Metabolic syndrome-associated osteoarthritis (MetS-OA) is a distinct osteoarthritis phenotype defined by the coexistence of MetS or its individual components. Despite the high prevalence of MetS-OA, its pathogenic mechanisms are unclear. The aim of this study was to determine the role of cellular senescence in the development of MetS-OA. Analysis of the human osteoarthritis initiative (OAI) dataset was conducted to investigate the MRI subchondral bone features of MetS-human OA participants. Joint phenotype and senescent cells were evaluated in two MetS-OA mouse models: high-fat diet (HFD)-challenged mice and STR/Ort mice. In addition, the molecular mechanisms by which preosteoclasts become senescent as well as how the senescent preosteoclasts impair subchondral bone microenvironment were characterized using in vitro preosteoclast culture system. Humans and mice with MetS are more likely to develop osteoarthritis-related subchondral bone alterations than those without MetS. MetS-OA mice exhibited a rapid increase in joint subchondral bone plate and trabecular thickness before articular cartilage degeneration. Subchondral preosteoclasts undergo senescence at the pre- or early-osteoarthritis stage and acquire a unique secretome to stimulate osteoblast differentiation and inhibit osteoclast differentiation. Antagonizing preosteoclast senescence markedly mitigates pathological subchondral alterations and osteoarthritis progression in MetS-OA mice. At the molecular level, preosteoclast secretome activates COX2-PGE2, resulting in stimulated differentiation of osteoblast progenitors for subchondral bone formation. Administration of a selective COX2 inhibitor attenuated subchondral bone alteration and osteoarthritis progression in MetS-OA mice. Longitudinal analyses of the human Osteoarthritis Initiative (OAI) cohort dataset also revealed that COX2 inhibitor use, relative to non-selective nonsteroidal antiinflammatory drug use, is associated with less progression of osteoarthritis and subchondral bone marrow lesion worsening in participants with MetS-OA. Our findings suggest a central role of a senescent preosteoclast secretome-COX2/PGE2 axis in the pathogenesis of MetS-OA, in which selective COX2 inhibitors may have disease-modifying potential. This work was supported by the National Institutes of Health grant R01AG068226 and R01AG072090 to MW, R01AR079620 to SD, and P01AG066603 to XC.