RESUMO
Physical frailty and cognitive impairment are risk factors for adverse outcomes in older people with osteoarthritis of the knee (knee OA). This cross-sectional study was conducted to determine the prevalence and associated factors of frailty and cognitive frailty among community-dwelling older patients with knee OA in four representative cities of Thailand. Data composed of three parts, Part 1: Demographic data, Part 2: The assessment of frailty by Fried phenotype and cognitive function by MiniCog and Part 3: The assessment of factors associated with frailty. Of 780 elders (mean age, 69.4 ± 6.9 years) screened, 101 (12.9%) were classified to be frail, 511 (65.6%) pre-frail and 168 (21.5%) non-frail. The prevalence of cognitive frailty was 2.44%. The correlation between physical activity rated by the Global Physical Activity Questionnaire (GPAQ) and self-rated methods was high (kappa 0.721; p < 0.001). Self-rated physical activity yielded similar prevalence of frail (9.4%), pre-frail (69.1%) and non-frail (21.5%). In multivariate analysis, aging (OR 3.42; 95% CI 1.16-10.11), severe knee OA symptoms (OR 18.96; 95% CI 3.53-101.65), malnutrition (OR 2.50; 95% CI 1.23-5.09), and functional dependence (OR 3.94; 95% CI 1.19-13.03) were associated with frailty. The prevalence of frailty and pre-frailty was high in knee OA and associated with aging, severe knee OA symptoms, malnutrition, and functional dependence, whereas the prevalence of cognitive frailty was not uncommon in community-dwelling elderly. Physical activity rated by the GPAQ and self-rated methods were highly correlated. Self-rated physical activity may be used in community surveys of frailty.