Abstract
Background.
Knee osteoarthritis (OA) and frailty are two conditions that are associated with functional limitation and disability in elders, yet their relation to one another is not known.
Methods.
We included participants from two large, multicenter studies enriched with community dwelling older adults with knee OA (Multicenter Osteoarthritis Study and Osteoarthritis Initiative). Knee OA was defined radiographically (ROA) and symptomatically (SOA). Frailty was defined using the Study of Osteoporotic Fracture index as the presence of ≥2 of the following: (i) weight loss >5% between two consecutive visits; (ii) inability to arise from chair five times without support; (iii) poor energy. Cross-sectional and longitudinal associations of knee OA with prevalent and incident frailty, respectively, were examined using binomial regression with robust variance estimation, adjusting for potential confounders.
Results.
In the cross-sectional analyses, frailty was more prevalent among participants with ROA (4.39% vs 2.77%; PR 1.60 [1.07, 2.39]) and SOA (5.88% vs 2.79%; PR 1.92 [1. 35, 2.74]) compared with those without ROA or SOA, respectively. In the longitudinal analyses, risk of developing frailty was greater among those with ROA (4.73% vs 2.50%; RR 1.45 [0.91, 2.30]) and SOA (6.30% vs 2.83%; RR 1.66 [1.11, 2.48]) than those without ROA or SOA, respectively.
Conclusions.
Knee OA is associated with greater prevalence and risk of developing frailty. Understanding the mechanisms linking these two common conditions of older adults would aid in identifying novel targets for treatment or prevention of frailty.
Key Words: Osteoarthritis, Knee, Frailty.
Knee osteoarthritis (OA), the most common form of arthritis in older adults, is associated with significant lower extremity functional limitations and is one of the leading causes of disability in the United States and worldwide (1–3). Functional limitation and disability are also strongly associated with another geriatric condition, frailty, which affects 4%–17% of older adults in the community (4,5). Knee OA and frailty also share some common risk factors, such as female sex and obesity (6,7). Despite these commonalities, the association between knee OA and frailty has not been evaluated previously. Frailty is a state of compromised physiologic reserve and heightened susceptibility to adverse outcomes such as falls, fractures, institutionalization, and even mortality (4). Therefore, the presence of frailty in participants with knee OA would imply vulnerability to additional adverse consequences associated with frailty, beyond those that are associated with knee OA in isolation. As the prevalence of knee OA is rising, a parallel rise in these consequences would be expected for older adults, posing an additional public health burden. Further, identifying an association between knee OA and frailty would lead to mechanistic studies as part of a larger effort to understand factors related to development of frailty in chronic conditions. For example, knee OA is associated with decreased physical activity, which can lead to loss of muscle mass (due to disuse). This may contribute to the generalized loss of muscle mass in frailty.
In this study, we sought to evaluate the association between knee OA and frailty cross-sectionally and longitudinally in a large cohort of community-dwelling older adults.
Methods
Study Sample
Men and women ≥60 years were included from two large NIH-funded studies, designed to study risk factors for knee OA: Multicenter Osteoarthritis Study (MOST; N = 3,026) and Osteoarthritis Initiative (OAI; N = 4,796) (8). Both MOST and OAI are multicenter, longitudinal, and observational cohorts that have similar study populations, with similar age ranges at the baseline visits (age 50–79 years in MOST; age 45–79 years in OAI), inclusion criteria and factors used to identify those at high risk for knee OA, as described elsewhere (8). Both cohorts included comprehensive assessment of knee OA (using the same protocols with the same center reading the radiographs for both cohorts, and the same pain questionnaires), frailty measures and other pertinent potential confounders at each study visit, as explained in further detail below. The MOST study visits occurred at baseline, 30, 60, and 84 months; data up to the 60-month visit was used in these analyses to provide a comparable timeframe as that which was available in the OAI. The OAI had annual study visits; we used data up to the 48-month visit as those were publicly available at the time of these analyses.
Because weight-loss is a component of the frailty definition used (see as in what follows), we used the 30-month visit in MOST as the baseline study visit to enable determination of weight-loss between baseline and that first follow-up visit. Similarly, the baseline study visit for OAI was the 24-month visit to enable a similar timeframe over which weight-loss was determined in that cohort.
Exposure
Participants in both MOST and OAI have had bilateral weight-bearing, fixed-flexion posteroanterior knee x-rays at each study visit. We defined radiographic knee OA (ROA) as presence of Kellgren and Lawrence (KL) grade ≥2 in at least one knee, a standard, validated definition of knee OA (9). The 30-month study visit in MOST and the 24-month visit in OAI were used to define the ROA status as that was the baseline for these analyses, as described earlier. To better understand whether burden of OA as reflected by symptoms and to test whether a “dose–response” relation of OA severity may be more likely associated with frailty, we also examined the following definitions of knee OA: (i) symptomatic knee OA (SOA), defined as ROA in at least one knee plus pain in the same knee; (ii) severity of ROA, based upon the highest KL grade of either knee, with those knees that had undergone knee replacement (KR) surgery being considered KL grade 4. However, because participants who have undergone KR surgery may have a different profile for physical activity and muscle strength than those with grade 4 OA, we excluded participants with KR in a sensitivity analysis; and (iii) number of knees with ROA, categorized as bilateral ROA, unilateral ROA, or no ROA in either knee.
Outcome
Frailty was defined using Study of Osteoporotic Fracture (SOF) (Ensrud’s) index (5) as presence of ≥2 out of three of the following criteria: (i) weight loss >5% between baseline and the subsequent follow-up visit; (ii) inability to rise from chair five times without using support (hereafter referred to as difficulty with chair stands); and (iii) poor energy derived from the SF12 questionnaire response of “little at a time” or “none at a time” to the question “in the past 4 weeks, did you have a lot of energy?”. Weight loss was defined using weight change between the baseline to the 30-month visit in MOST and the baseline to 24-month visit in OAI for the cross-sectional analyses. For the longitudinal analyses, weight change was ascertained between the 30- to 60-month visits in MOST and the 24- to 48-month visits in OAI. Difficulty with chair stands and poor energy from the 30-month visit in MOST and the 24-month visit in OAI were used for cross-sectional analyses; the respective 60- and 48-month visits were used for the longitudinal analyses. Although the outcome is prevalent frailty in the cross-sectional analyses, in the longitudinal analyses the outcome is incident (new-onset) frailty at follow up among those who did not have frailty at the study baseline.
Confounders
Selected based on literature review and biologic plausibility, potential confounders included age, sex, body mass index (BMI), education, physical activity (Physical Activity Scale for Elderly), comorbidities (Modified Charlson Comorbidity index) and smoking status (current/former, yes vs no).
Statistical Analyses
The cross-sectional associations of knee OA (four definitions analyzed separately) with prevalent frailty as well as each of the individual frailty components were performed using binomial regression with robust variance estimation to calculate prevalence ratios (PRs) in both cohorts combined. Given that women and obese individuals are at greater risk for both knee OA and frailty, interaction by sex and BMI with knee OA on its relation to frailty was also tested. For sensitivity analyses, the above associations were tested after excluding participants with KR. Additionally, the longitudinal association of all four definitions of knee OA with incident frailty at the subsequent follow-up visit was examined by calculating risk ratios (RRs). All analyses were adjusted for potential confounders and indicator variable for type of study (MOST vs OAI). All analyses were performed using SAS 9.3 (SAS Inc., Cary, NC). A two-sided p value of <.05 was considered significant.
The Institutional Review Board at Boston University Medical Campus approved the study.
Results
In the combined MOST and OAI cohorts, we excluded 2,867 participants (1,032 MOST; 1,835 OAI) who were <60 years. There were 685 participants with missing frailty assessment at study baseline. Among 3,707 eligible participants for the cross-sectional analyses, in MOST and OAI combined (mean age 68 years, 61% women and mean BMI 29kg/m2), ROA and SOA was identified in 2,371 and 1,225 participants, respectively. Participants with ROA had higher mean BMI, were slightly less physically activity, and had more comorbidities compared with those without ROA. Participant characteristics by ROA status at study baseline are shown in Table 1. Prevalent and incident frailty was identified in 141 and 116 participants, respectively. The frequency of the individual frailty components at baseline for this study is shown in Table 1.
Table 1.
Participant Characteristics by Radiographic Knee Osteoarthritis (OA) Status at the Study Baseline for Combined MOST and OAI Cohorts
| Characteristics | All Participants (N = 3,707) | Radiographic Knee OA | |
|---|---|---|---|
| Yes (N = 2,371) | No (N = 1,336) | ||
| Mean age (years) (SD) | 68 (5.4) | 68 (5.4) | 68 (5.3) |
| % Women | 61 | 61 | 59 |
| Mean BMI (kg/m2) (SD) | 29 (5.1) | 30 (5.3) | 28 (4.6) |
| High school education or less (%) | 37 | 37 | 35 |
| Mean physical activity (PASE) score (SD) | 143 (73.2) | 141 (72.1) | 146 (74.3) |
| Never smokers (%) | 52 | 52 | 52 |
| Mean modified Charlson’s comorbidity index (SD) | 0.64 (1.1) | 0.66 (1.1) | 0.59 (1.1) |
| Presence of frequent knee pain (%) | 1,491 (40.2) | 1,147 (48.5) | 344 (25.7) |
| Radiographic severity of knee OA | |||
| KL = 0 | 848 | n/a | 848 |
| KL = 1 | 479 | n/a | 479 |
| KL = 2 | 838 | 838 | n/a |
| KL = 3 | 850 | 850 | n/a |
| KL = 4 | 692 | 692 | n/a |
| MOST study (%) | 42 | 42 | 47 |
| Weight loss (%) | 546 (14) | 362 (15) | 184 (14) |
| Inability to perform repeat chair stands | 180 (5) | 140 (6) | 40 (3) |
| Poor energy | 426 (11) | 299 (12) | 127 (9) |
| Frailty (%) | 141 (4) | 104 (4) | 37 (3) |
Notes: BMI = body mass index; KL = Kellgren and Lawrence; MOST = Multicenter Osteoarthritis Study; n/a = not available; OAI = Osteoarthritis Initiative; PASE = Physical Activity Scale for Elderly.
In the cross-sectional analyses, participants with ROA and SOA had 60% and two-fold higher prevalence of frailty, respectively, than their counterparts (Table 2). The prevalence of frailty was also related to disease burden, such that it increased with increasing (ie, more severe) KL grades (p = .002 for linear trend), and was higher among participants with ROA in both knees compared with those without ROA in either knee (PR 2.02; 95% CI 1.33–3.08) (Table 2). Participants with ROA in one knee (vs none) did not have increased prevalence of frailty (PR 0.95; 95% CI 0.55–1.63). With respect to the relation of ROA to individual frailty components, presence of ROA was associated with increased prevalence of difficulty with chair stands (PR 2.10 [1.48–3.02]) and poor energy (PR 1.36 [1.10–1.67]), but not weight loss (PR 1.10 [0.93–1.33]). No interaction by sex (p = .56) or BMI (p = .64) with ROA on the prevalence of frailty was found. Sensitivity analyses excluding participants with KR did not substantially change the results (data not shown).
Table 2.
Cross-sectional and Longitudinal Association of Knee Osteoarthritis (OA) With Frailty in Community Dwelling Older Adults
| Prevalence of Frailty (Cross-sectional Analyses) | |||
|---|---|---|---|
| Knee OA Status | N (%) | Crude PR | Adjusted* PR (95% CI) |
| Radiographic knee OA | |||
| No (ref) (N = 1,336) | 37 (2.77) | 1.0 | 1.0 |
| Yes (N = 2,371) | 104 (4.39) | 1.58 | 1.60 (1.07, 2.39) |
| Symptomatic knee OA | |||
| No (ref) (N = 2,475) | 69 (2.79) | 1.0 | 1.0 |
| Yes (N = 1,225) | 72 (5.88) | 2.11 | 1.92 (1. 35, 2.74) |
| Radiographic severity of knee OA | |||
| KL = 0 (ref) (N = 848) | 19 (2.23) | 1.0 | 1.0 |
| KL = 1 (N = 479) | 18 (3.72) | 1.67 | 1.68 (0.85, 3.30) |
| KL = 2 (N = 838) | 29 (3.50) | 1.57 | 1.66 (0.91, 3.02) |
| KL = 3 (N = 850) | 33 (3.88) | 1.74 | 1.89 (1.05, 3.40) |
| KL = 4 (N = 692) | 42 (6.07) | 2.72 | 2.52 (1.41, 4.49) |
| p for trend = .002 | |||
| Number of knees with OA | |||
| None (ref) (N = 1,341) | 37 (2.77) | 1.0 | 1.0 (Ref) |
| Unilateral (N = 931) | 22 (2.37) | 0.86 | 0.95 (0.55, 1.63) |
| Bilateral (N = 1,425) | 81 (5.65) | 2.04 | 2.02 (1.33, 3.08) |
| Risk of Frailty (Longitudinal Analyses) | |||
| N (%) | Crude RR | Adjusted* RR (95% CI) | |
| Radiographic knee OA | |||
| No (ref) (N = 1,121) | 28 (2.50) | 1.0 | 1.0 |
| Yes (N = 1,859) | 88 (4.73) | 1.94 | 1.45 (0.91, 2.30) |
| Symptomatic knee OA | |||
| No (ref) (N = 2,053) | 58 (2.83) | 1.0 | 1.0 |
| Yes (N = 921) | 58 (6.30) | 2.31 | 1.66 (1.11, 2.48) |
| Radiographic severity of knee OA | |||
| KL = 0 (ref) (N = 701) | 18 (2.57) | 1.0 | 1.0 |
| KL = 1(N = 406) | 10 (2.46) | 0.96 | 0.79 (0.35, 1.79) |
| KL = 2 (N = 705) | 26 (3.69) | 1.45 | 1.16 (0.62, 2.17) |
| KL = 3 (N = 677) | 31 (4.58) | 1.82 | 1.20 (0.65, 2.24) |
| KL = 4 (N = 491) | 31 (6.31) | 2.56 | 1.50 (0.82, 2.87) |
| Number of knees with OA | |||
| None (ref) (N = 1,121) | 28 (2.50) | 1.0 | 1.0 |
| Unilateral (N = 783) | 34 (4.34) | 1.78 | 1.51 (0.89, 2.56) |
| Bilateral (N = 1,069) | 54 (5.05) | 2.08 | 1.36 (0.82, 2.25) |
Notes: CI = confidence interval; KL = Kellgren and Lawrence; PR = prevalence ratio; RR = risk ratio.
*Adjusted for age, sex, body mass index, education, physical activity, comorbidities, smoking, and study indicator.
In the longitudinal analyses, components of the frailty definition were missing in 603 participants. Among the 2,980 participants eligible for the longitudinal analyses, incidence of frailty was 45% and 66% greater among those with ROA and SOA, respectively, compared with their counterparts (Table 2), although the result for ROA did not reach statistical significance. Increasing magnitudes of effect for risk of frailty related to worsening KL grade was also noted, however, the results did not reach statistical significance. Similarly, a non-significant increased risk of frailty was noted with unilateral and bilateral ROA when compared with those without OA.
Discussion
In this study comprising data from two large well-characterized cohorts, knee OA was associated with greater prevalence and incidence of frailty in community-dwelling older adults, consistent with our hypothesis. Also, the burden of knee OA, both related to symptoms as well as severity and extent of the radiographic disease, had a greater relation to frailty prevalence.
To our knowledge, this is the first study to examine the relation between knee OA and frailty. One prior cross-sectional study that included only those with knee or hip OA (ie, lacking participants without OA) found a high prevalence of frailty (22.4%) and prefrailty (48.3%) among community-dwelling older adults, which is in-line with our results (10). However, without a comparison group (no OA group), conclusions regarding the relation of frailty to OA would be difficult. Another recent study performed in a cohort of older men noted that hip OA was associated with frailty, complementing our study’s findings in knee OA (11).
Although an association between knee OA and frailty is perhaps not surprising, the underlying mechanisms are not understood. Identifying potential mechanisms through which knee OA and frailty are associated would provide novel targets for management and prevention of frailty in older adults. To that end, two potential mechanisms are of interest. First, because knee pain and quadriceps muscle weakness related to knee OA can lead to decreased physical activity (12), which may subsequently contribute to loss of muscle mass (as a result of disuse) and frailty, physical activity needs to be examined as a mediator between knee OA and frailty. Another potential mechanism is through synergism between knee OA and systemic inflammation. Systemic inflammation has consistently been associated with frailty (13,14), but not with knee OA (15,16). Interestingly, the two major risk factors for knee OA are considered to be pro-inflammatory states: aging, due to cell senescence and release of proinflammatory cytokines, and obesity, due to release of proinflamamtory cytokines from adipose tissue (17,18). It is conceivable that systemic inflammation is present in knee OA only intermittently and, when present, may act in conjunction with knee OA in the development of frailty.
The frailty definition used in this study also deserves discussion because several definitions exist in the literature (4,5,19). In epidemiologic studies, the frailty phenotype has been described using the Cardiovascular Health Study index, which defines frailty by the presence of three out of five of the following criteria (4): (i) unintentional weight loss ≥10% per year, (ii) general feeling of exhaustion, (iii) weak grip strength, (iv) slow gait speed, and (v) low physical activity. However, use of the Cardiovascular Health Study index has limited utility in clinical practice given the practical issues related to assessing these criteria in a clinic visit (eg, grip strength, timed walk, and physical activity). The definition of frailty from the SOF was developed based on a similar construct (5). The SOF index is considered to be more practical for use clinically and in large cohort studies. It predicts adverse outcomes such as falls, hospitalization, disability, and mortality, and has equivalent predictive ability to the Cardiovascular Health Study index (20). We therefore chose to use the SOF index given its demonstrated validity and feasibility of data acquisition from these two large cohorts. The inclusion of repeat chair stands in the SOF frailty definition as a measure of muscle weakness in knee OA also merits comment. It could be argued that local symptoms related to knee OA (eg, pain or instability) may affect ability to perform chair stands. The best approach to addressing this concern is by using an alternate definition of frailty which does not include repeat chair stands as one of the functional measures to validate our results, which is beyond the scope of this article. However, although pain and instability can be associated with functional limitation in knee OA, pain is not necessarily well-correlated with function in knee OA (21,22), just as pain and radiographic severity are not well-correlated in knee OA. Further, only half of our ROA participants had SOA, and only a minority reported instability. In addition, because muscle weakness, particularly quadriceps weakness, is common in knee OA, it is likely that performance in repeat chair stands also reflects lower extremity weakness, the intended domain being measured in the SOF index. Pain and instability themselves can also lead to decreased physical activity and therefore generalized loss of muscle mass and thus contribute to development of frailty. Of note, in an article by Wise and colleagues (11), an increased prevalence and risk of frailty was noted among participants with hip OA using the Cardiovascular Health Study definition of frailty. Because hip OA and knee OA share many of the same symptoms, the results from Wise and colleagues provide complementary support for frailty being present in lower extremity OA using two different frailty definitions. Thus, we believe that, as intended in the original SOF article, performance on the repeat chair stands can reasonably represent muscle weakness in this clinical population as well. Because not all individuals with knee OA have functional limitations, such difficulties may be an indication of increased propensity to adverse consequences such as frailty.
We acknowledge that our study has limitations. Despite studying two large cohorts of older adults with/at risk for knee OA, we were still limited by power, in particular for the longitudinal evaluation due to the low incidence of frailty. Although baseline samples of the combined cohorts of MOST (n = 3,026) and OAI (n = 4,796) was 7,822, the sample analyzed for this study was reduced markedly by limiting the study to participants ≥60 years, the age group at risk for frailty. The timeframe over which weight loss was assessed was not identical in the two cohorts, but we chose study visits to enable relatively comparable timeframes. Given the observational nature of the study, despite adjusting for plausible potential confounders, residual confounding is still possible.
Our study also has several strengths. First, the assessment of knee OA was comprehensive. It is reassuring that the findings were similar across the different characterizations of knee OA, enabling insights regarding both structural and symptomatic disease, as well as in the cross-sectional and longitudinal analyses. Second, we used a validated assessment of frailty that can be feasibly applied in clinical practice to identify frailty in patients with knee OA, which would indicate that they may be at increased risk for adverse outcomes. Third, the use of data from two cohorts that had acquired radiographs, symptoms, and frailty components in a comparable and standardized manner enabled inclusion of a large number of study participants to comprehensively examine these associations that would not otherwise have been possible. The study participants included both older men and older women from different geographic regions of the United States, aiding this study’s generalizability.
In summary, our study findings demonstrate an association between two common geriatric conditions that carry substantial risk for adverse outcomes and are costly. Elucidation of mechanisms underlying this association may enable the development and testing of novel prevention and treatment strategies.
Funding
Dr. Misra was supported by Arthritis Foundation Postdoctoral Fellowship Award and Rheumatology Research Foundation Investigator Award. Support for the Multicenter Osteoarthritis (MOST) Study is by NIA U01-AG18820 (Felson, PI), U01-AG19069 (Nevitt, PI), U01-AG18947 (Lewis, PI), U01-AG18832 (Torner, PI). This study is also supported by AR47785 (MRCR P60). Dr. Neogi is supported by AF Innovative Research grant, K23 AR055127 and R01AR062506-01A1.
Acknowledgment
Funding for this study was received through Arthritis Foundation Clinical to Research Transition Award.
References
- 1. Dillon CF, Rasch EK, Gu Q, Hirsch R. Prevalence of knee osteoarthritis in the United States: arthritis data from the Third National Health and Nutrition Examination Survey 1991-94. J Rheumatol. 2006;33:2271–2279. [PubMed] [Google Scholar]
- 2. Guccione AA, Felson DT, Anderson JJ, et al. The effects of specific medical conditions on the functional limitations of elders in the Framingham Study. Am J Public Health. 1994;84:351–358. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Murray CJ, Vos T, Lozano R, et al. Disability-adjusted life years (DALYs) for 291 diseases and injuries in 21 regions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380:2197–2223. :10.1016/S0140-6736(12)61689-4 [DOI] [PubMed] [Google Scholar]
- 4. Fried LP, Tangen CM, Walston J, et al. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001;56:M146–M156. [DOI] [PubMed] [Google Scholar]
- 5. Ensrud KE, Ewing SK, Cawthon PM, et al. A comparison of frailty indexes for the prediction of falls, disability, fractures, and mortality in older men. J Am Geriatr Soc. 2009;57:492–498. :10.1111/j.1532-5415.2009.02137.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Felson DT, Anderson JJ, Naimark A, Walker AM, Meenan RF. Obesity and knee osteoarthritis. The Framingham Study. Ann Intern Med. 1988;109:18–24. [DOI] [PubMed] [Google Scholar]
- 7. Blaum CS, Xue QL, Michelon E, Semba RD, Fried LP. The association between obesity and the frailty syndrome in older women: the Women’s Health and Aging Studies. J Am Geriatr Soc. 2005;53:927–934. :10.1111/j.1532-5415.2005.53300.x [DOI] [PubMed] [Google Scholar]
- 8. Felson DT, Niu J, Yang T, et al. Physical activity, alignment and knee osteoarthritis: data from MOST and the OAI. Osteoarthritis Cartilage. 21:789–795. :10.1016/j.joca.2013.03.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Kellgren JH, Lawrence JS. Radiological assessment of osteo-arthrosis. Ann Rheum Dis. 1957;16:494–502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Miguel Rde C, Dias RC, Dias JM, et al. Frailty syndrome in the community-dwelling elderly with osteoarthritis. Rev Bras Reumatol. 2012;52:331–347. :S0482-50042012000300004 [pii] [PubMed] [Google Scholar]
- 11. Wise BL, Parimi N, Zhang Y, et al. Frailty and hip osteoarthritis in men in the MrOS cohort. J Gerontol A Biol Sci Med Sci. 2014;69:602–608. :10.1093/gerona/glt126 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Ling SM, Fried LP, Garrett ES, Fan MY, Rantanen T, Bathon JM. Knee osteoarthritis compromises early mobility function: the Women’s Health and Aging Study II. J Rheumatol. 2003;30:114–120. [PubMed] [Google Scholar]
- 13. Leng SX, Xue QL, Tian J, Walston JD, Fried LP. Inflammation and frailty in older women. J Am Geriatr Soc. 2007;55:864–871. :10.1111/j.1532-5415.2007.01186.x [DOI] [PubMed] [Google Scholar]
- 14. Li H, Manwani B, Leng SX. Frailty, inflammation, and immunity. Aging Dis. 2011;2:466–473. [PMC free article] [PubMed] [Google Scholar]
- 15. Spector TD, Hart DJ, Nandra D, et al. Low-level increases in serum C-reactive protein are present in early osteoarthritis of the knee and predict progressive disease. Arthritis Rheum. 1997;40:723–727. :10.1002/1529-0131(199704)40:4<723::AID-ART18>3.0.CO;2-L [DOI] [PubMed] [Google Scholar]
- 16. Vlad SC, Neogi T, Aliabadi P, Fontes JD, Felson DT. No association between markers of inflammation and osteoarthritis of the hands and knees. J Rheumatol. 2011;38:1665–1670. :10.3899/jrheum.100971 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Freund A, Orjalo AV, Desprez PY, Campisi J. Inflammatory networks during cellular senescence: causes and consequences. Trends Mol Med. 2010;16:238–246. :10.1016/j.molmed.2010.03.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Shoelson SE, Herrero L, Naaz A. Obesity, inflammation, and insulin resistance. Gastroenterology. 2007;132:2169–2180. :10.1053/j.gastro.2007.03.059 [DOI] [PubMed] [Google Scholar]
- 19. Rockwood K, Andrew M, Mitnitski A. A comparison of two approaches to measuring frailty in elderly people. J Gerontol A Biol Sci Med Sci. 2007;62:738–743. :62/7/738 [pii] [DOI] [PubMed] [Google Scholar]
- 20. Kiely DK, Cupples LA, Lipsitz LA. Validation and comparison of two frailty indexes: The MOBILIZE Boston Study. J Am Geriatr Soc. 2009;57:1532–1539. :10.1111/j.1532-5415.2009.02394.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. White DK, Niu J, Zhang Y. Is symptomatic knee osteoarthritis a risk factor for a trajectory of fast decline in gait speed? Results from a longitudinal cohort study. Arthritis Care Res (Hoboken). 2013;65:187–194. :10.1002/acr.21816 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. White DK, Felson DT, Niu J, et al. Reasons for functional decline despite reductions in knee pain: the multicenter osteoarthritis study. Phys Ther. 2011;91:1849–1856. :10.2522/ptj.20100385 [DOI] [PMC free article] [PubMed] [Google Scholar]