Dear Editor,
Based on previous case reports, a 2023 report by the UK Commission on Human Medicines recommended increased warnings on the potential risk for sexual dysfunction from isotretinoin use.1,2 This report also noted that available data is limited, conflicting, and from studies that are poorly designed to assess causality.2 Given the importance of isotretinoin as a treatment for severe acne and the potential relevance of sexual dysfunction to populations most at risk of acne, this is an important clinical question.3 To complement available data from prior case series, we conducted a cohort study evaluating whether isotretinoin exposure is associated with sexual dysfunction.
Using the TriNetX US Collaborative Network (56 healthcare organizations) between August 16th, 2003 and August 16th, 2023, we identified male acne patients, ages 14-40 years old using International Classification of Diseases (ICD), 10th revision codes (L70.0, L70.1, L70.5, L70.8, L70.9) from 2003-2023.3 We created three cohorts: 1. Acne managed with isotretinoin (RxNorm:6064): patients managed with isotretinoin with no prior tetracycline-class antibiotic exposure; 2. Acne managed with tetracycline-class antibiotic (ATC:J01A): patients managed with tetracycline-class antibiotic, with no isotretinoin exposure (to control for baseline acne severity); 3. Acne managed without systemic medications: patients with an encounter for acne and no exposure to isotretinoin or tetracycline-class antibiotics. The index date was defined as the first prescription for isotretinoin (cohort 1), tetracycline-class antibiotic (cohort 2), or first diagnosis of acne (cohort 3). The isotretinoin cohort was 1:1 propensity score matched using a greedy nearest-neighbor matching algorithm based on potential confounding variables such as hypertension, diabetes, alcohol related disorders, depression, anxiety, and certain medications (Table 1).4, 5 Clinical outcomes of interest within one year of the index event included erectile dysfunction (ICD-10-CM: N52.0, N52.1, N52.2, N52.8, or N52.9), sexual dysfunction (ICD-10-CM: F52, N53, or R37), decreased libido (ICD-10-CM: R68.82), and phosphodiesterase-5 inhibitor (PDE5i) prescription (RxNorm: 136411, 358263, or 306674). In all cohorts, we excluded patients with pulmonary heart diseases (ICD-10-CM: I27) to minimize confounding PDE5i prescriptions for such conditions, as well as patients with clinical outcomes of interest prior to the defined time window for all analyses performed.4 Adjusted risk ratios (aRR) with 95% confidence intervals (CI) were tabulated by the proprietary TriNetX analytic feature.
Table 1.
Baseline demographic characteristics of acne patients treated with isotretinoin, tetracycline-class antibiotics, and without systemic medications after propensity score matching.
| Characteristic | Isotretinoin (n = 13600) |
Tetracycline-class Antibiotics (n = 13600) |
p-value | Isotretinoin (n = 13598) |
Acne Managed Without Systemic Medications (n = 13598) |
p-value |
|---|---|---|---|---|---|---|
| Age at the initiation of the drug treatment, mean (SD), y | 19.05 (5.07) | 19.11 (5.13) | 0.34 | 19.04 (5.05) | 18.99 (4.98) | 0.35 |
| Ethnicity | ||||||
| Not Hispanic or Latino | 9149 (67.27%) | 9140 (67.21%) | 0.91 | 9146 (67.26%) | 9146 (67.26%) | >0.99 |
| Hispanic or Latino | 1363 (10.02%) | 1369 (10.07%) | 0.90 | 1364 (10.03%) | 1370 (10.08%) | 0.90 |
| Race | ||||||
| White | 9877 (72.63%) | 9907 (72.85%) | 0.68 | 9875 (72.62%) | 9891 (72.74%) | 0.83 |
| Asian | 743 (5.46%) | 724 (5.32%) | 0.61 | 742 (5.46%) | 746 (5.49%) | 0.92 |
| Black or African American | 407 (2.99%) | 426 (3.13%) | 0.50 | 407 (2.99%) | 407 (2.99%) | >0.99 |
| American Indian or Alaska Native | 91 (0.67%) | 89 (0.65%) | 0.88 | 91 (0.67%) | 102 (0.75%) | 0.43 |
| Native Hawaiian or Other Pacific Islander | 30 (0.22%) | 22 (0.16%) | 0.27 | 29 (0.21%) | 26 (0.19%) | 0.69 |
| Comorbidities | ||||||
| Overweight and obesity | 384 (2.82%) | 359 (2.64%) | 0.35 | 384 (2.82%) | 359 (2.64%) | 0.35 |
| Generalized anxiety disorder | 270 (1.99%) | 252 (1.85%) | 0.43 | 269 (1.98%) | 237 (1.74%) | 0.15 |
| Disorders of lipoprotein metabolism and other lipidemias | 247 (1.82%) | 254 (1.87%) | 0.75 | 246 (1.81%) | 224 (1.65%) | 0.31 |
| Essential (primary) hypertension | 190 (1.40%) | 199 (1.46%) | 0.65 | 189 (1.39%) | 163 (1.20%) | 0.16 |
| Major depressive disorder, recurrent | 152 (1.12%) | 108 (0.79%) | <0.05 | 152 (1.12%) | 116 (0.85%) | 0.03 |
| Alcohol related disorders | 86 (0.63%) | 80 (0.59%) | 0.64 | 86 (0.63%) | 85 (0.63%) | 0.94 |
| Persistent mood [affective] disorders | 74 (0.54%) | 53 (0.39%) | 0.06 | 74 (0.54%) | 60 (0.44%) | 0.23 |
| Type 2 diabetes mellitus | 50 (0.37%) | 41 (0.30%) | 0.34 | 50 (0.37%) | 47 (0.35%) | 0.76 |
| Testicular hypofunction | 34 (0.25%) | 22 (0.16%) | 0.11 | 34 (0.25%) | 21 (0.15%) | 0.08 |
| Ischemic heart diseases | 10 (0.07%) | 10 (0.07%) | >0.99 | 10 (0.07%) | 10 (0.07%) | >0.99 |
| Metabolic syndrome | 10 (0.07%) | 14 (0.10%) | 0.41 | 10 (0.07%) | 10 (0.07%) | >0.99 |
| Medications | ||||||
| finasteride | 49 (0.36%) | 37 (0.27%) | 0.19 | 49 (0.36%) | 44 (0.32%) | 0.60 |
| dutasteride | 10 (0.07%) | 10 (0.07%) | >0.99 | 10 (0.07%) | 10 (0.07%) | 1.00 |
| propranolol | 56 (0.41%) | 57 (0.42%) | 0.92 | 55 (0.40%) | 53 (0.39%) | 0.85 |
| metoprolol | 31 (0.23%) | 26 (0.19%) | 0.51 | 30 (0.22%) | 22 (0.16%) | 0.27 |
| atenolol | 19 (0.14%) | 15 (0.11%) | 0.49 | 19 (0.14%) | 14 (0.10%) | 0.38 |
| lisinopril | 59 (0.43%) | 62 (0.46%) | 0.78 | 58 (0.43%) | 51 (0.38%) | 0.50 |
| hydrochlorothiazide | 27 (0.20%) | 25 (0.18%) | 0.78 | 27 (0.20%) | 21 (0.15%) | 0.39 |
| trazodone | 122 (0.90%) | 92 (0.68%) | 0.04 | 122 (0.90%) | 93 (0.68%) | 0.05 |
| venlafaxine | 46 (0.34%) | 41 (0.30%) | 0.59 | 44 (0.32%) | 40 (0.29%) | 0.66 |
| desvenlafaxine | 10 (0.07%) | 10 (0.07%) | >0.99 | 10 (0.07%) | 10 (0.07%) | 1.00 |
| duloxetine | 29 (0.21%) | 26 (0.19%) | 0.69 | 29 (0.21%) | 25 (0.18%) | 0.59 |
| milnacipran | 0 (0.00%) | 10 (0.07%) | <0.05 | 0 (0.00%) | 10 (0.07%) | <0.05 |
| levomilnacipran | 0 (0.00%) | 10 (0.07%) | <0.05 | 0 (0.00%) | 0 (0.00%) | N/A |
| esketamine | 15 (0.11%) | 11 (0.08%) | 0.43 | 15 (0.11%) | 20 (0.15%) | 0.40 |
| sertraline | 244 (1.79%) | 230 (1.69%) | 0.52 | 244 (1.79%) | 225 (1.66%) | 0.38 |
| paroxetine | 29 (0.21%) | 26 (0.19%) | 0.69 | 29 (0.21%) | 30 (0.22%) | 0.90 |
| citalopram | 85 (0.63%) | 92 (0.68%) | 0.60 | 85 (0.63%) | 71 (0.52%) | 0.26 |
| fluoxetine | 191 (1.40%) | 172 (1.27%) | 0.32 | 191 (1.41%) | 180 (1.32%) | 0.57 |
| escitalopram | 183 (1.35%) | 162 (1.19%) | 0.26 | 184 (1.35%) | 145 (1.07%) | 0.03 |
| mirtazapine | 38 (0.28%) | 34 (0.25%) | 0.64 | 38 (0.28%) | 37 (0.27%) | 0.91 |
| bupropion | 130 (0.96%) | 115 (0.85%) | 0.34 | 131 (0.96%) | 106 (0.78%) | 0.10 |
| vilazodone | 10 (0.07%) | 0 (0.00%) | <0.05 | 10 (0.07%) | 10 (0.07%) | 1.00 |
| vortioxetine | 10 (0.07%) | 10 (0.07%) | >0.99 | 10 (0.07%) | 10 (0.07%) | 1.00 |
| fluvoxamine | 10 (0.07%) | 11 (0.08%) | 0.83 | 10 (0.07%) | 11 (0.08%) | 0.83 |
| viloxazine | 0 (0.00%) | 0 (0.00%) | N/A | 0 (0.00%) | 0 (0.00%) | N/A |
| nefazodone | 0 (0.00%) | 0 (0.00%) | N/A | 0 (0.00%) | 0 (0.00%) | N/A |
| maprotiline | 0 (0.00%) | 0 (0.00%) | N/A | 0 (0.00%) | 0 (0.00%) | N/A |
Among 13,600 patients treated with isotretinoin who were matched with 13,600 patients treated with oral tetracycline-class antibiotics (Table 1), there were no significant differences in risk of erectile dysfunction (aRR [95% CI]) = (1.0 [0.55-1.8]), sexual dysfunction (0.74 [0.39-1.38]), decreased libido (1.0 [0.42-2.4]), or PDE5iP use (1.7 [0.88-3.2]) (Table 2). Similar findings were observed comparing those treated with isotretinoin and those with acne managed without systemic medications, with no significant differences in risk of erectile dysfunction (aRR [95% CI]) = (0.82 [0.46-1.4]), sexual dysfunction (1.3 [0.64-2.7]), decreased libido (1.0 [0.42-2.4]), or PDE5i use (1.1 [0.62-1.9]) (Table 2).
Table 2.
Risk ratios for sexual health outcomes among male acne patients treated with isotretinoin, tetracycline-class antibiotics, and without systemic medications.
| Isotretinoin | Tetracyclines | Isotretinoin vs. Tetracyclines | |||
|---|---|---|---|---|---|
| Outcome | Risk, % | Risk, % | Risk difference, % (95% CI) | Risk ratio (95% CI) | p-value |
| Erectile dysfunction | 0.16 | 0.16 | 0 (−0.096, 0.096) | 1.0 (0.55, 1.8) | >.99 |
| Sexual dysfunction | 0.13 | 0.17 | −0.045 (−0.14, 0.047) | 0.74 (0.39, 1.38) | 0.34 |
| Decreased libido | 0.070 | 0.070 | 0 (−0.065, 0.064) | 1.0 (0.42, 2.4) | >.99 |
| PDE5i Prescription (Sildenafil, Tadalafil, Vardenafil | 0.18 | 0.11 | 0.074 (−0.018, 0.17) | 1.7 (0.88, 3.2) | 0.11 |
| Isotretinoin | No Systemic Medications | Isotretinoin vs. No Systemic Medications | |||
| Outcome | Risk, % | Risk, % | Risk difference, % (95% CI) | Risk ratio (95% CI) | p-value |
| Erectile dysfunction | 0.16 | 0.20 | −0.037 (−0.14, 0.064) | 0.82 (0.46, 1.4) | 0.47 |
| Sexual dysfunction | 0.13 | 0.096 | 0.029 (−0.050, 0.11) | 1.3 (0.64, 2.7) | 0.47 |
| Decreased libido | 0.074 | 0.074 | 0 (−0.064, 0.064) | 1.0 (0.42, 2.4) | >.99 |
| PDE5i Prescription (Sildenafil, Tadalafil, Vardenafil | 0.18 | 0.17 | 0.015 (−0.085, 0.12) | 1.1 (0.62, 1.9) | 0.77 |
This study suggests that isotretinoin exposure among male patients with acne does not appear to be associated with an increased risk of erectile dysfunction, sexual dysfunction, decreased libido, or PDE5i use. A strength of the study is the matched cohort design, including efforts to control for potential confounders such as comorbid depression and anxiety.1, 4 Limitations include the potential for residual confounding, ascertainment bias as not all patients will seek care for sexual dysfunction, and sample size limitations resulting in imprecision in our estimates. Due to challenges in outcome definitions, the present study only evaluates sexual health outcomes among male patients with acne and future studies are needed to evaluate female sexual health outcomes. Although further studies are warranted to replicate these findings and evaluate whether they generalize to other populations, the present study provides reassuring data for clinicians and patients.
Acknowledgements:
John S. Barbieri (JSB) created and supervised the study. Christopher J. Thang (CJT) and David Garate (DG) conducted the study. JSB and George Golovko advised in study design, implementation, and statistical analysis. CJT and DG created both tables. JSB assisted in data interpretation and table design. CJT, DG, GG, and JSB drafted the manuscript. JSB is corresponding author. JSB is supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the National Institutes of Health under award number 1K23AR078930. No funding sources supported this work.
Conflicts of Interest:
John S. Barbieri has received consulting fees from Dexcel Pharma for work unrelated to the current submission. The authors have no other conflicts to declare.
References
- 1.Healy D, Le Noury J, Mangin D. Enduring sexual dysfunction after treatment with antidepressants, 5α-reductase inhibitors and isotretinoin: 300 cases. Int J Risk Saf Med. 2018;29(3-4):125–134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Commission on Human Medicines. Report of the commission on human medicines isotretinoin expert Working Group. GOV.UK. April 26, 2023. https://www.gov.uk/government/publications/report-of-the-commission-on-human-medicines-isotretinoin-expert-working-group. [Google Scholar]
- 3.Barbieri JS, Weiner DM, Kakpovbia E, Nagler AR. Validating the optimal classification approach using International Classification of Diseases, 10th Revision codes to identify dermatology patients with acne. J Am Acad Dermatol. 2021;84(5):1491–1493. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Hatzimouratidis K. Sildenafil in the treatment of erectile dysfunction: an overview of the clinical evidence. Clinical Interventions in Aging. 2006. Jan 1;1(4):403–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Nackeeran S, Havanur A, Ory J, Althof S, Ramasamy R. Erectile Dysfunction is a Modifiable Risk Factor for Major Depressive Disorder: Analysis of a Federated Research Network. J Sex Med. 2021;18(12):2005–2011 [DOI] [PubMed] [Google Scholar]