Citation
Clomiphene citrate versus testosterone replacement therapy in male hypogonadism: a systematic review of literature and meta-analysis
Constantinou BT, Benedicto BC, Galligani L, et al.
Eur J Clin Pharmacol. 2026;82(8). Published July 11, 2026.
PMID: 42435198 DOI: 10.1007/s00228-026-04134-3
Study at a Glance
- Design
- Systematic review and meta-analysis
- Population
- Adult men with testosterone deficiency, excluding primary hypogonadism. The authors report 1,512 participants or treatment-arm entries, comprising 764 in clomiphene-based groups and 748 receiving TRT. Shared comparator groups mean these totals should not be assumed to represent unique individuals. No reliable pooled age was reported.
Question Addressed
In adult men with testosterone deficiency, how do clomiphene-based treatments compare with testosterone replacement therapy (TRT) for raising serum testosterone and improving libido? The review also examined gonadotropin concentrations, but its inclusion of enclomiphene trials means that the findings do not apply exclusively to clomiphene citrate (Constantinou et al. 2026).
Plain Language Summary
This review found no statistically clear difference in the increase in blood testosterone between clomiphene-based treatment and testosterone gel. However, results varied considerably across studies, and the certainty of the testosterone evidence was very low. The findings therefore do not establish that the treatments are equally effective. The review also combined clomiphene with enclomiphene, a related but different medicine [2].
Testosterone injections produced a larger testosterone increase in 1 small comparison. A separate analysis suggested better sexual desire scores with TRT, but this finding came from a small body of observational evidence with concerns about repeated counting of participants. It cannot establish that TRT provides better overall sexual function or quality of life [2,3,6].
Fertility remains an important consideration. An included enclomiphene trial measured sperm concentrations and found that they were maintained with enclomiphene but reduced with testosterone gel. This is evidence about sperm production, rather than proof of improved pregnancy or live-birth rates. Current European guidance describes these alternative medicines as off-label treatments with limited supporting evidence. Treatment decisions require attention to symptoms, the cause of low testosterone, fertility goals, and monitoring [1,4,5].
Methods
The authors used independent screening and data extraction, with disagreements resolved by an additional reviewer. Eligibility generally required baseline total testosterone of 300 ng/dL or less, although the study table includes an exception using 350 ng/dL. Randomized trials and observational cohorts were combined using mean differences with 95% confidence intervals (CIs) (Constantinou et al. 2026).
The analyses included:
- TRT formulation subgroups: Injections, gel, and mixed injection/gel treatment.
- Sensitivity analyses: Repeating the analysis after removing individual comparisons, together with assessment of which studies contributed most to heterogeneity.
- Consistency and publication bias: I² statistics, prediction intervals, funnel plots, and an Egger test for the testosterone outcome.
- Gonadotropins: Descriptive reporting of FSH and LH because heterogeneity and unstable estimates made pooled conclusions unreliable. (Constantinou et al. 2026)
Effect estimates below follow the article's convention of clomiphene-based treatment minus TRT. Positive testosterone differences favor a larger increase with clomiphene-based treatment; negative libido differences indicate lower post-treatment libido scores with clomiphene-based treatment. These are published aggregate comparisons, rather than pooled estimates adjusted for patient-level confounders (Constantinou et al. 2026).
Key Findings
- Overall testosterone: Mean difference 6.64 ng/dL (95% CI −35.35 to 48.63; p = 0.7566 in Figure 3), with substantial heterogeneity (I² = 80.6%). The prediction interval was −135.91 to 149.19 ng/dL, indicating considerable uncertainty about the effect in another comparable study.
- Testosterone gel: Mean difference 24.29 ng/dL (95% CI −18.91 to 67.48; I² = 76.1%), across 8 comparisons and 1,012 reported arm entries. No statistically significant difference was detected.
- Injectable testosterone: The single comparison, comprising 31 men per group, favored injections: mean difference −290.50 ng/dL (95% CI −518.78 to −62.22). This estimate should not be generalized to all injectable preparations or dosing regimens.
- Mixed TRT formulations: Mean difference −13.40 ng/dL (95% CI −113.78 to 86.99; I² = 93.3%). The test for differences between formulation subgroups was significant (p = 0.0262), but the sparse injection evidence limits interpretation.
- Libido: Mean difference −0.54 points on the reported 1–5 libido scale (95% CI −0.87 to −0.21; abstract p = 0.009; I² = 18.8%), favoring TRT. Figure 4 contains 3 comparisons from 2 publications and repeats the 31-person clomiphene group. The reported 199 arm entries therefore do not represent 199 unique participants. The prediction interval, −1.27 to 0.19, crossed no difference.
- FSH and LH: Heterogeneity was extreme, at I² = 90.6% and 95.6%, respectively. The authors declined to present definitive pooled estimates. Differences in gonadotropins cannot establish differences in pregnancy outcomes.
- Certainty and safety: GRADE certainty was very low for testosterone, FSH, and LH, and low for libido. Comparative hematologic and prostate-related safety outcomes were not quantitatively synthesized. (Constantinou et al. 2026; Ramasamy et al. 2014)
Limitations and Strengths Affecting Interpretation
Strengths
- Broad search: Multiple databases, grey literature, and reference lists were searched without language restrictions.
- Structured appraisal: The authors assessed randomized and nonrandomized studies with separate risk-of-bias tools and applied GRADE.
- Attention to variability: Formulation subgroups, sensitivity analyses, and prediction intervals exposed differences that an overall average could conceal.
- Appropriate caution: The conclusion explicitly states that failure to detect a difference does not establish noninferiority. (Constantinou et al. 2026)
Limitations
- Bias and confounding: All 5 randomized studies had some risk-of-bias concerns; all 6 nonrandomized studies were judged to have serious/high risk of bias. Age, treatment selection, fertility intentions, and follow-up could influence comparisons.
- Different interventions: Combining racemic clomiphene and enclomiphene introduces indirectness. No clomiphene-only sensitivity analysis was provided.
- Clinical heterogeneity: Diagnostic criteria, TRT preparations, doses, assessment timing, and populations varied. The review does not establish effects specifically in older men with rigorously defined late-onset hypogonadism.
- Dependence between comparisons: Reuse of a shared clomiphene group in separate TRT comparisons raises concerns about participant counts, weighting, and precision. The report does not explain an adjustment for this dependence.
- Limited patient-important outcomes: The libido result does not establish overall sexual-function benefit. The review provides no pooled pregnancy, live-birth, quality-of-life, or long-term comparative safety estimate.
- Reporting inconsistencies: The testosterone CI differs in the GRADE narrative, and the description of absent semen outcomes conflicts with an included trial. These discrepancies reduce confidence in data handling and interpretation. (Constantinou et al. 2026; Kim et al. 2016; Ramasamy et al. 2014)
Read this review as a preliminary comparison of biochemical responses with a limited signal for libido. It does not demonstrate therapeutic equivalence, establish comparative long-term safety, or prove a fertility benefit (Constantinou et al. 2026).
Authors' Conclusions
The authors concluded that serum testosterone responses did not differ significantly between clomiphene-based treatment and testosterone gel, while libido outcomes favored TRT. They emphasized that the analysis did not test noninferiority and that heterogeneity, risk of bias, and limited studies prevent firm conclusions. They called for better randomized trials measuring fertility, sexual function, and quality of life (Constantinou et al. 2026).
Commentary
The most useful contribution is the distinction between testosterone formulations. The overall estimate conceals a larger increase with injections in 1 small comparison and no clear difference with gel. The underlying retrospective comparison also found broadly similar overall treatment satisfaction despite differences in achieved testosterone concentrations. Thus, the biochemical findings cannot establish which treatment provides the greatest patient-perceived benefit (Constantinou et al. 2026; Ramasamy et al. 2014).
The libido signal is consistent with an included prospective cohort in which testosterone treatment improved several symptom domains, while the clomiphene group's libido score declined from 3.75 to 3.2 (p = 0.04). However, the meta-analysis largely re-expresses these small observational datasets, and its 3 libido comparisons arise from only 2 publications. The repeated clomiphene group creates a concern about statistical independence that needs clarification before the pooled precision is accepted. A claim of established TRT superiority for sexual function would exceed this evidence (Dadhich et al. 2017; Constantinou et al. 2026; Ramasamy et al. 2014).
Fertility requires a separate interpretation from serum testosterone. In 2 phase III trials reported together, enclomiphene maintained sperm concentrations over 16 weeks, whereas testosterone gel reduced spermatogenesis. The review's statement that no included study assessed semen parameters conflicts with that publication. Nevertheless, sperm concentration is a surrogate outcome, and this review supplies no pooled evidence of improved conception or live birth. Furthermore, enclomiphene results cannot automatically be attributed to racemic clomiphene (Kim et al. 2016; Constantinou et al. 2026).
Guidelines help define the population in which these findings might be considered. The Endocrine Society guideline requires compatible symptoms and consistently low testosterone, confirmed with repeat morning testing, and advises against testosterone therapy when near-term fertility is planned. Current EAU guidance describes anti-estrogen treatment as off-label, with poor supporting evidence and a requirement for an intact hypothalamic-pituitary-gonadal axis. It also prioritizes management of reversible contributors to functional hypogonadism. The review therefore provides limited comparative information within a broader diagnostic and treatment assessment, rather than support for treating an isolated low result (Bhasin et al. 2018; European Association of Urology 2026; Constantinou et al. 2026).
Comparative safety remains unresolved. An included retrospective study reported polycythemia in 11.2% of testosterone-treated men versus 1.7% of clomiphene-treated men (p = 0.0003); the difference remained significant after adjustment for age, site, smoking history, and pretreatment hematocrit. The testosterone group was older and treated longer, however, and the review did not pool safety outcomes. This association supports attention to hematocrit but cannot establish an overall safety advantage or exclude uncommon adverse effects of clomiphene (Wheeler et al. 2017; Constantinou et al. 2026).
Clinically, the findings support assessing biochemical response and symptom response separately, with fertility goals and the cause of hypogonadism influencing treatment choice. The absence of a significant testosterone difference versus gel should not be presented as equivalence. The libido result is exploratory, and treatment-specific fertility or safety claims require evidence beyond this analysis. The main research need is a sufficiently powered randomized comparison of clearly specified clomiphene and TRT regimens, with independent treatment groups and standardized symptom, reproductive, and safety outcomes (Bhasin et al. 2018; European Association of Urology 2026; Constantinou et al. 2026).
References
1. Bhasin S, Brito JP, Cunningham GR, Hayes FJ, Hodis HN, Matsumoto AM, Snyder PJ, Swerdloff RS, Wu FC, Yialamas MA. Testosterone therapy in men with hypogonadism: an Endocrine Society clinical practice guideline. Journal of Clinical Endocrinology & Metabolism. 2018;103(5):1715–1744. DOI: PubMed: PMID
Supporting quote: “We recommend confirming the diagnosis by repeating the measurement of morning fasting total T concentrations.”
2. Constantinou BT, Benedicto BC, Galligani L, Schiavon BN, Mendes CMO, Petri MMG, Uzueli B, Maia RS, da Sardenberg RAS, Fraietta R, Shiomi da Cruz JA. Clomiphene citrate versus testosterone replacement therapy in male hypogonadism: a systematic review of literature and meta-analysis. European Journal of Clinical Pharmacology. 2026;82(8):204. DOI: PubMed: PMID
Supporting quote: “this should not be interpreted as evidence of non-inferiority, as no formal non-inferiority analysis was conducted.”
3. Dadhich P, Ramasamy R, Scovell J, Wilken N, Lipshultz L. Testosterone versus clomiphene citrate in managing symptoms of hypogonadism in men. Indian Journal of Urology. 2017;33(3):236–240. DOI: PubMed: PMID
Supporting quote: “qADAM subscore for libido was lower following treatment (3.75-3.2, P = 0.04)”
4. European Association of Urology. Male hypogonadism. In: EAU Guidelines on Sexual and Reproductive Health. Current online edition, accessed November 17, 2026. . DOI and PubMed record: not assigned to this online chapter.
Supporting quote: “The evidence published so far is poor; all these products are off-label treatments”
5. Kim ED, McCullough A, Kaminetsky J. Oral enclomiphene citrate raises testosterone and preserves sperm counts in obese hypogonadal men, unlike topical testosterone: restoration instead of replacement. BJU International. 2016;117(4):677–685. DOI: PubMed: PMID
Supporting quote: “Enclomiphene citrate treatment maintained sperm concentrations in the normal range.”
6. Ramasamy R, Scovell JM, Kovac JR, Lipshultz LI. Testosterone supplementation versus clomiphene citrate for hypogonadism: an age matched comparison of satisfaction and efficacy. Journal of Urology. 2014;192(3):875–879. DOI: PubMed: PMID
Supporting quote: “Men receiving testosterone injections reported greater libido than men on clomiphene citrate (4 vs 3, p = 0.04)”
7. Wheeler KM, Smith RP, Kumar RA, Setia S, Costabile RA, Kavoussi PK. A comparison of secondary polycythemia in hypogonadal men treated with clomiphene citrate versus testosterone replacement: a multi-institutional study. Journal of Urology. 2017;197(4):1127–1131. DOI: PubMed: PMID
Supporting quote: “This significance remained on logistic regression after correcting for age, site, smoking history and pretreatment hematocrit.”
Study at a Glance
- Design. Systematic review and random-effects meta-analysis, reported as 11 studies comprising 5 randomized controlled trials and 6 nonrandomized comparative studies.
- Sources searched. PubMed/MEDLINE, Embase, Scopus, Cochrane CENTRAL, and Google Scholar, supplemented by reference-list searches, without language restrictions.
- Period covered. Database inception through August 2025; published online on July 11, 2026.
- Interventions. Oral clomiphene or enclomiphene compared with testosterone gel, injections, or mixed formulations. Doses and diagnostic criteria differed between studies.
- Duration. Treatment and follow-up periods varied and were incompletely summarized, preventing a dependable common duration from being assigned.
- Outcomes. Between-group differences in testosterone change; libido questionnaire scores; follicle-stimulating hormone (FSH) and luteinizing hormone (LH).
- Evidence assessment. RoB 2 for randomized trials, ROBINS-I for nonrandomized studies, and GRADE for certainty. PROSPERO registration: CRD420251252199. (Constantinou et al. 2026)