Arthritis & Rheumatism, Volume 62,
November 2010 Abstract Supplement

Abstracts of the American College of
Rheumatology/Association of Rheumatology Health Professionals
Annual Scientific Meeting
Atlanta, Georgia November 6-11, 2010.

Frailty and 3-Year Mobility Outcome in Persons with or at Higher Risk for Knee Osteoarthritis.

Cahue4,  September, Hayes2,  Karen W., Song4,  Jing, Dunlop3,  Dorothy D., Chmiel4,  Joan S., Hochberg5,  Marc C., Colbert1,  Carmelita J.

Northwestern University, LaGrange, IL
Northwestern University, Downers Grove, IL
Northwestern University, Chicago, IL
Northwestern University
University of Maryland, Baltimore, MD


Frailty has been examined in the context of aging, but little is known about its role in health outcomes in persons with knee osteoarthritis (OA). Risk of knee OA increases with age, and knee OA is a leading cause of chronic disability in older individuals; however, how best to identify persons who are likely to decline and/or benefit from intervention is unknown. The Osteoarthritis Initiative (OAI) cohort represents the continuum of knee OA from the stage of high risk prior to radiographic change through advanced disease and is uniquely suited to our goals: 1) determine the prevalence of frailty in this population; and 2) evaluate whether frailty at baseline predicts subsequent 3-year mobility outcome.


OAI data from public release data set 0.2.2 were used. We adapted the frailty index of Ensrud et al (Arch Int Med 2008), including inability to rise five times from a seated position without using arms and self-reported exhaustion (CES-D) (but not the weight criterion due to varied implications of weight loss in this population and because BMI was the basis of enrollment for a subset of the OAI cohort). Mobility outcomes were: baseline-to-186-year increase in difficulty in WOMAC mobility items (walking, ascending stairs, descending stairs); poor SF12 physical component outcome (persons were grouped by quintile of baseline function, and "poor" defined as moving into a worse quintile group or remaining in the worst 2 groups at 3 years). We used logistic regression to analyze the relationship between frailty at baseline (>= 1 item) (vs. not frail, reference) and baseline-to-186-year outcome, adjusting for age, gender, race, education, comorbidities, and K/L grade (worse of the two knees) in the overall sample and in BMI strata.


In 4138 persons, mean age was 61 years, mean BMI was 28, and 58% were women. At baseline, 14% of women and 11% of men were frail. Frailty was also associated with being African-American, having a lower education level, and poorer self-assessed health. By three years, 35% had worse WOMAC mobility, and 52% had poor SF12 physical component outcome. In the overall sample (Table), frailty at baseline was associated with worse 3-year WOMAC mobility and poor SF12 physical component outcome in adjusted analyses. These results persisted in the normal weight and overweight strata but were not significant in the obese stratum (Table).


A simple frailty index, easily usable in a clinical setting, was independently associated with worse 3-year mobility in persons at high risk for or with knee OA. These results also suggest that a frailty index may not function similarly across BMI strata in this population. Further work to refine the ability to capture the frailty phenotype—decreased reserve and resistance to stressors—should occur, as a simple but potentially powerful means of ultimately aiding the prevention of adverse outcomes in persons with knee OA.

Table. Frailty and 3-Year Mobility Outcome, Odds Ratios (OR) (95% CI)

 WOMAC Mobility Item Worsening OR (95% CI)Poor SF12 Physical Component Outcome OR (95% CI)
Overall N = 4138**1.73 (1.44, 2.08)1.56 (1.29, 1.89)1.79 (1.48, 2.17)1.63 (1.34, 1.99)
BMI CategoriesUnadjustedAdjusted*UnadjustedAdjusted*
Normal (BMI <25) N = 1027**2.06 (1.36, 3.11)1.79 (1.16, 2.75)1.55 (1.02, 2.37)1.48 (0.96, 2.28)
Overweight (25<=BMI<30) N = 1027**1.91 (1.42, 2.57)1.71 (1.25, 2.32)2.30 (1.67, 3.17)2.21 (1.59, 3.08)
Obese (BMI >=30) N = 1483**1.33 (1.00, 1.76)1.27 (0.94, 1.70)1.39 (1.03, 1.88)1.23 (0.89, 1.68)
*Adjusted for age, gender, race, education, comorbidities, K/L grade
**SF-12 outcome sample sizes; Overall n = 3968; BMI Categories; Normal n = 967; Overweight n = 1569; Obese n = 1432

To cite this abstract, please use the following information:
Cahue, September, Hayes, Karen W., Song, Jing, Dunlop, Dorothy D., Chmiel, Joan S., Hochberg, Marc C., et al; Frailty and 3-Year Mobility Outcome in Persons with or at Higher Risk for Knee Osteoarthritis. [abstract]. Arthritis Rheum 2010;62 Suppl 10 :186
DOI: 10.1002/art.27955

Abstract Supplement

Meeting Menu