We included adult patients (age > 20 yr) with respiratory diseases who were newly started on glucocorticoids from January 2003 through December 2008 at three hospitals affiliated with Seoul National University, Korea: Seoul National University Hospital, Seoul National University Bundang Hospital, and Seoul National University Boramae Hospital. The respiratory diseases treated with glucocorticoid included idiopathic interstitial pneumonia, endobronchial tuberculosis, sarcoidosis, and hypersensitivity pneumonitis. Patients with asthma and chronic obstructive pulmonary disease (COPD) were not included, not only because they frequently used inhaled steroid, but also because their use of systemic steroid was usually short term.

Patients with respiratory diseases treated with systemic glucocorticoid exceeding a prednisolone-equivalent dose of 20 mg/day for at least for 4 weeks were included in the analysis. Patients with pre-existing diabetes or an initial random glucose level exceeding 200 mg/dL before initiating steroid therapy were excluded. We also excluded patients who were treated with steroid for other than lung diseases, such as malignancy, rheumatoid arthritis, transplantation, and nephrotic syndrome.

The demographic characteristics, clinical findings, and laboratory results were obtained from a retrospective review of the medical records. In addition, detailed information on the use of glucocorticoid (daily dose, duration of treatment, and cumulative dose) was collected.

Steroid-induced DM (S-DM) was defined as a fasting glucose concentration above 126 mg/dL or a random glucose concentration exceeding 200 mg/dL at least twice after beginning steroid treatment.

The chi-squared test was used to compare categorical variables, and the t-test to compare continuous variables between patients with and without S-DM. The variables analyzed included age, sex, body mass index (BMI), underlying lung diseases, daily and cumulative dose of steroid, duration of treatment, and initial random serum glucose level. To identify predictors of S-DM, multiple logistic regression models were constructed that included any variables with P < 0.10. Statistical significance was considered if P < 0.05. All statistical analyses were performed using SPSS® (Version 12.0, Chicago, IL, USA).

The study protocol was approved by the institutional review boards of the Seoul National University Hospital (H-0906-059-284), Seoul National University Bundang Hospital (B-1001-092-103), and Seoul National University Boramae Hospital (06-2010-4).

The analysis included 231 patients with respiratory diseases who started taking systemic glucocorticoid during the study period. Their median age was 55 yr (range 22-85 yr), and 139 were female (60.2%). The baseline clinical characteristics and laboratory findings of the study population are summarized in Table 1. Respiratory diseases requiring glucocorticoid treatment included idiopathic interstitial pneumonia (125 patients), endobronchial tuberculosis (60 patients), sarcoidosis (43 patients), and hypersensitivity pneumonitis (three patients).

The median random glucose before initiating glucocorticoid in the 231 patients was 101 mg/dL (range 59-198 mg/dL). The median cumulative prednisolone-equivalent dose of glucocorticoid was 4,965 mg (range 560-32,585 mg), and the median duration of steroid treatment was 193 days (range 28-1,869 days). S-DM was diagnosed in 34 (14.7%) out of 231 patients.

The patients who developed S-DM were older than the patients who did not (65 vs 53 yr, P < 0.001). In addition, idiopathic interstitial pneumonitis was more common (79.4% vs 49.7%, P = 0.001), and endobronchial tuberculosis was less common (5.9% vs 29.4%, P = 0.004) among patients with S-DM than among those without S-DM. However, neither the cumulative dose nor the duration of glucocorticoid use was associated with the development of S-DM (Table 1).

Multiple logistic regression analysis revealed that only older age (odds ratio [OR] 1.05, 95% confidence interval [CI] 1.02-1.09) was an independent risk factor for developing S-DM (Table 2).

The median time of onset of S-DM was 91.5 days (range 15-1,733 days); 18 patients (52.9%) were diagnosed in the first 3 months, and five patients (14.5%) after 1 yr. Among the 23 patients followed for longer than 6 months after the diagnosis of S-DM, 10 were treated with oral hypoglycemic drugs and five with insulin. The other eight patients were observed without treatment. Treatment for S-SM could be stopped after a median of 202 days (range 47-1,621 days) in seven of the 15 patients in whom hypoglycemic agents and insulin were initiated.