Association Between 5-Alpha Reductase Inhibitor Use and The Risk of Depression: A Meta-Analysis Tuo Deng1, 2, 3 #, Xiaolu Duan1, 2, 3#, Zihao He1, 2, 3, Zhijian Zhao1,2, 3, Guohua Zeng1, 2, 3 * Purpose: To explore the association between 5α-reductase inhibitors (5ARIs) use and risk of depression based on published literature through a meta-analysis. Materials and methods: A comprehensive literature search was conducted by searching Pubmed, Embase, Cochrane Library, CBM, CNKI, and VIP databases up to June, 2019. Summarized risk ratios (RRs) with 95% confidence intervals (CIs) were calculated to evaluate the strength of association between 5ARIs and depression. Subgroup analyses were performed according to population, 5ARI types, degree of depression, and publication date. Registered in PROSPERO under number CRD42018096147. Results: A total of 6 clinical studies with 265672 participants were included in our meta-analysis. The application of 5ARIs could significantly increase the risk of depression based on both pooled unadjusted (95% CI: 1.28-2.78, RR = 1.89, P = .001) and multivariable adjusted RRs (95% CI: 1.01-1.17, RR = 1.09, P = .03). In subgroup anal- yses, dutasteride was associated with depression significantly (95% CI: 1.37-1.70, RR = 1.53, P < .001), while finasteride was not. As to the degree of depression, 5ARIs mainly caused mild depression (95% CI: 1.91-2.33, RR = 2.11, P < .001), instead of moderate or severe depression. Conclusion: We concluded that 5ARIs could potentially increase the risk of depression. Clinicians need to care- fully consider the use of 5ARIs for benign prostatic hyperplasia and androgenic alopecia patients, especially those exhibiting risk factors for depression or those who have a previous history of depression. More studies with larger sample size and comprehensive study design are needed to further verify our outcomes. Keywords: association; 5α-reductase inhibitors; depression; meta-analysis INTRODUCTION Benign prostatic hyperplasia (BPH) is a major con-tributor to lower urinary tract symptoms (LUTS) due to bladder outlet obstruction in elderly men. Both European Association of Urology (EAU) and American Urological Association (AUA) guidelines recommend 5α-reductase inhibitors (5ARIs) as the primary phar- macological treatment for LUTS secondary to BPH. (1,2) 5ARIs could lower the conversion of testosterone to dihydrotestosterone (DHT) through targeting the 5α-re- ductase enzyme family,(3) so 5ARIs are commonly used for BPH and androgenic alopecia. Two equally efficacious 5ARIs are available for clinical use: finas- teride and dutasteride. Finasteride inhibits only type 2 5α-reductase, whereas dutasteride inhibits both types 1 and 2. The most relevant adverse effect of 5ARIs is sexual dysfunction, including reduced libido, erectile dysfunc- tion (ED) and ejaculation disorders.(4-6) Some studies(7-9) 1Department of Urology, Minimally Invasive Surgery center, The First Affiliated Hospital of Guangzhou Medical University, Guangzhou, China. 2Guangdong Key Laboratory of Urology, Guangzhou, China. 3Guangzhou Institute of Urology, Guangzhou, China. *Correspondence: Department of Urology, Minimally Invasive Surgery Center, The First Affiliated Hospital of Guangzhou Medical University. Guangzhou Institute of Urology. Guangdong Key Laboratory of Urology. Address: Kangda Road 1#, Haizhu District, Guangzhou, Guangdong, China, 510230. Telephone: +86-020-34294145, Email: gzgyzgh@vip.tom.com. #: These authors contribute equally to this work. Received December 2019 & Accepted June 2020 also reported a significant increase in depressive symp- toms among patients exposed to Propecia (finasteride 1 mg), which might even exist after discontinuation of the medication. These findings resulted in the addition of depression to the professional labels for Propecia in the United States. A recent population-based matched cohort study(10) indicated that the use of 5ARIs was sig- nificantly associated with increased risk of depression. However, another large population-based study showed that the risk of depression did not increase with 5ARIs. (11) Above all, the association between 5ARIs use and the risk of depression is still controversial. So far, it is difficult to draw a solid conclusion from published studies reporting depression after 5ARIs be- cause most of them are case series;(12-14) few controlled studies were published, and their results remained con- tradictory.(8,10,11) Additionally, depression is not includ- ed in the adverse effects of 5ARIs in either EAU or AUA guidelines due to inadequate levels of evidence. Urology Journal/Vol 18 No. 2/ March-April 2021/ pp. 144-150. [DOI: 10.22037/uj.v16i7.5866] REVIEW (1,2) Consequently, we performed this meta-analysis to clarify the association between 5ARIs use and the risk of depression based on current original controlled stud- ies, hoping to provide some references for clinicians and 5ARIs users. MATERIALS AND METHODS This meta-analysis was conducted in accordance with the meta-analysis of observational studies in epidemi- ology (MOOSE) guidelines. The protocol of this anal- ysis was registered in PROSPERO, and the registration number is CRD42018096147. Search strategy A comprehensive electronic literature search using the Pubmed, Embase, Cochrane Library, CBM, CNKI, and VIP databases was performed to identify controlled studies investigating the association between 5ARIs use and the risk of depression. The date limit of this search was from the inception of these databases to June 2019. Search terms were “‘5-alpha reductase inhibitors’ or ‘5α-reductase inhibitors’ or ‘5-a reductase inhibitors’ or ‘5ARI’ or ‘5-ARI’ or ‘finasteride’ or ‘dutasteride’” in combination with “‘depression’ or ‘depressive’”. References of relevant studies were also checked to identify potential records. No language restrictions ex- isted in this search. Inclusion and exclusion criteria Only controlled clinical studies exploring the associa- tion between 5ARIs use and the risk of depression were included in this meta-analysis. Accordingly, studies without the control group of non-5ARI users were ex- cluded. Meanwhile, studies as abstracts, case reports, conference proceedings, reviews, animal experiments, or repeated publications were also excluded. Relevant studies’ search and screen, quality assessment and data extraction were performed by two reviewers (T.D. and X.D.) independently. Discrepancies were re- solved via open discussion. Study quality assessment and data extraction The level of evidence (LOE) of all eligible studies was assessed by the criteria provided by the Oxford Cen- tre for Evidence-based Medicine.(15) The quality of non-randomized controlled studies included was evalu- ated using the Newcastle-Ottawa scale (NOS).(16) Data from all eligible studies were attentively extracted as follows: study country, population, institution and period, research methodology, type of 5ARIs, diagnos- tic criteria of depression, characteristics of participants, follow-up time, and related outcomes. Authors of rele- vant studies were contacted to obtain incomplete data. Statistics analysis For the included case-control studies, odds ratios (ORs) Association between 5ARIs and depression-Deng et al. Table 1. Baseline characteristics of included studies. Abbreviations: 5ARI 5α-reductase inhibitor, LOE: level of evidence, NOS: Newcastle-Ottawa scale, NM: not mentioned. a Newcastle-Ottawa scale points, one star means one point. Subgroups Number of Included Studies No. Participants Heterogeneity RR (95% CI) I2 P Study population European population 2 183229 99% < .001 2.04 (1.00-4.17) USA population 3 82403 97% < .001 1.85 (1.01-3.38) 5ARI type Finasteride 3 99923 94% < .001 1.34 (0.96-1.87) Dutasteride 1 93790 NA 1.53 (1.37-1.70) Degree of depression Mild depression 2 3631 0% .32 2.11 (1.91-2.33) Moderate and 2 3631 85% .01 4.74(0.14-162.14) severe depression Study publication Before 2015 2 3631 62% .10 3.93 (1.72-8.98) date After 2015 3 262001 98% < .001 1.39 (1.00-1.91) Abbreviations 5ARI: 5α-reductase inhibitor, RR: risk ratio, CI: confidence interval, NA: not applicable. Bold numbers mean the P-value is < .05. Table 2. Results of subgroup analyses Review 145 of 5ARIs use on the risk of depression were extract- ed and converted to risk ratios (RRs) based on the for- mula RR=OR/((1-P"0" )+(P"0" ×OR)) (P0 indicates the incidence of the outcome of interest in the nonex- posed group) according to the Cochrane Handbook. (17) Summarized unadjusted RRs with 95% confidence intervals (CIs) were calculated to assess the strength of association between 5ARIs and the risk of depres- sion. Available adjusted RRs of depression risk and mean differences (MDs) of the Beck Depression Inven- tory-second edition (BDI-II) score in eligible studies were also pooled as references. Chi-square test-based Q- and I2- statistic was used to test the heterogeneity among included studies.(18) The fixed-effect model was used when no significant heterogeneity existed with a P value > 0.10. Otherwise, the random-effect model was applied. All results in this meta-analysis were consid- ered significant with a two-sided P value < 0.05. Sub- group analyses were performed based on the study pop- ulation, type of 5ARI, degree of depression, and study publication date. Sensitivity analyses were conducted by excluding every single eligible study in turn. The publication bias among eligible studies was assessed through the inverted funnel plot visual inspection and the Egger’s test. All statistical analyses were conducted by RevMan (version 5.3; Cochrane Collaboration, Ox- ford, UK) and STATA (version 13.0; StataCorp, Col- lege Station, Texas, USA) software. RESULTS Characteristics and quality assessments of eligible studies After article reviewing and screening carefully, six controlled clinical studies(8,10,11,19,20,21) were included in this meta-analysis (Figure 1). And a total of 265672 participants were involved. Table 1 showed the base- line characteristics of all eligible studies. Among them, 4(8,10,11,20) were cohort studies and 2(19,21) were cross-sec- tional study. As to the study regions, 3 studies(8,11,20) were conducted in the USA, and the other 3 were per- formed in Canada(10), Poland(19), and Italy(21) respective- ly. Regarding to the study publication date, 2 studies(8,19) were published before 2015, while 4 studies(10,11,20,21) were published after 2015. Among the six included studies, the depression rates among 5ARI users ranged from 1.95% to 75.41%, with an average rate of 3.68% (3737/101513); whereas the average rate of depression among non-5ARI users was 2.98% (4886/164119) with a range of 1.37% to 17.6%. LOEs of all 6 included articles were listed in Table 1. Among the 6 clinical studies, 4(8,10,11,20) of them were considered as high quality with a NOS score more than 6 stars. Meta-analysis Unadjusted RRs of 5ARIs use on the risk of depression could be extracted or calculated from 5 clinical stud- ies(8,10,11,19,20). Result of the meta-analysis showed that the use of 5ARIs could significantly increase the risk of depression (95% CI: 1.28-2.78, RR = 1.89, P = .001), with significant heterogeneity among them (I2 = 98%, P < .001) (Figure 2). No obvious publication bias was detected through either inverted funnel plot or Egger’s test (t = 1.44, P = .245). Multivariable adjusted RRs were available in 2 clinical studies(11,20). And the summarized adjusted RR and its 95%CI also indicated that 5ARIs use could signif- icantly increase the risk of depression (95% CI: 1.01- 1.17, RR = 1.09, P = .03) without heterogeneity (I2 = 0%, P = .53) (Figure 3). Mean differences of BDI-II score were also available Figure 1. Flow diagram of study selection. Association between 5ARIs and depression-Deng et al. Vol 18 No 2 March-April 2021 146 in 2 clinical studies(8,21). However, no significant dif- ference was found in BDI-II scores between 5ARI and non-5ARI groups (95% CI: -8.62 to 26.25, MD = 8.81, P = .32) (Figure 4). Subgroup analysis Subgroup analyses of unadjusted RRs of 5ARIs use on the risk of depression were conducted according to the study populations, 5ARI type, degree of depression, and study publication date. Table 2 showed the results of all subgroup analyses. Positive association between 5ARIs use and increased risk of depression was only found in USA population (95% CI: 1.01-3.38, RR = 1.85, P < .05), dutasteride (95% CI: 1.37-1.70, RR = 1.53, P < .001), mild depression (95% CI: 1.91-2.33, RR = 2.11, P < .001), and study published before 2015 (95% CI: 1.72-8.98, RR = 3.93, P = .001) subgroups. Sensitivity analysis Sensitivity analyses of summarized unadjusted RRs of 5ARIs use on the risk of depression were conducted to evaluate the stability and reliability of our results by ex- cluding every single eligible study in turn. As shown in Table 3, no matter which eligible study was excluded, the pooled result remained significant, which means our results are stable and reliable. However, in the sensitiv- ity analyses, we did not find the source of heterogeneity among the five included studies, cause the exclusion of any single study could not reduce the heterogeneity. DISCUSSION In our meta-analysis, we included 6 clinical studies with 265672 participants. We found that the application of 5ARIs may increase the risk of depression. From our subgroup analyses, dutasteride was associated with the existence of depression, while this relationship could not be observed with finasteride. As to the degree of depression, 5 ARIs mainly caused mild depression, instead of moderate or severe depression. Sensitivity analyses indicated that our results are stable and reli- able. According to our unadjusted results from 5 studies, BPH or androgenic alopecia patients having a history of 5ARIs had a significant higher tendency to suffer from depression. In a study conducted by Unger et al, an in- crease in the existence of depression was detected in finasteride users.(20) A large observational study based on the General Practice Research Database also report- ed similar results, showing a probable positive relation- ship between 5ARIs and depression.(22) Several clinical researches discovered the occurrences of depression in their patients receiving 5ARIs and their findings should not be ignored, which need further necessary analysis. Although with different affinities with 5α-reductase, fi- nasteride and dutasteride had similar mechanisms when causing potential risk of depression.(23) First, 5α-reduc- tase participates in the synthesis of some neuroactive steroids.(24) These are not only produced by the central nervous system itself, but also by the gonads and adrenal glands and then transported to the brain.(25) 5ARIs, in- cluding finasteride and dutasteride, can pass the blood- brain barrier and inhibit the activity of 5α-reductase, so the concentration of a variety of neuroactive steroids reduces.(26-28) Second, γ-aminobutyric acid (GABA) is an important inhibitory neurotransmitter. 5α-reductase promotes the formation of allopregnanolone, which is responsible for depression, tension and anxiety, owing to its binding to GABA receptor.(29,30) Therefore, the Table 2. Results of subgroup analyses Excluded study Included participants Heterogeneity RR (95% CI) I2 P Irwig et al, 2012(8) 265542 98% < .001 1.67 (1.13-2.48) Pietrzyk et al, 2015(19) 262091 97% < .001 1.56 (1.12-2.17) Unger et al, 2016(20) 251697 97% < .001 2.27 (1.49-3.46) Welk et al, 2017(10) 85944 99% < .001 2.21 (1.18-4.13) Hagberg et al, 2017(11) 197254 98% < .001 1.91 (1.20-3.03) Abbreviations: RR: risk ratio, CI: confidence interval. Bold numbers mean the P-value is < 0.05. Figure 2. Forest plot of unadjusted RR and 95% CIs of 5ARI use for risk of depression. Association between 5ARIs and depression-Deng et al. Review 147 application of 5ARIs decreases the secretion of allo- pregnanolone and suppresses GABA’s function. Third, in laboratory tests, levels of neuroactive steroids were lower in BPH patients who received finasteride.31 This phenomenon was also observed among patients with depression,32 further proving the potential association between 5ARIs and depression. Furthermore, some ex- periments were conducted on animals. Rodents tended to have anxiolytic and depressive behaviors after being given finasteride. They were also shown to have a lower level of plasma allopregnanolone compared to controls. (33,34) Except for the pharmacological pathways, 5ARIs have other adverse events, such as loss of libido and ED, which may also lead to depression. Taken togeth- er, these clinical findings and the experimental research provided some evidence for the increased risk caused by 5ARIs, confirming our results to some extent. After adjusting for confounding factors, our results still showed that 5ARIs use was significantly associated with an increased risk of depression. BPH and depres- sion could share some kinds of co-risk factors, including old age, smoking, and the presence of chronic disease. (35) Apart from this, BPH itself could induce depression due to LUTS. Pietrzyk and colleagues confirmed the association between LUTS and depression, indicating that the severity of urinary urgency, frequency, and in- creased nocturia really influenced male patients’ quali- ty of life drastically.(19) In our meta-analysis, we adjust- ed our data for confounding factors based on 2 original studies; however, we did not know the details in the factors included in their researches. In addition, pooled mean differences of BDI-II score were also calculated by combining two studies, however, no significant dif- ference was found between the two groups. Since these data from only 2 articles were not adequate for a valid conclusion, these results should be taken into thorough consideration. In our subgroup analyses, dutasteride was shown to in- crease the risk of depression significantly, while finas- teride did not. There are 3 kinds of 5α-reductase in this family: 5α-reductase 1, 2, and 3. Dutasteride inhibits both type 1 and 2; however, finasteride has a specific affinity with only type 2, probably explaining its weak relationship with depression. Nevertheless, because only 3 articles were included in the subgroup finasteride and 1 for dutasteride, this result was not reliable enough to provide any guidance in the application of drugs. Be- sides, our data suggested that 5ARIs could only evoke mild depression, but not moderate or severe. 2 studies were included in this subgroup(8,19) and both used the Beck Depression Inventory to evaluate the severity of depression. Perhaps more clinical studies with larger sample sizes were necessary to ensure the accuracy and validity of our results. Our meta-analysis had several limitations. Firstly, only 6 articles were included for us to reach the pooled re- Figure 3. Forest plot of multivariable adjusted RR and 95% CIs of 5ARI use for risk of depression. Figure 4. Forest plot of mean difference in BDI-II scores between 5ARI and non-5ARI groups. Supplementary material 1: PRISMA checklist. Association between 5ARIs and depression-Deng et al. Vol 18 No 2 March-April 2021 148 sults. The lack of original studies was a great obstacle to conduct a comprehensive meta-analysis. Subsequently, the number of studies in each subgroup was no more than 3, meaning our results cannot affect the current guidelines. Secondly, all studies were observational and no RCTs met our inclusion criteria. Selection bias and recall bias were apparent in retrospective studies, and for data extracted from databases, it was hard to guarantee the accuracy of diagnoses, because research- ers could only judge the existence of disease according to the code recorded. It was also not realistic to achieve detailed information through recalling or scanning da- tabases, such as types of the drugs, severity of diseases and so on. Thirdly, BPH and depression do share sever- al similar risk factors, and LUTS from BPH could also cause depression. Although 2 included studies consid- ered the confounding factors and adjusted for them, we still could not reach a valid conclusion due to the lack of enough evidences. CONCLUSIONS We finally conclude that 5ARIs could potentially in- crease the risk of depression. Based on several large ob- servational studies and FDA’s suggestions, clinicians need to carefully consider the use of 5ARIs for BPH and androgenic alopecia patients, especially those at risk for depression. More studies with a larger sample size and comprehensive study design are needed neces- sary to further verify our outcomes. ACKNOWLEDGMENTS This study was financed by grants from National Natu- ral Science Foundation of China (No. 81802821), Nat- ural Science Foundation of Guangdong Province (No. 2017A030310547), and China Postdoctoral Science Foundation (No. 2018T110859 and No. 2017M612636). CONFLICT OF INTEREST The authors declare that they have no conflicts of in terest. REFERENCES 1. Gravas S, Cornu JN, Drake MJ, et al. European Association of Urology guidelines on treatment of non-neurogenic male LUTS 2017. Available at: http://uroweb.org/ guideline/treatment-of-non-neurogenic-male- luts/. Accessed January 20, 2018. 2. McVary KT, Roehrborn CG, Avins AL, et al. American Urological Association guidelines statement on management of Benign Prostatic Hyperplasia (BPH) 2014. Available at: http:// www.auanet.org/guidelines/benign-prostatic- hyperplasia-(2010-reviewed-and-validity- confirmed-2014). Accessed January 20, 2018. 3. Traish AM. 5α-reductases in human physiology: an unfolding story. Endocr Pract. 2012; 18: 965-75. 4. McConnell JD, Roehrborn CG, Bautista OM, et al. The long-term effect of doxazosin, finasteride, and combination therapy on the clinical progression of benign prostatic hyperplasia. N Engl J Med. 2003; 349: 2387- 98. 5. Naslund MJ, Miner M. A review of the clinical efficacy and safety of 5alpha-reductase inhibitors for the enlarged prostate. Clin Ther. 2007; 29: 17. 6. Roehrborn CG, Siami P, Barkin J, et al. The effects of combination therapy with dutasteride and tamsulosin on clinical outcomes in men with symptomatic benign prostatic hyperplasia: 4-year results from the CombAT study. Eur Urol. 2010; 57: 123. 7. Rahimi-Ardabili B, Pourandarjani R, Habibollahi P, Mualeki A. Finasteride induced depression: a prospective study. BMC Clin Pharmacol. 2006; 6: 7. 8. Irwig MS. Depressive symptoms and suicidal thoughts among former users of finasteride with persistent side effects. J Clin Psychiatry. 2012; 73: 1220-3. 9. Traish AM, Mulgaonkar A, Giordano N. The dark side of 5areductase inhibitors’ therapy: sexual dysfunction, high Gleason grade prostate cancer and depression. Korean J Urol. 2014; 55: 367-79. 10. Welk B, McArthur E, Ordon M, Anderson KK, Hayward J, Dixon S. Association of Suicidality and Depression With 5α-Reductase Inhibitors. JAMA Intern Med. 2017; 177: 683-91. 11. Hagberg KW, Divan HA, Nickel JC, Jick SS. Risk of Incident Antidepressant- Treated Depression Associated with Use of 5α-Reductase Inhibitors Compared with Use of α-Blockers in Men with Benign Prostatic Hyperplasia: A Population-Based Study Using the Clinical Practice Research Datalink. Pharmacotherapy. 2017; 37: 517-27. 12. Altomare G, Capella GL. Depression circumstantially related to the administration of finasteride for androgenetic alopecia. J Dermatol. 2002; 29: 665-9. 13. Rahimi-Ardabili B, Pourandarjani R, Habibollahi P, Mualeki A. Finasteride induced depression: a prospective study. BMC Clin Pharmacol. 2006; 6: 7. 14. Melcangi RC, Santi D, Spezzano R, et al. Neuroactive steroid levels and psychiatric and andrological features in post-finasteride patients. J Steroid Biochem Mol Biol. 2017; 171: 229-35. 15. Kurahashi N, Iwasaki M, Sasazuki S, Otani T, Inoue M, Tsugane S. Soy product and isoflavone consumption in relation to prostate cancer in Japanese men. Cancer Epidemiol Biomarkers Prev. 2007; 16: 538-45. 16. Stang A. Critical evaluation of the Newcastle- Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol. 2010; 25: 603-5. 17. Higgins J, Thomas J (editors). Cochrane Handbook for Systematic Reviews of Interventions. Version 6. The Cochrane Collaboration. Available at: https://training. cochrane.org/handbook. Accessed May 7, 2020. 18. Higgins JP Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta- analyses. BMJ. 2003; 327: 557-60. 19. Pietrzyk B, Olszanecka-Glinianowicz M, Association between 5ARIs and depression-Deng et al. Review 149 Vol 18 No 2 March-April 2021 138 Owczarek A, et al. Depressive symptoms in patients diagnosed with benign prostatic hyperplasia. Int Urol Nephrol. 2015; 47: 431- 40. 20. Unger JM, Till C, Thompson IM Jr, et al. Long- term Consequences of Finasteride vs Placebo in the Prostate Cancer Prevention Trial. J Natl Cancer Inst. 2016; 108. pii: djw168. 21. Catalano A, Martino G, Bellone F, et al. Neuropsychological Assessment in Elderly Men with Benign Prostatic Hyperplasia Treated with Dutasteride. Clin Drug Investig. 2019; 39: 97-102. 22. Clifford GM, Farmer RDT. Drug or symptom- induced depression in men treated with alpha1- blockers for benign prostatic hyperplasia? A nested case-control study. Pharmacother Drug Saf. 2002; 11: 55-61. 23. Traish AM, Melcangi RC, Bortolato M, Garcia-Segura LM, Zitzmann M. Adverse effects of 5α-reductase inhibitors: What do we know, don’t know, and need to know? Rev Endocr Metab Disord. 2015; 16:177-98. 24. Celec P, Ostatníková D, Hodosy J. On the effects of testosterone on brain behavioral functions. Front Neurosci. 2015; 9: 1-17. 25. Paul SM, Purdy RH. Neuroactive steroids. FASEB J. 1992; 6: 2311-22. 26. Pinacho-Garcia LM, Valdez RA, Navarrete A, Cabeza M, Segovia J, Romano MC. The effect of finasteride and dutasteride on the synthesis of neurosteroids by glioblastoma cells. Steroids. 2020; 155:108556. 27. Litim N, Morissette M, Caruso D, Melcangi RC, Di Paolo T. Effect of the 5α-reductase enzyme inhibitor dutasteride in the brain of intact and parkinsonian mice. J Steroid Biochem Mol Biol. 2017; 174:242-56. 28. Clark RV, Hermann DJ, Cunningham GR, Wilson TH, Morrill BB, Hobbs S. Marked suppression of dihydrotestosterone in men with benign prostatic hyperplasia by dutasteride, a dual 5alpha-reductase inhibitor. J Clin Endocrinol Metab. 2004; 89: 2179-84. 29. Majewska MD, Harrison NL, Schwartz RD, Barker JL, Paul SM. Steroid hormone metabolites are barbiturate-like modulators of the GABA receptor. Science. 1986; 232: 1004-7. 30. Melcangi RC, Caruso D, Abbiati F, et al. Neuroactive steroid levels are modified in cerebrospinal fluid and plasma of post- finasteride patients showing persistent sexual side effects and anxious/depressive symptomatology. J Sex Med. 2013; 10: 2598- 603. 31. Uzunova V, Sheline Y, Davis JM, et al. Increase in the cerebrospinal fluid content of neurosteroids in patients with unipolar major depression who are receiving fluoxetine or fluvoxamine. Proc Natl Acad Sci USA. 1998; 95: 3239-44. 32. Romeo E, Ströhle A, Spalletta G, et al. Effects of antidepressant treatment on neuroactive steroids in major depression. Am J Psychiatry. 1998; 155: 910-13. 33. Rhodes ME, Frye CA. Inhibiting progesterone metabolism in the hippocampus of rats in behavioral estrus decreases anxiolytic behaviors and enhances exploratory and antinociceptive behaviors. Cogn Affect Behav Neurosci. 2001; 1: 287-96. 34. Frye CA, Walf AA. Changes in progesterone metabolites in the hippocampus can modulate open field and forced swim test behavior of proestrous rats. Horm Behav. 2002; 41: 306- 15. 35. Djernes JK. Prevalence and predictors of depression in populations of elderly: a review. Acta Psychiatr Scand. 2006; 113: 372-87. Association between 5ARIs and depression-Deng et al. Vol 18 No 2 March-April 2021 150