BACKGROUND: Inhaled corticosteroids have been recommended as an important part of asthma therapy in children; however, there have been concerns about long-term side effects of these medications. In 1994, Allen1 published a meta-analysis of trials of inhaled steroids that suggested that inhaled corticosteroids (beclomethasone dipropionate) were not associated with growth delay. Because of some methodologic criticisms of that meta-analysis and because of 3 newer studies on the topic, the authors of this Cochrane review decided to reexamine the literature.
POPULATION STUDIED: The authors performed an exhaustive search of the literature using the methods typically employed by Cochrane reviewers, including searching the Cochrane Airways Group Asthma Trials registry, searching bibliographies of trials on the subject, and contacting colleagues and researchers in the field. The authors selected only studies involving children (aged <18 years) with asthma who had not been taking inhaled or oral steroids for at least 3 months. These studies had to be randomized controlled trials comparing beclomethasone with nonsteroidal medication and had to have data from which linear growth velocity could be calculated. Interestingly, there was no overlap between the studies evaluated (regardless of inclusion) for this review and the studies included in the meta-analysis by Allen. Only 3 studies were found that met the inclusion criteria. The patients in these studies were diagnosed with clinically stable asthma in the mild to moderate category. No information on the ages of the subjects was available from the primary authors. All of these trials used 200 mg of beclomethasone delivered by diskhaler (a dry powder inhaler) for 7 to 12 months.
STUDY DESIGN AND VALIDITY: The review methodology used was standard for the Cochrane Collaboration. The authors assessed study quality by examining randomization adequacy, allocation concealment, blinding, and description of withdrawals. They performed the final study selection independently and resolved disagreement by consensus. They both abstracted the data and contacted the primary authors of the original studies to fill in data where needed. The appropriate subgroup and sensitivity analyses were planned and performed when necessary.
OUTCOMES MEASURED: The main outcome measure was change in growth velocity (measured in cm/yr).
RESULTS: There were some important differences between the trials that could have implications for the generalizability and validity of the review, such as the definition of asthma and the 10% to 25% dropout rates. The dropout rates were adequately explained in each study, but only one study used intention-to-treat analysis to compensate for the dropout rates. There was a mean reduction in growth velocity of 1.54 cm per year (95% confidence interval, 1.15-1.94 cm/yr), corresponding to a reduction in growth velocity of 25%. The sensitivity analyses performed for methodologic quality, publication bias, and statistical model did not reveal any significant concerns for the validity of the meta-analysis, and there was no significant heterogeneity between the studies.
The authors of this study found a decrease in growth velocity with chronic administration of inhaled beclomethasone in children with asthma. The lack of published data about other inhaled steroids makes it difficult to generalize this finding. This reduction does seem, however, to be independent of the usual confounders of height (eg, severity of asthma, parental height, and so forth). In addition, there were no data in this review concerning final adult height to address the clinical impact of steroids on the age-related change in growth velocity. Given the concerning results of this study, clinicians should counsel families about the beneficial effects of inhaled steroids on controlling asthma and their possible negative effects on growth and should minimize the dose of any required inhaled steroid therapy.