David Mark06.01.09
My grandfather was a big man with a slow step. He smoked cheap cigars. On Fridays and Saturdays he wore a brown suit, walked to shul for the Sabbath prayers, and walked home again. On Sunday mornings he sat on the patio drinking a glezele tey mit tsuker (glass of tea with sugar) and reading Foverts (Yiddish newspaper). Mondays through Fridays he put on his butcher's whites and took the bus to work, where he spent the day turning forequarters of kosher beef into parcels for the customers.
By the time I was old enough to remember, my grandfather had osteoarthritis in his knees and his right hand. He couldn't even open his hand to shake hands. At work, he held his hand in warm water until the joints loosened enough to hold a knife. He could not tie his own shoes or button his own shirt, so my grandmother had to do that for him.
When one person in a family has arthritis, the whole family has custody of the arthritis. Or maybe the arthritis has custody of them. There is the help needed in the kitchen, in the yard, loss of income when a job is no longer possible, weight gain due to inability to exercise, compromised vacation and retirement plans, not being able to help care for grandchildren, and the potential cost and burden of surgery and recovery.
The Centers for Disease Control and Prevention (CDC), Washington, D.C., conduct frequent surveys of the health status of people living in the U.S., which include questions on joint pain. The results are tabulated in an annual report "Health, United States." Figure 1 shows some of the joint pain results from the 2006 survey, as contained in the 2008 report. The exact survey question was: "During the past 30 days, have you had any symptoms of pain, aching or stiffness in or around a joint?"
By 60 years of age, more than 40% of men and 50% of women answered that they had had joint pain. These percentages translate to 25 million people with osteoarthritis and an almost equal number with other arthritis types. Joint by joint, knee pain was the most common, followed by shoulder, finger and hip pain. For any joint, women had a higher incidence than men. Because the survey is limited to the non-institutionalized population it under-measures the prevalence of arthritis in people ≥ 75 years. Further, Asian and Hispanic respondents were less likely to say they had joint pain, while American Indians were more likely to. Amount of education had little impact, but respondents living below poverty level were more likely to have arthritis.
The prevalence of osteoarthritis in the U.S. is expected to worsen, as obesity increases the risk of developing osteoarthritis and the prevalence of obesity, as defined by BMI > 30, increased from 13% in 1960 to 35% in 2006. The relationship makes sense from a biomechanical stress perspective, but hand osteoarthritis also correlates with obesity, suggesting a hormonal contribution too. It has been hypothesized that the hormone leptin, produced by fat cells, may contribute to osteoarthritis risk. This might also explain why women are more likely to develop osteoarthritis, as they usually have a higher percent body fat than men. Bariatric surgery leading to major weight loss often completely relieves osteoarthritis pain.
Glucosamine. There has been more news than expected on glucosamine since the publication of GAIT (Glucosamine Arthritis Intervention Trial). Interestingly, the Glucosamine Unum in Die Efficacy (GUIDE) trial reported less knee pain at six months, while another trial also with >100 subjects reported glucosamine as not significantly better than placebo for hip pain at 24 months. Both trials used glucosamine sulfate.
GAIT was criticized for using glucosamine hydrochloride instead of the more studied sulfate form. The rationale from GAIT researchers was that because both HCl and sulfate are disassociated from glucosamine prior to absorption, it should not matter. One trial out of Russia reported a benefit for glucosamine HCl, and two trials out of China did head-to-head comparisons of glucosamine HCl and glucosamine sulfate and reported no differences between the two. Unfortunately, neither had a placebo group. Bruyere et al reported on a post-trial follow-up. Subjects were contacted an average of five years after the glucosamine sulfate clinical trial had ended. Researchers found that knee replacements were more common in the placebo group (14.5%) than in the glucosamine group (6%).
Chondroitin. The chondroitin category, perhaps not surprisingly, has also received more attention since GAIT, and the news has been interesting, especially as support for 800 mg daily continues to gain traction. GAIT concluded that chondroitin sulfate at 1200 mg daily was no better at pain relief compared to placebo. But two trials concurrent with GAIT reported less knee joint-space narrowing over years of taking 800 mg of chondroitin daily. In 2007, Mazieres et al reported on 307 patients getting either chondroitin at 1000 mg per day or placebo for 24 weeks. The results either favored or trended for chondroitin. As with GAIT, the study was handicapped by a high placebo group response (56% using the OARSI test). The GAIT researchers reported that patients with less severe OA at trial enrollment had a statistically significant improvement for knee swelling for chondroitin over placebo. GAIT II reported 35% less loss of knee cartilage over two years in the chondroitin group compared to control, but the difference was not statistically significant because of the smaller than planned number of subjects completing the study. Finally, the "STOPP" trial enrolled 622 patients to take either chondroitin at 800 mg per day or placebo for two years. More than two-thirds of subjects completed the trial. Importantly, the "STOPP" chondroitin trial reached its two-year endpoint with three times as many patients as GAIT and was strongly positive for less cartilage loss in knee joints, even though subjects were only taking 800 mg daily.
Specifically, joint space loss was 0.150/year in the control group compared to 0.035/year in the chondroitin group, representing a 77% improvement. And pain relief occurred significantly faster in the chondroitin group compared to the control group. According to clinicaltrials.gov, there are six more chondroitin clinical trials either underway or starting soon, so potential remains for more good (or bad) news.
Glucosamine + Chondroitin. Given that more than 100 clinical trials have looked at either glucosamine or chondroitin, remarkably few looked at the two ingredients in combination. Perhaps this is because ingredient companies foot the research costs and few make both products. Unfortunately GAIT poked holes in the combination theory, as glucosamine (1500 mg) plus chondroitin (1200 mg) was found to be no better than either alone for pain relief. And jarringly, the combination was worse than placebo for loss of knee cartilage over two years, whereas glucosamine or chondroitin alone were better than placebo (none of the differences were statistically significant, but still).
A small study published in a 2007 edition of Osteoarthritis Cartilage reported no benefit for glucosamine and chondroitin in combination in terms of pain or knee strength at six months.
There are two more small trials from about 10 years back showing a pain relief benefit for the combination, but that's it.
Rose Hip (Rosa canina). A recent meta-analysis identified three randomized clinical trials of rose hip powder, at 5 grams per day, for osteoarthritis. A total of 306 subjects were enrolled. Patients getting the rose hip product had less pain and used less pain medication. All "responders to therapy" were 61% for rose hip powder and 42% for placebo (the latter is typical for arthritis pain trials). All of the clinical trials were sponsored by Hyben-Vital International, the manufacturer of a rose hip powder branded as LitoZin in some countries. The proposed mechanism of action includes antioxidant function and anti-inflammatory action of a specific galactolipid called "GOPO." But GOPO content is not specified.
Green-Lipped Mussel (GLM). New Zealand green-lipped mussels (Perna canaliculus) are both a food and a source for a dietary supplement for arthritis. Other arthritis products such as glucosamine, chondroitin, collagen and omega 3 fatty acids are also sourced from animals (although glucosamine and omega 3 are also available in vegetarian forms), so the idea of an animal product with a health benefit is not entirely foreign.
The story begins with an observation that coastal-living Maori had a lower risk of arthritis than inland dwellers. Whether true or not, this led to tests of freeze-dried, powdered mussels, identification and then solution via stabilization of an auto-oxidation problem akin to that seen for stabilized rice bran, and finally, a lipid extract of GLM promising to deliver the active fraction in a smaller volume without risk of triggering protein-based allergies. A recent review captured four clinical trials: two powder, one extract and one comparing extract to powder. The last did not have a placebo control. The review concluded that GLM may be of benefit in mild to moderate osteoarthritis, but because of the variation in product tested and experimental design, no definitive conclusion was reached. There is also an 8-week, open-label trial of the extract with 60 patients, but because of the known large placebo-response seen in osteoarthrititis trials it is difficult to judge the value of the results.
The proposed mechanism of action for GLM extract is the EPA and DHA content having an anti-inflammatory activity, possibly by inhibiting the lipoxygenase (LOX) and cyclooxygenase (COX) pathways of prostaglandin synthesis. The review cites evidence from animal models of adjuvant- or collagen-induced arthritis, and in fact it is these animal experiments, which are used in marketing claims boasting that GLM extract is far, far more potent than fish oils or plant oils. Given dosing of 200 mg per day (50 mg per softgel, 4 times daily) it is difficult to understand how the omega 3 fatty acids can be responsible for the claimed benefits, as osteoarthritis is not thought to be responsive to omega 3 fatty acids. Animal studies at Hong Kong Polytechnic University are reporting preliminary evidence for GLM extract affecting protein and cytokine expression, suggesting that other molecules in GLM extract may be active through other mechanisms.
Cat's Claw & Maca (Reparagen). Vincaria (Ua de Gato, translation: cat's claw) is a traditional-use, multi-functional herbal medicine used in parts of Peru. An osteoarthritis trial published in 2001 compared 100 mg per day of freeze-dried cat's claw to placebo. Improvement was apparent in one week. More recently, a larger, longer trial was completed comparing a cat's claw and maca combination (Reparagen) to glucosamine sulfate (no placebo). This well designed and executed trial reported WOMAC and VAS scores during a period of 8 weeks. The two treatments had near-identical effects on knee pain, stiffness and function. Benefits for both were seen at one week and continued to improve. The Reparagen patients used 10-15% less acetaminophen as pain relief medication during the trial.
Boswellia. The cumulative evidence for Boswellia serrata (frankincense) extracts recently crossed the line from wishful to something more substantial. The active ingredients are tentatively identified as boswellic acids, although there is dispute whether 11-keto-β-boswellic acid (KBA) or 3-acetyl-11-keto-β-boswellic acid (AKBA) are both active, or equally active. There are two commercialized boswellic extract ingredients with supporting clinical evidence-WokVel and 5-LOXIN. Both claim action as inhibitors of the 5-lipoxygenase enzyme (LOX) in the pathway of production of inflammatory leukotrienes. The old clinical work used 700-1000 mg per day of crude extracts with low boswellic acid content. However, the new products have higher concentration and lower amounts needed.
In vitro and animal toxicology testing, plus the minimal adverse events reported in clinical trials suggest that the safety profile is good. A concise review in a 2008 edition of the British Medical Journal by Dr. E. Ernst captured all of the clinical trials with control groups, including Sander et al on rheumatoid arthritis and Kimmakar, Sontakke and Sengupta on osteoarthritis. The rheumatoid arthritis results trended favorable and the osteoarthritis trial results were statistically significant. In vitro and animal work shows inhibition of LOX and also of metalloproteinase enzymes implicated in the degeneration of joint tissues.
Collagen. Ten grams a day for 24 weeks. That's what it took for young (20-year-old) athletes to feel improvements in joint pain, compared to what was reported by the placebo group. Measurements at 6, 12 and 18 weeks did not pick up the difference, although there was a trend at 18 weeks. Previous studies with the same collagen hydrolysate, summarized in a 2006 review, reported similar improvements in older subjects with knee osteoarthritis, however most of those studies were either without a control group or difficult to interpret. However, a very recent placebo-controlled trial in osteoarthritis patients offered more confirmation.
The mechanism for collagen supporting healthy joint cartilage may be stimulus, supply of raw materials or both. In vitro work suggests chondrocytes exposed to hydrolyzed collagen increase production of type II cartilage. In vivo work confirms that some of the collagen is absorbed as peptide chains rather than first being reduced to free amino acids. A single-dose human trial (Walrand 2008) demonstrated that 10 grams of collagen hydrolysate resulted in elevated plasma levels of collagen amino acids.
Eggshell membrane. Chicken eggs are assembled from the inside out. The yolk accrues to full size inside a membrane, then albumin is added outside the yolk and inside a double layer membrane, and then the calcified shell is added. Eggshell membrane can be separated, dried and powdered, and is currently being sold as a dietary supplement ingredient, called NEM ("Natural Eggshell Membrane"). The composition of eggshell membrane is primarily collagen, plus hyaluronic acid, chondroitin and other substances.
Two unpublished, open-label clinical trials enrolled a modest number of subjects with joint pain (any joint, not limited to osteoarthritis) for 30 days at 500 mg per day. There were improvements from baseline, but again given the large placebo effect for joint pain, not conclusive. Most recently, NEM was tested in a randomized, multicenter, double-blind, placebo-controlled study, which investigated the effectiveness of NEM for the relief of pain and stiffness associated with moderate osteoarthritis of the knee. Published in the April edition of the journal Clinical Rheumatology, the study investigated 67 patients who were randomly assigned to receive either 500 mg of NEM or placebo daily for 8 weeks. Supplementation with NEM produced an absolute rate of response that was statistically significant (up to 27%) versus placebo at all time points for both pain and stiffness, but was not significantly improved for function and overall WOMAC scores, although trending toward improvement. Rapid responses were seen for mean pain subscores (16% reduction) and mean stiffness subscores (13%) occurring after only 10 days of supplementation.
Methylsulfonylmethane (MSM) was a hot new ingredient a few years ago, but nothing has been reported since a 2006 study showing knee pain and function improvements after dosing with 6000 mg per day.
A root extract from Mojave yucca (Yucca schidigera) has traditional use for arthritis along with anti-inflammatory applications. A review of composition was published in 2006, which noted that human evidence for arthritis stems from three uncontrolled trials published in the 1970s, but they were not of sufficient scientific rigor.
Boron's laurels rest on one very small clinical trial conducted in the 1980s and an observation from the same investigator (Newnham) that countries with high dietary boron intake have low prevalence of arthritis. Most of the boron supplements on the market are at 3 mg per day for bone health-a few mention joint health too. Boron intake from foods is estimated at 1-2 mg per day.
Cetylated fatty acids have been evaluated in four osteoarthritis clinical trials, all with favorable results. But two were topical application, two were oral, and of the latter, only one (conducted in India) has been published so far. The authors theorize that LOX inhibition may be involved.
Hops products are usually marketed for mild calming, sedating or sleep promoting qualities, but there is one unpublished, small clinical trial of a hops extract in osteoarthritis patients. Over a 14-day period, 1000 mg per day reduced some of the WOMAC pain scores on some of the days. The proposed mechanism is as an anti-inflammatory and possibly a COX inhibitor.
We all know 41-50 count shrimp means so many per pound, but how about 1000 count? That's the size (less than half a gram each) of the shrimp-like Antarctic krill processed to make krill oil. This relatively new dietary supplement ingredient is a source of omega 3 fatty acids and the carotenoid astaxanthin. One clinical trial published in 2007 provided subjects with 300 mg krill oil per day for 30 days, and measured changes in C-reactive protein (inflammation biomarker) and WOMAC (a standard arthritis score). Improvements were better than what was seen in the placebo group. Interpretation of the study is weakened by the fact that enrollees, all selected for elevated CRP, either had cardiovascular disease, and/or osteoarthritis, and/or rheumatoid arthritis. The dogma for fish oil sourced omega 3 treatments is that it helps with rheumatoid arthritis, but not osteoarthritis.
A double-blind, placebo-controlled study published in the journal Nutrition Research (Farid 2007) showed that Pycnogenol, an antioxidant plant extract from the bark of the French maritime pine tree, improved physical function by 52% in patients suffering from osteoarthritis. In the study, 35 volunteers (average age 42) were randomly assigned a daily dose of Pycnogenol (50 mg, 3times daily) or placebo for 3 months. Patients reported arthritic pain using the WOMAC index after 30, 60 and 90 days. Participants also were instructed to indicate the frequency and dosage of NSAIDS and COX-2 inhibitor usage.
After 2 months of supplementation, physical function and pain scores improved in the Pycnogenol group. After 3 months in the Pycnogenol group, there was a reduction of 43% in pain, 35% in stiffness, 52% in physical function subscales and 49% composite WOMAC. The placebo group showed no significant scores throughout the entire study. Additionally, further reduction in the number of NSAIDS and COX-2 inhibitor pills and number of days taking medication was noted in the Pycnogenol group.
A recent study published in Redox Report (Belcaro 2008) showed Pycnogenol's anti-inflammatory potency further to improving osteoarthritis symptoms and pain by significantly lower plasma levels of CRP. With disease progression of osteoarthritis, the inflammation may reach a level where it no longer is limited to the affected joint and stresses the organisms, which in turn increases the inflammatory marker CRP in the blood.
This study investigated a subset 55 patients from a previous osteoarthritis study with 156 patients who had significantly elevated CRP levels. Treatment consisted of two tablets daily of either 50 mg Pycnogenol or placebo. Blood specimens were drawn at baseline of the initial study and again after three-month treatment. Results showed that Pycnogenol significantly lowered CRP from average 3.9 mg/L at baseline to 1.1 mg/L, reflecting essentially healthy levels. In the placebo group a marginal lowered CRP level was detected. Other blood parameters indicative of acute inflammation likewise decreased with Pycnogenol, such as fibrinogen (lowered 37%) and reactive oxygen species (lowered by 30%).
These and other ingredients have break-through potential, but obviously more and better clinical evaluation is needed. At the very least, marketing efforts should identify published work with full reference information, not just the name of the university. Citing reviews is not a substitute for citing the original work. And in vitro and animal work should not be commingled with human results in such a way as to mislead readers into thinking all results are from human studies.
It's nice to think that everything that works at all works fast, but where is the evidence? Most glucosamine studies made an initial measurement at 4 to 6 weeks (GAIT was at 6 weeks), so no help there. Four studies, however, made weekly assessments starting at week 1. Beginnings of reductions in pain were seen at the end of the first week, with progressive improvement during subsequent weeks. Thus, ingredients paired with glucosamine may really contribute to faster relief activity or just may be reflecting the quicker than expected action of the glucosamine component of the mixture.
Ingredients combined with glucosamine and touted as promoting fast action include hyaluronic acid, Chinese skullcap, willow bark extract, black catechu and antioxidant vitamins. While in truth these combination products may start working within a week, they should not be marketed as 6X, 8X or 12X faster than glucosamine given the evidence that glucosamine alone may work faster than is usually supposed.
In 2004, FDA rejected two health claim applications for glucosamine. Most of the evidence was rejected on a procedural basis-as the subjects already had osteoarthritis when they were enrolled, none of the positive results supported the theory that glucosamine reduced the risk of developing osteoarthritis. There was hope that the second part of GAIT, on slowing loss of knee cartilage, would support a reapplication, but the results were equivocal.
Since the glucosamine rejection (no application was ever made for chondroitin) there have been several regulatory actions against companies that went beyond the allowable "structure/function" language for osteoarthritis health statements. For instance, in 2008 the National Advertising Division (NAD) of the Council of Better Business Bureaus issueda statement advising Park Labs to revise advertising for Reparagen, not so much for lack of evidence, but for statements that it was guaranteed to help within 7 days, and that it worked 8X faster than glucosamine and chondroitin.
NAD also started a review process of advertising claims for Supple Dietary Supplement Drink, a product of Supple Beverages. But it halted the effort when it learned that Supple was already the subject of a federal government investigation.
In recent years FDA has issued numerous Warning Letters and at times seized products for label claims about treating arthritis. Best practical advice here is have a modicum of evidence behind the product or the product's ingredients, avoid treatment claims and avoid specific superiority claims over other products.
GAIT was started in 2000 as a to-be-definitive clinical trial of A) glucosamine 1500 mg, B) chondroitin 1200 mg, C) the two in combination, D) a COX-2 inhibitor (Celebrex), and E) a placebo group. Measurements included pain relief and slowing loss of knee cartilage. Subjects included patients with moderate to severe osteoarthritis of the knee. Eight years and $13 million later the results were equivocal.
GAIT I, the pain relief part, reported on reduction of pain at 6 months. Only in the subset of patients who started with severe OA was glucosamine and chondroitin in combination significantly better than placebo. GAIT II, the joint space narrowing evaluation at 24 months, was reported as not statistically significant.
In the original proposal, the investigators had anticipated a pain relief placebo group response of 30% to 40%. As a result, trial size was set accordingly. Recently published "placebo effect" research reports that as the expectation of successful treatment increases, placebo success increases. In GAIT, four of five treatments were ones for which the subjects had high expectations. Furthermore, positive responses in placebo groups rise when trials are large or long, both descriptive of GAIT. In fact, the percentage of patients reporting pain relief was A) 64%, B) 65%, C) 67%, D) 70% and E) 60%. That high E response (placebo group) squeezed the ability of the treatment groups to be statistically significant.
In contrast to pain measurement, placebo effects are thought to be small to non-existent on objective measures, such as cartilage loss. The problem plaguing GAIT II was the small group size. The planned study group size was 159 subjects per group based on an expected joint space loss of 0.200 mm per year in untreated patients. Five studies published after GAIT started averaged a loss of 0.106 mm per year. In GAIT II the placebo group came in at 0.083 mm per year, and only about 70 subjects per group completed the two-year study. In the end, the glucosamine effect was actually very good-a 92% reduction compared to placebo, but the groups were too small to achieve statistical significance.
Arthritis has not been cured by modern medicine. And presently it is not adequately treated. Clinicaltrials.gov lists 695 trials for osteoarthritis that are planned, ongoing or recently concluded, so lack of success is not for lack of trying.
The disappointing results of GAIT I and II were not the feared nails in the coffin for glucosamine or chondroitin everyone thought they'd be. And while raw material costs continue to pressure dietary supplement sales, researchers have stepped up for chondroitin to show that 800 mg a day is sufficient.
Further, the observation that combining glucosamine and chondroitin may not have any advantage over either alone opens the market to inventive reformulating with other ingredients.
Anti-inflammatory ingredients as being potentially therapeutic-not just palliative-got a left-handed boost from GAIT II, wherein Celebrex relieved pain at 6 months (as expected for a COX-2 inhibitor), but surprisingly at 2 years had also resulted in one-third less loss of cartilage thickness in arthritic knees. Five years ago, before COX-2 inhibitor drugs imploded in a mass of serious side effects, many herbal pain reliever products were claiming COX-2 inhibition. Today, with licofelone, the furthest advanced combined COX/LOX inhibitor drug not yet approved for use, a few non-drug products are claiming LOX inhibition as mechanism. Proof, as always, will depend on well designed, well executed, and well reported clinical trials.
About the author: David Mark, PhD, is president of dmark consulting LLC (est. 2004), a provider of research management services to functional food and dietary supplement companies. He can be reached at 978-897-0890; E-mail: david@dmarknutrition.com; Website: www.dmarknutrition.com.
References furnished upon request.
By the time I was old enough to remember, my grandfather had osteoarthritis in his knees and his right hand. He couldn't even open his hand to shake hands. At work, he held his hand in warm water until the joints loosened enough to hold a knife. He could not tie his own shoes or button his own shirt, so my grandmother had to do that for him.
When one person in a family has arthritis, the whole family has custody of the arthritis. Or maybe the arthritis has custody of them. There is the help needed in the kitchen, in the yard, loss of income when a job is no longer possible, weight gain due to inability to exercise, compromised vacation and retirement plans, not being able to help care for grandchildren, and the potential cost and burden of surgery and recovery.
The Sobering Statistics
The Centers for Disease Control and Prevention (CDC), Washington, D.C., conduct frequent surveys of the health status of people living in the U.S., which include questions on joint pain. The results are tabulated in an annual report "Health, United States." Figure 1 shows some of the joint pain results from the 2006 survey, as contained in the 2008 report. The exact survey question was: "During the past 30 days, have you had any symptoms of pain, aching or stiffness in or around a joint?"
The prevalence of osteoarthritis in the U.S. is expected to worsen, as obesity increases the risk of developing osteoarthritis and the prevalence of obesity, as defined by BMI > 30, increased from 13% in 1960 to 35% in 2006. The relationship makes sense from a biomechanical stress perspective, but hand osteoarthritis also correlates with obesity, suggesting a hormonal contribution too. It has been hypothesized that the hormone leptin, produced by fat cells, may contribute to osteoarthritis risk. This might also explain why women are more likely to develop osteoarthritis, as they usually have a higher percent body fat than men. Bariatric surgery leading to major weight loss often completely relieves osteoarthritis pain.
New Evidence for Established Ingredients
Glucosamine. There has been more news than expected on glucosamine since the publication of GAIT (Glucosamine Arthritis Intervention Trial). Interestingly, the Glucosamine Unum in Die Efficacy (GUIDE) trial reported less knee pain at six months, while another trial also with >100 subjects reported glucosamine as not significantly better than placebo for hip pain at 24 months. Both trials used glucosamine sulfate.
GAIT was criticized for using glucosamine hydrochloride instead of the more studied sulfate form. The rationale from GAIT researchers was that because both HCl and sulfate are disassociated from glucosamine prior to absorption, it should not matter. One trial out of Russia reported a benefit for glucosamine HCl, and two trials out of China did head-to-head comparisons of glucosamine HCl and glucosamine sulfate and reported no differences between the two. Unfortunately, neither had a placebo group. Bruyere et al reported on a post-trial follow-up. Subjects were contacted an average of five years after the glucosamine sulfate clinical trial had ended. Researchers found that knee replacements were more common in the placebo group (14.5%) than in the glucosamine group (6%).
Chondroitin. The chondroitin category, perhaps not surprisingly, has also received more attention since GAIT, and the news has been interesting, especially as support for 800 mg daily continues to gain traction. GAIT concluded that chondroitin sulfate at 1200 mg daily was no better at pain relief compared to placebo. But two trials concurrent with GAIT reported less knee joint-space narrowing over years of taking 800 mg of chondroitin daily. In 2007, Mazieres et al reported on 307 patients getting either chondroitin at 1000 mg per day or placebo for 24 weeks. The results either favored or trended for chondroitin. As with GAIT, the study was handicapped by a high placebo group response (56% using the OARSI test). The GAIT researchers reported that patients with less severe OA at trial enrollment had a statistically significant improvement for knee swelling for chondroitin over placebo. GAIT II reported 35% less loss of knee cartilage over two years in the chondroitin group compared to control, but the difference was not statistically significant because of the smaller than planned number of subjects completing the study. Finally, the "STOPP" trial enrolled 622 patients to take either chondroitin at 800 mg per day or placebo for two years. More than two-thirds of subjects completed the trial. Importantly, the "STOPP" chondroitin trial reached its two-year endpoint with three times as many patients as GAIT and was strongly positive for less cartilage loss in knee joints, even though subjects were only taking 800 mg daily.
Specifically, joint space loss was 0.150/year in the control group compared to 0.035/year in the chondroitin group, representing a 77% improvement. And pain relief occurred significantly faster in the chondroitin group compared to the control group. According to clinicaltrials.gov, there are six more chondroitin clinical trials either underway or starting soon, so potential remains for more good (or bad) news.
Glucosamine + Chondroitin. Given that more than 100 clinical trials have looked at either glucosamine or chondroitin, remarkably few looked at the two ingredients in combination. Perhaps this is because ingredient companies foot the research costs and few make both products. Unfortunately GAIT poked holes in the combination theory, as glucosamine (1500 mg) plus chondroitin (1200 mg) was found to be no better than either alone for pain relief. And jarringly, the combination was worse than placebo for loss of knee cartilage over two years, whereas glucosamine or chondroitin alone were better than placebo (none of the differences were statistically significant, but still).
A small study published in a 2007 edition of Osteoarthritis Cartilage reported no benefit for glucosamine and chondroitin in combination in terms of pain or knee strength at six months.
There are two more small trials from about 10 years back showing a pain relief benefit for the combination, but that's it.
Rose Hip (Rosa canina). A recent meta-analysis identified three randomized clinical trials of rose hip powder, at 5 grams per day, for osteoarthritis. A total of 306 subjects were enrolled. Patients getting the rose hip product had less pain and used less pain medication. All "responders to therapy" were 61% for rose hip powder and 42% for placebo (the latter is typical for arthritis pain trials). All of the clinical trials were sponsored by Hyben-Vital International, the manufacturer of a rose hip powder branded as LitoZin in some countries. The proposed mechanism of action includes antioxidant function and anti-inflammatory action of a specific galactolipid called "GOPO." But GOPO content is not specified.
Green-Lipped Mussel (GLM). New Zealand green-lipped mussels (Perna canaliculus) are both a food and a source for a dietary supplement for arthritis. Other arthritis products such as glucosamine, chondroitin, collagen and omega 3 fatty acids are also sourced from animals (although glucosamine and omega 3 are also available in vegetarian forms), so the idea of an animal product with a health benefit is not entirely foreign.
The story begins with an observation that coastal-living Maori had a lower risk of arthritis than inland dwellers. Whether true or not, this led to tests of freeze-dried, powdered mussels, identification and then solution via stabilization of an auto-oxidation problem akin to that seen for stabilized rice bran, and finally, a lipid extract of GLM promising to deliver the active fraction in a smaller volume without risk of triggering protein-based allergies. A recent review captured four clinical trials: two powder, one extract and one comparing extract to powder. The last did not have a placebo control. The review concluded that GLM may be of benefit in mild to moderate osteoarthritis, but because of the variation in product tested and experimental design, no definitive conclusion was reached. There is also an 8-week, open-label trial of the extract with 60 patients, but because of the known large placebo-response seen in osteoarthrititis trials it is difficult to judge the value of the results.
The proposed mechanism of action for GLM extract is the EPA and DHA content having an anti-inflammatory activity, possibly by inhibiting the lipoxygenase (LOX) and cyclooxygenase (COX) pathways of prostaglandin synthesis. The review cites evidence from animal models of adjuvant- or collagen-induced arthritis, and in fact it is these animal experiments, which are used in marketing claims boasting that GLM extract is far, far more potent than fish oils or plant oils. Given dosing of 200 mg per day (50 mg per softgel, 4 times daily) it is difficult to understand how the omega 3 fatty acids can be responsible for the claimed benefits, as osteoarthritis is not thought to be responsive to omega 3 fatty acids. Animal studies at Hong Kong Polytechnic University are reporting preliminary evidence for GLM extract affecting protein and cytokine expression, suggesting that other molecules in GLM extract may be active through other mechanisms.
Cat's Claw & Maca (Reparagen). Vincaria (Ua de Gato, translation: cat's claw) is a traditional-use, multi-functional herbal medicine used in parts of Peru. An osteoarthritis trial published in 2001 compared 100 mg per day of freeze-dried cat's claw to placebo. Improvement was apparent in one week. More recently, a larger, longer trial was completed comparing a cat's claw and maca combination (Reparagen) to glucosamine sulfate (no placebo). This well designed and executed trial reported WOMAC and VAS scores during a period of 8 weeks. The two treatments had near-identical effects on knee pain, stiffness and function. Benefits for both were seen at one week and continued to improve. The Reparagen patients used 10-15% less acetaminophen as pain relief medication during the trial.
Boswellia. The cumulative evidence for Boswellia serrata (frankincense) extracts recently crossed the line from wishful to something more substantial. The active ingredients are tentatively identified as boswellic acids, although there is dispute whether 11-keto-β-boswellic acid (KBA) or 3-acetyl-11-keto-β-boswellic acid (AKBA) are both active, or equally active. There are two commercialized boswellic extract ingredients with supporting clinical evidence-WokVel and 5-LOXIN. Both claim action as inhibitors of the 5-lipoxygenase enzyme (LOX) in the pathway of production of inflammatory leukotrienes. The old clinical work used 700-1000 mg per day of crude extracts with low boswellic acid content. However, the new products have higher concentration and lower amounts needed.
In vitro and animal toxicology testing, plus the minimal adverse events reported in clinical trials suggest that the safety profile is good. A concise review in a 2008 edition of the British Medical Journal by Dr. E. Ernst captured all of the clinical trials with control groups, including Sander et al on rheumatoid arthritis and Kimmakar, Sontakke and Sengupta on osteoarthritis. The rheumatoid arthritis results trended favorable and the osteoarthritis trial results were statistically significant. In vitro and animal work shows inhibition of LOX and also of metalloproteinase enzymes implicated in the degeneration of joint tissues.
Collagen. Ten grams a day for 24 weeks. That's what it took for young (20-year-old) athletes to feel improvements in joint pain, compared to what was reported by the placebo group. Measurements at 6, 12 and 18 weeks did not pick up the difference, although there was a trend at 18 weeks. Previous studies with the same collagen hydrolysate, summarized in a 2006 review, reported similar improvements in older subjects with knee osteoarthritis, however most of those studies were either without a control group or difficult to interpret. However, a very recent placebo-controlled trial in osteoarthritis patients offered more confirmation.
The mechanism for collagen supporting healthy joint cartilage may be stimulus, supply of raw materials or both. In vitro work suggests chondrocytes exposed to hydrolyzed collagen increase production of type II cartilage. In vivo work confirms that some of the collagen is absorbed as peptide chains rather than first being reduced to free amino acids. A single-dose human trial (Walrand 2008) demonstrated that 10 grams of collagen hydrolysate resulted in elevated plasma levels of collagen amino acids.
Eggshell membrane. Chicken eggs are assembled from the inside out. The yolk accrues to full size inside a membrane, then albumin is added outside the yolk and inside a double layer membrane, and then the calcified shell is added. Eggshell membrane can be separated, dried and powdered, and is currently being sold as a dietary supplement ingredient, called NEM ("Natural Eggshell Membrane"). The composition of eggshell membrane is primarily collagen, plus hyaluronic acid, chondroitin and other substances.
Two unpublished, open-label clinical trials enrolled a modest number of subjects with joint pain (any joint, not limited to osteoarthritis) for 30 days at 500 mg per day. There were improvements from baseline, but again given the large placebo effect for joint pain, not conclusive. Most recently, NEM was tested in a randomized, multicenter, double-blind, placebo-controlled study, which investigated the effectiveness of NEM for the relief of pain and stiffness associated with moderate osteoarthritis of the knee. Published in the April edition of the journal Clinical Rheumatology, the study investigated 67 patients who were randomly assigned to receive either 500 mg of NEM or placebo daily for 8 weeks. Supplementation with NEM produced an absolute rate of response that was statistically significant (up to 27%) versus placebo at all time points for both pain and stiffness, but was not significantly improved for function and overall WOMAC scores, although trending toward improvement. Rapid responses were seen for mean pain subscores (16% reduction) and mean stiffness subscores (13%) occurring after only 10 days of supplementation.
Not Quite Proven, but Promising
Methylsulfonylmethane (MSM) was a hot new ingredient a few years ago, but nothing has been reported since a 2006 study showing knee pain and function improvements after dosing with 6000 mg per day.
A root extract from Mojave yucca (Yucca schidigera) has traditional use for arthritis along with anti-inflammatory applications. A review of composition was published in 2006, which noted that human evidence for arthritis stems from three uncontrolled trials published in the 1970s, but they were not of sufficient scientific rigor.
Boron's laurels rest on one very small clinical trial conducted in the 1980s and an observation from the same investigator (Newnham) that countries with high dietary boron intake have low prevalence of arthritis. Most of the boron supplements on the market are at 3 mg per day for bone health-a few mention joint health too. Boron intake from foods is estimated at 1-2 mg per day.
Cetylated fatty acids have been evaluated in four osteoarthritis clinical trials, all with favorable results. But two were topical application, two were oral, and of the latter, only one (conducted in India) has been published so far. The authors theorize that LOX inhibition may be involved.
Hops products are usually marketed for mild calming, sedating or sleep promoting qualities, but there is one unpublished, small clinical trial of a hops extract in osteoarthritis patients. Over a 14-day period, 1000 mg per day reduced some of the WOMAC pain scores on some of the days. The proposed mechanism is as an anti-inflammatory and possibly a COX inhibitor.
We all know 41-50 count shrimp means so many per pound, but how about 1000 count? That's the size (less than half a gram each) of the shrimp-like Antarctic krill processed to make krill oil. This relatively new dietary supplement ingredient is a source of omega 3 fatty acids and the carotenoid astaxanthin. One clinical trial published in 2007 provided subjects with 300 mg krill oil per day for 30 days, and measured changes in C-reactive protein (inflammation biomarker) and WOMAC (a standard arthritis score). Improvements were better than what was seen in the placebo group. Interpretation of the study is weakened by the fact that enrollees, all selected for elevated CRP, either had cardiovascular disease, and/or osteoarthritis, and/or rheumatoid arthritis. The dogma for fish oil sourced omega 3 treatments is that it helps with rheumatoid arthritis, but not osteoarthritis.
A double-blind, placebo-controlled study published in the journal Nutrition Research (Farid 2007) showed that Pycnogenol, an antioxidant plant extract from the bark of the French maritime pine tree, improved physical function by 52% in patients suffering from osteoarthritis. In the study, 35 volunteers (average age 42) were randomly assigned a daily dose of Pycnogenol (50 mg, 3times daily) or placebo for 3 months. Patients reported arthritic pain using the WOMAC index after 30, 60 and 90 days. Participants also were instructed to indicate the frequency and dosage of NSAIDS and COX-2 inhibitor usage.
After 2 months of supplementation, physical function and pain scores improved in the Pycnogenol group. After 3 months in the Pycnogenol group, there was a reduction of 43% in pain, 35% in stiffness, 52% in physical function subscales and 49% composite WOMAC. The placebo group showed no significant scores throughout the entire study. Additionally, further reduction in the number of NSAIDS and COX-2 inhibitor pills and number of days taking medication was noted in the Pycnogenol group.
A recent study published in Redox Report (Belcaro 2008) showed Pycnogenol's anti-inflammatory potency further to improving osteoarthritis symptoms and pain by significantly lower plasma levels of CRP. With disease progression of osteoarthritis, the inflammation may reach a level where it no longer is limited to the affected joint and stresses the organisms, which in turn increases the inflammatory marker CRP in the blood.
This study investigated a subset 55 patients from a previous osteoarthritis study with 156 patients who had significantly elevated CRP levels. Treatment consisted of two tablets daily of either 50 mg Pycnogenol or placebo. Blood specimens were drawn at baseline of the initial study and again after three-month treatment. Results showed that Pycnogenol significantly lowered CRP from average 3.9 mg/L at baseline to 1.1 mg/L, reflecting essentially healthy levels. In the placebo group a marginal lowered CRP level was detected. Other blood parameters indicative of acute inflammation likewise decreased with Pycnogenol, such as fibrinogen (lowered 37%) and reactive oxygen species (lowered by 30%).
These and other ingredients have break-through potential, but obviously more and better clinical evaluation is needed. At the very least, marketing efforts should identify published work with full reference information, not just the name of the university. Citing reviews is not a substitute for citing the original work. And in vitro and animal work should not be commingled with human results in such a way as to mislead readers into thinking all results are from human studies.
'Starts Working in 7 Days'
It's nice to think that everything that works at all works fast, but where is the evidence? Most glucosamine studies made an initial measurement at 4 to 6 weeks (GAIT was at 6 weeks), so no help there. Four studies, however, made weekly assessments starting at week 1. Beginnings of reductions in pain were seen at the end of the first week, with progressive improvement during subsequent weeks. Thus, ingredients paired with glucosamine may really contribute to faster relief activity or just may be reflecting the quicker than expected action of the glucosamine component of the mixture.
Ingredients combined with glucosamine and touted as promoting fast action include hyaluronic acid, Chinese skullcap, willow bark extract, black catechu and antioxidant vitamins. While in truth these combination products may start working within a week, they should not be marketed as 6X, 8X or 12X faster than glucosamine given the evidence that glucosamine alone may work faster than is usually supposed.
Recent Regulatory Actions
In 2004, FDA rejected two health claim applications for glucosamine. Most of the evidence was rejected on a procedural basis-as the subjects already had osteoarthritis when they were enrolled, none of the positive results supported the theory that glucosamine reduced the risk of developing osteoarthritis. There was hope that the second part of GAIT, on slowing loss of knee cartilage, would support a reapplication, but the results were equivocal.
Since the glucosamine rejection (no application was ever made for chondroitin) there have been several regulatory actions against companies that went beyond the allowable "structure/function" language for osteoarthritis health statements. For instance, in 2008 the National Advertising Division (NAD) of the Council of Better Business Bureaus issueda statement advising Park Labs to revise advertising for Reparagen, not so much for lack of evidence, but for statements that it was guaranteed to help within 7 days, and that it worked 8X faster than glucosamine and chondroitin.
NAD also started a review process of advertising claims for Supple Dietary Supplement Drink, a product of Supple Beverages. But it halted the effort when it learned that Supple was already the subject of a federal government investigation.
In recent years FDA has issued numerous Warning Letters and at times seized products for label claims about treating arthritis. Best practical advice here is have a modicum of evidence behind the product or the product's ingredients, avoid treatment claims and avoid specific superiority claims over other products.
What Went Wrong with GAIT?
GAIT was started in 2000 as a to-be-definitive clinical trial of A) glucosamine 1500 mg, B) chondroitin 1200 mg, C) the two in combination, D) a COX-2 inhibitor (Celebrex), and E) a placebo group. Measurements included pain relief and slowing loss of knee cartilage. Subjects included patients with moderate to severe osteoarthritis of the knee. Eight years and $13 million later the results were equivocal.
GAIT I, the pain relief part, reported on reduction of pain at 6 months. Only in the subset of patients who started with severe OA was glucosamine and chondroitin in combination significantly better than placebo. GAIT II, the joint space narrowing evaluation at 24 months, was reported as not statistically significant.
In the original proposal, the investigators had anticipated a pain relief placebo group response of 30% to 40%. As a result, trial size was set accordingly. Recently published "placebo effect" research reports that as the expectation of successful treatment increases, placebo success increases. In GAIT, four of five treatments were ones for which the subjects had high expectations. Furthermore, positive responses in placebo groups rise when trials are large or long, both descriptive of GAIT. In fact, the percentage of patients reporting pain relief was A) 64%, B) 65%, C) 67%, D) 70% and E) 60%. That high E response (placebo group) squeezed the ability of the treatment groups to be statistically significant.
In contrast to pain measurement, placebo effects are thought to be small to non-existent on objective measures, such as cartilage loss. The problem plaguing GAIT II was the small group size. The planned study group size was 159 subjects per group based on an expected joint space loss of 0.200 mm per year in untreated patients. Five studies published after GAIT started averaged a loss of 0.106 mm per year. In GAIT II the placebo group came in at 0.083 mm per year, and only about 70 subjects per group completed the two-year study. In the end, the glucosamine effect was actually very good-a 92% reduction compared to placebo, but the groups were too small to achieve statistical significance.
Conclusions
Arthritis has not been cured by modern medicine. And presently it is not adequately treated. Clinicaltrials.gov lists 695 trials for osteoarthritis that are planned, ongoing or recently concluded, so lack of success is not for lack of trying.
The disappointing results of GAIT I and II were not the feared nails in the coffin for glucosamine or chondroitin everyone thought they'd be. And while raw material costs continue to pressure dietary supplement sales, researchers have stepped up for chondroitin to show that 800 mg a day is sufficient.
Further, the observation that combining glucosamine and chondroitin may not have any advantage over either alone opens the market to inventive reformulating with other ingredients.
Anti-inflammatory ingredients as being potentially therapeutic-not just palliative-got a left-handed boost from GAIT II, wherein Celebrex relieved pain at 6 months (as expected for a COX-2 inhibitor), but surprisingly at 2 years had also resulted in one-third less loss of cartilage thickness in arthritic knees. Five years ago, before COX-2 inhibitor drugs imploded in a mass of serious side effects, many herbal pain reliever products were claiming COX-2 inhibition. Today, with licofelone, the furthest advanced combined COX/LOX inhibitor drug not yet approved for use, a few non-drug products are claiming LOX inhibition as mechanism. Proof, as always, will depend on well designed, well executed, and well reported clinical trials.
About the author: David Mark, PhD, is president of dmark consulting LLC (est. 2004), a provider of research management services to functional food and dietary supplement companies. He can be reached at 978-897-0890; E-mail: david@dmarknutrition.com; Website: www.dmarknutrition.com.
Disclosure: Dr. Mark has as clients dietary supplement companies, which market arthritis-targeting products containing chondroitin, fish oil, glucosamine, hyaluronic acid, MSM and willow bark.
References furnished upon request.