Objective. Sarcopenic obesity is a body composition category in which obesity is accompanied by low skeletal muscle mass, offsetting the increase in body weight caused by increased adipose tissue. The purpose of this study was to analyze the association between knee osteoarthritis (OA) and 4 different categories of body composition: normal, sarcopenic nonobesity, nonsarcopenic obesity, and sarcopenic obesity.Methods. This was a cross-sectional study using the data from 2,893 participants in the Fifth Korean National Health and Nutrition Examination Survey. Radiographic knee OA was defined as a Kellgren/ Lawrence grade of >2. Appendicular skeletal muscle mass (ASM) and whole-body fat mass were measured using dual x-ray absorptiometry. Sarcopenia was defined as a skeletal muscle mass index (ASM/body weight [%]) below -2SD of the value in sex-matched young reference groups. Nonsarcopenic obesity was defined as a body mass index (BMI) >27.5 kg/m 2 .Results. The prevalence of each body composition category was as follows: 83.5% normal, 4.3% sarcopenic nonobesity, 9.2% nonsarcopenic obesity, and 3.0% sarcopenic obesity. Compared with nonsarcopenic obesity participants, participants with sarcopenic obesity were significantly older, had lower ASM, higher whole-body fat mass, and higher waist circumference. Conclusion. Sarcopenic obesity was more closely associated with knee OA than was nonsarcopenic obesity, although both groups had equivalent body weight. This finding supports the importance of the systemic metabolic effect of obesity in knee OA.Osteoarthritis (OA) is the most common articular disorder in the elderly, with ϳ40% of those over the age of 65 years affected at the knee joint (1). Although the exact pathophysiology of OA is unclear, the association between OA and obesity is well known (2-4). The biomechanical effect of increased body weight is one explanation for this association, especially in weightbearing joints such as the knee (5).The association between OA and obesity may be explained by factors other than biomechanical stress. Obesity is associated with OA not only in the weightbearing joints, but also in the joints of the hand (6), where mechanical load or factors such as grip strength (7) are unlikely to play an important role. Adipose tissue is an endocrine organ that secretes various adipokines (8). Data obtained in vitro have suggested a role of adipokines, including leptin, adiponectin, resistin, and visfatin, in the pathophysiology of OA (9). Recently, leptin, acting synergistically with other inflammatory cytokines, was shown to have a catabolic effect on articular cartilage metabolism via the up-regulation of proteolytic enzymes (10). In addition, extreme obesity induced in leptin (receptor)-deficient mice did not cause knee OA in the absence of leptin signaling (11).