Twenty-one studies comprising a total of 810 patients were included.
General effect: there was a significant but small tendency for corticosteroid therapy to be associated with being shorter than one's expected height (Z=2.328, p=0.0100, mean r=-0.023). The effect varied with the type of drug used: oral prednisone was significantly associated with decreased height (Z=2.137, p=0.0164, r=-0.295), BDP with normal stature (Z=7.395, p=2.17E-13, mean r=0.432), and 'other' oral corticosteroids showed a significant but weak association with reduced height (Z=9.107, p=2.44E-18, mean r=-0.260). Any growth-retarding effect of BDP appeared to be age-related: the correlation (r) between the size of BDP's effect on growth and patient age was -0.568 (Z=4.058, p=0.0000249). The predicted BDP-growth effect was calculated to be zero at age 11.2 years, indicating a cut-off point beyond which inhaled BDP was more likely to cause growth impairment.
Effects of therapy duration: greater duration of BDP therapy was associated with no growth impairment. (Z=1.977, r=0.273, p=0.0240). This lack of association remained when the effects of age were taken into consideration.
Effects of dosage: increased dosage of BDP was associated with an absence of growth impairment (Z=2.144, r=0.263, p=0.0160).
Effects of illness severity: greater dosages of BDP tended to be used in more severe cases though the trend was not significant (p=0.159). Patients with severe asthma showed a significantly weaker positive effect for attaining normal stature (Z=2.318, p=0.0102) than those with less severe asthma. Overall, however, there did not appear to be any significant associations between growth impairment and any combination of illness severity and BDP dosage. The number of studies reporting no effect, which would render the meta-analysis non significant, has been calculated by the authors: 2 studies reporting no general effect of therapy on growth would be required.