Abstract
The aim of this study was to examine the use of pedometers as a tool to promote daily physical activity levels in patients with COPD.
A systematic review meta-analysis of pedometer physical activity promotion in patients with COPD was conducted. Medline/PubMed, Cochrane Library, Web of Science and CINAHL were searched from inception to January 2019. The search strategy included the following keywords: physical activity promotion, pulmonary rehabilitation and daily physical activity. The eligibility criteria for selecting studies were randomised controlled trials reporting pedometer physical activity promotion in patients with COPD.
Improvements in steps per day were found with pedometer physical activity promotion either standalone (n=12, mean 0.53 (95% CI 0.29–0.77); p=0.00001) or alongside pulmonary rehabilitation (n=7, 0.51 (0.13–0.88); p=0.006). A subgroup analysis reported significant differences in the promotion of physical activity based on baseline physical activity levels and the type of instrument used to assess levels of physical activity.
Future trials should consider the way in which pedometers are used to promote physical activity to inform clinical practice in the setting of pulmonary rehabilitation.
Abstract
Pedometer based physical activity promotion as a standalone intervention or alongside pulmonary rehabilitation induces meaningful improvements in daily physical activity levels (steps per day) in patients with COPD. http://bit.ly/2LnxM2o
Introduction
Interventions to promote levels of daily physical activity are becoming important in the management of patients with COPD [1, 2]. Studies comparing the levels of physical activity in patients with COPD with healthy age-matched controls have reported significantly lower levels in those with COPD [3–5]. In addition, low levels of physical activity in patients with COPD are associated with an increased risk of hospitalisation and mortality [3, 6, 7]. Therefore, effective approaches to improve daily physical activity are needed in patients with COPD.
Pulmonary rehabilitation has shown substantial improvements in exercise tolerance; however, these findings have not consistently progressed into improvements in daily physical activity [8]. One reason for this may link to physical activity being a complex health behaviour, with the determinants of physical activity influenced by personal, interpersonal, environmental, regional and/or national and global factors [9].
Physical activity promotion through the use of pedometers encompasses the stimulation of patients towards higher levels of daily physical activity by modifying their behaviour, with many versions of this intervention also using elements of the self-regulatory theory [10]. This theory involves a process of guiding an individual's own thoughts, behaviours and feelings towards achieving specific goals [11]. Incorporating the use of pedometers as a real-time feedback tool for improving daily steps allows patients the ability to follow individualised physical activity goals, which can be assessed and improved alongside techniques of motivational interviewing [12].
Implementing behaviour strategies using pedometer feedback can be done a number of ways, including face-to-face contact between patients and clinicians, group contact during rehabilitation sessions and through electronic information and communication technologies (tele-coaching) [13].
Studies have, however, provided inconsistent findings towards the implementation of pedometer-based feedback and motivational interviewing as part of physical activity promotion [14, 15]. Moreover, when the same intervention was added to standard care pulmonary rehabilitation, results remained inconclusive [1]. The most updated systematic review and meta-analysis has found high levels of heterogeneity regarding physical activity promotion, both as a standalone intervention and alongside pulmonary rehabilitation [1]. The existence of such heterogeneity is predominantly due to both methodological variables (types of goal setting, provided feedback and length of intervention) and patient demographics (severity and baseline physical activity levels). Hence, the aim of this systematic review and meta-analysis was to elucidate on aspects of physical activity promotion related to the way that pedometers are used to optimise physical activity in patients with COPD. In this context, we investigated the optimal frequency of goal setting, the type of patient feedback, the optimal length of interventions, the type of instrument used for assessing physical activity, and associations between baseline activity levels and the magnitude of improvement in daily physical activity.
Review objective
The aim of this review was to systematically review and meta-analyse aspects of physical activity promotion, specifically regarding how pedometers are used to optimise physical activity in interventions which incorporate the use of pedometers as a key component for improving levels of daily physical activity in patients with COPD.
Methods
The Cochrane Handbook for Systematic Reviews of Interventions and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [16] guidelines for reporting systematic review and meta-analyses were followed when conducting and reporting this prospectively registered systematic review (identifier CRD42018103893; www.crd.york.ac.uk/prospero/).
Eligibility criteria
The review team conducted a computerised literature search beginning in March 2018 in the following databases: Medline/PubMed, Cochrane Review, Web of Science and CINAHL. The final search of the literature took place on 18 January, 2019. Pre-piloted literature searches prior to the final search strategy were conducted based on two previously published systematic reviews on a related topic [1, 17]. The full search strategy can be found in table 1. It included a wide range of modalities; using terms associated with “chronic obstructive pulmonary disease”, “physical exercise training”, “physical activity promotion, physical activity counselling” and “randomized controlled trial”. Bibliographic details of all articles from the different databases were stored in the reference software file EndNote.
Search criteria for computerised literature search conducted in PubMed
On completion of the literature search, all stored references were exported from EndNote to the systematic review management software programme, Covidence. Eligible studies published in the English language were included if they fulfilled the predetermined PICOS criteria. 1) Population/participants: individuals with COPD defined by spirometry (i.e. forced expiratory volume in 1 s (FEV1)/forced vital capacity (FVC) <0.7). 2) Interventions or exposures: patients with COPD who were enrolled onto a programme of physical activity promotion, which included the use of a tool that provides real-time feedback on steps per day (i.e. pedometer screen). This included standalone interventions or those incorporated into pulmonary rehabilitation. 3) Comparison or control groups: patients not receiving any physical activity promotion intervention. 4) Outcomes of interest: the effect of physical activity promotion on steps per day as a measure of daily physical activity. 5) Setting: certified research studies. 6) Study design: randomised controlled trials (RCTs), both arms (intervention plus control).
Data extraction
After removing the duplicates and based on the inclusion criteria, two authors (M. Armstrong and N. Chynkiamis) independently and blinded, reviewed the title and abstract of trials and assessed the full text of articles. Any possible disagreement between both authors during the study selection process was discussed with a third author (I. Vogiatzis) for resolution.
For each eligible study, a pre-designed standardised Microsoft Excel form was used to collect data by a single author (M. Armstrong) on the following subheadings: author information (including name of first author and date of publication), blindness, participant characteristics (including age, FEV1 % pred, FVC, 6-min walk distance (6MWD), baseline daily steps, total lung capacity and residual volume, intervention details, physical activity measurements, primary outcomes and results). Two blinded reviewers (M. Armstrong and N. Chynkiamis) screened all articles independently, any disagreements were sent to a third independent author (I. Vogiatzis) to make a majority agreement.
Quality assessment
Quality assessment was performed using the PEDro quality scale, which is an 11-item scale assessing internal and external validity of clinical trials [18]. Two authors (M. Armstrong and N. Chynkiamis) independently reviewed the following domains employed by this scale: eligibility criteria, random allocation, concealed allocation, baseline similarity, blinding (subject, therapist and assessor), and measures recorded from at least 85% of participants, full intention to treat, group comparison and point measure. The higher the given score, the better the quality. Cut-off points of the scale were excellent (9–10), good (6–8), fair (4–5) and poor (<3) [18].
Data synthesis
Meta-analyses were undertaken using Review Manager (RevMan v.5.3; Cochrane Collaboration, Oxford, UK). Change scores or end of intervention values with the corresponding standard deviation for the outcomes of interest were used to obtain the overall effect size represented by standard mean difference with 95% confidence interval, with a threshold p<0.05 considered as significant. Heterogeneity in this meta-analysis was assessed by I2 value as follows: 0–40%, might not be important; 30%–60%, moderate heterogeneity; 50%–90% substantial heterogeneity; and 75%–100% considerable heterogeneity [19]. A fixed-effects model was used for the meta-analysis; however, if statistical heterogeneity was noted (I2 >40%), meta-analyses were performed using the random effects model. Sensitivity analysis was used if substantial heterogeneity (I2 >75%) was reported in meta-analyses.
Results
The search strategy yielded 2582 potentially relevant articles. After removing 714 duplicates and screening 1868 abstract/titles, 55 articles remained for the full-text screening. On completion of full-text screening, 38 studies were excluded. Therefore, 17 studies were considered eligible for inclusion in this systematic review and meta-analysis. One article provided three different comparisons, resulting in three RCTs. A full PRISMA flow diagram of the screening process is shown in figure 1. Participants were individually randomised in all included trials (i.e. there were no cluster RCTs). Characteristics of included RCTs are summarised in table 2 and all were published between 2006 and 2018.
PRISMA flow diagram for database search and study selection process.
Characteristics of included studies
Characteristics of included subjects
All of the included trials comprised 1677 patients (45% male), with a median (range) sample size of 72 (16–343). Included patients had a mean (range) age of 66 (54–75) years and average FEV1 % pred ranged from 43 to 78, indicative of mild-to-moderate COPD [34]. Patients were reported as physically inactive at baseline with an average mean (range) value of 4365 (1557–7161) steps·day−1.
Characteristics of included/excluded trials
A total of 38 studies were excluded from this review on completion of full-text screening. The reasons for exclusion include: the wrong intervention (n=11), duplicates (n=9), wrong study design (n=6), wrong outcomes (n=6), wrong comparators (n=2), no full-text availability (n=2) and no reported data for daily steps (n=2).
Quality assessment
Table 3 provides a summary of the risk of bias decision made for each category for the included studies. In line with the PEDro scale, the quality of included studies ranged from good to excellent (mean (interquartile range) PEDro score 9.29 (1)); suggesting a low risk of bias towards the main outcome measure.
Qualitative synthesis of included studies using PEDro scale for the quality of randomised controlled trials
Meta-analyses of included studies
When observing the effects of physical activity promotion, there was a positive effect on steps per day compared with usual care (n=12 RCTs; 0.53 (0.29–0.77), p<0.00001) (figure 2) [13, 14, 20, 21, 24, 25, 27, 28, 30, 32, 33], which equated to an improvement of ∼1000 steps·day−1. A positive effect on steps per day was also found when pedometer physical activity promotion was added to pulmonary rehabilitation versus pulmonary rehabilitation alone (n=7 RCTs; 0.51 (0.13–0.88), p=0.006) (figure 2) [14, 15, 22, 23, 26, 30, 31]. However, the pooled analysis of pedometer physical activity promotion compared with usual care reported considerable heterogeneity (I2=77%).
Effect sizes of pedometer-based physical activity (PA) promotion on steps per day in patients with COPD. PR: pulmonary rehabilitation.
Moreover, the increases in daily physical activity induced by pedometer physical activity promotion (both alone and alongside pulmonary rehabilitation), were comparable among studies that provided: 1) weekly or infrequent goal setting; 2) an intervention length <3 months or >3 months; and 3) remote or face-to-face contact following overall or subgroup analysis (all p<0.05) (table 4). In contrast, studies employing accelerometers to measure physical activity were less effective compared with those employing pedometers. Furthermore, patients with greater baseline physical activity levels (>4000 steps·day−1) exhibited greater improvements in daily physical activity compared with those with lower baseline physical activity levels (<4000 steps·day−1) (table 4).
Subgroup analysis on physical activity (PA) outcomes of included studies
Sensitivity analysis removing a single study [27] from the pooled analysis of pedometer-based physical activity promotion reduced heterogeneity (I2=60%). The sensitivity analysis did not statistically affect the pooled analysis of the remaining 11 studies in pedometer-based physical activity promotion (0.44 (0.25–0.63); p<0.05).
Discussion
Summary of the main findings
This systematic review and meta-analysis of 19 RCTs provides evidence that pedometer physical activity promotion as a standalone intervention compared with usual care or alongside pulmonary rehabilitation compared with pulmonary rehabilitation alone improves steps per day by a magnitude that is within the minimal important difference (MID) of 600–1100 steps·day −1 reported by Demeyer et al. [35] (table 3). Moreover, this meta-analysis suggests that pedometer physical activity promotion was more effective in patients with greater baseline physical activity levels and when pedometers were used to measure improvements in physical activity compared with accelerometers (table 4).
The addition of a sensitivity analysis reducing levels of heterogeneity provided no additional effects to the pooled analysis of pedometer physical activity promotion.
Interpretation of the results
Previous literature surrounding the effects of physical activity promotion has reported inconclusive evidence of the effectiveness of this intervention on steps per day. In agreement with the findings of our review, Qui et al. [1] found that physical activity promotion improved steps per day compared with usual care in nine studies. However, significant heterogeneity (I2=81%) may have affected the overall analysis of those studies. The increase in steps per day reported as a result of pedometer physical activity promotion seems much larger than those from other methods including exercise training as part of pulmonary rehabilitation, health monitoring, long-term oxygen therapy or neuromuscular electrical stimulation [12, 17].
However, Lahham et al. [17] reported that physical activity promotion was not an effective standalone intervention towards improving steps per day. A number of disparities are apparent between review articles. First, Lahham et al. [17] based their analysis of physical activity promotion on a subgroup analysis of subjective and objective measures. Both our study and that of Qui et al. [36] only included studies reporting objective measures of daily physical activity due to limited validity and inaccuracy of subjective measures of activity levels in patients with COPD [37]. Secondly, the number of included studies varied across separate meta-analyses. In our review, a total of 12 studies with an average total sample size of 120 were included in the pooled analysis of pedometer physical activity promotion. Meanwhile, Lahham et al. [17] reported only two studies on objective measures of physical activity, with an average total sample size of 17. With the significant benefits of collecting and reporting objective measures of physical activity in both healthy individuals and patients with COPD, and a much greater sample size across pooled analyses, our review and that of Qui et al. [1] could be argued to have more valid findings for patients with COPD than Lahham et al. [17]. Benefits of pedometer physical activity promotion have also been reported in patients with type 2 diabetes [36]. A meta-analysis including 11 RCTs reported a significant increase in physical activity with an average magnitude of improvement of 1822 steps·day −1, which is greater than we found in patients with COPD (figure 3).
Comparison between improvements in daily steps shown by various implementation modalities. Vertical dashed line indicates the lower levels of minimal important difference in patients with COPD. PA: physical activity.
When observing the effects of physical activity promotion alongside pulmonary rehabilitation, the present study and that by Lahham et al. [17] and Qui et al. [1] have shown statistically significant effects on steps per day. Lahham et al. [17] stated that providing persistent and individualised feedback on activity levels in conjunction with pulmonary rehabilitation, achieved significant effects that exceeded both physical activity promotion alone and pulmonary rehabilitation alone. However, Lahham et al. [17] were unable to include a recent RCT [29]. Within that study, the authors reported evidence questioning the effectiveness of physical activity promotion on daily physical activity in patients attending pulmonary rehabilitation [29]. It was determined that the routine use of this intervention should not be included in standard care pulmonary rehabilitation because levels of daily physical activity were greater after pulmonary rehabilitation alone when compared with baseline measures [29]. These results were based upon this study being the first to include a large sample size, suggesting other studies were underpowered. In addition, that study [29] scored highly on the PEDro scale, suggesting it had a low level of bias and the results reported were of high quality. The present study and that by Qui et al. [1] have been able to incorporate the study by Nolan et al. [29] into separate meta-analyses. A contrast in reporting physical activity between our study and that by Qui et al. [1] has provided two interpretations of the study by Nolan et al. [29]. Qui et al. [1] provided accelerometer step counts from baseline as a measure of steps per day from the study by Nolan et al. [29], reporting a small positive effect on physical activity levels. In contrast, we have chosen to report steps per day from pedometer step counts, resulting in a neutral effect on physical activity. We agree with Qui et al. [1] that accelerometers provide a more accurate measure of physical activity; however, the majority of studies in our meta-analysis have primarily used pedometers to report physical activity levels, so this may falsify results [29].
Our meta-analysis also suggests a number of important principles surrounding the way in which pedometers have been used for promoting physical activity. An overall analysis of patients with greater baseline physical activity levels (>4000 steps·day −1) showed greater improvements in steps per day compared with those with lower baseline physical activity (≤4000 steps·day−1) (figure 3). Of further interest is the influence that baseline physical activity had on the effects of pedometer physical activity promotion alongside pulmonary rehabilitation (table 4). In studies that implemented physical activity promotion alongside pulmonary rehabilitation, an insignificant effect on steps per day was reported when patients had a baseline physical activity ≤4000 steps·day −1 [14, 15, 31]. It must be outlined that there was only a small number of studies in this subgroup analysis with a small mean sample size; however, such differences in effect size warrants closer scrutiny.
Osadnik et al. [38] proposed that patients with COPD who exhibit greater exercise capacity prior to pulmonary rehabilitation are more likely to achieve greater improvements in daily physical activity. They reported clinically meaningful improvements in steps per day with patients reporting a 6MWD >350 m compared with <350 m (707±1780 versus 157±1694 steps·day −1). This higher likelihood of improvement in physical activity in patients preserving a greater exercise tolerance may also provide an explanation for those patients exhibiting a higher baseline physical activity. However, in contrast with this notion, a recent study from Gulart et al. [39] suggests that patients with lower values of FEV1 and steps per day were more likely to achieve MID in steps per day. This finding was attributed to the notion that patients with more severe disease have a greater potential for improvement as they are further from their “maximal” capacity, compared with patients with less severe disease.
In addition, our meta-analysis has found that the primary measure of physical activity (i.e. through accelerometers or pedometers) may have marked influences on the effects of physical activity promotion (table 4). Specifically, significant improvements in steps per day were shown in those studies reporting physical activity via a pedometer compared with an accelerometer (figure 3). The finding in physical activity outcomes may be due to accelerometers being a validated tool for measuring steps per day in patients with COPD and therefore pedometers may overestimate physical activity [40]. However, a number of previous studies, including Qui et al. [1], disagree with this finding. Both that study [1] and a meta-analysis in patients with type 2 diabetes [41] have shown no significant differences between accelerometers and pedometers. Consideration must be made in relation to these comparisons being indirect, with such confirmation potentially required through a future 1 versus 1 design.
In contrast, pedometer physical activity promotion in patients with type 2 diabetes presents different findings. For instance, it has been reported that patients with type 2 diabetes should initially set their own activity goals, before they set to increase their goals with the assistance of healthcare professionals [36]. We were unable to confirm this hypothesis among patients with COPD as many of the reported studies do not provide definitive step goal descriptions. In addition, studies have shown that the use of step diaries alongside pedometers as a source of motivation were imperative to increase physical activity levels [36, 42, 43].
Finally, it has become evident that regardless of the way pedometers are used (i.e. frequency of goal setting, type of patient feedback, length of intervention, the instrument used to assess physical activity) or the baseline activity levels, the improvement in steps per day is within the MID (figure 3) [35]. This finding has strong implications for the use of pedometers as part of the comprehensive management of patients with COPD.
Quality of the evidence
The overall quality of evidence from included studies was good, in line with the PEDro scale for quality assessment. The inability to blind subjects reduced the overall quality of evidence and increased the risk of bias towards the intervention procedure and may increase the chances of a placebo effect when using the pedometer. Future research reporting the effects of physical activity promotion may improve quality scoring by blinding all subjects from the intervention procedure. However, a concern remains that blinding patients from the intervention would require a pedometer being issued to a control group, which may present the control group with a level of physical activity promotion as they are able to monitor their daily steps. A number of studies were unable to blind any members of the study from patient allocation [14, 21, 24, 25, 28, 30]. In any clinical trial, blinding of at least the researcher is desirable and the blinding of subjects is warranted in order to decrease bias within the findings. When blinding is not used or the subject group status is easily detectable, subjects will generally try to fulfil the perceived expectation of the researcher [44].
Strength and limitations
This systematic review and meta-analysis is the first to include two recently published RCTs reporting pedometer-based physical activity promotion implemented either alone [20] or alongside a combined pulmonary rehabilitation programme [31]. Moreover, we are the first to report that, regardless of how pedometers are used in the implementation of physical activity promotion, they can provide improvements in daily physical activity (steps per day) which exceed the MID. Several limitations should be noted. First, some heterogeneity existed in the outcomes of pedometer physical activity promotion, which was partially explained by our findings on the modalities of pedometer use. Secondly, we cannot be certain of the specific improvement a pedometer intervention can have alongside pulmonary rehabilitation on daily physical activity without knowing the exact progression of exercise training for individual patients during pulmonary rehabilitation. Finally, despite a comprehensive search of the literature using the main scientific search databases, there is still a possibility that studies eligible for inclusion may have been missed. The search restriction on English written studies and the failure to search for unpublished studies and/or abstracts/conference papers may have resulted in selection and publication bias.
Conclusion
In conclusion, our systematic review and meta-analysis provides evidence that pedometer-based physical activity promotion promotes steps per day when it is used as an intervention alone or alongside pulmonary rehabilitation, including two recently published RCTs [20, 31]. Future trials should concentrate on high-quality study designs, with specific thought towards the optimal way of using pedometers during physical activity promotion (i.e. consider frequency of goal setting, type of patient feedback, length of intervention and instrument used for assessing physical activity). This review has found further evidence that patients benefit more from physical activity promotion when baseline levels of physical activity are >4000 steps·day−1. Therefore, consideration of baseline daily physical activity levels and/or exercise tolerance [38], should feature prominently in future studies. Furthermore, future studies should investigate the combined benefits of pulmonary rehabilitation, physical activity promotion and cognitive behavioural therapy for those patients with severe COPD who are anxious and depressed and therefore exhibit limitations in improving daily physical activity. Moreover, future studies could incorporate the addition of semi-automated tele-coaching as delivered by Demeyer et al. [7], as a low maintenance approach to providing continued support towards daily physical activity feedback [38].
Footnotes
This study is registered at www.crd.york.ac.uk/prospero/ with identifier CRD42018103893.
Provenance: Submitted article, peer reviewed.
Author contributions: M. Armstrong conducted the study, collected and analysed the data, and wrote the manuscript. N. Chynkiamis supported the collection of data. I. Vogiatzis, A. Winnard, S. Boyle and C. Burtin contributed to the review/editing of the manuscript. All authors read and approved the final manuscript.
Conflict of interest: M. Armstrong has nothing to disclose.
Conflict of interest: A. Winnard has nothing to disclose.
Conflict of interest: N. Chynkiamis has nothing to disclose.
Conflict of interest: S. Boyle has nothing to disclose.
Conflict of interest: C. Burtin has nothing to disclose.
Conflict of interest: I. Vogiatzis has nothing to disclose.
- Received April 4, 2019.
- Accepted July 11, 2019.
- Copyright ©ERS 2019.
This article is open access and distributed under the terms of the Creative Commons Attribution Non-Commercial Licence 4.0.
References
