Hormone Therapy

Association of cannabis abuse/dependence on risks of erectile dysfunction and testosterone deficiency using a large claims database analysis.

TL;DR

Cannabis abuse/dependence is associated with increased risk of erectile dysfunction, PDE5-i prescription, and testosterone deficiency, with significantly shorter time to development of both ED and TD in a large propensity score-matched claims database analysis.

Key Findings

Cannabis abuse/dependence was significantly associated with increased risk of erectile dysfunction at 3 months to 1 year follow-up.

  • ED occurred in 0.9% of cannabis abuse/dependence patients vs. 0.2% of controls (RR = 3.99, 95% CI [3.05, 5.21])
  • 29,442 patients in each cohort after propensity score matching on 49 pertinent factors
  • Patients were drawn from the TriNetX claims database between 2005-2024
  • Patients previously diagnosed with or treated for ED were excluded prior to cohort formation

Cannabis abuse/dependence was significantly associated with increased risk of testosterone deficiency at 3 months to 1 year follow-up.

  • TD occurred in 0.2% of cannabis patients vs. 0.1% of controls (RR = 2.19, 95% CI [1.45, 3.31])
  • TD was defined as a testosterone level <300 ng/dL
  • The association with TD was not sustained at the 3-5 year follow-up interval
  • Patients with prior TD diagnoses or treatment were excluded at baseline

Cannabis abuse/dependence was significantly associated with increased PDE5 inhibitor prescription at 3 months to 1 year follow-up.

  • PDE5-i prescription occurred in 0.8% of cannabis patients vs. 0.2% of controls (RR = 3.80, 95% CI [2.86, 5.04])
  • This association was a secondary outcome of the study
  • No association between cannabis abuse/dependence and testosterone replacement therapy (TRT) prescription was found at either time interval

Cannabis abuse/dependence was associated with significantly shorter time to development of both erectile dysfunction and testosterone deficiency on Kaplan-Meier survival analysis.

  • Hazard ratio for time to ED development was 1.65 (95% CI [1.47, 1.85])
  • Hazard ratio for time to TD development was 1.34 (95% CI [1.07, 1.69])
  • Kaplan-Meier survival analysis was performed to assess differences in time to development of ED and TD across both cohorts

At 3-5 years follow-up, cannabis abuse/dependence was only significantly associated with erectile dysfunction and not with TD, PDE5-i prescription, or TRT.

  • ED occurred in 1.61% of cannabis patients vs. 1.34% of controls (RR = 1.20, 95% CI [1.01, 1.43]) at 3-5 years
  • The 3-5 year analysis included patients from 2005-2020, with 17,211 patients in each matched cohort
  • 17,948 patients with cannabis abuse/dependence and 839,496 without were identified in the 2005-2020 dataset before matching

In men under 40 years old, cannabis abuse/dependence was associated with ED and PDE5-i prescription at 3 months to 1 year but not at 3-5 years.

  • Subgroup analyses were performed for patients younger than 40 years and 40 years or older
  • In those <40 years old, only ED and PDE5-i prescription were significant at the 3-month to 1-year interval
  • Neither ED nor other outcomes were significant at 3-5 years in the under-40 subgroup

The study identified a large sample of men with cannabis abuse/dependence diagnoses from the TriNetX national claims database.

  • 30,964 patients with cannabis diagnoses and 1,473,182 without were identified between 2005-2024
  • Cohorts were propensity score-matched on 49 pertinent factors, yielding 29,442 patients per cohort
  • Men ≥18 years old were included, and those with prior prostatectomy, ED, or TD diagnoses were excluded
  • The study used ICD diagnosis codes as proxies for cannabis use, which the authors note as a limitation

Have a question about this study?

Citation

Davis R, Hershenhouse J, Maas M, Loh-Doyle J, Asanad K. (2025). Association of cannabis abuse/dependence on risks of erectile dysfunction and testosterone deficiency using a large claims database analysis.. The journal of sexual medicine. https://doi.org/10.1093/jsxmed/qdaf043