Fall Risk and Utilization of Balance Training for Adults with Symptomatic Knee Osteoarthritis: Secondary Analysis from a Randomized Clinical Trial

Abstract

Title:

Fall Risk and Utilization of Balance Training for Adults with Symptomatic Knee Osteoarthritis: Secondary Data Analysis from a Randomized Clinical Trial

Background and Purpose:

Knee osteoarthritis (KOA) is a common disease that hinders activity participation in older adults. Associated symptoms and physiological changes can increase risk of falling in individuals with KOA. Balance training can decrease fall risk in older adults. There is limited evidence regarding utilization of balance training in physical therapy (PT) for this population. This secondary data analysis investigated the proportion of participants at high fall risk in the PhysicAl THerapy vs. INternet-based Exercise Training for Patients with Osteoarthritis (PATH-IN) study and the frequency with which balance training was utilized as an intervention in PT.

Methods:

PATH-IN study participants (N=344) performed the Four-Stage Balance Test and the Timed Up and Go Test (TUG) during baseline assessment. Participants were randomly allocated to PT, an internet-based exercise program, or a control group. Participants were classified as being at high fall risk if they did not progress to the single leg stance (SLS) during the Four-Stage Balance Test, were unable to maintain SLS for 5 seconds, or took longer than 13.5 seconds to complete the TUG. The proportion of participants at high fall risk was calculated for all participants and separately for those allocated to PT. Additionally, PT notes were coded for balance training and the frequency of balance training utilization was calculated.

Results and Discussion:

Upon enrollment, 35.5% (N=122) of all participants and 36.2% (N=50) of those allocated to PT were at high fall risk. Of participants allocated to PT with documentation available for coding (N=118), 35.5% (N=42) were at high fall risk. Balance training was provided to 62.7% (N=74) during at least one PT session. Of those classified as high fall risk, 33.3% (N=14) did not receive balance training.

Conclusions:

The finding of high fall risk in more than 1/3 of all participants with KOA is consistent with previous reports of a higher risk of falling in this population. Many PT participants did receive some balance training, however 1/3 of participants at high fall risk did not. Balance training for individuals with KOA at high fall risk may be underutilized.

INTRODUCTION

Knee osteoarthritis (KOA) is a degenerative joint disease that involves the breakdown of cartilage and bone within the knee joint, which can result in significant pain, swelling, and stiffness.¹ Damage to joint cartilage increases joint laxity, allowing excessive motion within the knee.² This deterioration in joint integrity can result in a cycle of pain, physical impairments, and disability in individuals with KOA.² The progressive deterioration within the joint capsule may be further exacerbated during the aging process or after injury to the knee joint.²

KOA is a highly common joint disorder in the United States, with a prevalence of 37.4% of participants age 60 years or older in a national survey.³ Risk factors associated with the incidence of KOA include age over 50 years, female sex, obesity (body mass index [BMI] over 30 kg/m²), and prior trauma to the knee.⁴ Due to the aging population and rising obesity rates in the United States, the prevalence of individuals affected by this disease is likely to continue to rise.⁵

Reduced muscular strength and loss of physical function are common impairments found in individuals with KOA.⁶ In addition, individuals with KOA frequently experience a loss of proprioception, potentially due to articular cartilage damage and impaired mechanoreceptors.⁷-⁹ These deficits contribute to decreased neuromuscular control, which can decrease an individual’s ability to safely perform functional activities and increase risk for falling.⁶ Doré et al¹⁰ reported participants with symptomatic KOA had a 39% higher odds of future falls. Tsonga et al⁶ reported a frequency of falls in older adults with KOA to be 63.2% over the course of one year. Additionally, individuals who report a history of falls with a diagnosis of KOA are more likely to have a poorer health-related quality of life compared to individuals with a history of falls or KOA alone.¹¹ The rapidly aging population and increasing prevalence of arthritis will likely compound this public health problem.

Due to the increased risk for falling in this population, balance training and fall prevention programs may be important management strategies for individuals with symptomatic KOA. The Centers for Disease Control¹² and the American and British Geriatrics Societies Clinical Practice Guidelines¹³ describe balance exercises as effective and essential components in a falls prevention program. Several studies have demonstrated the additive benefit of balance training to standard exercise programs for individuals with KOA.⁸,¹⁴−¹⁶ Recently, Takacs et al¹⁶ reported decreased pain and fear of movement, and improved physical function in individuals with KOA following a 10-week dynamic balance training program. Other trials have reported positive effects of balance training on outcomes including the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) total and subscale scores, physical function tests, global rating of change, and proprioceptive function in individuals with KOA.⁸,⁹,¹⁵-¹⁷ Furthermore, balance training interventions have decreased fear of movement and depression symptoms in this population.¹⁵,¹⁶ Despite the increased risk of falling for individuals with KOA and the benefit of balance training in older adults, information is lacking about utilization of this intervention as a component of physical therapy (PT) for this population.¹⁸,¹⁹

The aim of this secondary data analysis was to: (1) determine the proportion of participants in the PhysicAl THerapy vs. INternet-Based Exercise Training for Patients with Knee Osteoarthritis (PATH-IN) study at high risk for falling, and (2) the frequency with which balance training was included in the PT intervention relative to fall risk. In addition, this analysis summarizes the interventions utilized for balance training by the physical therapists. Assessment of current PT practice patterns for individuals with KOA at risk for falling may help physical therapists recognize the potential to integrate balance training into a comprehensive plan of care for individuals with KOA, and potentially reduce risk for falling in this population.

METHODS

Participants

This secondary data analysis included participants from a randomized control trial of adults with symptomatic KOA.²⁰ Participants (N=350) had a prior diagnosis of KOA and current symptoms (pain, aching, or stiffness on most days of the week) in at least one knee. Participants were excluded from the PATH-IN trial if they had a falls history deemed by a physical therapist investigator to impose risk for injury with participation in an independent home exercise program. Participants were randomly allocated to PT, an internet-based exercise program, or a control group in a 2:2:1 ratio. The sample for this study included all participants who had data to assess fall risk, as described below, prior to randomization (N=344). PT intervention analysis included participants allocated to PT with complete documentation of the interventions provided during sessions, also described below, (N=118). The PATH-IN research trial was approved by the Institutional Review Boards (IRB) of the University of North Carolina at Chapel Hill and Duke University Medical Center.

Measures and Fall Risk Determination

All participants who consented to participate in the PATH-IN trial provided demographic and personal information via self-report (age, college education, sex, race, marital status, body mass index, and duration of joint pain). In addition, prior to randomization, they completed a battery of self-report questionnaires and physical function tests conducted by trained examiners. Self-report questionnaires included the WOMAC, a widely used health status questionnaire which assesses pain, stiffness, and physical function in individuals with KOA.²¹ Total WOMAC scores (0–96) were determined by the summation of scores for the pain (0–20), stiffness (0–8), and physical function (0–68) subscales.²¹ Higher WOMAC total scores indicate worse pain, stiffness, and physical function limitations.²¹ Physical function measures included the Four-Stage Balance Test²² and the Timed Up and Go Test (TUG)²³. The trained examiners were blinded to group assignment of participants.

An evidence-based classification of high fall risk was based on previously established cut-off scores for the single leg stance (SLS) and the TUG. The SLS was the final component of the Four-Stage Balance Test, which assesses static balance in four progressively challenging positions held for a maximum of 10 seconds each.²² Vellas et al²⁴ reported an inability to perform SLS for 5 seconds is predictive for injurious falls in older adults. The TUG is a useful outcome measure to assess functional mobility and dynamic balance.²³ Shumway-Cook et al²⁶ reported that taking longer than 13.5 seconds to complete the TUG can accurately predict falls in community-dwelling older adults. This cut-off for the TUG was confirmed by Zasadzka et al²⁵ as an indicator for falls in older adults with lower extremity osteoarthritis. Therefore, participants were classified as high falls risk if they did not progress to the SLS during the Four-Stage Balance Test, were unable to maintain SLS for 5 seconds, or took longer than 13.5 seconds to complete the TUG. All other participants were classified as low fall risk.

PT Intervention

Prior to participant enrollment, PATH-IN study investigators met with the physical therapists who would deliver PT to participants to provide orientation to the project, as well as training in the PT intervention and use of the study database. The physical therapists received a handout summarizing evidence-based interventions¹⁸,¹⁹ to guide the content of the PT sessions (Appendix) and were instructed to use clinical judgment to tailor PT sessions to the needs of each participant. During the first PT session, the physical therapists completed a standardized, electronic evaluation form, along with documentation of any treatment provided during that visit. The electronic evaluation form included pre-populated data from the baseline assessments completed prior to randomization. Pre-populated data included participants’ TUG score, as well as semi-tandem, tandem, and SLS times. The evaluation form indicated that a TUG score of ≥13.5 seconds may predict risk for falling. During follow-up visits, physical therapists documented specific exercises and other interventions provided during sessions. Participants received up to 8 one-hour sessions based on the physical therapist’s assessment of the need for skilled therapy and participant willingness to attend sessions.

APPENDIX.

PATH-IN guidance for physical therapy visits²¹

1. Programs, both in the clinic and at home, should be comprehensive and functional, focusing on core and lower body function, but can be tailored to meet the functional abilities, needs and deficits of each participant.

2. Each visit should emphasize therapeutic exercise and include muscle strengthening, stretching/flexibility/range of motion, and aerobic exercise.

3. Education on activity pacing, joint protection and pain management

4. A home program should be recommended during the 1st visit and should be progressed over the course of treatment.

5. Home programs should emphasize the following:

   1. Strengthening Exercises

      1. Recommend performing strengthening exercises 2–3 times per week
      2. Include functional exercises, such as gait or stair training and neuromuscular education
   2. Stretching/flexibility/range of motion Exercises

      1. Recommend performing range of motion exercises daily
   3. Aerobic Exercises

      1. Promote “lifestyle” physical activity
      2. Encourage moderate intensity exercise
      3. Episodes of activity should last at least 10 minutes, if the participant is able
      4. Episodes should be spread out throughout the week with a long-term goal of working up to a total of 150 minutes of activity per week
      5. Aerobic exercise can be weight-bearing, reduced weight-bearing or non-weight- bearing.

6. Modalities for pain management can be included during the clinic visit and as part of the home program. Modalities should be used conservatively, taking no more than 25% of the time of each clinic visit.

7. If appropriate, manual therapy can be provided during the clinic visit.

8. Shoes should be assessed during the 1st visit, and shoe recommendations should be provided, if appropriate.

9. If limb length inequality or frontal plane knee malalignment is suspected, treatment with shoe lifts or shoe wedges, respectively, should be attempted.

After all PT sessions were completed, a chart review was completed to identify use of balance training. Progress notes in the study database were manually reviewed by the study authors, and each visit was coded as including balance training or not. If an exercise listed was unfamiliar, the physical therapist who provided the intervention was contacted for further description and purpose of the exercise to allow for proper coding. Interventions were coded as balance training if they included narrowing the base of support, movement of the center of mass, reaching with an upper or lower extremity to the limit of stability, or overcoming external perturbations. Interventions that focused on improving strength, flexibility, or aerobic capacity were not coded as balance training. A list was compiled of interventions coded as balance training (Table 1).

Table 1.

Interventions Coded as Balance Training Activities.

Balance Activities
Standing with Feet Together
Lateral Weight Shifting
Anterior and Posterior Weight Shifting
Semi Tandem Stance
Tandem Stance
Marching in Place
Unilateral Stance
Lateral Walking
Backwards Walking
Standing on a compliant surface (foam, trampoline, BOSU ball)
Closed-chain Exercises on compliant Surface (mini-squats, heel raises)
Cone Step Overs (forward or lateral)
Step Up onto compliant surface
Figure Eight Walking
Grape Vine Walking
Walking with Head Movements
Tandem Walking
Wii Balance Games
Rockerboard/Slantboard/Wobbleboard Activities
Throwing/Catching/Reaching for Objects on compliant Surface
Throwing/Catching/Reaching for Objects with smaller Base of Support (semi-tandem, tandem, SLS*)
Agility Ladder Exercises
SLS* with contralateral lower extremity reaches in multiple directions

^*SLS=single leg stance, also referred to as unilateral stance

Data Analysis

Data analysis included (1) the frequency and percentage of PATH-IN participants at high risk for falling at baseline (total sample and for PT group separately) and (2) the frequency and percentage of participants in the PT group for whom balance training was utilized during at least one visit relative to fall risk. Descriptive statistics were computed for total sample and for PT group separately, overall and by risk for falling status. Means and standard deviations were calculated for continuous variables and percentages for categorical variables. Statistical significance between groups was not determined because this work was intended to be a descriptive analysis of a subpopulation of the full trial, and the study was not powered for testing these statistical hypotheses.

RESULTS

The mean age of participants who completed assessments for fall risk (N=344) was 65.2 years (SD = 11.2). The average WOMAC Total Score was 32.1 (SD = 17.8). Individuals who were identified as being at high fall risk (N=122) tended to be older, less likely to have a college degree, more likely to be female, of nonwhite race, not be working, not be married, have a higher BMI and have a higher WOMAC total score (Table 2).

Table 2.

Participant Characteristics of Total Sample Who Completed Falls Assessment Measures, Overall and by Falls Risk Status.

Characteristic	Total Sample (N=344)	Low Falls Risk (N=222)	High Falls Risk* (N=122)
Age at baseline, years	65.2 (11.2)	63.6 (11.1)	68.1 (10.8)
College Education, %	59.3	71.2	37.7
Female Sex, %	71.8	67.6	79.5
Non-white Race, %	26.2	23.4	31.4
Working, %	40.7	45.9	31.1
Married, %	61.9	69.4	48.4
BMI, kg/m2	31.4 (8.0)	29.5 (6.6)	34.8 (9.1)
Pain Duration, years	13.2 (11.7)	13.1 (12.4)	13.3 (10.4)
WOMAC Total Score, units	32.1 (17.8)	27.8 (16.1)	40 (18.1)

Values are mean ± standard deviation unless indicated otherwise. BMI=Body Mass Index, WOMAC= Western Ontario & McMaster Universities Arthritis Index.

^*High falls risk is single leg stance <5 seconds or Timed Up and Go Test ≥ 13.5 seconds; low falls risk includes all others

Upon enrollment, 35.5% (N=122) of 344 participants in the PATH-IN trial, including those allocated to PT, were classified as being at high fall risk. Of the subpopulation of participants in the PT group, 36.2% (N=50) of the 138 participants were classified as being at high fall risk.

Of the 118 participants included in the PT group that had complete documentation of interventions provided to enable coding for balance training, 64.4% (N=76) were at low fall risk and 35.5% (N=42) were at high fall risk (Table 3). Regardless of fall risk, 62.7% (N=74) received balance training during at least one visit. Twenty-eight of the 42 PT group participants classified as being at high fall risk received balance training during at least one visit (66.7%). Exactly one-third of the participants who were classified as high fall risk in the PT group did not receive balance training.

Table 3.

Frequency and Percentage to Whom Balance Training was Provided for Participants Allocated to the PT Intervention.

Intervention	PT Participants (N=118)	Low Risk for Falling (N=76)	High Risk for Falling* (N=42)
Received balance training	74 (62.7%)	46 (60.5%)	28 (66.7%)
Did not receive balance training	44 (37.3%)	30 (39.5%)	14 (33.3%)

^*High risk for falling is single leg stance <5 seconds or Timed Up and Go Test ≥ 13.5 seconds; low fall risk includes all others

DISCUSSION

The findings from this secondary data analysis convey that a majority of participants at high risk for falling who participated in PT did receive balance training during at least one session (66.7%). Regardless of fall risk, use of some type of balance training in PT was relatively common (62.7%). Balance training in this study included a wide variety of activities, which illustrates an opportunity to adapt the intervention to match the specific impairments of an individual. Clinical practice guidelines for older adults recommend balance training to reduce fall risk¹²,¹³, but one-third of participants in the PT group identified as at high fall risk did not receive specific balance training.

The proportion of participants allocated to the PT group at high fall risk (36.2%) was consistent with the proportion of all PATH-IN participants at high fall risk (35.5%). Prior studies of fall risk among adults with knee OA have reported a range of 48% to 63.2%.⁶,²⁷ The occurrence of high fall risk reported in this study may be lower because participants were excluded from the PATH-IN trial if deemed by a physical therapist investigator to impose risk for injury with participation in an independent home exercise program based on prior fall history.²⁰ However, even with this restriction, a little over one third of participants were at high fall risk. This study confirms a high proportion of individuals with KOA are at an increased risk for falling, which warrants follow up to decrease this risk and associated consequences.

There are some limitations to the results of this study. This study assessed PT practice pattern during a clinical trial, therefore these practice patterns may not reflect typical clinical practice. Consistent with guidelines for conservative management of KOA,¹⁸,¹⁹ guidance for PT interventions placed greater emphasis on strength and aerobic exercise than balance training, which may have biased the interventions selected by the physical therapists. However, the physical therapists were instructed to use their clinical judgment and tailor interventions to individual needs and deficits. Another limitation is the potential for error in coding of balance training. Some interventions may improve more than one impairment, such as strength and balance; however, interventions were coded as balance training only if they met the criteria above. Therefore, interventions that could improve balance may not have been counted if they were incorporated into an exercise that did not meet the criteria or were perceived to target another impairment. This study utilized a limited number of physical therapists and PT clinics, limiting the generalizability of the findings. As previously mentioned, participants were excluded from the PATH-IN trial if they had a history of falls deemed by a study investigator to impose risk for potential injury with participation in an independent home exercise program. Therefore, the occurrence of fall risk reported in this study may not be fully representative of the population with KOA.

Despite multiple studies demonstrating benefit from balance training in this population,⁸,⁹,¹⁴–¹⁷ actual implementation in clinical settings remains unclear. A larger scale study is indicated to more broadly assess and understand PT intervention selection for individuals with KOA at risk for falling.

CONCLUSION

Consistent with previous findings, a substantial proportion of individuals with KOA in this trial were at increased risk for falling. The majority of participants allocated to PT received balance training during at least one visit. However, one-third of participants at high falls risk in the PT group did not perform specific balance training during any PT session. Balance training may be underutilized for those with KOA at risk for falling. Since balance training can reduce fall risk in community-dwelling older adults, further research is warranted on its effectiveness to reduce fall risk among individuals with OA, as well as best strategies for implementation.