In Reply: Dr Di Comite and Mr Rossi suggest that cumulative prednisone dosage be determined to assess the contribution of this drug to the incidence of diabetes, indicating that (1) a lower prednisone dosage would likely be required for those taking hydroxychloroquine and (2) those patients who had never taken hydroxychloroquine might have had more exposure to prednisone based on their greater age, rheumatoid arthritis duration, and disability. In our time-to-event analyses, we adjusted for cumulative duration of prednisone use; our time-varying Cox proportional hazards analyses included adjustment for prednisone use (yes/no) in each cycle.
Wasko MCM, Hubert HB, Lingala VB, Elliott JR, Luggen ME, Fries JF, Ward MM. Risk of Diabetes in Patients With Rheumatoid Arthritis Taking Hydroxychloroquine—Reply. JAMA. 2007;298(20):2368-2370. doi:10.1001/jama.298.20.2369