tatively and qualitatively. Future research in this area should assess the reliability of Seven of the 14 studies assessed criterion validity of the submaximal exercise tests with higher quality study designs and submaximal tests with questionnaires or other submaximal report data in sufficient detail to allow for meta-analysis. Future tests.25,35,38,43–46 Difficulties in assessing criterion validity were: studies could also evaluate the concurrent validity of submaximal low reproducibility, and operationalisation variability of the exercise tests, compared to maximal tests, in people with chronic criterion at issue. However, there is no appropriate reference pain, fibromyalgia and chronic fatigue disorders. However, the lack standard. This could have led to underestimation of the test of studies of maximal testing of people with chronic pain, validity. fibromyalgia and chronic fatigue disorders may be due to difficulties with such tests.27 Concurrent validity with other physiological None of the included studies mentioned blinding of outcome measures, such as heart rate variability could also be investigated. measurement. This should not have an effect on reliability if the test Heart rate variability is related to emotional arousal48 and might be was done in accordance to the test protocol. However, validity of important in the assessment of physical capacity in this population. the submaximal tests could be overestimated if researchers were aware of the results of the submaximal tests before assessment of In conclusion, there is moderate evidence of the reliability, the questionnaires. This leads to potential for bias in the review. validity and acceptability of the evaluated submaximal exercise tests in people with chronic pain, fibromyalgia and chronic fatigue The stop criteria of the study protocols were comparable: heart disorders. There is no evidence, however, about maximal exercise rate too high or too low, signs of serious cardiovascular or tests in this population. pulmonary difficulties, and chest pain. Only one study added ‘fatigue’ as a stop criterion.41 This could have led to a higher What is already known on this topic: Guidelines recom- dropout rate at the entry of the study, compared to results of other mend graded activity in the treatment of chronic pain, fibro- studies. It is remarkable, however, that these higher dropout rates myalgia and chronic fatigue disorders. Self-reports of physical are only presented at the start of the study and not at the end. disability often do not correlate with pain severity, so objective Protas et al41 hypothesise that this is based on psychosocial fear- assessment of physical capacity is recommended. avoidance associated with pretesting rather than a true indication What this study adds: Although little is known about maxi- of physical deconditioning. Smeets and van Soest35 suggested mal exercise tests in this population, moderate evidence exists strict adherence to the testing protocol and extensive training of that several submaximal exercise tests are reliable, valid and the health care providers to increase the acceptability of the acceptable in people with chronic pain, fibromyalgia and exercise tests. Practical experiences show that acceptability of chronic fatigue disorders. treadmill and bicycle tests is lower in psychosomatic institutions than in outpatient settings. This is attributed to disease severity eAddenda: Appendices 1 and 2 can be found online at and other demographic features. doi:10.1016/j.jphys.2014.06.011 Ethics approval: Nil.

150 Ratter et al: Psychometrics of exercise tests in chronic pain patients Competing interests: There are no conflicts of interests. 24. Strong J, Unruh AM, Wright A. Pain: A Textbook for Therapists. London: Churchill Livingstone; 2001:123–147. Source(s) of support: No sources of support. 25. Wittink H, Rogers W, Sukiennik A, Carr DB. Physical functioning: Self-report and Acknowledgements: We are grateful to our friends, family and performance measures are related but distinct. Spine. 2003;28:2407–2413. colleagues. 26. Nielens H, Cornet D, Rigot F. Maximal versus submaximal testing protocols to evaluate fitness of patients with chronic low back pain. Med Sci Sports Exerc. Correspondence: Julia Ratter, Physiotherapy, Hospital Rivieren- 2002;34(Suppl 1). abstract 270. land Tiel, The Netherlands. Email: [email protected] 27. Wittink H, Hoskins Michel T, Wagner A, Sukiennik A, Rogers W. Deconditioning in patients with chronic low back pain: fact or fiction? Spine. 2000;25:2221–2228. References 28. Hodselmans AP, Dijkstra PU, van der Schans C, Geertzen JHB. Test-retest reliability 1. Kumar P, Clark M. Kumar and Clark’s Clinical Medicine. 8th ed. Edinburgh: Saunders; of psychophysical lift capacity in patients with non-specific chronic low back pain 2012. and healthy subjects. J Rehabil Med. 2007;39:133–137. 2. Turk DC, Okifuji A. Pain terms and taxonomies. In: Loeser D, Butler SH, Chapman JJ, 29. Rudy TE, Lieber SJ, Boston JR, Gourley LM, Baysal E. Psychosocial predictors of eds. In: Bonica’s Management of Pain. 3rd ed. Philadelphia: Lippincott Williams & physical performance in disabled individuals with chronic pain. Clin J Pain. 2003;1: Wilkins; 2001:18–25. 18–30. 3. Main CJ, Spanswick CC. Pain Management: An Interdisciplinary Approach. Edinburgh: 30. Hodselmans AP, Dijkstra PU, Geertzen JHB, van der Schans C. Exercise capacity in Churchill Livingstone; 2001:93. non-specific chronic low back pain patients: a lean body mass-based A˚ strand bicycle test; reliability, validity and feasibility. J Occup Rehabil. 2008;18:282–289. 4. Goldenberg DL. Diagnosis and differential diagnosis of fibromyalgia. Am J Med. 2009;122:14–21. 31. Kelly M, Gagne R, Newman JD, Olney C, Gualtieri C, Trail D. Assessment of fibromy- algia and chronic fatigue syndrome: a new protocol designed to determine work 5. Kato K, Sullivan PF, Evenga˚ rd B, Pedersen L. Chronic widespread pain and its capability: Chronic pain abilities determination (CPAD). Irish Med J. 2008;101: comorbidities: A population-based study. Arch Intern Med. 2006;166:1649–1654. 277–278. 6. White KP, Speechley M, Harth M, Ostbye T. Co-existence of chronic fatigue 32. Keller A, Hellesnes J, Brox JI. Reliability of the isokinetic trunk extensor test, syndrome with fibromyalgia syndrome in the general population: A controlled Biering-Sorensen test and A˚ strand bicycle test: assessment of intraclass correla- study. Scand J Rheumatol. 2000;29:44–51. tion coefficient and critical difference in patients with chronic low back pain and healthy individuals. Spine. 2001;26:771–777. 7. O’Connell NE, Wand BM, Marston L, Spencer S, DeSouza LH. Non-invasive brain stimulation techniques for chronic pain-review: A report of a Cochrane systematic 33. A˚ strand P, Rohdahl K. Textbook of work physiology, physiological bases of exercise. 3rd review and meta-analysis. Eur J Phys Rehabil Med. 2011;47:309–326. edn. New York: McGraw-Hill; 1986:354–391. 8. Afari N, Buchwald D. Chronic fatigue syndrome: a review. Am J Psychiatr. 34. Skinner SJ. Get with the Program: Exercise Prescription for Chronic Fatigue 2003;160:221–236. Syndrome and Fibromyalgia. ACSM’s health and fitness journal. 2005;9:16–21. 9. Berger A, Dukes E, Martin S, Edelsbeerg J, Oster G. Characteristics and healthcare 35. Smeets RJ, van Soest M. The usability of a modified A˚ strand bicycle test to assess costs of patients with fibromyalgia syndrome. Int J Clin Prac. 2007;61:1498–1508. the aerobic capacity in patients with musculoskeletal pain and healthy controls. Disabil Rehabil. 2009;31:1988–1995. 10. Breivik H, Eisenberg E, O’Brien T. The individual and societal burden of chronic pain in Europe: the case for strategic prioritisation and action to improve knowledge 36. Terwee CB, Mokkink LB, Knol DL, Ostelo RW, Bouter LM, de Vet HC. Rating the and availability of appropriate care. BMC Public Health. 2013;13:1229. methodological quality in systematic reviews of studies on measurement proper- ties: a scoring system for the COSMIN checklist. Qual Life Res. 2012;21:651–657. 11. International Association for the Study of Pain. EFIC’s declaration on chronic pain as a major healthcare problem, a disease in its own right. 2001. http://www.iasp- 37. Cook DJ, Mulrow CD, Haynes RB. Systematic reviews: synthesis of best evidence for pain.org [Accessed November 3, 2012]. clinical decisions. Ann Intern Med. 1997;126:376–380. 12. Lillefjell M, Krokstad S, Espnes GA. Prediction of function in daily life following 38. Viitanen JV. Feasibility of fitness tests in subjects with chronic pain (fibromyalgia): multidisciplinary rehabilitation for individuals with chronic musculoskeletal pain; discordance between cycling and 2-km walking tests. Rheumatol Int. 2001;21: a prospective study. BMC Musculoskel Disord. 2007;8:65. 1–5. 13. Wyller VB. The chronic fatigue syndrome - an update. Acta Neurol Scand. 2007;115: 39. van Santen M, Bolwijn P, Verstappen F, Bakker C, Hidding A, Houben H, et al. A 7–14. randomized clinical trial comparing fitness and biofeedback training versus basic treatment in patients with fibromyalgia. J Rheumatol. 2002;29:575–581. 14. Kent P, Mjøsund HL, Petersen DHD. Does targeting manual therapy and/or exercise improve patient outcomes in nonspecific low back pain? A systematic review BMC 40. van Santen M, Bolwijn P, Landewe´ R, Verstappen F, Bakker C, Hidding A, et al. High Med. 2010;8:22. or low intensity aerobic fitness training in fibromyalgia: does it matter? J Rheu- matol. 2002;29:582–587. 15. Vowles KE, Gross RT, Sorrell JT. Predicting work status following interdisciplinary treatment of chronic pain. Eur J Pain. 2004;4:351–358. 41. Protas EJ, Mayer TG, Dersh J, Keeley J, Gatchel RJ, McGeary D. Relevance of aerobic capacity measurements in the treatment of chronic work-related spinal disorders. 16. World Health Organization. World Health Organization supports global effort to Spine. 2004;29:2158–2166. relieve chronic pain. 2004. http://www.who.int/mediacentre/news/releases/2004/ pr70/en/ [Accessed March 7, 2012]. 42. Smeets RJ, Hijdra HJ, Kester AD, Hitters MW, Knottnerus JA. The usability of six physical performance tasks in a rehabilitation population with chronic low back 17. Carleton RN, Abrams MP, Kachur SS, Asmundson GJ. Waddell’s symptoms as pain. Clin Rehabil. 2006;20:989–998. correlates of vulnerabilities associated with fear-anxiety avoidance models of pain: pain-related anxiety, catastrophic thinking, perceived disability, and treat- 43. Harding VR, Williams AC, Richardson PH, Nicholas MK, Jackson JL, Richardson IH, et ment outcome. J Occup Rehabil. 2009;19:364–374. al. The development of a battery of measures for assessing physical functioning of chronic pain patients. Pain. 1994;58:367–375. 18. Edmonds M, McGuire H, Price J. Exercise therapy for chronic fatigue syndrome. Cochrane Database Syst Rev. 2004. CD003200. 44. Sato D, Kaneko F, Okamura H. Reliability and validity of the Japanese-language version of the physical performance test (PPT) battery in chronic pain patients. 19. Nijs J, Paul L, Wallman K. Chronic fatigue syndrome: an approach combining self- Disabil Rehabil. 2006;28:397–405. management with graded exercise to avoid exacerbations. J Rehabil Med. 2008;40:241–247. 45. Simmonds MJ, Olson SL, Jones S, Hussein T, Lee CE, Novy D, et al. Psychometric characteristics and clinical usefulness of physical performance tests in patients 20. Sumathipala A. What is the evidence for the efficacy of treatments for somatoform with low back pain. Spine. 1998;23:2412–2421. disorders? A critical review of previous intervention studies. Psychosom Med. 2007;69:889–900. 46. Mannerkorpi K, Svantesson U, Broberg C. Relationships between performance- based tests and patients’ ratings of activity limitations, self-efficacy, and pain in 21. Turk DC. Clinical effectiveness and cost effectiveness of treatments for patients fibromyalgia. Arch Phys Med Rehabil. 2006;87:259–264. with chronic pain. Clin J Pain. 2002;18:355–365. 47. Taylor S, Frost H, Taylor A, Barker K. Reliability and responsiveness of the shuttle 22. Malliou P, Gioftsidou A, Beneka A, Godolias G. Measurements and evaluations in walking test in patients with chronic low back pain. Physiother Res Int. 2001;6: low back pain patients. Scand J Med Sci Sports. 2006;16:219–230. 170–178. 23. Abdel-Moty AR, Maguire GH, Kaplan SH, Johnson P. Stated versus observed 48. Nickel P, Nachreiner F. Sensitivity and diagnosticity of the 0.1-hz component of performance levels in patients with chronic low back pain. Occup Ther Health heart rate variability as an indicator of mental workload. Hum Factors. 2003;45: Care. 1996;10:3–23. 575–590.

Journal of Physiotherapy 60 (2014) 169 Journal of PHYSIOTHERAPY journal homepage: www.elsevier.com/locate/jphys Appraisal Clinical Practice Guidelines Chronic pain Management of chronic pain pain assessment, early management and care planning in non- specialist settings (contained within Annex 2). This is followed by Date of latest update: December 2013. Date of next update: Will evidence for a variety of management approaches, including: be considered for review in three years. Patient group: Adults with supported self-management; pharmacological therapies (eg, chronic, non-malignant pain. Intended audience: Health care opioids, non-opioid analgesics); psychological-based intervention professionals involved in the assessment and management of (eg, multidisciplinary pain management programs, cognitive adults with chronic pain. Additional versions: The main guideline behavioural therapy); physical therapies (eg, manual therapy, is accompanied by a patient version and three treatment pathways. exercise, traction, electrotherapy); complementary therapies (eg, Expert working group: The expert working groups was comprised acupuncture, herbal medicine); and dietary therapies. Evidence for of a 24-member committee of medical professionals (consultants, manual therapy is presented for low back pain and neck pain. pharmacologist, psychologist, physiotherapist) and patient repre- Exercise approaches covered include: aerobic, strengthening, sentatives from Scotland and Australia. Funded by: Not stated. movement facilitation, stabilisation exercises, walking, T’ai Chi, Consultation with: Targeted consultation occurred with expert yoga, Pilates, aquatic and general exercise therapy. Care pathways reviewers invited to comment on draft versions. In addition, draft are also provided for people with neuropathic pain, and for people guidelines were presented at a national open meeting and with chronic pain using strong opioids. The guidelines concluded available on the SIGN website. Approved by: The Scottish with recommendations for future research. Intercollegiate Guidelines Network (SIGN), Healthcare Improve- ment Scotland. Location: The guidelines and associated docu- Sandra Brauer ments are available at: http://www.sign.ac.uk. The University of Queensland, Australia Description: These guidelines are published in a 71-page document beginning with a one-page summary of key recom- http://dx.doi.org/10.1016/j.jphys.2014.06.006 mendations and the associated supporting levels of evidence. Evidence for the assessment and planning of care for people with chronic pain is presented first; this includes a pathway for chronic Urinary incontinence Urinary incontinence: the management of urinary incontinence in women Date of latest update: September 2013. Date of next update: Not incontinence in women. They replace a previous guideline published stated. in 2006, so it is indicated whether recommendations in the current Patient group: Women aged over 18 years with any type of urinary version are an update, or are the same as in the previous guidelines. incontinence. Intended audience: Health care professionals The full guidelines include: recommendations, detailed discussion involved in the management of women with urinary incontinence. of the evidence underpinning the recommendations, care pathways Additional versions: The current guidelines are an update of the and all references. They outline the evidence for assessment and National Institute for Health and Care Excellence (NICE) clinical investigation, including history taking, physical examination and guideline 40 (2006): Urinary continence: the management of pelvic floor muscle assessment. Evidence for management of urinary urinary incontinence in women. Additional documents include incontinence is then presented. This begins with conservative appendices, a summary document and a version for the public. management, including: physical therapies, lifestyle interventions Expert working group: The expert working group was comprised and neurostimulation, and discusses evidence for preventative use of a 10-member committee of medical professionals (consultants, and the optimal sequence and timescale for conservative therapies. GPs, nurse, physiotherapist) and a consumer representative, all Other management options discussed include pharmacological from the UK. Funded by: Not stated. Consultation with: The therapies, invasive procedures and surgical interventions. The guidelines were developed in consultation with the National evidence is underpinned by over 1000 references. Collaborating Centre for Women’s and Children’s Health, which is based at the Royal College of Obstetricians and Gynaecologists, UK. Sandra Brauer Approved by: National Institute for Health and Care Excellence The University of Queensland, Australia (NICE). Location: The guidelines and additional documents are available at: http://guidance.nice.org.uk/CG171 http://dx.doi.org/10.1016/j.jphys.2014.06.001 Description: These guidelines are published in a 387-page document and provide evidence for the management of urinary 1836-9553/ß 2014 Published by Elsevier B.V. on behalf of Australian Physiotherapy Association.

Journal of Physiotherapy 60 (2014) 136–143 Journal of PHYSIOTHERAPY journal homepage: www.elsevier.com/locate/jphys Research Weight training is not harmful for women with breast cancer-related lymphoedema: a systematic review Vincent Singh Paramanandam a, Dave Roberts b a Physiotherapy Department, Tata Memorial Hospital, India; b Faculty of Health and Life Sciences, Oxford Brookes University, United Kingdom KEYWORDS ABSTRACT Breast neoplasm Question: Is weight-training exercise intervention harmful to women with or at risk of breast cancer- Lymphoedema related lymphoedema? Design: Systematic review with meta-analysis of randomised trials. Weight training Participants: Women with or at risk of breast cancer-related lymphoedema. Intervention: Progressive Exercise weight-training exercise. Outcome measures: The primary outcomes were severity (volume difference) and incidence of arm lymphoedema. Secondary outcomes included muscle strength of the upper and lower limbs, quality of life and body mass index. Results: Eleven studies from eight trials involving 1091 women were included. Weight-training exercise of low to moderate intensity with relatively slow progression significantly improved the upper limb strength (SMD 0.93, 95% CI 0.73 to 1.12) and lower limb strength (SMD 0.75, 95% CI 0.47 to 1.04) without increasing the arm volume (SMD –0.09, 95% CI –0.23 to 0.05) or incidence of breast cancer-related lymphoedema (RR 0.77, 95% CI 0.52 to 1.15). No significant effects were noted for body mass index (SMD –0.10, 95% –0.31 to 0.11). Some aspects of quality of life may improve with weight training. Participants in all trials used pressure garments and received supervision; no trials used high-intensity weight training. Conclusions: Weight training appears to be safe and beneficial in improving limb strength and physical components of quality of life in women with or at risk of lymphoedema. Pressure garments, supervision and limiting the intensity of the weight training may each be important, but this could not be confirmed with this review. Registration: PROSPERO CRD42012002737. [Paramanandam VS, Roberts D (2014) Weight training is not harmful for women with breast cancer-related lymphoedema: a systematic review. Journal of Physiotherapy 60: 136–143] ß 2014 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Introduction feelings of discomfort, pain, heaviness in the arm, disfigurement, Breast cancer is the most common cancer and the leading cause psychosocial disturbance and elevated risk of infection, so BCRL is of cancer deaths among women,1 accounting for 23% of total cancer considered to be the most feared complication of breast cancer.9,10 cases and 14% of cancer deaths. Early detection and recent advances in breast cancer treatment have improved the 5-year Published reports on the prevalence of BCRL range from 2 to relative survival rate to above 80%.2,3 Despite this, cancer treatments cause many long-term functional impairments and 83%, although this wide variance is due in part to discrepancies in considerably reduce the quality of life.4 Some of the post- treatment complications are: fatigue, weakness, loss of muscle the definition, diagnostic threshold and measurement methods extensibility, limited shoulder range of motion, upper body pain, used.11 The onset of BCRL is unpredictable and can even occur pulmonary complications, neuropathy, body composition and many years after surgery.12–14 breast cancer-related lymphoedema (BCRL).5,6 It was believed that exercise could adversely affect the BCRL is a chronic swelling of the arm, hand and associated trunk quadrant. It usually develops after damage to the axillary lymph lymphoedema-prone arm in women with breast cancer, until nodes due to breast cancer therapies. Surgical removal of lymph nodes, which is considered to be important for prognosis, causes the seminal work by McKenzie revealed no exacerbation or new permanent damage to the lymphatic pathways.7 In addition, many patients are treated with external beam radiation and this may cases of lymphoedema among women with breast cancer who lead to constriction of the lymphatic vessels due to fibrosis, and participated in dragon boat racing.15 However, a prospective study delay the growth of newer lymphatic vessels after the lymph node by Johansson and colleagues16 reported an acute increase in arm excision.8 Thus, overall lymphatic drainage may be reduced significantly and lead to BCRL.7 This condition is associated with volume within 24 hours following weight training. Additionally, a study by Lane and colleagues17 assessed the effect of exercise on BCRL by lymphoscintigraphy and revealed that the lymphoede- matous hand had more similar lymphatic clearance to that of the controls during upper body exercises. However, exercises did not markedly increase the uptake of radiopharmaceuticals in the axilla and showed backflow. Hence, the authors concluded that exercise might increase the chance of BCRL. On the contrary, recent studies found no harmful effects of exercise on BCRL.18,19 http://dx.doi.org/10.1016/j.jphys.2014.07.001 1836-9553/ß 2014 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/3.0/).

Research 137 Historically, people with or at risk of lymphoedema have been Box 1. Inclusion criteria. advised not to do strenuous activity or exercises with the operated- side arm in order to reduce the risk of causing or exacerbating Design lymphoedema.11 Guidelines advise to not lift heavy weights or  randomised trial children and to avoid doing repeated activities.2,20 Recent studies,  peer reviewed however, have reported that weight training did not induce or  published in English after 2001 exacerbate BCRL when it was performed under supervision with Population slow progression.21,22 This type of exercise results in robust  women with breast cancer diagnosis with or at risk of functional, physiological, psychological and clinical benefits.4 developing lymphoedema Progressive weight training is intended to elicit benefits in Intervention health and performance by challenging skeletal muscles with  weight-training exercises controlled physiological stress to the onset of muscle fatigue. These Outcomes weight-training sessions are followed by an optimal interval of  lymphoedema onset or exacerbation rest, ranging from 48 to 72 hours; this allows physiological  limb strength adaptation to occur.23,24 Aside from local effects at the arm, weight  quality of life training has many other benefits, including: a reduction in cancer-  body mass index related fatigue,25 and improvement in body weight, psychological Comparison well being,26 bone density,27 body image28 and survival.29 Some  sham exercise narrative19 and systematic4,11,18,30,31 reviews have been published  light exercise on this topic. However, these reviews included studies with mixed  conservative therapies such as light stretching exercise interventions30 or included non-randomised studies.4,18  no-intervention control Furthermore, at least two more randomised trials have been  only lower body exercises and education published since these previous reviews.4,18,31 Therefore, this present review was considered to be necessary and sought to to the total score because it is related to external validity. answer these research questions: The overall score is therefore calculated as the number of the remaining 10 items that the study achieves. Considering the nature 1. Is weight-training exercise safe for women with or at risk of of intervention studied in the included papers, blinding of lymphoedema after breast cancer? participants and therapists would be impractical, so scores above eight would not be anticipated. The PEDro scale can detect 2. Does weight-training exercise improve muscle strength, quality potential bias with fair to good reliability34 and is a valid measure of life and body mass index in this population? of methodological quality of trials.35 Method Design Only randomised trials were included in the review because Identification and selection of studies they eliminate more sources of potential bias than other study The following databases were searched electronically from designs. The publication year to post 2001 was limited due to inception to July/August 2012: PubMed, EMBASE, PsycINFO, advances in the management of breast cancer. CINAHL, AMED, Cochrane, PEDro, SPORTDiscus and Web of Science. Date restriction, female gender limit and peer review Participants were applied to the results where possible. In addition, reference This review included studies of women of any age who had or lists of the identified studies and previous reviews were searched for any potential articles. Furthermore, distinguished authors from were at risk of developing lymphoedema during or following this research area were contacted through email for any missed breast cancer treatment. Breast cancer treatment was defined as and relevant studies. Three key terms, ‘weight training’, ‘lym- any type of breast surgery, along with one of the following phoedema’ and ‘breast neoplasm’, were used to generate an procedures to the axilla: axillary lymph node dissection, axillary exhaustive list of key words. Appendix 1 (see eAddenda) shows the lymph node sampling or sentinel lymph node dissection with or full search strategies. without radiotherapy to the breast and/or axilla. Studies involving women with lymphoedema following local recurrence or metas- Eligibility assessment of each study was conducted in a non- tasis were excluded. blinded and standardised manner by a single researcher (VP) under the supervision of the second author (DR) in three stages and every Interventions effort was undertaken to avoid subjective bias.32 In the first stage, To be eligible for this review, trials were required to have articles obtained through the database searches were compared for duplicate entries using the de-duplicating facility of reference studied the effects of weight training or resistance exercises. management softwarea and were manually cross checked. The Studies with mixed exercises (apart from warm-up and cool- titles and abstracts of the remaining articles were examined for down), which could possibly moderate the effect of weight eligibility against the pre-defined criteria, as presented in Box 1. training, were not considered for inclusion. The above-mentioned Articles that were not definitely excluded by this screening were intervention was required to have been assessed against no obtained in full text for further assessment. Eligible studies intervention or against any of the control interventions listed in underwent data extraction. Study selection is reported according Box 1. to PRISMA guidelines.33 Outcomes Characteristics of included studies The primary outcome was BCRL, analysed as either the Quality incidence or severity of lymphoedema identified by comparing The quality of the included studies was assessed using the the volume difference between the operated-on and contralateral arms. Volume could be measured directly using the water PEDro scale,34 which consists of 11 items that address external displacement method or non-invasive optoelectronic scanning validity, risk of bias (internal validity) and interpretability. (ie, perometry), or calculated from a series of circumferential Although there are 11 items, the first item does not contribute measurements using a measuring tape. Additionally, studies that used a simple circumference measurement of the arm were also considered for this review. The reported difference could either be

]G$DT)1_erugiF([I138 Paramanandam and Roberts: Weight training in lymphoedema Titles and abstracts screened (n = 333) Results Papers excluded after screening titles/abstracts Flow of studies through the review (n = 295) After screening of the search results, 11 papers reporting eight Potentially relevant papers retrieved for trials were included in the review. Figure 1 depicts the flow of evaluation of full text (n = 38) studies through this review. In the eleven included papers, seven were from the United States of America.21,22,39–43 Among these Papers excluded after evaluation of full text seven papers, three of them39,41,42 were from a single trial called (n = 27) Weight Training for Breast Cancer Survivors (WTBS); they were • participants did not have lymphoedema considered as a single trial in the present review. Another three papers from the United States of America21,22,43 were from a trial related to breast cancer (n = 1) named Physical Activity and Lymphoedema (PAL); this trial was • study design was not randomised (n = 11) conducted with two distinctive objectives with adequate pow- • inappropriate outcome measure (n = 5) er.21,22 Thus, they were considered as two independent trials for • intervention was not weight training (n = 9) the present review. The last trial from the United States of America • full text was not available (n = 1) was a study by Anderson and colleagues,40 which included 30 minutes of walking with the resistance training. It was included Papers included in systematic review (n = 11) in the present review in view of the fact that the walking component would give negligible aerobic activity to the upper Figure 1. Flow of studies through the review. limb. The other four trials were from Canada,26 Norway,44 Australia45 and the Republic of Korea.46 absolute or relative. Absolute volume difference is the change of arm volume on the operated side, and relative change is the Characteristics of the included studies volume difference between the operated-on and contralateral arms. For analysing the onset of lymphoedema, the definition of Quality lymphoedema and the diagnostic cut-off values varies between The individual items achieved by each of the included trials are studies; therefore, studies with a predefined criterion for lymphoedema diagnosis were regarded to be eligible for inclusion. presented in Table 1. As discussed above, blinding of participants The secondary outcome measures (muscle strength of upper and and therapists is impractical, so no trials achieved this. All the lower limbs, quality of life and body mass index) were also included trials met the external validity item by specifying the included for analysis, if reported. eligibility criteria and source of participants. Similarly, all of the included trials met four other criteria that contribute to the Data analysis total score: random allocation, similar groups at baseline, reporting of between-group difference, and reporting of point Data extraction was performed by a single researcher (VP) estimates and variability. Three trials39,40,46 did not report using under the supervision of the second author (DR) using forms a valid method of allocation concealment; three trials26,40,46 failed developed and pilot tested for this review.36 Additionally, three to use blinded outcome assessors; three trials did not analyse authors of the included studies were contacted through emails for by intention to treat; 39,40,46 and three trials had > 15% loss to further data because they were presented in dichotomous format. follow-up.21,40,46 However, only one author21 replied and provided the required data. Participants The included trials provided data on 1091 participants, who had Meta-analyses were performed wherever appropriate data were available, and narrative syntheses are presented other- undergone either modified radical mastectomy or breast conser- wise.32,37 The continuous outcomes in the included studies were vation surgery along with different axillary node management. The typically reported with different scales, so standardised mean mean age of participants ranged from 49 to 57 years. Two trials21,46 differences (SMD) were calculated with a random-effects model enrolled women with BCRL and six trials22,26,39,40,44,45 enrolled and reported with a 95% CI. Lymphoedema incidence data were women at risk of developing BCRL, as presented in Table 2. pooled and reported as relative risk with a 95% CI.38 Additionally, subgroup analysis was attempted wherever sufficient data were Intervention available to compare slow progressive and moderate-intensity All of the trials provided the exercise intervention, at least exercise groups. partly, under supervision in an institutional setting, although in two studies21,22 the institution was in a community setting, for example a YMCA fitness centre. The supervision was provided by either physiotherapists or certified exercise professionals, al- though one trial did not provide any clear details about the Table 1 PEDro scores of the included studies (n = 8). Study Random Concealed Groups Participant Therapist Assessor < 15% Intention-to-treat Between-group Point Total allocation allocation similar at blinding blinding blinding dropouts analysis difference estimate and (0 to 10) baseline reported variability reported Ahmed39,41,42 Y NY N N YY N Y Y6 N Y Y4 Anderson40 Y N Y N N NN Y Y Y7 Y Y Y8 Courneya26 Y Y Y N N NY N Y Y4 Y Y Y8 Kilbreath45 Y Y Y N N YY Y Y Y7 Y Y Y8 Kim46 Y N Y N N NN Sagen44 Y Y Y N N YY Schmitz21 Y Y Y N N YN Schmitz22 Y Y Y N N YY Y = yes, N = no.

Table 2 Research 139 Summary of the included studies (n = 8). Experimental intervention Outcomes Study Participants Control intervention Follow-up Country Group, supervised and institution Lymphoedema Ahmed39,41,42 Breast cancer,  4 mth after based, 6 mth  no intervention Muscle strength treatment (MRM/BCS with AND), Quality of life USA may be on hormone therapy, no  upper body: 0 to 0.25 kg, if no BCRL  written information about 6 mth (next 6 mth partial contraindication to weight symptoms lymphoedema awareness and crossover) training ACS prevention exercises  lower body: greatest weight that the Lymphoedema n = 46, mean age = 52 yr participant could use for the set  quarterly newsletter with tips Quality of life protocol about nutrition and physical 18 mth Anderson40 Stage I to III breast cancer, 4 to 12 activity USA wk after surgery (MRM/BCS with  8 to10 reps, 1 to 3 sets over the first 2 Lymphoedema AND or SND) to 3 wk Aerobic exercise group: Muscle strength  60% of VO2max, 3/wk, for 6 wk, Quality of life n = 104, mean age = 54 yr Low intensity 6 mth then 70% for 6 wk, then 80% Courneya26 Stage I to IIIA breast cancer, Individual, supervised and institution  15 min for 3 wk, then increase Lymphoedema Strength Canada beginning first-line adjuvant based, 18 mth Quality of life chemotherapy after MRM/BCS by 5 min every 3 wk 8 mth with AND  warm up: 5 min aerobic, Cool down: Usual care group: 10 min stretch  asked not to initiate an exercise Lymphoedema n = 164, mean age = 49 yr Quality of life  upper body: 50% of 1RM for 2 wk, 0.5 program 2 mth Kilbreath45 Stage I to III breast cancer, 4 to 6 to 1 kg increment weekly Australia wk after surgery (SND or AND)  written arm exercises: brief, Lymphoedema  lower body: 50% of 1RM for 2 wk, 0.5 active-assisted, and active Body mass index n = 160, mean age = 53 yr to 2.5 kg increment weekly overhead, in frontal/sagittal 2 yr planes Kim46 BCRL, confirmed by  increment if 12 reps for 2 sessions Lymphoedema Korea circumference and  walk: 30 min, moderate to somewhat  if BCRL identified on fortnightly Muscle strength lymphoscintigraphy monitoring, referral for a Quality of life hard on RPE compression garment and 1 yr n = 40, mean age = 51 yr possibly other therapy Moderate intensity Sagen44 Early-stage breast cancer, 2 d  manual lymphatic drainage Norway after mastectomy or breast- Supervised, institution based, during  compression therapy conserving surgery with AND chemotherapy  remedial exercise Schmitz21,43 (levels I and II), with any USA adjuvant treatment  warm-up/cool-down: 5 min aerobic,  avoid heavy/strenuous stretch activities for 6 mth n = 204, mean age = 55 yr  training: 60 to 70% of 1RM  written information about Unilateral breast cancer, 1 to 5 y  8 to 12 reps, 2 sets these restrictions after diagnosis, with BCRL  increment weight by 10% confirmed by 10% volume  physiotherapy 1/wk: 6 passive difference between both arms if > 12 reps manual techniques for flexibility and common toxicity criteria and light massage of the affected Moderate intensity shoulder, arm and scar for 6 mth n = 141, mean age = 57 yr Supervised, institution based, 8 wk  physiotherapy 1/wk: 6 passive manual techniques for  written arm exercises: brief, flexibility and light massage of active-assisted and active overhead, the affected shoulder, arm and in frontal/sagittal planes scar for 6 mth  upper body: 1 to 1.5 kg initially  8 to 15 reps, 2 sets  increment weight until hard on RPE Low intensity Partly supervised and partly institution based, 8 wk  manual lymphatic drainage  compression therapy  remedial exercise  upper body: 0.5 kg for 2 wk, then 1 kg for 6 wk  10 reps, 2 sets Low intensity Supervised and institution based, 6 mth  training: 0.5 kg for 2 wk  increment as tolerated  15 reps, unlimited sets Low intensity Partly supervised, group and community based, 1 yr  training: little to no resistance, initially  10 reps, 2 sets initially  increment to 3 sets in first 5 wk  increment weight by 0.25 kg if tolerated Low intensity

140 Paramanandam and Roberts: Weight training in lymphoedema Table 2 (Continued ) Participants Experimental intervention Control intervention Outcomes Study Follow-up Country Unilateral breast cancer, 1 to 5 y Partly supervised, group and  physiotherapy 1/wk: 6 passive Schmitz22,43 after diagnosis, with  2 lymph community based, 1 yr manual techniques for Lymphoedema nodes removed, no BCRL and no flexibility and light massage of Muscle strength USA contraindications to exercise  training: little to no resistance, the affected shoulder, arm and Quality of life initially scar for 6 mth 1 yr  10 reps, 2 sets initially n = 154, mean age = 56 yr  increment to 3 sets in first 5 wk  increment weight by 0.25 kg if tolerated Low intensity ACS = American Cancer Society, AND = axillary node dissection, BCRL = breast cancer related lymphoedema, BCS = breast conserving surgery, MRM = modified radical mastectomy, RPE = rating of perceived exertion scale, SND = sentinel lymph node dissection, 1RM = one repetition maximum. supervisor.45 Four trials21,22,39,46 were conducted in groups, one Body mass index was reported only in three studies,21,22,39 as presented in Table 2. implied that the intervention was delivered on an individual basis,40 and the remaining three trials26,44,45 did not report Effect of weight training whether the intervention was group based or not. Two of the Severity of BCRL included trials26,45 were multi-centre trials. The weight-training Although the best estimate of the overall effect on lymphoe- program was categorised as low intensity (based on low weights dema severity favoured weight training, this was not statistically and/or slow progression) in six trials21,22,39,44–46 and moderate significant (SMD –0.09, 95% CI –0.23 to 0.05), as presented in intensity in two trials,26,40 as presented in Table 2. Figure 2. See Figure 3 on the eAddenda for a more-detailed forest plot. The trials in these forest plots are arranged to illustrate the The study by Courneya and colleagues26 compared three subgroup analysis, which identified no considerable difference between the low-intensity and moderate-intensity subgroups. groups: a weight training group, an aerobic training group and a Incidence of BCRL usual care group. Wherever applicable, two comparisons were Although the best estimate of the overall effect on lymphoe- presented: weight training versus aerobic training, and weight dema incidence favoured weight training, this was not statistically significant (RR 0.77, 95% CI 0.52 to 1.15), as presented in Figure 4. training versus usual care. However, the comparison of weight See Figure 5 on the eAddenda for a more-detailed forest plot. Again, subgroup analysis identified no considerable difference between training versus aerobic training was not included in quantitative the low-intensity and moderate-intensity subgroups. pooling to avoid overestimation of effect. Outcomes Upper limb muscle strength Five trials21,22,26,40,46 measured volume using the water Meta-analysis of four comparisons21,22,26,39 with upper limb displacement method and the other three trials39,46,47 estimated strength as the outcome showed better results in the weight- volume using circumference measures, although one of these39 training group than the controls, which was statistically significant (SMD 0.93, 95% CI 0.73 to 1.12). The low-intensity and moderate- only reported a single circumference measure. Six intensity subgroups again had similar results. This meta-analysis trials21,22,26,39,44,45 reported inter-limb volume difference, whilst G[(Figur]e_4)TD$I others reported volume change with treatment in the ipsilateral arm. Only two studies21,22 included clinician diagnosis based on the Common Toxicity Criteria of the US National Cancer Institute as a primary outcome. All the included studies reported quality of life a)2_TrugiF([D$IG]e s either primary or secondary outcomes using various scales. Subgroup SMD (95% CI) Subgroup Risk Ratio (95% CI) Study Random Study Random Low intensity Sagen44 Low intensity Kim46 Ahmed39 Schmitz21 Kilbreath45 Schmitz22 Sagen44 Subtotal (I2 = 0%, Effect p = 0.19) Schmitz21 Schmitz22 Moderate intensity Anderson40 Subtotal (I2 = 0%, Effect p = 0.30) Courneya26 Moderate intensity Subtotal (I2 = 0%, Effect p = 0.73) Courneya26 Total (I2 = 0%, Effect p = 0.20) Subtotal (Effect p = 0.31) Subgroup difference p = 0.66 Total (I2 = 0%, Effect p = 0.20) -1 -0.5 0 0.5 1 Subgroup difference p = 0.51 1 10 200 Favours training Favours control 0.005 0.1 Favours training Favours control Figure 2. Standardised mean difference (SMD) (95% CI) of effect of weight-training Figure 4. Relative risk (95% CI) of the onset of breast cancer-related lymphoedema exercise versus control on the severity of breast cancer-related lymphoedema, with weight-training exercise versus control, pooling data from six trials (n = 860). showing data from six trials (n = 776).

]GIF$DT)6_erugi([ Research 141 Subgroup SMD (95% CI) Study Random Low intensity Ahmed39 Schmitz21 Schmitz22 Subtotal (I2 = 0%, Effect p < 0.001) Moderate intensity Courneya26 Subtotal (Effect p < 0.001) Total (I2 = 0%, Effect p < 0.001) Subgroup difference p = 0.87 -1 -0.5 0 0.5 1 1.5 2 Favours control Favours training G]Figure[$D_(T8)I Figure 6. Standardised mean difference (SMD) (95% CI) of effect of weight-training exercise versus control on chest press, pooling data from four trials (n = 450). Subgroup SMD (95% CI) Study Random Low intensity Ahmed39 Schmitz21 Schmitz22 Subtotal (I2 = 65%, Effect p < 0.001) Moderate intensity Courneya26 Subtotal (Effect p < 0.001) Total (I2 = 51%, Effect p < 0.001) Subgroup difference p = 0.86 -1 -0.5 0 0.5 1 1.5 2 Favours control Favours training Figure 8. Standardised mean difference (SMD) (95% CI) of effect of weight-training exercise versus control on leg press, pooling data from four trials (n = 455). is presented in Figure 6. See Figure 7 on the eAddenda for a Quality of life more-detailed forest plot. In addition, a study by Kilbreath and colleagues45 reported individual muscle group strength contrary Six trials provided data related to quality of life. Three to other studies, which reported bench press, so it was not trials26,39,40 reported global quality of life scores whereas the included in the overall effect estimate. Although one result in this rest21,22,46 reported only individual domains of the quality of life study (horizontal flexion strength) favoured the control group, it was not statistically significant and the other shoulder move- scale. The forest plot in Figure 12 therefore presents pooling by ments tested showed some improvement with weight-training exercise. these two subgroups, without a single overall result. A more Lower limb muscle strength detailed forest plot is available in Figure 13 on the eAddenda. The Meta-analysis of lower limb strength data from the same four T_$[IG]1(Figu)r0eDglobal quality of life score showed a positive trend towards the trials21,22,26,39 also showed significantly better results in the weight-training group than the controls (SMD 0.75, 95% CI 0.47 to Subgroup SMD (95% CI) 1.04). This meta-analysis is presented in Figure 8. See Figure 9 on Study Random the eAddenda for a more-detailed forest plot. The low-intensity and moderate-intensity subgroups again had similar results. Low intensity Ahmed39 Body Mass Index Schmitz21 The overall effect based on three studies21,22,39 that reported Schmitz22 Subtotal (I2 = 0%, Effect p = 0.36) body mass index revealed no significant benefit of weight training (SMD –0.10, 95% CI –0.31 to 0.11), as presented in Figure 10. See Total (I2 = 0%, Effect p = 0.36) Figure 11 on the eAddenda for a more-detailed forest plot. All three of these trials used a low-intensity intervention, so no subgroup -1 -0.5 0 0.5 1 analysis was performed. Favours training Favours control Figure 10. Standardised mean difference (SMD) (95% CI) of effect of weight-training exercise versus control on body mass index, pooling data from three trials (n = 343).

F([$g1IGu]re2_)TDi142 Paramanandam and Roberts: Weight training in lymphoedema Subgroup SMD (95% CI) daily activities.21 Another important finding is that weight training Study Random improved muscle strength irrespective of adjuvant treatment status.26 A review by Cheema and colleagues4 suggested that upper Global quality of life score body function and strength are of the utmost importance in breast cancer survivors post-surgery. Improved arm strength might give Ahmed39 women a sense of control over their daily activities and prevent a spiral of disuse atrophy and associated impairments. Anderson40 Although a recent meta-analysis showed a significant reduction Courneya26 in body mass index as a result of physical activity intervention in people with breast cancer,49 the pooled effect in the present review Subtotal (I2 = 0%, Effect p = 0.23) was inconclusive. This lack of effect may be due to the low intensity of the exercise interventions delivered in these studies, which may Physical health need a prolonged period of training to be effective. Ahmed39 Variation among the quality of life scales and inconsistent reporting in the included studies precluded the overall pooling of Kim46 data. In view of the fact that weight-training exercise generally improves physical function and health, global measures of quality Schmitz21 of life might not be sensitive enough to detect changes specific to weight training.26,40 Schmitz22 The selection was conducted by the first author according to a Subtotal (I2 = 22%, Effect p = 0.007) pre-planned and well-defined protocol, under supervision from the second author. No blinding methods were employed and there -1 -0.5 0 0.5 1 1.5 was no blinding of authors and affiliations. Consequently, the risk Favours control Favours training of selection bias could be an issue in the present review. Therefore, to limit this bias, the list of selected studies was consulted with Figure 12. Standardised mean difference (SMD) (95% CI) of effect of weight-training experts in this field via email before the final selection was made. exercise versus control on quality of life: global score from three trials (n = 325) and Clinical heterogeneity among these studies limited the scope of Physical Health domain score from four trials (n = 351). statistical synthesis; therefore, to avoid misleading outcome and interpretation, a narrative synthesis along with the meta-analysis weight-training group. The Physical Health domain score demon- was conducted. In most of the outcomes, both the narrative and strated a significant overall improvement (SMD 0.34, 95% CI 0.09 to quantitative synthesis produced similar results. 0.58) in the weight-training group compared to the control group. In conclusion, weight training is a safe and effective exercise Heterogeneity and Publication bias modality in women with or at risk of developing BCRL. It improves the strength of the affected arm and physical components of Substantial statistical heterogeneity was noted in the low- quality of life without causing negative effects. Additionally, intensity subgroup in the meta-analysis of lower limb strength weight training helps to maintain the body mass index. Compres- (Tau2 = 0.09, chi2 = 5.78, df = 2, p = 0.06, I2 = 65%). When the study sion garments may be worn during exercise, and close monitoring by Ahmed and colleagues39 was excluded from analysis (not and supervision by a trained professional at the beginning of shown in Figure 8), however, the heterogeneity reduced to treatment is recommended. Weight-training exercise with low to moderate (Tau2 = 0.04, chi2 = 2.10, df = 1, p = 0.15, I2 = 52%). That moderate intensity, and slow to regular progressive exercise may study may have varied due to the absence of methodological be used in the beginning, but these need to be progressed features to control bias, which included allocation concealment, according to the symptom response. Although the intensity of blinding and attrition. initial intervention is recommended to be low, there does not need to be any upper weight limit as long as patients are symptom free. Discussion In recent years the role of weight training in BCRL has been the Overall findings of this review revealed that supervised weight- focus of many researchers. Nevertheless, many aspects of weight training exercise does not increase the risk or severity of BCRL and training in breast cancer and BCRL need further research. Although it it improves muscle strength of the limbs, as well as physical is slow progressive exercise, low-intensity exercise is recommended components of quality of life. These findings are similar to the to protect the arm from adverse effects. There is a lack of trials conclusions of recent reviews,18,19 although the present review comparing moderate or high-intensity training against slow additionally provides the statistical pooling of data, which is progressive training. Furthermore, there is no evidence to suggest generally considered to be more precise.48 that high-intensity weight training is harmful to the arm with, or at risk of BCRL. Although supervision and compression garments are The finding that weight training does not increase the risk or featured in the reviewed studies, their effectiveness needs to be severity of BCRL is very relevant to physiotherapists managing confirmed. women with BCRL, because weight training has many physical, psychological and clinical benefits. This finding does contradict What is already known on this topic: Breast cancer is some other studies. For example, the lymphatic function study by common among women. Many women treated for breast Lane and colleagues17 showed increased lymphoedema with cancer develop lymphoedema. Some physiological studies exercise training, but this study was not a prospective clinical trial. suggest that weight training may promote lymphoedema in this population. Participants in all trials used pressure garments and received What this study adds: Weight training does not increase the supervision, and no trials used high-intensity weight training. onset or severity of lymphoedema in women after breast Pressure garments, supervision and limiting the intensity of the cancer. Weight training also improves upper and lower limb weight training may each be important, but the present review could strength and improves aspects of quality of life related to not confirm this. Previous reviews18,19 suggested that supervision physical function. may not only help in learning the exercise program appropriately, but also in alleviating the fear of developing BCRL among women. Footnotes: a Zotero, Roy Rosenzweig Center for History and New Media Overall, muscle strength improved significantly more with weight training than the control. Furthermore, this improvement was significant even when the control groups did aerobic exercise.26 According to the theoretical assumptions of included studies, weight training may provide adequate strength to protect the arm from accidental injuries by reducing the relative stress of

Research 143 eAddenda: Figures 3, 5, 7, 9, 11 and 13 and Appendix 1 can be 25. Cramp F, Daniel J. Exercise for the management of cancer-related fatigue in adults. Cochrane Database Syst Rev Online. 2008;(2):CD006145. http://dx.doi.org/10.1002/ found online at doi:10.1016/j.jphys.2014.07.001 14651858.CD006145.pub2. Ethics approval: Not applicable. 26. Courneya KS, Segal RJ, Mackey JR, Gelmon K, Reid RD, Friedenreich CM, et al. Effects of Aerobic and Resistance Exercise in Breast Cancer Patients Receiving Adjuvant Competing interests: Nil. Chemotherapy: A Multicenter Randomized Controlled Trial. J Clin Oncol. 2007;25(28):4396–4404. http://dx.doi.org/10.1200/JCO.2006.08.2024. Source(s) of support: Nil. 27. Schwartz A, Winters-Stone K, Gallucci B. Exercise Effects on Bone Mineral Density Acknowledgements: Nil. in Women With Breast Cancer Receiving Adjuvant Chemotherapy. Oncol Nurs Forum. 2007;34(3):627–633. http://dx.doi.org/10.1188/07.ONF.627-633. Correspondence: Vincent Paramanandam, Physiotherapy De- 28. Milne H, Wallman K, Gordon S, Courneya K. Effects of a combined aerobic and partment, Tata Memorial Hospital, India. Email: vinsu24@gmail. resistance exercise program in breast cancer survivors: a randomized controlled trial. Breast Cancer Res Treat. 2008;108(2):279–288. http://dx.doi.org/10.1007/ com s10549-007-9602-z. References 29. Irwin ML, McTiernan A, Manson JE, Thomson CA, Sternfeld B, Stefanick ML, et al. Physical activity and survival in postmenopausal women with breast cancer: 1. Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer statistics. CA results from the women’s health initiative. Cancer Prev Res Phila Pa. Cancer J Clin. 2011;61(2):69–90. http://dx.doi.org/10.3322/caac.20107. 2011;4(4):522–529. http://dx.doi.org/10.1158/1940-6207.CAPR-10-0295. 2. Cancer Research UK. Breast cancer - UK incidence statistics: Cancer Research UK. 30. Chan DNS, Lui LYY, So WKW. Effectiveness of exercise programmes on shoulder 2011. Available at: http://info.cancerresearchuk.org/cancerstats/types/breast/ mobility and lymphoedema after axillary lymph node dissection for breast cancer: incidence/. Accessed December 15, 2011 systematic review. J Adv Nurs. 2010;66(9):1902–1914. http://dx.doi.org/10.1111/ j.1365-2648.2010.05374.x. 3. Coleman MP, Quaresma M, Berrino F, Lutz J-M, De Angelis R, Capocaccia R, et al. Cancer survival in five continents: a worldwide population-based study (CON- 31. McNeely ML, Campbell K, Ospina M, Rowe BH, Dabbs K, Klassen TP, et al. Exercise CORD). Lancet Oncol. 2008;9(8):730–756. http://dx.doi.org/10.1016/S1470- interventions for upper-limb dysfunction due to breast cancer treatment. The 2045(08)70179-7. Cochrane Coll-aboration, McNeely ML, eds. Cochrane Database Syst Rev. 2010. Available at: http://summaries.cochrane.org/CD005211/exercise-interventions- 4. Cheema B, Gaul CA, Lane K, Fiatarone Singh MA. Progressive resistance training in for-upper-limb-dysfunction-due-to-breast-cancer. Accessed December 20, 2011. breast cancer: a systematic review of clinical trials. Breast Cancer Res Treat. 2008;109(1):9–26. http://dx.doi.org/10.1007/s10549-007-9638-0. 32. Centre for Reviews and Dissemination. Systematic Reviews CRD’s guidance for undertaking reviews in health care. York, UK: CRD, University of York; 2009.. 5. Hayes S, Sipio TD, Rye S, Lopez JA, Saunders C, et al. Prevalence and prognostic Accessed July 20, 2012 http://www.york.ac.uk/inst/crd/pdf/Systematic_ significance of secondary lymphedema following breast cancer. Lymphat Res Biol. Reviews.pdf. 2011;9(3):135–141. http://dx.doi.org/10.1089/lrb.2011.0007. 33. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JPA, et al. The 6. Hayes SC, Johansson K, Stout NL, Prosnitz R, Armer JM, Gabram S, et al. Upper-body PRISMA Statement for Reporting Systematic Reviews and Meta-Analyses of Studies morbidity after breast cancer: incidence and evidence for evaluation, prevention, That Evaluate Health Care Interventions: Explanation and Elaboration. PLoS Med. and management within a prospective surveillance model of care. Cancer. 2009;6(7):e1000100. http://dx.doi.org/10.1371/journal.pmed.1000100. 2012;118(8 Suppl):2237–2249. http://dx.doi.org/10.1002/cncr.27467. 34. Maher CG, Sherrington C, Herbert RD, Moseley AM, Elkins M. Reliability of the 7. Woods M. Lymphoedema care. Oxford: Blackwell, c; 2007. PEDro Scale for Rating Quality of Randomized Controlled Trials. Phys Ther. 8. Britton TMB, Purushotham AD. Understanding breast cancer-related lymphoe- 2003;83(8):713–721. dema. Surg Edinb Univ Press. 2009;7(2):120–124. 35. De Morton NA. The PEDro scale is a valid measure of the methodological quality of 9. Hayes SC, Cornish B, Newman B. Comparison of methods to diagnose lymphoe- clinical trials: a demographic study. Aust J Physiother. 2009;55(2):129–133. dema among breast cancer survivors: 6-month follow-up. Breast Cancer Res Treat. 36. Higgins JP, Deeks JJ. Chapter 7: Selecting studies and collecting data. In: Higgins JP, 2005;89(3):221–226. http://dx.doi.org/10.1007/s10549-004-2045-x. Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions. 5.1.0 ed. 10. Hayes SC, Janda M, Cornish B, Battistutta D, Newman B. Lymphedema after breast 5.1.0. The Cochrane Coll-aboration; 2011. Available at: http://www.cochrane- cancer: incidence, risk factors, and effect on upper body function. J Clin Oncol. handbook.org/. Accessed December 22, 2011 2008;26(21):3536–3542. 11. Cheifetz O, Haley L. Management of secondary lymphedema related to breast 37. Verbeek J, Ruotsalainen J, Hoving JL. Synthesizing study results in a systematic cancer. Can Fam Physician. 2010;56(12):1277–1284. review. Scand J Work Environ Health. 2012. http://dx.doi.org/10.5271/sjweh.3201. 12. McLaughlin SA, Wright MJ, Morris KT, Giron GL, Sampson MR, Brockway JP, et al. Prevalence of Lymphedema in Women With Breast Cancer 5 Years 38. Deeks JJ, Higgins JP, Altman DG. Chapter 9: Analysing data and undertaking meta- After Sentinel Lymph Node Biopsy or Axillary Dissection: Objective Measurements. analyses. In: Higgins JP, Green S, editors. Cochrane Handbook for Systematic Reviews J Clin Oncol. 2008;26(32):5213–5219. http://dx.doi.org/10.1200/JCO.2008.16. of Interventions. 5.1.0 ed. 5.1.0. The Cochrane Coll-aboration; 2011. Available at: 3725. http://www.cochrane-handbook.org/. Accessed December 22, 2011. 13. Petrek JA, Senie RT, Peters M, Rosen PP. Lymphedema in a cohort of breast carcinoma survivors 20 years after diagnosis. Cancer. 2001;92(6):1368–1377. http://dx.doi.org/ 39. Ahmed RL, Thomas W, Yee D, Schmitz KH. Randomized Controlled Trial of Weight 10.1002/1097-0142(20010915). 92:6<1368::AID-CNCR1459>3.0.CO;2-9. Training and Lymphedema in Breast Cancer Survivors. J Clin Oncol. 14. Johansson K, Branje E. Arm lymphoedema in a cohort of breast cancer survivors 10 2006;24(18):2765–2772. http://dx.doi.org/10.1200/JCO.2005.03.6749. years after diagnosis. Acta Oncol Stockh Swed. 2010;49(2):166–173. http:// dx.doi.org/10.3109/02841860903483676. 40. Anderson RT, Kimmick G, McCoy T, Hopkins J, Levine E, Miller G, et al. A random- 15. McKenzie DC. Abreast in a Boat – a race against breast cancer. CMAJ Can Med Assoc J. ized trial of exercise on well-being and function following breast cancer surgery: 1998;159(4):376–378. the RESTORE trial. J Cancer Surviv. 2012;6(2):172–181. http://dx.doi.org/10.1007/ 16. Johansson K, Tibe K, Weibull A, Newton RC. Low intensity resistance exercise for s11764-011-0208-4. breast cancer patients with arm lymphedema with or without compression sleeve. Lymphology. 2005;38(4):167–180. 41. Ohira T, Schmitz KH, Ahmed RL, Yee D. Effects of weight training on quality of life in 17. Lane KN, Dolan LB, Worsley D, McKenzie DC. Upper extremity lymphatic function recent breast cancer survivors. Cancer. 2006;106(9):2076–2083. http://dx.doi.org/ at rest and during exercise in breast cancer survivors with and without lymph- 10.1002/cncr.21829. edema compared with healthy controls. J Appl Physiol. 2007;103(3):917–925. http://dx.doi.org/10.1152/japplphysiol.00077.2007. 42. Schmitz KH, Ahmed RL, Hannan PJ, Yee D. Safety and Efficacy of Weight Training in 18. Kwan ML, Cohn JC, Armer JM, Stewart BR, Cormier JN. Exercise in patients Recent Breast Cancer Survivors to Alter Body Composition, Insulin, and Insulin- with lymphedema: a systematic review of the contemporary literature. J Cancer Like Growth Factor Axis Proteins. Cancer Epidemiol Biomarkers Prev. 2005;14(7): Surviv Res Pract. 2011;5(4):320–336. http://dx.doi.org/10.1007/s11764-011- 1672–1680. http://dx.doi.org/10.1158/1055-9965.EPI-04-0736. 0203-9. 19. Schmitz KH. Balancing lymphedema risk: exercise versus deconditioning for breast 43. Speck RM, Gross CR, Hormes JM, Ahmed RL, Lytle LA, Hwang W-T, Schmitz KH. cancer survivors. Exerc Sport Sci Rev. 2010;38(1):17–24. http://dx.doi.org/10.1097/ Changes in the Body Image and Relationship Scale following a one-year strength JES.0b013e3181c5cd5a. training trial for breast cancer survivors with or at risk for lymphedema. Breast Cancer 20. Breast Cancer Care. Risk of lymphoedema j Breast Cancer Care. 2011. Available at: Res Treat. 2010;121(2):421–430. http://dx.doi.org/10.1007/s10549-009-0550-7. http://www.breastcancercare.org.uk/breast-cancer-information/treating-breast- cancer/lymphoedema/risk-lymphoedema. Accessed December 16, 2011 44. Sagen A˚ rn, Ka˚ resen R, Risberg M. Physical activity for the affected limb and arm 21. Schmitz KH, Ahmed RL, Troxel A, Cheville A, Smith R, Lewis-Grant L, et al. Weight lymphedema after breast cancer surgery. A prospective, randomized controlled Lifting in Women with Breast-Cancer-Related Lymphedema. N Engl J Med. trial with two years follow-up. Acta Oncol. 2009;48(8):1102–1110. http:// 2009;361(7):664–673. http://dx.doi.org/10.1056/NEJMoa0810118. dx.doi.org/10.3109/02841860903061683. 22. Schmitz KH, Ahmed RL, Troxel AB, Cheville A, Smith R, Lewis-Grant L, et al. Weight Lifting for Women at Risk for Breast Cancer-Related Lymphedema: A Randomized 45. Kilbreath SL, Refshauge K, Beith J, Ward LC, Simpson JM, Hansen RD. Upper limb Trial. JAMA J Am Med Assoc. 2010;304(24):2699–2705. http://dx.doi.org/10.1001/ progressive resistance training and stretching exercises following surgery for early jama.2010.1837. breast cancer: a randomized controlled trial. Breast Cancer Res Treat. 23. American College of Sports Medicine. ACSM’s guidelines for exercise testing and 2012;133(2):667–676. http://dx.doi.org/10.1007/s10549-012-1964-1. prescription. 8th ed. London: Lippincott Williams & Wilkins; 2009. 24. Garber CE, Blissmer B, Deschenes MR, Franklin BA, Lamonte MJ, Lee IM, et al. 46. Kim DS, Sim Y-J, Jeong HJ, Kim GC. Effect of active resistive exercise on breast American College of Sports Medicine position stand. Quantity and quality of cancer-related lymphedema: a randomized controlled trial. Arch Phys Med Rehabil. exercise for developing and maintaining cardiorespiratory, musculoskeletal, and 2010;91(12):1844–1848. http://dx.doi.org/10.1016/j.apmr.2010.09.008. neuromotor fitness in apparently healthy adults: guidance for prescribing exercise. Med Sci Sports Exerc. 2011;43(7):1334–1359. http://dx.doi.org/10.1249/ 47. Kilbreath SL, Refshauge KM, Beith JM, Ward LC, Simpson JM, Hansen RD. Progres- MSS.0b013e318213fefb. sive resistance training and stretching following surgery for breast cancer: study protocol for a randomised controlled trial. BMC Cancer. 2006;6:273. http:// dx.doi.org/10.1186/1471-2407-6-273. 48. Bartolucci AA, Hillegass WB. Overview, Strengths, and Limitations of Systematic Reviews and Meta-Analyses. In: Chiappelli F, ed. In: Evidence-Based Practice: Toward Optimizing Clinical Outcomes. Berlin, Heidelberg: Springer Berlin Heidel- berg; 2010:17–33. Accessed February 12, 2014 http://link.springer.com/10.1007/ 978-3-642-05025-1_2. 49. Fong DYT, Ho JWC, Hui BPH, Lee AM, Macfarlane DJ, Leung SSK, et al. Physical activity for cancer survivors: meta-analysis of randomised controlled trials. BMJ. 2012;344. http://dx.doi.org/10.1136/bmj.e70.

Journal of Physiotherapy 60 (2014) 178 Journal of PHYSIOTHERAPY journal homepage: www.elsevier.com/locate/jphys Appraisal Media Well-designed website for new clinicians provides information and develops problem-solving eLearnSCI website: www.elearnSCI.org Supporting evidence Each Physiotherapy sub-module includes a list of references. These eLearnSCI was created to prepare students and junior clinicians in various health-care professions to work with patients who have spinal lists present a small number (8 to 31) of relevant publications, many of cord injuries (SCI). The website contains a module that presents an which are reviews or widely accepted clinical practice guidelines. Most of overview of basic principles of SCI management, and an additional the references are current and appropriate, but some important ones are module addressing prevention. The remainder of the website is missing. For example, the reference list for Gait is missing a landmark organised by profession. Separate learning modules exist for doctors, study comparing weight-supported treadmill training to conventional nurses, occupational therapists and assistive technologists, physiothera- gait training,1 and does not include a recent report of a useful clinical pists, psychologists, social workers, and peer counselors that present prediction rule for predicting ambulation outcomes.2 Despite these information relevant to each profession. Links between the modules minor shortcomings, however, the content in the Physiotherapy module provide easy access to information across disciplines. This multi- of eLearnSCI reflects the best evidence currently available. professional content with links between professional modules appro- priately exposes new clinicians to the complex and multi-faceted nature Global perspective of SCI, and underscores need for multiple disciplines to work as a team. eLearnSCI was developed by a large team of experts from around the The Physiotherapy module provides a broad foundation of informa- world. The images, videos, and case studies present SCI survivors and tion that is essential for therapists working with patients with SCI. A therapists from a variety of countries, with both high- and low- sub-module on ‘The health essentials’ presents basic information on resourced settings represented. This diversity of perspective makes for the neurological and physiological sequelae of SCI, neurological an interesting site with broad relevance. The excellent content and classification, and medical and surgical management. Additional skillful presentation of this site will make it an important resource for sub-modules give an overview of physiotherapy management, and physical therapists working in a variety of settings. explain principles of assessment, goal-setting, program design, functional training, gait training, upper limb function, therapeutic Living document exercise, motor skill assessment and training, fitness, respiratory care, The website is structured in a way that encourages feedback: each and pain management. Each of these sub-modules includes an overview and basic information, with links to activities that provide slide includes a ‘comment’ button, and submitted suggestions receive more detailed explanations and develop clinical reasoning skills. The prompt responses. For example, a form was updated to a newer version module is practical and comprehensive, imparting sufficient informa- within days of the suggested change. A more complex suggestion tion and problem-solving strategies to enable a novice physiotherapist regarding the addition of content on pressure relief training quickly to work effectively with patients with SCI. The one notable exception is received a positive response expressing the intention to expand this pressure reliefs in a wheelchair. In terms of health and survival, the material. This responsiveness to comments makes it clear that the website ability to perform effective pressure reliefs may be the most important can be expected to remain current and continue to improve over time. functional skill for a patient with SCI to master. Although the Physiotherapy module has links to some information on pressure Something for everyone ulcers and pressure reliefs in the Nursing and Occupational Therapy This informative and useful website was created primarily to equip modules, this important topic could easily be missed by a novice clinician who doesn’t know to seek it out. students and junior clinicians to work effectively with SCI survivors. It accomplishes this goal through simple explanations of the most critical Effective teaching strategies concepts related to SCI, and by providing a variety of excellent learning It is clear that the website was designed to teach new clinicians how activities that develop clinical problem-solving skills. The site’s concise format and focus on clinically relevant topics makes it ideal for busy to work effectively with patients with SCI, as opposed to simply clinicians and students. For those who have the time and inclination to conveying information. Several features of the site support this goal. pursue additional knowledge, there are links to select websites that Critical, need-to-know concepts are clearly explained in simple provide more in-depth coverage of certain topics such as neurological language. The presentations in eLearnSCI are richly and effectively classification. illustrated with drawings, photographs, and videos. Clinical reasoning is explicitly explained and reinforced in each sub-module with case Although eLearnSCI is designed for health professionals who are studies and numerous interactive activities that engage the learner in unfamiliar with SCI, it may have broader appeal. Even seasoned analysis and application of the concepts presented. For example, the therapists might discover new ideas among the useful tips from functional training units include videos of patients having difficulty clinicians around the world. If nothing else, an experienced therapist performing various tasks such as rolling from supine, or negotiating a may come across an unfamiliar technique for performing a particular step in a wheelchair. Learners are encouraged to analyse the videos functional task, gain fresh ideas for therapeutic exercise, or learn new and come up with their own ideas about causes and solutions for the ways to clearly explain concepts. problems shown. In addition to the interactive problem-solving activities, each sub-module concludes with self-assessment ques- Martha F Somers tions. For both the self-assessment questions and the interactive Rangos School of Health Sciences, Duquesne University, Pittsburgh, activities, feedback is provided immediately after the answers are submitted. Pennsylvania, USA References 1. Dobkin B, Apple D, Barbeau H, et al. Neurology. 2006;66:484–493. 2. Van Middendorp JJ, Hosman AJF, et al. Lancet. 2011;377:1004–1010. http://dx.doi.org/10.1016/j.jphys.2014.07.002 1836-9553/ß 2014 Australian Physiotherapy Association. Published by Elsevier B.V. All rights reserved.