The Effect of Reactive Balance Training on Falls in Daily Life PDF

PTJ: Physical Therapy & Rehabilitation Journal | Physical Therapy, 2023;103:1–15 https://doi.org/10.1093/ptj/pzac154 Advance access publication date November 9, 2022 Review The Effect of Reactive Balance Training on Falls in Daily Life: An Updated Systematic Review and Meta-Analysis Augustine Joshua Devasahayam , PhD1 ,* , Kyle Farwell1 ,2 , Bohyung Lim1 ,2 , Abigail Morton3 , Natalie Fleming4 ,5 , David Jagroop, CSEP-CEP, MHSc1 , Raabeae Aryan, PhD1 ,6 , Tyler Mitchell Saumur, PhD1 ,6 ,7 , Avril Mansﬁeld , PhD1 ,7 ,8 1 KITE-Toronto Rehabilitation Institute, University Health Network, Toronto, Ontario, Canada Downloaded from https://academic.oup.com/ptj/article/103/1/pzac154/6812867 by guest on 02 January 2024 2 Faculty of Kinesiology and Physical Education, University of Toronto, Toronto, Ontario, Canada 3 Faculty of Health, University of Waterloo, Waterloo, Ontario, Canada 4 École de Readaptation, Université de Sherbrooke, Sherbrooke, Quebec, Canada 5 École de Kinésiologie et de Loisir, Université de Moncton, Moncton, New Brunswick, Canada 6 Rehabilitation Sciences Institute, University of Toronto, Toronto, Ontario, Canada 7 Evaluative Clinical Sciences, Hurvitz Brain Sciences Program, Sunnybrook Research Institute, Toronto, Ontario, Canada 8 Department of Physical Therapy, University of Toronto, Toronto, Ontario, Canada *Address all correspondence to Dr Devasahayam at: [email protected] Abstract Objective. Reactive balance training (RBT) is an emerging approach to reducing falls risk in people with balance impairments. The purpose of this study was to determine the effect of RBT on falls in daily life among individuals at increased risk of falls and to document associated adverse events. Methods. Databases searched were Ovid MEDLINE (1946 to March 2022), Embase Classic and Embase (1947 to March 2022), Cochrane Central Register of Controlled Trials (2014 to March 2022), and Physiotherapy Evidence Database (PEDro; searched on 22 March 2022). Randomized controlled trials of RBT were included. The literature search was limited to the English language. Records were screened by 2 investigators separately. Outcome measures were number of participants who reported falls after training, number of falls reported after training, and the nature, frequency, and severity of adverse events. Authors of included studies were contacted to obtain additional information. Results. Twenty-nine trials were included, of which 17 reported falls and 21 monitored adverse events. Participants assigned to RBT groups were less likely to fall compared with control groups (fall risk ratio = 0.76; 95% CI = 0.63–0.92; I 2 = 32%) and reported fewer falls than control groups (rate ratio = 0.61; 95% CI = 0.45–0.83; I 2 = 81%). Prevalence of adverse events was higher in RBT (29%) compared with control groups (20%). Conclusion. RBT reduced the likelihood of falls in daily life for older adults and people with balance impairments. More adverse events were reported in RBT than control groups. Impact: Balance training that evokes balance reactions can reduce falls among people at increased risk of falls. Older adults and individuals with balance problems were less likely to fall in daily life after participating in RBT compared with traditional balance training. Lay Summary. If you are an older adult and/or have balance problems, your physical therapist may prescribe reactive balance training rather than traditional balance training in order to reduce your likelihood of falling in daily life. Keywords: Accidental Falls, Exercise Therapy, Mobility Limitation, Physical Therapy, Postural Balance Received: January 31, 2022. Revised: May 11, 2022. Accepted: August 19, 2022 © The Author(s) 2022. Published by Oxford University Press on behalf of the American Physical Therapy Association. All rights reserved. For permissions, please email: [email protected] 2 Does Reactive Balance Training Prevent Falls? Introduction condition that increases the risk of falling; (4) included a con- Falls are a common cause of injuries and deaths among trol group that did not complete RBT; (5) randomly allocated people with balance impairments, such as older adults.1 Fall- participants to RBT and control groups; and (6) reported related injuries result in significant health care costs to older information about falls in daily life after the intervention adults and pose an economic burden to society.2 Exercise that and/or adverse events. RBT was defined as a training method includes balance training can reduce the rate of falls by about where participants intentionally experience repeated loss of 25%.3 However, individualized exercise approaches seem to balance, with a goal of evoking balance reactions such as be more effective in reducing falls risk than others (eg, based stepping or reach-to-grasp responses, so that participants can on the frequency, intensity, type, and duration of training).3 practice and improve control of balance reactions. The loss of Effective balance control involves the ability to sustain balance can either be caused by an external force (eg, a moving postures, move between postures, and to react to perturba- platform, push or pull from a therapist) or by the participant’s tions (loss of balance).4 A fall happens when a person fails inability to maintain balance during voluntary movement. to respond to a balance perturbation5 ; therefore, impaired Data Sources and Search Strategy Downloaded from https://academic.oup.com/ptj/article/103/1/pzac154/6812867 by guest on 02 January 2024 balance reactions may increase the risk of falls. Indeed, a meta- analysis of 61 studies including 9536 older adults reported The literature search was conducted in MEDLINE ALL (in that impaired reactive stepping was a significant risk factor for Ovid, including Epub Ahead of Print, In-Process & Other falls among older adults.6 “Conventional” balance exercises Non-Indexed Citations, Ovid MEDLINE Daily; 1946 to mostly involve maintaining balance during static postures March 21, 2022), Embase (in Ovid, including Embase Classic; or during movement.3 Reactive balance training (RBT) is a 1947 to March 21, 2022), Physiotherapy Evidence Database type of balance training where participants repeatedly expe- (PEDro; searched on March 22, 2022), and Cochrane Central rience balance perturbations, and execute balance reactions Register of Controlled Trials (Ovid; 2014 to March 21, to prevent a fall.7 RBT has been shown to improve control 2022) databases by an information specialist. Because the of balance reactions, and therefore it might help individuals terminology related to RBT (also referred to as perturbation- respond to a loss of balance and prevent a fall in daily based balance training) did not have standardized keywords life.8–10 Indeed, we previously reported in a meta-analysis or Medical Subject Headings (MeSH) terms, a search of small randomized controlled trials that RBT reduces the strategy involving multiple relevant keywords was used. A rate of falls in daily life by almost half among people with sample search strategy is provided as supplementary material increased fall risk, compared with other types of exercise or (Suppl. Appendix A). no intervention.7 This previous review7 was completed in Selection Process 2015; more randomized controlled trials of RBT have been published since then. The titles of articles retrieved from the databases were The purpose of this study was to determine the effect of screened by 2 reviewers independently, and clearly ineligible RBT, compared with either no or an active non-RBT control titles were removed. The abstracts of the remaining articles intervention, on falls in daily life in people at increased risk of were screened again by 2 reviewers independently to falls. Clinical interventions must balance intervention efficacy determine eligibility for this review. The full text of the article and harms.11 Therefore, our secondary aim was to document was read by the reviewers if they were unable to determine the nature, frequency, and severity of adverse events due to eligibility from the abstract. Disagreements between reviewers RBT, and to determine if there is an increased prevalence regarding inclusion of a study were resolved via discussion, of adverse events with RBT compared with other types of or consultation with the study team if necessary. The exercise. reference lists of relevant review articles were also screened by the 2 reviewers using steps mentioned above to identify studies that might be eligible for this review. Authors of Methods articles, including study protocols, that met all but the last Study Design inclusion criterion (reported data on falls in daily life and/or adverse events), were contacted to determine if these data This study is an update of a previous systematic review existed. Bibliographic references were managed using Endnote and meta-analysis,7 registered with PROSPERO database (version X5; Thomson Reuters, Philadelphia, PA, USA). (CRD42020220552),12 conducted according to the Cochrane Handbook for Systematic Reviews of Interventions,13 and Data Collection Process reported using the PRISMA (Preferred Reporting Items for The following data were extracted from the articles selected Systematic Reviews and Meta-Analyses) statement for report- for this review: details of the population studied; details of ing systematic reviews and meta-analyses of randomized the RBT and control interventions; number of participants controlled studies.14 allocated to RBT and control groups; falls monitoring dura- tion; number of participants in each group reporting 1 or Eligibility Criteria more falls after completing the interventions; total number Studies that met the following criteria were included in this of falls per group reported after the interventions; and the review: (1) published in English; (2) an experimental investi- nature, frequency, and severity of adverse events reported gation of RBT; (3) included people at increased risk of falls in RBT and control groups. The trial protocols and sec- due to impaired balance control, such as apparently healthy ondary publications were accessed to obtain the required older adults (60 years or older), people with neurological information, if necessary and available. The corresponding conditions (eg, stroke or Parkinson disease), people with authors of included studies were also contacted for additional chronic obstructive pulmonary disease, people with lower details and to obtain any missing information. The data were extremity amputations or joint replacement, or any other extracted and compiled into a Microsoft Excel spreadsheet. Devasahayam et al 3 The 3 outcomes of interest from studies included in this Reporting Bias Assessment review were: (1) number of participants who experienced 1 Funnel plot asymmetry was used to examine bias in the or more falls after the end of the intervention; (2) number results of the meta-analyses of risk of falls and rate of falls of falls in daily life after the end of the intervention; and (3) in accordance with the recommendations of Sterne et al.20 nature, frequency, and severity of adverse events during the Two independent reviewers evaluated the possible sources of intervention. Adverse events were deemed to be severe if they asymmetry on funnel plots due to selective nonreporting or were life threatening or resulted in permanent harm, moderate nonpublication of findings from the studies, poor method- if they required hospitalization or resulted in withdrawal from ological quality in studies with inflated effects, true hetero- the intervention (such that the participant no longer received geneity between studies (eg, due to difference in intensity the benefits of the intervention), or mild otherwise. of interventions), sampling variation, and random chance.21 Disagreement between reviewers during the reporting bias Methodological Quality and Risk of Bias assessment was resolved by consensus after discussion or by Assessment an independent third reviewer. Downloaded from https://academic.oup.com/ptj/article/103/1/pzac154/6812867 by guest on 02 January 2024 Studies included in this review were rated for methodological and reporting quality by 2 reviewers independently using Certainty Assessment the PEDro scale.15 Disagreement during data extraction and Certainty of evidence from studies that contributed data to quality assessment of the included articles was resolved by the meta-analysis of risk of falls was assessed separately by consensus after discussion between reviewers. If an agreement 2 reviewers using 5 Grading of Recommendations Assess- was not reached between 2 reviewers, the final scores were ment, Development and Evaluation (GRADE) domains (risk assigned by an independent third reviewer. The final PEDro of bias, inconsistency, indirectness, imprecision, and other scores were classified as good for 6 or greater, fair for 4 to considerations).22 The labels from GRADE certainty of evi- 5, and poor for 3 or less, on the basis of previous sensitivity dence assessment—that is, not serious, serious, and very seri- analyses with cutoffs set at 4 and 6.15,16 ous—were used to assess the body of evidence we evaluated Risk of bias was evaluated independently by 2 reviewers through meta-analysis.23–25 These labels correspond to high, for all studies that assessed risk of falls using the revised moderate, and low certainty of evidence for randomized con- Cochrane Risk of Bias Tool in 5 domains: (1) risk of bias trolled studies.23–25 The Cochrane Handbook for Systematic arising from the randomization process; (2) risk of bias due Reviews of Interventions was used to conduct the certainty to deviations from the intended interventions; (3) missing assessment and prepare the summary of findings table through outcome data; (4) risk of bias in measurement of the outcome; GRADEpro GDT software (GRADEpro guideline develop- and (5) risk of bias in selection of the reported results.17,18 ment tool; McMaster University and Evidence Prime Inc, The studies were rated as “high risk” of bias if at least 1 Hamilton, Ontario, Canada).26 Disagreement between the of 5 domains was at a high risk of bias or if there were reviewers during certainty assessment was resolved by con- “some concerns” of bias for multiple domains; rated as “some sensus after discussion or by an independent third reviewer. concerns” if at least 1 of 5 domains had “some concerns” of bias with no other domains at “high risk” of bias; and rated Role of the Funding Source as “low risk” of bias if all 5 domains were at “low risk” of The funding agencies played no role in the conception, design, bias.17,18 Disagreement between reviewers during the risk of conduct, data analysis, interpretation of findings, or reporting bias assessment was resolved by consensus after discussion, of this study. or by an independent third reviewer if consensus could not be reached. Results Data Analysis Study Selection Data extracted from included studies were analyzed using For the purposes of this review, a “record” refers to a doc- Review Manager (RevMan) software (version 5.4.1; The ument indexed in a database, a “report” refers to a doc- Cochrane Collaboration, Copenhagen, Denmark). Hetero- ument supplying information about a study (eg, paper or geneity was assessed using the chi-square test of heterogeneity abstract), and a “study” refers to an investigation of 1 or more and the I2 statistic.19 Data from included studies were interventions in a group of participants27 ; a study can have pooled using random-effects modeling when heterogeneity several reports (eg, protocol paper, primary publication, sec- was significant (>30%).13 A subgroup meta-analysis was ondary publication[s]). A total of 6614 records were identified performed that excluded studies with a no-intervention through database searching and an additional 284 records control group. The effects of RBT on falls in daily life were identified from other sources. The records included in compared with control interventions were reported as fall our previous review7 were also screened (n = 9). risk ratios and fall rate ratios, where the fall risk ratio After removing duplicates, the remaining records (n = 4159) compared the number of participants who experienced 1 or were screened for eligibility through a title and abstract more falls between groups, and the fall rate ratio compared search. Full texts of remaining reports (n = 589) were assessed the number of falls between groups. For the overall effects, for eligibility. One study included in the previous review 95% CIs were calculated , and the significance level was was removed from this review due to pseudo-random set at.05 for all analyses. The total numbers of participants group allocation.28 After screening, 44 reports of 29 studies with adverse events during RBT and control group training were included in this review (Fig. 1). All reports for each were summarized in percentages and compared using the included study are listed in Supplementary Appendix B. The Fisher exact test through SAS software (SAS Institute Inc, characteristics of the studies included in this review are Cary, NC, USA). presented in Table 1. 4 Does Reactive Balance Training Prevent Falls? Downloaded from https://academic.oup.com/ptj/article/103/1/pzac154/6812867 by guest on 02 January 2024 Figure 1. Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) ﬂow diagram showing the selection of studies included in the review. RBT = reactive balance training. Study Characteristics Risk of bias, assessed using the revised Cochrane Risk of Participants in studies included in this review were community- Bias Tool17,18 in the studies that reported risk of falls and living older adults9,28–40 ; adults with Parkinson disease,41–47 rate of falls (n = 17 in total), was high in 12 studies, found stroke,8,48–51 cerebral palsy,52 and spinal cord injury53 ; to have some concerns for risk of bias in 1 study, and low and long-term care residents.54,55 All studies had RBT in 4 studies (Suppl. Appendix D). Considering the findings as either a sole or a significant component of the exper- from each of the 5 domains, a high risk of bias arose from imental intervention. Three studies did not provide any the randomization process (7 of 17), due to deviations from intervention to the control group,40,41,44 and the remaining the intended interventions (3 of 17), missing outcome data studies used non-RBT exercise for the control interven- (7 of 17), and a risk of bias in measurement of the outcome tion. The loss of balance during RBT was caused by (6 of 17). manual perturbations,8,40–43,47,50–53 challenging balance tasks to evoke internal perturbations,8,31,53 single-axis or multidirectional perturbations on a treadmill or a Results of Synthesis movable platform,9,30,32–36,44–46,48,49,54,55 slip and trip Effect of RBT on Falls in Daily Life perturbations on a walkway with movable tiles/platforms and Seventeen studies monitored falls in daily life after the inter- tripping obstacles,34,37,38 and mediolateral tilt perturbations vention (Tab. 1); the falls-monitoring duration ranged from while riding a stationary bicycle.29 The frequency of RBT 2 weeks to 1 year. There were 1241 participants in 15 studies ranged from 1 training session total33,37,38 to 3 times per that reported the number of participants who experienced 1 week.9,35,39–41,43,44,47,50,53–55 The duration of RBT training or more falls in daily life after the intervention (635 were ranged from 15 minutes38 to 1.5 hours52 per session; most assigned to the RBT groups and 606 to the control groups). studies (21 of 29) conducted RBT for 30 to 60 minutes per Eleven of the 15 studies with fall risk data reported that session. fewer participants in the RBT groups experienced falls fol- lowing training compared with the control groups (Tab. 2). Participants who completed RBT were less likely to fall than those in the control group; the overall risk ratio for all 15 Methodological Quality and Risk of Bias in Studies studies combined was 0.76 (95% CI = 0.63–0.92; P =.005; The methodological and reporting quality assessed using I2 = 32%) (Fig. 2A). The risk ratio was unchanged when PEDro scale was good (≥6) in 21 studies, fair (4 to 5) in only studies with an active exercise control intervention were 7 studies, and poor (≤3) in 1 study (Suppl. Appendix C). included in the analysis (13 studies; risk ratio = 0.78; 95% Of 29 studies, 21 reported conducting an intention-to-treat CI = 0.68–0.89; P =.0003; I2 = 6%) (Suppl. Appendix E). analysis. None of the included studies masked participants to There were 1281 participants in 16 studies that reported the intervention they received. In 1 study,43 the therapists did the number of falls experienced by participants in daily life not know whether they were providing the experimental or after completing the intervention (648 were assigned to RBT control intervention. groups, and 633 to the control groups). Twelve of the 16 Table 1. Characteristics of the Included Studiesa Fall Monitoring of Study Location Sample RBT Intervention Control Intervention Monitoring Adverse Duration Events Toole et al41 (2000) USA People with Parkinson disease 15 posterior manual pulls per session, in No intervention None Yesb addition to lower extremity resistance training and standing balance exercises; three 1-h Devasahayam et al sessions/wk for 10 wk Shimada et al55 (2004) Japan Frail long-term-care residents at Usual exercise plus perturbed walking on a Individualized “usual” exercise 6 mo No high risk of falls treadmill; 600 min over 6 mo (1–3 times/wk) (eg, physical therapy for pain, stretching, resistance training) Protas et al44 (2005) USA Men with Parkinson disease Gait and step training with treadmill-based No intervention 2 wk No (Hoehn and Yahr stages 2–3) perturbations to stance, with 25–35 perturbations/session; three 1-h sessions/wk for 8 wk Marigold et al50 (2005) Canada People with chronic stroke (≥12 Agility training that included manual Stretching and “Tai-chi”-like 12 mo Yes mo after stroke, >50 y old) multidirectional push perturbations, in sets of weight-shifting exercises; three 10 perturbations, typically with 3 sets/session; 1-h sessions/wk for 10 wk three 1-h sessions/wk for 10 wk Maki et al39 (2008) Canada Older adults attending a falls Multidirectional moving platform Voluntary step and reach-to-grasp 12 mo Yesb prevention program perturbations to stance to train reactive training; three 30-min sessions/wk stepping and grasping; three 30-min for 6 wk sessions/wk for 6 wk Mansfield et al9 (2010) Canada Healthy older adults (64–80 y old) Moving platform perturbations to stance, with Stretching and relaxation 12 mo Yes 48–64 perturbations/session; perturbation exercises; three 30-min magnitude increased when participants could sessions/wk for 6 wk consistently respond to the current magnitude with a single step; three 30-min sessions/wk for 6 wk Smania et al43 (2010) Italy People with idiopathic Parkinson Balance training program that included Joint mobilization, muscle 1 mo Yes disease and postural instability external perturbations (push/pull from a stretching, and motor (Hoehn and Yahr stages 3–4) physical therapist); 10 min/session; three coordination exercises not 50-min sessions/wk for 7 wk designed to improve balance control; three 50-min sessions/wk for 7 wk Parijat and Lockhart38 USA Healthy older adults (>65 y old) 12 slip perturbations caused by the movement Unperturbed overground walking; None Yesb (2012) of a ovable platform that slid forward at the one 15- to 20-min session transition between double and single support during overground walking; one 15- to 20-min session Lurie et al32 (2013) USA Older adults at risk of falling (age Standard physical therapy + treadmill-induced Individualized “standard” 3 mo No range: 65–96 y) perturbations; 1–19 sessions with a mean physical therapy, including duration of 44 min strength, flexibility, balance, and mobility exercises, 3–17 sessions with a mean duration of 43 min Halvarsson et al40 (2013) Sweden Community-dwelling older adults External perturbations provided by the 12 mo No (≥65 y old) with a fear of falling instructor or partner and “typical” balance and/or with an experience of a fall exercises with a focus on maintaining balance; during the previous 12 mo three 45-min sessions/wk for 12 wk Pai et al37 (2014) USA Community-dwelling older adults 24 slips to overground walking via a pair of 10 unperturbed overground 12 mo No (≥65 y old) side-by-side low-friction movable platforms walking trials; 1 session that slid forward by 0.9 m; 1 session 5 (Continued) Downloaded from https://academic.oup.com/ptj/article/103/1/pzac154/6812867 by guest on 02 January 2024 6 Table 1. Continued. Fall Monitoring Study Location Sample RBT Intervention Control Intervention Monitoring of Adverse Duration Events Morgan et al52 (2015) Australia Adults with cerebral palsy and Strength and balance training (30 min of Guided relaxation, seated yoga, None Yes able to walk independently with each), including dynamic gait activities and Tai-chi; one 1.5-h session/wk or without a gait aid and practice of anticipatory and reactive for 8 wk dynamic balance strategies in standing and steppingb ; one 1.5-h session/wk for 8 wk Schlenstedt et al42 (2015) Germany People with idiopathic Parkinson Manual perturbations (shoulder pulls by Lower extremity resistance None Yes disease (Hoehn and Yahr stages the physical therapist); two 60-min training; two 60-min sessions/wk 2.5–3.0) sessions/wk for 7 wk for 7 wk Shen and Mak45 (2015) Hong People with idiopathic Parkinson Treadmill walking with multidirectional Hip, knee, and ankle 12 mo Yes Kong disease, stable on perturbations; overground walking with strengthening exercises using antiparkinsonian medications, manual perturbations provided by the dynamometers and leg-press and able to walk independently for physical therapist; 12 wk, with 2 phases of rowing machines; 12 wk, with 2 10 m, and able to follow 4-wk laboratory-based training phases of 4-wk laboratory-based instructions (60 min/session) separated by 4 wk of training (60 min/session) home-based training (20 min/session) separated by 4 wk of home-based training (20 min/session) Kumar and Pathan51 (2016) India People with subacute stroke (≤3 Multidirectional manual perturbations to Conventional physical therapy, None Yesb mo after stroke), 40–70 y old, and sitting, kneeling, and standing (10 including range-of-motion ambulatory with or without aids perturbations/posture) and “conventional” exercises, resistance training, physical therapy; six 45-min sessions/wk weight bearing/shifting, reaching, for 4 wk activities of daily living, and gait training; six 45-min sessions/wk for 4 wk Kurz et al36 (2016) Israel Community-dwelling older adults Multidirectional (mediolateral or Unperturbed treadmill walking; 12 mo Yes (≥70 y old) anteroposterior) perturbations, delivered two 20-min sessions/wk for 12 wk during treadmill walking in a random order using a treadmill mounted on a movable platform; 14–42 perturbations/session; displacement: 1–18 cm; velocity: 0.1–3.2 m/s; acceleration: 0.5–16 m/s/s; two 20-min sessions/wk for 12 wk Gandolfi et al47 (2017) Italy Adults with Parkinson disease Multidirectional manual perturbations, Balance exergames using the None Yesb (Hoehn and Yahr stages 2.5–3) delivered by the physical therapist, Nintendo Wii; three 50-min completed as part of a sensory integration sessions/wk for 7 wk balance-training program that included weight bearing, transfers, and maintaining balance on unstable surfaces; 10 exercises in total, 6 of which included external perturbations; each exercise was repeated 5–10 times over 5 min; three 50-min sessions/wk for 7 wk (Continued) Does Reactive Balance Training Prevent Falls? Downloaded from https://academic.oup.com/ptj/article/103/1/pzac154/6812867 by guest on 02 January 2024 Table 1. Continued. Fall Monitoring of Study Location Sample RBT Intervention Control Intervention Monitoring Adverse Duration Events Devasahayam et al Steib et al46 (2017) Germany People with Parkinson disease Treadmill walking with perturbations; the Unperturbed treadmill walking; None Yes (Hoehn and Yahr stages 1–3.5) treadmill was mounted on a tiltable two 35-min sessions/wk for 8 wk platform; continuous, random perturbations of ≤3 cm were delivered via lifting of 3 pneumatic actuators; two 35-min sessions/wk for 8 wk Mansfield et al8 (2018) Canada Community-dwelling adults with Challenging balance tasks to evoke Traditional balance training, 12 mo Yes chronic stroke (>6 mo after internal perturbations or external manual including overground walking, stroke) perturbations (lean and release or sit-to-stand, heel raises, walking push/pull from physical therapist); 60 with an object, tap-ups, reaching, perturbations/session, with intensity set weight shifting, and standing with such that participants needed assistance to a reduced base of support; two recover, took multiple steps, or used an 1-h sessions/wk for 6 wk + two upper extremity response ∼50% of the 1-h “booster” sessions (3 mo and time; two 1-h sessions/wk for 6 wk + two 9 mo after initial training period) 1-h “booster” sessions (3 mo and 9 mo after initial training period) Aviles et al54 (2019) USA Independent residents of senior Up to 40 perturbations to stance/session, Group Tai-chi training; 12 unique 6 mo Yes housing facilities (≥70 y old) delivered using a modified treadmill; sequences from the Yang Short perturbations were delivered over 40 ms at Form; three 30-min sessions/wk 0.22 m/s for forward belt movement or for 4 wk between 0.22 and 1.07 m/s for backward belt movement; three 30-min sessions/wk for 4 wk Handelzalts et al49 (2019) Israel People with stroke and receiving Multidirectional (mediolateral or 36 voluntary weight-shifting None Yes rehabilitation anteroposterior) perturbations, delivered exercises using force platform in a random order using a treadmill biofeedback and 10–12 min of mounted on a movable platform; 36 unperturbed treadmill walking; perturbations to stance and 30 twelve 30-min sessions over 2.5 perturbations during treadmill wk walking/session; progressed to a higher intensity if participants were able to recover from 24 consecutive perturbations without a “fall”; for anteroposterior perturbations: distance = 10.44–29.06 cm, velocity = 0.18–0.54 m/s, and acceleration = 1.27–2.44 m/s/s; for mediolateral perturbations: displacement = 5.68–17.41 cm, velocity = 0.20–0.58 m/s, and acceleration = 0.73–1.94 m/s/s; twelve 30-min sessions over 2.5 wk (Continued) 7 Downloaded from https://academic.oup.com/ptj/article/103/1/pzac154/6812867 by guest on 02 January 2024 Table 1. Continued. 8 Fall Monitoring of Study Location Sample RBT Intervention Control Intervention Monitoring Adverse Duration Events Okubo et al34 (2019) Australia Physically active 30 slip and 30 trip perturbations in total, Target step training (stepping on None Yes community-dwelling older adults induced during overground walking on an target tiles) with a specific (65–90 y old) instrumented walkway; gait speed was cadence; 30 trials in one 40-min 40% to 90% of the participant’s usual session speed, the trip board height was 7 or 14 cm, and the slip distance was 20–40 cm; three 40-min sessions over 2 d Esmaeili et al48 (2020) Canada People with chronic unilateral 36 perturbations/session, delivered during Walking on a treadmill at a None Yesb stroke (>6 mo) walking via acceleration/deceleration of 1 comfortable speed (n = 8); nine belt on a split-belt treadmill; nine 35- to 35- to 70-min sessions over 3 wk, 70-min sessions over 3 wk, depending on depending on gait speed gait speed Lurie et al35 (2020) USA Older adults (≥65 y old) at risk of Anteroposterior perturbations to walking Individualized “usual care,” 12 mo No falling or stance, delivered using a treadmill, and including strength, flexibility, “usual care” (see description of control balance, and mobility training, intervention); the duration of RBT and the patient education, and home number of perturbations were at the exercise prescription; two or three discretion of the treating physical 45-min sessions/wk for 4–6 wk therapist; two or three 45-min sessions/wk (varied on the basis of participant for 4–6 wk (varied on the basis of need) participant need) Rieger et al33 (2020) The Healthy community-dwelling Treadmill walking with 16 anteroposterior 8 min of unperturbed treadmill None Yesb Netherlands older adults perturbations (8 belt accelerations and 8 walking at 1 m/s; 1 session belt decelerations at foot contact) at either low (8 m/s/s for 0.11 s) or high (10 m/s/s for 0.13 s) intensity; 1 session Batcir et al29 (2021) Israel Community-dwelling older adults Mediolateral tilt perturbations, delivered Riding a stationary bicycle at an None Yesb (≥70 y old) while riding a stationary bicycle without intensity to yield a heart rate upper extremity support; two 20-min between 60% and 80% of sessions/wk for 10 wk maximum; two 20-min sessions/wk for 10 wk Unger et al53 (2021) Canada Adults with an incomplete spinal Multidirectional manual perturbations, Conventional balance training, 6 mo Yes cord injury (C or D on the delivered during the same balance tasks as including static and dynamic American Spinal Association those completed by the control group balance tasks; three 1-h Impairment Scale), >1 y after (∼45 perturbations/session); three 1-h sessions/wk for 8 wk injury, and able to stand sessions/wk for 8 wk independently Rogers et al31 (2021) USA Apparently healthy older adults Two RBT groups completed 40 lateral Hip-strengthening exercises 12 mo Yesc (≥65 y old) perturbations/session using a waist-pull (control group 1) or flexibility system, with the force of the perturbation and relaxation exercises (control initially set 15% higher than participants’ group 2); 3 times/wk for 12 wk multistep threshold and increased as participants improved; 1 RBT group completed only RBT, and 1 completed RBT and hip strengthening exercises in alternate sessions; 3 times/wk for 12 wk Wang et al30 (2022) USA Apparently healthy older adults 40 slips while walking on a treadmill; 1 30 min of treadmill walking; 1 6 mo No (≥65 y old) session session Does Reactive Balance Training Prevent Falls? a RBT = reactive balance training. b Information not reported in the publication and obtained from the study authors. c Adverse events were reported, but the specific numbers of events were not reported separately for both groups, so this study could not be included in the analysis of adverse event data. Downloaded from https://academic.oup.com/ptj/article/103/1/pzac154/6812867 by guest on 02 January 2024 Devasahayam et al 9 Table 2. Falls After the Intervention in the Included Studiesa Follow-Up Reactive Balance Training Control Fall Risk Fall Rate Study Duration Ratio Ratio No. of No. of Sample No. of No. of Sample (mo) Fallers Falls Size Fallers Falls Size Shimada et al55 (2004) 6 5 8 15 6 11 11 0.61 0.53 Marigold et al50 (2005) 12 11 25 19 16 75 21 0.76 0.37 Protas et al44 (2005) 0.5 5 10 9 6 23 9 0.83 0.43 Maki et al39 (2008) 12 2 4 4 3 4 4 0.67 1.00 Mansfield et al9 (2010) 12 5 6 16 4 4 15 1.17 1.41 Smania et al43 (2010) 1 8b 36 28 19b 111 27 0.41 0.31 Halvarsson et al40 (2013) 12 18 NA 34 2 NA 21 5.56 NA Lurie et al32 (2013) 3 5 10b 26 11 32b 33 0.58 0.40 Pai et al37 (2014) 12 14 17 102 26 32 96 0.51 0.50 Downloaded from https://academic.oup.com/ptj/article/103/1/pzac154/6812867 by guest on 02 January 2024 Shen and Mak45 (2015) 12 6 10 22 13 18 23 0.48 0.58 Kurz et al36 (2016) 12 9b 12b 21 9b 18b 19 0.90 0.60 Mansfield et al8 (2018) 12 19 53 41 23 64 42 0.85 0.85 Aviles et al54 (2019) 6 NA 4 16 NA 5 16 NA 0.80 Lurie et al35 (2020) 12 128b 686b 218 145b 603b 212 0.86 1.11 Rogers et al31 (2021) 12 NA 10 31 NA 19 32 NA 0.54 Unger et al53 (2021) 6 4 10 10 7 21 10 0.57 0.48 Wang et al30 (2022) 6 18 37 70 15 28 63 1.08 1.19 Total 148.5 257 938 682 305 1068 654 a NA = data not available. b Data received from authors. Figure 2. Results of meta-analysis for risk of falls (A) and rate of falls (B). (A) Forest plot of studies with data on risk of falls. The analysis included data from 15 studies with a total of 635 reactive balance training (RBT) participants and 606 control participants. The P value for heterogeneity was.11. (B) Forest plot of studies with data on rate of falls. The analysis included data from 16 studies with a total of 648 RBT participants and 633 control participants. The P value for heterogeneity was 30%. d Studies with 1 or more of the following methodological concerns: no concealed allocation; no masking of participants, therapists, and assessors; possibility of selective reporting; and no intention-to-treat analysis.8,9,30–32,35–37,39,43–45,50,53–55 increased risk of falls (risk ratio: 0.60; rate ratio: 0.52). there are not enough studies to compare fall rates between Repeated task-specific motor experience, such as reactive studies with shorter versus longer falls-monitoring durations. stepping, is necessary to learn perceptual-motor abilities that Evidence suggests that people who complete RBT retain are essential for recovery from loss of balance.57 Because improvements in reactive balance control and the ability to avoiding a fall requires the whole body to react, exercise resist in-lab falls for up to 12 months after the end of train- interventions that intend to reduce the risk of falls must evoke ing.8,61 Future work should determine how long the effects task-specific total body balance reactions. A large Cochrane of RBT are retained, and how often retraining is required. review3 including 25,160 individuals aged 60 or more years Although the methodological and reporting quality of the from 116 studies reported that interventions with a total included studies was good on average, there was a high risk of weekly dose of 3 or more hours of task-specific balance and bias arising from the randomization process, missing outcome functional exercises were effective in reducing falls (rate ratio: data, and bias in measurement of the outcome. Not masking 0.58) compared with controls. These findings suggest that therapists who deliver interventions could introduce bias in task-specific tailored RBT programs with well-defined dosage the treatment effects.62 However, it is not possible to mask prescription should be developed for people with balance therapists in exercise studies where the therapist delivers the impairments. exercise intervention. One study attempted to mask therapists Unlike other types of exercise, dosage is not well defined to the aims of the study,43 but the therapists were still aware in balance prescription (Tab. 1). Specifically, the intensity of of the nature and specific details of the intervention. Further- balance training is often assessed using metrics that are not more, having different therapists provide the 2 interventions, appropriate for balance training, such as the participant’s rate as was the case for Smania et al,43 may introduce another of perceived exertion and/or the therapist’s perception of the source of variability between groups (ie, the therapists may participant’s ability (ie, safety).58 For example, Halvarsson differ in their treatment approaches). et al40 prescribed exercises at 5 different levels with each The prevalence of adverse events was higher during RBT reflecting progressive demands on the postural control system, compared with control interventions. The between-group dif- whereas Smania et al43 challenged participants during each ference in adverse events seemed to be greatest for subjective session with 10 different exercises before increasing the com- or psychological events, although these data were not sta- plexity of the tasks; however, both studies did not quantify tistically analyzed due to the low numbers of each type of the perturbation intensity. It is possible that the dosage of adverse event (Tab. 3). Seven RBT participants reported fear RBT was not sufficient in some studies included in our review or anxiety related to the perturbations, whereas no control to impact fall risk in daily life.59 Furthermore, the amount participants reported fear or anxiety related to the control of training has to reach an asymptotic level within 1 session intervention. This finding highlights the need to build trust in order to demonstrate performance gains across sessions.60 between the therapist and client to alleviate these anxieties.63 Therefore, there is a need to characterize the intensity of RBT Therapists report that, as clients gradually overcome these (such as perturbation force) and determine dose–response anxieties with training, RBT can help to build self-efficacy and effect of RBT in future studies at different stages of recovery a sense of achievement.63 from balance impairments. Findings regarding adverse events should be interpreted There was variability in the fall monitoring duration for with caution. Fifteen studies that met most of our inclusion studies. Whereas most included studies monitored falls for at criteria were not included in the review because they did not least 12 months after the intervention, 4 monitored falls for monitor adverse events, and 7 trials that were included in 6 months, and 3 for less than 6 months. Two of the studies the current review did not monitor adverse events. Therefore, with short falls-monitoring durations included people with approximately half of RBT trials do not monitor or report Parkinson disease, and a large number of falls (>2.5 falls adverse events. This is consistent with 2 other reviews, where per person per month) were reported in the short time period only 43% and 44% of exercise interventions reported adverse when falls were monitored.43,44 It is possible that the effect of events.64,65 Even among those trials that did report adverse RBT on risk of falls in daily life declines over time. However, events, it is possible that not all events were reported; some 12 Does Reactive Balance Training Prevent Falls? study authors only report certain types of adverse events or on maintaining balance, and also experienced external per- only report adverse events above a certain threshold (eg, only turbations, which were provided by the instructor or partner. severe adverse events, or only those events reported by a The authors did not report the prescribed intensity of the certain proportion of participants).66–69 Seven RBT studies perturbations.40 This could have been a major factor (clinical that monitored adverse events did not report any; given the heterogeneity) influencing the effect of RBT on the risk of known adverse events associated with exercise, it is likely falls in daily life.75 Because of the above-mentioned factor that reports of low or zero adverse events in exercise studies and because none of the tests available for publication bias are either due to selective nonreporting or underreporting of has been compared against a gold standard,76 we were not adverse events. All harms need to be reported in clinical trials able to assess publication bias in our study. Therefore, our so that clinicians and patients are aware of both the risks and confidence in the estimate of the reduction in the risk and benefits of treatments,70 and reporting of harms is one of the rate of falls with RBT is low to support a recommendation. required items in the CONSORT checklist.71 If the risks of Further research that addresses the limitations highlighted in an intervention are unknown, therapists may assume harm this study is likely to change the estimate. and avoid prescribing RBT.63 Alternatively, clinicians may We were unable to confirm the prevalence of adverse events Downloaded from https://academic.oup.com/ptj/article/103/1/pzac154/6812867 by guest on 02 January 2024 prescribe RBT inappropriately assuming there are no risks for 8 studies (Tab. 1). From our correspondence with study if adverse events are underreported. Although most adverse authors, we noted that several studies did not report mild events reported with RBT were minor and commonly reported and/or expected adverse events associated with exercise in the with other exercise interventions (eg, muscle soreness, joint publication. It is possible that most of the minor intervention- pain), the occurrence of the adverse event often led to modi- related adverse events, such as muscle or joint pain, were fying the intervention until the issue resolved (eg, skipping a reported in the following session (ie, the participants are session or reducing the training intensity). Additionally, pain asked “How did you feel after the last session?”); therefore, was often reported as worsening of preexisting joint pain (eg, there is a lack of certainty about adverse events for studies for participants with osteoarthritis or prior joint injury); it is with only 1 session or without longer term intervention. important for therapists to know that certain patients may Such underreporting of adverse events could have biased our be at increased risk of adverse events. Because occurrence of results. adverse events may lead to modifying the intervention, it is important for therapists to be aware of these possible adverse events. Conclusions Control interventions may have been less “intense” than Older adults and individuals with balance impairments had RBT. For example, Mansfield et al9 and Marigold et al50 a lower likelihood of falls in daily life after participating in used low-intensity activities like stretching or relaxation for RBT compared with control interventions. The certainty of the control groups in their studies. In the study of Okubo evidence was “low” to “moderate” due to a lack of concealed et al,56 RBT participants completed 120 minutes of training allocation, a lack of masking procedures, a possibility of on a trip and slip walkway, with 30 trips and slips in 3 selective outcome reporting, and incomplete outcome data sessions, whereas control participants completed 40 minutes due to attrition, indicating the need for studies with a lower of walking on the same path with no perturbations in 1 risk of bias. Participants of RBT reported more adverse events session. Likewise, Shimada et al55 and Lurie et al35 assigned than control groups; however, the prevalence of adverse events “usual care” to the control groups while conducting RBT in for both RBT and control groups is likely underestimated addition to “usual care” for the experimental groups. Because because only about half of RBT studies monitored adverse the control interventions were often less challenging than RBT, events. with either lower dose of therapy or no intervention, the effect size estimate for RBT on fall risk could be inflated. Furthermore, the lower intensity or duration of control inter- Author Contributions ventions compared with RBT could, in part, explain the Concept/idea/research design: A.J. Devasahayam, A. Mansfield increased prevalence of adverse events during RBT compared Writing: A.J. Devasahayam, D. Jagroop, A. Mansfield with control interventions. Future RBT studies should take Data collection: A.J. Devasahayam, K. Farwell, B. Lim, A. Morton, steps to include clinically meaningful control interventions N. Fleming, D. Jagroop, R. Aryan, T.M. Saumur, A. Mansfield that are comparable to the experimental intervention while Data analysis: A.J. Devasahayam investigating the effect of RBT on fall risk.72 In addition, it is Project management: A.J. Devasahayam, A. Mansfield necessary to follow best practices for adverse event reporting Fund procurement: A. Mansfield Clerical/secretarial support: D. Jagroop to learn about both the risks and benefits of RBT for people Consultation (including review of manuscript before submitting): with balance impairments.73 A.J. Devasahayam, K. Farwell, B. Lim, A. Morton, N. Fleming, D. Jagroop, R. Aryan, T.M. Saumur, A. Mansfield Limitations We noted asymmetry in the funnel plots generated to deter- mine the presence of biases in meta-analysis of fall risk and Data Availability Statement rate. However, we were not able to confirm whether the Data are made available on reasonable request from the corresponding asymmetry in the funnel plots was due to publication bias. author at the KITE-Toronto Rehabilitation Institute, University Health Funnel plots determine whether studies with lower precision Network, Canada. (eg, higher SE) differ from studies with higher precision (lower SE).74 The SE was the highest in the study by Halvarsson Acknowledgments et al40 compared with the 12 other studies included in the The authors thank Jessica Babineau, BA, MLIS, information specialist meta-analysis of risk of falls (Fig. 2A). In this study,40 par- at the Toronto Rehabilitation Institute – University Health Network, ticipants performed “typical” balance exercises with a focus for developing the search strategy and conducting the literature search. Devasahayam et al 13 Funding 15. Maher CG, Sherrington C, Herbert RD, Moseley AM, Elkins M. Reliability of the PEDro scale for rating quality of randomized A. Mansfield was supported by a New Investigator Award from the controlled trials. Phys Ther. 2003;83:713–721. Canadian Institutes of Health Research (MSH-141983). 16. Maher CG. A systematic review of workplace interventions to prevent low back pain. Aust J Physiother. 2000;46:259–269. 17. Higgins JP, Altman DG, Gøtzsche PC et al. The Cochrane Systematic Review Registration Collaboration’s tool for assessing risk of bias in randomised trials. This study is an update of a previous systematic review and meta- BMJ. 2011;343:d5928. analysis, registered with PROSPERO database (CRD42020220552). 18. Sterne JAC, Savović J, Page MJ et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ. 2019;366:l4898. 19. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327:557–560. Disclosures and Presentations 20. Sterne JA, Sutton AJ, Ioannidis JP et al. Recommendations for examining and interpreting funnel plot asymmetry in meta- The authors completed the ICMJE Form for Disclosure of Potential analyses of randomised controlled trials. BMJ. 2011;343:d4002. Downloaded from https://academic.oup.com/ptj/article/103/1/pzac154/6812867 by guest on 02 January 2024 Conflicts of Interest and reported no conflicts of interest. 21. Egger M, Smith GD, Schneider M, Minder C. Bias in meta-analysis The authors have created a reactive balance training toolkit, which detected by a simple, graphical test. BMJ. 1997;315:629–634. is partially monetized. 22. Guyatt G, Oxman AD, Akl EA et al. GRADE guidelines: 1. The first version of this manuscript was submitted to the preprint Introduction—GRADE evidence profiles and summary of findings server for health sciences (medRxiv) on February 10, 2022 (https://doi.o tables. J Clin Epidemiol. 2011;64:383–394. rg/10.1101/2022.01.27.22269969). 23. Schünemann HJ, Cuello C, Akl EA et al. GRADE guidelines: 18. How ROBINS-I and other tools to assess risk of bias in nonrandomized studies should be used to rate the certainty of a References body of evidence. J Clin Epidemiol. 2019;111:105–114. 1. Rubenstein LZ. Falls in older people: epidemiology, risk factors and 24. Zhang Y, Alonso-Coello P, Guyatt GH et al. GRADE guidelines: strategies for prevention. Age Ageing. 2006;35:ii37–ii41. 19. Assessing the certainty of evidence in the importance of out- 2. Burns ER, Stevens JA, Lee R. The direct costs of fatal and non- comes or values and preferences—risk of bias and indirectness. fatal falls among older adults—United States. J Saf Res. 2016;58: J Clin Epidemiol. 2019;111:94–104. 99–103. 25. Zhang Y, Coello PA, Guyatt GH et al. GRADE guidelines: 20. 3. Sherrington C, Fairhall N, Kwok W et al. Evidence on physical Assessing the certainty of evidence in the importance of outcomes activity and falls prevention for people aged 65+ years: systematic or values and preferences—inconsistency, imprecision, and other review to inform the WHO guidelines on physical activity and domains. J Clin Epidemiol. 2019;111:83–93. sedentary behaviour. Int J Behav Nutr Phys Act. 2020;17:144. 26. Schünemann H, Brożek J, Guyatt G, Oxman A. Handbook for 4. Berg K, Wood-Dauphine S, Williams J, Gayton D. Measuring Grading the Quality of Evidence and the Strength of Recommenda- balance in the elderly: preliminary development of an instrument. tions Using the GRADE Approach. The GRADE Working Group. Physiother Can. 1989;41:304–311. 2013. Accessed December 28, 2022. https://gdt.gradepro.org/app/ 5. Maki BE, McIlroy WE. Postural control in the older adult. Clin handbook/handbook.html. Geriatr Med. 1996;12:635–658. 27. Page MJ, Moher D, Bossuyt PM et al. PRISMA 2020 explanation 6. Okubo Y, Schoene D, Caetano MJD et al. Stepping impairment and elaboration: updated guidance and exemplars for reporting and falls in older adults: a systematic review and meta-analysis of systematic reviews. BMJ. 2021;372:n160. volitional and reactive step tests. Ageing Res Rev. 2021;66:101238. 28. Rosenblatt NJ, Marone J, Grabiner MD. Preventing trip related 7. Mansfield A, Wong JS, Bryce J, Knorr S, Patterson KK. Does falls by community-dwelling adults: a prospective study. J Am perturbation-based balance training prevent falls? Systematic Geriatr Soc. 2013;61:1629–1631. review and meta-analysis of preliminary randomized controlled 29. Batcir S, Livne Y, Lev Lehman R et al. Development and piloting trials. Phys Ther. 2015;95:700–709. of a perturbation stationary bicycle robotic system that provides 8. Mansfield A, Aqui A, Danells CJ et al. Does perturbation- unexpected lateral perturbations during bicycling (the PerStBiRo based balance training prevent falls among individuals with system). BMC Geriatr. 2021;21:71. chronic stroke? A randomised controlled trial. BMJ Open. 2018; 30. Wang Y, Wang S, Liu X, Lee A, Pai YC, Bhatt T. Can a single session 8:e021510. of treadmill-based slip training reduce daily life falls in community- 9. Mansfield A, Peters AL, Liu BA, Maki BE. Effect of a perturbation- dwelling older adults? A randomized controlled trial. Aging Clin based balance training program on compensatory stepping and Exp Res. 2022;34:1593–1602. grasping reactions in older adults: a randomized controlled trial. 31. Rogers MW, Creath RA, Gray V et al. Comparison of lateral Phys Ther. 2010;90:476–491. perturbation-induced step training and hip muscle strengthening 10. Wang Y, Wang S, Bolton R, Kaur T, Bhatt T. Effects of task-specific exercise on balance and falls in community-dwelling older adults: obstacle-induced trip-perturbation training: proactive and reactive a blinded randomized controlled trial. J Gerontol A Biol Sci Med adaptation to reduce fall-risk in community-dwelling older adults. Sci. 2021;76:e194–e202. Aging Clin Exp Res. 2020;32:893–905. 32. Lurie JD, Zagaria AB, Pidgeon DM, Forman JL, Spratt KF. Pilot 11. Gutenbrunner C, Nugraha B. Decision-making in evidence-based comparative effectiveness study of surface perturbation treadmill practice in rehabilitation medicine: proposing a fourth factor. Am training to prevent falls in older adults. BMC Geriatr. 2013;13:49. J Phys Med Rehabil. 2020;99:436–440. 33. Rieger M, Papegaaij S, Pijnappels M, Steenbrink F, van Dieën 12. Booth A, Clarke M, Ghersi D, Moher D, Petticrew M, Stewart L. J. Transfer and retention effects of gait training with anterior- An international registry of systematic-review protocols. Lancet. posterior perturbations to postural responses after medio-lateral 2011;377:108–109. gait perturbations in older adults. Clin Biomech. 2020;75:104988. 13. Higgins JP, Thomas J, Chandler J et al. Cochrane Handbook for 34. Okubo Y, Sturnieks DL, Brodie MA, Duran L, Lord SR. Effect of Systematic Reviews of Interventions. Hoboken, NJ: John Wiley & reactive balance training involving repeated slips and trips on bal- Sons; 2019. ance recovery among older adults: a blinded randomized controlled 14. Liberati A, Altman DG, Tetzlaff J et al. The PRISMA statement trial. J Gerontol A Biol Sci Med Sci. 2019;74:1489–1496. for reporting systematic reviews and meta-analyses of studies that 35. Lurie JD, Zagaria AB, Ellis L et al. Surface perturbation training evaluate health care interventions: explanation and elaboration. to prevent falls in older adults: a highly pragmatic, randomized J Clin Epidemiol. 2009;62:e1–e34. controlled trial. Phys Ther. 2020;100:1153–1162. 14 Does Reactive Balance Training Prevent Falls? 36. Kurz I, Gimmon Y, Shapiro A, Debi R, Snir Y, Melzer I. Unexpected cord injury or disease: a randomized clinical trial. Front Neurol. perturbations training improves balance control and voluntary 2021;12:49. stepping times in older adults—a double blind randomized control 54. Aviles J, Allin LJ, Alexander NB, Van Mullekom J, Nussbaum MA, trial. BMC Geriatr. 2016;16:58. Madigan ML. Comparison of treadmill trip-like training versus tai 37. Pai Y-C, Bhatt T, Yang F, Wang E, Kritchevsky S. Perturbation chi to improve reactive balance among independent older adult training can reduce community-dwelling older adults’ annual fall residents of senior housing: a pilot controlled trial. J Gerontol A risk: a randomized controlled trial. J Gerontol A Biol Sci Med Sci. Biol Sci Med Sci. 2019;74:1497–1503. 2014;69:1586–1594. 55. Shimada H, Obuchi S, Furuna T, Suzuki T. New intervention 38. Parijat P, Lockhart TE. Effects of moveable platform training in program for preventing falls among frail elderly people: the effects preventing slip-induced falls in older adults. Ann Biomed Eng. of perturbed walking exercise using a bilateral separated treadmill. 2012;40:1111–1121. Am J Phys Med Rehabil. 2004;83:493–499. 39. Maki BE, Cheng KC-C, Mansfield A et al. Preventing falls in 56. Okubo Y, Schoene D, Lord SR. Step training improves reaction older adults: new interventions to promote more effective change- time, gait and balance and reduces falls in older people: a system- in-support balance reactions. J Electromyogr Kinesiol. 2008;18: atic review and meta-analysis. Br J Sports Med. 2017;51:586–593. 243–254. 57. Henry FM. Specificity vs generality in learning motor skills. Proc Downloaded from https://academic.oup.com/ptj/article/103/1/pzac154/6812867 by guest on 02 January 2024 40. Halvarsson A, Franzén E, Farén E, Olsson E, Oddsson L, Ståhle Coll Phys Educ Assoc. 1958;61:126–128. A. Long-term effects of new progressive group balance training 58. Farlie MK, Robins L, Keating JL, Molloy E, Haines TP. Intensity of for elderly people with increased risk of falling—a randomized challenge to the balance system is not reported in the prescription controlled trial. Clin Rehabil. 2013;27:450–458. of balance exercises in randomised trials: a systematic review. J 41. Toole T, Hirsch M, Forkink A, Lehman D, Maitland C. The effects Physiother. 2013;59:227–235. of a balance and strength training program on equilibrium in 59. Halvarsson A, Oddsson L, Olsson E, Farén E, Pettersson A, Ståhle parkinsonism: a preliminary study. NeuroRehabilitation. 2000;14: A. Effects of new, individually adjusted, progressive balance group 165–174. training for elderly people with fear of falling and tend to fall: a 42. Schlenstedt C, Paschen S, Kruse A, Raethjen J, Weisser B, Deuschl randomized controlled trial. Clin Rehabil. 2011;25:1021–1031. G. Resistance versus balance training to improve postural control 60. Hauptmann B, Reinhart E, Brandt SA, Karni A. The predictive in Parkinson’s disease: a randomized rater blinded controlled study. value of the leveling off of within session performance for proce- PLoS One. 2015;10:e0140584. dural memory consolidation. Cogn Brain Res. 2005;24:181–189. 43. Smania N, Corato E, Tinazzi M et al. Effect of balance training on 61. Pai Y-C, Yang F, Bhatt T, Wang E. Learning from laboratory- postural instability in patients with idiopathic Parkinson’s disease. induced falling: long-term motor retention among older adults. Age Neurorehabil Neural Repair. 2010;24:826–834. (Dordr). 2014;36:9640. 44. Protas EJ, Mitchell K, Williams A, Qureshy H, Caroline K, Lai 62. Armijo-Olivo S, Fuentes J, da Costa BR, Saltaji H, Ha C, Cummings EC. Gait and step training to reduce falls in Parkinson’s disease. GG. Blinding in physical therapy trials and its association with NeuroRehabilitation. 2005;20:183–190. treatment effects: a meta-epidemiological study. Am J Phys Med 45. Shen X, Mak MK. Technology-assisted balance and gait training Rehabil. 2017;96:34–44. reduces falls in patients with Parkinson’s disease: a randomized 63. Jagroop D, Houvardas S, Danells CJ, et al. Rehabilitation controlled trial with 12-month follow-up. Neurorehabil Neural clinicians’ perspectives of reactive balance training [published Repair. 2015;29:103–111. online November 23, 2021].s Disabil Rehabil. https://doi.o 46. Steib S, Klamroth S, Gaßner H et al. Perturbation during treadmill rg/10.1080/09638288.2021.2004246. training improves dynamic balance and gait in Parkinson’s disease: 64. James KA, von Heideken J, Iversen MD. Reporting of adverse a single-blind randomized controlled pilot trial. Neurorehabil Neu- events in randomized controlled trials of therapeutic exercise ral Repair. 2017;31:758–768. for hip osteoarthritis: a systematic review. Phys Ther. 2021; 47. Gandolfi M, Geroin C, Dimitrova E et al. Virtual reality tel- 101:pzab195. erehabilitation for postural instability in Parkinson’s disease: a 65. von Heideken J, Chowdhry S, Borg J, James K, Iversen MD. multicenter, single-blind, randomized, controlled trial. Bio Med Res Reporting of harm in randomized controlled trials of therapeutic Int. 2017;2017:7962826. exercise for knee osteoarthritis: a systematic review. Phys Ther. 48. Esmaeili V, Juneau A, Dyer J-O et al. Intense and unpredictable 2021;101:pzab161. perturbations during gait training improve dynamic balance abil- 66. Liu C-J, Latham N. Adverse events reported in progressive resis- ities in chronic hemiparetic individuals: a randomized controlled tance strength training trials in older adults: 2 sides of a coin. Arch pilot trial. J Neuroeng Rehabil. 2020;17:79. Phys Med Rehabil. 2010;91:1471–1473. 49. Handelzalts S, Kenner-Furman M, Gray G, Soroker N, Shani G, 67. Bernhardt J, Langhorne P, Lindley RI et al. Efficacy and safety of Melzer I. Effects of perturbation-based balance training in subacute very early mobilisation within 24 h of stroke onset (AVERT): a persons with stroke: a randomized controlled trial. Neurorehabil randomised controlled trial. Lancet. 2015;386:46–55. Neural Repair. 2019;33:213–224. 68. Patterson KK, Wong JS, Prout EC, Brooks D. Dance for the 50. Marigold DS, Eng JJ, Dawson AS, Inglis JT, Harris JE, Gylfadottir rehabilitation of balance and gait in adults with neurological S. Exercise leads to faster postural reflexes, improved balance and conditions other than Parkinson’s disease: a systematic review. mobility, and fewer falls in older persons with chronic stroke. J Am Heliyon. 2018;4:e00584. Geriatr Soc. 2005;53:416–423. 69. Sivendran S, Latif A, McBride RB et al. Adverse event reporting in 51. Kumar C, Pathan N. Effectiveness of manual perturbation exercises cancer clinical trial publications. J Clin Oncol. 2014;32:83–89. in improving balance, function and mobility in stroke patients: a 70. Jull A, Biggs R. Adverse event reporting and trial registration randomized controlled trial. J Nov Physiother. 2016;6:284. in venous leg ulcer trials published since the 2001 CONSORT 52. Morgan P, Murphy A, Opheim A, Pogrebnoy D, Kravtsov S, statement revision: a systematic review. J Tissue Viability. 2020;29: McGinley J. The safety and feasibility of an intervention to 155–160. improve balance dysfunction in ambulant adults with cerebral 71. Ioannidis JP, Evans SJ, Gøtzsche PC et al. Better reporting of harms palsy: a pilot randomized controlled trial. Clin Rehabil. 2015;29: in randomized trials: an extension of the CONSORT statement. 907–919. Ann Intern Med. 2004;141:781–788. 53. Unger J, Chan K, Lee JW et al. The effect of perturbation-based 72. Dobkin BH. Confounders in rehabilitation trials of task-oriented balance training and conventional intensive balance training on training: lessons from the designs of the EXCITE and SCILT reactive stepping ability in individuals with incomplete spinal multicenter trials. Neurorehabil Neural Repair. 2007;21:3–13. Devasahayam et al 15 73. World Health Organization. World Alliance for Patient Safety: 75. Gagnier JJ, Morgenstern H, Altman DG et al. Consensus-based rec- WHO Draft Guidelines for Adverse Event Reporting and Learning ommendations for investigating clinical heterogeneity in systematic Systems: From Information to Action. Geneva: World Health reviews. BMC Med Res Methodol. 2013;13:106. Organization; 2005. 76. Rothstein HR, Sutton AJ, Borenstein M. Publication Bias in Meta- 74. Lau J, Ioannidis JPA, Terrin N, Schmid CH, Olkin I. The case of Analysis: Prevention, Assessment and Adjustments. Hoboken, NJ: the misleading funnel plot. BMJ. 2006;333:597–600. John Wiley & Sons; 2006. Downloaded from https://academic.oup.com/ptj/article/103/1/pzac154/6812867 by guest on 02 January 2024

The Effect of Reactive Balance Training on Falls in Daily Life PDF

Document Details

Tags

Related

Summary

Full Transcript