BackgroundIt is not elucidated that there is treatment-related damage in elderly patients with antineutrophil cytoplasmic antibody (ANCA)–associated vasculitis (AAV).MethodsElderly (≥ 75 years of age) patients were enrolled from two nationwide prospective inception cohort studies. The primary outcome was 12-month treatment-related Vasculitis Damage Index (VDI) score. Secondary outcomes included serious infections within 6 months, total VDI score, remission, and relapse. Patient characteristics and outcomes were compared across three different initial glucocorticoid (GC) dose groups: high-dose, prednisolone (PSL) ≥ 0.8 mg/kg/day; medium-dose, 0.6 ≤ PSL < 0.8 mg/kg/day; and low-dose, PSL < 0.6 mg/kg/day.ResultsOf the 179 eligible patients, the mean age was 80.0 years; 111 (62%) were female. The mean Birmingham Vasculitis Activity Score was 16.1. Myeloperoxidase-ANCA findings were positive in 168 (94%) patients, while proteinase 3-ANCA findings were positive in 11 (6%). The low-dose group was older and had higher serum creatinine levels than the other groups. There were no statistically significant intergroup differences in remission or relapse, whereas serious infection developed more frequently in the high-dose (29 patients [43%]) than the low-dose (13 patients [22%]) or medium-dose (10 patients [19%]) groups (p = 0.0007). Frequent VDI items at 12 months included hypertension (19%), diabetes (13%), atrophy and weakness (13%), osteoporosis (8%), and cataracts (8%). Logistic regression analysis revealed that GC dose at 12 months (odds ratio, 1.14; 95% confidence interval, 1.00–1.35) was a predictor for diabetes.ConclusionA reduced initial GC dose with rapid reduction might be required to ensure the safe treatment of elderly AAV patients.