Volume 26, Issue 5 , Page 322, June 2003
Effect of a topical herbal cream on osteoarthritis of the hand and knee: a pilot study
Article Outline
Abstract
Objective
To compare the efficacy of an herbal ointment to a sham ointment for relieving pain and stiffness associated with osteoarthritis of the hand and knee.
Method
Single-blind, randomized controlled clinical trial conducted in 3 chiropractic medicine practices. Subjects were solicited from patients attending the 3 centers. Selection of subjects with osteoarthritis of the hand or knee was based on the criteria developed by the American College of Rheumatology. Subjects were randomized to an active (n = 17) or a sham (n = 19) group. Subjects applied the herbal ointment to the affected joint(s) for 42 consecutive days and recorded level of pain and stiffness daily on visual analog scales.
Results
Thirty subjects completed the study (15 active and 15 sham). Significant differences in pain (P = .003) and stiffness (P = .0008) were found between the first 21 days of application and the last 21 days of application for the experimental group but not for the sham group. Gain scores between the experimental and sham groups were significant for pain (P = .026) and stiffness (P = .042).
Conclusion
An herbal ointment showed significant improvement in pain and stiffness for patients with hand and knee osteoarthritis who applied the ointment to the affected joint(s) for 42 consecutive days.
Keywords: Osteoarthritis, Herbal Medicine, Complementary and Alternative Medicine
Introduction
An estimated 40 million Americans are affected by arthritic diseases and the number is expected to rise to 60 million by 2020.1 Osteoarthritis (OA), the most common type of arthritis, is a major cause of disability and suffering. The term osteoarthritis was coined at the beginning of the 20th century, following observation of osteophytosis by pathologists and radiologists.2 Previously considered a degenerative disease due to aging and trauma, OA is becoming considered a metabolically active reparative process responsive to treatment.3 OA is thought to be a collection of similar diseases affecting joints rather than a single disease.4 This disease typically affects the knee5 and hands and is the primary reason for joint replacement surgery.6 The prevalence of hand OA is about 3.3% for men and 6.8% for women.7 The knee is affected in about 15% of people over 55,3 and it is estimated that by the age of 65, 80% of the population will have detectable changes typical of OA. Of those patients with OA, 60% with radiographic changes have pain, while 15% to 30% have mobility problems.8 The most common clinical features of OA include pain, stiffness, swelling, and inflammation, and the condition can be diagnosed by radiograph. Further, crepitus, bony enlargement, deformity, instability, restricted movement, warmth, effusion, synovial thickening, along with muscle weakness or wasting represent signs of this condition.3 Risk factors include advancing age, repetitive motion, family history, obesity, and injury.9
As no cure is available for OA, treatment centers on reducing symptoms. Such treatment can include exercises or orthoses but usually involves analgesics, such as acetaminophen and nonsteroidal anti-inflammatory drugs (NSAIDs) such as aspirin or ibuprofen. Based on current and predicted estimates of OA rates, an increase in the number of people suffering side effects from analgesics and NSAIDs will also occur. Consequently, with the frequency and severity of side effects from NSAIDs and analgesics, suggestions for less toxic treatments of OA are warranted.10 Natural remedies, because of their relative safety, may reduce the dependency on NSAIDs and analgesics and could have an important role in the treatment of OA, even if they are only moderately effective.10 A recent clinical trial compared piroxicam gel with homeopathic SRL gel and found the homeopathic gel to be as effective and as well tolerated as the NSAID gel.11 The homeopathic SRL gel contained comfrey, poison ivy, and marsh tea in homeopathic tinctures for external use. The piroxicam (Feldene) gel contained 0.5% piroxicam. Subjects applied 1 g of gel 3 times a day for 4 weeks. Mean pain reduction was 16.5 mm in the homeopathic group and 8.1 mm in the piroxicam group.
A study of topical capsaicin in osteoarthritis of the hands found tenderness to be reduced by 40% compared with the control group.12 Twenty-one patients applied 0.075% capsaicin cream 4 times a day for 4 weeks.
Arthritis Relief Plus (ARP) (Tabe Herbals, Surfers Paradise, Queensland, Australia) is a new herbal ointment containing only natural medicinal herbs used for external application. The ointment has a certificate of listing with the Australian Therapeutic Goods Administration and with the American Food and Drug Administration13 and has been promoted as having therapeutic effects based on the action of the herbs on bone and joint tissue to relieve pain, inflammation, and thus stiffness.13 The ointment is prepared commercially in Australia and contains the following herbs: brewer’s yeast, comfrey, common fallow laurel, English oak, fenugreek, kelp, Solomon’s seal, tumeric, and winter cherry. The cream also contains glycerol, castor oil, coconut oil, menthol, and capsaicin. The quantity of capsaicin is 0.015%, as compared with the available capsaicin creams, which come in concentrations of 0.025% and 0.075%.
The cream was initially formulated in Australia to treat the common joint and tendon diseases seen in racehorses and racing dogs. After several years of experience, Australian Rules football players used the cream with good results. The cream was then introduced for general use with various forms of arthritis, especially osteoarthritis.
Anecdotally, the ointment has been used with positive effects in the management of OA for several years in both Australia and India. However, no published clinical trials exist. The purpose of this study was to compare the efficacy of ARP to a placebo in the relief of osteoarthritic pain and stiffness of the hands and knees in patients diagnosed with OA.
Methods
This was a single-blind, randomized controlled clinical trial with subjects blind to the type of ointment they used. Subjects with OA of the hand or knee were solicited from patients of 3 chiropractic physicians located in the midwestern United States. Over a 6-week period, all patients with hand or knee pain and stiffness were screened for OA, and 36 patients met requirements of the study. Inclusion in the study was based on the American College of Rheumatology (ACR) criteria.14 The ACR standards classify subjects as having symptomatic conditions, based on duration and frequency of selected symptoms, and involves specific OA criteria for the hands and knees, which have been shown to be more sensitive and specific than radiographic findings.7
The exclusion criteria included open lesions on the knees or hands, intra-articular steroid injection into the hand or knee joints within 4 weeks preceding entry into the study, having engaged in disability-related litigation, or prior joint replacement surgery.
All nondrug modalities of treatment (exercise, splints, nutritional supplements, physical therapy) and all drug treatments remained constant during the study (ie, no new drug therapy, vitamin or mineral supplement, herbal product, or exercise program was started during the study). The study was approved by the University Institutional Review Board, and signed informed consent for participation was obtained from each subject.
Using a table of random numbers, as recommended by Borg and Gall,15 the randomization schedule was generated before the start of the study. The 36 subjects were randomized to either an experimental (n = 17) or a placebo (n = 19) group. Of the original cohort, 30 subjects completed the study (15 active and 15 placebo). Two subjects dropped out of the experimental group; 1 subject’s spouse was hospitalized with a long-term illness, and the other subject had watering of the eyes, which she thought was due to the herbal product. In the placebo group, 4 subjects dropped out, all for nonresponse to treatment. Seventeen subjects had OA of the hand, and 13 subjects had OA of the knee. In the experimental group, 7 subjects were diagnosed with hand OA, and 8 subjects were diagnosed with knee OA. In the placebo group, 10 and 5 subjects had hand and knee OA, respectively. The experimental group (11 women and 4 men) and the placebo group (12 women and 3 men) were given identical instructions. The research protocol was given orally and in written form to each subject, so that consistency in ointment application and compliance could be strengthened.
For 42 consecutive days, each evening prior to applying either the placebo or the experimental ointment, subjects recorded their level of pain and stiffness on 100-point visual analog scales (VAS). Separate scales were used, 1 for pain and 1 for stiffness. The pain scale was a 100-mm line with no pain on the left and the worst pain imaginable on the right. The stiffness scale was a 100-mm line with no stiffness on the left and severe stiffness on the right. While a VAS is not a commonly used outcome measure for stiffness, Bellamy and Buchanan16 include the Likert scale and the VAS as outcome measures for stiffness in their discussion of clinical evaluation tools for the rheumatic diseases.
The ointments (placebo and active) were gently massaged around the affected joint for 2 to 3 minutes. Massaging for several minutes was used to facilitate penetration of the cream by increasing local blood supply and by encouraging local movement. The preparations were well matched for color, smell, and consistency. Neither the examiner nor patient could distinguish the preparations. The placebo contained molasses, arrowroot powder, neem powder, and capsaicin and menthol in concentrations identical to the active cream. Molasses was too dark, so 15% arrowroot powder was added. Neem powder was used to give an herbal smell similar to the experimental cream. The containers were labeled identically, with the patient number being the only distinguishing feature. After application, the joint was covered with a cotton cloth and/or an ace bandage to protect the bed linens from being stained. The area was washed each following morning on awaking.
The subjects were asked to report any side effects from the ointments and were personally seen every 2 weeks during the course of the study. During these visits, the patients were asked about compliance and correctness of ointment application. All participants were compliant with the treatment program. The final follow-up visit with the clinician occurred within 1 week of ending the study. During this visit, all the VAS scales were collected, and the subject was asked about adverse effects.
Clinical experience indicates improvement with ARP ointment does not occur until the ointment has been used for 3 weeks (L. McKay, written communication, November 15, 2000), so data from the VAS scales for pain and stiffness (dependent variables) were separated and analyzed by 2 categories: days 1 through 21 (baseline) versus days 22 through 42.
Pain and stiffness raw scores were summed for each subject over the 21-day period, and the average of all summed scores across all subjects in the group was obtained. The first 21 days of application was used as the baseline, and the mean for pain and for stiffness was compared with the mean for pain and stiffness of the last 21 days of application. Paired t tests were used to compare these within-group mean values between the first 21 days and the last 21 days for pain and stiffness in each condition.
Raw scores were summed for each subject in each group for the baseline phase and the treatment phase, and the difference (improvement or lack of improvement) for each subject was obtained. The mean of these difference scores was then obtained for each group (placebo/active) across the pain and stiffness dimensions. These mean values were then compared using independent t tests. An alpha level of P < .05 was used to determine significance.
Results
The subjects’ general characteristics are shown in Table 1. Significant differences in both mean pain (P = .003) and mean stiffness (P = .0008) were found between the first 21 days of application and the last 21 days of application for the experimental group but not for the placebo group (Table 2). Analysis of mean gain scores between the experimental group and the placebo group yielded significant differences in both pain (P = .026) and stiffness (P = .042) (Table 3). No significant differences were found between sex for either pain or stiffness and no significant differences were found between hands or knees for pain or stiffness.
Table 1. General characteristics of subjects by group
| Placebo group mean SD | Active group mean SD | |
|---|---|---|
| Age (y) | 60.40 | 63.5 |
| 10.75 | 12.09 | |
| Height (cm) | 168.5 | 163.8 |
| 7.6 | 6.8 | |
| Weight (kg) | 88.3 | 78.4 |
| 30.3 | 29.3 |
Table 2. Within group change in pain and stiffness between days 1-21 and days 22-42
| Variable | Mean SD | N | t | df | P |
|---|---|---|---|---|---|
| Placebo Pain | 968.5 | ||||
| Days 1-21 | 373.9 | ||||
| 15 | 1.21 | 14 | .247 | ||
| Placebo pain | 869.5 | ||||
| Days 22–42 | 427.4 | ||||
| Experimental pain | 992.4 | ||||
| Days 1-21 | 492.4 | ||||
| 15 | 3.56 | 14 | .003∗ | ||
| Experimental pain | 615.0 | ||||
| Days 22-42 | 350.4 | ||||
| Placebo stiffness | 947.2 | ||||
| Days 1-21 | 386.5 | ||||
| 11 | .179 | 10 | .861 | ||
| Placebo stiffness | 927.6 | ||||
| Days 22-42 | 470.1 | ||||
| Experimental stiffness | 1045.3 | ||||
| Days 1-21 | 277.1 | ||||
| 14 | 4.31 | 13 | .0008∗ | ||
| Experimental stiffness | 682.2 | ||||
| Days 22-42 | 405.3 |
∗ P < .05. |
Table 3. Between groups chance in pain and stiffness scores
| Variable | Mean | t value | df | P |
|---|---|---|---|---|
| Placebo pain | 85.9 | |||
| −2.34 | 28 | .027∗ | ||
| Experimental pain | 444.1 | |||
| Placebo stiffness | 87.9 | |||
| −2.16 | 22 | .042∗ | ||
| Experimental stiffness | 362.9 |
∗ P < .05. |
Discussion
Pain decreased by 38.0% in the experimental group compared with only 10.2% in the placebo group, and stiffness decreased by 34.7% in the experimental group and by 2.1% in the placebo group. Further, in dividing the data by duration of treatment, it was possible to determine the effectiveness of the experimental ointment in comparison with the placebo. This separation also reduced any possible placebo effect that may have threatened the data. By comparing within-group first half to last half data and between-group scores, it is evident that the experimental ointment reduced both pain and stiffness, and the pain and stiffness reduction was equally apparent in both hands and knees, as well as in both men and women.
In drug trials, NSAID effectiveness across studies is about the same, with patients reporting a 30% reduction in pain and 15% improvement in function.17 The results of the current study suggest the herbal cream is as effective as the NSAIDs but without the attendant side effects. However, due to the small sample size of this study, further study with larger groups is required to confirm this finding.
Most topical creams for osteoarthritis have a short-term effect. These creams may be rubefacients, counterirritants, and NSAIDs. Several neuropeptides play a major role in pain transmission within the central nervous system. The most important of these are substance P and glutamate. Both are released from C fibers following a noxious stimulus. Local application of capsaicin initially stimulates the release of these neurotransmitters and is algesic. Continued application blocks their release and leads to desensitization and analgesia.17 Capsaicin cream depletes substance P locally when applied 2 to 4 times daily in a strength of 0.025% to 0.050%.18
The herbal cream is different in action from the other topical therapies in that its activity is delayed in onset and the effects are prolonged. In a concentration of 0.015%, the capsaicin did not produce burning or irritation of the skin in any of the subjects. Further, the active and placebo creams contained equal amounts of capsaicin and menthol. This suggests the effects observed were not mediated by the capsaicin component.
A weakness of the study was the small sample size. Another possible weakness was the lack of blinding of the observers. While prestudy preparation emphasized treating all subjects the same, the lack of clinician blinding may have introduced a potential bias.
Conclusion
The results do not suggest that all OA sufferers can benefit from topically applied herbal ingredients; however, the data show that a significant change occurred as a result of following a regimen of experimental ointment application. Since the ointment is somewhat messy, assurance of continuous compliance was based only on the word of each patient. If compliance suffered due to this factor, it may be assumed that stricter compliance could result in greater benefits. Ultimately, the beneficial result of this study was determining the effectiveness of a nonprescription external herbal method for managing arthritic pain and stiffness. Indeed, such intervention not only benefits the patient in terms of comfort, it may help to reduce side effects and contraindications of drug use. Further study is warranted comparing the herbal ointment to standard drug therapy and as a complementary treatment to standard drug therapy. A larger, double-blind, placebo-controlled clinical trial is in the planning stages.
Acknowledgements
The authors wish to thank Peter Anderson of Tabe Herbals for supplying the experimental and placebo creams.
References
- . Health in the later years. 3rd ed. Boston (MA): McGraw Hill; 2000; 267, 269
- . Osteoarthritis (time to shift the paradigm). Br Med J. 1999;318:1299–1300
- . ABC of rheumatology (osteoarthritis). Br Med J. 1995;310:457–460
- . Osteoarthritis. Lancet. 1997;350:503–509
- A randomized trial of acupuncture as an adjunctive therapy in osteoarthritis of the knee. Rheumatology. 1999;38:346–354
- . Radiographic assessment and psychological variables as predictors of pain and functional impairment in osteoarthritis of the knee or hip. Arthritis Rheum. 1988;31:204–209
- . Hand osteoarthritis in the elderly. Application of clinical criteria. Scand J Rheumatol. 1996;25:34–36
- . Comparison of an anti-inflammatory dose of ibuprofen, an analgesic dose of ibuprofen, and acetaminophen in the treatment of patients with osteoarthritis of the knee. N Engl J Med. 1991;325:87–91
- . Comparison of clinical, radionuclide, and radiographic features of osteoarthritis of the hands. Ann Rheum Dis. 1991;50:623–626
- . Glucosamine and chondroitin for treatment of osteoarthritis. A systematic quality assessment and meta-analysis. JAMA. 2000;283:1469–1475
- . A randomized controlled trial comparing topical piroxicam gel with homeopathic gel in osteoarthritis of the knee. Rheumatology. 2000;39:714–719
- . Effect of topical capsaicin in the therapy of painful osteoarthritis of the hands. J Rheumatol. 1992;19:604–607
- . Maximizing the arthritis treatment: breaking the cycle of pain. Dallas: F.E. Watson; 1999; 4-12
- . Criteria for classification of clinical osteoarthritis. J Rheumatol. 1991;18(Suppl 27):10–12
- . Educational research. An introduction. New York: Longman; 1983;
- . Clinical evaluation in the rheumatic diseases. In: Koopman WJ editors. 13. Arthritis and allied conditions. A textbook of rheumatology. 1:Baltimore: Williams & Wilkins; 1997;p. 47–79
- . Published trials of nonmedicinal and noninvasive therapies for hip and knee osteoarthritis. Ann Intern Med. 1994;121:133–140
- . Capsaicin and pain mechanisms. Br J Anaesth. 1995;75:157–168
PII: S0161-4754(03)00009-5
doi:10.1016/S0161-4754(03)00009-5
© 2003 National University of Health Sciences. Published by Elsevier Inc. All rights reserved.
Volume 26, Issue 5 , Page 322, June 2003
