Chronic obstructive pulmonary disease (COPD) is a common, incurable but manageable disease. Characteristic to the condition is airway and/or alveolar abnormalities, leading to chronic respiratory symptoms and limited airflow that is often triggered by exposure to toxic gases (Mirza et al. 2018; Vogelmeier et al. 2017). The presentation is often a slow, progressive onset of symptoms in people older than 40 (Viviers & Van Zyl-Smit 2015). The Global Burden of Disease Study reported a prevalence of 251 million COPD cases globally in 2016 (World Health Organization 2017). Despite lacking information regarding COPD prevalence in South Africa, it is predicted that by 2020 the disease will become one of the three leading causes of death attributed to non-communicable disease globally (Viviers & Van Zyl-Smit 2015).

South Africans are exposed to COPD risk factors such as cigarette smoke, inorganic dust, indoor use of biomass fuel for heating and cooking and HIV (World Health Organization 2017). Symptoms include a cough, sputum production and progressive worsening of shortness of breath. Functional tasks such as walking or stair climbing lead to air trapping and increased hyperinflation, exacerbating breathlessness (Khachi, Barnes & Antoniou 2010; O’Donnell et al. 2007; Viviers & Van Zyl-Smit 2015). This increased breathlessness incites anxiety, which leads to more breathlessness, causing a chain reaction where physical activities are avoided (O’Donnell et al. 2007).

Common systemic presentations in COPD linked to increased mortality include impairments in respiratory and peripheral muscle function and mass (Barreiro & Gea 2015). Such dysfunction is heterogeneously distributed among muscle groups; however, the strength of upper limbs and respiratory muscles is better preserved than that of lower limbs (Maltais et al. 2014). People with COPD struggle to participate in physical activity (Ek & Ternestedt 2008), mainly because of dyspnoea and leg fatigue (Maltais et al. 2014). This further reduces exercise tolerance and quality of life (QoL) and increases healthcare utilisation (Barreiro & Gea 2015).

Chronic obstructive pulmonary disease has very few treatment options (pharmacological or non-pharmacological) that influence mortality (Keen & Medarov 2017). Pulmonary rehabilitation is effective for improving exercise tolerance, dyspnoea and fatigue (World Health Organization 2018). As lower limb dysfunction commonly occurs in people with COPD who are struggling with severe dyspnoea (Vonbank et al. 2012), international guidelines recommend combining endurance and strength training as a part of pulmonary rehabilitation (Iepsen et al. 2015; Spielmanns et al. 2017a; World Health Organization 2018). Strength training has a decreased cardiovascular response compared to endurance exercise. The added benefit of strength training is that it has a lower level of oxygen consumption and minute ventilation, resulting in decreased dyspnoea during training. Strength training is thus considered an effective and feasible treatment option in the clinical setting for people with COPD, although optimal training strategies and mechanisms of improvement are still under investigation (Daabis, Hassan & Zidan 2017).

A proposed complementary modality to strength training in people with COPD is whole-body vibration training (WBVT) (Gloeckl, Heinzelmann & Kenn 2015). The WBVT is conducted on a vibrating platform moving in sinusoidal oscillations and reportedly improves muscle activation when implemented simultaneously with static and dynamic exercises (Spielmanns et al. 2017a). Strength training incorporating WBVT is reported to be a safe and time-saving intervention that may be more effective than strength training alone to result in improved exercise performance and other person-centred outcomes in people with COPD (Spielmanns et al. 2017a). Vibrations are applied at different frequencies, which cause the muscle relaxation period to be interrupted. This contributes to higher output in muscle power (Spielmanns et al. 2017a) and there is some evidence that WBVT effectively improves lower limb strength in people with COPD (Koczulla et al. 2014).

Four systematic reviews (Cardim et al. 2016; Gloeckl et al. 2015; Yang et al. 2016; Zhou et al. 2018) have been published dealing with the effect of WBVT on pulmonary function, functional exercise capacity (FEC) and QoL in people with COPD. The most recent review (Zhou et al. 2018) showed that WBVT was beneficial for improving FEC, but limited evidence was found regarding the effects on pulmonary function and QoL. Only two systematic reviews (Cardim et al. 2016; Gloeckl et al. 2015) assessed muscle strength as an outcome in people with COPD and found no difference between the WBVT and control groups. However, these reviews based their findings on the same single randomised controlled trial (RCT) (Pleguezuelos et al. 2013). The most recent systematic review (Zhou et al. 2018) did not assess muscle strength. Since the publication of the three systematic reviews up to 2016, more RCTs have been published, and include lower limb muscle strength as an outcome. Therefore, the purpose of this systematic review was to update the current evidence regarding the effect of WBVT on muscle strength, FEC, pulmonary function and QoL in people with COPD.

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were used to conduct this systematic review (Liberati et al. 2009).

This review aimed to determine the effectiveness of lower limb strength training combined with WBVT, compared to lower limb strength training alone, on various patient-centred outcomes in COPD. To our knowledge, this was the first review to focus on lower limb strength training as a comparative element. Results demonstrated limited evidence that combining strength training and WBVT may improve lower limb strength (measured as leg press strength) more effectively than strength training alone after at least 12 weeks in adults older than 40 with mild-to-moderate COPD. Results further suggest that combining strength training with WBVT improves FEC (measured as 6MWD) significantly more than strength training alone in the short term (≤ 12 weeks) for adults with mild-to-severe COPD. Similar to our review’s findings, previous systematic reviews (Cardim et al. 2016; Gloeckl et al. 2015; Yang et al. 2016; Zhou et al. 2018) revealed that WBVT had beneficial effects on FEC. Evidence was lacking or limited for pulmonary function and QoL in all these reviews.

Skeletal and lower limb muscle dysfunction is considered one of the most troublesome systemic manifestations of COPD and is characterised by limb muscle atrophy, weakness and decreased endurance (Chen et al. 2018). Although the relationship between limb muscle dysfunction and COPD is still poorly understood, the loss of type I muscle fibres, reduced oxidative enzyme activity, reduced capillary-to-fibre ratio, nutritional depletion and inflammation have been suggested as causal factors (Maltais et al. 2014). Peripheral muscle atrophy is also closely linked to increased hospitalisations and decreased life expectancy in people suffering from COPD (Swallow et al. 2007). A mid-thigh muscle cross-sectional area of < 70 cm² is associated with a fourfold increase in mortality after adjusting for age, sex and lung function (FEV1) (Maltais et al. 2014). International clinical practice guidelines recommend that exercise and strength training of lower limb muscles should be a mandatory component of pulmonary rehabilitation for people with COPD to improve lower limb muscle strength, function and mass (Ries et al. 2007). Such training may prevent disease progression and improve prognosis regardless of the severity of pulmonary disease (Jaitovich & Barreiro 2018).

Across the RCTs assessing lower limb strength and included in this review, participant groups performing squats (with or without WBVT) or conventional progressive resistance programmes (even without squats and/or WBVT) demonstrated improved knee extension strength. Thus, regarding targeted knee extension strength, it seems that squatting may be effective regardless of adding WBVT. Furthermore, a conventional resistance programme may be preferable to combined squatting and WBVT. In terms of more general lower limb strength (leg press), it remains unclear whether combining squats with WBVT holds additional benefits over squatting alone because of the non-comparable training approaches noted in the RCTs. It may, for example, simply be that squatting was more effective than the calisthenics (bodyweight exercises, similar to gymnastics) employed by Spielmanns et al. (2017a). In other populations, increased muscle strength is a major benefit of WBVT, especially in the elderly (Lau et al. 2011). The current review also included a mostly elderly sample. Reasons for discrepant results may be differences in the WBVT operation protocols, exercise protocols, the specific pathology and variations in the type of squats.

Another reason for the observed results may be the peripheral muscle dysfunction associated with COPD, along with the different muscle strength assessments used in the various studies: a leg press test (a closed-chain and multi-joint or muscle movement) versus a pure knee extension test (an open-chain and single muscle or joint movement). Quadriceps weakness is a particular concern in people with COPD; the weakness becoming more prevalent with increasing disease severity (Kharbanda, Krishnan & Ramakrishna 2015). Quadriceps femoris neuromuscular activation is greater with isolated knee extension compared to a multi-joint leg press (Andersen et al. 2006). It may thus be that assessing this muscle in isolation in adults with COPD has less potential for demonstrating improvement than using a more functional assessment involving additional muscles. A leg press also more closely resembles the squatting exercises performed on the vibration plate, involving more muscles (gluteus maximus, quadriceps femoris, gastrocnemius, soleus, hamstrings and adductor muscles) than isolated knee extension (mainly rectus femoris quadriceps muscle), thus constituting a better indication of overall lower limb strength improvement.

Functional exercise capacity was measured using 6MWD in all five RCTs. This outcome improved significantly following 3 weeks of combined strength training and WBVT in two studies (Gloeckl et al. 2012, 2017). The improvements exceeded the minimum clinically important difference (MCID) of 14 m – 30 m (Bohannon & Crouch 2017) and may thus be considered clinically significant. Notably, the training programmes in these two studies were comparable between groups. These studies included both participants with severe to very severe COPD, supporting the potential value and feasibility of the modality in advanced disease stages. Similarly, Spielmanns et al. (2017a) found a statistically significant difference between WBVT and control favouring WBVT, but after a longer period (12 weeks). This study included patients with mild-to-severe COPD and results also exceeded the MCID. However, it again remains unknown whether the beneficial effect can truly be attributed to the added WBVT, or whether the entire intervention was more effective than calisthenics in this particular study. Two studies (Salhi et al. 2015; Spielmanns et al. 2017b) found no statistically significant between-group differences in FEC after 12 weeks (including for VO₂ max). Salhi et al. (2015), however, did find a statistically significant intragroup improvement for both the WBVT and control groups. It thus seems that both a WBVT-and-squat-containing strength regime and a resistance programme without these components may be equally beneficial.

Despite the lack of evidence noted for improved isolated muscle strength, the observed improvements in FEC may be because of other neuromuscular adaptations associated with WBVT such as increased muscle power and the myostatic stretch reflex elicited by mechanical vibration (Salhi et al. 2015). Indeed, such changes – rather than muscle hypertrophy – have been associated with improvements in 6MWD (Gloeckl et al. 2015). A possible reason why two RCTs (Salhi et al. 2015; Spielmanns et al. 2017b) did not find a significant exercise capacity improvement in the WBVT group is that WBVT may not be specific enough to increase peripheral muscle strength or walking distance reliably, especially in higher functioning cohorts (Salhi et al. 2015). For example, participants in the studies conducted by both Salhi et al. (2015) and Spielmanns (2017b) had among the highest baseline 6MWD values and thus less room to show improvement compared to the studies that did show significant improvements.

In healthy elderly adults, positive, moderate correlations have been demonstrated between the strength and function of the respiratory muscles (measured using a vacuum manometer) and lower limb muscles (measured as knee extension peak torque), as well as functional capacity (measured using 6MWD) (Simões et al. 2010). In addition, WBVT combined with static squatting has been found to improve cardiorespiratory response (VO₂, heart rate and oxygen saturation) in COPD (Lage et al. 2018) and respiratory muscle function and strength (inspiratory and expiratory pressure) (Pessoa et al. 2017). Tonic contractions generated by vibratory stimuli, along with isometric abdominal muscle contraction to maintain static positioning during training, are proposed mechanisms for the increased strength in expiratory muscles (Pessoa et al. 2017). In our review, pulmonary function was measured using FEV1 (% predicted) in three RCTs (Gloeckl et al. 2012; Salhi et al. 2015; Spielmanns et al. 2017b). None of these studies found significant differences in FEV1, either within or between groups. This suggests limited benefits of the intervention and control procedures on pulmonary function, at least in the short term (3–12 weeks). Chronic obstructive pulmonary disease level of severity across these studies ranged from mild (stage I) to very severe (stage IV); thus, this lack of improvement does not seem to be associated with COPD severity. It may be that in people with chronic respiratory diseases, neither WBVT nor lower limb strengthening (alone or combined) may be effective in improving lung function in a time period shorter than 3 months (Yang et al. 2016) – a finding also noted in studies in the cystic fibrosis population (Rietschel et al. 2008).

Although the our review found that leg press peak force improved significantly in the WBVT groups, no improvements were observed for knee extension peak torque. Considering the positive correlation previously observed between knee extension peak torque and lung function (Simões et al. 2010), this finding may perhaps be further supported for the proposed relationship.

Four RCTs assessed QoL, using three different measurement tools at 3 and 12 weeks. Although QoL improvements were noted at 12 weeks, these were not always of clinical relevance. For example, the significant (control group) improvements in CAT scores in the work of Spielmanns et al. (2017a) did not exceed the MCID of two points (Kon et al. 2014). This lack of clinical improvement may be because of the CAT not being specific enough to sense changes in exercise capacity, as only two of the eight questions related to physical activity (Spielmanns et al. 2017a). The improvements noted in Spielmanns et al. (2017b) for some individual CRDQ domains, on the contrary, did exceed the 0.5-point MCID (Schünemann et al. 2005) but should be interpreted cautiously because of the high participant dropout rate. Overall, given that the QoL measures may have lacked sensitivity and that different aspects of QoL were measured between studies, it is difficult to conclude whether adding WBVT to strength training is beneficial for improving QoL.

Overall, it is difficult to draw conclusions from our review regarding the best WBVT and strength training protocols to recommend to clinicians, as these varied vastly across studies. The intervention groups in all studies received a combination of WBVT and squat exercises to be performed on the vibrating platform. However, the type, number of squats and squatting positions differed between studies, as did the dosages. It has not been investigated yet if one squatting modality is superior to another, but it is speculated that a combination of both static and dynamic exercises may be beneficial to improve leg muscle function (Gloeckl et al. 2015). Our results also revealed wide variation in the type and dosage of additional exercise training prescribed to participants, and the specific WBVT device settings differed between studies. Current literature recommends aiming for a frequency of > 20 Hz for a side alternating platform, and < 35 Hz for a vertical platform, to improve muscle strength and function. These frequency ranges seem to increase neuromuscular activation of the lower limb muscles most effectively (Gloeckl et al. 2015). Vertical vibration may have a better long-term effect compared to side alternating motion (Yang et al. 2016) and WBVT duration sets of 30–180 s have been recommended depending on the specific training target. It is also suggested that WBVT amplitudes less than 6 mm do not result in significant treatment effects as compared to 6 mm – 10 mm, but the evidence remains insufficient regarding which amplitude is optimal (Yang et al. 2016).

Strengths of our review include a comprehensive and systematic search of eight scientific databases. The review was conducted according to a documented, stepwise process. Each step involved at least two independent authors, thus limiting potential errors because of cross-checking of information and increased objectivity. We identified five eligible RCTs (level II evidence) that were of a high methodological quality, although no studies were of excellent quality (i.e. > 8/10) (Armijo-Olivo et al. 2015). None of the studies were able to blind the participants, therapists or the assessors. Although inherent to the intervention employed, this could have led to observer or detection bias. Gloeckl et al. (2012) and Spielmanns et al. (2017a) did not apply concealed allocation, thus decreasing internal validity by causing selection bias, which may influence the type of treatment that the person will receive in the trial. Participant dropout (Spielmanns et al. 2017b) may have led to attrition bias and therefore potentially less reliable findings. The results of our review are limited by the fact that heterogeneity prohibited meta-analysis; however, performing meta-analyses of unsuitable data may have resulted in misleading findings. In addition, no conclusions can be drawn regarding long-term effects, as no studies had follow-up periods exceeding 12 weeks. The limited number of included studies and small sample sizes may affect the overall validity of our review results. Language bias is possible as only English papers were included. Finally, the fact that all studies were conducted in high-income countries limits generalisability to middle- and lower-income countries.

The clinical implications of this review’s findings should be considered carefully for physiotherapy practice, including the South African context. Given that South Africa is a high-middle income country with extreme levels of income inequality, and considering the high cost of WBVT devices, such equipment may fall outside of budgetary constraints for many South African clinics or physiotherapists – especially in rural settings. The current evidence should therefore be contemplated in terms of the specific goals of therapy and the needs of the person.

For example, if the goal of therapy is to improve muscle strength, the current evidence is insufficient to recommend WBVT over strength training alone, which is an effective modality for people with COPD (Barreiro & Gea 2015). Given that strength training without WBVT is more affordable, this is an important consideration for middle- or lower-income countries. On the contrary, if the primary goal is to improve FEC, combining WBVT with strength training may be considered as an even more effective modality than strength training alone, should resources allow. However, the substantial heterogeneity among the various studies makes it difficult to state guidelines for clinical practice and further research is required.

Future research should implement standardised exercise regimes and specifications regarding the settings and type of vibration plate. Similarly, standardising follow-up intervals and including long-term follow-up would provide greater consistency regarding results and the effect of combined therapy in the long-term management of a chronic condition such as COPD. Furthermore, greater attention should be paid to using valid, context-appropriate measurement tools that are specific enough to identify changes in COPD-related outcomes to ensure that results are valid.

In conclusion, level II evidence suggests that short-term FEC may be improved more effectively by supplementing strength training with WBVT for adults with mild-to-severe COPD. Combining strength training with WBVT may not be more effective than strength training alone for improving isolated knee extension strength in adults with severe COPD. However, limited evidence (single study with a small sample size) suggests that leg press strength may improve after 12 weeks in those with mild-to-moderate COPD. Combined training was not found to be more effective than strength training alone for improving pulmonary function and evidence remain limited for QoL. Even though the current evidence suggests that if physiotherapists have the resources, and their primary goal of therapy is to improve FEC, they may consider combining strength training with WBVT, no recommendation can be made regarding optimal dosing and protocols. The results of our systematic review remain limited by the small number of available RCTs for inclusion and small sample sizes.