Osteoarthritis (OA) is a progressive joint disease characterized by varying degrees of pain and degeneration. The disease affects the structural and functional integrity of articular cartilage—the thin load-bearing tissue lining the ends of long bones—as well as the adjacent bone and other joint tissues. The clinical symptoms of OA include joint pain, stiffness, and swelling, which may lead to muscle weakness and impaired physical function. OA is generally diagnosed radiographically by bony changes, including osteophyte formation, cysts, subchondral sclerosis, and joint space narrowing. These clinical and radiographic features, which can often manifest independent of each other, are most commonly diagnosed in the knee, hip, and hand joints. Elderly individuals are most at risk for OA, and the incidence of OA increases significantly with age. Other risk factors include joint trauma, body mass index (i.e., obesity), and heritable genetic factors. Of all joint diseases, OA is the most common, affecting an estimated 15% of the U.S. population (5).
Under normal physiological conditions, articular cartilage provides a nearly frictionless surface for the transmission and distribution of joint loads, exhibiting little or no wear over decades of use (9). This remarkable function is determined by the structure and composition of the extracellular matrix (ECM), which is primarily water. The remaining solid matrix is largely composed of a crosslinked network of collagen (mainly type II) and proteoglycans such as aggrecan (Fig. 1). Osteoarthritis is characterized by an imbalance between the anabolic and catabolic activities of the cartilage cell population, the chondrocytes, suggesting that alterations in cellular metabolism contribute to the onset and progression of the disease (6). Chondrocytes regulate their metabolic activity in response to mechanical, chemical, or electrical signals in their local microenvironment. In doing so, articular cartilage can alter its structure and composition to accommodate the physical demands of the body.
Adult articular cartilage has a limited ability for self-repair, because the tissue is avascular, aneural, alymphatic, and sparsely populated by cells. Nevertheless, the components of the ECM are in a state of slow turnover, maintained in homeostatic balance by the catabolic and anabolic events of the chondrocytes. These activities are controlled through the processing of both genetic and environmental information (e.g., growth factors, cytokines, and ECM composition). In particular, recent studies have shown that the mechanical stress environment of the joint is an important factor that influences (and presumably regulates) the activity of the chondrocytes in vivo (8).
Hypothesis 1: Moderate Mechanical Loading Is Necessary for Cartilage Homeostasis
In vitro, in vivo, and epidemiological studies provide a wide range of approaches to understand the effect of mechanical loading on cartilage metabolism and homeostasis. Each level of study provides specific advantages and disadvantages that need to be taken into account in the interpretation of results taken from each of these different levels. For example, in vitro models are advantageous for examining mechanisms governing how specific mechanical or biochemical factors affect cartilage metabolism. Some disadvantages of in vitro models, though, are the inability to replicate the various characteristics of the in vivo loading environment and the relatively brief time span of investigation (e.g., hours to months). In vivo animal studies improve upon these deficiencies by emulating physiologically relevant loading conditions to study long-term (e.g., weeks to years) tissue changes associated with growth, remodeling, or aging. However, in these studies it is difficult to accurately determine and control the cartilage loading history or the potential interaction of systemic or local soluble mediators (e.g., hormones, cytokines, enzymes) on cartilage metabolism. These disadvantages also apply to clinical and epidemiological studies. Nonetheless, clinical studies of patient populations may provide the best means for evaluating the physiological relevance of mechanical factors on integrated disease pathways and treatments.
Evidence From In Vitro Studies: Mechanical Loading of Cartilage Explants
During normal daily activities, articular cartilage is exposed to cyclic bouts of mechanical stress consisting of both static (i.e., constant) and dynamic (i.e., time-varying) components. Therefore, to understand how these in vivo joint loading patterns affect the cellular regulation of catabolic and anabolic activities, investigators have studied the biological response of cartilage explants to both “static” and “dynamic” loading.
Cartilage explants subjected to static compression exhibit a significant suppression of metabolic activity that is dependent on the magnitude of applied stress and particularly affects the biosynthesis of aggrecan and collagen (8). In contrast to static loading, dynamic compression of articular cartilage explants at certain frequencies can increase chondrocyte anabolism. Cyclic compression at frequencies of 0.01 to 1 Hz and strain amplitudes of 1 to 5% stimulate aggrecan biosynthesis above control values (Fig. 2A). Lower frequencies and amplitudes, however, do not appear to affect biosynthesis rates. Several studies indicate that above a certain threshold frequency (∼0.01—0.1 Hz), aggrecan biosynthesis increases in a dose-dependent manner with loading frequency. Cyclic loading also increases the synthesis of other components of the ECM, such as cartilage oligomeric matrix protein and fibronectin (8). It is important to note, however, that even under carefully defined explant loading configurations, it is difficult to separate the influence of various biophysical parameters on cell physiology, because many of the mechanical and physicochemical factors that are known to affect chondrocytes are inextricably coupled to one another within the cartilage ECM. Furthermore, significant spatial and temporal variations may exist in the stress–strain and fluid–flow environments of the chondrocyte because of local variations in fluid transport and tissue deformation, as well as heterogeneities in cell and tissue properties (8).
Evidence From In Vivo Studies: Animal Studies of Altered Physical Activity
Animal studies of altered joint loading clearly illustrate that biomechanical factors affect cartilage metabolism and play a role in the development of OA. Some of the most consistent animal models of cartilage degeneration and OA involve surgically impairing joint stability, such as transection of the anterior cruciate ligament or removal of the meniscus. However, changes in joint loading through altered physical activity levels, without altering joint stability, may elicit either an anabolic or catabolic response depending on the intensity, duration, and age of exercise onset (2).
For example, moderate exercise, particularly in younger animals, produces anabolic changes in cartilage, such as increased proteoglycan content, decreased proteoglycan extractability, and increased cartilage thickness (Fig. 2B). These anabolic changes enhance the load-bearing properties of cartilage and may help explain how lifelong increases in physical activity levels in dogs do not increase the incidence of OA. In rodents, however, the effect of lifelong exercise on cartilage metabolism and integrity may depend on whether exercise is forced or voluntary. Mice forced to exercise on a treadmill developed more severe cartilage lesions than sedentary controls; however, no difference in lesion occurrence was found between wheel-exercised and sedentary mice despite weekly wheel-running distances being two to five times greater than treadmill distances. These findings may reflect differences in joint loading between wheel and treadmill exercise, or they may indicate the influence of other systemic factors, such as those related to stress or cardiovascular fitness levels, on cartilage metabolism.
High-intensity exercise or a sudden increase in exercise at an older age produces catabolic changes in cartilage that eventually lead to OA (2). These degenerative changes include a decreased collagen network structure, site-specific proteoglycan loss, and reduced cartilage stiffness. Severe inactivity (e.g., immobilization) also produces catabolic changes, such as reduced cartilage thickness and proteoglycan content. These changes, however, are reversible on return to normal ambulation if the period of immobilization is not too extended. This reversal of cartilage catabolism appears to be possible for changes in cartilage structure and composition induced by immobilization, but not by other altered loading conditions, because immobilization primarily affects the proteoglycan component and does not irreversibly alter the collagen network structure.
Evidence from Epidemiological Studies: The Role of Physical Activity
The risk of OA is increased by numerous factors related to physical activity, such as occupation, type and level of sports participation, and exercise intensity levels. In particular, occupations and sports that involve heavy loading and twisting of the knee are potent risk factors for OA (5). However, elevated daily physical activity levels do not appear to be a risk factor for OA (2). When accounting for known risk factors of OA (e.g., age, body mass index, knee injury, smoking, and education), habitual physical activity levels in middle age do not correlate with future incidence of OA. Retrospective and prospective cohort studies also generally support the conclusion that moderate, habitual exercise of uninjured joints does not increase the risk of OA (10).
Of particular interest is the question of whether running increases the risk of OA. Prospective cohort studies indicate that moderate intensity long-distance running is not harmful to knee joints. However, results from retrospective longitudinal and cohort studies are mixed. This discrepancy may in part be explained by the different diagnostic criteria (i.e., radiographic vs clinical) used among the studies. In addition, the average running distance and speed were notably different between the studies supporting a positive or negative effect of running on OA. A comparison of these studies suggests that recreational and mid-intensity running does not increase the risk for OA; however, competitive, high-intensity running, especially if practiced for many years, probably increases OA risk (10) (Fig. 2C).
The importance of mechanical loading in maintaining healthy joints has been recognized for quite some time, as captured in a popular quote by L. Sokoloff (as quoted within (10)): “Cartilage can survive in a large range of solicitations, but below or beyond, it will suffer.” More than 30 yrs later, in vitro, in vivo, and epidemiological studies continue to support the hypothesis that moderate mechanical loading is necessary to maintain healthy articular cartilage. If joints are insufficiently loaded (e.g., immobilization or purely static loading), cartilage metabolism shifts in favor of catabolism. Similarly, acute (i.e., traumatic) or chronic (i.e., lifestyle) high-intensity joint loading leads to cartilage degeneration and OA. Recent studies suggest that a primary way in which mechanical stress balances catabolic and anabolic processes to maintain cartilage homeostasis is by modulating joint inflammation.
Hypthesis 2: Osteoarthritis Develops From a Complex Interaction of Inflammatory and Biomechanical Pathways
Although originally characterized as a noninflammatory disease, current research indicates that proinflammatory mediators play a central role in the pathogenesis of OA (Fig. 3). Proinflammatory cytokines, such as interleukin-1 (IL-1) and tumor necrosis factor-α (TNF-α), are robust mediators of cartilage matrix degradation and joint inflammation (6). These cytokines disturb the normal remodeling activities of the chondrocytes by increasing the production of matrix metalloproteinases and aggrecanases that cleave the collagen and proteoglycan networks, respectively, as well as diminishing anabolism of ECM components. The proinflammatory effects of these cytokines appear to be mediated through the production of prostaglandin E2 (PGE2) and nitric oxide (NO), which are elevated in the tissues of osteoarthritic joints. Chondrocytes produce and respond to PGE2 and NO, allowing for both autocrine (i.e., intracellular) and paracrine (i.e., intercellular) feedback and regulation of these mediators. The importance of these inflammatory components of OA is also indicated by the symptomatic relief afforded by nonsteroidal antiinflammatory drugs, which target the cyclooxygenase (COX) enzymes that are responsible for the synthesis of prostanoids such as PGE2. Although the factor(s) responsible for the initiation and progression of cartilage catabolism in OA are not known, mechanical stress is a potential stimulus. Mechanical stress may act directly on the chondrocyte to increase the production of proinflammatory mediators, such as PGE2 and NO (7). Furthermore, the presence of proinflammatory cytokines and mediators (e.g., IL-1, TNF-α, PGE2, and NO) can have a significant effect on how mechanical stress affects cartilage metabolism (7).
Obesity and Osteoarthritis: Biomechanical and Inflammatory Factors Converge
Obesity is a major risk factor for developing OA of the knee and, to a lesser extent, the hip and feet (Fig. 4) (5). This relationship appears to be causal as obesity precedes the development of radiographic knee OA. The traditional explanation of obesity-induced OA is that the added axial load on the joints accelerates normal joint “wear and tear” that occurs with aging. Recent studies suggest that the onset of OA may be determined by a shift in load-bearing to less frequently loaded regions of the cartilage and subsequent progression of OA is caused by increased loading of these regions. In particular, varus/valgus laxity and malalignment of the knee are risk factors for the development and progression, respectively, of OA (5). Malalignment of the knee appears to increase the progression of OA by altering the local loading environment of the knee. For example, varus malalignment increases the load transmitted through the medial tibio-femoral compartment and is associated with an increased risk of OA progression in the medial compartment. However, mechanical factors alone do not seem to be sufficient to explain the relationship between OA incidence and obesity. Lifelong runners likely load their joints to a much greater extent than the added loads associated with obesity, yet running does not increase the risk of OA (10). Moreover, obesity also increases the risk of OA in non–weight bearing joints (e.g., hand), suggesting that the link between obesity and OA involves complex interactions of pathological changes in biomechanical and physiological factors at multiple levels (4).
Obesity, and in particular abdominal obesity, is associated with numerous metabolic disorders (e.g., insulin resistance/hyperinsulinemia, dyslipidemia, type 2 diabetes mellitus, and hypertension) that have collectively been called Metabolic Syndrome X (Fig. 4). Recent evidence suggests that this syndrome is an inflammatory condition (3): as body mass index (BMI) is positively correlated with a number of proinflammatory mediators, such as C-reactive protein (CRP), TNF-α, IL-1, IL-6, and NO. Abdominal fat appears to be the primary tissue source responsible for producing many of these inflammatory mediators and, as such, they have recently been called “adipokines.” In vitro, these adipokines produce catabolic changes within cartilage (6). Although the relationship between central body fat distribution and OA has not been established, longitudinal studies show that weight loss and decreased body fat reduce the risk of symptomatic OA (5).
It is difficult to assess the independent effects of abnormal joint loading and systemic inflammation on the pathogenesis of OA because BMI, physical activity levels, and inflammation co-vary (3). Recent studies from the Arthritis, Diet, and Activity Promotion Trial (ADAPT) illustrate the integrated nature of inflammation, diet, exercise, and OA (12). High levels of the soluble receptors of TNF-α were associated with lower physical function and increased knee OA (both symptomatic and radiographic) in older, obese adults. By participating in a diet-induced weight-loss and moderate exercise program, overweight and obese adults with knee OA lowered their concentrations of CRP, IL-6, and the soluble receptor for TNF-α1 and reported improved function and pain.
Systemic inflammation may mediate OA onset and progression via direct and indirect pathways. Elevated levels of cytokines circulating in the blood stream diffuse into the synovial fluid and, presumably, inflammatory mediators such as TNF-α and IL-1 act directly on chondrocytes to upregulate catabolic processes. Alternatively, obesity may indirectly mediate the development of OA by accelerating muscle senescence—a relative weakening of the muscles that is positively correlated with mild systemic inflammation and age (14). As a result, elevated systemic inflammation may accelerate quadriceps weakness, which is a potential risk factor for developing knee OA (5). Thus, inflammation may mediate OA onset and progression via both direct (e.g., as a soluble mediator upregulating chondrocyte catabolism that is amplified by excessive loading) and indirect (e.g., as a mediator of neuromuscular weakness) pathways.
Osteoarthritis: Better to be “Fit and Fat” or “Thin and Sedentary”?
Exercise is an effective nonpharmacological treatment for restoring physical function and reducing pain in individuals with OA (12). Exercise also has a protective effect against developing severe OA—after adjusting for known risk factors (e.g., age, BMI, knee injury, physical work stress, and smoking)—moderate recreational physical exercise is associated with a decreased risk of severe knee OA (11). The mechanism(s) governing the benefits of exercise for preventing and treating OA are not well understood, although greater cardiorespiratory fitness may play a central role by lowering systemic inflammation (14), improving neuromuscular coordination and strength, and improving pain coping and other psychosocial factors (1).
Many exercise effects appear to counteract the deleterious effects of obesity on OA. To what extent do cardiorespiratory fitness and obesity—and the interaction between these two factors—determine one’s risk of developing OA? This question has already been addressed for cardiovascular disease, and the answer appears to be that cardiorespiratory fitness trumps obesity: obese individuals with at least a moderate cardiorespiratory fitness level have lower disease rates compared with normal-weight, unfit peers (15). Comparable results using laboratory measures of cardiorespiratory fitness (e.g., maximal exercise treadmill test) have not yet been performed for OA patients. Using qualitative measures of physical activity as a surrogate for cardiorespiratory fitness, results from longitudinal studies suggest that moderate physical activity has either no effect or a protective effect on the risk of OA (2,5,13). Future prospective studies are needed to thoroughly evaluate the interaction between obesity and fitness to understand the role of each factor in determining OA risk.
It is important to note that OA is not a single entity, but rather a multifaceted family of diseases that involve a number of genetic and environmental factors. Nonetheless, there is strong in vivo evidence for the role of biomechanics in OA, and nearly all hypotheses on the etiopathogenesis of OA involve a role for mechanical factors in regulating the physiology and pathology of the joint. Furthermore, physical activity, with its effects on both local and systemic mechanical and inflammatory factors, may be a common modality for treating OA associated with a suite of other metabolic diseases. Thus, future therapies and interventions for OA may involve combinations of physical and pharmacologic therapies to address mechanisms involved in the disease process.
Supported by National Institutes of Health grants AR51672, EB01630, AR50245, AG15768, and AR48182. The authors thank Drs. Virginia Kraus, William Kraus, Lori Setton, Dan Schmitt, Beverley Fermor, Brice Weinberg, David Pisetsky, and Frank Keefe for many insightful discussions.
1. American College of Sports Medicine Position Stand. Exercise
and physical activity for older adults. Med. Sci. Sports Exerc.
2. Arokoski, J.P., J.S. Jurvelin, U. Vaatainen, and H.J. Helminen. Normal and pathological adaptations of articular cartilage to joint loading. Scand. J. Med. Sci. Sports.
3. Das, U.N. Is metabolic syndrome X
an inflammatory condition? Exp. Biol. Med. (Maywood
). 227:989–997, 2002.
4. Eaton, C.B. Obesity
as a risk factor for osteoarthritis: mechanical versus metabolic. Med. Health R. I.
5. Felson, D.T., R.C. Lawrence, P.A. Dieppe, R. Hirsch, C.G. Helmick, J.M. Jordan, R.S. Kington, N.E. Lane, M.C. Nevitt, Y. Zhang, M. Sowers, T. McAlindon, T.D. Spector, A.R. Poole, S.Z. Yanovski, G. Ateshian, L. Sharma, J. A. Buckwalter, K.D. Brandt, and J.F. Fries. Osteoarthritis: new insights. Part 1: the disease and its risk factors. Ann. Intern. Med.
6. Goldring S.R., and M.B. Goldring. The role of cytokines in cartilage matrix degeneration in osteoarthritis. Clin. Orthop. Relat. Res.
Oct(427 Suppl):S27–S36, 2004.
7. Guilak F., B. Fermor, F.J. Keefe, V.B. Kraus, S.A. Olson, D.S. Pisetsky, L.A. Setton, and J.B. Weinberg. The role of biomechanics
and inflammation in cartilage injury and repair. Clin. Orthop. Relat. Res.
8. Guilak F., R.L. Sah, and L.A. Setton. Physical regulation of cartilage metabolism. In: Basic Orthopaedic Biomechanics,
edited by V.C. Mow and W.C. Hayes. Philadelphia: Lippincott-Raven, 1997.
9. Guilak F., L.A. Setton, and V.B. Kraus. Structure and function of articular cartilage. In: Principles and Practice of Orthopaedic Sports Medicine,
edited by W.E.J. Garrett, K.P. Speer, and D.T. Kirkendall. Philadelphia: Lippincott Williams & Wilkins, 2000, pp. 53–73.
10. Lequesne M.G., N. Dang, and N.E. Lane. Sport practice and osteoarthritis of the limbs. Osteoarthritis Cartilage
11. Manninen P., H. Riihimaki, M. Heliovaara, and O. Suomalainen. Physical exercise
and risk of severe knee osteoarthritis requiring arthroplasty. Rheumatology (Oxford)
12. Miller, G.D., W.J. Rejeski, J.D. Williamson, T. Morgan, M.A. Sevick, R.F. Loeser, W.H. Ettinger, and S.P. Messier. The Arthritis
, Diet and Activity Promotion Trial (ADAPT): design, rationale, and baseline results. Control Clin. Trials.
13. Rogers, L.Q., C.A. Macera, J.M. Hootman, B.E. Ainsworth, and S.N. Blair. The association between joint stress from physical activity and self-reported osteoarthritis: an analysis of the Cooper Clinic data. Osteoarthritis Cartilage
14. Roubenoff, R. Exercise
and inflammatory disease. Arthritis Rheum.
15. Wei, M., J.B. Kampert, C.E. Barlow, M.Z. Nichaman, L.W. Gibbons, R.S. Paffenbarger, Jr., and S.N. Blair. Relationship between low cardiorespiratory fitness and mortality in normal-weight, overweight, and obese men. JAMA