Osteoarthritis (OA) is the most common joint disorder in the United States. The most common cause of disability in the United States [1], the overall prevalence of OA is estimated at 26.9 million adults greater than 25 years of age [2] and continues to increase [3]. One study estimated the lifetime risk of developing symptomatic knee OA to be 44.7%, increasing if one has a history of a knee injury (56.8%) or obesity (66%) [4]. Other studies estimate prevalence of radiographically apparent OA in adults over age 45 to be 19.2%-27.8% [5]. Knee pain alone led to 36 million ambulatory care visits in 2008 [3].

In addition to direct health care expenditures related to diagnosis and treatment, knee OA is an important confounding morbidity in many other chronic conditions such as cardiovascular disease, diabetes, and obesity. This is primarily driven by the limited ability to exercise secondary to pain and loss of function. The American College of Sports Medicine (ACSM) recommends healthy adults ages 18-65 participate in moderate-intensity aerobic physical activity for a minimum of 30 minutes/session, 5 days per week or vigorous-intensity aerobic physical activity for a minimum of 20 minutes/session 3 days per week in order to promote and maintain health [6,7]. The benefit of exercise in all individuals is well supported [8-10], including those with knee OA [11-15]. However, the best method of reaching current ACSM goals in patients with knee OA is not known. A Cochrane review on aquatic exercise provided gold level evidence that aquatic exercise likely reduces pain and increases function over 3 months; however, the review stated that more research was needed to understand which type of aquatic exercise, how often, and for how long might be beneficial [16].

The purpose of this study is to compare the effect of three exercise modalities (underwater treadmill, land treadmill, and upright cycling) on patient reported symptoms of knee OA while exercising at moderate intensity exercise for a goal of 30 minutes three times a week.

Design

In this randomized, prospective study we evaluated the knee symptoms of a participant with knee OA when exercising at the current American College of Sports Medicine recommendations of 30 minutes of moderate intensity exercise during underwater treadmill walking, land based treadmill walking, and cycling.

Enrollment and randomization

Upon obtaining University Institutional Review Board (IRB) approval, participants were recruited via radio advertisements, newspaper ads, and fliers posted in local clinics. Participants were screened via telephone interview to meet eligibility criteria:

Inclusion criteria: Ages 30-70 and knee pain with radiographically confirmed knee OA.

Exclusion criteria: Medical conditions contraindicating moderate aerobic exercise (as determined through prescreening questions, i.e., heart condition, asthma, history of stroke), inability to exercise via treadmill or exercise cycle, contraindication to radiography, pregnancy, history of recent joint injection (steroid, hyaluronic acid, etc.) within 6 weeks of study, history of previous joint arthroplasty, history of inflammatory joint disease, and inability to sign informed consent.

After obtaining informed consent, participants underwent bilateral AP standing knee radiographs to confirm the presence and degree of knee OA. X-rays were graded using the Kellgren and Lawrence (K/L) scale for knee OA severity [17]. Subjects with K/L level ≥ 1 were enrolled.

Each eligible participant was then randomized and prospectively assigned into one of three exercise groups: underwater treadmill (Hydroworx 2000 series pool), land based treadmill (True CS 6.0 or Nautilus NTR 700), or upright exercise cycle (Schwinn Windsprint), by drawing a card, with their respective study arm named, out of a bucket.

Exercise sessions

Exercise sessions were completed three times weekly for 8 weeks. Exercise sessions were directly monitored by a member of the research team with licensed medical professionals in close proximity. Automated external defibrillators were present in the Department of Physical Therapy and ready for use if needed. Each participant participated in three monitored exercise sessions per week and completed between 10 to 40 minutes of exercise with a goal of achieving 30 minutes of moderate aerobic exercise. While recognizing the ACSM goal of 30 minutes moderate exercise per session 5 day/week, limiting exercise to 3 days per week was chosen to avoid increasing exercise in the study participants too quickly. The exercise progression is presented in Table 1. All participants participated in no more than routine activities only for the 8-week study period outside of the study protocol (no strength training, cardio training, etc.). Participants were compensated $100 for completion of this research. Participants who failed to complete the study were compensated in proportion to their duration of participation. Participants were progressed according to their symptoms and cardiovascular endurance.

Table 1: Exercise Progression: Duration of time spent exercising per session. All participants exercised three days per week for a previously determined amount of time. Participants gradually increased exercise duration over first 3 weeks of study to achieve target METs (3-6) and RPE (4-6) in “Training” period of exercise session by Week 4 (Session 10).

Exercise intensity

The goal of the exercise progression was to achieve 30 minutes of moderate aerobic exercise by week 4 of the program. Participants progressively increased exercise time over the first 3 weeks; participants who were not inhibited by pain or increased symptoms reached a training level of up to 5 minutes of warm-up, 30 minutes of moderate aerobic exercise, and 5 minutes of cool down for a total of 40 minutes of exercise 3 times per week for the final 5 weeks of the program (Table 1).

Moderate aerobic exercise has been objectively defined as activity that generates energy expenditure of 3.0 to 6.0 Metabolic Equivalents (METs) [6]. In this study, moderate aerobic exercise was also defined by a participant’s subjective Rate of Perceived Exertion (RPE) of 4-6 on a 10-point scale. For the purpose of this study, METs and RPE were used as guidelines to determine exercise intensity. A participant may have an RPE of 4-6 with METs outside the 3.0 to 6.0 range. Participants were permitted (vs. encouraged) to exercise to tolerance (RPE) even if MET intensity was outside the preferred range due to possible cardiovascular limitations or high exercise tolerance. From a conditioning standpoint, the goal was to have participants capable of exercising at a moderate aerobic exercise level (METs) by week 4 of the study through a gradual progression over the first 3 weeks. Modality-specific exercise instructions and measurements of associated Metabolic Equivalents (METs) were determined based upon previous studies and presented in Table 2 [18].

Table 2: Exercise progression per modality. Methods for completing exercise sessions for each modality. Baseline METs criteria for exercise modalities based upon previous studies [7,18].

Participants in each group were asked whether or not they have unusual/persistent fatigue, increased weakness, stiffness, or excessive joint swelling/pain lasting for >24 hours after their previous exercise session. If a participant answered “yes” to any of these questions, they were not permitted to exercise that day [19]. If they were unable to exercise for 2 consecutive sessions, they were asked to see their primary care provider prior to continuing with the research study. Subjects were asked to discontinue exercise by the supervising health care provider during a session if they developed any type of musculoskeletal disorder that prohibited their ability to exercise. In addition, any symptoms of chest pain or excessive shortness of breath prompted discontinuation of exercise.

Data collection

Each participant completed three health status measurement questionnaires: the Western Ontario and McMaster Universities (WOMAC) Index of Osteoarthritis, the Knee Osteoarthritis Outcome Score (KOOS), and the SF12 Health Survey based on their more symptomatic knee. Questionnaires were completed at the time of randomization and then weekly. Reviewers and the statistician were blinded to subject group assignment. An intention to treat paradigm was utilized. Descriptive statistics were constructed for changes in the WOMAC and KOOS subscales in the form of means and standard deviations. The three groups were compared via a one-way analysis of variance (ANOVA). Pairwise comparisons of the three groups were performed via the ANOVA standard error and a Bonferroni correction to the significance level (0.05/3=0.01667).

Demographics and participation

Seventy-four adults ages 30 to 70 were recruited over 9 months to participate in this study. Six were deemed ineligible, as their radiographs did not demonstrate OA. Seven subsequently withdrew prior to initiating exercise. Sixty-one participants initiated the exercise program, randomized into one of three exercise modalities (Figure 1). The basic demographics of these cohorts are presented in Table 3. Of these 61 participants, 60 (98%) demonstrated OA in their knees bilaterally, with an average K/L score of 2.17. Ultimately, 42 participants (70%) completed the 8-week program. Study completion was highest for the water treadmill cohort (80%) as compared to the land treadmill (62%) and upright cycle (65%) cohorts, though this difference was not statistically significant (p=0.33). No adverse events occurred during the exercise sessions (Table 4).

Figure 1: Participant flow study diagram. Flow of participants from pre-screening and study enrollment through completion of the study.

Table 3: Baseline participant demographics by exercise cohort.

Table 4: Causes of study withdrawal: Participant-provided reasons for study withdrawal.

One individual in the land treadmill cohort withdrew due to severe gastroenteritis and was subsequently re-enrolled from the start of the exercise progression, bringing their cohort total to 21 participants. Causes of program withdrawal by exercise modality are detailed in Table 4. Participant withdrawal due to pain did not correlate with K/L scores. Enrolled participants based their survey responses on the more symptomatic knee at baseline and no participant in this study reported increasing pain in the opposite knee during this study.

Outcome measures

All participants were encouraged to exercise at a “moderate” level, defined as 3.0-6.0 METs and a subjective Rate of Perceived Exertion (RPE) of 4-6, deferring to RPE if both ratings could not be simultaneously satisfied. Overall, all participants exercised to a moderate level as defined by RPE; however, 4 participants (3 cycle, 1 land treadmill) failed to reach a moderate exercise level as defined by METs. Similarly, 5 individuals (2 cycle, 3 water treadmill) exercised above a standard METs level in order to reach a moderate level on RPE.

In comparing weekly survey results, all cohorts demonstrated an improvement in WOMAC scores; however, there was no statistically significant difference noted between exercise groups. Baseline WOMAC and change in score (95% confidence interval) at completion of the exercise program was 74.8 and -46.4 (-65.3, -27.4) for the water treadmill; 60.2 and -31.2 (-50.7, -11.7) for the land treadmill; 71.9 and -30.3 (-50.4, -10.2) for the upright cycle (Table 5). Using greater than 25% improvement in WOMAC as a threshold for clinically significant change, 80% of the water treadmill group improved by the conclusion of their participation in the study (either at point of drop-out or at study completion), compared with 60% cycle and 62% land treadmill.

Table 5: KOOS and WOMAC survey results: Statistically significant P-values presented in bold. WT=Water Treadmill, LT=Land Treadmill.

Likewise, an improvement in KOOS scores was observed over baseline for each exercise modality, but none exhibited statistically significant improvement in comparison to other modalities. Change from baseline KOOS score (95% confidence interval) at the completion of the study was -13.05 (-18.98, -7.11) for the water treadmill, -7.26 (-13.60, -0.92) for the land treadmill, and -9.71 (-16.13, -3.28) for the upright cycle (Table 5). Neither physical nor mental component summary scales demonstrated statistical significance between baseline and end of study values (Table 6).

Table 6: SF-12 results PCS=Physical composite score MCS=Mental composite score.

This study showed that an 8-week exercise program improved symptoms in participants with knee OA. Mean changes in KOOS scores were -13.05, -7.26, and -9.71 for the water treadmill, land treadmill, and upright cycle respectively, indicative of improvements in OA symptoms, pain, stiffness, and daily functionality. These results are consistent with other studies [16,20], which found that exercise is beneficial in alleviating symptoms of osteoarthritis. However, there was no statistically discernable difference noted based on training device

Though other studies have compared aquatic- versus landbased exercise for patients with knee OA, this study differed from those studies both in types of exercises performed and in outcomes measured. In one study [21], participants performed both resistance and aerobic exercises either in a pool or on land, and researchers evaluated the changes in knee range of motion, thigh girth, subjective pain, and time for a 1-mile walk. Another study evaluated the effects of land- and aquatic-based walking on gait kinematics and overall pain levels with movement [22]. Lastly, a third study studied cohorts of land- and water-based participants who participated in 60-minute group exercise sessions that focused on both resistance and aerobic methods of full-body conditioning over a 12-week period. Researchers primarily evaluated changes in pain levels in these two groups, as well as changes in activities of daily life, knee range of motion, and ability to perform the six-minute walk test [23]. The results of these studies were mixed; two studies found that subjective pain levels were lower in aquatic exercise groups when compared with land cohorts, while one showed no significant difference between land and aquatic groups. Both studies that demonstrated a documented difference in subjective pain scales utilized a continuous visual analog scale, while the one that did not show a statistically significant change utilized the KOOS score. None of these studies investigated the relative benefit of the exercise cycle in comparison to the water- and land-based treadmill.

Our study evaluated individuals as they participated in moderate aerobic exercise as recommended by ACSM guidelines, comparing the outcomes between land-based treadmill walking, water-based treadmill walking, and cycling. Evaluation was based on outcomes measurements involving exercise ability, effects upon markers such as pain, stiffness, and functions of daily living, and participants’ ability to sustain an exercise regimen for 8 weeks.

Study limitations

This study had several limitations. First, there was no control group to with which to compare the effects. Participants’ improvement was compared against their baseline survey findings, but may have been more markedly differentiated when compared to a non-exercising control group. Additionally, the conclusion that these three separate means of exercise were not statistically different may be Beta (Type II) error. Conversely, this conclusion may be accurate with no statistically significant difference. Increasing the sample size and the duration of the study may aid in the differentiation of the exercise modality and therefore yield a more statistically significant result. The study had a sizable dropout rate (30%). This group was not found to be statistically different from those who completed the 8-week program in terms of gender, BMI, and K/L score. While this attenuation rate may speak to the difficulty in encouraging individuals with OA to commit to regular exercise, it may have impacted the ability of this study to reach statistical significance among exercise modalities. Lastly, each subject participated in three monitored exercise sessions per week and completed between 10 to 40 minutes of exercise with a goal of achieving 30 minutes of moderate aerobic exercise. While recognizing the ACSM goal of 30 minutes moderate exercise per session 5 days/week, limiting exercise to 3 days per week was chosen to avoid increasing exercise in the study participants too quickly. Further progression towards the ACSM goal of 5 days per week should be considered for future study, although advancement of exercise is known to be highly variable across individuals [6].

Though it cannot be “cured”, the symptoms of OA can be effectively managed by regular moderate exercise, regardless of modality. Regardless of type, performance of 30 minutes of moderate exercise three times weekly is beneficial in the management of knee OA. However, it may be difficult to encourage individuals with OA to continue regular exercise to the point of noticeable benefit. Medical practitioners should encourage their patients with OA to reach these exercise goals set forth by the ACSM by whatever means they are able to sustain exercise.

Financial Disclosure/Conflict of Interest Statement

Grant support for this project from the Ben Franklin Fund and Hydroworx Corporation. No authors have any additional conflicts of interest to disclose.

IRB Statement

Institutional Review Board at Penn State Hershey Medical Center approved the study protocol.

Public Trials Registry

Trial was registered at Clinicaltrials.gov, study number NCT01359124.

1. U.S. Centers for Disease Control and Prevention (2010)Prevalence of Doctor-Diagnosed Arthritis and Arthritis-Attributable Activity Limitation — United States, 2007–2009.
2. Lawrence RC, Felson DT, Helmick CG, Arnold LM, Choi H, et al. (2008) Estimates of the prevalence of arthritis and other rheumatic conditions in the United States, Part II. Arthritis Rheum 58: 26-35.
3. Helmick CG, Felson DT, Lawrence RC, Gabriel S, Hirsch R, et al. (2008) Estimates of the prevalence of arthritis and other rheumatic conditions in the United States: Part I. Arthritis Rheum 58: 15-25.
4. Murphy L, Schwartz T, Helmick C, Renner JB, Tudor G, et al. (2008) Lifetime risk of symptomatic knee osteoarthritis. Arthritis Rheum 59: 1207-1213.
5. Zhang Y, Jordan JM (2008) Epidemiology of osteoarthritis. Rheum Dis Clin N Am 34: 515-529.
6. Haskell WL, Lee IM, Pate RR, Powell KE, Blair SN, et al. (2007) Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Med Sci Sports Exerc39: 1423-1434.
7. Whaley MH, Brubaker PH, Otto RM (2006) ACSM’s Guidelines for Exercise Testing and Prescription.7thedtn.Lippincott Williams and Wilkins, USA.
8. Dunn AL, Jewell JS (2010)The effect of exercise on mental health. Curr Sports Med Rep 9: 202-207.
9. Heckman GA, McKelvie RS (2008)Cardiovascular aging and exercise in healthy older adults. Clin J Sport Med18: 479-485.
10. Lee I (2010) Physical activity and cardiac protection. Curr Sports Med Rep9: 214-219.
11. Bosomworth NJ (2009) Exercise and knee osteoarthritis: benefit or hazard? Can Fam Physician 55: 871-878.
12. Dunlop D, Semanik P, Song J, Sharma L, Nevitt M, et al. (2010) Moving to maintain function in knee osteoarthritis: evidence from the osteoarthritis initiative. Arch Phys Med Rehab91: 714-721.
13. Ettinger W, Burns R, Messier SP, Applegate W, Rejeski WJ, et al. (1997) A randomized trial comparing aerobic exercise and resistance exercise with a health education program in older adults with knee osteoarthritis. JAMA 277: 25-31.
14. Fransen M, McConnell S, Harmer AR, Van der Esch M, Simic M, et al. (2008) Exercise for osteoarthritis of the knee. Cochrane Database Syst Rev 8: CD004376.
15. Lange AK, Vanwanseele B, Singh M (2008) Strength training for treatment of osteoarthritis of the knee: a systematic review. Arthritis Rheum59: 1488-1494.
16. Bartels EM, Lund H, Hagen KB, Dagfinrud H, Christensen R, et al. (2007) Aquatic exercise for the treatment of knee osteoarthritis. Cochrane Database Syst Rev 17: CD005523.
17. Kellgren JH, Lawrence JS (1957) Radiologic assessment of osteo-arthrosis. Ann Rheum Dis16: 494-502.
18. Ainsworth BE, Haskell WL, Whitt MC, Irwin ML, Swartz AM, et al. (2000) Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sports Exerc32: S498-S516.
19. McDermott A, Mernitz H (2006) Exercise and older patients: Prescribing guidelines. AmFam Physician 74:437-444.
20. Roddy E, Zhang W, Doherty M (2005) Aerobic walking or strengthening exercise for osteoarthritis of the knee? A systematic review. Ann Rheum Dis 64: 544-548.
21. Wyatt FB, Milam S, Manske RC, Deere R (2001) The effects of aquatic and traditional exercise programs on persons with knee osteoarthritis. J Strength Cond Res 15: 337-340.
22. Roper JA, Bressel E, Tillman MD (2013) Acute Aquatic Treadmill Exercise Improves Gait and Pain in People With Knee Osteoarthritis. Arch Phys Med Rehabil 94: 419-425.
23. Wang TJ, Lee SC, Liang SY, Tung HH, Wu SF, et al. (2011) Comparing the efficacy of aquatic exercises and land-based exercises for patients with knee osteoarthritis. J ClinNurs 20: 2609-2622.