[Glucocorticoid-induced Diabetes in Severe Acute Respiratory Syndrome: the Impact of High Dosage and Duration of Methylprednisolone Therapy]

Objective: Over-dose glucocorticoid was frequently used in patients with severe acute respiratory syndrome (SARS), the aim of the present study is to investigate the frequency and risk factors of corticosteroid-induced diabetes in this situation.

Methods: One hundred thirty-three cases of SARS admitted from May to June 2003 in our hospital were included in the study. Maximal dose, average daily dose, treatment course and fasting plasma glucose level (FPG) were recorded. Corticosteroid-induced diabetes was diagnosed if FPG was equal to or higher than 7 mmol/L twice or more after the administration of glucocorticoids.

Results: Ninety five patients (71.9%) were prescribed with glucocorticoid among the 132 patients. Thirty-three patients (36.3%) were diagnosed as corticosteroid-induced diabetes according to the FPG. Significant differences were found in daily maximal dosage, duration of treatment and average dosage of methylprednisolone between the patients with and without corticosteroid-induced diabetes (275 mg/d vs 136 mg/d, 24 d vs 16 d, 139 mg/d vs 91 mg/d, P < 0.01). As compared with the lowest tertile, the frequency of diabetes in SARS patients treated with highest tertile of maximal daily dosage, treatment duration, average dosage or total dosage were significantly higher (64.7% vs 13.0, 61.9% vs 17.4%, 62.5% vs 21.7%, and 59.2% vs 13.6%, P < 0.05 for all comparisons). If the patients were treated with an average dose less than 90 mg/d and treatment duration shorter than 15 days, the diabetes incidence was 10.5%. After adjusting age and sex, the daily maximal dosage of methylprednisolone was the only factor that might predict the occurrence of diabetes.

Conclusions: Over-dose administration of methylprednisolone in SARS patients leads to a high frequency of diabetes. Decreasing the daily maximal dosage may be beneficial to the reduction of corticosteroid-induced of diabetes.