LETTER TO THE EDITOR e3028 J Clin Endocrinol Metab, August 2020, 105(8):e3028–e3029 https://academic.oup.com/jcem doi:10.1210/clinem/dgaa320 ISSN Print 0021-972X ISSN Online 1945-7197 Printed in USA © Endocrine Society 2020. All rights reserved. For permissions, please e-mail: journals. permissions@oup.com Received 22 April 2020. Accepted 22 May 2020. First Published Online 8 June 2020. Corrected and Typeset 23 June 2020. Letter to the Editor: “Rate and Extent of Recovery from Reproductive and Cardiac Dysfunction Due to Androgen Abuse in Men” Jon J. Rasmussen 1,2 and Caroline Kistorp 1,3 1 Department of Endocrinology, University Hospital of Copenhagen, Rigshospitalet, Copenhagen, Denmark; 2 Department of Internal Medicine, Endocrinology Unit, Holbæk University Hospital, Holbæk, Denmark; and 3 Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark ORCiD numbers: 0000-0003-2898-1771 (J. J. Rasmussen); 0000-0002-3019-6775 (C. Kistorp). I llicit use of anabolic androgenic steroids (AASs) among men has become a public health concern, al- though, until recently, knowledge on the side-effects were limited. Clinical studies among illicit AAS users are few, but in recent years we and others have reported long-lasting cardiovascular and reproductive side- effects among former AAS users including lower plasma total testosterone (TT), lower testicular volume, erectile dysfunction, and decreased libido (1-7). In a recent cross-sectional study, Shankara-Narayana et al. report no difference in serum TT between former AAS users and non-AAS users (8). The authors conclude that testicular dysfunction in AAS users, besides tes- ticular volume, is fully reversible within 7 to 18 months following AAS cessation. The cross-sectional study by Shankara-Narayana et al. is in many aspects compar- able to the study we previously conducted with regards to study design, number of participants, age and body composition of participants, AAS exposure in AAS users, and elapsed duration since AAS cessation among former AAS users (1-5). However, the conclusion by the authors of fully re- versible testicular function among former AAS users seems questionable. The results of the study by Shankara- Narayana et al. may simply confrm our previous fndings of long-lasting decreased testicular volume and im- paired endogenous testosterone production in previous AAS users (1). The reported data in the present study clearly suggest that former AAS users had lower serum TT than non-AAS users, with mean ± standard error of mean (SEM) of 21.5 ± 13.2 versus 30.2 ± 16.0 nmol/L (SI units), respectively (8). Thus, suggesting an absolute difference in serum TT of approximately 10 nmol/L be- tween former AAS users and non-AAS users, which is an even greater numerical difference in plasma TT than we noted in our study, with a signifcant difference of 4.4 nmol/L (P = .009) between the corresponding groups (1). However, Shankara-Narayana et al. did not fnd any statistical difference, which is somewhat surprising. The conficting fndings between these studies might simply be due to differences in the statistical analyses. Shankara-Narayana et al. did not disclose standard de- viations (SDs) of serum TT in their paper nor was serum TT long-transformed and nonparametric statistical tests were not applied, although the values of the corres- ponding SEMs suggest a large data variance and skew distribution of serum TT in all 3 groups, which may ex- plain the non-signifcant difference. In their description of the statistical analyses, the authors stated that they performed covariance analyses on reproductive and car- diac outcomes using potentially confounding covariates, but none of these covariates were mentioned, and the results of these analyses were not disclosed in this paper. Accordingly, the results of the study by Shankara- Narayana et al. may just confrm our previous results of long-lasting impaired endogenous testosterone produc- tion in former AAS users. To continue the development of this novel research feld of potential complications to AAS use, we encourage the authors to disclose the SDs, log- transformations, and the nonparametric statistical tests of Downloaded from https://academic.oup.com/jcem/article/105/8/e3028/5854800 by guest on 29 April 2023