Sage Journals: Discover world-class research

Abstract

We sought to quantify whether health-related quality of life (HRQoL) is improved through exercise training in people with peripheral artery disease (PAD) and to clarify which prescriptions were optimal for improving HRQoL when compared to usual care. We conducted a systematic search (PubMed, CINAHL, Cochrane Central Register of Controlled Trials; 1966 – 31 August 2014). We only included randomized controlled trials (RCTs) of exercise training versus usual medical care in persons with PAD that included the Walking Impairment Questionnaire (WIQ) and Short-Form Health Survey component summary scores as outcomes. Of 15 RCTs, 1257 participants were studied: 543 participated in supervised exercise, with only 61 undertaking resistance training and 316 unsupervised exercise. When compared to controls, participants who completed any form of exercise training significantly improved their WIQ speed [mean difference (MD) 9.60 (95% CI 6.98 to 12.23, p<0.00001)]; WIQ distance [MD 7.41 (95% CI 4.49 to 10.33, p<0.00001)] and WIQ stair-climbing [MD 5.07 (95% CI 3.16 to 6.99, p<0.00001)]. Walking also significantly improved the Short-Form Physical Component Summary (SF-PCS) score when compared to controls [MD 1.24 (95% CI 0.48 to 2.01, p=0.001)], but not the Mental Component Summary (SF-MCS) score [MD –0.55 (95% CI –1.27 to 0.18, p=0.14)]. Exercise training improves the SF-PCS dimension, as well as perceived walking distance, speed and stair-climbing as measured by the WIQ, but not the SF-MCS score. Future studies should aim to blind assessors of such subjective measures, and study alternative modes and prescriptions of exercise alternative to walking.

Keywords

claudication exercise training peripheral artery disease quality of life walking impairment

Introduction

The health-related quality of life (HRQoL) burden in persons with peripheral artery disease (PAD) is as great in magnitude as in patients with other forms of cardiovascular disease.¹ In addition, HRQoL has been shown to predict long-term survival in persons with PAD,² and outcomes from the Walking Impairment Questionnaire (WIQ) have been shown to predict mortality in PAD.³ In addition, persons with PAD have limited exercise capacity, reduced functional performance, and poor cardiorespiratory and musculoskeletal fitness,⁴ which have also been linked to higher all-cause and cardiovascular mortality.⁵ Previous systematic reviews and meta-analyses have shown exercise improves walking ability, fitness and functional performance^4,6–8; however, these reviews have not shown whether exercise training programs within randomized controlled trials (RCTs) also improve HRQoL and perceived walking ability. Although there was a review completed in 2010,⁹ the search was current until 2008 and data from non-RCTs was considered. Moreover, at that time of publication only five RCT’s were available, so meta-analysis was not justified.

There has also been no meta-analysis on RCTs for exercise and PAD that has quantified the magnitude of change in HRQoL and WIQ outcomes attributable to exercise training. We therefore sought to conduct an updated systematic review, while also undertaking appropriate meta-analyses, with a particular focus on quality of life and perceived walking ability outcomes. The aims were to quantify the magnitude of change in exercise training participants versus sedentary controls for improvements in HRQoL and the WIQ, and to identify whether specific prescriptive elements improve HRQoL and WIQ more so than others.

Methods

Search strategy

Potential studies were identified by conducting a systematic search using PubMed, www.ncbi.nlm.nih.gov/pubmed (1966 to 31 August 2014). CINAHL and the Cochrane Central Register of Controlled Trials were also searched (1966 to 31 August 2014). The search strategy included key terms such as peripheral arterial disease, intermittent claudication, peripheral vascular disease, lifestyle therapy, physical training and exercise training. These were combined with a sensitive search strategy to identify RCTs. Reference lists of papers found were scrutinized for new references. All identified papers were assessed independently by two reviewers (BP and GP); a third reviewer (NS) was consulted to resolve disputes. Searches of published papers were also conducted until 31 August 2014. Study quality was assessed using a modified Physiotherapy Evidence Database (PEDro) score.¹⁰ As supervision has previously been deemed an important component of an exercise program for this population,^11,12 supervision was added to the quality criteria score for a maximum score out of 11.

Inclusions

RCTs of exercise training of greater than 4 weeks’ program duration in persons diagnosed with PAD were included. There were no language restrictions. Studies were included if they had a control group that had been placed on usual medical care, with or without exercise advice, and if they measured quality of life outcomes including the WIQ or component scores from the Short-Form (SF) 12, 20 or 36 Health Surveys. HRQoL, as measured via the SF surveys, and perceived walking ability, as measured via WIQ, are both validated questionnaires for measuring aspects of quality of life in PAD.^13,14 The Physical and Mental Component Summary scores from the SF-36, SF-20 and SF-12 have been shown to correlate highly¹⁵ and therefore were merged into one analysis. The WIQ is a PAD-specific measure of self-reported walking limitations that has been validated for use within PAD cohorts.¹⁴ The questionnaire consists of four sub-categories: claudication severity, distance, walking speed and stair-climbing. Owing to only two studies measuring the claudication severity score, it was not suitable for inclusion in this meta-analysis.

Exclusions

Animal studies, review papers and non-RCTs were excluded. Studies that did not have any of the desired outcome measures or used participants without diagnosed PAD in either exercise or control groups were excluded. Several authors were contacted to provide missing data or to clarify if data was duplicated in multiple publications. If authors did not respond, or were unable to provide data, the study was excluded. If a partial data set was deemed to have been previously published, and this was confirmed by the corresponding author, the more complete data set was used in these analyses. Studies using interventions other than exercise, or in addition to exercise (e.g. electro-acupuncture, ultrasound, surgery), were excluded.

Data synthesis

Information on outcome measures was archived in a database. The outcome measures were changes in three items of the WIQ (speed, distance and stair-climbing) and SF-36 Physical and Mental Component Summary scores. The mean difference (MD) was calculated for the outcome measures by subtracting the baseline from post-intervention values (e.g. using the formula MD = post-mean – pre-mean). The standardized mean difference (SMD) calculated as percentage change from baseline was used when different methods to establish the same outcomes were used. If studies reported median and standard error, range or inter-quartile range, then the median was substituted for the mean when the sample size exceeded 25 and their measures of variability were converted to standard deviation as per Hozo et al.¹⁶

Statistical analysis

Meta-analyses were completed for continuous data by using the change in the mean and standard deviation of outcome measures. It is an accepted practice to only use post-intervention data for meta-analysis but this method assumes that random allocation of participants always creates intervention groups matched at baseline for age, disease severity, etc. Change in post-intervention mean was calculated by subtracting the baseline from post-intervention values. Change in the standard deviation of post-intervention outcomes was calculated by using RevMan 5.0 (Nordic Cochrane Centre, Copenhagen, Denmark). Data required were (1) 95% confidence interval data for pre–post intervention change for each group or, when this was unavailable, (2) actual p-values for pre–post intervention change for each group or, if only the level of statistical significance was available, (3) we used default p-values (e.g. p<0.05 becomes p=0.049, p<0.01 becomes p=0.0099, and p = not significant becomes p=0.05). A random effects inverse variance was used with the effects measure of mean difference. Heterogeneity was quantified using the Cochrane’s Q test.¹⁷ Egger plots were provided to assess the risk of publication bias (supplementary files). We used a 5% level of significance and 95% confidence intervals; figures were produced using RevMan 5.0.

Results

Studies included in the review

Our initial search identified 16,606 manuscripts; examination of the latest editions of relevant journals yielded a further 63 manuscripts. Out of 16,669 studies, 62 were excluded at first inspection as duplicates; 16,473 were removed after reading titles or abstracts; 46 studies were not studies of exercise therapy in PAD patients. Out of the remaining 88 studies, 73 studies did not meet the inclusion criteria, leaving 15 studies for inclusion in this analysis (Figure 1).

Figure 1.

Literature search flow chart.

Fifteen studies^18–32 (18 intervention groups) met the inclusion criteria for our analyses; three of these studies^20,24,30 had more than one intervention group, which are separated in Forest plots. The total number of patients in our analyses was 1257: 543 took part in supervised exercise, 61 in a resistance training program, 316 completed an unsupervised exercise program, and 337 were assigned to usual medical care. The exercise prescriptive elements for each trial are outlined in Table 1. Training was often interval-based due to claudication pain and ranged from 20 to 60 minutes in duration, averaging 46±12 minutes, including rest time. Walking to moderate–maximum claudication pain was the most common prescription (nine of 15 studies). Training frequency ranged from 1 to 7 days weekly, with the most common frequency being 3 days per week. All studies provided complete supervision for the intervention. In the trials that reported it, average compliance to the supervised exercise programs was high: mean 83.1±9.2%^{18–20,28,29,31} or median 79.1%.^22,24 In addition, for those trials that compared supervised exercise to unsupervised exercise, there were no significant differences between the groups for exercise compliance.^20,29 The authors note that unsupervised exercise compliance was rigorously quantified with a step activity monitor in Gardner (2011);²⁰ however, it was self-reported in the earlier trial.²⁹

Table 1.

Included studies and intervention characteristics.

Author, year/ref.	Control activities	Exercise prescription
		Training modality	Claudication pain level during exercise	Exercise intensity	Training progression	Maximum volume	Frequency (days / week)	Program length (weeks)	Supervision
Treadmill/walking trials
Gardner, 2001¹⁸	Usual medical care Advised to walk^a	Walking	Mod–max pain	NR	Yes	40 min	3↓ →2 at 6 months	76	All
Gardner, 2011²⁰	Usual medical care Advised to walk	A: Unsupervised + quantified walking B: Supervised + quantified walking	A: Mod–max pain B: Mod–max pain	A: NR B: Light	Yes	A: 45 min B: 40 min	3	12	All
Gardner, 2012¹⁹	Usual medical care Advised to walk	Walking	Mod–max pain	Moderate	Yes	15↑→40 min	3	24	All
Kakkos, 2005²²	Advised to walk	Walking	Mod–max pain	NR	Yes	60 min incl rest	3	24	All
McDermott, 2004²³	Usual medical care Advised to walk^a	Walking	Mod–max pain	Moderate	Yes	30↑→40 min	3	12	All
McDermott, 2013²⁵	Education Sedentary control^b	Walking	Mod–max pain	Moderate	NR	45 min incl rest	1	24	All
Murphy, 201232	Usual medical care Cilostazol	Walking + cilostazol	Mild–mod pain	NR	Yes	60 min incl rest	3	26	All
Nicolaï, 2010²⁶	Usual medical care Advised to walk	A: Walking B: Walking with accelerometer	Mod pain	NR	NR	A: 30 min B: 30 min	3	52	All
Regensteiner, 1997²⁹	Usual medical care Advised to walk	Walking	Mild–mod pain	NR	Yes	55 min incl rest	3	12	All
Tsai, 2002³¹	Usual medical care Advised to walk	Walking	Mild–mod pain	NR	Yes	40 min	3	12	All
Lower extremity aerobic trials
Guidon, 2013²¹	Usual medical care Advised to walk	LEA exercise	Mod–max pain	Vigorous	Yes	60 min incl rest	2	12	All
Multi-modal exercise trials
Patterson, 1997²⁷	Education + advised to walk	Walking + AC and cycling + education			Yes	60 min	3	12	All
Pinto, 1997²⁸	Advised to walk	Walking, AC and cycling	Mild pain	NR	NR	20 min incl rest	3	12	All
Progressive resistance training trials
Regensteiner, 1996³⁰	Usual medical careAdvised to walk	A: WalkingB: PRT	A: Mod–max painB: No pain	A: NRB: Moderate	Yes	A: 60 min incl restB: 3 × 6 reps 5 exercises	3	12	All
McDermott, 2009²⁴	Nutrition educationSedentary control^b	A: WalkingB: PRT	A: Mod–max painB: No pain	A: ModerateB: Moderate	Yes	A: 40 minB: 3 × 8 reps 3 exercises	3	24	All

Mod, moderate; max, maximum; NR, not reported; ↑, increasing; ↓, decreasing; →, progressing to; A/B, intervention A or B; incl, including; LEA, lower extremity aerobic exercise; AC, arm crank; PRT, progressive resistance training; reps, repetitions.

Studies that only stated control group were prescribed usual medical care were deemed to have included advice to walk unsupervised unless otherwise stated.

Control groups were only classed as sedentary control if the paper specifically said participants were advised not to exercise or it was stated the control group was kept sedentary or non-exercising.

Walking Impairment Questionnaire (WIQ)

WIQ speed

Eleven studies (14 intervention groups) measured WIQ speed. Overall, a significant improvement was observed in exercise training participants versus controls: MD 9.60 (95% CI 6.98 to12.23, p<0.00001) (Figure 2). Both of the trials24,30 that used progressive resistance training (PRT) as an intervention showed that moderate intensity PRT did not significantly improve perceived WIQ speed. However, walking to some form of claudication pain three times a week for 30–40 minutes was the only other exercise prescription studied.

Figure 2.

Mean difference in WIQ speed: exercise versus control groups.

WIQ speed sub-analysis

When non-aerobic intervention groups were removed for the purpose of sub-analysis, the MD increased marginally: 10.27 (95% CI 7.47 to 13.07, p<0.00001) (Figure 3).

Figure 3.

WIQ speed sub-analysis.

WIQ distance

Ten studies (13 intervention groups) measured WIQ distance. Overall, an improvement was observed in exercise training participants versus controls: MD 7.41 (95% CI 4.49 to 10.33, p<0.00001) (Figure 4). The one trial that used PRT as an intervention showed that moderate intensity PRT did not significantly improve perceived WIQ distance. However, again, walking to some level of claudication pain was the only other exercise prescription studied. Sub-analysis was not indicated.

Figure 4.

Mean difference in WIQ distance: exercise versus control groups.

WIQ stair-climbing

Ten studies (13 intervention groups) measured WIQ stair-climbing ability. Overall, an improvement was observed in exercise training participants versus controls: MD 5.07 (95% CI 3.16 to 6.99, p<0.00001) (Figure 5). Both of the trials that used PRT as an intervention showed that moderate intensity PRT did not significantly improve perceived WIQ stair-climbing ability. All other trials used a prescription of walking to a level of claudication pain as the prescription and only three of these trials significantly improved the WIQ stair-climbing score. Sub-analysis was not indicated.

Figure 5.

Mean difference in WIQ stair-climbing: exercise versus control groups.

Short-Form Physical Component Summary (SF-PCS) score

Nine studies (11 intervention groups) measured SF-PCS; there were insufficient data reported within the studies to conduct individual SF physical domain analyses. A small improvement was observed in exercise training participants versus controls: MD 1.24 (95% CI 0.48 to 2.01, p=0.001) (Figure 6). The one trial that used PRT as an intervention showed that moderate intensity PRT did not significantly improve perceived SF-PCS. One trial used a variety of modes of aerobic exercise and did not improve SF-PCS. All other trials used walking to various levels of claudication pain, three times a week.

Figure 6.

Mean difference in SF-PCS scores: exercise versus control groups.

Short-Form Mental Component Summary (SF-MCS) score

Only six studies (six intervention groups) measured SF-MCS, which was not significantly different between exercise training participants versus controls: MD –0.55 (95% CI –1.27 to 0.18, p=0.14) (Figure 7).

Figure 7.

Mean difference in SF-MCS scores: exercise versus control groups.

Study quality assessment

Study quality assessment is presented in Table 2. Overall, the quality of the included trials was modest, with, on average, seven of the 11 quality criteria being present (mean 6.7±1.2, range 5–9/11). Common limitations were concealment of randomization, blinding of subjects, therapists and assessors, and, importantly, measurement of key outcomes in greater than 85% of trial participants. Ten of the 15 studies reported a >15% drop-out rate; however, nine of these studies addressed this concern by implementing an intention-to-treat analysis. Eleven studies (73.3%) in total completed an intention-to-treat analysis. No studies blinded therapists administering exercise interventions or participants. Only four studies (26.7%) blinded assessors who measured at least one key outcome. Thus, the assessor’s belief in intervention efficacy in 11 of the 15 trials (73.3%) may have biased such subjective treatment outcomes.³³ This lack of blinding of the outcome assessors is a serious limitation of the RCTs included in this review, particularly because the outcomes assessed are subjective and present greater opportunities for bias.³⁴

Table 2.

PEDro quality assessment.

Author, year/ref.	Eligibility criteria specified^a	Random allocation	Concealed allocation	Baseline similarity	Blinding of all assessors	Measures of key outcomes >85% of initial subjects	Intention-to-treat analysis	Between-group statistics	Point measures + measures of variability	Supervised vs unsupervised^b	Final score (×/11)
Gardner, 2001¹⁸	1	1	0	1	0	0	1	1	1	1	6
Gardner, 2011²⁰	1	1	1	1	0	0	1	1	1	1	7
Gardner, 2012¹⁹	1	1	1	1	0	0	1	1	1	1	7
Guidon, 2013²¹	1	1	1	1	0	0	0	1	1	1	6
Kakkos, 2005²²	1	1	1	1	0	0	1	1	0	1	6
McDermott, 2004²³	1	1	0	1	0	0	1	1	1	1	6
McDermott, 2009²⁴	1	1	1	1	1	1	1	1	1	1	9
McDermott, 2013²⁵	1	1	0	1	1	1	1	1	1	1	8
Murphy, 201232	1	1	0	1	1	1	1	1	1	1	8
Nicolaï, 2010²⁶	1	1	1	1	1	0	1	1	1	1	8
Patterson, 1997²⁷	1	1	0	1	0	0	0	1	1	1	5
Pinto, 1997²⁸	1	1	0	1	0	0	0	1	1	1	5
Regensteiner, 1996³⁰	1	1	0	1	0	1	0	1	1	1	6
Regensteiner, 1997²⁹	1	1	0	1	0	1	1	1	1	1	7
Tsai, 2002³¹	1	1	0	1	0	0	1	1	1	1	6

Eligibility criteria not included in final score.

Supervision included in final score. Intervention was classed as supervised if ≥1 session per week was supervised.

Heterogeneity

The Cochrane I² scores from analyses in Figures 2 and 3 showed zero heterogeneity. Analyses in Figures 4 –7 exhibited low to high evidence of between-study heterogeneity, ranging from 14% to 67%.

Egger plots

Egger plots indicated negligible risk of publication bias for all analyses.

Discussion

Despite several recent systematic reviews^4,6,35 and Cochrane meta-analyses,^7,12 this is the first meta-analysis limited to RCTs to analyse perceived walking impairment, general health (both Physical and Mental Component Summary) and quality of life outcomes for exercise interventions in PAD. In analysing the findings, it is important to take into account the quality of the trials included in this review and the limited study of exercise modes other than walking. Despite this, our findings suggest that walking training to claudication pain leads to small but significant improvements in perceived walking distance, speed and stair-climbing ability in PAD as measured by the WIQ. In addition, walking to claudication pain also improves the SF-36 Physical, but not Mental Component Summary score. These findings are encouraging, as HRQoL in PAD has been shown to predict long-term survival.³

Unfortunately, to date, studies that have measured the WIQ and/or the SF-36 have mostly been studies using walking to some level of claudication pain as a prescription. Only two studies^28,27 looked at a combination of aerobic modes and only two studies^24,30 have examined the effect of resistance training. In the studies that utilized exercise modalities other than walking, the efficacy of exercise training produced mixed results. More studies looking at various modes and elements of exercise prescription are needed to assess the maximum likely treatment effect of exercise training on HRQoL in PAD and to identify which specific prescriptive elements improve HRQoL the most.

Strengths and limitations

Analyses in Figures 2 and 3 showed zero heterogeneity, suggesting the analyses of WIQ speed were appropriate. Analyses in Figures 4 –8 exhibited moderate to high evidence of between-study heterogeneity, suggesting results of data pooling of WIQ distance, stair-climbing and the SF Physical and Mental Components must be interpreted with caution. Nevertheless, Egger plots showed minimal evidence of publication bias. This lack of bias may be due to similarity of training modes used in most of the trials examined. It is therefore unlikely that unpublished negative or neutral datasets exist for the outcome measures and significance levels suggest unpublished data would not change presented findings. A serious limitation of the literature analysed in this paper is that most (73%) of the investigators assessing outcome measures were aware of intervention group assignment. Whilst it is difficult to blind participants and trainers in an exercise training study, unless ‘sham’ exercise is used, the importance of blinding outcome assessors cannot be overlooked. The importance of blinding assessors for subjective measures presented in this paper cannot be under-estimated. If the outcomes assessors believe in one treatment more so than another, they could influence more generous responses to that intervention.36 The outcomes presented in this paper are entirely subjective outcomes, answers which are based on the patient’s perception, which present large opportunities for bias.³⁴ Future studies must at least blind the outcome assessors to the participants’ intervention to minimise outcome assessment bias, especially with subjective outcomes of interest such as the WIQ and Short-Form Health Surveys. The results from this study should be interpreted keeping the lack of assessor blinding in mind. Only one study with >15% withdrawal failed to conduct an intention-to-treat analysis; this would have had minimal impact on the overall findings. The PEDro scale has no assessment for quantifying compliance to the exercise programs, nor has an optimal compliance to exercise regimes for PAD been established. Therefore, we are unable to objectively comment on how compliance to the exercise program may have affected the overall findings. However, with such a high rate of compliance to the exercise programs, the authors doubt that compliance to programs affected the results of this analysis. Further limitation in this field of study is the lack of trials employing alternative modes of aerobic exercise or resistance training whilst measuring these HRQoL outcomes. More studies employing alternative aerobic and resistance training prescriptions are required in order to make further comment on the maximum likely treatment effect of specific exercise prescriptions on these outcomes. The authors note that the addition of a further nine RCTs since Guidon’s 2010 review⁹ has made meta-analysis possible.

Conclusions

Walking training to various levels of claudication pain improves perceived walking speed, distance and stair-climbing performance as measured by the WIQ, and self-reported physical function (SF-36) in people with PAD. Walking to claudication pain was the most common prescription studied and consequently yielded the most improvements in these outcomes. Further studies employing alternative modes and prescriptions of exercise are needed to measure the effect of multiple modes of exercise on HRQoL and walking impairment outcomes.

Footnotes

Acknowledgements

The authors would like to thank included study authors who provided additional information.

Declaration of conflicting interest

The authors declare that there is no conflict of interest.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Regensteiner

Hiatt

Coll

. The impact of peripheral arterial disease on health-related quality of life in the Peripheral Arterial Disease Awareness, Risk and Treatment: New Resources for Survival (PARTNERS) Program. Vasc Med 2008; 13: 15–24.

Issa

Hoeks

Scholte op Reimer

. Health-related quality of life predicts long-term survival in patients with peripheral artery disease. Vasc Med 2010; 15: 163–169.

Jain

Liu

Ferrucci

. The Walking Impairment Questionnaire stair-climbing score predicts mortality in men and women with peripheral arterial disease. J Vasc Surg 2012; 55: 1662–1673.

Parmenter

Raymond

Dinnen

Singh

. A systematic review of randomized controlled trials: Walking versus alternative exercise prescription as treatment for intermittent claudication. Atherosclerosis 2011; 218: 1–12.

Gardner

Montgomery

Parker

. Physical activity is a predictor of all-cause mortality in patients with intermittent claudication. J Vasc Surg 2008; 47: 117–122.

Parmenter

Raymond

Fiatarone Singh

. The effect of exercise on fitness and performance-based tests of function in intermittent claudication: a systematic review. Sports Med 2013; 43: 513–524.

Watson

Ellis

Leng

. Exercise for intermittent claudication. Cochrane Database Syst Rev 2008; (4): CD000990.

Leng

Fowler

Ernst

. Exercise for intermittent claudication. Cochrane Database Syst Rev 2000; (2): CD000990.

Guidon

McGee

. Exercise-based interventions and health-related quality of life in intermittent claudication: a 20-year (1989–2008) review. Eur J Cardiovasc Prev Rehabil 2010; 17: 140–154.

10.

Maher

Sherrington

Herbert

Moseley

Elkins

. Reliability of the PEDro scale for rating quality of randomized controlled trials. Phys Ther 2003; 83: 713–721.

11.

Bendermacher

Willigendael

Teijink

Prins

. Supervised exercise therapy versus non-supervised exercise therapy for intermittent claudication. Cochrane Database Syst Rev 2006; (2): CD005263.

12.

Fokkenrood

Bendermacher

Lauret

Willigendael

Prins

Teijink

. Supervised exercise therapy versus non-supervised exercise therapy for intermittent claudication. Cochrane Database Syst Rev 2013; (8): CD005263.

13.

Brazier

Harper

Jones

. Validating the SF-36 health survey questionnaire: new outcome measure for primary care. BMJ 1992; 305: 160–164.

14.

McDermott

Liu

Guralnik

Martin

Criqui

Greenland

. Measurement of walking endurance and walking velocity with questionnaire: validation of the walking impairment questionnaire in men and women with peripheral arterial disease. J Vasc Surg 1998; 28: 1072–1081.

15.

Ware

Jr Kosinski

Keller

. A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity. Med Care 1996; 34: 220–233.

16.

Hozo

Djulbegovic

Hozo

. Estimating the mean and variance from the median, range, and the size of a sample. BMC Med Res Methodol 2005; 5: 13.

17.

Higgins

JPT

Green

. Cochrane Handbook for Systematic Reviews of Interventions, Version 5.1.0, 2011. Available at: www//Cochrane-handbook.org (accessed 29 March 2012).

18.

Gardner

Katzel

Sorkin

. Exercise rehabilitation improves functional outcomes and peripheral circulation in patients with intermittent claudication: a randomized controlled trial. J Am Geriatr Soc 2001; 49: 755–762.

19.

Gardner

Montgomery

Parker

. Optimal exercise program length for patients with claudication. J Vasc Surg 2012; 55: 1346–1354.

20.

Gardner

Parker

Montgomery

Scott

Blevins

. Efficacy of quantified home-based exercise and supervised exercise in patients with intermittent claudication: a randomized controlled trial. Circulation 2011; 123: 491–498.

21.

Guidon

McGee

. One-year effect of a supervised exercise programme on functional capacity and quality of life in peripheral arterial disease. Disabil Rehabil 2013; 35: 397–404.

22.

Kakkos

Geroulakos

Nicolaides

. Improvement of the walking ability in intermittent claudication due to superficial femoral artery occlusion with supervised exercise and pneumatic foot and calf compression: a randomised controlled trial. Eur J Vasc Endovasc Surg 2005; 30: 164–175.

23.

McDermott

Criqui

Greenland

. Leg strength in peripheral arterial disease: associations with disease severity and lower-extremity performance. J Vasc Surg 2004; 39: 523–530.

24.

McDermott

Ades

Guralnik

. Treadmill exercise and resistance training in patients with peripheral arterial disease with and without intermittent claudication: a randomized controlled trial. JAMA 2009; 301: 165–174.

25.

McDermott

Liu

Guralnik

. Home-based walking exercise intervention in peripheral artery disease: a randomized clinical trial. JAMA 2013; 310: 57–65.

26.

Nicolai

Teijink

Prins

. Multi centre randomised controlled trial of supervised exercise therapy with or without feedback versus walking advice for intermittent claudication. J Vasc Surg 2010; 52: 348–355.

27.

Patterson

Pinto

Marcus

Colucci

Braun

Roberts

. Value of a supervised exercise program for the therapy of arterial claudication. J Vasc Surg 1997; 25: 312–318; discussion 318–319.

28.

Pinto

Marcus

Patterson

Roberts

Colucci

Braun

. On-site versus home exercise programs: psychological benefits for individuals with arterial claudication. J Aging Phys Act 1997; 5: 311–328.

29.

Regensteiner

Meyer

Krupski

Cranford

Hiatt

. Hospital vs home-based exercise rehabilitation for patients with peripheral arterial occlusive disease. Angiology 1997; 48: 291–300.

30.

Regensteiner

Steiner

Hiatt

. Exercise training improves functional status in patients with peripheral arterial disease. J Vasc Surg 1996; 23: 104–115.

31.

Tsai

Chan

Wang

Jeng

Hsieh

Kao

Chen

Liu

. The effects of exercise training on walking function and perception of health status in elderly patients with peripheral arterial occlusive disease. J Intern Med 2002; 252(5): 448–455.

32.

Murphy

Cutlip

Regensteiner

. Supervised exercise versus primary stenting for claudication resulting from aortoiliac peripheral artery disease: six-month outcomes from the claudication: exercise versus endoluminal revascularization (CLEVER) study. Circulation 2012; 125: 130–139.

33.

Guyatt

Rennie

. Users’ guides to the medical literature. JAMA 1993; 270: 2096–2097.

34.

Schulz

Chalmers

Altman

. The landscape and lexicon of blinding in randomized trials. Ann Intern Med 2002; 136: 254–259.

35.

Parmenter

Raymond

Fiatarone Singh

. The effect of exercise on haemodynamics in intermittent claudication: a systematic review of randomized controlled trials. Sports Med 2010; 40: 433–447.

36.

Schulz

Grimes

. Blinding in randomised trials: hiding who got what. Lancet 2002; 359: 696–700.

Exercise training for health-related quality of life in peripheral artery disease: A systematic review and meta-analysis

Abstract

Keywords

Introduction

Methods

Search strategy

Inclusions

Exclusions

Data synthesis

Statistical analysis

Results

Studies included in the review

Walking Impairment Questionnaire (WIQ)

WIQ speed

WIQ speed sub-analysis

WIQ distance

WIQ stair-climbing

Short-Form Physical Component Summary (SF-PCS) score

Short-Form Mental Component Summary (SF-MCS) score

Study quality assessment

Heterogeneity

Egger plots

Discussion

Strengths and limitations

Conclusions

Footnotes

Acknowledgements

Declaration of conflicting interest

Funding

References