The aim of the study was to investigate all-cause and disease-specific mortality in a large population-based cohort of oral glucocorticoid (GC) users.
This was a retrospective, matched cohort study. Information on dispensed prescriptions was obtained from the Swedish Prescribed Drug Register. The cause of death was obtained from the Swedish Cause-of-Death Registry. Patients receiving prednisolone ≥5 mg/day (or equivalent dose of other GC) for ≥21 days between 2007-2014 were included. For each patient, one control subject matched for age and sex was included. The study period was divided into 3-month periods and patients were divided into groups according to a defined daily dose (DDD) of GC used per day. The groups were: Non-users (0 DDD per day), low-dose users (>0 but <0.5 DDD per day), medium-dose users (0.5-1.5 DDD per day) and high-dose users (>1.5 DDD per day). Hazard ratios (HRs), unadjusted and adjusted for age, sex and comorbidities, were calculated using a time-dependent Cox proportional hazard model.
Cases (n=223 211) had significantly higher all-cause mortality compared to controls (HR adjusted for age, sex and comorbidities 2.08, 95% confidence interval 2.04 to 2.13). After dividing the cases into subgroups, adjusted HR was 1.31 (1.28 to 1.34) in non-users, 3.64 (3.51 to 3.77) in low-dose users, 5.43 (5.27 to 5.60) in medium-dose users and, 5.12 (4.84 to 5.42) in high-dose users. The highest adjusted hazard ratio was observed in high-dose users for deaths from sepsis 6.71 (5.12 to 8.81) and pulmonary embolism 7.83 (5.71 to 10.74).
Oral GC users have an increased mortality rate compared to the background population, even after adjustment for comorbidities. High-dose users have an increased risk of dying from sepsis, and pulmonary embolism compared to controls. Whether the relationship between GC exposure and the excess mortality is causal remains to be elucidated.