Background
NB: For an explanation of some of the terms used in this review, please see the Glossary of Terms ( Table 1).
Balance is defined as the ability to maintain the projection of the body's centre of mass (CoM) within manageable limits of the base of support, as in standing or sitting, or in transit to a new base of support, as in walking (Winter 1995). The base of support is composed of the area between all points of contact of the body with another surface; points of contact also include extensions of the body through assistive devices e.g. walking sticks and frames. Balance is an integral component of daily (functional) activities, however, balance control is very complex and multifactorial. The task being undertaken and the environment in which it is taking place both affect an individual's ability to control balance, by altering the biomechanical and information processing needs (Huxham 2001). Balance may be measured when the body has a constant, or static, base of support, or during movement from one base of support to another. It can be analysed directly by quantifying the position of the CoM in relation to the base of support. Alternatively, balance can be measured indirectly through observation, self reporting or other reporting methods such as objective tests of functional activities.
However, the ability to undertake functional activities is complex and multifaceted involving not only balance but other factors such as strength, proprioception, integrity of the neuromuscular system, pain, vision and in some instances fear of falling.
Physiological changes related to ageing include, for example, cognitive impairment (Nevitt 1989), reductions in muscle strength (Daubney 1999; Doherty 1993), proprioception (Skinner 1984), joint range of motion (Mills 1994) and reaction time (Stelmach 1994), and changes in sensory systems (Berg 1989). These factors potentially negatively affect balance control and impact on the functional ability of the older person.
Diminished ability to maintain balance may be associated with an increased risk of falling (Berg 1989). In older adults, falls commonly lead to injury, loss of independence, associated illness and early death (Baker 1985; Berg 1989; Tinetti 1988). An extensive review of published trials of interventions to reduce the incidence of falling in elderly people has been published (Gillespie 2004). Although some exercise interventions with balance and muscle strengthening components have been shown to reduce falls (Campbell 1997; Robertson 2001; Wolf 1996), it is not clear which element or combination of elements is necessary to achieve this result.
Biofeedback and visual feedback have been used to improve balance control by addressing internal factors that are thought to contribute towards balance (Geiger 2001; Walker 2000). However, such interventions have tended to focus on single components of balance control; a multifactorial approach may be more appropriate.
Although there have been traditional literature reviews describing studies designed to improve balance in the elderly (Chandler 1996) there is still uncertainty surrounding the efficacy of exercise interventions, the effectiveness of the dosage (frequency, duration or intensity of delivery), setting in which the intervention takes place, level and type of supervision, or indeed who is most likely to benefit.
Objectives
To present the best evidence for effectiveness of exercise interventions designed to improve balance in older people living in the community or in institutional care.
The following hypothesis was tested:
- Exercise interventions designed to improve balance are as effective as usual care, attention or recreational activities.
Methods
Criteria for considering studies for this review
Types of studies
We included randomised controlled trials and quasi-randomised trials (e.g. randomised by date of birth or hospital record number) testing exercise interventions designed to improve balance in older people against a control group. Control group activity included either usual activities or attention (receiving the same attention or contact with the investigators such as attending appointments) or recreational activities.
Types of participants
We included trials with participants described as older adults, elderly, geriatric, aged, seniors or all over the age of 60, and studies that separately randomised and analysed the group described above. The participants could have included frail older people, or healthy older people, of either gender, living in the community or in institutional care. Participant characteristics of interest included age, gender, functional status at entry and residential status.
In order not to broaden the scope of this review too widely, we excluded trials of interventions targeting individuals with specific conditions such as; stroke, Parkinson's disease, multiple sclerosis, labyrinthitis, Meniere's disease, amputation of upper or lower limbs, cognitive impairments, dementia, osteoporosis, rheumatoid arthritis, osteoarthritis, hip fracture or Alzheimer's disease.
Types of interventions
Trials were included where participants were randomised to receive the following: a single exercise intervention or a multiple exercise intervention and a control group (usual activities, receiving the same attention as the intervention group or recreational activities). Trials comparing two or more exercise interventions and a control group were also included.
Exercise interventions designed to improve balance were defined as those in which participants exercise their muscles (and neuromuscular responses) against an external force, as a consequence of voluntary movement, or in response to an unexpected perturbation/stimulus in order to maintain the body's centre of mass within manageable limits of the base of support or in transit to a new base of support. Examples of exercise interventions include: walking, cycling, balance training, and tai chi.
The exercise interventions could take place in the home, institutional dwelling, community, gymnasium or clinic setting and could be self-supervised (for example using exercise sheets/video), individually supervised or as part of a supervised group, the supervisor could include for example, self, peer, physical trainer or healthcare professional.
Types of outcome measures
The main outcome of interest was balance, defined as the ability to maintain the body's centre of mass within manageable limits of the base of support, as in maintaining a standing or sitting position, or in transit to a new base of support, as in walking or moving. Outcome measures were classified according to the dimensions of the ICF (International Classification of Functioning, Disability and Health) (WHO 2001): impairment, activity limitation or participation restriction.
To be included, studies must have reported primary outcome measures that are direct or indirect measures of balance performance.
- Direct measures (ICF dimension impairment) include force platform indicators (centre of pressure behaviour or position) (Winter 1995).
- Indirect measures of balance (ICF dimension activity limitation) based on quantification of functional abilities included, but were not restricted to: Functional Reach Test (Duncan 1990), temporal parameters of gait, one legged standing time, timed up and go test (Podsiadlo 1991).
- Indirect measures of balance based on observation were restricted to the Berg Balance Scale (Berg 1992).
We excluded timed walking tests such as distance walked in 3, 6 or 12 minutes, as these are indicators of aerobic capacity rather than balance ability. Trials that focused on fall rates, numbers of fallers, or other surrogate measures of balance, for example muscle strength or global functional ability, and did not report balance as a primary outcome, were excluded; these have been reviewed elsewhere (Gillespie 2004; Latham 2004).
Information was sought on the level of adherence or compliance with the exercise intervention, the magnitude and duration of effect, and adverse events associated with the exercise intervention.
Search methods for identification of studies
We searched the Cochrane Bone, Joint and Muscle Trauma Group Specialised Register of trials (Feb 2006), the Cochrane Central Register of Controlled Trials (The Cochrane Library 2006, Issue 1), MEDLINE (1966 to 28th Feb 2006), EMBASE (from 1980 to 28th Feb 2006), PEDro - The Physiotherapy Evidence Database (http://www.pedro.fhs.usyd.edu.au/ accessed 28th Feb 2006), OTseeker - The Occupational Therapy Systematic Evaluation of Evidence Database (http://www.otseeker.com accessed 28th Feb 2006), CINAHL (from 1982 to 28th Feb 2006), AMED (from 1985 to 24th March 2006), and reference lists of articles. No language restrictions were applied. Further trials were identified by contact with institutions and experts in the field.
In MEDLINE (OVID ONLINE) the first two phases of the optimal trial search strategy (Robinson 2002) were combined with one subject specific search (Appendix 1, column 1), and the less precise third phase of the optimal trial search strategy was combined with a more precise subject specific search (Appendix 1, column 2). Search strategies are also shown for The Cochrane Library (Appendix 2), CINAHL (Appendix 3), EMBASE (Appendix 4), AMED (Appendix 5), PEDro (Appendix 6), and OTseeker (Appendix 7).
Data collection and analysis
Selecting trials for inclusion
The search strategy identified 1297 titles for potential inclusion. From the title, abstract, and descriptors, pairs of members of the review team independently reviewed the results of the literature searches to identify potentially relevant trials for full review. From the full text of 158 papers that appeared to meet the selection criteria, 34 trials were selected for inclusion. Disagreement was resolved by consensus or third party adjudication.
Data collection
Three pairs of members of the review team used a customised data extraction tool, tested prior to use, to independently extract data. Disagreement was resolved by consensus or third party adjudication. We contacted authors of studies where there was inadequate reporting of data to enable clarification and where appropriate to allow pooling.
Assessment of methodological quality
Methodological quality was independently assessed for each study by three pairs of members of the review team using a modification of the Cochrane Bone Joint & Muscle Trauma Group's quality assessment tool (Madhok 2006). The final scoring scheme for 15 aspects of trial quality included items from the Cochrane Bone Joint & Muscle Trauma Group's quality assessment tool (items denoted by 'M'), items from the Delphi list (Verhagen 1998) (items denoted by 'D') and items from the Maastricht-Amsterdam consensus list for Methodological Quality Assessment (Bellamy 1997) (items denoted by 'MAC') (see Table 2). Any disagreement was resolved by consensus or third party adjudication.
Data analysis
Where available and appropriate, quantitative data for the outcomes listed in the inclusion criteria are presented in the Analyses (01.01 to 07.16). Where studies reported standard errors of the means (SEMs), standard deviations were obtained by multiplying standard errors of means by the square-root of the sample size. For each trial, relative risk and 95% confidence intervals were calculated for dichotomous outcomes, and weighted mean differences (WMD) and 95% confidence intervals calculated for continuous outcomes (reporting mean and standard deviation or standard error of the mean). Standardised mean differences (SMD) and 95% confidence intervals were calculated when combining results from studies using different ways of measuring the same concept. Change scores have been reported separately as these cannot be incorporated into meta analyses of standardised mean differences.
Where appropriate, results of comparable groups of trials were pooled using the fixed effect model and 95% confidence intervals calculated. Heterogeneity between comparable trials was tested using a standard chi-squared test and considered statistically significant at P < 0.1 after due consideration of the value of I squared. In the presence of heterogeneity the results of comparable groups of trials were pooled using the random effect model and 95% confidence intervals calculated.
Sensitivity and sub-group analysis
It was anticipated that sensitivity analyses would be undertaken, when indicated, to investigate the effects of methodological quality, for example, allocation concealment and intention-to-treat analysis. Where cluster randomised trials were combined with each other or with other studies in a meta-analysis, sensitivity analyses were performed to investigate the effect clustering had on the results.
Where the data allowed, we also planned separate outcome analyses to test the following hypotheses:
- exercise interventions are equally effective in males and females;
- exercise interventions are equally effective in young old (mean age 60-75 years) and older old (mean age over 75 years);
- exercise interventions are equally effective in frail and less frail;
- effectiveness is not dependent on the duration and (or) intensity of exercise interventions;
- effectiveness is not dependent on the setting in which the exercise intervention is delivered;
- effectiveness is not dependent on the level or type of supervision of the exercise intervention.
Results
Description of studies
See: Characteristics of included studies; Characteristics of excluded studies.
In all 158 full papers were considered for inclusion in this review. Of these 83 were excluded for reasons given in the 'Characteristics of excluded studies' table; 20 in the 'Studies awaiting assessment' section; and 16 in the 'Classification pending' section (NB a number of papers related to the same study). There were 39 papers included in this review which described 34 studies.
For the 34 included trials there were 2883 participants at entry. One study was published only as an abstract (McGarry 2001). There was great variation across the trials in the characteristics of participants, design and content of the exercise interventions, and the outcomes assessed. A brief summary is provided here and more detailed information is provided in the 'Characteristics of included studies' table. Trials took place in North America (n = 18), Europe (n = 7), Asia (n = 6) and Australasia (n = 3).
Participants
The participants in 27 trials were defined as healthy older people (Boshuizen 2005; Brouwer 2003; Buchner 1997a; Buchner 1997b; Cress 1999; Crilly 1989; Islam 2004; Jessup 2003; Johansson 1991; Lichtenstein 1989; Lord 1995; Lord 2003; Lord 2005; MacRae 1994; McGarry 2001; McMurdo 1993; Nelson 2004; Okumiya 1996; Paillard 2004; Ramsbottom 2004; Reinsch 1992; Rooks 1997a; Schoenfelder 2004; Shigematsu 2002; Suzuki 2004; Wolf 1997; Wolfson 1996) and participants in the remaining seven trials had general frailty and/or functional limitations (Krebs 1998; Rubenstein 2000; Sauvage 1992; Shimada 2004; Sihvonen 2004; Wolf 2001; Zhang 2006). Participants were residing in institutions (hospital or residential care facilities) in five trials (Crilly 1989; McMurdo 1993; Sauvage 1992; Schoenfelder 2004; Sihvonen 2004); the community in 27 trials (Boshuizen 2005; Brouwer 2003; Buchner 1997a; Buchner 1997b; Cress 1999; Islam 2004; Jessup 2003; Johansson 1991; Krebs 1998; Lichtenstein 1989; Lord 1995; Lord 2003; Lord 2005; MacRae 1994; McGarry 2001; Nelson 2004; Okumiya 1996; Paillard 2004; Ramsbottom 2004; Reinsch 1992; Rooks 1997a; Rubenstein 2000; Shigematsu 2002; Suzuki 2004; Wolf 1997; Wolfson 1996; Zhang 2006); and types of residence was mixed in two trials (Shigematsu 2002; Wolf 2001).
Nine trials included only women (Crilly 1989; Jessup 2003; Johansson 1991; Lichtenstein 1989; Lord 1995; MacRae 1994; Shigematsu 2002; Sihvonen 2004; Suzuki 2004) and four trials only men (Nelson 2004; Okumiya 1996; Rubenstein 2000; Sauvage 1992). The other 21 trials included both men and women in varying proportions; in the majority of trials, the proportion of women was typically greater.
The average age of participants was 60-75 years in 14 trials (Jessup 2003; Johansson 1991; Krebs 1998; Lord 1995; MacRae 1994; McGarry 2001; Nelson 2004; Paillard 2004; Ramsbottom 2004; Reinsch 1992; Rooks 1997a; Rubenstein 2000; Sauvage 1992; Zhang 2006) and over 75 years in the remaining 20 trials.
Exercise interventions
All the exercise interventions described were land-based. We categorised exercise interventions of included studies based on the taxonomy of exercise interventions developed by ProFaNE (Lamb 2006). (NB numbers of participants indicated are at entry to the trial. For information on numbers in each group, see the 'Characteristics of included studies' table or the Analyses).
Gait, balance, co-ordination and functional tasks
Eleven trials involving 694 participants at entry investigated the effects of exercise programmes involving gait, balance, co-ordination and functional task activities on balance performance. The content of the exercise programmes was varied. Brouwer 2003 (38 participants) included low resistance exercises against gravity, using theraband for legs and trunk, reaching, weight shifting, marching on spot, and a home exercise programme. Crilly 1989 (50 participants) included exercise aimed at improving breathing, single and double limb balance, co-ordination, flexibility, strength and relaxation. Islam 2004 (43 participants) included balance exercises designed to challenge the visual (e.g. opened /closed eyes), vestibular (e.g. move head), somatosensory (e.g. stand on foam) and muscular (e.g. standing on one leg, bending body in different directions) systems. Lichtenstein 1989 (50 participants) included stretching, "static balance" (e.g. standing on one leg), "active balance" (e.g. using tandem heel/toe gait, walking along a line), "response exercises" (e.g. performing manoeuvres in response to changing colour signals), walking and cool-down and relaxation. MacRae 1994 (80 participants) included a stand up, step up routine designed to improve strength and balance with warm up and cool down. McGarry 2001 (22 participants) included the "Get off your Rocker" balance class, including single leg stance, exercises with Swiss balls and tandem walking. Reinsch 1992 (107 participants) included stand-ups and step-ups and functional exercises. Sihvonen 2004 (28 participants) included dynamic exercise on a force platform and training device with visual feedback on movement of the Centre of Pressure (COP). Wolf 1997 (72 participants) included standing on a force platform using exercise to move a target via a cursor on screen, this involved performing excursions outside the base of support with eyes open and closed. Wolf 2001 (94 participants) included exercise in sitting, standing and walking, in a variety of situations to test balance. Wolfson 1996 (110 participants) included exercise on a PRObalancemaster with COP feedback, standing and sitting, exercises using gym balls with eyes open and eyes closed, with and without perturbations, and gait on foam and narrow beams.
Strengthening (including resistance or power training)
Six trials involving 513 participants at entry investigated the effects of exercise programmes involving strengthening, including resistance or power training, on balance performance. Boshuizen 2005 (110 participants) included strengthening exercises of lower limbs with theraband and increasing resistance in sitting and standing. Buchner 1997a (51 participants) included free weights and gym equipment. Cress 1999 (56 participants) included combined endurance and resistance exercises. Krebs 1998 (132 participants) included the 'strong for life programme' a 35 minute video of 11 exercises, resistance using elastic bands, functional movement patterns similar to proprioceptive neuromuscular facilitation (PNF) for the arms and legs. Rooks 1997a (131 participants) included stair climbing with resistance, seated knee extension, standing, standing knee extension. Wolfson 1996 (110 participants) included stretching and progressive resistive exercise with sand bags for the hip and knee.
3D exercise (including tai chi, qi gong, dance, yoga)
Four trials involving 193 participants at entry investigated the effects of 3D exercise programmes on balance performance. Buchner 1997b (106 participants) included exercise involving dance movement to music. Shigematsu 2002 (38 participants) included aerobic dance exercise to music. Wolf 1997 (72 participants) included ten forms of tai chi quan. Zhang 2006 (49 participants) included a simplified form of 24 forms of tai chi plus 11 easy forms for home exercise.
General physical activity
Only two trials involving 91 participants at entry investigated the effects of general physical activity on balance performance. McMurdo 1993 (49 participants) included exercises that were performed seated and following a warm-up, exercises were designed to put joints in upper and lower limbs through their full range of movements. As the study progressed participants were encouraged to sustain muscle contractions for longer and increase the number of repetitions. In Okumiya 1996, 42 participants completed a warm up, light aerobic exercise aimed at improving neuromotor co-ordination and muscle-strengthening exercises, followed by a cool down.
General physical activity (walking)
Four trials involving 290 participants at entry investigated the effects of walking on balance performance. Buchner 1997b (106 participants) involved participants walking outdoors. Paillard 2004 (21 participants) included individual walking programmes determined by lactate levels during a VO2 max test. Rooks 1997a (131 participants) included participants walking at their own pace on level ground. Shimada 2004 (32 participants) involved gait training on a bilateral separated treadmill.
General physical activity (cycling)
Buchner 1997b (106 participants at entry) investigated the effects of cycling on a static cycle on balance performance.
Multiple intervention types (combinations of the above)
Twelve trials involving 1559 participants at entry investigated the effects of multiple exercise types on balance performance. Jessup 2003 (18 participants) included strength training that began with 8 to 10 repetitions at 50% of the pre-test one repetition maximum (1RM) score and progressed to 75% RM. This programme also included load-bearing walking, stair-climbing and balance-training exercises while wearing weighted vests; and balance-training exercises in walking. Johansson 1991 (34 participants) involved walking in different directions at different speeds, combined with exercises involving movement of the arms, neck and trunk and exercise to music including weight transfer exercises while sitting and standing and rising from and sitting down in a chair. Lord 1995 (197 participants) involved an exercise programme aimed at improving strength, flexibility, co-ordination, and balance; the individualised exercise regimes were based on the participants' falls risk profile. In Lord 2003 (280 participants), the exercise programme included a warm-up period, conditioning period including aerobic exercises, specific strengthening exercises, and activities for balance, hand-eye and foot-eye coordination, and flexibility. In Lord 2005 (620 participants), the programme was based on falls risk profile, individualised exercises aimed at improving strength and balance and/or vision if a problem. It included a peripheral warm up, conditioning, strength, flexibility, coordination and balance. Nelson 2004 (72 participants) included exercise for balance and strength using free weights working at 7/8 on a 10-point Borg Scale, tandem walks, turning, plus 120 minutes of physical activity per week. Ramsbottom 2004 (22 participants) included free weights to strengthen and develop power in shoulder, hip adductors, abductors, flexors and extensors, the knee flexor and extensors. Exercise progressed by increasing the number of repetitions. The programme also included functional mobility, stretching and balance exercises. Rubenstein 2000 (59 participants) involved Progressive Resistance Exercise (PRE) for the muscles of the hip, knee and ankle, endurance training on a bike and treadmill and indoor walking and balance training. Sauvage 1992 (14 participants) included PRE and aerobic conditioning (>70% exercise stress tested maximal heart rate) using gym equipment and ergometers. Schoenfelder 2004 (81 participants) included strength and endurance training plus 10 minutes walking. Suzuki 2004 (52 participants) included an exercise centred falls prevention programme that included home based exercise aimed at enhancing muscle strength, balance and gait and also included resistance exercise and Tai Chi. Wolfson 1996 (110 participants) included exercise on a PRObalancemaster with centre of pressure feedback, standing and sitting including exercise with gym balls with eyes open and eyes closed, and with and without perturbations. It also included gait on foam and narrow beams, stretching and PRE with sand bags for the hip and knee.
Exercise delivery: settings
The exercise interventions took place in a variety of settings; in institutions - five trials (Crilly 1989; McMurdo 1993; Schoenfelder 2004; Shimada 2004; Sihvonen 2004); home - three trials (Krebs 1998; Nelson 2004; Wolf 2001); community - 13 trials (Boshuizen 2005; Lichtenstein 1989; Lord 1995; Lord 2003; Lord 2005; Okumiya 1996; Paillard 2004; Ramsbottom 2004; Reinsch 1992; Rooks 1997a; Shigematsu 2002; Suzuki 2004; Zhang 2006); and gymnasium or clinic - 14 trials (Brouwer 2003; Buchner 1997a; Buchner 1997b; Cress 1999; Islam 2004; Jessup 2003; Johansson 1991; MacRae 1994; McGarry 2001; Rubenstein 2000; Sauvage 1992; Wolf 1997; Wolf 2001; Wolfson 1996). Wolf 2001 was a factorial design with 2 arms of the study involving different interventions taking place in different settings.
The interventions were delivered mainly as part of supervised groups (28 trials); or individually supervised - four trials (Shimada 2004; Sihvonen 2004; Wolf 2001; Wolfson 1996); or self-supervised (for example using exercise sheets/video) - three trials (Krebs 1998; Nelson 2004; Okumiya 1996).
The supervisors were healthcare professionals or fitness instructors in 19 trials (Boshuizen 2005; Brouwer 2003; Crilly 1989; Islam 2004; Johansson 1991; Lord 1995; Lord 2003; Lord 2005; MacRae 1994; McGarry 2001; Okumiya 1996; Ramsbottom 2004; Rooks 1997a; Rubenstein 2000; Shigematsu 2002; Shimada 2004; Wolf 1997; Wolf 2001; Zhang 2006). The background of the supervisor was not stated in 12 trials (Buchner 1997a; Buchner 1997b; Cress 1999; Jessup 2003; Johansson 1991; Lichtenstein 1989; McMurdo 1993; Paillard 2004; Sauvage 1992; Sihvonen 2004; Suzuki 2004; Wolfson 1996).
Exercise delivery: duration
The duration of the exercise programmes ranged from 4 weeks (Sihvonen 2004) to 12 months (Lord 1995; Lord 2003; Lord 2005; Reinsch 1992) with the most frequent being 3 months. The frequency of the individual sessions ranged from once every two weeks (Suzuki 2004) to every day (Zhang 2006), the most common being three times per week. The duration of each session ranged from 15 mins (Schoenfelder 2004) to 90 mins (Jessup 2003), the most frequent being one hour.
Exercise delivery: compliance
The definition of adherence or compliance with the exercise intervention and the method of recording and reporting varied considerably across trials and thus these data are difficult to interpret. Typically adherence or compliance was reported as the median or mean percentage of actual sessions completed compared to the total available sessions. This was reported in 15 trials and ranged from 65% (Crilly 1989) to 95% (Sauvage 1992).
Further details are provided in the 'Characteristics of included studies' tables.
Comparison interventions
We compared exercise interventions with a control group. The control group was usual activities in 24 trials (Boshuizen 2005; Buchner 1997a; Buchner 1997b; Cress 1999; Crilly 1989; Islam 2004; Jessup 2003; Johansson 1991; Krebs 1998; Lichtenstein 1989; Lord 1995; Lord 2003; Lord 2005; McGarry 2001; Okumiya 1996; Paillard 2004; Rooks 1997a; Rubenstein 2000; Sauvage 1992; Shigematsu 2002; Shimada 2004; Sihvonen 2004; Suzuki 2004; Zhang 2006) and attention or recreational activities in 10 trials (Crilly 1989; MacRae 1994; McMurdo 1993; Nelson 2004; Ramsbottom 2004; Reinsch 1992; Schoenfelder 2004; Wolf 1997; Wolf 2001; Wolfson 1996).
Exercise was categorised as:
01: Gait, balance, co-ordination and functional tasks;
02: Strengthening;
03: 3D exercise;
04: General physical activity;
05: General physical activity (walking);
06: General physical activity (cycling);
07: Multiple intervention types (combinations of the above).
Outcomes
We only included trials that reported primary outcome measures that were direct or indirect measures of balance performance, however a wide variety of outcomes (11 broad categories described below) were assessed in these trials and often they utilised different methods of data collection and reporting.
Direct measures
Force platform and sway indicators
Force platforms allow the measurement of the movement of the centre of pressure, or limits of stability, under different conditions. Force platforms or sway meters were used in 16 trials (Brouwer 2003; Buchner 1997b; Crilly 1989; Islam 2004; Jessup 2003; Lichtenstein 1989; Lord 1995; Lord 2003; Lord 2005; McMurdo 1993; Paillard 2004; Ramsbottom 2004; Sauvage 1992; Sihvonen 2004; Wolf 1997; Wolfson 1996). Typically when these tests are performed under static conditions (e.g. quiet stance, one leg stance) lower values indicate better balance ability but when performed under dynamic conditions (e.g. leaning forwards, backwards and sideways) higher values indicate better balance ability.
Indirect quantifiable measures
Functional reach
The distance an individual can reach forward beyond arms length while maintaining a fixed base of support in standing (Duncan 1990) was used in six trials (Cress 1999; McGarry 2001; Okumiya 1996; Ramsbottom 2004; Shigematsu 2002; Shimada 2004). Higher values indicate better balance ability.
Timed up-and-go
Timed up-and-go (i.e. time to stand, walk 3 m, turn, and return to sitting), measured in seconds (Podsiadlo 1991), was used in four trials (Boshuizen 2005; McGarry 2001; Okumiya 1996; Ramsbottom 2004). Lower values indicate better balance ability.
Gait speed
Gait speed, time to walk a known pre-determined distance, was used as an outcome in 17 trials (Boshuizen 2005; Brouwer 2003; Buchner 1997a; Buchner 1997b; Cress 1999; Johansson 1991; Krebs 1998; Lichtenstein 1989; MacRae 1994; Nelson 2004; Rooks 1997a; Sauvage 1992; Schoenfelder 2004; Shimada 2004; Suzuki 2004; Wolfson 1996; Zhang 2006). This was expressed in different units of measurement; velocity (e.g. m/s, cm/min, m/min), or time (s) taken to complete the required distance. A higher value of velocity indicates faster mobility and thus better balance ability, whereas a higher time to complete a required distance indicates slower mobility. The distance walked varied from 2 m (Nelson 2004) to 30 m (Johansson 1991) and was typically measured at the participant's preferred pace of walking usually this was from a standing start and finish but sometimes included acceleration and deceleration distances.
Single legged stance
Single legged stance is the ability to balance on one leg measured as the time before placing the opposite leg on the ground. This outcome measure was used in 13 trials (Buchner 1997a; Johansson 1991; Lichtenstein 1989; MacRae 1994; Nelson 2004; Reinsch 1992; Rooks 1997a; Rubenstein 2000; Shigematsu 2002; Shimada 2004; Suzuki 2004; Wolfson 1996; Zhang 2006). This test was undertaken in a variety of conditions; eyes open and eyes closed and was in some cases measured subject to ceiling effects with a maximum time allowed ranging from 15 s (Rubenstein 2000) to one minute (Suzuki 2004). Higher values indicate better balance ability.
Parallel stance
Parallel stance is the ability to stand with both feet placed beside each other measured as the time before loss of balance and movement of either leg. This outcome measure was used in two trials (Buchner 1997a; Schoenfelder 2004). Higher values indicate better balance ability.
Tandem stance
Tandem stance is the ability to stand with one foot placed in front of the other and touching heel to toe measured as the time before loss of balance and movement of either leg. This outcome measure was used in four trials (Boshuizen 2005; Buchner 1997a; Rooks 1997a; Schoenfelder 2004). Higher values indicate better balance ability.
Tandem walk
Tandem walk is the ability to walk with one foot placed in front of the other and touching heel to toe, measured as the time taken to walk a set distance or the number of steps taken before loss of balance occurs. This outcome measure was used in three trials (Nelson 2004; Rooks 1997a; Suzuki 2004). Higher values indicate better balance ability.
Tilt boards
The ability to maintain balance whilst standing on a tilt board that allows movement only in the antero-posterior direction or multiple directions, measured in time to loss of balance, was used in two trials (Buchner 1997a; Buchner 1997b). Higher values indicate better balance ability.
Balance beams
The ability to walk on wide (17 cm) and narrow (8.5 cm) beams, measured as distance completed before loss of balance (m), or speed of walking (m/s), was used in 4 trials (Buchner 1997a; Buchner 1997b; Cress 1999; Johansson 1991). Higher values indicate better balance ability.
Indirect observational measures
Berg Balance Scale
The Berg Balance Scale is a 56 point scale comprising 14 items of activities of daily living deemed safe for elderly people to perform, each item is scored 0-4 (Berg 1992). This was used in three trials (McGarry 2001; Sihvonen 2004; Wolf 2001). Higher values indicate better balance ability.
Risk of bias in included studies
Methodological quality assessment scores for each item are given in Table 3 and Table 4, a brief summary is provided here.
Randomisation
The reported method of randomisation included random number tables, block randomisation using permuted blocks, and stratification. However, 13 trials did not state the method of randomisation (Boshuizen 2005; Brouwer 2003; Buchner 1997b; Cress 1999; Islam 2004; Krebs 1998; Johansson 1991; McGarry 2001; Okumiya 1996; Paillard 2004; Sauvage 1992; Schoenfelder 2004; Wolf 1997). Of the seven trials that were cluster randomised (Lichtenstein 1989; McMurdo 1993; MacRae 1994; Lord 1995; Lord 2003; Reinsch 1992; Shigematsu 2002), none reported that an adjustment had been made for cluster randomisation. The units of randomisation were place of residence or community centre but the analysis in the primary reporting was by individual.
Allocation
Allocation concealment was adequate in 10 trials (Jessup 2003; Lichtenstein 1989; Lord 2003; Lord 2005; MacRae 1994; Ramsbottom 2004; Rubenstein 2000; Shigematsu 2002; Wolf 2001; Wolfson 1996); unclear in 22 trials (Boshuizen 2005; Brouwer 2003; Buchner 1997a; Buchner 1997b; Cress 1999; Crilly 1989; Islam 2004; Johansson 1991; Krebs 1998; Lord 1995; McMurdo 1993; Nelson 2004; Okumiya 1996; Paillard 2004; Reinsch 1992; Rooks 1997a; Sauvage 1992; Schoenfelder 2004; Sihvonen 2004; Suzuki 2004; Wolf 1997; Zhang 2006); and not used in two trials (McGarry 2001; Shimada 2004).
Blinding
It is difficult to ensure blinding of participants in trials of exercise interventions. In an attempt to minimise bias, 10 trials used attention or recreational control groups (the participants received matching periods of attention or recreational activity) (Crilly 1989; MacRae 1994; McMurdo 1993; Nelson 2004; Ramsbottom 2004; Reinsch 1992; Schoenfelder 2004; Wolf 1997; Wolf 2001; Wolfson 1996). Seventeen trials stated that assessors for all outcomes were blind to the group allocation (Boshuizen 2005; Buchner 1997a; Buchner 1997b; Krebs 1998; Jessup 2003; Johansson 1991; Lord 1995; Lord 2005; McMurdo 1993; Nelson 2004; Okumiya 1996; Rubenstein 2000; Sauvage 1992; Schoenfelder 2004; Suzuki 2004; Wolf 2001; Wolfson 1996); however, 17 trials did not report the status of blinding of assessors (Brouwer 2003; Cress 1999; Crilly 1989; Islam 2004; Lichtenstein 1989; Lord 2003; MacRae 1994; McGarry 2001; Paillard 2004; Ramsbottom 2004; Reinsch 1992; Rooks 1997a; Shigematsu 2002; Shimada 2004; Sihvonen 2004; Wolf 1997; Zhang 2006).
Follow-up and exclusions
Most trials (n = 29) did not have any follow up beyond the end of the programme of exercise intervention. For those five trials reporting follow-up (Brouwer 2003; Buchner 1997b; Schoenfelder 2004; Wolf 2001; Wolfson 1996) the duration varied from 6 weeks (Brouwer 2003) to 1 year (Wolf 2001). Twenty-nine trials reported losses that ranged from 3% of participants (Johansson 1991) to 48% of participants (Wolf 2001), however, five trials did not report whether any losses were incurred (McGarry 2001; Paillard 2004; Shigematsu 2002; Wolf 1997; Wolfson 1996). Most trials included only participants that completed the entire trial in their analysis whereas eight trials stated that they used intention-to-treat analysis (Buchner 1997a; Buchner 1997b; Cress 1999; Ramsbottom 2004; Reinsch 1992; Rooks 1997a; Suzuki 2004; Wolf 2001).
Study size
Nine trials had more than 100 participants at entry (Buchner 1997a; Buchner 1997b; Krebs 1998; Lord 1995; Lord 2003; Lord 2005; Reinsch 1992; Rooks 1997a; Wolfson 1996) but most were small. Ten trials had less than 40 participants at entry (Brouwer 2003; Jessup 2003; Johansson 1991; McGarry 2001; Paillard 2004; Ramsbottom 2004; Sauvage 1992; Shigematsu 2002; Shimada 2004; Sihvonen 2004). The smallest sample was Sauvage 1992 with only 14 participants.
Effects of interventions
Gait, balance co-ordination and functional tasks versus control (Analyses 01.01 to 01.16)
Eleven trials involving 694 participants at entry investigated the effects of exercise programmes involving gait, balance, co-ordination and functional task activities versus control. Sixteen different outcome measures representative of direct, indirect quantifiable and observational measures were used to evaluate these interventions. Statistically significant differences were observed in only five of these measures.
01.01 Anterior-Posterior (A-P) stability during stance (quiet and dynamic) with eyes open
Three studies (Brouwer 2003; Crilly 1989; Wolf 1997) provided data on AP stability during stance with eyes open. A meta-analysis of standardised differences in means gave a statistically significant decrease of 0.71 standard deviations immediately after the exercise intervention (95% CI -1.33 to -0.09) indicating better balance ability in the exercise groups (n = 116). Two studies reported follow up. For Brouwer 2003 (n = 30) there was a tendency towards better balance ability of the exercise group (P = 0.09) (95% CI -1.37 to 0.10) at 6 weeks post intervention, while Wolf 1997 (n = 35) reported a statistically significant difference of 0.96 standard deviations at 4 months post intervention (CI -1.67 to -0.26) in favour of the exercise group.
0 1.02 Medio-lateral (ML) stability during stance (quiet and dynamic) with eyes open
Three studies (Brouwer 2003; Crilly 1989; Wolf 1997) (n = 116) provided data on this outcome measure. A meta-analysis of standardised differences in means indicated a tendency towards better balance ability in the exercise groups immediately after the exercise intervention (P = 0.19) (95% CI -0.98 to 0.20). This tendency was maintained at 6 weeks post intervention (Brouwer 2003) (P = 0.07) (95% CI -1.42 to 0.06). However, Wolf 1997 (n = 35) observed a statistically significant increase of 1.09 standard deviations at 4 months post intervention (95% CI 0.37 to 1.81) indicating better balance ability in the control group.
01.05 Functional base of support during a dynamic test
One study (Wolfson 1996), involving analysis of 35 participants, provided data on this outcome measure. A statistically significant difference of 0.12 cm, (95% CI 0.05 to 0.19) was observed immediately post intervention and a difference of 0.08 cm (95% CI 0.01 to 0.15) was maintained at 6 months post intervention (n = 33) indicating better balance ability in the exercise group.
01.06 Loss of balance during sensory organisation test
Only one study (Wolfson 1996) provided data on this outcome measure and tendency towards better balance ability in the exercise group (P = 0.06) (95% CI -2.24 to 0.04) was observed immediately post intervention (n = 53) and there was a statistically significant difference in the number of errors of 1.10 (95% CI -2.16 to -0.04) at 6 months post intervention (n = 47) indicating better balance ability in the exercise group.
01.07 Maximum excursion of limits of stability test
Only one small study (Islam 2004) (n = 29) used this measure. Statistically significant differences in maximum excursion were observed for leaning forward (20.10 mm, 95% CI 8.66 to 31.54), right (19.00 mm, 95% CI 9.02 to 28.98) and left (12.80 mm, 95% CI 4.10 to 21.50) indicating better balance ability in the exercise group. A tendency towards better balance was observed for the exercise group for leaning backwards.
01.10 Single leg stance time eyes open
Four studies (Johansson 1991; MacRae 1994; Reinsch 1992; Wolfson 1996) provided data on 164 participants' single legged stance time with eyes open. A meta-analysis of standardised differences in means gave a statistically significant increase of 0.33 standard deviations immediately after the exercise intervention (95% CI 0.02 to 0.64). However this difference does not appear to be maintained in one study (Wolfson 1996) (n = 37) at 6 months follow up.
01.16 Berg Balance Scale
Three studies (McGarry 2001; Sihvonen 2004; Wolf 2001) provided data using the Berg Balance Scale. A statistically significant difference of 2.72 points (Weighted Mean Difference (WMD)) across the three studies (n = 126) was observed immediately after the exercise intervention (95% CI 0.94 to 4.50), indicating better balance ability in the exercise groups. In one study (Wolf 2001) at 4 weeks post intervention (n = 77) there is still a tendency towards better balance ability in the exercise group but this does not appear to be maintained at 1 year post intervention (n = 49).
There was no evidence of effect on average across all other outcome measures.
02. Strengthening exercise versus control (Analyses 02.01 to 02.11)
Six trials (Boshuizen 2005; Buchner 1997a; Cress 1999; Krebs 1998; Rooks 1997a; Wolfson 1996) investigated the effect of exercise programmes involving strengthening exercises versus controls. Eleven different outcome measures representative of direct and indirect quantifiable measures were used to evaluate these interventions. Statistically significant differences were observed in four of these measures.
02.03 Tilt board time
One study (Buchner 1997a) provided data on ability to stand on a tilt board. A statistically significant mean difference of 4.00 s (95% CI -7.89 to -0.11) was observed for omnidirectional tilt indicating better balance ability in the control group. However, no statistically significant differences were observed for anteroposterior tilt.
02.04 Single legged stance eyes open
Three studies (Buchner 1997a; Rooks 1997a; Wolfson 1996) provided data on single legged stance time with eyes open (n = 170). A meta-analysis of standardised differences in means gave a statistically significant increase of 0.39 standard deviations immediately after the exercise intervention (95% CI 0.08 to 0.70) indicating better balance ability across the exercise groups.
02.06 Tandem walk
A trend towards better balance in the exercise group (P = 0.10) (95% CI -4.40 to 0.40) was observed for one study (Rooks 1997a) in time to walk 10 feet.
02.08 Functional Reach Test
One study (Cress 1999) reported data on the functional reach of participants (n = 49) immediately post intervention. A statistically significant difference of 4.33 cm (95% CI 8.00 to 0.66) was observed indicating better balance ability in the exercise group.
02.10 Gait speed
Five studies (Boshuizen 2005; Buchner 1997a; Cress 1999; Krebs 1998; Wolfson 1996) provided data on gait speed (n = 304). A meta-analysis of standardised differences in means gave a statistically significant increase of 0.25 standard deviations immediately after the exercise intervention (95% CI 0.2 to 0.48), indicating better balance ability across the exercise groups.
There was no evidence of effect on average across all other outcome measures.
03. 3D exercise versus control (Analyses 03.01 to 03.16)
Four studies (Buchner 1997b; Shigematsu 2002; Wolf 1997; Zhang 2006) involving 265 participants at entry investigated the effects of exercise programmes involving 3D exercise versus controls. Fifteen different outcome measures representative of direct and indirect quantifiable measures were used to evaluate these interventions. Statistically significant differences were observed in only one of these measures.
03.10 Single leg stance eyes open
Two studies (Shigematsu 2002; Zhang 2006) provided data on this outcome measure (n = 85). A trend towards better balance ability in the exercise group was observed immediately after the exercise intervention (P = 0.09) (95% CI -0.17 to 2.44).
03.14 and 03.15 Walking on beams
One study (Buchner 1997b) provided data for participants walking on a wide and narrow balance beams (n = 52). For the wide balance beam a statistically significant difference in velocity of 0.10 m/s (95% CI -0.16 to 0.04) was observed immediately post intervention indicating better balance ability in the control group and this trend was observed at 3 months follow up (95% CI -0.17 to 0.03). However for the narrow balance beam, a trend towards better balance ability in the exercise group (95% CI -0.01 to 1.61) was observed immediately post intervention and this was maintained at 3 months follow up.
There was no evidence of effect on average across all other outcome measures.
04. General physical activity versus control (Analyses 04.01 to 04.03)
Two studies (McMurdo 1993; Okumiya 1996) involving 83 participants at entry investigated the effects of exercise programmes involving 3D exercise versus controls. Three different outcome measures representative of direct and indirect quantifiable measures were used to evaluate these interventions. Statistically significant differences were observed in two of these measures.
04.02 Functional Reach Test
One study (Okumiya 1996) provided data for 42 participants performing the functional reach test. A statistically significant difference of 11.80 cm (95% CI 7.75 to 15.85) was observed immediately post intervention indicating better balance ability in the exercise group.
04.03 Timed up-and-go test
One study (Okumiya 1996) provided data for 42 participants performing the timed up-and-go test. A statistically significant difference in time taken -3.90 s (95% CI -5.83 to -1.97) was observed immediately post intervention indicating better balance ability in the exercise group.
There was no evidence of effect on average in the other outcome measure postural sway double stance.
05. General physical activity (walking) versus control (Analyses 05.01 to 05.15)
Four studies (Buchner 1997b; Paillard 2004; Rooks 1997a; Shimada 2004) involving 235 participants at entry investigated the effects of exercise programmes involving walking as general physical activity versus controls. Sixteen different outcome measures representative of direct and indirect quantifiable measures were used to evaluate these interventions. Statistically significant differences were observed in five of these measures.
05.03 Area during narrow stance eyes closed
One study (Buchner 1997b) provided data for participants in quiet narrow stance with eyes closed. A statistically significant difference of 102 mm
05.04 Angular radius during narrow stance eyes closed
There was a tendency towards better balance in the control group immediately and 3 months after the intervention for angular radius during narrow stance with eyes open (Buchner 1997b) (P = 0.10) (95%CI -0.26 to 3.06).
05.06 Dynamic balance lateral axis
One study (Paillard 2004) provided data for this outcome measure (n = 21). No statistically significant difference was observed immediately post intervention but a statistically significant difference of -2.5 degrees (95% CI -4.01 to -0.99) was observed between the groups at 3 months post intervention, indicating a better balance ability in the control group.
05.11 Tandem walk
One study (Rooks 1997a) provided data on this outcome measure (n = 69). A statistically significant difference of 2.30 s (95% CI 0.55 to 4.05) was observed immediately post intervention indicating better balance ability in the exercise group.
05.12 Tandem stance
One study (Rooks 1997a) provided data on this outcome measure (n = 69). A statistically significant difference of 12.90 s (95% CI 3.91 to 21.89) was observed immediately post intervention, indicating better balance ability in the exercise group.
05.13 Functional Reach Test
One study (Shimada 2004) provided data for 26 participants performing the functional reach test. A statistically significant difference of 10.92 cm (95% CI 5.03 to 16.81) between the groups was observed immediately post intervention indicating better balance ability in the exercise group.
05.15 & 05.16 Walking on wide and narrow beams
Buchner 1997b provided data for ability to walk on a wide beam. A trend towards better balance ability in the exercise group was observed immediately post intervention (n = 52) (95% CI -0.01 to 0.13) and at 3 months post intervention (n = 51) (95% CI -0.01 to 0.17).
Buchner 1997b also provided data for ability to walk on a narrow beam. A statistically significant difference of 0.50 m/s (95% CI -0.07 to 1.07) was observed immediately post intervention (n = 52) indicating a better balance ability in the exercise group but this was not maintained at 3 months post intervention (n = 51).
There was no evidence of effect on average across all other outcome measures.
06. General physical activity (cycling) versus control (Analyses 06.01 to 06.09)
One study (Buchner 1997b) involving 106 participants at entry investigated the effects of exercise programmes involving cycling as general physical activity versus controls. Nine different outcome measures representative of direct and indirect quantifiable measures were used to evaluate these interventions. There was no statistically significant effect on average across all outcome measures for which data was provided. However, a trend towards better balance ability in the exercise group was observed for the following outcome measures; 06.01 area during narrow stance eyes open; 06.02 angular radius during narrow stance eyes open; 06.07 self paced gait velocity. A trend towards better balance ability in the control group was observed for the following outcomes; 06.03 area narrow stance eyes closed, 06.04 angular radius during narrow stance eyes closed.
07. Multiple exercise types versus control (Analyses 07.01 to 07.16)
Twelve studies (Jessup 2003; Johansson 1991; Lord 1995; Lord 2003; Lord 2005; Nelson 2004; Ramsbottom 2004; Rubenstein 2000; Sauvage 1992; Schoenfelder 2004; Suzuki 2004; Wolfson 1996) involving 1559 participants at entry investigated the effects of exercise programmes involving multiple exercise types versus controls. Sixteen different outcome measures, representative of direct and indirect quantifiable measures, were used to evaluate these interventions. Statistically significant differences were observed in seven of these measures.
07.01 Functional base of support during a dynamic test
One study (Wolfson 1996) provided data for this outcome measure. A statistically significant difference of 0.09 proportion of foot length (95% CI 0.03 to 0.15) was observed between the groups immediately post intervention (n = 32) and 0.09 proportion of foot length (95% CI 0.02 to 0.16) at 6 months post intervention (n = 26) indicating better balance ability in the exercise group.
07.03 Total distance travelled by COP during quiet stance
One study (Sauvage 1992) provided data for this outcome measure for 14 participants under eyes open and eyes closed conditions. Statistically significant differences were observed between the groups immediately post intervention of 97.15 mm (95% CI 18.59 to 175.71) with eyes open and 212.52 mm (95% CI 114.79 to 310.25) with eyes closed indicating better balance ability in the exercise group.
07.05 Body sway
Two studies (Jessup 2003; Ramsbottom 2004) provided data for this outcome measure. A statistically significant decrease of 0.87 standard deviations across the two studies (n = 35) was observed immediately after the exercise intervention (95% CI -1.60 to -0.13) indicating better balance ability in the exercise group.
07.12 Tandem stance time
One study (Schoenfelder 2004) provided data for this outcome for 67 participants. A statistically significant difference of 0.80 s (95% CI -0.47 to 2.07) was observed indicating better balance ability in the exercise group.
07.13 Tandem walking (number of steps)
One study (Suzuki 2004) provided data for the number of steps able to be made during tandem walking (n = 39). A statistically significant difference of 3.39 steps (95% CI 1.75 to 5.03) was observed between the groups immediately post intervention indicating better balance ability in the exercise group.
07.14 Tandem walking time
One study (Nelson 2004) provided data for tandem walk time. A statistically significant difference of 8.10 s (95% CI 2.49 to 13.71) was observed between the groups (n = 70) immediately post intervention indicating better balance ability in the exercise group.
07.15 Functional Reach Test
Two studies (Ramsbottom 2004; Suzuki 2004) provided data on distance reached during the functional reach test. A statistically significant difference of 5.80 (95% CI 3.37 to 8.23) across the two studies (n = 60) was observed immediately after the exercise intervention indicating better balance ability in the exercise group.
There was no evidence of effect on average across all other outcome measures.
08. Sensitivity analyses for effect of clustering (Analyses 08.01 to 08.02)
08.01 Gait, balance, co-ordination, functional task exercise versus control (01.10) Single leg stance time eyes open
Two cluster randomised studies were removed from the meta-analysis (MacRae 1994; Reinsch 1992). Two studies (Johansson 1991; Wolfson 1996) provided data on 72 participants single legged stance time with eyes open. A meta-analysis of standardised differences in means demonstrated a trend towards an improvement in balance in favour of exercise immediately after the exercise intervention (P = 0.07, 95% CI -0.04 to 0.90).
08.02 Gait, balance, co-ordination, functional task exercise versus control (01.14) Self paced gait speed
One cluster randomised study was removed from the meta-analysis (MacRae 1994). Three studies (Brouwer 2003; Johansson 1991; Wolfson 1996) provided data on 117 participants self paced gait speed. A meta-analysis of standardised differences in means gave a statistically significant increase of 0.45 standard deviations immediately after the exercise intervention (P = 0.02, 95% CI 0.08 to 0.82).
Discussion
Summary of main results
Statistically significant improvements were observed in balance ability assessed across a variety of outcome measures for exercise interventions compared to control (usual, recreational or attentional activity). Exercise interventions were heterogeneous and were categorised into seven types. They took place mainly in gym/clinic or community settings in supervised groups delivered predominantly by healthcare professionals or fitness instructors. The duration of the exercise programmes ranged from 4 weeks to 12 months, the most frequent being 3 months. The frequency of the individual sessions ranged from once every two weeks to every day, typically three times per week for one hour.
Gait, balance, co-ordination and functional tasks exercise interventions (Analysis 01) showed statistically significant improvements compared with control in direct measures of balance including: AP stability during stance (quiet and dynamic) with eyes open; maximum excursion of locus of support (LOS) test; and functional base of support distance during a dynamic test. Statistically significant improvements in favour of the exercise intervention were also observed in single legged stance time with eyes open (indirect quantifiable measure) and the Berg Balance score (indirect observational measure). A trend towards an improvement in balance in favour of exercise was observed for two other measures.
Strengthening exercise interventions (Analysis 02) showed statistically significant positive effects on time for omnidirectional tilt (direct measure) and on the following indirect quantifiable measures of balance; functional reach, single legged stance with eyes open, tandem stance, and gait speed. A trend towards an improvement in balance was observed for tandem walk.
3D exercise, including Tai Chi and dance (Analysis 03), demonstrated statistically significant improvements in favour of control in walking on a wide balance beam (indirect quantifiable measure). A trend towards an improvement in balance in favour of exercise was observed for single leg stance eyes open. No other improvements were observed in other outcomes relating to direct measures or indirect observational measures.
General physical activity (Analysis 04) demonstrated statistically significant improvements in indirect quantifiable measures of balance; timed up-and-go test and functional reach.
Walking as a general physical activity (Analysis 05) showed statistically significant improvements in favour of exercise for the indirect quantifiable balance measures; tandem walk, tandem stance, walking on wide beams and functional reach. However, statistically significant improvements in favour of the control was observed for area during narrow stance (direct quantifiable balance measure).
There was trend towards improvement of balance in favour of cycling (Analysis 06) this lack of significant improvement could be due to the task specific nature of cycling, any improvements in neuromuscular control and conditioning not translating directly into balance ability. However these results are based on one study.
Interventions of multiple exercise types (Analysis 07) demonstrated statistically significant improvements in direct quantifiable balance measures including: body sway, centre of pressure excursion during quiet stance and functional base of support. Indirect quantifiable measures such as functional reach, tandem stance and tandem walking also improved.
Indirect quantifiable measures of balance such as the functional reach test (FRT), timed up and go test and tandem walking require minimal equipment and are easy to use in the clinical and community environments. The interventions examined demonstrated clinically important improvements compared with control in some of these measures, for example differences of 11.80, 10.92 and 5.80 cm in FRT for Analyses 04.02, 05.13 and 07.15 respectively.
In the main where differences were observed immediately post intervention there was limited evidence to suggest that these effects were maintained over longer periods of time.
Overall completeness and applicability of evidence
The 34 studies included in this review were predominantly in the English language and originate mainly from North America and Europe (n = 25). Whilst this may be seen to limit the applicability of the evidence to these healthcare systems and social environments the evidence has potential generalisability. The majority of participants were healthy community dwelling women and may not have had impairment or activity limitation at baseline. This may have impacted on the capacity of these mainly small studies to reveal any differences, whether positive or negative, between the exercise intervention and control groups. However the majority of participants were on average over 75 years and some studies included participants described as frail or with activity limitations.
The interventions investigated included many commonly utilised categories of exercise such as gait, balance, function, muscle strengthening, walking, cycling, tai chi and dance. The definition of adherence or compliance with the exercise intervention and the method of recording and reporting varied considerably across trials and thus these data are difficult to interpret. However none of the studies included information indicating enthusiasm for uptake of exercise or long term uptake among participants in the programmes.
The wide range of interventions and outcome measures reported across the studies made it difficult to combine outcomes in meta-analysis. The lack of longer term follow-up of outcomes made it difficult to determine any lasting effects. Furthermore the lack of standardisation of measures and their relative validity limit the interpretation of these results. For example, direct measures of balance use force platforms and sway meters. Typically when these tests are performed under static conditions (e.g. quiet stance, one leg stance) lower values indicate better balance ability but when performed under dynamic conditions (e.g. leaning forwards, backwards and sideways) higher values indicate better balance ability. However, there are difficulties in the interpretation of this type of data as in some populations an increased sway under static conditions may indicate better dynamic control whereas less sway may indicate that the individual has an over stiffened system in an effort to maintain stability. Furthermore, for some timed measures of balance, authors applied ceiling effects stipulating a maximum time allowed for the test, and this was not adjusted for in the analysis.
Quality of the evidence
The overall quality of current evidence about the effectiveness of exercise interventions designed to improve balance in older people living in the community or in institutional care is mixed. Of the 34 included trials there were 2883 participants at entry, only nine trials had more than 100 participants at entry, most were small, ten trials had less than 40 participants. Many studies identified demonstrate a range of methodological weaknesses that are exacerbated by inadequate reporting resulting in a large number of studies being placed in 'Studies awaiting assessment'. The main weaknesses were the lack of information about randomisation methods and allocation concealment, blinding of assessors and intention-to-treat analyses. There was limited follow-up data to demonstrate the extent to which the effects of programmes were maintained. Some included studies reported findings based on change scores. This requires measurement of the outcome twice and can result in bias for outcomes that are difficult to measure precisely as the measurement error may be larger than true between person baseline variability. These issues make it difficult to draw firm conclusions.
Potential biases in the review process
There are several potential sources of bias in this review. Although we attempted to extract direct and indirect measures of balance there is a possibility that the measures reported are a biased representation of those collected by the study authors (selective reporting). Indeed there were 11 broad categories of outcome measures used across the studies, some of which were used under a variety of conditions, e.g. eyes open, eyes closed, different surfaces. Only eight studies undertook intention-to-treat analyses the remainder reporting the results for only those participants who completed all post-treatment assessments. The seven studies that were cluster randomised trials did not appear to make adjustments for the cluster effect. As a result, these studies may have overly narrow confidence intervals and will receive more weight than is appropriate in a meta-analysis. Sensitivity analyses were performed to account for this effect. An effect was observed in only two comparisons where a significant difference became a trend (08.01 for 01.10) and a trend became significant (08.02 for 01.14) both in favour of exercise. There are several studies for which the data were inadequately reported and therefore were unable to be included in the analyses.
Agreements and disagreements with other studies or reviews
The objective of this review was to present the best evidence for effectiveness of exercise interventions designed to improve balance in older people living in the community or in institutional care. The general direction of findings presented is in keeping with those of other related systematic reviews: 'Progressive resistance training for physical disability in older people' (Latham 2004) and 'Interventions for preventing falls in elderly people' (Gillespie 2004) where the positive effects of exercise on balance were secondary findings.
Authors' conclusions
Implications for practice A substantial amount of research has been undertaken to assess the effects of exercise on balance and functional ability in older people. Exercise appears to have statistically significant beneficial effects on balance ability compared to usual activity. Interventions involving gait, balance, co-ordination and functional exercises; muscle strengthening; and multiple exercise types, appear to have the greatest impact on indirect measures of balance. However, there was limited evidence that effects were long lasting. The strength of evidence contained within these randomised controlled trials is limited. The failure across the included studies to apply a core set of standardised outcome measures to determine balance ability restricts the capacity to compare or pool different trials from which firm conclusions regarding efficacy can be made. Further standardisation in timing of outcome assessment is also required as is longer term follow-up of outcomes to determine any lasting effects of regular physical activity. |
Implications for research The most important implication for research in this area is for better documenting and reporting of study design and execution. In future trials, the CONSORT statement should be used as a guide for both designing and reporting (www.consort-statement.org). The benefits of exercise interventions on balance may be relatively small. Sample sizes should be reported and have adequate power to answer the research question allowing the detection of clinically significant differences between groups. Reporting should include the method of randomisation and treatment allocation concealment and an intention-to-treat analysis performed. The history and reasons for drop-outs and exclusions throughout the trial should be ascertained and reported so that factors affecting exercise adherence can be further explored. Ideally trials should wherever possible, follow-up participants for at least one year, rather than focussing on immediate post intervention. To enable comparison and pooling of the results of RCTs, we suggest that future trials report means with standard deviations for continuous measures or number of events and total numbers analysed for dichotomous measures. Furthermore, we recommend that a consensus of outcome measures for evaluating the effects of interventions on balance ability that have been demonstrated as sensitive to change be developed similar to the work being undertaken by ProFaNE for falls prevention. |
Acknowledgements
We would like to thank Lesley Gillespie for her valuable assistance with the search strategy, Joanne Elliott, Bill Gillespie, Peter Herbison, Vicki Livingstone, Rajan Madhok, Lindsey Shaw, Janet Wale and Wiebren Zijlstra for critical comments and the stimulating discussions that resulted from these, Mark Waters for input into the protocol, Sallie Lamb for assistance with the taxonomy of exercise interventions, Philip Rowe and Danny Rafferty for advice on outcome measures and Daniel Soule for drafting the plain language summary. In addition, we would like to thank the valuable contributions of Jayne Elms who co-ordinated data collection, searching and retrieval of papers and additional information, entered data into RevMan and commented on drafts and Pam Dawson who screened search results, appraised quality and extracted data from papers.
Data and analyses
- Top of page
- Background
- Objectives
- Methods
- Results
- Discussion
- Authors' conclusions
- Acknowledgements
- Data and analyses
- Appendices
- What's new
- History
- Contributions of authors
- Declarations of interest
- Sources of support
- Index terms
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||
Appendices
Appendix 1. Search strategy for MEDLINE
| ||||||
Appendix 2. Search strategy for The Cochrane Library
| |||
Appendix 3. Search strategy for CINAHL
| |||
Appendix 4. Search strategy for EMBASE
| |||
Appendix 5. Search strategy for AMED
| |||
Appendix 6. Search strategy for PEDro
| |||
Appendix 7. Search strategy for OT seeker
| |||
What's new
Last assessed as up-to-date: 26 July 2007.
| |||||||||
History
Protocol first published: Issue 4, 2004
Review first published: Issue 4, 2007
Contributions of authors
TEH - conceived the review, coordinated data collection, searching and retrieval of papers and additional information, screened all search results, appraised quality and extracted data from all papers, entered data into RevMan, analysed and interpreted data and wrote review. TEH is the guarantor for this review.
LR - conceived the review, screened search results, appraised quality, extracted data from papers, assisted in interpretation of data and critically commented on drafts.
AJ - screened search results, appraised quality and extracted data from papers and commented on drafts
PMHB & VAB - screened search results, appraised quality and extracted data from papers.
Declarations of interest
None known.
Sources of support
Internal sources
- Glasgow Caledonian University, UK.
- University of Northumbria, UK.
- Bell College, UK.
External sources
- Scottish Funding Council, UK.
- Scottish Executive Health Department, UK.
- NHS Education for Scotland, UK.
- Chief Scientist Office, UK.
Index terms
Medical Subject Headings (MeSH)
Breathing Exercises; Dancing; Exercise [*physiology]; Gait [physiology]; Muscle Strength [physiology]; Postural Balance [*physiology]; Randomized Controlled Trials as Topic; Tai Ji; Yoga
MeSH check words
Aged; Female; Humans; Male; Middle Aged
* Indicates the major publication for the study