PT - JOURNAL ARTICLE AU - Richard T. Wang AU - Cheri Silverstein AU - J. Wes Ulm AU - Ivana Jankovic AU - Ascia Eskin AU - Ake Lu AU - Vanessa Rangel Miller AU - Rita M. Cantor AU - Ning Li AU - Robert Elashoff AU - Ann S. Martin AU - Holly Peay AU - Stanley F. Nelson TI - Online Self-Report Data for Duchenne Muscular Dystrophy confirms natural history and can be used to assess for therapeutic benefits AID - 10.1101/012344 DP - 2014 Jan 01 TA - bioRxiv PG - 012344 4099 - http://biorxiv.org/content/early/2014/12/08/012344.short 4100 - http://biorxiv.org/content/early/2014/12/08/012344.full AB - To assess the utility of online patient self-report outcomes in a rare disease, we attempted to observe the effects of corticosteroids in delaying age at fulltime wheelchair use in Duchenne muscular dystrophy (DMD) using data from 1,057 males from DuchenneConnect, an online registry. Data collected were compared to prior natural history data in regard to age at diagnosis, mutation spectrum, and age at loss of ambulation. Because registrants reported differences in steroid and other medication usage, as well as age and ambulation status, we could explore these data for correlations with age at loss of ambulation. Using multivariate analysis, current steroid usage was the most significant and largest independent predictor of improved wheelchair-free survival. Thus, these online self-report data were sufficient to retrospectively observe that current steroid use by patients with DMD is associated with a delay in loss of ambulation. Comparing commonly used steroid drugs, deflazacort prolonged ambulation longer than prednisone (median 14 years and 13 years, respectively). Further, use of Vitamin D and Coenzyme Q10, insurance status, and age at diagnosis after 4 years were also significant, but smaller, independent predictors of longer wheelchair-free survival. Nine other common supplements were also individually tested but had lower study power. This study demonstrates the utility of DuchenneConnect data to observe therapeutic differences, and highlights needs for improvement in quality and quantity of patient-report data, which may allow exploration of drug/therapeutic practice combinations impractical to study in clinical trial settings. Further, with the low barrier to participation, we anticipate substantial growth in the dataset in the coming years.