+ MODEL Journal of Bodywork & Movement Therapies (2016) xx, 1e9 Available online at www.sciencedirect.com ScienceDirect journal homepage: www.elsevier.com/jbmt RANDOMIZED CONTROLLED TRIAL The effectiveness of Pilates on balance and falls in community dwelling older adults Sharon Josephs, PT, DPT, OCS a,*, Mary Lee Pratt, PT, DPT, MAA, OCS a,b, Emily Calk Meadows, DPT a, Stephanie Thurmond, PT, DPT, ScD, CFMT, COMT a, Amy Wagner, PT, DPT, GCS a a b School of Physical Therapy, University of the Incarnate Word, USA Pyramid Plaza Physical Therapy, USA Received 31 October 2015; received in revised form 16 January 2016; accepted 23 January 2016 KEYWORDS Balance; Falls; Pilates; Physical therapy; Older adults; Balance confidence Summary Purpose: The purpose of this study was to determine whether Pilates is more effective than traditional strength and balance exercises for improving balance measures, balance confidence and reducing falls in community dwelling older adults with fall risk Method: Thirty-one participants with fall risk were randomly assigned to the Pilates group (PG) or the traditional exercise group (TG) Both groups participated in 12 weeks of exercise, times/week for h Results: There was significant improvement in the Fullerton Advanced Balance Scale for both the PG (mean difference Z 6.31, p < 05) and the TG (mean difference Z 7.45, p Z 01) The PG also showed significant improvement in the Activities-Specific Balance Confidence Scale (mean difference Z 10.57, p Z 008) Conclusion: Both Pilates and traditional balance programs are effective at improving balance measures in community dwelling older adults with fall risk, with the Pilates group showing improved balance confidence ª 2016 Elsevier Ltd All rights reserved * Corresponding author Pilates and Physical Therapy Center of San Antonio, 211 Post Oak Way, San Antonio, TX, 78230, USA Tel.: þ1 210 410 0751 E-mail addresses: sharonjosephsPT@gmail.com, josephs@uiwtx.edu (S Josephs) http://dx.doi.org/10.1016/j.jbmt.2016.02.003 1360-8592/ª 2016 Elsevier Ltd All rights reserved Please cite this article in press as: Josephs, S., et al., The effectiveness of Pilates on balance and falls in community dwelling older adults, Journal of Bodywork & Movement Therapies (2016), http://dx.doi.org/10.1016/j.jbmt.2016.02.003 + MODEL Introduction Falls are a common problem in older adults There are significant consequences from falling including injury, decreased mobility and independence, increased health care costs and psychological concerns from fear of falling (Centers for Disease Control and Prevention, 2015a,b; Kannus et al., 2005; Powell and Myers, 1995) The Centers for Disease Control and Prevention (2015a,b) reported that in 2013 in the United States, 2.5 million nonfatal fall injuries in older adults were treated in the emergency department and over 734,000 of those were hospitalized The direct medical costs of falls in older adults, adjusted for inflation, were over $34 billion in the United States (Centers for Disease Control and Prevention, 2015a) Powell and Myers (1995) reported a study where 57% of older adults living in the community reported fear of falling and 30% reported avoidance of activities due to fear Avoidance of activities leads to further physical frailty and loss of independence As the population is aging, there is a need to develop effective fall prevention programs (Centers for Disease Control and Prevention, 2015a; Kannus et al., 2005) The 2012 Cochrane Review concluded that multiple component group exercise significantly reduced rate of falls (rate ratio 0.71, 95% CI, 0.63 to 0.82; 16 trials and 3622 participants) and risk of falling (risk ratio 0.85% 95% CI 0.76 to 0.96; 22 trials, 5333 participants) in older adults living in the community (Gillespie et al., 2012) A meta-analysis by Sherrington et al (2008) reported that the pooled estimate of the effect of exercise was that it reduced the rate of falling by 17% (rate ratio 0.83, 95% CI 0.75e0.91, 44 trials, 9603 participants) A meta-regression analysis showed that the greatest relative effects were seen when the balance portion of the program was challenging, the dose of exercise was greater than 50 h for the entire program, equating to twice a week for 25 weeks, and the program did not include a walking component The inclusion of these relatively important components improved the reduction in rate of falls to 42% (rate ratio 0.58, 95% CI 0.48e0.69) The optimal exercise prescription, however, has not yet been determined (Gillespie et al., 2012; Sherrington et al., 2008; Arnold et al., 2008) Also, patients need to continue exercise programs long term; otherwise the benefits disappear (Wolf et al., 2001) A problem discussed in the literature is the difficulty recruiting and retaining older patients in a moderate intensity exercise program (Means et al., 2005) Pilates is an exercise method that is well suited for the older adult The equipment is adaptable and the intensity can be modified to assist or resist depending on the level of the patient (Anderson and Spector, 2000) The potential benefits of Pilates include: improved posture, core and extremity strength, flexibility, motor control, balance and body awareness (Endelman, 2009) Research using Pilates’ equipment to improve balance in older adults, however, is still in its infancy stage, with only studies reported and none specifically on patients with fall risk (Kaesler et al., 2007; Siqueira Rodrigues et al., 2010; Bird et al., 2012; Bird and Fell, 2014) Kaesler et al (2007) reported a study with a small sample size of high functioning subjects, aged 66e71 and no control group and S Josephs et al showed an improvement in the Timed Up and Go (TUG) and balance sway measures following a Pilates-inspired exercise program, times per week for weeks Siqueira Rodrigues et al (2010) reported a randomized controlled trial of 52 women, aged 60e72, where the Pilates group (n Z 25) showed improvement in the Tinetti test for balance and the Latin America Development Group for Elderly score of personal autonomy compared to a no treatment group of age-matched controls (n Z 27) Their intervention was times per week for weeks Bird et al (2012) reported a randomized crossover design trial of 32 subjects over the age of 60 The Pilates exercises were performed for weeks They found an improvement in TUG, Four Square Standing test and balance sway measures pre to post Pilates for each group, but no difference between the groups They suggested that the lack of between group differences might be due to the small sample size and/or the crossover design Bird and Fell (2014) then did a follow-up study 12 months after the initial intervention Postural sway, dynamic balance and functional improvements seen after the initial Pilates training were maintained 12 months later in all participants, with increased benefits in participants that continued Pilates exercise Balance confidence refers to a person’s perception of their balance ability and is important to study as low balance confidence leads to avoidance of activities and physical frailty Kendrick et al (2014) in a Cochrane review of 24 studies and 1692 participants found a small to moderate reduction in fear of falling following an exercise intervention (standardized mean differences 0.37, 95% confidence interval 0.18 to 0.56) without increasing the frequency of falls Six of the 24 studies used the Activities-Specific Balance Confidence Scale (ABC), a questionnaire that measures balance confidence There are many limitations in the current balance literature Many traditional balance studies fail to describe their exercise protocols in enough detail to allow them to be reproduced (Arnold et al., 2008) Pilates studies have yet to investigate the effectiveness of Pilates exercise using Pilates equipment to reduce rates of falls, improve balance confidence, study patients with documented fall risk or compare Pilates to a traditional strength and balance program The literature indicates that this older population needs to continue the program long term (Wolf et al., 2001) However, balance exercise programs provided by physical therapists have difficulty meeting the current recommended exercise dosage under the current funding model (Shubert, 2011) Group based therapy is billed at a fraction of the rate of individual therapy, and Pilates is a form of exercise that patients can continue following discharge The purpose of this study was to investigate the effectiveness of Pilates group exercise versus traditional strength and balance group exercise for improving balance, reducing falls and improving balance confidence in community dwelling older adults with fall risk Methods and procedures This study was a single blind randomized controlled trial The study was granted approval by the University of the Incarnate Word Institutional Review Board Please cite this article in press as: Josephs, S., et al., The effectiveness of Pilates on balance and falls in community dwelling older adults, Journal of Bodywork & Movement Therapies (2016), http://dx.doi.org/10.1016/j.jbmt.2016.02.003 + MODEL Pilates effect on balance and falls in community dwelling older adults Subjects and screening procedure Thirty-nine subjects, 65 years of age and older living in the community, were recruited through local physicians in the area, seniors groups at churches and community centers, word of mouth and notices posted in the local libraries Potential subjects signed an informed consent form and underwent a screening evaluation by a blind student or volunteer assessor The assessors were trained in the assessment procedure and the order of testing was standardized The screening included a demographic questionnaire, fall history and medical history form, blood pressure (BP) and heart rate (HR), Timed up and Go (TUG), Fullerton Advanced Balance Scale (FAB) and Activities-Specific Balance Confidence Scale (ABC) A fall was defined as an unexpected event in which the participant came to rest on the ground, floor or lower level (Lamb et al., 2005) Inclusion criteria were: 65 years of age or older living in the community; impaired balance as defined by at least one of the following: a fall in the past year, TUG >13.5 s or FAB 25; and ability to follow instructions as assessed by the ability to complete the questionnaires without assistance Subjects were not screened for ability, such as use of an assistive device for walking, but only that they met the inclusion criteria of history of fall or meeting the cutoff for balance compromise with the TUG or FAB Exclusion criteria included: participation in a Pilates program within the last year; significant health problem that would keep the subject from participating; vestibular conditions and progressive neurological conditions Subjects were excluded if they were not community dwelling, e.g if they lived in an institution such as an assisted living facility Outcome measures The TUG is a widely used test of balance and has been found to be reliable and valid and normative values have been reported for community dwelling older adults (Steffen et al., 2002; Lin et al., 2004) A TUG >13.5 s is associated with fall risk (Shumway-Cook et al., 2000) The TUG is considered to be appropriate for older people who are frailer or who use walking aids (Lin et al., 2004) The test was performed by measuring the time it took the subject to stand up from an armchair, walk a distance of m at their usual pace with their usual assistive device, turn, walk back to the chair and sit down (Podsiadlo and Richardson, 1991) The FAB is a newer test and a more comprehensive assessment of the multiple dimensions of balance as identified in the systems theory of postural control (Rose et al., 2006; Sibley et al., 2015) The Berg balance scale is a more widely used balance test, but it has been criticized for its ceiling effect in community dwelling older adults (Pardasaney et al., 2012) The FAB includes more difficult static and dynamic balance tasks to make it less prone to ceiling effects and more sensitive to evaluate the effectiveness of an intervention conducted in a higher functioning group of patients (Rose et al., 2006) It has been shown to be reliable and valid, but normative values have not been reported (Rose et al., 2006) A score of less than or equal to 25 has been associated with fall risk (Hernandez and Rose, 2008) The FAB was performed as per the documented instructions (Rose, 2008), with one exception In the reactive postural control test where the assessor removes their hand support from the patient to test their postural reaction, the documented test does not inform the patient that the assessor will be removing their hand We chose to inform our patients that we would be removing our hand, but not when in order to maintain the patients’ trust and for patient safety The patient also completed the ABC, a 16-item questionnaire that is a psychological measure of balance confidence in performing specific activities (Powell and Myers, 1995) Each item is scored from 0% (no confidence) to 100% (full confidence in the ability to perform the activity without losing balance) The ABC has been shown to be reliable and valid (Powell and Myers, 1995) Scores on the ABC showed an association with tandem stance time (Cyarto et al., 2008), one-leg stance time (Cyarto et al., 2008), TUG (Cyarto et al., 2008; Hatch et al., 2003) and Berg Balance Scale (Hatch et al., 2003) Procedures Patients that met the inclusion and exclusion criteria were randomly assigned to either the Pilates exercise group or the traditional exercise group using a randomization table Groups were separated based on the participant group allocation Both groups exercised 60 min, times per week for 12 weeks Participants that missed a session due to illness or scheduling conflict made up the missed session, such that each subject completed 24 sessions of exercise The groups were a maximum of participants The exercise classes were taught by of physical therapists, both Board Certified Clinical Specialists in Orthopaedics and comprehensively certified Pilates instructors The therapist supervising a group continued with that group for the entire 12 weeks as much as possible, with the other physical therapist substituting on occasion Both therapists taught both Pilates’ groups and traditional groups over the course of the study Additional student volunteers assisted with spotting for safety purposes The Pilates group performed exercises as per Appendix The Pilates program utilized the Reformer, Cadillac and Chair apparatus Exercises were individually made more challenging by altering spring tension, reducing base of support, adding complexity to the exercise and altering surface stability Each exercise was performed for 10 repetitions The traditional group performed exercises as per Appendix Elastic resistance bands, ankle weights, foam balance pads, boxes of varying heights and half foam rollers were props that were used with the traditional group Repetitions of the resistance exercises were progressed individually When the patient could perform 20 repetitions, the resistance was increased: elastic resistance bands were progressed to the next color; ankle weights were progressed by 0.5 pounds; box heights were progressed by inches All exercises for both groups were performed in a challenging but pain-free way Modifications were made to exercises in consideration of a patient with spine or peripheral joint pain by reducing the range of motion performed during an exercise, reducing the resistance or changing position Occasionally, an exercise was discontinued if a pain-free modification could not be Please cite this article in press as: Josephs, S., et al., The effectiveness of Pilates on balance and falls in community dwelling older adults, Journal of Bodywork & Movement Therapies (2016), http://dx.doi.org/10.1016/j.jbmt.2016.02.003 + MODEL S Josephs et al Figure Summary of participants found All patients were asked to work at level 12e14 on the Borg Perceived Exertion Scale, which rates exertion subjectively from to 20 (Borg, 1970) It has been shown to be a valid measure of exertion (Day et al., 2004; Eston and Evans, 2009) On non-program days and daily following discharge from the program, patients were asked to perform home exercises The same 15e20 home exercise program was given to both groups in order to minimize confounding effects of the home exercise program The home lower extremity strength exercises were drawn from both Pilates mat exercises and traditional physical therapy exercises The home program also included standing balance exercises See Appendix Patients were given a home exercise handout and a monthly calendar to record their home exercise participation This was done in anticipation of the patients performing their home exercises on non-program days, and continuing for weeks following discharge from the program; their exercise dose would then exceed the recommended 50 h The calendar was also used to record falls during the 1-year follow-up after discharge The follow-up assessment occurred following completion of the 12 weeks of exercise BP, HR, the TUG, FAB and ABC were re-tested by an assessor who was blinded to group assignment, using the standardized testing procedure Following completion of the program, the patients were telephoned once per month to obtain information on any falls that occurred during the previous month The patients were also asked about their current participation in all types of exercise (home program and other exercise such as walking and group fitness classes) This data is currently being collected and will be reported in a follow-up study Data analysis All data was analyzed at the 0.05 alpha level using Microsoft Excel and SPSS version 19 software The pre-test to posttest within group, between group, and interactions data were analyzed with paired t-test, independent t-test and  factorial ANOVA respectively Results Thirty-one patients met the inclusion criteria and subjects were excluded due to lack of fall risk as defined above Seven patients dropped out of the study and the remaining 24 patients completed all 24 sessions of exercise and the follow-up evaluation See Fig Of the 24 participants, 18 were female and were male The age range was 65e85, with an average of 75.6 in the Pilates group and 74.5 in the traditional group Ten out of 13 patients in the Pilates group and out of 11 in the traditional group had fallen in the past year Three patients in each group met the inclusion criteria only because of the fall they had in the past year, while not meeting the fall risk criteria on the TUG and FAB All subjects were similar at Please cite this article in press as: Josephs, S., et al., The effectiveness of Pilates on balance and falls in community dwelling older adults, Journal of Bodywork & Movement Therapies (2016), http://dx.doi.org/10.1016/j.jbmt.2016.02.003 + MODEL Pilates effect on balance and falls in community dwelling older adults baseline for age (p Z 693), sex for males and females (p Z 094), education (p Z 812), number of falls (p Z 703) and co-morbidities (p Z 274) However, there were differences in number of males (38.46% of PG, 9.1% of TG) and amount of activity (77% physically active in PG, 45% physically active in TG) between groups at baseline Table presents the demographic data In the traditional group, there was a significant difference (mean difference Z 7.45, p Z 01) between the pre-test and post-test scores of the FAB In the Pilates group there was a significant difference in the pre-test and post-test scores of the FAB (mean difference Z 6.31, p < 05) and the ABC (mean difference Z 10.57, p Z 008) respectively Table presents the outcome data There were no significant between group differences on any of the selected tests A  factorial ANOVA showed no interaction between groups for pre and post intervention scores, and type of intervention Follow-up data on number of falls is still being collected for future analysis Table Demographic characteristics of participants at baseline for both groups Mean Æ SD age, years (range) Number of males Number of females Mean height, inches (range) Mean weight, pounds (range) Education (# of patients) Less than high school High school graduate Some college/technical school College graduate Graduate school Living situation (# of patients) Alone With spouse With other Employment # of participants working Retired Walks with an assistive device Yes No Exercise regularly Yes No Mean Æ SD number of falls (range) # that fell in the past year Mean Æ SD # of comorbidities (range) Mean Æ SD # of medications (range) Pilates group Traditional group 75.6 Æ 6.2 (66e85) 65.2 (58e72) 158 (107e230) 74.5 Æ 6.9 (65e83) 10 63.8 (58e70) 3 5 9 9 10 1.5 Æ 1.3 (0e4) 10 2.7 Æ 1.5 (0e5) 5.8 (0e13) 1.8 Æ 2.2 (0e7) 3.8 Æ 3.0 (1e10) 3.8 (1e9) 169 (101e235) Discussion Both the Pilates and the traditional group improved in their FAB scores, with neither group showing significant improvement over the other The traditional balance exercise literature indicates that effective fall prevention exercise programs include multiple components and challenging balance exercises, so both of the exercise programs in this study were designed to include core and lower extremity strength, flexibility and challenging balance exercises Both exercise groups were taught with an emphasis on whole body alignment and core stabilization, similar to Pilates’ principles of exercise Therefore, it is not surprising that the study found no between group differences The Pilates group showed improvement in balance confidence, as measured by the ABC, while the traditional group did not This is an interesting finding considering that the balance measure, FAB, improved in both groups, with neither group showing significantly more improvement than the other One possible explanation for this is that the patients in the Pilates group had to learn new skills on foreign equipment, and the learning of this novel skill lead to improved balance confidence Means et al (2005) discussed the problem of attrition in balance studies Their study reported a 17% attrition rate during the intervention and discussed the difficulty in recruiting and retaining older patients in a moderate intensity exercise program This study had a 19% attrition rate in the Pilates group: patients dropped out prior to starting any exercise and dropped out due to a family crisis There was a 27% attrition rate in the traditional group: of the patients dropped out following weeks or less of exercise, dropped out as he did not feel he was improving and dropped out due to a knee injury that happened while on vacation Studies have shown that the benefits of exercise quickly diminish once the program is completed (Wolf et al., 2001) Therefore, there is a need to develop programs that the patients can continue long-term as a wellness program Long-term commitment to exercise also means that patients need to enjoy the program It is the experience of the investigators that Pilates is an exercise method that older adults enjoy and may offer an option for patients to participate in a program that can be initiated as a physical therapy intervention and continued long term as a wellness program Shubert (2011) discussed the difficulty of providing the recommended 50-h dose of strength and balance exercises in physical therapy due to the current funding models Group based exercise is more cost-effective than one-on-one therapy as it is billed at a fraction of the rate However, group exercise presented a problem as challenging balance exercises also meant that there was a risk of falling during the exercise Additional student volunteer spotters were used to ensure the safety of the patients Group based programs are also social and this motivating factor may improve adherence compared to home programs (Phillips et al., 2004) The addition of home exercises has also been suggested as a method to try to reach the recommended dose (Martin et al., 2013) and home exercises were included for both groups on non-program days as well as following discharge from the program Please cite this article in press as: Josephs, S., et al., The effectiveness of Pilates on balance and falls in community dwelling older adults, Journal of Bodywork & Movement Therapies (2016), http://dx.doi.org/10.1016/j.jbmt.2016.02.003 + MODEL S Josephs et al Table Outcome variables at baseline and post intervention for both groups Test Pilates group pre-test average Æ SD Pilates group post-test average Æ SD Mean difference pvalue Traditional pre-test average Æ SD Traditional group post-test average Æ SD Mean difference pvalue TUG (seconds) FAB (0e40 points) ABC (0e100 points) 14.38 Æ 6.54 18.54 Æ 10.08 63.08 Æ 24.08 13.07 Æ 4.59 24.85 Æ 12.54 73.65 Æ 22.47 À1.31 6.31 10.57 NS