Inhaled, nasal and topical corticosteroid burden and association with corticosteroid-associated adverse events

Patients with type 2 inflammation-associated diseases are often treated with multiple formulations of corticosteroids. Long-term risks of using combinations of local corticosteroids are poorly characterised.

Using Danish nationwide databases, investigate associations between local corticosteroid exposure and 11 corticosteroid exposure-associated conditions compared to non-users, and eventual dose-response relationship for exposed individuals.

Systemic corticosteroid-naïve adults who redeemed inhaled, nasal and/or topical corticosteroids 2000-2012 were matched with non-initiators and followed for up to 23 years. Overall (inhaled, nasal and topical) exposure was calculated for 6-month periods as low (<1.25 daily prednisolone-equivalent mg), moderate (1.25-2.5 mg) or high (>2.5 mg). Competing risk Cox models with time-varying effects were used, presented as HR per additional 6-month period of exposure.

Overall, 292,393 local corticosteroid users were included (mean age 45, 47.7% female). Corticosteroid exposure-associated condition incidences were generally low, but all conditions had higher subdistribution-HRs in users, ranging from 1.4 to 3.8 for cataracts and obesity (p<0.001).For local corticosteroid users, increased risk per 6-month exposure was observed for osteoporosis (low-high dose: HRs 1.01-1.10), obesity (1.03-1.03) and pneumonia (1.03-1.15). Risk of cardiovascular disease, heart failure, cataract and gastroesophageal reflux disease increased with high-dose exposure only at HRs 1.03, 1.02, 1.04 and 1.04 per additional 6-month period of exposure, respectively.Stratified by formulation, high-dose inhaled and topical corticosteroids demonstrated the strongest associations.

In individuals using local corticosteroids, the overall corticosteroid burden, including shorter periods of high-dose exposure, associates with corticosteroid exposure-associated adverse events albeit with generally low incidence and hazard.