Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Arthritis Rheum. Author manuscript; available in PMC Jan 7, 2010.
Published in final edited form as:
PMCID: PMC2802836

Is Obesity a Risk Factor for Progressive Radiographic Knee Osteoarthritis?




To examine whether obesity increases the risk of progression of knee osteoarthritis (OA).


We used data from the Multicenter Osteoarthritis Study, a longitudinal study of persons with or at high risk of knee OA. OA was characterized at baseline and 30 months using posteroanterior fixed-flexion radiographs and Kellgren/Lawrence (K/L) grading, with alignment assessed on full-extremity films. In knees with OA at baseline (K/L grade 2 or 3), progression was defined as tibiofemoral joint space narrowing on the 30-month radiograph. In knees without OA at baseline (K/L grade 0 or 1), incident OA was defined as the development of radiographic OA at 30 months. Body mass index (BMI) at baseline was classified as normal (<25 kg/m2), overweight (25–<30 kg/m2), obese (30–<35 kg/m2), and very obese (≥35 kg/m2). The risk of progression was tested in all knees and in subgroups categorized according to alignment. Analyses were adjusted for age, sex, knee injury, and bone density.


Among the 2,623 subjects (5,159 knees), 60% were women, and the mean ± SD age was 62.4 ± 8.0 years. More than 80% of subjects were overweight or obese. At baseline, 36.4% of knees had tibiofemoral OA, and of those, only one-third were neutrally aligned. Compared with subjects with a normal BMI, those who were obese or very obese were at an increased risk of incident OA (relative risk 2.4 and 3.2, respectively [P for trend < 0.001]); this risk extended to knees from all alignment groups. Among knees with OA at baseline, there was no overall association between a high BMI and the risk of OA progression; however, an increased risk of progression was observed among knees with neutral but not varus alignment. The effect of obesity was intermediate in those with valgus alignment.


Although obesity was a risk factor for incident knee OA, we observed no overall relationship between obesity and the progression of knee OA. Obesity was not associated with OA progression in knees with varus alignment; however, it did increase the risk of progression in knees with neutral or valgus alignment. Therefore, weight loss may not be effective in preventing progression of structural damage in OA knees with varus alignment.


Obesity is a strong risk factor for incident (new-onset) tibiofemoral knee osteoarthritis (OA). However, in knees that already show evidence of OA, findings on the relationship between body mass index (BMI) and OA progression are inconsistent. Some studies have showed that a high BMI is associated with a high risk of knee OA progression, but other studies have failed to confirm this association (14). In general, the magnitude of association between BMI and the risk of progressive OA is weaker than the magnitude of association between BMI and the risk of incident disease. Patients who experience pain from knee OA might be motivated to lose weight if it could be shown that such weight loss not only alleviated pain but also prevented the progression of structural damage. Therefore, clarification of the relationship between obesity and the risk of OA progression has more practical implications than does a study of the effect of obesity on incident disease (2).

A study that determines whether the relationship between BMI and incident OA is different from the relationship between BMI and progressive OA requires that a large number of participants have incident disease and a large number of participants have progressive disease. Few studies have had large numbers of both.

If obesity is not a potent risk factor for disease progression, there are several potential explanations. One explanation is that obesity might not affect disease in individuals with malalignment, which is a potent risk factor for disease progression (5), because focal load may be sufficient by itself to damage these knees, and the excess load conferred by obesity may not matter. Such a phenomenon has been described in a small cohort in which obesity had no effect on OA progression overall and in subsets of knees with neutral and severe varus alignment; an effect of obesity on progression was seen among knees with moderate varus malalignment (6). Because severe malalignment might cause inevitable disease progression, the lack of effect of obesity in such knees might be neither surprising nor relevant to most persons with knee OA. The absence of an effect of obesity in knees with neutral alignment is unexpected, however, and would have cautionary implications regarding the effect of weight loss in persons with knee OA. If the alignment environment is critical to the effect of obesity, this might also be relevant to the effect of obesity on incident (new-onset) disease, a matter that has not been studied. Finally, no study has had sufficient numbers of persons with valgus alignment to test whether obesity affects progression in this group.

If obesity does not cause progression in any at-risk subgroup, other explanations need to be explored, including failure to adjust adequately for other causes of progression, problems with the definition of progression, and even the possibility that obesity simply does not increase the risk of progression, even though it does cause new-onset disease.

Using data from the Multicenter Osteoarthritis Study (MOST), a large longitudinal cohort study of risk factors for knee OA, we examined the relationship between BMI and the risk of incident and progressive radiographic tibiofemoral knee OA. In addition, we examined the effects of obesity among knees in different alignment milieux.


Study design and subjects

MOST is a prospective epidemiologic study of individuals ages 50 –79 years. Its overall goals are to identify risk factors for incident symptomatic knee OA and progressive OA in a sample of individuals who either have OA or are at high risk of developing OA. Persons considered to be at high risk included those who were overweight or obese, had knee pain, aching, or stiffness on most of the last 30 days, had a history of knee injury that made walking difficult for at least 1 week, or had previous knee surgery. High risk from obesity was defined based on persons who weighed more than the Framingham Heart Study median weight for their age- and sex-specific group (7).

All subjects were recruited from 2 communities in the US, Birmingham, Alabama and Iowa City, Iowa, through mass mailing of letters and study brochures supplemented by media and community outreach campaigns. Each center also recruited ethnic minorities according to their representation in the recruitment population.

Subjects were excluded if they screened positive for rheumatoid arthritis, had ankylosing spondylitis, psoriatic arthritis, or chronic reactive arthritis, had renal problems resulting in a need for hemodialysis or peritoneal dialysis, had a history of cancer (except for nonmelanoma skin cancer), had bilateral knee replacement surgery, were unable to walk without the help of another person or walker, or were planning to move out of the area in the next 3 years.

The protocol for MOST was approved by the institutional review boards at the University of Iowa, the University of Alabama, Birmingham, the University of California, San Francisco, and Boston University Medical Center.

Evaluation of risk factors

Sociodemographic variables (i.e., age, sex, and race) and information regarding a history of knee injury were collected at the baseline examination. We assessed knee injury at the baseline examination by asking subjects whether they had ever had an injury to a knee severe enough to limit their ability to walk for at least 2 days. At the baseline clinic examination, participants were weighed (balance beam scale) without shoes or heavy clothes, and had their height measured (stadiometer) without shoes. Total hip bone mineral density (BMD) was assessed at the baseline clinic visit using QDR 4500A dual x-ray absorptiometry scanners (Hologic, Bedford, MA). We included this as a confounder, because bone density was higher in those with knee OA than in those without it in the MOST study (as in other studies), and because it could confound the relationship between weight and OA incidence or progression. Full limb radiographs of both legs were obtained at baseline, using the method described by Sharma et al (5). The hip– knee–ankle (HKA) mechanical axis was defined as the angle formed by the intersection of a line from the center of the femoral head to the center of the femoral notch in the knee, and a second line from the center of the talus to the center of the tibial spines. The interobserver intraclass correlation coefficient for HKA was 0.95 (P < 0.0001). The 2 observers measuring the HKA worked independently and were unaware of other risk factors. Neutral alignment was defined as an angle of 179 –181°, and malalignment was defined as an angle ≤178° (varus) or ≥182° (valgus).


At baseline and the 30-month followup visit, all subjects underwent weight-bearing posteroanterior (PA) fixed flexion knee radiography, using the protocol described by Peterfy et al (8) and a plexiglas positioning frame (SynaFlexer; Synarc, San Francisco, CA). Lateral weight-bearing films of each knee were also obtained at baseline and 30 months using the Framingham OA Study protocol (9).

A musculoskeletal radiologist and a rheumatologist experienced in reading study films, both of whom were blinded to clinical data, graded the PA films for Kellgren/Lawrence (K/L) grades and individual radiographic features including joint space narrowing (JSN) and osteophyte scores. JSN and osteophytes were scored on a scale of 0 –3, separately for each compartment, using the Osteoarthritis Research Society International atlas (10). Lateral films were read using the protocol described by LaValley et al (9), in which JSN is scored in the medial and lateral tibiofemoral compartments. In previous studies (3), and as confirmed by preliminary readings in this study, we observed that knees often showed JSN changes longitudinally but did not show enough narrowing to move from one full grade to the next (e.g., 1 to 2). When joint space narrowed but did not change a full grade on the 0 –3-point scale, readers were instructed to use one-half grades, a strategy we have validated (10). Radiographs were read paired with the time sequence known. Each subject’s films were read by both readers working independently. If readers disagreed on the presence of incidence (for knees without radiographic OA at baseline) or progression (for knees with OA at baseline) using the definitions below, the film readings were adjudicated by a panel of 3 experienced readers to decide whether incidence or progression had occurred.

In this article, we focused on radiographic OA of the tibiofemoral joint. For knees without tibiofemoral OA (K/L grade 0 or 1) at baseline, we defined incident OA as new-onset K/L grade ≥2 at the 30-month followup. For knees with OA (K/L grade 2 or 3), we defined progression of JSN as any increase in the JSN score ≥one-half the grade in any location, based on either the PA or lateral view in the tibiofemoral joint (6). Knees with a K/L grade of 4 or a JSN grade of 3 either medially or laterally were not eligible for progression of JSN and were excluded from analysis.

Statistical analyses

We divided BMI into 4 categories: <25 kg/m2 (normal [referent]), 25–29.9 kg/m2 (overweight), 30 –34.9 kg/m2 (obese), and ≥35 kg/m2 (very obese). We calculated the risk of incident OA over the 30-month period by dividing the number of knees in which incident OA developed the total number of knees with K/L grade <2 at the baseline examination, for each BMI category. In a multiple binomial regression model, we adjusted for age, sex, race, BMD (sex-specific tertiles), and knee injury. We used generalized estimating equations to account for correlations between 2 knees within a subject. We tested a dose-response relationship between BMI and risk of incident OA by adding BMI as a continuous variable in the regression model. We also assessed whether the relationship of BMI with the risk of incident knee OA varied across 3 strata of alignment categories: varus (≤178°), neutral (179 –181°), and valgus (≥182°).

We calculated the risk for progressive OA by dividing the number of knees that had radiographic worsening during the followup period by the total number of knees with OA (K/L grade 2 or 3). We used the same approach described above to evaluate whether such an association varied according to alignment status at baseline.

We examined whether the effect of obesity was modified by alignment. Specifically, we grouped the knees into 3 categories, neutral, varus, and valgus, and examined the association between BMI and OA progression in each stratum separately. Although incidence in other studies has usually been defined to include new-onset osteophytes on radiographs, our definition of incidence required a narrowing of joint space on paired films. To make sure that our different results for incidence and progression were not attributable to different outcome definitions, we examined progressive JSN starting at grade 0, to create a definition of radiographic worsening that exactly paralleled our definition of progression.


Of the 3,026 subjects who had baseline knee radiographs, 30-month radiographs were available for 2,660 (87.9%). Among the 366 subjects not having 30-month radiographs, 33 (9.0%) died during followup, 7 (1.9%) withdrew consent, 2 had inflammatory arthritis that developed after the baseline examination, 41 (11.2%) had bilateral total knee replacement by the 30-month followup, and 283 (77.3%) did not attend to have a knee radiograph obtained. We further excluded knees from persons with missing information on height (n = 2), alignment measures (n = 141), PA view knee radiograph (n = 2), and knees with osteonecrosis at either baseline or followup (n = 16). Of the remaining 2,623 subjects (5,159 knees), 60% were women, and 85% were white. The mean ± SD age at baseline was 62.4 ± 8.0 years. More than 80% of subjects were overweight or obese. Radiographic OA in the tibiofemoral compartment at baseline was common (36.4% of knees), and of knees with radiographic OA, only one-third were neutrally aligned (Table 1).

Table 1
Characteristics of the MOST subjects*

Although many knees without OA had varus alignment (Table 1), both malalignment in general and extreme malalignment were more common in knees with OA at baseline (Figure 1). Among knees without radiographic OA at baseline, 40.6% had varus alignment, whereas among those with OA at baseline, 60.8% had varus alignment (P < 0.001 by chi-square test).

Figure 1
Distribution of alignment (by degree) in A, knees without tibiofemoral osteoarthritis, and B, with tibiofemoral osteoarthritis at baseline.

The risk of incident tibiofemoral knee OA over the followup period was 6.2%. Compared with subjects with normal BMI, the risk ratios for incident radiograph OA were 1.8, 2.4, and 3.2 for overweight, obese, and very obese subjects, respectively (P for trend < 0.001). Such an effect was not modified by knee alignment status at baseline (Table 2). In knees without OA at baseline, obesity increased the risk for the development of both mild knee OA (K/L grade 2) and moderate to severe disease (K/L grades 3 and 4).

Table 2
BMI and the risk of incident tibiofemoral osteoarthritis*

More than half of the knees with baseline OA (54.1%) progressed over the 30-month period. As shown in Table 3, we observed no association between obesity and the risk of OA progression. However, the effect of obesity on OA progression was modified by knee alignment; obesity had no effect on OA progression among knees with varus malalignment, whereas, among the knees with neutral alignment, high BMI was significantly associated with an increased risk of OA progression. Although it was statistically significant, the effect of obesity in valgus-aligned knees was intermediate between the risk in knees with varus or neutral alignment. More than half the knees with OA came from extremities with varus alignment, accounting for the null overall effect of obesity on progression in knees with OA.

Table 3
BMI and the risk of progressive tibiofemoral osteoarthritis*

We performed several additional analyses. In one, we limited progression to a full grade of worsening of JSN, and this yielded similar results. Analyses restricted to medial joint space loss and varus and neutral malalignment yielded similar findings also. Our findings of valgus alignment and progression were limited by the number of knees exhibiting lateral joint space loss (only 13% of knees with progression showed progression in that compartment). We could not determine definitively, for example, whether obesity had no effect on lateral progression in those with valgus alignment (among persons who were very obese, 58 knees were in valgus malalignment, and the adjusted risk ratio for lateral progression was 1.4 [range 0.9 –2.3], which represents a modest but insignificant increase in risk). Our analysis of the effect of neutral versus valgus alignment in terms of obesity and lateral joint space loss was similarly limited.

We performed additional analyses in which we looked at knees with joint space narrowing scores of 0 at baseline and the effect of obesity on the risk of narrowing to a score of ≥1. This paralleled our definition of progression. Our results in this analysis were identical to those shown for incidence (Table 2), with obesity increasing the risk of joint space loss regardless of alignment status.


In our study, obesity was associated with an increased risk of incident radiographic OA of the tibiofemoral joint. However, we observed no overall effect of BMI on the risk of progressive knee OA, but we report that this effect was modified by knee alignment status. A high BMI was associated with an increased risk of progressive knee OA among knees with neutral alignment but not among knees with varus alignment. Obesity had a modest, nonsignificant effect on OA progression among valgus-malaligned knees.

A different effect of obesity on incident versus progressive knee OA has been reported by several investigators (1,2,4). In one study of risk factors for incident versus progressive knee OA, in which obesity conferred a much higher risk of incidence than progression, Cooper et al noted, “Most currently recognized risk factors for prevalent knee OA (obesity, knee injury, and physical activity) influence incidence more than radiographic progression” (2). In a systematic review, Belo and colleagues (12) concluded that except for generalized OA and the level of serum hyaluronic acid, to date no single risk factor has consistently been associated with an increased risk of OA progression. One possible exception was the Rotterdam Study (11), in which obesity was associated with an increased risk of OA incidence (odds ratio [OR] 3.3 for the overweight group), and the relationship of obesity to progression depended on how progression was defined (in the overweight group, the OR was 1.4 –3.2 for different radiograph definitions of progression).

Our findings that obesity had no effect on progression in OA knees with varus alignment is not in full agreement with the results of a previous smaller study (6) that did not include knees with valgus alignment. Our current findings from a much larger sample probably make better biologic sense than the prior study, in that they suggest that obesity has its greatest effect among neutrally aligned knees, in which the excess load from obesity would act without the influence of focal loading from malalignment, to increase the risk of progression. The absence of an effect of obesity in knees with varus alignment could be explained. As noted by Andriacchi, “Varus knees always show a preponderant and often large adduction moment, producing varus stress across the knee and accelerating medial progression” (13). Our findings suggest that this stress is sufficient by itself to produce progression, and that the excess load conferred by obesity may not be necessary as an additional factor. Knees with valgus alignment do not necessarily produce comparable valgus stress: some show excess varus moment, others show valgus moments, and yet others have neutral dynamic moments. That might explain why the effects of obesity on OA progression in extremities with valgus alignment are intermediate between extremities with varus or neutral alignment.

Why did we not find the same malalignment effect in knees without OA? Among non-OA knees, obesity increased the risk of incident OA regardless of the malalignment status of the extremity. There are several potential explanations, all of which have relevance to our understanding of the effects of malalignment. First, as shown in Figure 1, malalignment among non-OA knees was less severe than that in OA knees, especially varus malalignment, suggesting that malalignment occurs with the development of knee OA. Second, malalignment in knees with OA may not be the same as in knees without it. For instance, studies of malalignment and its role in incident disease have shown conflicting results (14,15), and a lifetime of compensatory changes in the extremity may counteract stresses induced by malalignment in a healthy extremity and knee. In contrast, malalignment in OA knees represents evidence of joint damage and in some schemes (16) is actually used to measure disease severity. In some ways, severe malalignment in more advanced stages of OA might be a proxy for severe disease, and the absence of an obesity effect in this group might speak to the irreversibility of OA once it has reached the stage of varus malalignment. If so, our findings have implications well beyond the effects of obesity.

We noted that even in neutrally aligned knees, the magnitude of the association between obesity and the risk of progressive OA is smaller than that of the association between obesity and the risk of incident OA (for very obese individuals, the OR for progression is 1.8, and the OR for incidence is 3.2). Although a lower risk of incident disease versus progressive disease among subjects with a normal BMI is one explanation, other potential explanations may also account for such a phenomenon. For example, knees with OA not associated with obesity must have been exposed to risk factors other than obesity. If these risk factors were not identified, we would not be able to control for these unknown confounders. Furthermore, in our study, OA progression was assessed at fixed time points; therefore, we could not accurately define when the disease progressed. Therefore, OA knees in obese persons may have progressed faster than OA knees in nonobese persons. However, at the end of a long followup period, OA knees in both obese and nonobese individuals reached end-stage disease.

The failure to demonstrate that obesity increases the overall risk of OA progression in our study and others does not eliminate opportunities for weight loss trials aimed at slowing disease progression, especially among knees in extremities with neutral or valgus alignment. When such trials are conducted, we suggest that persons with varus malalignment should be excluded, especially if the focus of the trial is prevention of the progression of structural disease.

Our findings have implications both for other observational studies of progressive knee OA and for trials testing weight loss in OA. All studies of risk factors for OA disease progression are likely to face difficulties detecting risk factors in the face of varus malalignment. In all cases, such factors will be tested among persons with knee OA, and many of these persons will have varus malalignment. Large ongoing observational studies of knee OA may be limited in their ability to identify important risk factors for disease progression, because these risk factors operate differently in an environment of malalignment, or because the effects of risk factors on progression are muted compared with the effects on incidence. This, in turn, will limit our ability to select persons at high risk of progression who might be ideal candidates for trials.

Among the important limitations of our study is that we studied only 2 time points, so that the evolution of knee OA and its relationship to obesity could not be thoroughly characterized. Also, our analysis focused on structural (radiographic) progression and not on symptoms. Obese persons with knee OA generally have more severe symptoms than nonobese persons with the disease (17). Furthermore, weight loss may alleviate knee pain in persons with knee OA. However, the effects of weight loss on symptoms in knee OA trials have been modest at best (18,19), and our findings may help explain that observation. Because knee OA symptoms are driven by activity-based loading of the knee, the effects of weight loss on symptoms should parallel those on knee structure. Weight loss has many positive health effects, and we do not mean to suggest that it is not worthwhile, only that it may not delay the progression of structural damage in some persons with knee OA. In an observational study such as MOST, substantial weight loss is extremely uncommon and was a major reason we chose to focus on the structural effects of obesity.

In summary, in our analyses, obesity did not increase the risk of progressive OA among knees with existing OA. However, our analyses were not uniformly negative but showed that obesity increases the risk of progression, especially in neutrally aligned extremities.


Supported by the NIH (grant AR-47785). Drs. Torner, Nevitt, Lewis, and Felson’s work was supported by the National Institute on Aging (grants 1-U01-AG18832, 1-U01-AG19069, 1-U01-AG18947, and 1-U01-AG18820, respectively). The Multicenter Osteoarthritis (MOST) Study is a cooperative study of knee osteoarthritis funded by the National Institute on Aging.



Dr. Felson had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Zhang, Torner, Nevitt, Lewis, Sharma, Felson.

Acquisition of data. Torner, Nevitt, Lewis, Aliabadi, Sack, Clancy, Sharma, Felson.

Analysis and interpretation of data. Niu, Zhang, Nevitt, Sharma, Felson.

Manuscript preparation. Niu, Zhang, Torner, Nevitt, Lewis, Sack, Sharma, Felson.

Statistical analysis. Niu, Zhang.


1. Dieppe PA, Cushnaghan J, Shepstone L. The Bristol ‘OA500’ study: progression of osteoarthritis (OA) over 3 years and the relationship between clinical and radiographic changes at the knee joint. Osteoarthritis Cartilage. 1997;5:87–97. [PubMed]
2. Cooper C, Dieppe PA, Snow S, Kellingray S, Coggon D, Stuart B. Determinants of incidence and progression of radiographic knee osteoarthritis [abstract] Arthritis Rheum. 1997;40:1797. [PubMed]
3. Felson DT, McLaughlin S, Goggins J, Lavalley MP, Gale ME, Totterman S, et al. Bone marrow edema and its relation to progression of knee osteoarthritis. Ann Intern Med. 2003;139 (5 Pt 1):330–6. [PubMed]
4. Dieppe P, Cushnaghan J, Tucker M, Browning S, Shepstone L. The Bristol ’OA500 study’: progression and impact of the disease after 8 years. Osteoarthritis Cartilage. 2000;8:63–8. [PubMed]
5. Sharma L, Song J, Felson DT, Cahue S, Shamiyeh E, Dunlop DD. The role of knee alignment in disease progression and functional decline in knee osteoarthritis. JAMA. 2001;286:188–95. [PubMed]
6. Felson DT, Goggins J, Niu J, Zhang Y, Hunter DJ. The effect of body weight on progression of knee osteoarthritis is dependent on alignment. Arthritis Rheum. 2004;50:3904–9. [PubMed]
7. Felson DT, Anderson JJ, Naimark A, Walker AM, Meenan RF. Obesity and knee osteoarthritis: the Framingham Study. Ann Intern Med. 1988;109:18–24. [PubMed]
8. Peterfy CG, Li J, Duryea J, Lynch JA, Miaux Y, Genant HK. Nonfluoroscopic method for flexed radiography of the knee that allows reproducible joint-space width measurement [abstract] Arthritis Rheum. 1998;41:S361.
9. LaValley MP, McLaughlin S, Goggins J, Gale D, Nevitt MC, Felson DT. The lateral view radiograph for assessment of the tibiofemoral joint space in knee osteoarthritis: its reliability, sensitivity to change, and longitudinal validity. Arthritis Rheum. 2005;52:3542–7. [PubMed]
10. Felson D, Nevitt MC, Yang M, Clancy M, Niu J, Torner JC, et al. A new approach yields high rates of x-ray progression in knee osteoarthritis. J Rheumatol. 2008;35:2047–54. [PMC free article] [PubMed]
11. Reijman M, Pols HA, Bergink AP, Hazes JM, Belo JN, Lievense AM, et al. Body mass index associated with onset and progression of osteoarthritis of the knee but not of the hip: the Rotterdam Study. Ann Rheum Dis. 2007;66:158–62. [PMC free article] [PubMed]
12. Belo JN, Berger MY, Reijman M, Koes BW, Bierma-Zeinstra SM. Prognostic factors of progression of osteoarthritis of the knee: a systematic review of observational studies. Arthritis Rheum. 2007;57:13–26. [PubMed]
13. Andriacchi TP. Dynamics of knee malalignment. Orthop Clin North Am. 1994;25:395–403. [PubMed]
14. Hunter DJ, Niu J, Felson DT, Harvey WF, Gross KD, McCree P, et al. Knee alignment does not predict incident osteoarthritis: the Framingham osteoarthritis study. Arthritis Rheum. 2007;56:1212–8. [PubMed]
15. Brouwer GM, van Tol AW, Bergink AP, Belo JN, Bernsen RM, Reijman M, et al. Association between valgus and varus alignment and the development and progression of radiographic osteoarthritis of the knee. Arthritis Rheum. 2007;56:1204–11. [PubMed]
16. Derek T, Cooke V, Kelly BP, Harrison L, Mohamed G, Khan B. Radiographic grading for knee osteoarthritis. A revised scheme that relates to alignment and deformity. J Rheumatol. 1999;26:641–4. [PubMed]
17. Paradowski PT, Englund M, Lohmander LS, Roos EM. The effect of patient characteristics on variability in pain and function over two years in early knee osteoarthritis. Health Qual Life Outcomes. 2005;3:59. [PMC free article] [PubMed]
18. Messier SP, Loeser RF, Miller GD, Morgan TM, Rejeski WJ, Sevick MA, et al. Exercise and dietary weight loss in overweight and obese older adults with knee osteoarthritis: the Arthritis, Diet, and Activity Promotion Trial. Arthritis Rheum. 2004;50:1501–10. [PubMed]
19. Fransen M. Dietary weight loss and exercise for obese adults with knee osteoarthritis: modest weight loss targets, mild exercise, modest effects. Arthritis Rheum. 2004;50:1366–69. [PubMed]
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • MedGen
    Related information in MedGen
  • PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...