Background/Purpose: Both knee osteoarthritis (OA) and the frailty syndrome affect older adults and both are associated with functional limitation and disability. Frailty in elders is a state of increased vulnerability to adverse outcomes, such as falls, fractures, hospitalization and even death. While frailty is perceived to occur in thin older adults, it has been shown to be present in those who are obese, a common feature of those with knee OA. If knee OA and frailty are associated, then by extension those with knee OA might be at risk for not only the known OA outcomes of pain, functional limitation, and disability, but also the adverse outcomes related to frailty. We therefore examined the cross-sectional association of knee OA and frailty in community-dwelling older adults using data from two large cohorts.

Methods: The Multicenter Osteoarthritis (MOST) Study and the Osteoarthritis Initiative (OAI) are two NIH-funded longitudinal observational studies of individuals with or at high risk for knee OA. We included subjects from these two studies who had knee x-rays read and information on frailty parameters available at baseline and at the 30-mo (MOST) or 24-mo (OAI) follow-up visits. Prevalent knee OA was defined at the follow-up visit described above as: 1) Radiographic knee OA (ROA): Kellgren and Lawrence (KL) grade ≥2; 2) Symptomatic knee ROA: presence of ROA plus frequent knee pain; 3) Severity of ROA: highest KL grade of either knee, with replaced knees considered to be KL grade 4; 4) Number of knees with ROA: subjects categorized as having no knee ROA, unilateral knee ROA, or bilateral knee ROA. Frailty was defined using the Study of Osteoporotic Fractures (Ensrud) index as presence of 2 of 3 of the following criteria: 1) Weight loss >5% between baseline and the follow-up visit; 2) Inability to rise from chair 4 times without using support at the follow-up visit; 3) Poor energy from the SF12 questionnaire at the follow-up visit. We evaluated the cross-sectional association of knee OA (4 definitions) with prevalent frailty at the follow-up visit using Poisson regression to calculate prevalence ratios (PR), adjusting for age, sex, BMI, physical activity, education, smoking, co-morbidities (modified Charlson score), race and study site.

Results: Among 7822 participants (3026 MOST, 4796 OAI; mean age 62±8.81, 59% women, mean BMI 29±5.44), there were 213 (186 poor energy, 116 inability to rise from chair and 145 weight loss) prevalent frail subjects.  Prevalence of frailty was higher in those with radiographic knee OA and symptomatic knee OA, and increased in prevalence with increasing x-ray severity and with number of knees involved with OA (Table).

Conclusion: Frailty is present more frequently in persons with knee OA than those without. Further research is needed to explore whether knee OA predisposes to frailty, and whether early management of knee OA might prevent frailty and its related adverse outcomes.