The 25 million Americans suffering from osteoarthritis (OA) represent a significant social and economic burden [
In the context of knee OA, pain and low muscle power (the product of force and velocity) lead to functional limitations such as reduced community mobility. Reduced activity due to these factors may lead to further weakness and altered cartilage nutrition, contributing to further joint pathology, impairments, functional limitations, and disability. In particular, velocity of muscle contraction declines with aging, leading to a more precipitous drop in muscular power than in strength [
In older adults without symptomatic knee OA, a land-based weighted vest program has been shown to be well tolerated and effective in improving lower limb muscle power and chair rise speed [
A solution to the problem of delivering power training to those with symptomatic knee OA may be to deliver it in an aquatic environment, as submersion in water is known to reduce knee symptoms [
Consequently, patients with symptomatic knee OA may be able to tolerate power training in an aquatic environment better than in a land-based environment. However, there have been mixed results regarding the effectiveness of exercise in an aquatic environment. A 12-week aquatic strength-training program in 55–75-year-old individuals with unilateral knee arthroplasty was well tolerated and significantly improved knee extensor and flexor power, thigh muscle cross-sectional areas, and stair climb and walking times compared with those randomized to control [
Thus, although evidence suggests that
This study enrolled adults age 50 and older, residing within commuting distance of our institution, with frequent knee symptoms (pain, aching, or stiffness) on most of the last 30 days, knee OA by ACR clinical criteria (using history and physical examination to determine the presence of pain in the knee plus three of the following: over 50 years of age, less than 30 minutes of morning stiffness, crepitus on active motion, bony tenderness, bony enlargement, and no palpable warmth of synovium) [
Mobility limitation inclusion criteria [
Sex | Decade | 400 m walk time (sec) |
---|---|---|
Men |
50’s | >250.2 |
60’s | >289.9 | |
70’s+ | >290.8 | |
| ||
Women | 50’s | >315.9 |
60’s | >305.2 | |
70’s+ | >292.5 |
Volunteers were asked “Do you have health problems that affect your walking or ability to exercise such as severe back pain, heart disease, diseases of the muscles or nerves, or problems with your eyesight that affect your walking?” and excluded if they reported causes for mobility limitation other than knee symptoms. No participants were unwilling to be in a 1.2-meter deep pool, or had a history of bilateral knee replacement, lower limb amputation, myocardial infarction or stroke in the past year, lower limb surgery in the last six months that affected walking ability, concurrent participation in another research study, or medical conditions that affected walking ability or ability to follow the protocol (e.g., Alzheimer’s or other type of dementia, multiple sclerosis, Parkinson’s disease, severe cardiovascular disease, congestive heart failure, severe dysrhythmias, severe emphysema, severe asthma, skin disease that would be adversely affected by aquatic exposure, inability to attend visits, or understand instructions). Those who met eligibility criteria other than mobility limitation were invited to a clinical visit, where volunteers participated in an institutional review board-approved informed consent process and then 400-meter walk time was assessed. Those who met inclusion criteria were scheduled for an initial pool visit and given a study description form for their usual physician to review and sign if the volunteer was medically safe to participate in the aquatic power training protocol.
The aquatic power training program involved two training sessions per week of one hour duration for six weeks in a 1.2-meter deep therapy pool, heated to approximately 34.5° to 36° Celsius. A maximum of three subjects were trained at one time by one of two exercise specialists certified in the aquatic power training protocol. The subjects were encouraged to perform a total of eight exercises (Table
Aquatic power training exercises.
Exercise | Sets | Repetitions |
---|---|---|
Step-ups on the pool stairs or underwater riser | 3 | 10 on each leg |
|
2 | |
Step-downs on pool stairs or underwater risers | 3 | 10 on each leg |
|
2 | |
Row with foam water dumbbells | 3 | 10 |
|
2 | |
Hip extension leg lifts | 3 | 10 on each leg |
|
2 | |
Hip abduction | 3 | 10 |
|
2 | |
Hip adduction | 3 | 10 |
|
2 | |
Plantar flexion heel raises | 3 | 10 |
Calf stretches—30 seconds each stretch | 2 | |
|
2 | |
Dorsiflexion toe raises | 3 | 10 |
The exercise specialist provided verbal and tactile cues both to ensure proper form for each exercise and to encourage rapid contractions for power training. Between each exercise set, subjects walked 2.4 meters forwards and backwards as quickly as possible in chest-deep water. The exact level of submersion in water depended on the subject’s height. At the conclusion of each training session, subjects were asked to complete 3–5 minutes of aquatic walking forwards and backwards as a cool-down before exiting the pool. Subjects’ Borg ratings of perceived exertion (RPE) on a scale of 6 to 20 were monitored after each set of repetitions. If a participant had reported an RPE ≥17, then the subject would have been asked to rest until the RPE went back to 13. However, no subject reported an RPE greater than 14 during the study. Specialists recorded subject attendance, number of repetitions of each exercise performed, and presence of knee pain prior to and following each session. Participants’ ratings of knee pain severity were assessed during training sessions for safety and to adjust the exercise intensity to avoid exacerbating symptoms.
Attempts to optimize compliance and retention included providing reminder phone calls as needed prior to appointments, compensation for attending study visits, and parking vouchers or bus passes at each study visit.
Tolerability (primary outcome measure) was assessed with ratings of knee pain and exertion, need for modification of the protocol, and attendance. Symptomatic and physical functional outcome measurements were conducted at baseline, following 6 weeks of training and again 6 weeks after completion of training to assess for durability of effects (12 weeks after baseline) as shown in Figure
Enrollment and progression of subjects and timing of study measurements.
Subjects were instructed to safely ascend a standard 8-stair flight (total vertical distance = 1.441 meters), with handrails on both sides as quickly as possible. If necessary for balance, the handrails could be used on either side. Timing was started when the subject initiated foot movement to begin stair ascent and was stopped when both feet arrived on the top (eighth) step. Time was recorded to the nearest 0.01 second. Times for two trials, attempted on the same day, were averaged. The reliability for this test has been reported to be excellent (ICC = 0.97) [
The long distance corridor walk (LDCW) is a measure of timed gait during an unassisted walk and is sensitive to changes in community mobility. The protocol was based on the protocol used in the Health, Aging, and Body Composition (Health ABC) Study and the Osteoarthritis Initiative [
Self-reported difficulty with physical activity was measured using the Late Life Function and Disability Instrument: Function Component (LLFDI, Boston University, Boston, MA) [
The knee injury and osteoarthritis outcome score (KOOS) is an extension of the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), the most commonly used outcome instrument for assessment of patient-relevant treatment effects in OA. This instrument has been found to be a reliable, and responsive measure in older adults with knee OA, and sensitive to changes in pain and knee-related activities of daily living and quality of life [
In addition to assessing for knee-specific pain that might relate to the exercise protocol, overall pain that might be modified by the aquatic environment was assessed with the SF-36, a well-validated tool that combines the dimensions of impairments and functional limitations [
Outcome measures were continuous. Tolerance of the intervention was assessed descriptively for pain and exertion ratings, modifications, and attendance. Tests of normality were performed, followed by construction of linear mixed models for repeated measures to test for changes in outcome measures across time points. Dunnett-Hsu’s method was used to adjust for multiple comparisons in comparing each of the followup measurements (Week 6 and Week 12) with the baseline measurement. Symptomatic and functional outcome measures were summarized with least squares means and standard errors (SE) as well as absolute ranges. Effect sizes were calculated as the mean difference in the parameter between time points divided by the standard deviation of the mean difference in that parameter. Statistical analyses were conducted using SAS 9.2 with an alpha level of 0.05.
Although designed to assess tolerability, this study also was initiated to obtain evidence regarding whether aquatic power training would improve muscle power. In a prior study of land-based power training [
Figure
Overall attendance at training visits was 92.2% (hours attended/hours scheduled), with 15 participants attending all 12 visits, the median number of visits attended (interquartile range 11-12 visits).
Twenty-one of 29 subjects completed the study per protocol. The remaining 8 required modification of the protocol for tolerance. Two shorter subjects performed the exercises on the underwater staircase due to feeling uncomfortable with the 1.2-meter pool depth. Other subjects who required a modification to the protocol reported Achilles tendon pain (1), knee pain (2), or muscle pain (1) when attempting the standard protocol or could not complete the exercises at a velocity judged by the trainer to be consistent with power training. In such cases, trainers reduced the number of repetitions and sets. However, one subject required an increase in repetitions due to insufficient intensity.
The ratings on the verbal analogue scale of knee pain prior to and following each aquatic training session were normally distributed. The mean ± SD pretraining and posttraining pain ratings were
Baseline measurement results and change at each followup time point are summarized in Table
Baseline and weeks 6 and 12 changes (LS mean ± SE).
Variable | Week 0 (range) | Week 6 change | Effect size | Adjusted |
Week 12 change | Adjusted |
---|---|---|---|---|---|---|
400-meter walk time (sec) |
|
|
0.06 | 0.9805 | − |
0.8828 |
Stair climb time (sec) |
|
−0.5 ± 0.2 | −0.56 | 0.0036 | − |
0.0013 |
Stair climb power (W) |
|
|
0.62 | 0.0075 |
|
0.0051 |
LLFDI: basic lower limb function |
|
|
0.32 | 0.2279 |
|
0.3657 |
KOOS: knee pain subscale |
|
|
0.44 | 0.0743 |
|
0.0496 |
KOOS: knee-related activities of daily living |
|
|
0.56 | 0.0107 |
|
0.0070 |
KOOS: knee-related quality of life |
|
|
0.69 | 0.0068 |
|
0.0224 |
SF36: bodily pain score |
|
|
0.69 | 0.0040 |
|
0.0453 |
*Dunnett-Hsu adjustment; for these repeated measures analyses, 81 observations were used, with each participant contributing data at a maximum of 3 time points.
Six weeks after completion of the aquatic intervention, participants continued to demonstrate improvement in stair climb time and power, as well as in knee pain, activities of daily living, quality of life, and bodily pain. However, the effects on bodily pain and quality of life were attenuated in the context of 7 of the 29 subjects failing to return for the 12-week visit. There continued to be no significant effect on 400-meter walk time and basic lower limb function score on the LLFDI in comparison with baseline measures.
Overall, the results of this study demonstrated feasibility and tolerance of an aquatic-based power training intervention and potential for an immediate beneficial effect on lower limb power. The detection of improved knee function (stair climb time, KOOS knee-specific activities of daily living and quality of life (QOL) and SF-36 bodily pain score) supports the need for a controlled trial to assess the independent effects of the aquatic power training program. The durability of these findings when assessed 6 weeks following the conclusion of training suggests that short-term participation could potentially provide longer-term benefits as significant effects were maintained without continuation of training. However, the primary relevance of the findings of this study relates to subjects’ tolerance of the aquatic power training program, as well as the suitability of the eligibility criteria, outcomes, and intervention.
Despite the need to modify the protocol for some participants, the compliance of 92% (319 visits attended out of 348 total for 29 subjects), for the 6-week intervention time period, suggests that the intervention was well tolerated. The report of increased pain in only 4 of 34 subjects and retention of 29 of 34 subjects who started the program was considerably better than the rate of 50% in a land-based power training program that had similar eligibility criteria [
The eligibility criteria were selected to specifically target subjects with knee OA with clinically relevant needs—those with daily pain and with mobility limitations. A total of 39 people with symptomatic OA, who were screened, were ineligible to take part in the study due to a walking speed that was above average for age and sex (Table
As the intervention was designed with the intent to improve mobility, the 400-meter walk test was selected as a suitable primary outcome measure for mobility. However, this test may not be the optimal outcome measure for this aquatic training program. In retrospect, these findings are consistent with the principle of specificity of training, in which exercises that closely approximate the goal functional activity are the most effective in improving physical performance during that activity. The exercises involved stepping motions as well as targeting muscles necessary for stepping, but did not focus on walking. Similarly, in a land-based study of adults over age 70, power training improved muscle power but not walk time compared with a walking program [
Alternatively, underwater treadmill training could be added to increase the specificity for targeting gait speed. As the intervention was originally intended to improve gait speed in older adults with symptomatic knee OA and mobility limitations, water walking was included between the exercises. However, this exposure was of an insufficient training intensity to elicit a detectable improvement in gait speed. Therefore, we speculate that the increased intensity involved with underwater treadmill training might be better suited to achieving the goal of improving gait speed while maintaining the partial weight-bearing necessary for tolerance of the intervention.
As impairments in lower limb muscle power have been associated with mobility limitations [
Although the intervention may not have been well targeted towards walking, it did demonstrate potential for eliciting improvements in symptoms and quality of life. In a prior 6-week study of adults with a similar mean age and meeting similar inclusion criteria (ACR criteria for OA and frequent pain), Hinman et al. found that an aquatic physical therapy intervention resulted in improved WOMAC (a subset of KOOS) [
When assessing improvement in impairments and functional limitations, it is necessary to confirm that changes detected are not merely statistically significant, but also clinically meaningful and that the results are generalizable. Considering that the minimum clinically important difference (MCID) for improvement on the WOMAC pain scale (a subset of the KOOS) has been reported to be 2.1 points [
A second important finding of this study was that the significant improvements were maintained 6 weeks following discontinuation of the intervention, suggesting that the intervention may not need to be continuous to confer some residual benefits. This result is consistent with the previous report by Hinman et al., in which improvements in symptoms, function, and mobility were maintained 6 weeks after completion of the supervised intervention. The sustained effect in the study of Hinman et al. may have related to the fact that 84% of participants continued the intervention independently. Although participants were allowed to continue training during the 6 weeks following completion of the intervention in the present study, no subjects elected to continue the intervention—possibly due to the location of the pool in the medical center being inconvenient for subjects or due to the high water temperature. Interestingly, despite the lack of continued participation, subjects demonstrated sustained reduction in impairments and functional limitations.
The principal limitation of this pilot study was the absence of a control group. This study was conducted to assess the feasibility and tolerance of power training in an aquatic environment among adults with symptomatic knee OA, rather than primarily to assess efficacy. Because there was no control group, in addition to the potential for the power training intervention to have accounted for the results, an alternative possibility is that nonspecific effects of the aquatic environment may have contributed to the results. Given the universally soothing effect of warm water, it is possible that the reduction in pain may be partially attributable to (a) the effects of buoyancy on decreasing joint loading, (b) specific heat enabling increased joint range of motion, and/or (c) vasodilatation enhancing nutrition of joint tissues and clearance of edema and metabolic waste products [
Another limitation was the lack of a quantitative assessment of the degree to which subjects were power training. Interpretation of the results would benefit from incorporating an objective measure of the force and velocity of motion during the aquatic exercises. Our observation suggested that 21 of the 29 subjects engaged in exercise of sufficient velocity to be considered power training. Those who did not engage in what we believed to be true power training did not do so due to concern for exacerbation of joint pain. If the intervention were individualized, a progressive exercise program, in which, subjects who have difficulty with the power training start at a low intensity (below the threshold for pain provocation) and progressively increase the intensity to a level appropriate for power training, could potentially enable tolerance in a greater proportion of subjects. Lastly, although we found that 6 weeks following completion of the protocol, many of the benefits were sustained, no subjects elected to continue following completion of the supervised training. In order to assess whether continuation may result in greater benefits, future studies should incorporate a motivational intervention.
In conclusion, our data suggest that a 6-week aquatic rehabilitation program is both feasible and well tolerated by adults over age 50 with daily knee pain, clinical knee OA, and mobility limitations. Furthermore, this program may result in improved lower limb muscle power, symptoms, activities of daily living, and quality of life, but may not improve walking speed or self-reported lower limb function. The results suggest that benefits persist for at least 6 weeks following discontinuation of the program. Future studies should include a comparison group, an objective measure of the power training intensity, a greater duration of follow-up, and eventually a transition from aquatic to land-based training.
This research was supported by a National Institutes of Aging Paul B. Beeson Career Development Award in Aging Research (K23AG030945). Investigators retained full independence in the study design, data collection, analyses and interpretation of data, writing of this paper, and its submission for publication. The authors have no financial benefits to disclose. This research has not been previously presented.
The authors thank the participants and staff involved in this study as well as the National Institute on Aging for funding this study (Grant K23AG030945).