Although delayed ejaculation (DE) is typically characterized as a persistently longer than anticipated or desired time to ejaculation (or orgasm) during sexual activity, a timing-based definition of DE and its association with serum testosterone has not been established in a large cohort.

To examine in an observational study estimated intravaginal ejaculatory latency time (IELT) and masturbatory ejaculation latency time (MELT) in men self-reporting DE, assess the association of IELT and MELT with�serum testosterone levels, and determine whether correlation with demographic and sexual parameters exist.

Men who resided in the United States, Canada, and Mexico were enrolled from 2011 to 2013. Self-estimated IELT and MELT were captured using an Ejaculatory Function Screening Questionnaire in a sample of 988 men screened for possible inclusion in a randomized clinical trial assessing testosterone replacement therapy for ejaculatory dysfunction (EjD) and who self-reported the presence or absence of DE and symptoms of hypogonadism. Additional comorbid EjDs (ie, anejaculation, perceived decrease in ejaculate volume, and decreased force of ejaculation) were recorded. Men with premature ejaculation were excluded from this analysis. IELT and MELT were compared between men self-reporting DE and men without DE. The associations of IELT and MELT with serum testosterone were measured.

IELT, MELT, and total testosterone levels.

Sixty-two percent of screened men self-reported DE with or without comorbid EjDs; 38% did not report DE but did report at least one of the other EjDs. Estimated median IELTs were 20.0 minutes for DE vs 15 minutes for no DE (P < .001). Estimated median MELTs were 15.0 minutes for DE vs 8.0 minutes for no DE (P < .001). Ejaculation time was not associated with serum testosterone levels. Younger men and those with less severe erectile dysfunction had longer IELTs and MELTs.

Estimated ejaculation times during vaginal intercourse and/or masturbation were not associated with serum testosterone levels in this study; thus, routine androgen evaluation is not indicated in these men.

This large systematic analysis attempted to objectively assess the ejaculation latency in men with self-reported DE. Limitations were that ejaculation time estimates were self-reported and were queried only once; the questionnaire did not distinguish between failure to achieve orgasm and ejaculation; and assessment of DE was limited to heterosexual vaginal intercourse and masturbation.

IELT and MELT were longer in men with DE, and there was no association of ejaculation times with serum testosterone levels in this study population.