Longitudinal study of short-term corticosteroid use by working-age adults with diabetes mellitus: Risks and mitigating factors.

Abstract

This study assessed the frequency of short-term oral corticosteroid use in adults with diabetes, examined the incidence of fractures, venous thromboembolism (VTE), and hospitalization for sepsis after corticosteroid use, and evaluated whether preventative medications mitigated adverse events. A longitudinal study (2012-14) was conducted of 1 548 945 adults (aged 18-64 years) who received healthcare coverage through a large national health insurer. Incidence rate ratios (IRR) were calculated using conditional Poisson regression. Short-term oral corticosteroids were used by 23.9%, 20.8%, and 20.9% of adults with type 2 diabetes, type 1 diabetes, and no diabetes, respectively, during the 3-year period (P < 0.001). Baseline risks of fracture, VTE, and sepsis were greater for individuals with than without diabetes (P < 0.001). The combined effect of having diabetes and using corticosteroids was greater than the sum of the individual effects (synergy indices of 1.17, 1.23, 1.30 for fracture, VTE, and sepsis, respectively). The IRR for VTE in the 5-30 days after corticosteroid use was 3.62 (95% confidence interval [CI] 2.41-5.45). Fractures increased in the 5-30 days after corticosteroid use (IRR 2.06; 95% CI 1.52, 2.80), but concomitant use of ergocalciferol mitigated this risk (IRR 1.13; 95% CI 0.12, 11.07). The risk of hospitalization for sepsis was elevated with corticosteroid use (IRR 3.79; 95% CI 2.05, 7.01), but was mitigated by the concomitant use of statins. Short-term oral corticosteroid use is common in adults with diabetes and is associated with an elevated, but low, risk of adverse events. The findings suggest that preventative medications may mitigate risk.