Body mass index at diagnosis affects disease course in juvenile dermatomyositis

Abstract

Purpose: We investigated the hypothesis that obesity at diagnosis alters the disease course, response to treatment, and complication rate in juvenile dermatomyositis (JDM). Methods: Institutional Review Board approval was obtained to retrospectively review the charts of 73 patients with JDM seen in pediatric rheumatology clinic at Nationwide Children's Hospital over the past 23 years. Data on treatment and outcomes were collected up through the last clinic visit, or July 30, 2010, whichever came first. Body mass index (BMI) was calculated using 'Centers for Disease Control and Prevention - BMI calculator for children and teenagers'. We used the Wilcoxon two-sample test to compare numerical variables between the group of patients who were obese (BMI >85th percentile) and the group who were non-obese (BMI ≤85th percentile) at their initial visit. We used logistic regression to compare categorical variables between obese and non-obese groups in SAS 9.2. Results: Patient age ranged from 0-18 years. Twenty eight patients had age- and sex-specific BMI ≥ 85th percentile, and 45 had < 85th percentile. Average BMI of obese patients was 93rd percentile and of non-obese patients was 46th percentile. Comparatively fewer patients in the obese group received corticosteroid pulses at diagnosis (p=0.027) and developed osteonecrosis (p=0.050). Comparatively more patients in the obese group had elevated muscle enzymes at 12 months (p=0.027), developed calcinosis (p=0.047), received methotrexate for longer time (p=0.040), were treated with plaquenil (p=0.04), continued to take steroids at last follow up (p=0.028), and continued to have active disease at 5 years (p=0.006), 8 years (p=0.050) and 10 years (p=0.045) following diagnosis. Conclusion: JDM patients who were obese at diagnosis (BMI > 85th percentile) were less likely to receive pulse corticosteroids and had fewer bone complications. On the other hand, they had a more severe and prolonged disease course. It is not clear if these differences were related treatment choices, or to metabolic differences such as secretion of proinflammatory cytokines by adipose tissue, alteration of pharmacokinetics of medications, or altered sensitivity and response of the hypothalamicpituitary- adrenal axis including response to corticosteroids.