Does aquatic exercise improve commonly reported predisposing risk factors to falls within the elderly? A systematic review

This manuscript is written in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement which includes controlled and randomized studies [18, 19]. A systematic review was completed to assess the role of aquatic exercise for improving the commonly reported predisposing risk factors of falls in the elderly.

A computer based search was completed between January 2005 and May 2018 and the electronic databases sourced were: ISI Web of Knowledge, ProQuest, PubMed, Science Direct, SPORTDiscus, the Cochrane Central Register of Controlled Trials and the Google Academic Meta searcher. The PICO (Population, Intervention, Comparison and Outcome) framework was used to define the search strategy. The search strategy was developed to include a combination of controlled vocabulary (MeSH) and free text terms which can be found in Table 1. Reviews and commentaries were used to identify papers, but were not included within the synthesis of results. The search reviewed all available studies published between 2005 and 2018. In the mid-nineties, strong research emerged on the benefit of the aquatic environment in rehabilitation compared to the land environment. In particular, an article belong to Tovin [20] who compared the anterior cruciate ligament reconstruction water versus land exercise. However, during this time it was not shown that training in the aquatic environment could be as effective in older adults as land exercise in order to improve one’s physical condition. Until the past decade, no research groups considered that the limitation of the aquatic environment was due to the lack of control of the training load during the exercise. The year 2005 is estimated as an ideal date to begin reviewing literature as before this time, the scientific rigor of relevant studies was low as they were directed solely for therapeutic exercise such as knee injuries or hip osteoarthritis [21]. From studies where the intensity of the exercise is controlled in a systematic way through adapted material, depth, type of exercise and number of series, aquatic exercise can be better understood by healthcare professionals [22].

The study population consisted of healthy adults aged between 60 and 80 years old.

Data were extracted by one reviewer (EMCG) using a standardized template and verified by a second reviewer (JO). Information was extracted from each included trial on: authors, year of publication, study design, population characteristics (number of participants, gender and age), description of intervention and outcome measures. Disagreements or discrepancies on data extraction were resolved by discussion.

The methodological quality of all studies was assessed using the PEDro (Physiotherapy Evidence Database) appraisal instrument [23]. The 11 item scale is a valid and replicable measure used to critically appraise the quality of randomized controlled trials. Each study is scored out of ten, with a threshold of seven (or over) for a study to be considered high quality (item one on the scale indicates external validity).

The flexibility variable was recorded by the chair sit-and-reach and the back scratch test in one study [26, 30] whilst another study used the functional tests of Johnson and Nelson [35, 36]. The remaining studies utilised the sit and reach test [25, 27, 29].

Dynamic balance was measured using the berg balance scale [33], the timed up and go test [24, 25, 27, 28, 35], the 8 ft up and go test [29, 30, 34] and the walking in circle test [36]. Static balance was recorded with the Sharpened Romberg test [24, 28, 35] with both open and closed eyes, it was also registered by a Biodex Balance Stability System [32] and using a force platform where the subjects kept their arms on their hips to maintain position [36].

The variable of strength was predominantly recorded for upper limbs using a hand dynamometer [25, 27, 30, 32]. The maximal dynamic strength was evaluated using the 1-repetition maximum (1RM) bench press [16] bilateral elbow flexion [16, 26, 36] and bilateral knee flexion/extension [16]. The muscular endurance was evaluated during a bilateral knee extension and flexion with a load equivalent to 60% of 1RM maintaining a maximum possible number of repetitions [16]. An isokinetic machine was also used for lower limbs. In this case a knee flexion-extension is performed at speeds of 60°/s,120°/s [26]. In addition, a functional test like the 30 s sit-and-stand-up were used by a large number of researchers [26, 29, 31, 36].

Flexibility was mainly measured by the sit-and-reach test for the hamstrings muscles and lower back [25‐27, 29, 30] and the back-scratch test was used to assess the general shoulder range of motion [26, 30].

When divided into their respective experimental groups, nine of the studies sourced had sample sizes of between 13 and 17 subjects [16, 24‐27, 30‐33, 35]. The remaining studies tested between 20 and 66 subjects [28, 29, 34, 36]. Four studies assessed both male and female participants [27‐30], four studies only included women [16, 25, 31, 36] and three studies only assessed male participants [24, 32, 35]. The age range of the study populations was generally small, however in four studies the range was 15 years [16, 25, 29, 35] and in two studies the range was over 24 years [33, 36].

The temperature of the water varied between 28 °C and 32 °C within the majority of studies, [16, 25‐27, 29, 34‐36] with the exception of four studies, [24, 28, 30, 31] where the temperature was over 32 °C and two studies which do not specify the water temperature [32, 33]. The depth of the water in most of the interventions was from between waist to the midline of the chest [24, 26, 28] and at the level of the xiphoid process [16, 25, 29‐31, 33‐36] except in two studies [30, 32] where the depth reached 1.80 m. The shorter studies had a duration of two weeks [28] and six weeks [24, 30]. Most interventions were implemented for eight to 12 weeks [16, 26, 27, 29, 31‐35]. However, one study completed a 16 weeks intervention [36], and the longest intervention had a duration of 24 weeks [25]. The frequency of the intervention was generally between two sessions [16, 24, 30, 31, 33, 35, 36] and three sessions a week [25‐27, 29, 32, 34] although there is a study with a frequency of 5 days a week for 2 weeks during the intervention [28]. In terms of duration, most of the interventions were 60 min long, [25, 26, 28‐30, 32, 35, 36] however, one study had a duration of 90 min [28], another of 40 min [34] and four studies did not specify the duration of the intervention [17, 25, 32, 35].

In some cases, the intensity of training was described by referring to the Borg Rate of Perceived Exertion Scale, the same Borg scale which ranges from six to 20 points. A value of one represents a very light exercise and 20 a serious physical exertion [37]. Moderate to high intensity was mentioned in ten of the included results [16, 25‐27, 29‐31, 34‐36] and exercise intensity was not systematically recorded in the remaining studies [24, 28, 32, 33].

The control group, absent from any intervention, did not obtain any significant difference at the end of the study [25, 26, 28, 30‐36]. However, the aquatic intervention group obtained improvements (p ≤ 0.05) at the end of the study in the variable of flexibility [25‐27, 29, 30, 36], balance [24, 26, 28, 32, 35, 36] and strength [16, 25‐27, 29‐32, 36].

Regarding the four studies that include a third exercise programme in a land environment, two of them did not find differences between the aquatic and land groups after the intervention, in variables of flexibility, balance and strength [24, 28]. Nevertheless, the aquatic group obtained greater improvement over the terrestrial group in the variable of balance in one study [30]. Another study [27] compared two aquatic groups with different types of intervention: resistance material and without any material. The material group presented better results in gait speed, however, in variables of strength, mobility and flexibility, no significant differences were found between groups.

The studies did not specify what kind of materials they used except in one case [27] where the aquatic group used newly made water-resistance equipment (power leg and power hand DESCENTE Co., Ltd., Japan) and another study [36] where the aquatic group progressed at their own rate by adding surface area equipment (Aqua Flex paddles by MIZUNO Corporation, Osaka, Japan) and by opening the webbing of the Aqua Mitt gloves to increase surface area to reach the goal RPE.

The studies that introduce flexibility training do not indicate specific exercises that are performed [26, 27, 29]. Only five studies indicate the static and dynamic balancing exercises which they use [26, 28, 33, 35, 36]. When looking at strength training there is a general focus on large muscle groups from the upper or lower limbs [16, 24, 26, 27, 31, 32, 34]. Only two studies [31, 36] mention the execution of exercises that resemble activities of daily living (ADL). A schematic compression can be viewed in Table 3.

Studies which assessed balance within their training programmes found significant improvements (p ≤ 0.05) following the intervention [24, 28, 32, 34‐36] with the exception of two studies [26, 33]. The lack of positive results in balance training may be due to the lack of scientific rigor and the specificity of training. In this review, the exercises used most for balance training are alternating movements of the upper and lower limb [24, 26, 28, 33, 35, 36] however, their training variables such as the control of intensity or the specific exercises are not clearly reported in the methodology. Balance is considered the most important variable during the design of a fall prevention training programme [38], and consequently the specificity of the exercises and the control of training should appear in order to reproduce the results obtained in the studies. Therefore, certain criteria or principles that support the decisions made during the fall prevention program should be considered during the design of protocols to improve balance in older people. An example are those established by two experts [8, 38]: the reduction of the support base, the movement of the centre of gravity, the mobilization of the limbs and the reduction of the visual field [8, 38]. It has been recommended that due to the aquatic turbulence during the aquatic exercise, muscle are activated continuously to stabilize the position of the individual [16]. However, the depth of the water and velocity of exercise should be controlled in order to modify muscle activation during balance exercises [39].

One study [36] demonstrated that in a group of 66 subjects (60–89 years), during a 16-week programme which included strength training, in an elderly population it is possible to achieve a better static balance in the elderly. A similar study [24] evaluated the static and dynamic balance in two groups of 15 subjects (63–70 years) during two equally designed training programmes in both a land and aquatic environment. At six weeks, both groups achieved significant improvements in static and dynamic balance (p < 0.05) when compared to pre-test scores. Similarly, a significant gain (p < 0.05) was found for the Berg Balance Score in adults older than 68 years old with no significant difference from one environment to another (p > 0.05).

Although the studies reviewed did not find significant differences between land and water training, the results suggest that an aquatic environment offers a safe environment where the risk of falling is eliminated. It should, however, be noted that there is evidence to highlight the insufficiency of the Berg Balance Scale for predicting falls [40] and the Timed Up and Go Test is instead recommended by the U.S. Centres for Disease Control and Prevention as an easy to administer clinical test to evaluate a senior citizen’s fall risk [41].

One study assessed the effect of aquatic training on agility [20]. The authors found significant improvements in the intervention group receiving an aquatic exercise programme in comparison to the control group (p < 0.05). There is evidence that agility training programmes can reduce the risk of falls [42], largely due to improved balance, which has been examined as a variable for improving mobility in older adults. A study comparing fall prevention programmes in older adults [43] found that visual training had the greatest improvement on obstacle course performance (22%) following a 12-week training program. In addition, agility has been found to reduce the risk of falling through step training [44], and in multifactorial programs which focus on other variables such as balance and strength [45].

Aquatic interventions that trained large muscle groups of the upper and lower limbs found significant improvements in strength from pre-test to post-test (p < 0.05) [16, 25‐27, 29‐32, 36]. Although the methodologies of these protocols vary, all results indicate that an aquatic environment is beneficial for promoting muscle hypertrophy. Strength training cause changes in the nervous system through neuromuscular activation, resulting in improved muscular power and delayed onset of sarcopenia which may decrease the risk of falls [46]. Evidence for the effectiveness of lower limb strengthening exercises for the prevention of falls in elderly subjects has been previously reported [47‐49] and as lower extremity weakness is a clinically significant risk factor of a fall. The design of aquatic exercise programmes should focus on the lower limbs, respecting the principle of training specificity. Lower limb weakness in older adults is considered an important influencing factor of falls and its preservation promotes the maintenance of independence. More specifically, the maximum voluntary isometric strength and the rate of force generation of the hip abductors are considered a useful measure to distinguish between older adult fallers and non-fallers [50]. However, several studies use the handgrip dynamometer as a measure for strength in the elderly [25, 27, 32]. Although this outcome measure does not specifically assess lower limb strength, a recent study developed by the Cooperative Health Research in the Region of Augsburg [51] with 808 individuals (age ≥ 65 years) found a trend toward an indirect effect of grip strength on balance problems (p value = .043). In addition, handgrip strength has been estimated by other studies as an Indicator of Health-Related Quality of Life in Old Age and fall prevention [52, 53].

Axial stiffness can occur through ageing and is a loss of flexibility and elasticity in the neck and trunk that causes the imbalance of the posture, thus reducing balance and increasing the chance of falling. The ACSM (American College of Sports Medicine) [47] highlights the importance of flexibility training during aging for fall prevention; however often it remains neglected in training programmes.

Within this review, six studies assessed flexibility as an outcome following aquatic training [23, 25, 27, 29, 30, 36]. In all cases, flexibility was seen to improve significantly from pre-test to post-test (p ≤ 0.05), however the specific effect on each muscle group was not clearly described. Regarding the principle of specificity and the relationship of flexibility with the risk of falls, the findings indicate that the rigidity in the pelvic muscles causes an alteration in the gait of the elderly, which will contribute to an increased risk of falls [54]. In this case, a specific program of hip flexor stretches is needed, which may reverse variables that increase the risk of falls in the elderly [55].

Six studies [16, 26, 29, 31, 33, 34] agree with others found in the literature [56‐58] that a training programme that seeks to allow for significant changes in balance, strength or flexibility to prevent falls in older adults should be a minimum duration of 12 weeks. To make a positive impact, the duration of the session should be around 60 min and two to three times a week [16, 26, 29, 31‐34], as recommended by similar studies [48, 57, 59, 60]. The ACSM expresses an optimum frequency of three times a week at a moderate-to-high intensity to facilitate fall prevention training in the elderly [47]. This exercise prescription allows for physical recovery that induces physiological adaptations (48–72 h between sessions) [61]. However, studies with less frequency and duration also found significant improvements in balance, strength and flexibility tests [24, 27, 30, 32, 35, 36], after their respective interventions. These results indicate that the intensity of a programme will be a greater modifier of effect than the volume of training to reach physiological adaptations in the elderly and to improve functional capacity [16, 62, 63]. Therefore, it seems likely that frequency, duration and intensity of exercises, including those related to strength, balance, and flexibility are relevant training principles for reducing the risk of falls in the elderly.

The intensity of an exercise programme is an important factor for determining the physiological stress and subsequent adaption an individual can experience through training. If the frequency and intensity of the training programme is too high, muscular damage may occur and this can negatively affect adherence to programmes [64]. The Borg Scale is a valid and simple measure of controlling intensity that has been utilised within the studies analysed [16, 26, 27, 29, 30, 34, 36]. In addition, using Rate of Perceived Exertion (RPE) [37] to prescribe exercise intensity is considered appropriate for older adults [65, 66]. However, in many cases, the number of sets, repetitions or recovery periods was missing from the methodology, preventing the protocols being repeated. Only one study [30] explained the control of the intensity of the flexibility whereby the involved muscles were contracting for 60–90 s during stretching.

Intensity should be controlled like with land exercise but with consideration of the principles of hydrodynamics and the forces that influence aquatic exercise to improve physiological responses in older adults [67]. For example, in one study [17], three different types of training were compared over ten weeks: land training with machines (n = 14), training with elastic bands (n = 21) and water resistance training (n = 17) following the same protocol training but adapted to the environment. The three groups obtained significant differences (p < 0.05) in flexion of arms (push up), squats, crunches and in the fat-free mass. It is shown that controlled resistance training in the aquatic environment can be as beneficial for the improvement of physical condition and body composition as land training when the protocol is adapted to the specific nature of the environment. In this way, a table of the intervention of the studies is designed, which will serve for clinical protocols and researchers to know the variables of exercise prescription of each of the articles analysed (Table 3).

Generally, the methodological quality of the studies included was low. The mean PEDro score for all fourteen studies was 5.5/10 (Table 4). A score of seven (or over) is considered high quality [68], and therefore only on studies in this review had a high quality methodology [33]. Thirteen studies were considered ‘low quality’ when using a threshold of six or less out of ten [16, 24‐32, 34‐36]. Due to the nature some interventions, it was difficult to blind subjects, therapists and assessors and therefore consistently low scoring items on the scale were: three, five, six and seven.

It is difficult to advocate the use of one study over another due to a lack of repeated designs. This makes it difficult to establish a “fixed criteria” in the development of training programmes in the aquatic environment in order to reduce the risk of falls in the elderly.

Although the studies mention the reduction of fall risk with the results obtained, they do not present correlations between variables of strength, balance or flexibility that could suggest a strong clinical relationship. Well-written methodologies should express all relevant information on the training design (volume, intensity, frequency, recovery and type of exercise) in a clear and replicable way. They should also present correlations between falls and physical training variables such as agility, balance, strength and flexibility.

There is some evidence to support using aquatic exercise for improving predisposing, modifiable risk factors of falls. However, the quality of this evidence is low, and many interventions are poorly described. The lack of consistency between study methodologies makes intervention comparison difficult. In addition, the trial findings did not find a statistically significant relationship between training variables and falls. Resistance equipment is not included to increase exercise intensity, and few studies highlight the importance of execution speed during the aquatic exercises. Regarding the types of exercise, most of them refer to general exercises of upper and lower limbs; however, the exercises should be described in detail so that professionals can achieve the desired results. Regarding the flexibility variable, it is included during the programs directed to older adults but it is not discussed as a relevant variable. Flexibility is not treated in a specific way, obviating the importance of the pelvic musculature in the disturbance of the gait and therefore of the risk of falling. Researchers should avoid describing a superficial and uncritical methodology in order to design specific and valid aquatic programs for reducing the most modifiable predisposing risk factors of falls within the elderly.