- Email: [email protected]
Influence of upper- and lower-limb exercise training on cardiovascular function and walking distances in patients with intermittent claudication
Influence of upper- and lower-limb exercise training on cardiovascular function and walking distances in patients with intermittent claudication Richard D. Walker, PhD, Shah Nawaz, FRCS, Claire H. Wilkinson, BSc, John M. Saxton, PhD, A. Graham Pockley, PhD, and Richard F. M. Wood, MD, FRCS, Sheffield, United Kingdom Purpose: The effects of upper-limb (arm cranking) and lower-limb (leg cranking) exercise training on walking distances in patients with intermittent claudication was assessed. Methods: Sixty-seven patients (33 to 82 years old) with moderate to severe intermittent claudication were recruited, and the maximum power generated during incremental upper- and lower-limb ergometry tests was determined, as were pain-free and maximum walking distances (by using a shuttle walk test). Patients were randomly assigned to an upper-limb training group (n = 26) or a lower-limb training group (n = 26). An additional untrained group (n = 15) was recruited on an ad hoc basis in parallel with the main trial by using identical inclusion criteria. This group was subsequently shown to possess a similar demographic distribution to the two exercise groups. Supervised training sessions were held twice weekly for 6 weeks. Results: Both training programs significantly improved the maximum power generated during the incremental upper- and lower-limb ergometry tests (P < .001), which may reflect an increase in central cardiovascular function that was independent of the training mode. More importantly, pain-free and maximum walking distances also improved in both training groups (P < .001). The improvements in the training groups were similar; there were no changes in the untrained control group. These findings suggest that the symptomatic improvement after upper-limb exercise training may result, in part, from systemic cardiovascular effects rather than localized metabolic or hemodynamic changes. Conclusion: Carefully prescribed upper-limb exercise training can evoke a rapid symptomatic improvement in patients with claudication, while avoiding the physical discomfort experienced when performing lower-limb weight-bearing exercise. (J Vasc Surg 2000;31:662-9.)
Previous studies have shown that patients with intermittent claudication who exercise can increase their walking distance1-5; however, the mechanism by which this symptomatic improvement occurs is unclear. Several factors, including improvement in the From the Division of Clinical Sciences, Northern General Hospital (Drs Walker, Nawaz, Pockley, and Wood and Ms Wilkinson) and the Sheffield Institute of Sports Medicine and Exercise Science (Dr Saxton). Competition of interest: nil. Supported by the British Heart Foundation (PG/98105) and a grant from the Northern General Hospital NHS Trust Research Committee. Reprint requests: Professor Richard F.M. Wood, Division of Clinical Sciences (NGH), Clinical Sciences Centre, Northern General Hospital, Herries Rd, Sheffield S5 7AU, UK. Copyright © 2000 by The Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter. 0741-5214/2000/$12.00 + 0 24/1/104104 doi:10.1067/mva.2000.104104
662
collateral circulation, changes in the local metabolic efficiency of ischemic tissues, and enhanced cardiovascular function, may be involved.2,3 The effect of longterm exercise training on patients with intermittent claudication remains unclear, as does the form of exercise that these patients should be advised to undertake. To date, few studies have rigorously examined whether exercise-induced improvements in walking distance result from a generalized increase in central cardiovascular function or from more localized physiological changes in the lower limb. This study was designed to establish whether improving cardiovascular function with an upper-limb exercise training regimen has a beneficial effect on walking distance in patients with intermittent claudication. If this were the case, then the prescription of an upper-limb exercise program would avoid the physical discomfort associated with lower-limb exercise training. Two training schemes were devised, an upper-limb and a
JOURNAL OF VASCULAR SURGERY Volume 31, Number 4
Walker et al
663
Table I. Baseline characteristics of the study groups Variable Sex Men, % Women, % Age, years Duration of claudication, months Ankle-brachial pressure index Walking distances Pain-free, meters Maximum, meters Current angina, % Previous myocardial infarction, % Stroke, % Diabetes mellitus, % Smoking history Current, % Previous, % None, % Hyperlipidemia (cholesterol > 6.5 mmol/L), %
Untrained (n = 15)
60 40 71 (70 to 77) 24 (24 to 36) 0.71 (0.61 to 0.79) 142 (108 to 180) 272 (197 to 342) 33 30 7 26 13 69 18 40
Upper-limb (n = 26)
83 17 68 (63 to 70) 24 (18 to 33) 0.67 (0.56 to 0.73) 130 (82 to 220) 267 (188 to 340) 8 10 8 25 38 51 11 46
Lower-limb (n = 26)
83 17 70 (61 to 73) 24 (15 to 31) 0.64 (0.60 to 0.8) 147 (120 to 220) 228 (199 to 341) 15 15 8 13 42 41 17 50
Significance NSχ NSΚ NSΚ NSΚ NSΚ NSΚ NSχ NSχ NSχ NSχ NSχ NS
χ
Values are median (Q25 to Q75). Group differences were determined with the test indicated (NS, not significant, P > .05; , χ, chi-square; Kruskal-Wallis). Current smoking status was confirmed by means of CO analysis (Smokerlyzer, Bedfont, UK). Patients with diabetes mellitus were taking oral hypoglycemics or insulin.
Κ,
lower-limb program. Recruited patients were randomly assigned to either the upper- or the lower-limb program. The training schedules were devised to stimulate an equivalent cardiovascular response (as measured by means of the heart rate) for both modes of exercise. This was achieved by precisely controlling the work rate with calibrated upper- and lower-limb cycle ergometers. It would not have been possible to achieve a valid comparison with a group undertaking structured walking exercise. First, the relationship between walking speed and heart rate is nonlinear, rendering an accurate prescription of work rate difficult. Second, the primary output measurement of this trial was the ability to walk, and the learning curve associated with a walking program would make separation of any cardiovascular changes impossible. Previously, investigators have recommended that a 6-month study, in which participants exercised three times a week, should be used to achieve the maximum benefit from a training program.1 However, effects are also achievable in shorter, lower frequency studies.1 This study tested whether an upper-limb training program could elicit a change in walking distance. A 6-week program was used, because we were not attempting to develop an optimized training program for a claudicant group. SUBJECTS AND METHODS Patients. A total of 67 patients (33 to 82 years old) with stable intermittent claudication were recruit-
ed from the vascular clinic at the Northern General Hospital NHS Trust, Sheffield, UK. Patients were excluded from the trial if they had experienced symptoms for fewer than 12 months, undergone a revascularization process within the last 12 months, or had exercise-limiting angina, shortness of breath, or severe arthritis. The median ankle-brachial pressure index (ABPI) for all patients was 0.68 (range, 0.6 to 0.75). The ABPI was taken as the highest index value observed in the more symptomatic limb. Further demographic data are included in Table I. This study was approved by the North Sheffield Local Research Ethics Committee, and all patients gave their informed consent before entering the program. Determination of claudication and maximum walking distances. The functional capabilities of the lower limb are typically assessed by using treadmill walking distance measurements. However, in our clinical practice, less than half the patients with claudication who are referred to us are able to undertake a standard treadmill test at a walking speed of 4 km.h-1. This is mainly because of their concern about falling or an inability to cope with the walking speed required by the test. Because of these difficulties with the treadmill, patients were examined by means of a shuttle walk test.6 This test offers the advantages that it involves walking on flat ground at incrementally increasing speeds and it can accommodate the simultaneous assessment of more than one patient. For this test, patients walked back and forth, between
664 Walker et al
two cones spaced 10 m apart on a flat gymnasium floor. The speed at which the patient walked was controlled by means of a series of audible tones recorded on a cassette. The patients began walking on the first tone, and they aimed to reach the 10-m cone by the next tone. Patients then turned and returned to the first cone, and the process was repeated. The patients generally achieved the correct pace within three laps (30 m). The initial speed at which the patients walked was 3 km.h-1, and the interval between audible tones gradually decreased throughout the test, resulting in an incremental increase of 0.5 km.h-1 in the required walking speed every minute. The test was completed when patients could no longer maintain the required pace. The time and distance to claudication (CD) and the maximum walking distance (MWD) were determined. The shuttle test is a reliable measurement of the MWD in patients with chronic airways obstruction,6 and we have found it to have a low (less than 10%) variability in CD and MWD measurements in patients with claudication, as determined from three measurements made on separate days (S.N., unpublished observations). Assessment of cardiovascular fitness. The cardiovascular fitness of all patients was assessed before and after the training program by using frictionbraked upper-limb and lower-limb cycle ergometry (881E rehab trainer and 824E cycle ergometer, Monark, Varberg, Sweden). All patients were familiarized with the testing apparatus before commencement of the study. After a 2 minute warm-up against no resistance, the workload was increased by 10 W (upper limb) or 25 W (lower limb) at 3-minute intervals by increasing the frictional load applied to the flywheel, while patients maintained a constant rate of 50 revs.min-1. This process was repeated until the patient was unable to continue, and the final outputs from each test were recorded as the maximum arm power (MAP) and maximum leg power (MLP). The maximum achieved power was recorded as the power at which the patient failed, regardless of how long the patient had been exercising at this power. The workload increments applied in each test reflect the different muscle mass involved in the cranking action (upper limb versus lower limb), and these were determined on the basis of a pilot study in five patients with claudication. Heart rate was measured during the final minute of each power increment for each assessment, either manually or by using a continuous output heart rate monitor (Sports Tester PE-3000, Polar Electro, Finland). A linear
JOURNAL OF VASCULAR SURGERY April 2000
increase in heart rate occurs with work rate in healthy individuals, and this was also observed in our patients. Cardiovascular exercise training produces a decrease in heart rate at the submaximal work rates, defined at the preexercise assessment, because of improved stroke volume.7 The heart rate versus work rate relationship for incremental arm and leg cranking was examined before and after training to assess changes in the cardiovascular fitness of the patients. All patients were assessed by using both the upperand lower-limb ergometry protocols, irrespective of the training program to which they were subsequently assigned. Tests were performed in random order at least 1 day apart by an independent observer. Training programs. After the assessment of physical work capacity (arm and leg cranking) and walking distances, each patient was randomly assigned to either a lower-limb (n = 26) or upper-limb (n = 26) training group. Supervised exercise training was undertaken twice a week for 6 weeks. Members of an additional untrained group (n = 15) were recruited on an ad hoc basis in parallel with the main trial, by using identical inclusion criteria. This group was subsequently shown to possess a similar demographic distribution to the exercise groups. Untrained patients were given lifestyle advice, including encouragement to undertake regular exercise, but did not undertake any formal training sessions. This is the advice given to stable patients with claudication in the United Kingdom, and this was the standard against which we wished to assess our two exercise modalities. Upper-limb exercise induces a greater cardiovascular stress (greater heart rate, intra-arterial blood pressure, and pulmonary ventilation) for a given level of submaximal work than lower-limb exercise.7-11 Thus, the incremental upper-limb assessment was necessary as a means of defining training intensity for the upper-limb training group, and the lower-limb assessment was used as a means of prescribing training intensity for the lower-limb training group. The penultimate workload achieved in the respective assessment was used as the initial training intensity, because this strategy enabled a similar cardiovascular stress to be applied to each patient during the training period. By using the penultimate load, we ensured that the patients were working near to cardiac capacity, but only for short intervals during the training sessions. For each of the supervised training sessions, patients trained against this load in cycles of 2 minutes of exercise, followed by 2 minutes of rest, for a total exercise time of 20 minutes in a 40-minute session. Interval training regimens such as this in healthy patients enable a greater amount of higher-intensity
JOURNAL OF VASCULAR SURGERY Volume 31, Number 4
Walker et al
665
work to be performed in a given time than could be achieved with continuous exercise of a similar nature7 and, therefore, optimizes the stimulus for cardiovascular adaptations. The MAP or MLP workloads identified during the initial assessment were applied after 3 weeks of training to ensure a sufficient stimulus for continued improvement in cardiovascular fitness. The maximum power achieved by the arms and legs and the walking distances were determined immediately after the training program was completed. Quality of life assessment. All patients were assessed at the beginning and end of the study period with the SF36 (Medical Outcomes Study) questionnaire.12 Statistical analysis. Data are presented as medians and first and third quartiles (Q25 and Q75). The level of statistical significance was set at a P value of .05. Data typically displayed skewness, and comparisons of variables between groups were performed with the Wilcoxon signed rank test for paired data. Categorical variables were compared with the χ2 and Kruskal-Wallis tests. Patients dropping out of the trial were included in the baseline comparisons to test for differences between groups. They were not entered into the analysis of the effects of training because of the paired techniques used. RESULTS Patient attendance and compliance. Of the 67 patients recruited into this study, two patients dropped out of each of the upper- and lower-limb training groups. Shingles developed in one patient, and another patient was unable to arrange transportation. A third patient had chronic depression and did not continue because of anxiety about the training, and a fourth patient dropped out for unknown reasons. Attendance at the pretraining and posttraining assessments was 100% for all three groups, and attendance at the training sessions was 98% for the upper-limb group and 94% for the lower-limb group. Pretraining characteristics of control and patient groups. Before training, the MAP of the patients was 20 W (range, 15 to 35 W), and the MLP of the patients was 75 W (range, 50 to 100 W). The median CD and MWD of the patients was 142 m (range, 101 to 212 m) and 170 m (range, 120 to 301 m), respectively. Details of the CD and MWD for the individual groups are shown in Table I. The pretraining power outputs and walking distances of the three groups were not different. Posttraining cardiovascular characteristics Upper-limb ergometry assessments. The MAP
Fig 1. Effects of upper- and lower-limb exercise training on the maximum achieved arm power of individual patients (upper) and the group as a whole (lower). Data are presented as medians and 25th and 75th quartiles. *P < .001 vs pre-training, Wilcoxon signed rank test.
increased in all 24 patients who completed upperlimb training, whereas the MAP of only 19 of the 24 patients in the lower-limb training group improved (Fig 1). Overall, the improvements in both groups were statistically significant (P < .001; Fig 1). No significant change occurred in the untrained group, as a whole. Lower-limb ergometry assessments. Nineteen of the 24 patients in the upper-limb training group and 20 of the 24 patients in the lower-limb training group improved their maximum achieved power output in the lower-limb assessment. These improvements were of statistical significance for both groups (P < .001; Fig 2). No significant improvements were
JOURNAL OF VASCULAR SURGERY April 2000
666 Walker et al
Untrained
Fig 2. Effects of upper- and lower-limb exercise training on the maximum achieved leg power of individual patients (upper) and the group as a whole (lower). Data are presented as medians and 25th and 75th quartiles. *P < .001 vs pre-training, Wilcoxon signed rank test.
observed in the untrained group (P = 1.0). Upperlimb training improved lower-limb performance in this assessment. Heart rate response to exercise. The heart rate response to submaximal work loads during the upperand lower-limb assessments was reduced after training, confirming an improvement in the cardiovascular system after both modes of exercise training (Fig 3). Posttraining walking distances. The CD (pain-free walking distance) increased by 122% in the upper-limb training group (P < .001) and 93% in the lower-limb (P < .001) training group (Fig 4). The MWD increased by 47% in the upper-limb group (P = .003) and 50% in the lower-limb group
Fig 3. Heart rates at a given work rate for the upper-limb (upper) and lower-limb (lower) assessments before (diamond) and after (square) exercise training. Data are median values for the given group. Zero power reflects the internal resistance of the arm and leg ergometers. Statistical differences between the groups are indicated (Wilcoxon signed rank test).
(P < .001; Fig 5). The improvements in the claudication and MWDs observed in the upper- and lowerlimb groups were not significantly different from each other (Mann-Whitney U test). No change in either of these parameters occurred in the untrained group (Figs 4, 5). Quality of life. The “effect size” for each of the eight domains assessed in the SF36 questionnaire was calculated. This was achieved by dividing the difference between the pretreatment and posttreatment values by the interquartile range of the pretreatment value.12 The effect size is of value in demonstrating the importance of a treatment effect within a study.
JOURNAL OF VASCULAR SURGERY Volume 31, Number 4
Fig 4. Effects of upper and lower limb exercise training on pain-free (claudication) distance in individuals (upper) and the group as a whole (lower). Data are medians and 25th and 75th quartiles. *P < .001 vs pretraining, Wilcoxon signed rank test.
The “physical functioning” and “role limitationphysical” domains improved significantly after upper-limb and lower-limb training (Fig 6). No significant change in the other domains occurred. No significant changes in any of the quality of life domains occurred in the untrained group (Fig 6). DISCUSSION Exercise is well established as a treatment modality in patients with claudication, although the mechanisms by which walking performance improves are not well understood twice a day. This study demonstrates that training twice a week in a 6-week period significantly increases the walking distance of patients with intermittent claudication. This program is considerably shorter than the three-times-weekly protocols for 6 months that have been recommended pre-
Walker et al
667
Fig 5. Effects of upper and lower limb exercise training on maximum walking distances in individuals (upper) and the group as a whole (lower). Data are medians and 25th and 75th quartiles. *P < .001 vs pretraining, Wilcoxon signed rank test.
viously.1,3 Our findings suggest that training regimens based on short, higher-intensity work with interpolated rest periods are more efficient at producing rapid, symptomatic improvement. This may be because of the increased cardiovascular stress, which provides a better stimulus for exercise-induced adaptations. None of our patients reported any adverse effects from this training system, although 53.9% of patients in the leg training group experienced calf pain, which required a reduction in the load applied, during the later stages of exercise sessions. Unfortunately, the availability of exercise programs in the United Kingdom is limited. Additional problems occur with achieving high attendance rates, because a large proportion of patients are
JOURNAL OF VASCULAR SURGERY April 2000
668 Walker et al
Fig 6. Quality of life changes after exercise training. Data are presented as medians, and comparisons between pretraining and posttraining values were made with the Wilcoxon signed rank test. GH, general health; PF, physical functioning; RLP, role limitation–physical; RLE, role limitation–emotional; SF, social functioning; BP, bodily pain; EV, energy and vitality; MH, mental health.
unable to manage two to three sessions per week for prolonged periods.4,13,14 The excellent attendance and low dropout rate experienced in the present study was partly due to the comparatively short training program. Other contributory factors were group training sessions, a flexible approach to the time and day that each patient could attend, and the provision of regular advice and encouragement. The most important finding of this study is that both upper- and lower-limb exercise training elicits similar symptomatic improvements in patients with claudication. These symptomatic changes were accompanied by significant improvements in the “physical functioning” and “role limitation–physical” domains of the SF36 questionnaire in both training groups. This is in keeping with other studies, in which improvements in these domains after exercise training in patients with claudication were reported.15 However, in contrast to another study,16 we found no improvements in the “bodily pain” domain. This may occur because exercise training delays the onset of, rather than eliminates, claudication pain. Furthermore, our results provide an insight into the mechanisms by which cardiovascular exercise training may be beneficial, because they suggest that central cardiovascular adaptations may be a factor in the increased walking distances observed in the upper-limb training group. Increasing the walking distance of patients with
intermittent claudication with exercise training programs has been described previously.1,3,5,17 The physical discomfort encountered by performing lower-limb weight-bearing exercise is undoubtedly a major reason why this strategy has failed to achieve widespread popularity. The prescription of a painfree upper-limb program may provide symptomatic relief to a large number of patients with intermittent claudication. REFERENCES 1. Gardner AW, Poehlman ET. Exercise rehabilitation programs for the treatment of claudication pain. A meta-analysis. JAMA 1995;274:974-80. 2. Lundgren F, Dahllof AG, Schersten T, Bylund-Fellenius AC. Muscle enzyme adaptation in patients with peripheral arterial insufficiency: spontaneous adaptation, effect of different treatments and consequences on walking performances. Clin Sci 1989;11:485-93. 3. Lundgren F, Dahllof AG, Lundholm K, Schersten T, Volkmann R. Intermittent claudication—Surgical reconstruction of physical training? A prospective randomized trial of treatment efficiency. Ann Surg 1989;209:346-55. 4. Johnson EC, Voyles WF, Atterbom HA, Pathak D, Sutton MF, Greene ER. Effects of exercise training on common femoral artery blood flow in patients with intermittent claudication. Circulation 1989;80:11159-72. 5. Andriessen MP, Barendsen GJ, Wouda AA, dePater L. Changes in walking distance in patients with intermittent claudication during six months intensive physical training. Vasa 1989;18:63-8. 6. Singh SJ, Morgan MDL, Scott S, Walters D, Hardman AE.
JOURNAL OF VASCULAR SURGERY Volume 31, Number 4
7. 8. 9.
10.
11.
12.
Development of a shuttle walking test of disability in patients with chronic airways obstruction. Thorax 1992;47:1019-24. Astrand PO, Rodhal K. Textbook of work physiology: physiological bases of exercise. Singapore: McGraw-Hill; 1986. Bobbert A. Physiological comparison of three types of ergometry. J Appl Physiol 1960;15:1007-14. Astrand PO, Ekblom B, Messin R, Saltin B, Stenberg J. Intra-arterial blood pressure during exercise with different muscle groups. J Appl Physiol 1965;20:253-6. Schwade J, Blomqvist CG, Shapiro W. A comparison of the response to arm and leg work in patients with ischemic heart disease. Am Heart J 1967;94:203-8. Stenberg J, Astrand PO, Ekblom B, Royce J, Saltin B. Hemodynamic response to work with different muscle groups. J Appl Physiol 1967;22:61-70. Ware JE Jr, Keller SD, Gandek B, Brazier JE, Sullivan M. Evaluating translations of health status questionnaires. Methods from the IQOLA project. International Quality of Life Assessment. Int J Tech Assessment Health Care 1995; 11:525-51.
Walker et al
669
13. Hiatt WR, Wolfel EE, Meier RH, Regensteiner JG. Superiority of treadmill walking exercise versus strength training for patients with peripheral arterial disease. Circulation 1994;90:1866-73. 14. Womack CJ, Sieminski DJ, Katzel LI, Yataco A, Gardener AW. Improved walking economy in patients with peripheral arterial occlusive disease. Med Sci Sports Exercise 1997; 29:1286-90. 15. Regensteiner JG, Steiner JF, Hiatt WR. Exercise training improves functional status in patients with peripheral arterial disease. J Vasc Surg 1996;23:104-15. 16. Currie IC, Wilson YG, Baird RN, Lamont PM. Treatment of intermittent claudication. The impact on quality of life. Eur J Vasc Endovasc Surg 1995;10:356-61. 17. Smith KL, Skinner J. Exercise training for claudication patients. Sports Med Training Rehab 1992;22:135-7.
Submitted Jul 9, 1999; accepted Oct 18, 1999.
Previous studies have shown that patients with intermittent claudication who exercise can increase their walking distance1-5; however, the mechanism by which this symptomatic improvement occurs is unclear. Several factors, including improvement in the From the Division of Clinical Sciences, Northern General Hospital (Drs Walker, Nawaz, Pockley, and Wood and Ms Wilkinson) and the Sheffield Institute of Sports Medicine and Exercise Science (Dr Saxton). Competition of interest: nil. Supported by the British Heart Foundation (PG/98105) and a grant from the Northern General Hospital NHS Trust Research Committee. Reprint requests: Professor Richard F.M. Wood, Division of Clinical Sciences (NGH), Clinical Sciences Centre, Northern General Hospital, Herries Rd, Sheffield S5 7AU, UK. Copyright © 2000 by The Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter. 0741-5214/2000/$12.00 + 0 24/1/104104 doi:10.1067/mva.2000.104104
662
collateral circulation, changes in the local metabolic efficiency of ischemic tissues, and enhanced cardiovascular function, may be involved.2,3 The effect of longterm exercise training on patients with intermittent claudication remains unclear, as does the form of exercise that these patients should be advised to undertake. To date, few studies have rigorously examined whether exercise-induced improvements in walking distance result from a generalized increase in central cardiovascular function or from more localized physiological changes in the lower limb. This study was designed to establish whether improving cardiovascular function with an upper-limb exercise training regimen has a beneficial effect on walking distance in patients with intermittent claudication. If this were the case, then the prescription of an upper-limb exercise program would avoid the physical discomfort associated with lower-limb exercise training. Two training schemes were devised, an upper-limb and a
JOURNAL OF VASCULAR SURGERY Volume 31, Number 4
Walker et al
663
Table I. Baseline characteristics of the study groups Variable Sex Men, % Women, % Age, years Duration of claudication, months Ankle-brachial pressure index Walking distances Pain-free, meters Maximum, meters Current angina, % Previous myocardial infarction, % Stroke, % Diabetes mellitus, % Smoking history Current, % Previous, % None, % Hyperlipidemia (cholesterol > 6.5 mmol/L), %
Untrained (n = 15)
60 40 71 (70 to 77) 24 (24 to 36) 0.71 (0.61 to 0.79) 142 (108 to 180) 272 (197 to 342) 33 30 7 26 13 69 18 40
Upper-limb (n = 26)
83 17 68 (63 to 70) 24 (18 to 33) 0.67 (0.56 to 0.73) 130 (82 to 220) 267 (188 to 340) 8 10 8 25 38 51 11 46
Lower-limb (n = 26)
83 17 70 (61 to 73) 24 (15 to 31) 0.64 (0.60 to 0.8) 147 (120 to 220) 228 (199 to 341) 15 15 8 13 42 41 17 50
Significance NSχ NSΚ NSΚ NSΚ NSΚ NSΚ NSχ NSχ NSχ NSχ NSχ NS
χ
Values are median (Q25 to Q75). Group differences were determined with the test indicated (NS, not significant, P > .05; , χ, chi-square; Kruskal-Wallis). Current smoking status was confirmed by means of CO analysis (Smokerlyzer, Bedfont, UK). Patients with diabetes mellitus were taking oral hypoglycemics or insulin.
Κ,
lower-limb program. Recruited patients were randomly assigned to either the upper- or the lower-limb program. The training schedules were devised to stimulate an equivalent cardiovascular response (as measured by means of the heart rate) for both modes of exercise. This was achieved by precisely controlling the work rate with calibrated upper- and lower-limb cycle ergometers. It would not have been possible to achieve a valid comparison with a group undertaking structured walking exercise. First, the relationship between walking speed and heart rate is nonlinear, rendering an accurate prescription of work rate difficult. Second, the primary output measurement of this trial was the ability to walk, and the learning curve associated with a walking program would make separation of any cardiovascular changes impossible. Previously, investigators have recommended that a 6-month study, in which participants exercised three times a week, should be used to achieve the maximum benefit from a training program.1 However, effects are also achievable in shorter, lower frequency studies.1 This study tested whether an upper-limb training program could elicit a change in walking distance. A 6-week program was used, because we were not attempting to develop an optimized training program for a claudicant group. SUBJECTS AND METHODS Patients. A total of 67 patients (33 to 82 years old) with stable intermittent claudication were recruit-
ed from the vascular clinic at the Northern General Hospital NHS Trust, Sheffield, UK. Patients were excluded from the trial if they had experienced symptoms for fewer than 12 months, undergone a revascularization process within the last 12 months, or had exercise-limiting angina, shortness of breath, or severe arthritis. The median ankle-brachial pressure index (ABPI) for all patients was 0.68 (range, 0.6 to 0.75). The ABPI was taken as the highest index value observed in the more symptomatic limb. Further demographic data are included in Table I. This study was approved by the North Sheffield Local Research Ethics Committee, and all patients gave their informed consent before entering the program. Determination of claudication and maximum walking distances. The functional capabilities of the lower limb are typically assessed by using treadmill walking distance measurements. However, in our clinical practice, less than half the patients with claudication who are referred to us are able to undertake a standard treadmill test at a walking speed of 4 km.h-1. This is mainly because of their concern about falling or an inability to cope with the walking speed required by the test. Because of these difficulties with the treadmill, patients were examined by means of a shuttle walk test.6 This test offers the advantages that it involves walking on flat ground at incrementally increasing speeds and it can accommodate the simultaneous assessment of more than one patient. For this test, patients walked back and forth, between
664 Walker et al
two cones spaced 10 m apart on a flat gymnasium floor. The speed at which the patient walked was controlled by means of a series of audible tones recorded on a cassette. The patients began walking on the first tone, and they aimed to reach the 10-m cone by the next tone. Patients then turned and returned to the first cone, and the process was repeated. The patients generally achieved the correct pace within three laps (30 m). The initial speed at which the patients walked was 3 km.h-1, and the interval between audible tones gradually decreased throughout the test, resulting in an incremental increase of 0.5 km.h-1 in the required walking speed every minute. The test was completed when patients could no longer maintain the required pace. The time and distance to claudication (CD) and the maximum walking distance (MWD) were determined. The shuttle test is a reliable measurement of the MWD in patients with chronic airways obstruction,6 and we have found it to have a low (less than 10%) variability in CD and MWD measurements in patients with claudication, as determined from three measurements made on separate days (S.N., unpublished observations). Assessment of cardiovascular fitness. The cardiovascular fitness of all patients was assessed before and after the training program by using frictionbraked upper-limb and lower-limb cycle ergometry (881E rehab trainer and 824E cycle ergometer, Monark, Varberg, Sweden). All patients were familiarized with the testing apparatus before commencement of the study. After a 2 minute warm-up against no resistance, the workload was increased by 10 W (upper limb) or 25 W (lower limb) at 3-minute intervals by increasing the frictional load applied to the flywheel, while patients maintained a constant rate of 50 revs.min-1. This process was repeated until the patient was unable to continue, and the final outputs from each test were recorded as the maximum arm power (MAP) and maximum leg power (MLP). The maximum achieved power was recorded as the power at which the patient failed, regardless of how long the patient had been exercising at this power. The workload increments applied in each test reflect the different muscle mass involved in the cranking action (upper limb versus lower limb), and these were determined on the basis of a pilot study in five patients with claudication. Heart rate was measured during the final minute of each power increment for each assessment, either manually or by using a continuous output heart rate monitor (Sports Tester PE-3000, Polar Electro, Finland). A linear
JOURNAL OF VASCULAR SURGERY April 2000
increase in heart rate occurs with work rate in healthy individuals, and this was also observed in our patients. Cardiovascular exercise training produces a decrease in heart rate at the submaximal work rates, defined at the preexercise assessment, because of improved stroke volume.7 The heart rate versus work rate relationship for incremental arm and leg cranking was examined before and after training to assess changes in the cardiovascular fitness of the patients. All patients were assessed by using both the upperand lower-limb ergometry protocols, irrespective of the training program to which they were subsequently assigned. Tests were performed in random order at least 1 day apart by an independent observer. Training programs. After the assessment of physical work capacity (arm and leg cranking) and walking distances, each patient was randomly assigned to either a lower-limb (n = 26) or upper-limb (n = 26) training group. Supervised exercise training was undertaken twice a week for 6 weeks. Members of an additional untrained group (n = 15) were recruited on an ad hoc basis in parallel with the main trial, by using identical inclusion criteria. This group was subsequently shown to possess a similar demographic distribution to the exercise groups. Untrained patients were given lifestyle advice, including encouragement to undertake regular exercise, but did not undertake any formal training sessions. This is the advice given to stable patients with claudication in the United Kingdom, and this was the standard against which we wished to assess our two exercise modalities. Upper-limb exercise induces a greater cardiovascular stress (greater heart rate, intra-arterial blood pressure, and pulmonary ventilation) for a given level of submaximal work than lower-limb exercise.7-11 Thus, the incremental upper-limb assessment was necessary as a means of defining training intensity for the upper-limb training group, and the lower-limb assessment was used as a means of prescribing training intensity for the lower-limb training group. The penultimate workload achieved in the respective assessment was used as the initial training intensity, because this strategy enabled a similar cardiovascular stress to be applied to each patient during the training period. By using the penultimate load, we ensured that the patients were working near to cardiac capacity, but only for short intervals during the training sessions. For each of the supervised training sessions, patients trained against this load in cycles of 2 minutes of exercise, followed by 2 minutes of rest, for a total exercise time of 20 minutes in a 40-minute session. Interval training regimens such as this in healthy patients enable a greater amount of higher-intensity
JOURNAL OF VASCULAR SURGERY Volume 31, Number 4
Walker et al
665
work to be performed in a given time than could be achieved with continuous exercise of a similar nature7 and, therefore, optimizes the stimulus for cardiovascular adaptations. The MAP or MLP workloads identified during the initial assessment were applied after 3 weeks of training to ensure a sufficient stimulus for continued improvement in cardiovascular fitness. The maximum power achieved by the arms and legs and the walking distances were determined immediately after the training program was completed. Quality of life assessment. All patients were assessed at the beginning and end of the study period with the SF36 (Medical Outcomes Study) questionnaire.12 Statistical analysis. Data are presented as medians and first and third quartiles (Q25 and Q75). The level of statistical significance was set at a P value of .05. Data typically displayed skewness, and comparisons of variables between groups were performed with the Wilcoxon signed rank test for paired data. Categorical variables were compared with the χ2 and Kruskal-Wallis tests. Patients dropping out of the trial were included in the baseline comparisons to test for differences between groups. They were not entered into the analysis of the effects of training because of the paired techniques used. RESULTS Patient attendance and compliance. Of the 67 patients recruited into this study, two patients dropped out of each of the upper- and lower-limb training groups. Shingles developed in one patient, and another patient was unable to arrange transportation. A third patient had chronic depression and did not continue because of anxiety about the training, and a fourth patient dropped out for unknown reasons. Attendance at the pretraining and posttraining assessments was 100% for all three groups, and attendance at the training sessions was 98% for the upper-limb group and 94% for the lower-limb group. Pretraining characteristics of control and patient groups. Before training, the MAP of the patients was 20 W (range, 15 to 35 W), and the MLP of the patients was 75 W (range, 50 to 100 W). The median CD and MWD of the patients was 142 m (range, 101 to 212 m) and 170 m (range, 120 to 301 m), respectively. Details of the CD and MWD for the individual groups are shown in Table I. The pretraining power outputs and walking distances of the three groups were not different. Posttraining cardiovascular characteristics Upper-limb ergometry assessments. The MAP
Fig 1. Effects of upper- and lower-limb exercise training on the maximum achieved arm power of individual patients (upper) and the group as a whole (lower). Data are presented as medians and 25th and 75th quartiles. *P < .001 vs pre-training, Wilcoxon signed rank test.
increased in all 24 patients who completed upperlimb training, whereas the MAP of only 19 of the 24 patients in the lower-limb training group improved (Fig 1). Overall, the improvements in both groups were statistically significant (P < .001; Fig 1). No significant change occurred in the untrained group, as a whole. Lower-limb ergometry assessments. Nineteen of the 24 patients in the upper-limb training group and 20 of the 24 patients in the lower-limb training group improved their maximum achieved power output in the lower-limb assessment. These improvements were of statistical significance for both groups (P < .001; Fig 2). No significant improvements were
JOURNAL OF VASCULAR SURGERY April 2000
666 Walker et al
Untrained
Fig 2. Effects of upper- and lower-limb exercise training on the maximum achieved leg power of individual patients (upper) and the group as a whole (lower). Data are presented as medians and 25th and 75th quartiles. *P < .001 vs pre-training, Wilcoxon signed rank test.
observed in the untrained group (P = 1.0). Upperlimb training improved lower-limb performance in this assessment. Heart rate response to exercise. The heart rate response to submaximal work loads during the upperand lower-limb assessments was reduced after training, confirming an improvement in the cardiovascular system after both modes of exercise training (Fig 3). Posttraining walking distances. The CD (pain-free walking distance) increased by 122% in the upper-limb training group (P < .001) and 93% in the lower-limb (P < .001) training group (Fig 4). The MWD increased by 47% in the upper-limb group (P = .003) and 50% in the lower-limb group
Fig 3. Heart rates at a given work rate for the upper-limb (upper) and lower-limb (lower) assessments before (diamond) and after (square) exercise training. Data are median values for the given group. Zero power reflects the internal resistance of the arm and leg ergometers. Statistical differences between the groups are indicated (Wilcoxon signed rank test).
(P < .001; Fig 5). The improvements in the claudication and MWDs observed in the upper- and lowerlimb groups were not significantly different from each other (Mann-Whitney U test). No change in either of these parameters occurred in the untrained group (Figs 4, 5). Quality of life. The “effect size” for each of the eight domains assessed in the SF36 questionnaire was calculated. This was achieved by dividing the difference between the pretreatment and posttreatment values by the interquartile range of the pretreatment value.12 The effect size is of value in demonstrating the importance of a treatment effect within a study.
JOURNAL OF VASCULAR SURGERY Volume 31, Number 4
Fig 4. Effects of upper and lower limb exercise training on pain-free (claudication) distance in individuals (upper) and the group as a whole (lower). Data are medians and 25th and 75th quartiles. *P < .001 vs pretraining, Wilcoxon signed rank test.
The “physical functioning” and “role limitationphysical” domains improved significantly after upper-limb and lower-limb training (Fig 6). No significant change in the other domains occurred. No significant changes in any of the quality of life domains occurred in the untrained group (Fig 6). DISCUSSION Exercise is well established as a treatment modality in patients with claudication, although the mechanisms by which walking performance improves are not well understood twice a day. This study demonstrates that training twice a week in a 6-week period significantly increases the walking distance of patients with intermittent claudication. This program is considerably shorter than the three-times-weekly protocols for 6 months that have been recommended pre-
Walker et al
667
Fig 5. Effects of upper and lower limb exercise training on maximum walking distances in individuals (upper) and the group as a whole (lower). Data are medians and 25th and 75th quartiles. *P < .001 vs pretraining, Wilcoxon signed rank test.
viously.1,3 Our findings suggest that training regimens based on short, higher-intensity work with interpolated rest periods are more efficient at producing rapid, symptomatic improvement. This may be because of the increased cardiovascular stress, which provides a better stimulus for exercise-induced adaptations. None of our patients reported any adverse effects from this training system, although 53.9% of patients in the leg training group experienced calf pain, which required a reduction in the load applied, during the later stages of exercise sessions. Unfortunately, the availability of exercise programs in the United Kingdom is limited. Additional problems occur with achieving high attendance rates, because a large proportion of patients are
JOURNAL OF VASCULAR SURGERY April 2000
668 Walker et al
Fig 6. Quality of life changes after exercise training. Data are presented as medians, and comparisons between pretraining and posttraining values were made with the Wilcoxon signed rank test. GH, general health; PF, physical functioning; RLP, role limitation–physical; RLE, role limitation–emotional; SF, social functioning; BP, bodily pain; EV, energy and vitality; MH, mental health.
unable to manage two to three sessions per week for prolonged periods.4,13,14 The excellent attendance and low dropout rate experienced in the present study was partly due to the comparatively short training program. Other contributory factors were group training sessions, a flexible approach to the time and day that each patient could attend, and the provision of regular advice and encouragement. The most important finding of this study is that both upper- and lower-limb exercise training elicits similar symptomatic improvements in patients with claudication. These symptomatic changes were accompanied by significant improvements in the “physical functioning” and “role limitation–physical” domains of the SF36 questionnaire in both training groups. This is in keeping with other studies, in which improvements in these domains after exercise training in patients with claudication were reported.15 However, in contrast to another study,16 we found no improvements in the “bodily pain” domain. This may occur because exercise training delays the onset of, rather than eliminates, claudication pain. Furthermore, our results provide an insight into the mechanisms by which cardiovascular exercise training may be beneficial, because they suggest that central cardiovascular adaptations may be a factor in the increased walking distances observed in the upper-limb training group. Increasing the walking distance of patients with
intermittent claudication with exercise training programs has been described previously.1,3,5,17 The physical discomfort encountered by performing lower-limb weight-bearing exercise is undoubtedly a major reason why this strategy has failed to achieve widespread popularity. The prescription of a painfree upper-limb program may provide symptomatic relief to a large number of patients with intermittent claudication. REFERENCES 1. Gardner AW, Poehlman ET. Exercise rehabilitation programs for the treatment of claudication pain. A meta-analysis. JAMA 1995;274:974-80. 2. Lundgren F, Dahllof AG, Schersten T, Bylund-Fellenius AC. Muscle enzyme adaptation in patients with peripheral arterial insufficiency: spontaneous adaptation, effect of different treatments and consequences on walking performances. Clin Sci 1989;11:485-93. 3. Lundgren F, Dahllof AG, Lundholm K, Schersten T, Volkmann R. Intermittent claudication—Surgical reconstruction of physical training? A prospective randomized trial of treatment efficiency. Ann Surg 1989;209:346-55. 4. Johnson EC, Voyles WF, Atterbom HA, Pathak D, Sutton MF, Greene ER. Effects of exercise training on common femoral artery blood flow in patients with intermittent claudication. Circulation 1989;80:11159-72. 5. Andriessen MP, Barendsen GJ, Wouda AA, dePater L. Changes in walking distance in patients with intermittent claudication during six months intensive physical training. Vasa 1989;18:63-8. 6. Singh SJ, Morgan MDL, Scott S, Walters D, Hardman AE.
JOURNAL OF VASCULAR SURGERY Volume 31, Number 4
7. 8. 9.
10.
11.
12.
Development of a shuttle walking test of disability in patients with chronic airways obstruction. Thorax 1992;47:1019-24. Astrand PO, Rodhal K. Textbook of work physiology: physiological bases of exercise. Singapore: McGraw-Hill; 1986. Bobbert A. Physiological comparison of three types of ergometry. J Appl Physiol 1960;15:1007-14. Astrand PO, Ekblom B, Messin R, Saltin B, Stenberg J. Intra-arterial blood pressure during exercise with different muscle groups. J Appl Physiol 1965;20:253-6. Schwade J, Blomqvist CG, Shapiro W. A comparison of the response to arm and leg work in patients with ischemic heart disease. Am Heart J 1967;94:203-8. Stenberg J, Astrand PO, Ekblom B, Royce J, Saltin B. Hemodynamic response to work with different muscle groups. J Appl Physiol 1967;22:61-70. Ware JE Jr, Keller SD, Gandek B, Brazier JE, Sullivan M. Evaluating translations of health status questionnaires. Methods from the IQOLA project. International Quality of Life Assessment. Int J Tech Assessment Health Care 1995; 11:525-51.
Walker et al
669
13. Hiatt WR, Wolfel EE, Meier RH, Regensteiner JG. Superiority of treadmill walking exercise versus strength training for patients with peripheral arterial disease. Circulation 1994;90:1866-73. 14. Womack CJ, Sieminski DJ, Katzel LI, Yataco A, Gardener AW. Improved walking economy in patients with peripheral arterial occlusive disease. Med Sci Sports Exercise 1997; 29:1286-90. 15. Regensteiner JG, Steiner JF, Hiatt WR. Exercise training improves functional status in patients with peripheral arterial disease. J Vasc Surg 1996;23:104-15. 16. Currie IC, Wilson YG, Baird RN, Lamont PM. Treatment of intermittent claudication. The impact on quality of life. Eur J Vasc Endovasc Surg 1995;10:356-61. 17. Smith KL, Skinner J. Exercise training for claudication patients. Sports Med Training Rehab 1992;22:135-7.
Submitted Jul 9, 1999; accepted Oct 18, 1999.