authors’ reply Marie-France Beauchesne highlights the rise in serum osteocalcin levels and bone mineral density seen in our study and raises possible reasons for this result. The study did not obtain data on calcium and vitamin D intake but no patients received supplemental calcium or vitamin D during the study period. Exercise was broken down into light (32%), moderate (56%), and heavy (12%) average weekly physical exercise; 83% of the patients were non-smokers and 17% were current smokers; 31% drank no alcohol and, of the 69% consuming alcohol, the weekly median intake was 5 units (25th percentile 2.0; 75th percentile 14).

Although calcium and vitamin D intake is clearly important in assessing overall bone loss, our patients were not taking supplements and therefore this should not have influenced the rise in bone mineral density and osteocalcin levels seen in the study.

Of our study population 76% had been taking beclomethasone dipropionate prior to being enrolled in the study and, as both Eganet al 1-1 and Pauwelset al 1-2 reported either no change or increases in bone mineral density in patients on fluticasone propionate but decreases in those taking beclomethasone dipropionate, the type of inhaled corticosteroid may be important in determining bone mineral density and bone marker changes over time. Furthermore, Fordet al 1-3 found no differences in lumbar bone mineral density after a two year period when 500 μg fluticasone propionate given twice daily via a diskhaler was compared with placebo.

Future studies are therefore needed to differentiate between possible effects from specific inhaled corticosteroids. General statements concerning the safety and efficacy of inhaled corticosteroids need to take into account these differences.