ABSTRACT
PURPOSE: We evaluated associations between adiponectin and the risk of diabetes among patients with rheumatoid arthritis (RA), a systemic inflammatory disease associated with metabolic disturbance. METHODS: This prospective cohort study included adults with RA from the Veterans Affairs Rheumatoid Arthritis Registry. Adiponectin and inflammatory cytokines/chemokines were measured at enrollment on stored serum samples. Adiponectin levels were categorized, and clinical variables were described across categories (<10â µg/mL; 10-40â µg/mL; >40â µg/mL). Multivariable Cox proportional hazard models evaluated associations between adiponectin and incident diabetes adjusting for age, sex, race, smoking status, body mass index (BMI), disease-modifying therapy use, calendar year, and comorbidity. Testing for modification of effect in the context of elevated cytokines/chemokines was performed. RESULTS: Among 2595 patients included in the analysis, those with adiponectin levels >40â µg/mL (N = 379; 15%) were older and had lower BMI. There were 125 new cases of diabetes among 1689 patients without prevalent disease at enrollment. There was an inverse association between adiponectin and incident diabetes; however, the association was positive among patients with adiponectin levels >40â µg/mL. Patients with levels >40â µg/mL were at higher risk compared to those with levels 10-40â µg/mL (HR: 1.70 [1.34, 2.16] P < .001). Those with adiponectin levels >40â µg/mL had significantly higher levels of inflammatory cytokines with evidence of a modified effect of adiponectin on diabetes risk in the setting of inflammation. CONCLUSION: The relationship between adiponectin and incident diabetes risk is U-shaped in RA. Patients with very high adiponectin levels have greater systemic inflammation and an altered relationship between adiponectin and diabetes risk.