Arthritis is the leading cause of physical disability among older adults, affecting more than 70 million Americans, of whom the majority are women (1–4). The joint damage and chronic pain from osteoarthritis (OA), the most common form of arthritis, lead to muscle atrophy, decreased mobility, poor balance, and, eventually, physical disability (5–8). Traditional therapies include pharmacologic, surgical, and exercise interventions. Pharmacologic therapy includes the use of antiinflammatory medications that have potentially serious long-term side effects (9, 10). Recent evidence also casts doubt as to the effectiveness of arthroscopic surgery for adults with mild to moderate knee OA (11). Consequently, the need for a safe and effective treatment for knee OA has never been more evident.
Previous clinical exercise trials have demonstrated significant, yet modest, improvements in function and pain in older adults with knee OA (12–18). There is also evidence that obesity is strongly associated with knee OA (19–27) and that weight loss may prevent the onset of this degenerative joint disease (28). Accordingly, both the American College of Rheumatology (ACR) and the European League Against Rheumatism (EULAR) recommend weight loss and exercise for obese patients with knee OA (29, 30).
A nonrandomized trial conducted by Toda (31) demonstrated greater short-term (8-week) benefits of diet plus exercise compared with diet alone. Preliminary data from our laboratory suggest that a combined exercise and dietary weight loss approach may indeed be effective in improving clinical outcomes for older adults with OA (32). In addition, Huang and colleagues (33) reported that interventions that included diet and exercise improved weight loss, disability, and walking speed compared with pain therapy alone. Whereas these studies lend support to the efficacy of exercise and dietary weight loss in improving function in persons with knee OA, the lack of control groups, full factorial designs, short intervention periods, and small sample sizes limit the generalizability of these results. Interestingly, results from our current study suggest that the combination of diet and exercise is superior to either treatment alone in improving self-reported physical health measures of quality of life (34).
We present data from the Arthritis, Diet, and Activity Promotion Trial (ADAPT), which examined the effects of exercise and weight loss interventions, both separately and in combination, on self-reported physical function in overweight and obese older adults with knee OA. Moreover, we examined the effects of these interventions on the secondary outcome measures of weight loss, pain, mobility, and joint space narrowing.
- Top of page
- PATIENTS AND METHODS
Results of previous short-term studies suggest that the combination of dietary weight loss plus exercise is effective in improving self-reported physical function and mobility in osteoarthritic patients compared with exercise only (32), diet only (31), and pain therapy (ultrasound and transcutaneous electric stimulation) (33). Our statistical analysis indicated that all of the intervention groups and the healthy lifestyle (control) group tended to have improved function; however, it was when diet was combined with exercise that subjects realized the greatest and statistically significant benefit in function. Interestingly, the 3 intervention groups improved within the first 6 months, and then maintained these improvements for an additional 12 months. This was especially noteworthy among patients in the diet plus exercise group, who maintained a 24% improvement. Only the healthy lifestyle group experienced regression of function toward baseline values. Therefore, we suggest that the combination of diet plus exercise produces consistently better and clinically relevant improvements in self-reported physical function compared with either diet or exercise alone.
The effects of long-term exercise programs (aerobic or resistance) on self-reported physical function indicate significant, yet modest, beneficial effects (1–11%) with a gradual return toward baseline values (12, 13, 15). The improvement observed in our exercise-only group (13% improvement) compared favorably with that observed in previous studies (12, 13, 15), yet there was no significant difference relative to the healthy lifestyle control group. The surprising 13% improvement in the control group combined with the high functional level of the entire cohort (see Table 1) made attaining significant gains more difficult.
The decision to use an attention control versus a no-treatment control group in clinical trials research usually creates considerable debate among investigators. Equating the amount of contact time between intervention and attention control groups is difficult and costly. Recently, we observed declines over a 30-month period of 10% and 6% in lower extremity strength and balance, respectively, in older adults with knee OA (45). These data suggest that in the absence of any intervention, a decline in physical ability and physical performance is likely in this older, osteoarthritic population. The improvements in our control group suggest that the healthy lifestyle intervention was effective in slowing the decline in function commonly seen in no-treatment control groups.
The increase in disability and dependence in performing daily activities observed in older adults with OA is accelerated by mobility impairment (46). Although short-term exercise programs have demonstrated between 4% and 18% improvements in mobility in patients with knee OA (16–18), the effects of long-term exercise programs indicate more modest improvements (1–2%), with a return toward baseline values (15). Our physical performance (6-minute walk distance, stair-climb time) data indicate that exercise can result in clinically relevant (16% overall improvement) and statistically significant long-term gains in mobility, effectively slowing the increase in mobility impairment that is common in an older, osteoarthritic population (45, 46).
Radiographic progression of disease, as measured by joint space narrowing, was similar among the 4 study groups. The relatively short duration (18 months) of the intervention and the number of subjects per group (∼80) probably prevented the detection of meaningful differences in radiographic disease progression.
Often, older adults with knee OA are reluctant to exercise for fear of exacerbating their knee pain. In the diet plus exercise group, walking velocity significantly increased, suggesting an increase in the load placed on the lower extremity joints, yet knee pain significantly decreased (by 30%). For the exercise-only group, the improvement in walking velocity did not have an adverse effect on knee pain, which improved 6% from baseline. These data are consistent with results of 2 previous studies of older adults with knee OA who were enrolled in exercise or diet plus exercise programs (15, 32). Interestingly, diet produced a 16% improvement in knee pain; however, it was diet in combination with exercise that had the greatest and only significant effect on pain (see Table 4). From a clinical perspective, these results suggest that physicians can prescribe either diet or exercise for their overweight and obese patients with mild to moderate knee OA without the likelihood of worsening their symptoms. Furthermore, patients are most likely to realize the greatest improvements in pain when they combine diet with exercise. We acknowledge, however, that this assertion is based on mean values, and that there will be individuals whose symptoms may worsen with exercise.
The diet plus exercise and diet-only groups lost significantly more weight (5.7% and 4.9%, respectively) than did the control group. This weight loss is consistent with that observed in previous long-term weight loss studies (37, 47, 48). Williams and Foulsham (49) reported that the degree of weight loss was related to clinical improvement. While this held true for our self-reported measures, subjects in the exercise group, who experienced only a 3.7% weight loss, generally performed better on mobility measures than did the diet-only group. Change in muscle mass, which was not measured, may have contributed to the differences in mobility between the groups.
Relative to our prior experience with a combined facility- and home-based exercise clinical trial (15), overall adherence was lower in ADAPT but was similar to that observed in the facility-based group in another large-scale community-based exercise trial (50). These data underscore the difficulty in maintaining standard exercise and dietary weight loss programs in previously sedentary, overweight adults with mobility disability, and emphasizes the need for studies that focus on strategies that improve adherence to long-term exercise and dietary therapy programs.
Improvement in the diet and exercise groups was dependent on the type of outcome measure. More specifically, the diet-only group performed better than the exercise-only group on self-reported measures (although neither group was statistically different from the healthy lifestyle group), while the reverse was true for mobility measures. It appears logical, then, that combining the 2 interventions would lead to the diet plus exercise group's consistently superior results on both self-reported and mobility measures. Further research is needed to determine whether there is a significant dose response to both exercise and weight loss on measures of physical function and mobility.
For many years, physicians have recommended that overweight patients with knee OA exercise and lose weight, and this approach has been recommended in the ACR guidelines for the management of knee OA (29). However, our study is the first large, randomized, controlled clinical trial to demonstrate the relative and combined contributions of exercise and weight loss on function, pain, and mobility in patients with knee OA. This study indicates that combining modest weight loss with moderate exercise provides the best overall improvements in self-reported measures of function and pain and performance measures of mobility. These data also show that long-term weight loss through calorie restriction can be achieved in this population. Without the addition of exercise, however, dietary weight loss alone does not result in significant improvements in mobility (an important determinant of disability) or self-reported function and pain. Considering the side effects that often limit the use of OA drug therapy (9) and the possible ineffectiveness of surgical intervention in cases of mild to moderate knee OA (11), our results give strong support to the combination of exercise and weight loss as a cornerstone for the treatment of overweight and obese patients with knee OA.