Background The effect of glucocorticoids on organ damage in lupus has been suggested in previous studies, with some relations described with specific domains. Ideal prednisone doses have not yet been described.

Objectives To analyse the association of different doses of prednisone with damage attributable to steroids in an observational cohort of patients with systemic lupus erythematosus (SLE) with a follow-up of at lest 5 years.

Methods Demographic, clinical and therapeutical variables were extracted from our database. Five prednisone-related variables were constructed: the maximum dose (mg/day) received at least for 3 consecutive days during the 1 ^st year, the mean daily dose received at the end of the 1 ^st and 4 ^th year after the diagnosis and a categorical variable based on the mean daily dose at the end of the 1^st and 4^th years, with the following categories: no prednisone, up to 7.5 mg/d (low dose), up to 30 mg/d (medium dose) and over 30 mg/d (high dose). Damage attributable to steroids was considered if accrued at any time after the first 6 months up to the 5th year of follow up in any of the following SLICC damage index (SDI) domains: cataracts, avascular necrosis (AVN), osteoporosis and diabetes. We excluded those cases with such damage within the first 6 months. Activity was measured using SLEDAI, with a variable consisting of the mean maximum score during years 1 to 5.

Results 203/225 (90.2%) patients were women. The mean age (SD) at diagnosis was 35.5 (15) years. 18 (8%) patients had accrued damage attributable to steroids within 5 years after the diagnosis of SLE. The most frequent damage was osteoporosis (7 patients), cataracts (6 patients), AVN (4 patients) and diabetes (2 patients). 1 patient developed cataracts and osteoporosis. As expected, the mean (SD) SDI score increased over time from 0.2 (0.5) at 6 months to 0.6 (0.9) at 5 years. The maximum dose of prednisone received within the 1 ^st year was significantly higher in patients with damage attributable to steroids (42 vs. 25 mg/d, respectively, p=0.012). Likewise, the mean daily dose of prednisone after the 1^st and 4^th year was significantly higher in patients with damage: 20 vs. 11 mg/d (p=0.017) and 11 vs. 7 mg/d (p=0.04), respectively.

With regard to the different prednisone dose categories up to the end of the 4^th year, the univariate logistic regression showed an increased risk for damage in patients treated with medium doses vs. those taking no prednisone (OR 7.3, 95% CI 0.9-58), with no difference between the latter and patients receiving low doses (OR 1.7 95%CI 0.18-16). Patients taking high doses were very few (n=4) and were not analysed. Considering the prednisone dose categories at the end of the 1^st year, only patients taking high doses were at a higher risk of accruing damage than patients not taking prednisone (OR 20.76, 95% 2.32-193).

Conclusions Prednisone causes cataracts, osteoporosis, avascular necrosis and diabetes in patients with SLE. Long term doses above 7.5 mg/d should be avoided.

Acknowledgements Zonana-Nacach A et al. Arthritis Rheum 2000;43:1801-8. Thamer M et al. J Rheum 2009;36:560-4.

Disclosure of Interest None Declared