Whole-body vibration training in obese subjects: A systematic review

Objective (i) to determine the outcomes of whole-body vibration training (WBVT) on obese individuals, and the intervention settings producing such effects; (ii) identify potential improper or harmful use of WBVT. Design Systematic review. Data sources Medline, Scopus, Web of Science, PEDro and Scielo until July 2018. Eligibility criteria Full papers evaluating the effect of WBVT on body composition, cardiovascular status and functional performance in obese adults. Papers with PEDro score<4 were excluded. Study appraisal and synthesis Risk of bias and quality of WBVT reporting were assessed with PEDro scale (randomized controlled trials) or TREND checklist (non-randomized studies) and a 14-items checklist, respectively. Weighted acceleration, daily exposure and Hedges’ adjusted g were computed. Results We included 18 papers published 2010–2017. Typical interventions consisted in three sessions/week of exercises (squats, calf-raises) performed on platforms vibrating at 25–40 Hz (amplitude: 1–2 mm); according to ISO 2631–1:1997, daily exposure was “unsafe” in 7/18 studies. Interventions lasting ≥6 weeks improved cardiac autonomic function and reduced central/peripheral arterial stiffness in obese women; 10 weeks of WBVT produced significant weight/fat mass reduction, leg strength improvements as resistance training, and enhanced glucose regulation when added to hypocaloric diet. No paper evidenced losses of lean mass. Isolated cases of adverse effects were reported. Summary To date, WBVT is a promising adjuvant intervention therapy for obese women; long-term studies involving larger cohorts and male participants are required to demonstrate the associated safety and health benefits. The therapeutic use of WBVT in the management of obese patients is still not standardised and should be supported by an extensive knowledge on the causality between vibration parameters and outcomes.


Introduction
Obesity is a severe and growing health problem which increases the risk of debilitating and death-leading diseases [1,2].In obese individuals, hypertriglyceridemia and insulin resistance lead to impaired fasting glucose, high blood sugar levels, inflammation and visceral adipose tissue (VAT) accumulation [3].These elements contribute to several adverse cardiovascular outcomes, partly due to cardiac autonomic dysfunction [4][5][6][7]: hypertension, increased sympathovagal balance and arterial stiffness, reduced heart rate (HR) variability, endothelial dysfunction, and eventually enhanced risk of coronary heart disease, stroke and cardiovascular death [3,4,8,9].Cardiovascular complications are often associated with leg sarcopenia (loss of muscle mass), with consequent postural instability and increased risk of falls [10][11][12].Menopause and aging additionally impact upon the health of obese individuals, aggravating cardiovascular status [13][14][15], VAT accumulation, sarcopenia, and reducing bone mineral density [16].
In obese individuals, even a modest weight loss (5-10% of body weight) helps to alleviate cardiovascular risk [3].Thus, weight and VAT loss are primary treatment goals, conventionally achieved through dietary modifications [3,17,18], behavioural correction and/or exercise prescription [19][20][21].However, the success rate of therapy for obesity is very low: dieting may work in the short term, but severe dietary restriction alone reduces muscle mass and leads to a decline in physical fitness [22]; traditional exercise, such as aerobic and resistance training, improves heart rate variability, physical strength and body composition [23,24].However, the majority of obese people maintains a sedentary lifestyle and is reluctant to enrol and persist in conventional exercise programs due to physical limitations, musculoskeletal discomfort and lack of self-motivation [25].
In the last two decades, whole-body vibration training (WBVT) emerged as an alternative exercise modality for strength training [26][27][28].WBVT involves exercising on a vibrating platform.Vibrations mechanically generate rapid variations in the length of the muscle-tendon complex [29], stimulating repetitive eccentric-concentric muscular work and reflexive muscle contractions [30,31].WBVT was first recognized as an alternative to resistance exercise for its ability in enhancing force and power in skeletal muscle [28,32].Evidence that body vibrations slow down fat accumulation and reduce adipogenesis in rats [33,34] suggested a possible clinical use of WBVT in the treatment of obesity.Indeed, WBVT improved body composition [35,36], muscle strength [28] and cardiovascular function in various populations, including obese individuals [37,38].
To date no systematic review summarised the outcomes of WBVT on obese subjects: although WBVT is gaining growing interest as an exercise prescription for obese patients, there is no clear consensus about the vibration exposure (i.e.amplitude, frequency, duration), and exercise performed on the platform to obtain positive effects and avoid over-training or injuries [39].Therefore, this systematic review aimed at (i) defining the outcomes of WBVT on obese individuals and which combination of vibration and exercise setting enables to attain such effects, and (ii) identifying gaps of knowledge that may lead to improper use of WBVT with consequent harmful effects.

Methods
This systematic review was undertaken using the preferred reporting guidelines for systematic reviews and meta-analysis (PRISMA) [40].

Search strategy
A systematic literature search was performed in July 2018 on the following electronic databases (from 1990): Web of Science, PubMed MEDLINE, Scopus, Mendeley, PEDro, Scielo.Customised queries including keywords and Boolean logic with AND/OR operators were entered in the search engines in this form: ("whole-body vibration" OR "vibrating platforms" OR "vibration training") AND ("obese" OR "obesity"), with document type set to "Article".
The search was limited to full original articles written in English and to investigations on human subjects.Bibliographies of identified papers were hand searched for supplemental relevant items.

Eligibility criteria and study selection
Randomised control trials, quasi-experimental studies and observational case series were included.The PICO (P = Patients, I = Intervention, C = Comparison, O = Outcomes) method was used to define inclusion criteria: P = obese adults (age!18 years) with body mass index (BMI)!30kgÁm -2 ; studies including among them few pre-obese participants were included if the total sample BMI was on average !30kgÁm -2 ; studies assessing obese patients with documented comorbidities (i.e.Diabetes Mellitus) were also included; I = whole-body vibration training; C = (1) comparison between pre-and post-intervention, (2) comparison with no whole-body vibration under the same exercise condition, or with other forms of physical activity/intervention; O = body composition and cardiovascular status (primary outcomes), and/or measures of biomechanics/functional performance (secondary outcomes).
We excluded studies not primarily focused on the evaluation of WBVT, studies assessing the acute effects during WBVT, animal studies, studies including overweight (not obese) or non-adult participants, studies that used focal rather than whole-body vibration.Two reviewers (CF and MZ) independently screened titles and abstracts of the identified records and took decision about items retention in a blinded manner.Any disagreement between the examiners was resolved by a technical discussion involving a third reviewer (MT).

Methodological quality
The Physiotherapy Evidence Database (PEDro) scale [41] was used to evaluate whether the selected randomized controlled trials were scientifically sound (9-10 = excellent, 6-8 = good, 4-5 = fair, and <4 = poor).When the score was unavailable on the PEDro database, articles were rated independently by two researchers (MZ and CF, S1 Table ).Papers with poor PEDro score were excluded.The quality of non-randomized (quasi-experimental and case series) studies was assessed with the Transparent Reporting of Evaluations with Nonrandomized Designs (TREND) checklist [42].
The quality of whole-body vibration reporting was assessed using the checklist of the International Society of Musculoskeletal and Neuronal Interactions [43], consisting of 13 items regarding the vibration protocol plus an item addressing the supervision received during WBVT [44] (S3 Table ).

Data synthesis and analysis
A standardised data extraction and appraisal form was developed to collect the key features of each study: authors/year, study design, demographics of participants to WBVT, body composition (BMI), status and comorbidities, intervention details (vibration frequency, acceleration and displacement, duration and nature of performed exercises), main outcomes of WBVT grouped in the following domains: (i) body composition, (ii) cardiovascular parameters and (iii) hormonal concentrations, strength and functional changes.For each study, the weighted acceleration was derived by multiplying the amplitude of the vibration by the coefficient (at the stimulus frequency) of the weighting curve, and the daily exposure was computed according as per ISO 2631-1:1997, Annex B [45].
After reviewing the results of the selected studies, it was decided that a meta-analysis was not appropriate because the treatment protocols and the outcomes measures substantially varied across studies.To estimate the effect size of interventions yielding significant results, Hedges' adjusted g was computed based on the data provided in the examined articles; small, medium and large effects corresponded to g = 0.2, g = 0.5 and g = 0.8, respectively [46].

Description of the included studies
Study population.The sample size of obese participants ranged from n = 7 to n = 40 (mean age range: 20 to 59 years).A total of 321 subjects were involved, but we could not exclude subjects overlapping in studies conducted by the same research group.Sixteen out of 18 papers focused on obese women (Table 1); six of them investigated the effect of WBVT after menopause [26,49,50,52,57,59].Four studies assessed a small male sample (n = 4-10 [45,48,52,53]).The majority of studies involved obese participants with no further pathologies, apart from [62] and [61] who included patients diagnosed with type-2 Diabetes Mellitus and fibromyalgia, respectively; Wong et al. [49] and Figueroa et al. [48] included prehypertensive and hypertensive women with brachial systolic blood pressure (BP) higher than 120 mmHg.
Intervention.WBVT mainly consisted in a series of exercises performed on the platform, namely squats (at different degrees of knee flexion) and calf-raises (Table 2); two studies prescribed static upright standing sessions with a knee flexion angle ranging from 20 to 45 degrees [47,61].Interventions lasted 6 to 12 weeks, with a frequency of three sessions per week in 13/ 18 studies.A single trial evaluated the effect of a 24-week intervention and related long-term effects [58].Platforms used applied synchronous vertical vibrations in 9/18 studies and sidealternating (or rotating) vibrations in 2/18 studies; seven studies did not provide vibration type.Vibration frequency ranged from 12.5 to 60 Hz, while the most common choice was between 25 and 40 Hz, with a peak-to-peak displacement (amplitude) of 1-2 mm.Exercises bouts lasted 30 to 60 s, with a work:rest ratio from 1:1 to 1:2.According to ISO 2631-1:1997, daily exposure was "tolerable" in 5/18 studies, "unsafe" in 7/18 studies and not available due to lack of data in the remaining studies.

Quality of the included studies
Of the 16 included RCT papers, 9 were deemed with "fair", and 7 with "good" methodological quality according to the PEDro score.Eligibility (item 1 in the PEDro scale and item 3 in the TREND checklist) was fulfilled by 12/18 papers (S1 and S2 Tables); all papers presented between-groups comparison and point/variability estimates (items 10-11 in the PEDro scale and item 17 in the TREND checklist); 16 papers randomly allocated subjects; 17 papers reported similar demographics and anthropometrics at baseline.Rather, only few studies could dispose of blind assessors.
Twelve studies investigated the cardiovascular response to WBVT.Ten articles reported changes in arterial stiffness: systolic/diastolic BP decreased with a large effect in [26,49,55,56,60]; Figueroa et al. [48] observed a large-effect reduction in ankle systolic BP only in the hypertensive group.Carotid-femoral and/or brachial-ankle Pulse Wave Velocity (PWV) also decreased by 40-100 cmÁs -1 (large effects) [26,55,59,60].Blood flow velocity increased by 35 mlÁmin -1 with a small to medium effect [62].The augmentation index (AIx and AIx adjusted to 75 bpm, indicators of wave reflection computed as difference between the second and first systolic peak of aortic pulse pressure) reduced (large effects) [48,49,55].Wong et al. found that transit time of reflexion wave increased more in the group supplemented with L-Citrulline (large effect) than after WBVT alone [50].Positive effects of WBVT in improving cardiac autonomic modulation by decreasing sympathovagal balance and/or decreasing resting heart rate (5-10 bpm) were reported consistently with large effects in young [56], healthy [52], and pre-or stage 1-hypertensive postmenopausal women [49]: low-to-high respiratory frequency power ratio (or combination thereof) improved in [49,52,56]; total power did not significantly change in [49,52] but increased in [56]; R-R intervals decreased with a large effect [52].
Three studies found hormones and hematic lipids concentrations changes following the intervention: Bellia et al. reported a decrease of fasting insulin level (-30 pmolÁl -1 , large effect), a slight decrease in leptin levels (medium effect) and an increase in adiponectin levels (large effect) [63].Miyaki et al. described a reduction in LDL cholesterol and triglycerides concentration (large effect) [60].Nitric oxide (NOx) concentration increased with medium effects in [49].
[26] BW, % body fat and lean mass of arms and legs did not change (p>0.05).
(Continued) (11.0±15.4s) and sit-and-reach (6.5±4.8 cm) in the group undergoing WBVT and diet [54]; Yang et al. found an enhanced dynamic stability in terms of a larger decline in fall rate in WBVT (-45%) than in the placebo group (-25%) [47].Miyaki et al. reported an increase in peak oxygen uptake (large effect); Wilms et al. found a positive effect of WBVT in enhancing resting energy expenditure (large effect) [51].

Discussion
Six to twelve weeks of WBVT in obese individuals generally led to a reduction in fat mass and cardiovascular improvements.However, the results of the examined studies are various and sometimes inconsistent or inconclusive.Thus, they should be interpreted the light of the specific cohorts and intervention.

Outcomes of vibration training
Sinusoidal vibrations stimulate the primary endings of the muscle spindles, which in turn activate α-motor neurons and induce rapid eccentric-concentric involuntary contractions; this

Domain Reference Body composition Cardiovascular parameters Hormonal, hematic and functional parameters
[59] % Body fat decreased in both groups (g = 0.74, p).Lean mass index increased only in WBV+L (g = 1.00).
Three factors may contribute to fat mass reduction: (i) the acute exposure to vibrations activates the central sympathetic nervous system, whose innervation of white adipose tissue triggers lipolysis [65]; (ii) WBVT enhances glycemic control by improving insulin action and glucose regulation [49,63,66]; (iii) WBVT promotes GH release [67], which stimulates metabolism and is usually reduced in obese subjects [68].
The improved glycaemic control is crucial in postmenopausal women, whose hormonal changes lead to insulin resistance [69]: Bellia et al. found a 35% increase of insulin sensitivity following 8 weeks of WBVT with static squats [63]; additional effects on metabolic regulation were an increase of adiponectin and a decrease of leptin levels [63].In patients with type-2 Diabetes Mellitus, insulin-mediated glucose uptake in the skeletal muscle improved, probably due to increase in femoral artery blood flow [62].
Bone mineral density negatively correlates with aging and obesity [70]; after menopause, decreasing estrogen concentrations cause a further decline in bone mineral density, which can lead to osteoporosis.Fluid movement produced by vibrations is anabolic to the bone [30], as generates shear stresses on the plasma membranes of highly sensitive cells like resident osteocytes, bone lining cells and osteoblasts [71].
Cardiovascular effects.Eleven papers examined the role of WBVT in mitigating the adverse cardiovascular outcomes involved with obesity (i.e.increased sympathovagal balance and arterial stiffness), and complicated by aging, menopause and Diabetes.There is compelling evidence that at least 6-weeks of WBVT can reduce sympathovagal balance [49,52,56] and central/peripheral arterial stiffness [26,48,50,55,59,60] in obese women-only one paper addressed these topics in men [62].
An increase of Low Frequency (LF) and a decrease of High Frequency (HF) power of heart rate variability spectrum indicate a concurrent decrease in sympathetic and an improvement of cardiovagal modulation, respectively.This has important clinical implications since a lower sympathovagal balance (measured as LF/HF ratio or R-R duration) is associated with reduced cardiovascular risk and greater longevity in obese women [72,73].In contrast, conventional exercise such as resistance [74] or aerobic [75] training did not improve LF or LF/HF ratio in postmenopausal women.
L-Citrulline is a non-protein amino acid naturally found in watermelon which is efficiently converted to L-arginine, the substrate for endothelial production of NOx [76].Decrement in aortic systolic and diastolic BP by 5-10 mmHg corresponds to 30-40% reduction in the risk of death due to stroke and other cardiovascular complications [77].Importantly, these hemodynamic effects of WBVT were not observed in studies following traditional resistance training in overweight and obese women [78][79][80].Only high-intensity aerobic training, but not highintensity resistance training, was effective in reducing AIx [81].
The underlying mechanisms of these outcomes rely on the combination of several factors.First, WBVT increases the levels of circulating angiotensin-II (inhibitor of cardiovagal activity [82]) and the local production of vasodilatory substances, including NOx [83].NOx decreases systolic BP and AIx reducing the vascular tone of small arteries [84].Second, the mechanical oscillatory contractions during vibration serve as an active muscle pump and increase stroke volume, probably enhancing venous return and preload [85].Third, total peripheral resistance to blood flow increases during body vibration.As a compensation, more capillaries are opened to keep a necessary level of cardiac output, resulting in more efficient gas and material metabolism between the blood and muscle fibers [39].Fourth, friction forces applied by the mechanical vibration on the endothelial cells [38] also contribute in enhancing blood flow [30,38,83].The weekly repetition of such acute vascular effects is likely responsible for the improvements in arterial stiffness and wave reflection [56].
Functional and other effects.As individuals with obesity fatigue at a greater rate [86], daily motor performance may be hindered.Since increased muscle strength per se may offers protection from obesity [59], a safe, time-efficient and low-intensity exercise modality in the management of obese individuals could prevent vascular complications, muscle dysfunction and physical disability [55].Although the training protocols differed between studies, from young obese women to the elderly WBVT produced a similar positive effect to resistance training on muscle strength [47,52,53,55,58], and an improvement in sit-and-reach and sit-tostand functional tests [54].
Improved leg muscle blood flow can contribute to muscle mass increase in older adults [87].With short WBVT interventions (6 weeks), Figueroa et al. [56] proposed that muscle strength gains are primarily attributed to neural adaptations rather than muscle hypertrophy.Consistently, vibration training increases the efficiency of agonist/antagonist pairs and the synchronization of motor units, which implies that more fibers are contracted at once and more force can be produced [29].This positively affects balance control: previous observations in populations prone to fall, such as frail people [30,88,89], were confirmed in obese subjects by single leg balance improvements [54] and by a decrease in fall rate [47].
Lastly, the addition of vibrations to both static and dynamic exercises seemed to significantly increase oxygen uptake in obese women [58].Although WBVT might increase oxygen consumption and caloric output when added to an exercise program [30,90], the amount of energy turnover due to vibration is modest, estimated as 4.5 mlÁmin -1 Ákg -1 (frequency: 26 Hz, amplitude: 3 mm) [30,91].We argue that the potential mechanisms of increased oxygen uptake could be the higher metabolism due to hormonal and cardiovascular changes [39,67], increase of lean mass and muscle activation [91].

Side effects
The effects of vibration training on human body may depend on vibration settings (frequency, amplitude and duration) and on the exercise program (type of exercises, intensity and volume).WBVT stimulates reflexive muscle contractions "in a safe and gentle manner" [52] depending on the vibration settings: vibrations ranged from tolerable levels (6 papers, typically 1-2 mm at 25-40 Hz, 5.1 mÁs -2 ), to several times higher than what the ISO 2631-1:1997 considers a safe threshold for workers (8 papers, up to 42 mÁs -2 ).It is worth noting that ISO 2631-1:1997 thresholds were established to minimize health threats deriving from a continuous vibration exposure in the workplace, and may not be directly transferable to medical devices and WBVT [30,92].Six papers explicitly reported no unfavourable symptoms or adverse effects resulting from the vibration stimulus [47,50,54,55,61,93].However, isolated cases of lower leg phlebitis [58], mild knee pain [94] and back pain after two weeks of training [63] were reported.
In summary, while it is presumed that the benefits of using vibrating devices offset the risks generated by exposure [92], the use of WBVT for therapeutic purposes is still not standardized and the related potential adverse effects are uncertain [51], especially the implications on brain health due to chronic exposure to accelerations.Some guidelines can be drawn, though.First, Muir et al. proposed that vibrations delivered by medical devices can be considered reasonably safe on a basis of 15 minutes of exposure/day if enclosed within the boundaries of 30-50 Hz and 2.25-7.98g [92]; consistently, in all the reviewed papers the daily duration of training was lower than 10-15 min.Second, vibrations close to the main resonant natural frequency of the human body (5-20 Hz) can produce accelerations to the head equalling or exceeding the gforce delivered at the foot, and should be prudently avoided [39].Third, transmissibility to the cranium is markedly attenuated by flexing the knees and by using side-alternating rather than synchronous vibration [92]; thus, straight-legged stance has to be avoided.

Limitations
Five major flaws emerged in the assessed papers: (i) few comparison studies, (ii) small sample sizes, (iii) potential sex and regional biases, (iv) poor quality of WBVT reporting and (v) lack of long-term investigations.i.While evidence of WBVT effectiveness exists when compared to not-exercising control groups [26,48,49,53,55,61], relevant therapeutic information could come from the comparison of different WBVT settings and exercise modalities.Unfortunately, only three studies compared WBVT to other training [51,54,63], and no study addressed the effect of different vibration settings on obese individuals.
ii. Conventionally, the number of independent variables (k) can be considered adequate if the ratio k:n exceeds 1:10, where n is the sample size [95].No article included in this review met such requirement, as all papers but [57] involved less than 26 obese subjects.Small sample sizes reduce the relevance of the results by increasing the risk of type-II error and reducing the ability to validate hypotheses.Further, small sample size did not allow to evaluate different intervention settings and exercises, as admitted by [49,63,94].
iii.Hemodynamic effects [96], systolic BP and AIx [13,15], insulin and GH hormones concentration present gender-specific differences.This limits the generalization of results found on women, as WBVT may not have the same effects on obese men [93].Further, some results were attributed to very specific cohorts (i.e.Korean [94] or Hispanic [55]), and cannot be generalized to any obese population due to regional-specific features of minority racial/ethnic groups, as the higher incidence of reduced HR variability in Hispanics [97].
iv.Given the potential damage of WBVT training, supervised use of devices by trained therapists should be recommended [92].It is likewise essential to describe whether study participants were standing freely on the device or were holding on to some support, and the type of footwear, which influences the transmissibility to the body structures [43].However, a third of the examined paper did not mention any supervision; type of support was indicated in five studies and footwear in two; skidding and accuracy of vibration parameters was never reported.These details may have introduced additional variability and played a role as confounding factors.
v. Lastly, 11/18 papers presented results after 6 or 8 weeks, a relatively short period of intervention; how long beneficial effects persist after the intervention remains an open question, like the potential risk of WBVT-related side effects in the long term [63].Results by Vissers et al. [58], showing that patients treated with WBVT succeeded in maintaining a weight loss of 10% at 12 months, are encouraging but need to be more extensively confirmed.

Conclusions and recommendations for future research
Whole-body vibration training is a promising adjuvant intervention therapy for obese women.
WBVT could be prescribed without additional exercise in the initial stages of a weight loss program, due to the limited stress upon joints and the GH-mediated stimulation of metabolism, without provoking an excessive fatigue [63,85].Passive vibrations do not involve voluntary movement and require a lower contribution of central command [98].Thus, WBVT could be a useful mode of exercise for deconditioned obese with poor motivation [56]: when combined with dietary intervention or prescribed as alternative to traditional exercise training, WBVT is as effective as aerobic and resistance exercise in reducing fat mass [58,93] and moderating the deficit of the relative muscle strength [49,54].Lastly, WBVT may be effective in vascular health promotion and prevention in young obese women [50,55,59].
The positive potential of WBVT for obese individuals is partly hindered by methodological inconsistencies in the existing literature.The lack of results on obese males, and the small (and occasionally biased) cohorts suggest further research to standardise WBVT in the management of obesity.Two additional main directions of research emerged: first, it is mandatory to make light on the causality between the vibration parameters (frequency, amplitude, exposure, work: rest ratio) and the WBVT outcomes.Second, long-term studies are required to demonstrate safety and health benefit that can be achieved with WBVT for obese patients.

Table 3 . Outcomes of whole-body vibration training.
Hedges adjusted g effects size and statistical significance were reported for each variable.