Version May 2024
INTRODUCTION — Osteoarthritis (OA) is the most common form of arthritis. Globally, OA represented 62 percent of all arthritic conditions in 2017 to 2018 [1]. OA often affects the hands, hips, knees, feet, and spine. Pain is typically the symptom of OA that leads people to present to health care providers and subsequently receive a diagnosis of OA.
There is substantial morbidity associated with OA, including disability and reduced quality of life. OA is the leading cause of lower extremity disability in older adults, and hip and knee OA has accounted for 1.12 percent of all years lived with disability (YLD), an overall measure of disease burden. In Global Burden of Disease (GBD) studies, OA is consistently ranked among the leading contributors to global YLDs.
With the aging population and higher rates of obesity, this burden is expected to rise. The rapid increase in prevalence of OA will lead to a growing impact and major challenges for health care and public health systems.
The epidemiology and risk factors for OA will be reviewed here. Clinical manifestations, diagnosis, management, and investigational therapies are presented separately. (See "Clinical manifestations and diagnosis of osteoarthritis" and "Overview of the management of osteoarthritis" and "Pathogenesis of osteoarthritis" and "Management of hand osteoarthritis" and "Comorbidities that impact management of osteoarthritis" and "Management of knee osteoarthritis" and "Management of moderate to severe knee osteoarthritis" and "Overview of surgical therapy of knee and hip osteoarthritis".)
EPIDEMIOLOGY — The incidence and prevalence of osteoarthritis (OA) depend on the definition used, such as symptomatic, radiographic, self-reported, or doctor-diagnosed; it can be described pathologically, radiographically, or clinically. It should be noted that not everyone with radiographic OA is symptomatic. Due to the various definitions of OA that are used, the reported prevalence and incidence rates vary across studies. In general, prevalence estimates for symptomatic OA tend to be lower than radiographic OA as its presence is defined by a combination of symptoms such as pain and stiffness in addition to radiographic features.
●Incidence – Data modeling from the 2019 Global Burden of Disease study (GBD2019) estimated the age-standardized global incidence of overall OA as 492 cases per 100,000 [2]. Incidence of knee OA was greatest (350 per 100,000), followed by hand (80 per 100,000), other sites (43 per 100,000), and hip (19 per 100,000). Incidence rates vary according to regions, with estimates of 635 new cases per 100,000 in high sociodemographic index areas, compared with 423 per 100,000 in low sociodemographic index areas. The measurement of incidence in OA is difficult, as data from low- and middle-income countries is sparse [3], and estimates are predominantly modeled data.
The incidence of OA increases with age and is higher in females compared with males. Incidence rates of symptomatic hand, hip, and knee OA increase rapidly around 50 years of age and then level off after age 70 [4]. GBD2019 modeling estimated an incidence rate of OA of 1600 per 100,000 for those in the 50 to 69 year age group and 1100 per 100,000 for those 70 years and over [2].
The length of time over which incidence is calculated is not clearly specified or known in many reports, making comparisons difficult [4]. As examples, population-based data from the Johnston County OA Project reported age- and sex-standardized incidence rates for radiographic hip OA as 24 per 1000 person-years, symptomatic hip OA as 17 per 1000 person-years, and severe radiographic hip OA as 3.2 per 1000 person-years [5]. Another population-based study from the Fallon Community Health Plan in Massachusetts (United States) found that the age- and sex-standardized incidence rate was highest for knee OA (240 per 100,000 person-years), with intermediate rates for hand OA (100 per 100,000 person-years), and lowest rates for hip OA (88 per 100,000 person-years) [6]. Data from the Framingham OA study have estimated the crude incidence of radiographic hand OA to be approximately 35 percent in males and females, and the symptomatic hand OA incidence 4 percent for males and 9.7 percent for females over a nine-year period [7].
One estimate of the lifetime risk of developing symptomatic knee OA was approximately 40 percent in males and 47 percent in females [8], with similar estimates for hand OA (25 percent in males, 47 percent for females) [9]. The risk of symptomatic hand OA was greater in individuals with obesity (47 percent) than those without obesity (36 percent) [9]. Similarly, the lifetime risk of developing symptomatic hip OA has been estimated at approximately 25 percent [10].
●Prevalence – Globally, approximately 595 million people are affected by OA [2,11], with some studies reporting higher rates of 654 million in individuals 40 years of age and older. Prevalence in high-income North America has been estimated to be 51 million, which has increased by 89 percent from 1990 [2]. GBD2020 reports that an estimated 7.6 percent of the global population lived with OA [2,11], with 15 percent of those aged over 30 years living with some form of OA, taking into account differences between rural and urban areas and between high- and low- to moderate-income regions. This rate is destined to increase as the population ages and obesity rates increase [12].
The prevalence of OA increases with age: an estimated 2.9 percent in the 25- to 49-year age group, 23.2 percent in 50- to 69-year age group, and 38.4 percent in the 70+ age group. Prevalence in females (8.1 percent) is higher than in males (5.8 percent) [11]. Just under 10 percent of males and 14 percent of females aged 50 to 69 years report symptomatic hip and knee OA, increasing to 18 percent of males and 25 percent of females in the 70 years and older age group [2,13].
Knee — The knee is the most commonly reported site of OA, with an estimated 368 million prevalent cases in 2020 [2,11]. GBD2020 forecasting analyses estimate that by 2050 there will be 642 million individuals with knee OA. The age-standardized prevalence of radiographically confirmed symptomatic knee OA has been estimated at 4.3 percent globally, with rates being higher in females (4.8 percent) compared with males (2.8 percent) [14]. Knee OA accounts for approximately 80 percent of the disease's total burden [15]. The prevalence of knee OA in the United States is similar to that of Europe [16], with lower rates reported in southern Asia [14].
Fourteen million people in the United States have symptomatic knee OA, with more than half of those with knee OA being younger than 65 years of age [17]. United States data from surveys in 2007 to 2008 report 7 percent of adults over 25 as having symptomatic knee OA [18]. Black Americans have greater prevalence, severity progression, and worse pain and function compared with White Americans [19].
Regional differences in age-standardized prevalence of knee OA were seen in GBD2020: high-income Asia-Pacific 5.6 percent, east Asia 5 percent, eastern Europe 3.3 percent, central Asia 2.7 percent [11]. In the Framingham OA Study, the prevalence of symptomatic knee OA was found to be 7 percent [20]; in the Johnston County study, 17 percent reported symptomatic knee OA [21]. Similar rates have been reported in Sweden, with a prevalence of 15 percent for symptomatic knee OA.
Studies including Asian populations report a broad range of prevalence estimates for knee OA. Korean data report a prevalence of 4.5 percent for males and 19 percent for females [22]. A systematic review of Chinese data found the prevalence of symptomatic knee OA increased with age (3.1 percent in the 15- to 39-year age group, compared with 26.3 percent in those over 70 years of age) [23]. The prevalence of symptomatic knee OA was also found to be higher in females (19.1 percent) compared with males (10.9 percent). Using a methodology similar to that of the Framingham OA Study, studies from Chinese populations found that the prevalence of bilateral knee OA and lateral compartment disease was two to three times higher in the Chinese cohorts compared with the Framingham estimates [16]. Also in China, rates of symptomatic knee OA of 8 percent have been reported [18], and the age-standardized prevalence in GBD2020 for China was 5 percent [11]. Studies of prevalence in rural and urban areas report higher prevalence in rural areas, with prevalence of knee pain in India of 13 and 8 percent in rural and urban areas, respectively. Males older than 60 years in rural areas of China reported twice the rate of symptomatic knee pain compared with those in urban areas [16]. Higher rates in rural areas may be a result of workload, nutritional status, longevity [24], less education, and poorer socioeconomic status [19]. In addition, higher incidence rates of knee OA have been found in areas deemed most deprived by measures of socioeconomic status, including among those with lower levels of education [19].
Hip — Hip OA is less common than knee or hand OA, with an estimated 35.3 million (26.5 million to 45.1 million) cases globally in 2020 [11]. GBD2021 forecasting analyses estimate that by 2050 there will be 62.6 million (49.7 to 75.5) people with hip OA. The age-standardized prevalence of hip OA is estimated to be 0.42 percent, while the age-standardized prevalence of radiographically confirmed symptomatic hip OA has been reported to be 0.85 percent in adults over 18 years of age globally [14]. No sex differences have been identified in the rates of hip OA, and rates increase with age. GBD2019 data estimate global prevalence of hip OA to be 0.4 percent (females 0.41, males 0.39). Rates of radiographic hip OA are higher than symptomatic OA, with a prevalence of 1 to 6 percent being reported, depending on the definition used [16].
Prevalence of hip OA is generally higher in Europe and North America compared with Asia and Africa. GBD2020 data reported radiographically confirmed symptomatic hip OA prevalence estimates of almost 1 percent in North America, while estimates of approximately 0.3 percent were reported in southeast and east Asia, Oceania, and north Africa/the Middle East [14]. United States data from the Framingham study report an age-standardized prevalence of 2.4 percent for symptomatic hip OA [25]. The prevalence of symptomatic hip OA in Korea has been reported as 0.1 percent in males and 0.2 percent in females [22].
Hand — GBD2021 estimates there to be 189 million cases of hand OA, and forecasting analyses estimate that by 2050 there will be 279 million people with hand OA [11]. The prevalence of radiographically confirmed symptomatic hand OA was estimated to be 6.3 percent, with the rate for females (7.3 percent) slightly higher than for males (5.3 percent) [2]. The prevalence of radiographic hand OA in the United States has been reported to range from 27 to over 80 percent [16], and in Japan, a prevalence of radiographic hand OA of over 90 percent has been reported [26]. However, the rate of symptomatic hand OA is much lower at around 8 percent [27].
The prevalence of symptomatic hand OA increases with age, with one United States study reporting that 13 percent of males and 26 percent of females over 70 years reported symptomatic hand OA [16]. Interestingly, data from China suggest that symptomatic hand OA is rarely reported, irrespective of age or sex [16]. Regional differences of hand OA in GBD2021 ranged from 4 percent in eastern Europe and central Asia to less than 2 percent in east Asia and Oceania [11].
Spine — There are limited data on the prevalence of spinal OA [28]. One community-based study in the United States found that cervical facet joint OA was present in 19 percent of adults aged 45 to 64 years, and in 57 percent of adults aged 65 years and older [29]. A similar increase in facet joint OA prevalence with age is seen in the lumbar facet joint, with a prevalence of 24 percent in those younger than 40 years increasing to 69 percent of those 70 years and older [30]. These estimates probably underestimate true prevalence because of the insensitivity of radiography. A community-based population study found no association between facet joint OA identified on computed tomography (CT) scanning at any level and low back pain [30]. Sex differences in spinal OA are unclear. Some studies have reported a higher rate of spinal OA in males. As an example, in one study, males younger than 55 years had a greater risk of prevalent cervical spine OA (risk ratio [RR] 1.29, 95% CI 1.18-1.41) [31]. By contrast, another population-based study from Korea reported a rate of spinal OA of 5.6 percent in males and 16 percent in females [22]. Other studies have found no sex differences in spinal OA [32].
Multiple joints — The presence of OA in multiple joints is an often-overlooked component of the burden of OA, with varying definitions within the literature making it difficult to enable a global estimate [33]. Population-based data from Korea suggest that almost 11 percent of males and 23 percent of females had more than two joints involved [22]. Among a Swedish population, almost 27 percent of prevalent OA cases report OA in multiple joints [34]. One study attempted to define multiple-joint OA and apply the definitions to the Johnston County dataset, finding that 52 percent reported symptomatic hip or knee OA with one other site (spine, ankle, or foot), while 2 percent reported bilateral knee and spine OA. An analysis of multiple-joint OA in the Osteoarthritis Initiative found a greater frequency and progression in females [35]. Selected biomarkers of glucose metabolism appeared to be associated with prevalent multiple-joint OA and progression. Regardless of exact definition of multiple-joint OA, there are significant detrimental effects on general health and physical function when several joints are involved [36].
Mortality — OA-associated mortality is largely driven by cardiovascular death. Possible explanations for the excess mortality of OA include reduced levels of physical activity among persons with OA due to involvement of lower limb joints and presence of comorbid conditions, as well as adverse effects of medications used to treat symptomatic OA, particularly nonsteroidal anti-inflammatory drugs (NSAIDs), in addition to mortality associated with complications of joint replacement surgery [37].
A Swedish study has found an increased risk for cardiovascular death from ischemic heart disease and heart failure among those with hip or knee OA compared with those without OA, and this gap increases over time (hazard ratio [HR] 1.13 and 1.16, respectively). However, no increased mortality was found with hand OA or OA of other peripheral joints, and no consistent difference was seen between males and females [38,39]. A Canadian population-based cohort study of individuals with hip and knee OA found a significant association between greater OA-related disability and both all-cause mortality and risk of serious cardiovascular events. This study suggested that OA functional limitations, in particular walking disability, may be potential modifiable risk factors for serious cardiovascular events and death.
RISK FACTORS — Osteoarthritis (OA) is a complex interplay between mechanical, cellular, and biomechanical factors leading to end-stage pathology [21]. Multiple risk factors have been linked to OA, several of which are discussed in detail below.
Demographics
Age — Age is one of the strongest predictors of OA, with incidence of hand, hip, and knee OA increasing with age, especially after the age of 50 years [21,25,40]. A leveling-off occurs at all joint sites around the age of 80 years. Approximately 14 percent of adults aged 25 and older have symptomatic OA of at least one joint, while 34 percent of adults aged 65 and older have OA [41]. Worldwide estimates are that 10 percent of males and 18 percent of females aged over 60 years have symptomatic OA [13].
The mechanisms behind this increased risk with increasing age are not well known. Possible causes are sarcopenia, loss of proprioception, and joint laxity that may affect joint function and predispose the joint to injury. Changes affecting joint tissues include loss of normal bone structure, increased stiffness of ligaments and tendons, and meniscal degeneration. The role of aging in the pathogenesis of OA is discussed in detail separately. (See "Pathogenesis of osteoarthritis", section on 'Aging'.)
Sex — Female sex is associated with a higher prevalence and severity of OA. A large meta-analysis demonstrated a greater risk in females for prevalent and incident knee and hand OA as well as incident hip OA compared with males [31], and a recent meta-analysis showed that being female was a risk factor for knee OA, which was 1.04 times higher than males; although, this was not seen in cohort or case-control studies [42]. GBD2021 estimates showed higher prevalence in knee OA and hand OA in females than males but similar prevalence in hip OA and OA of other sites [11]. In addition, females were found to have more severe knee OA than males, particularly after menopause [21,31].
The Framingham Knee OA Study reported a 1.7 times higher incidence of OA of the knee in females compared with males (95% CI 1.5-2.7) [43]. Females were estimated to have a lifetime increased risk of symptomatic knee OA of 47 percent.
The reason for the increased risk of OA in females remains unclear, but it may be related to hormones, genetics, or other undetermined factors. (See "Pathogenesis of osteoarthritis", section on 'Risk factors'.)
Genetics — Numerous publications have emerged reviewing genetics in OA. Despite large numbers of patients with OA being compared with participants without OA, the genetic predisposition to the development of OA remains incompletely understood. All studies report limitations, and all suggest more population-based studies of well-documented cohorts are needed. Heterogeneity across the studies including the participants' age, ethnicity, sex, and phenotype distribution and the range of different clinical, biologic, and radiographic parameters collected alongside the genetic material make it difficult to pool data and also limit the ability to fully adjust for other potential causative factors. At this stage, most studies do not have sufficient data of phenotypes, other potential predictors, clinical and radiologic parameters to fully elucidate the causal association of genomics.
At least 30 percent of the risk of OA is thought to be genetically related, with genetic factors being typically stronger for hand and hip OA compared with knee OA [44]. Twin and family studies suggest that the influence of genetic factors is approximately 40 percent for OA of the knee, 60 percent for OA of the hip, 65 percent for OA of the hand, and approximately 70 percent for OA of the spine, independent of known environmental or demographic confounding factors. These suggest a heritability of OA of 50 percent or more, indicating that half the variation in susceptibility to disease in the population is explained by genetic factors [45]. It is also likely that the genes associated with pain and OA are different from each other [46].
While genetics play a significant role, no single gene is involved in the development of OA. Many genes could play a role in the disease onset, which could provide targets for future pharmacologic treatments.
●Demographic associations – Genes may operate differently between the two sexes, at different body sites, on different disease features within body sites, and between different populations. Multifactorial inheritance is more likely exerting influence with genes for specific sites rather than a generalized OA phenotype.
The differences between groups are demonstrated by a genome-wide association study (GWAS) that found a significant association of a signal in a region on chromosome 7q22 with OA in European-descent samples but not in Asian populations [44,47]. Similarly, two single-nucleotide polymorphisms (SNPs) on a region of human leukocyte antigen (HLA) locus on chromosome 6p were associated with knee OA in a Japanese population, but not in Han Chinese or European populations [44].
●SMAD3 – A meta-analysis reviewed the role of SMAD3 and identified 10 studies from six published articles comparing 5093 OA patients with 5699 controls [48]. An association was found between the G variant of SMAD3 rs12901499 polymorphism and an increased risk of both hip and knee OA in White patients but not among Asian patients. SMAD3 is located on chromosome 15Q22.33 and plays a critical role in joint homeostasis as a downstream mediator of transforming growth factor beta (TGFB) that regulates chondrocyte growth.
●ADAMTS polymorphisms – A review of studies of ADAMTS5 on chromosome 21q21.3 and ADAMTS4 on chromosome 1q23 found seven studies and had sufficient data to meta-analyze two SNPs for ADAMTS5 (rs226794 and rs2830585) but not for the range of different SNPs for ADAMTS4 [49]. Analysis among approximately 4000 OA patients and 3000 controls showed no association overall; however, a subgroup analysis found a statistically significant association between rs2830585 ADAMTS5 and OA in Asian patients but not White patients [49]. A recent systematic review and meta-analysis identified four studies (707 cases of OA with 859 controls) in which genetic modeling found single nucleotide polymorphism of ADAMTS14 (rs4747096) to be significantly associated with OA [50]. The ADAMTS family is part of the matrix-degrading enzymes the aggrecanases that degrade articular cartilage proteoglycans.
●Microribonucleic acids – Another emerging area of interest in OA pathophysiology is the study of microribonucleic acids (miRNAs), which are small noncoding RNA that regulate gene expression in human cells and are expressed differently in OA cartilage compared with non-OA cartilage [51]. A systematic review identified 37 articles describing these epigenetic factors with at least 22 different miRNAs implicated in OA being both increased and decreased and shown to influence a range of transcription and growth factors that regulate chondrogenesis and chondrocyte function [51]. As an example, miRNA-140 is expressed exclusively in chondrocytes and appears to be one of the major miRNAs implicated in OA. While there is overexpression of miRNA in OA, this is not considered causal as not specifically associated with OA alone and, therefore, is not considered useful for diagnosing OA. A range of different treatments using miRNAs suspended in lipid nanoparticles are being studied; however, the complex interactions between miRNAs and multiple target genes are still being elucidated.
●Additional genetic associations – Abnormalities in joint shape have been implicated in OA development [52]. Through genome-wide linkage analysis (GWLA) and candidate gene analysis studies, key genes identified included COL1A1 and growth differentiation factor 5 (GDF5).
GWLA has identified a region on chromosome 17q21 that codes for a range of candidate genes including COL1A1, HOX B, and DLX3 and on 13q22 for GDF5, ASPN, and TBX [52]. Of all the possible associations, only GDF5 has been consistently replicated across the studies. GDF5 is a member of the TGFB signaling family and a key regulator of joint morphogenesis. Mutations are associated with limb abnormalities and developmental dysplasia of the hip (DDH), hip OA, and some knee OA.
DDH is a recognized risk factor for developing hip OA, with individuals with DDH being more than 10 times more likely to develop OA than those without DDH [52]. Between 50 to 85 percent of total DDH susceptibility is reportedly due to genetic factors. DDH is also associated with female sex, first-born, breech presentation at birth, and family history.
Other childhood conditions affecting hip shape such as Perthes, slipped femoral epiphysis, and femoral acetabular impingement (FAI) have not been as well studied as DDH, and no consistent genetic associations have been identified. A range of other joint geometry parameters such as femoral neck length and neck-shaft angle have also been shown to increase risk of OA through statistical models, providing similar predictive value to sex, age, and body mass index (BMI) adding 7 percent to the variance.
Genetic variants of several groups of genes (eg, cartilage extracellular matrix structural genes and the genes related to bone density) have been implicated in disease pathogenesis. Many genes have been identified in playing a role in OA pathophysiologic pathways and thus may contribute to OA risk [21]. A GWAS meta-analysis of dual-energy X-ray absorptiometry (DEXA) derived hip shape identified eight SNPs to be independently associated with hip shape, including SOX9, PTHrP, RUNX, FGFR4, and DICER1 [53]. It was hypothesized that these findings may lead to a greater understanding of the relationship between hip OA and hip fracture.
●Genetic variance in OA – The availability of large numbers of data from biobanks have allowed analysis of the genetic variance in OA. Two publications of note that have emerged from the United Kingdom (UK) Biobank:
•A meta-analysis including 30,727 cases of OA and 297,121 controls using data across 16.5 million variants from the UK Biobank resource identified 9 novel OA variants [54]. Three of these loci showed an association with biologically relevant radiographic endophenotypes; however, radiographs were not routinely available. Five of the loci found genes differentially expressed in degraded compared with intact articular cartilage from OA patients who had undergone joint replacement.
Together, all established OA loci accounted for 26.3 percent of trait variance. Large effects in OA susceptibility were not detected, confirming the highly polygenic model that underpins OA risk. Higher BMI and other obesity-related measures showed a causal relationship with OA, but no convincing effect of others, including triglyceride level or type 2 diabetes, was associated. The study also reported a high genetic correlation between self-reported and hospital-diagnosed OA, thus corroborating the validity of self-reported OA for genetics studies, as well as a strong genome-wide association between hip and knee OA not previously supported. They concluded that a gold standard for OA genetic studies remains elusive.
•A subsequent analysis compared 77,052 patients with OA and 378,169 controls and identified 65 genetic variants, with 52 being novel [55]. The total heritability explained by OA loci was estimated to be 14.7 percent for knee, 51.9 percent for hip, 24.3 percent for hip and/or knee, and 22.5 percent OA at any site. Linkage disequilibrium regression analyses revealed significant correlations between OA and obesity, cognition, smoking, bone mineral density (BMD), and reproductive traits with potential protective effects of low-density lipoprotein (LDL) cholesterol and higher levels of education. In a subgroup analysis of specimens derived from joint replacement surgery, some associations were found with COL11A1 and COL11A2, genes associated with bone structure and cartilage proteins as well as FGFR3 and GDF5, genes associated with bone and cartilage development in animal models.
Additional information regarding the role of genetics in the pathogenesis of OA can be found separately. (See "Pathogenesis of osteoarthritis", section on 'Risk factors'.)
Joint injury — Joint injury is associated with the development of OA and is commonly referred to as post-traumatic OA. The knee is one of the most frequently injured joints. Knee injury at a young age (mean age of 27 years) is an independent risk factor for the development of knee OA at middle age (mean age of 39 years) [56,57]. In particular, rupture of the anterior cruciate ligament (ACL) is associated with early-onset knee OA in 13 percent of cases after 10 to 15 years [58]. When such rupture is associated with damaged cartilage, subchondral bone, collateral ligaments, and/or menisci, the prevalence of knee OA is higher, with estimates ranging from 21 to 40 percent [58]. In a systematic review, one study reported a symptomatic knee OA prevalence of 35 percent for the tibiofemoral joint and 15 percent for the patellofemoral joint at 10 or more years after an ACL tear [59]. A meta-analysis has shown that approximately 5 percent of new knee OA cases are related to a previous knee injury; thus, avoiding knee injury where possible could prevent 5 percent of new cases [60]. While only approximately 1 percent of the global population are estimated to have OA of the ankle joint, among those with ankle OA, prior joint trauma is the most common cause, accounting for between 20 and 78 percent of all cases of ankle OA [61].
Rates of OA development have been reported to be higher in patients after ACL reconstruction (44 percent) than in those who remained ACL deficient (37 percent), with an odds ratio (OR) of 1.29 [61]. In addition, a meta-analysis of six studies found that after ACL injury, irrespective of whether the injury was treated operatively or nonoperatively, the relative risk of developing even minimal OA was 3.89 compared with the contralateral control [62]. The relative risk of developing moderate to severe OA (grades III and IV) was 3.84 after a mean follow-up of 10 years compared with controls.
Additional information regarding the role of joint injury on the pathogenesis of OA can be found elsewhere. (See "Pathogenesis of osteoarthritis", section on 'Risk factors'.)
Anatomic factors — Anatomic factors such as joint shape and alignment have been associated with the development of OA, particularly for the knee and hip joints. As an example, a meta-analysis of high-quality cohort studies of radiographic and magnetic resonance imaging (MRI)-diagnosed OA found that malalignment of the knee joint is an independent risk factor for progression of OA of the knee [63]. In a prospective observational study including 230 patients with knee OA, medial progression of knee OA was fourfold greater in patients with varus alignment and lateral progression of OA was fivefold greater in those with valgus alignment [64].
Hip joint shape alterations such as femoroacetabular impingement (FAI) and acetabular dysplasia have been associated with high risk of early-onset hip OA and progression of hip OA [65], with FAI being the more prevalent of these shape alterations [66]. A more detailed discussion of FAI and acetabular dysplasia in the setting of hip OA is presented separately. (See "Management of hip osteoarthritis", section on 'Assessment for alterations in hip joint morphology'.)
Lifestyle factors
Obesity — Obesity represents one of the most important risk factors for both the incidence and progression of OA at weightbearing joints, such as the knee and hip, as well as for the hand. A meta-analysis of the impact of overweight and obesity on the population prevalence of OA found that the population-attributable risk percentage (PAR%), ie, the proportion of knee OA that would have been avoided if there was no obesity in the population, varied from 50 percent for symptomatic knee OA in the United States, where the prevalence of obesity is high, to 8 percent in China, where the obesity rate is lower [67]. Around 25 percent of cases of knee OA could be prevented in Europe and the United States if being overweight (BMI >25 but <30) is avoided. Lower rates of around 10 percent are seen in Asian countries where the prevalence of overweight is lower.
Body weight also influences the severity of the disease, with obese individuals experiencing more severe joint degeneration in the knees and a greater proportion requiring hip or knee joint replacement [68]. Increased weight has been associated with early articular cartilage damage identified by MRI well before symptoms develop, suggesting a causation between obesity and OA, rather than obesity developing as a result of lifestyle changes associated with OA [69]. As a modifiable risk factor, studies have demonstrated the implications of weight loss as an important approach for the management and prevention of OA. (See "Overview of the management of osteoarthritis", section on 'Nonpharmacologic therapy' and "Management of knee osteoarthritis", section on 'Weight loss'.)
The effect of obesity on OA is likely to be multifactorial, possibly due to increased load on the joint, decreased muscle strength, and altered biomechanics. In addition to these physical factors, obesity is characterized by a low-grade inflammatory state, which exerts effects on the joint tissue including cartilage, synovium, and bone [70]. Additional information regarding the role of obesity and the pathogenesis of OA can be found elsewhere. (See "Pathogenesis of osteoarthritis", section on 'Risk factors'.)
●Knee osteoarthritis – Individuals with a BMI >30 kg/m2 were 6.8 times more likely to develop knee OA than normal-weight controls [71]. A meta-analysis of 21 studies of BMI and risk of OA showed that BMI was positively associated with the development of knee OA [72]. A five-unit increase in BMI was associated with a 35 percent increased risk of knee OA (risk ratio [RR] 1.35, 95% CI 1.21, 1.51). This was more evident in females compared with males (males, RR 1.22, 95% CI 1.19-1.25 and females, RR 1.38, 95% CI 1.23-1.54).
Weight loss in OA can impart clinically significant improvements in pain and delay progression of joint structural damage [70]. Pain and functional disability associated with knee OA have been shown to improve significantly in a dose-response relationship after weight loss, with a reduction in weight of 7.7 percent leading to a clinical improvement in function scores [73]. A multicenter observational study evaluated the impact of weight loss on OA seen on MRI of the knee in 640 adults (mean age, 62.9 years; 398 females) who were overweight or obese or with mild to moderate OA [74]. When compared with those without weight loss, patients with a decrease in BMI of at least 5 percent over a 48-month period showed a decrease in rate of progression of knee cartilage and meniscal degeneration; the magnitude of effect was larger for those with a ≥10 percent decrease in BMI. Additionally, an analysis of a prospective population-based cohort aged 45 years and over identified a 31 percent reduction in the incidence of total knee replacement among those who had lost more than 7.5 percent of their body weight compared with their weight-stable counterparts [75]. (See "Management of knee osteoarthritis", section on 'Weight loss'.)
●Hip osteoarthritis – The relationship between obesity and hip OA is weaker than that for knee OA. A large population-based cohort study found that, compared with normal-weight subjects, the population classified as overweight or grade II obese had a 46 and 93 percent excess risk of hip OA, respectively [76]. Another meta-analysis found that a five-unit increase in BMI was associated with an 11 percent increased risk of hip OA (RR 1.11, 95% CI 1.07-1.16) [77].
●Hand osteoarthritis – There is evidence that demonstrates a positive association between increased BMI and the risk of hand OA [78]. A dose-response relationship has been demonstrated between BMI and hand OA, with one meta-analysis showing that a five-unit increase in BMI is related to an increased risk of radiographic hand OA (RR 1.06, 95% CI 1.02-1.10) [78]. In another study, fat percentage was associated with hand OA in males (OR 1.34, 95% 1.11-1.61) and females (OR 1.26, 95% CI 1.05-1.51) [79]. As hands are a non-weightbearing joint, this association between obesity and hand OA suggests a role for metabolic or inflammatory processes. (See "Pathogenesis of osteoarthritis", section on 'Risk factors'.)
●Shoulder osteoarthritis – As with hand OA, despite the conventional belief that the association between OA and obesity is limited to weightbearing joints, an increase in OA of the shoulder has been reported in people with a BMI greater than 25 [80]. In addition, the rate of shoulder arthroplasty was found to be greater among those with BMI greater than 30. A possible explanation could be as adipokines increase with the accumulation of additional adipose tissue, leptin in the synovial fluid may mediate increased cartilage destruction.
Occupation — Most studies of occupation and OA are limited to knee OA, although more generally, heavy physical workload was the most common occupational risk factor for OA in several anatomic locations. Certain occupational activities, especially those related to repetitive knee bending, appear to be risk factors for the development of knee OA. Other risk factors include kneeling; lifting; regular stair climbing, crawling, bending, and whole-body vibration; and repetitive movements. The mechanism for the association remains unclear, but joint loading and repetitive damage over time are implicated.
●Knee – Workers in several occupations appear to be at increased risk for knee OA, such as workers in construction, firefighting, agriculture, fisheries, forestry, and mining [81]. A meta-analysis of observational studies found that occupations with a heavy physical workload were associated with a higher risk of knee OA (odd ratio [OR] 1.65, 95% CI 1.43-1.91) [82]. Specifically, these occupations included agricultural workers, construction workers, miners, service workers, houseworkers, and cleaners, as they involved performing high levels of lifting, kneeling, climbing, squatting, and standing. Results from a British study revealed a more than a fivefold greater risk of knee OA among workers 55 years or older who were exposed to a combination of heavy lifting (>25 kg) and kneeling/squatting or climbing stairs [83]. A national Korean study found an increased risk of OA among machine operators and agricultural workers, with a higher rate of severe OA among the agricultural workers [84].
Other data have shown that the duration in occupations influences risk of OA. Males who worked for 11 to 30 years in building and construction work had a 3.7-fold greater risk of developing knee OA, after adjustment for confounding factors [85].
In occupations that involve prolonged kneeling or squatting, obesity leads to a significantly increased risk of knee OA (OR 14.7, 95% CI 7.2-30.2), compared with subjects with a BMI <25 kg/m2 who were not involved in occupations involving kneeling or squatting [86].
●Hip – Heavy physical workload or manual labor involving heavy loads has been shown to contribute to hip OA, with participation in such occupations doubling the risk of OA in males (RR 2.09, 95% CI 1.4-3.1) and increasing by approximately 40 percent in females (RR 1.41, 95% CI 1.0-1.9) [87]. Occupations involving bending, twisting, and reaching have been associated with increased symptomatic hip OA [88].
As with knee OA, the number of years in the occupation increased the risk, with farming for a period of one to nine years increasing the risk of hip OA by 4.5 times; farming 10 or more years increased the risk 9.3 times [89].
●Spine – Heavy physical workload is also an occupational risk factor that was found to be associated with spine and neck OA [81]. Compared with office workers, significantly greater risk of cervical spondylosis was found among meat carriers, dentists, and miners, possibly related to frequent extreme positions or high load being placed on the cervical spine [90].
Other possible associations
Nutrition — A review of evidence for the effect of dietary factors in OA has identified recommended intakes for dietary micronutrients.
●Polyunsaturated fatty acids – OA joints accumulate high levels of omega-6 fatty acids, which are the precursors of proinflammatory eicosanoids, hormone-like agents that mediate and regulate inflammation. Indirectly, these are reduced by omega-3 fatty acids, which also generate anti-inflammatory mediators [91]. The use of fish oil supplementation appears to be effective in pain reduction in rheumatoid arthritis; however, trials in OA are limited, and one randomized trial has failed to show benefit from high- compared with low-dose fish oil [92].
●Cholesterol – Studies have suggested that serum cholesterol may be implicated as a systemic risk factor for OA, with knee and hand OA associated with raised serum cholesterol. Reducing cholesterol accumulation with statins appears to have favorable effects in OA, and dietary strategies to reduce cholesterol and serum lipids are recommended [91].
●Vitamin D – The primary role of vitamin D is thought to be the regulation of bone metabolism and calcium homeostasis. Results from studies in knee OA suggest a positive association between vitamin D deficiency, cartilage loss, and OA progression, although trials have mixed findings, with a systematic review suggesting that vitamin D supplementation does not appear to provide clinically important benefits in either symptoms or structural progression of knee OA [93]. While vitamin D deficiency is unlikely to be a causal factor in OA, it may have positive effects on muscle strength, contributing to the stability of the joints [91].
●Vitamin K – With its role in bone and cartilage mineralization, a deficiency in vitamin K, as found in green leafy vegetables, is associated with increased incidence and progression of OA [94,95]. Vitamin K-dependent proteins found in bone and cartilage include the matrix Gla proteins GAS6 and osteocalcin. Vitamin K is an essential cofactor in the post-translational carboxylation of the Gla proteins, which is required for them to become functional. While data from clinical trials are limited, observational studies suggest that vitamin K could prevent OA, and maintaining an adequate intake is recommended.
However, warfarin is a commonly prescribed anticoagulant, with its effects occurring through the inhibition of the functioning of vitamin K. A nested case-control study comparing warfarin with direct oral anticoagulants, which do not antagonize vitamin K, identified patients undergoing hip or knee replacement, a measure of end-stage OA [96]. Warfarin was associated with a higher risk of hip and knee replacement, and longer duration of warfarin use was associated with greater risk of joint replacement compared with shorter duration of use. It is likely that warfarin leads to under-carboxylation of the vitamin K-dependent proteins, limiting their functionality.
Muscle weakness/strength — The quadriceps muscle contributes to stability of the knee joint, and clinically, weakness of this muscle is found in people with knee OA. There is evidence that exercise to strengthen the lower limb may improve symptoms [97], although conflicting results have been reported on the relationship between muscle strength or weakness and structural (radiographic) OA. As an example, a large cohort study including 3856 knees with OA found that quadriceps weakness in females, but not in males, was associated with increased risk for tibiofemoral and whole knee joint space narrowing over 30 months' follow-up [98]. Data from the Osteoarthritis Initiative found that isometric strength was significantly lower in symptomatic versus asymptomatic knees, whereas no association between strength and radiographic disease severity was observed [99]. In a systematic review and meta-analysis of 15 studies evaluating over 8000 patients, lower knee extensor strength was associated with an increased risk of pain (OR 1.35, 95% CI 1.10-1.67) and functional decline (OR 1.38, 95% CI 1.00-1.89); however, an increased risk of radiographic tibiofemoral joint space narrowing was not observed [100]. When stratified by sex, there was an increased risk of symptomatic decline in pain and progression to total knee replacement that was observed in females but not in males.
Conflicting results were found in another meta-analysis where knee extensor muscle weakness at baseline was a risk factor for later knee OA (OR 1.65, 95% CI 1.23-2.21), assessed as a combination of symptomatic, radiographic, and self-reported OA for both males and females [101].
In predicting deterioration of knee structure, muscle strength may not serve as the best surrogate measure of muscle function. Other measures, such as endurance or activation, may be independently related to disease [100].
Smoking — The association between smoking and OA is somewhat unclear and needs further examination. While a cohort of relatively young patients undergoing surgery for meniscal tears has found no significant association between current smoking and OA [102], a protective effect of smoking in OA has been observed in some epidemiologic studies. However, this is likely related to selection bias in which study participants are from hospital settings, where the rate of smoking may be higher than in the general population. As an example, a meta-analysis of 48 observational studies found a negative association between smoking and OA; however, when subgroup analyses were performed according to study type, no association was observed among the cohort or cross-sectional studies [103]. Further subgroup analyses revealed the protective association for smoking and OA was only observed for hospital-based case-control studies and not for community-based studies.
Another large meta-analysis of 38 independent observational studies including 481,744 participants found an inverse association between smoking and the risk for knee OA [104]. Those who had ever smoked had a significantly decreased risk of developing knee OA relative to those who had never smoked (RR 0.80, 95% CI 0.73-0.88). The lower risk for developing knee OA was more apparent in male smokers (RR 0.69, 95% CI 0.58-0.80) compared with female smokers (RR 0.89, 95% CI 0.77-1.02), and a dose-response analysis showed a linear decrease in knee OA with increased cigarette consumption.
Bone density — The relationship between bone mass and OA has been a subject of debate, with higher bone mass often being associated with increased risk of radiographic OA [105]. Most early studies reporting on the association between higher BMD and OA were cross-sectional designs that defined OA radiographically. Subsequent longitudinal studies have shown an association in most, but not all studies, between high BMD and the development of radiographic knee OA. Studies of people with high bone mass and OA in non-weightbearing joints such as the hands have shown an association with OA related to osteophyte growth rather than joint space narrowing reflecting cartilage loss [106]. While most studies of the association between high bone mass and OA assess radiographic OA, individuals with high bone mass were found to have a higher prevalence of self-reported joint replacement and use of nonsteroidal anti-inflammatory drugs (NSAIDs) compared with unaffected family controls, suggesting an increased risk of OA as ascertained by clinical, as opposed to radiographic, outcomes [105].
Higher spine and total hip BMD have been linked to progression of tibiofemoral cartilage defects in asymptomatic adults [107]. It is speculated that high-bone-mass individuals manifest a "bone-forming" tendency, which contributes to their risk of OA.
Physical activity — The relationship between sports participation and OA remains complicated and controversial and, in many cases, is based on low-quality evidence [108]. It is unclear if reported positive associations are a result of the participation itself or consequences of injury associated with the participation [17,109]. On the other hand, preserved cartilage thickness or prevention of weight gain through regular exercise may contribute to a protective effect from exercise [109]. Moderate daily recreational or sport activities, whatever the type of sport, are not a consistent risk factor for clinical or radiographic knee or hip OA [110]; thus, participation is encouraged to reduce the risk of other various age-related chronic conditions.
●Recreational running and walking – A review of studies of recreational running and walking does not suggest an association with OA. In prospective studies of middle-aged runners, long-distance running was not associated with accelerated clinical or radiographic OA of the knees compared with non-runner controls, when followed for 5 to 18 years [111,112]. In a study of 89,377 runners and walkers, neither running (including marathon running) nor walking were associated with increased risk of OA or hip replacement [109]. Similarly, observational data from the Framingham cohort did not show any additional risk or benefit from recreational walking or jogging on incident radiographic knee, even in subjects with a BMI >30 kg/m2 [111].
●Other (non-running) exercise – Limited evidence suggests that recreational non-running exercise may increase the risk for OA. In a retrospective study of 117 athletes, OA on radiographic examination was twice as prevalent among soccer players and weight lifters when compared with runners (29 percent of soccer players, 31 percent of weight lifters, and 14 percent of runners) [113]. In the largest cohort study of runners and walkers to date, in contrast to running alone, other non-running exercise, such as exercise performed in gyms or circuit training, was associated with increased risk for both OA and hip replacement [109].
●Vigorous exercise and elite athletes – There is some evidence to suggest that higher-paced activities and higher exposure to vigorous exercise may increase the risk of OA [108]. A meta-analysis found an increased risk of OA with both elite and non-elite sport participation (RR 1.37, 95% CI 1.14-1.64), although the quality of evidence was rated as very low. When limiting the data to elite athletes, there was also an increased risk of OA after sports exposure in elite sports people compared with a control group (RR 1.31, 95% CI 1.09-1.57). However, when sensitivity analysis was performed in which only cohort studies were included, this association was no longer significant. The risk was higher in soccer (RR 1.42, 95% CI 1.14-1.77) and greater among elite soccer players, but lower and not significant in runners (RR 0.86, 95% CI 0.53-1.41).
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Osteoarthritis".)
INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)
●Beyond the Basics topics (see "Patient education: Osteoarthritis symptoms and diagnosis (Beyond the Basics)" and "Patient education: Osteoarthritis treatment (Beyond the Basics)")
SUMMARY AND RECOMMENDATIONS
●Epidemiology – Osteoarthritis (OA) is a common condition, with prevalence increasing with age. The prevalence is higher in high-income countries compared with low-income countries. (See 'Epidemiology' above.)
The involvement of multiple joints occurs in approximately 25 percent of those with OA, contributing to the increased burden of disease. (See 'Multiple joints' above.)
●Risk factors – OA is a complex interplay between mechanical, cellular, and biomechanical factors leading to end-stage pathology. Multiple risk factors have been linked to OA, including:
•Age – Age is one of the strongest predictors of OA, with incidence of hand, hip, and knee OA increasing with age, especially after the age of 50 years. (See 'Age' above.)
•Sex – Female sex is associated with a higher prevalence and severity of OA. (See 'Sex' above.)
•Genetics – Genetics play a role in the development of OA, but no single gene is responsible. (See 'Genetics' above.)
•Joint injury – Joint injury is associated with the development of OA and is commonly referred to as post-traumatic OA. (See 'Joint injury' above.)
•Anatomic factors – Anatomic factors such as joint shape and alignment have been associated with the development of OA. (See 'Anatomic factors' above.)
•Obesity – Obesity represents one of the most important risk factors for both the incidence and progression of OA at weightbearing joints, such as the knee and hip, as well as for the hand. Body weight also influences the severity of the disease. (See 'Obesity' above.)
•Lifestyle factors – In addition to obesity, certain occupational activities and higher-paced physical activity are modifiable risk factors that contribute to the development of OA. (See 'Obesity' above and 'Occupation' above and 'Physical activity' above.)
●Other possible associations – While some types of physical activity may increase the risk of OA, moderate recreational or sport activities are not a consistent risk factor. Other possible risk factors for OA that are less well defined include nutritional factors, muscle weakness, smoking, and bone density. (See 'Nutrition' above and 'Muscle weakness/strength' above and 'Smoking' above and 'Bone density' above and 'Physical activity' above.)
ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Kenneth C Kalunian, MD, who contributed to an earlier version of this topic review.
آیا می خواهید مدیلیب را به صفحه اصلی خود اضافه کنید؟
