Department of Medicine, Endocrine and Diabetes Unit (B.A.), University of Wuerzburg, D-97080 Germany; and Division of Medical Sciences (W.A.), University of Birmingham, Queen Elizabeth Hospital, Birmingham, B15 2TH, United Kingdom
Address correspondence to: Dr. Wiebke Arlt, M.D., Division of Medical Sciences, University of Birmingham, Queen Elizabeth Hospital, Edgbaston, Birmingham, B15 2TH, United Kingdom. E-mail w.arlt{at}bham.ac.uk.
To the editor:
Recent double-blind, randomized crossover studies have demonstrated significant beneficial effects of dehydroepiandrosterone (DHEA) replacement on mood, sexuality, and health-related quality of life (HRQoL) in patients with adrenal insufficiency (1, 2). HRQoL in patients with both primary and secondary adrenal insufficiency is significantly impaired (3, 4). In a cohort of 88 Addison patients, Lovas et al. (3) found significantly impaired HRQoL scores, similar to those usually seen in chronic heart failure, with a significant impact on the incidence of disablement pensions (41% in Addisons vs. only 17% in the general population for ages 4067 yr; P < 0.001).
In the March issue of the JCEM, Lovas et al. (5) have published the results of a 9-month, randomized, parallel group clinical trial investigating the effects of DHEA replacement in women with adrenal failure, failing to find significant improvements in HRQoL during DHEA administration. However, this outcome could have been easily predicted because their study is severely underpowered. Statistical power calculations show that a parallel study aiming to obtain significant effects similar to those reported by recent crossover studies (1, 2), which used 2439 patients, would require the inclusion of at least 100 patients to ascertain adequate statistical power. Power calculations are of particular importance for parallel studies using HRQoL measurements, because HRQoL scores generally show broad interindividual variability. In addition, the study presented by Lovas et al. (5) is the result of a multicenter approach, reporting the results of DHEA treatment in 15 patients from five different centers. For a reliable power calculation, this would push the number needed to treat beyond even 100 because a multicenter study necessarily shows a greater variability in results, again, in particular with regard to HRQoL scores.
Results from underpowered studies are definitely noteworthy if they show significant effects, like the recent study by Johannsson et al. (6) on DHEA replacement in women with secondary adrenal failure. However, negative results from an underpowered study like the paper by Lovas et al. (5) do not contribute to furthering our knowledge in this field. By contrast, this paper may adversely affect it because it will be cited as negative evidence against DHEA replacement in adrenal insufficiency, despite its lack of statistical power.
Results from an adequately powered parallel trial on DHEA replacement including 106 patients with primary adrenal insufficiency will hopefully be available soon; preliminary results published in abstract form (7, 8) reported significant improvements in HRQoL scores in both men and women with primary adrenal failure. The scientific community needs the results from well-designed, adequately powered, larger scale studies to securely establish the role of DHEA in the standard replacement regimen for adrenal insufficiency.
Received March 27, 2003.
References
HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |