Bronchodilator treatment and asthma death: a new analysis of a British case-control study

We investigated the relationship between asthma mortality and long-acting beta(2)-agonists (LABA), including interactions with age, inhaled corticosteroids (ICS) and social deprivation. We used a new, expanded dataset of recorded medication extracted blind from the anonymised primary care records of an earlier British case-control study. The cases were 532 asthma deaths aged < 65 occurring between 1994 and 1998 and the controls were 532 asthma admissions, matched for age, hospital, and index date (date of death/asthma admission). The exposure periods prior to the index date were current (≤ 2 months) or recent ( > 2-6 months). We found no evidence of an overall association with current (OR = 0.89 [95% confidence interval 0.61-1.30]) or recent (1.08 [0.76-1.53]) mention of LABA, but there was some evidence of a positive interaction with age. Among controls with mention of LABA, a concurrent mention of ICS (within 1 month) was common (85% and 93% for the two respective periods) which limited our power to investigate any interaction between LABA and ICS. There was no coherent evidence of effect modification by social deprivation. In a population based case-control study where prescription of LABA without concomitant ICS was uncommon there was no evidence of an overall association between LABA and asthma death.