Article Text

  1. P. Y. Liu,
  2. R. S. Swerdloff,
  3. C. Wang
  1. LABioMed at Harbor-UCLA Medical Center, Torrance, CA


Context Male hormonal contraceptive methods are approaching market. Factors predicting the reversibility of these methods have not been systematically examined.

Objective To determine the time and potential modulating covariables required for sperm density to recover to normal after androgen or androgen-progestin therapy sufficient for contraceptive purposes and to characterize other semen parameters as sperm density recovers.

Design An integrated multivariate stepwise and best-subset Cox proportional hazards analysis of all published studies of at least 3 months' treatment duration. Individual subject, not just summary, data were directly collected from individual investigators concerning studies published 1990-2005, which assessed subjects monthly until sperm density recovered to normal (to 20 M/mL).

Setting 29 separate studies spanning the globe.

Participants 1,549 normal (by physical, blood, and semen exam) men aged 19-51 years of predominantly Caucasian (two-thirds) or Asian (one-third) descent. This represents 1,283.5 patient years of treatment and 705 patient years of recovery and includes > 97% of all the published data.

Intervention(s) Men were treated with testosterone (T) with or without progestin for 3 to 18 months.

Results Sperm density recovered to normal after 3.4 (3.2-3.5 95% CI) months in 50% of all subjects (Kaplan-Meier). Slower recovery occurred with very prolonged treatment, Caucasian ethnicity, and with longer-acting T preparations. For a typical male, aged 35 years of Caucasian descent, treated for 1 year with intramuscular T undecanoate, the median recovery time was 4.6 (4.1-5.0) months. These times were 4.1 (3.9-4.6), 5.3 (4.9-6.3), and 3.0 (2.8-3.6), months if the male was of Asian descent, treated for 1.5 years, or treated with T enanthate, respectively. Faster recovery was predicted by greater age, faster suppression of spermatogenesis, higher baseline sperm density, and lower baseline blood LH; however, although significant, these adjusted-effect sizes were small. From multivariate model analyses, the probability of recovery within 6, 12, 16, and 24 months was 67 (61-72), 90 (85-93), 96 (92-98), and 100%, respectively. Recovery to lower sperm thresholds (3 M/mL and 10 M/mL) tracked recovery to 20 M/mL and sperm morphology and motility closely followed recovery of sperm density.

Conclusion We have identified important predictors and quantified expected probabilities of recovery. This is critical for the utilization of male hormonal contraceptive regimens.

Statistics from

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.