Introduction

Numerous studies have demonstrated the effectiveness of tricyclic and SSRI antidepressants in the treatment of rapid ejaculation (RE).^{1, 2, 3, 4} Although doses used for RE are substantially lower than those used for clinical depression, daily administration of such drugs is still often associated with unpleasant side effects. Such adverse effects may reduce compliance, particularly for couples who engage in sexual activity infrequently.

So as to attenuate the adverse effects of daily administration, we recently demonstrated that 25 mg of the tricyclic antidepressant clomipramine, when administered as a single dose either 6 or 12 h prior to anticipated sexual activity, could delay ejaculation on average by 2.5 and 6 min, respectively.^{4, 5} However, as with many pharmacotherapeutic treatments, not all patients respond positively to clomipramine treatment, and, more specifically, when taken 4–6 h prior to intercourse, only about half of the participants responded with increases in ejaculatory latency greater than 60 s.⁶ Although follow-up treatment on four 'nonresponders' with daily administration of clomipramine (20 or 30 mg) resulted in acceptable increases in ejaculatory latencies, it would be clinically advantageous to know the patient's potential responsiveness to an 'as needed' 25 mg clomipramine regimen, prior to the commencement of treatment.

The present study attempted to identify a priori factors that might enable the prediction of patients' response or nonresponse undergoing treatment for RE with 25 mg clomipramine taken 'as needed.' In this analysis, we have identified three such factors that might readily be used by the clinician to predict treatment efficacy under this regime.

Methods

Subjects

In response to a newspaper interview describing the project with one of the authors, 39 men were initially screened by phone to determine their eligibility for the study. Of these, 27 agreed to participate and were deemed eligible based on the following inclusion criteria. Participants had to be 18 y of age or older, in a stable heterosexual relationship, and reporting a minimum 6-month history of RE. In addition, they had to be free of current or past psychiatric disorders or substance abuse (alcohol or drugs), and of diseases, medication use, and surgical procedures that might impact sexual function. All patients provided written consent prior to participation, as approved by the Medical Ethics Committee of Erasmus Medical Center (Erasmus MC, The Netherlands).

The ejaculatory disorder of these men was assessed through a retrospective questionnaire combined with a structured diagnostic interview.⁷ Specifically, these men reported an intravaginal ejaculation latency time (IELT) under 150 s, a lack of ejaculatory control (response of '1' or '2' on a seven-point scale, where 1=none at all), and a moderate level of sexual dissatisfaction (mean=4.0 on a seven-point scale, where 1=very dissatisfied). These criteria, consistent with the DSM-IV-TR characterization of rapid ejaculation,⁸ reliably discriminate rapid ejaculators from sexually functional men.⁹

Rationale for and definition of relevant variable domains

We initially identified five domains of theoretical and/or empirical relevance to RE that might differentiate responders from nonresponders. Measures relevant to these domains were taken prior to treatment or under placebo and included: initial IELT, sexual satisfaction, arousal/excitement during sexual stimulation, ejaculatory frequency, and erectile functioning. The first two of these were fundamental to the RE classification of the participant and were, to a large extent, independent measures of the severity of the RE problem.^{9, 10} Specifically, (1) IELT, estimated in seconds, ranged from 0 to 150 and (2) self-reported 'satisfaction with sex life/sexual relationship,' rated on a seven-point Likert scale, provided a general measure of distress related to the dysfunctional condition.

The remaining three variables were included on the basis of their previous theoretical or empirical relevance to RE. Specifically, men with RE are purportedly characterized by hyperarousal to sexual stimulation,¹¹ and therefore we included a measure of self-reported arousal (seven-point scale) obtained during psychosexual stimulation in the lab. In addition, self-reported frequency of ejaculation/orgasm has been related to RE, with past research suggesting that RE men exhibit lower frequencies of sexual activity and orgasm relative to sexually functional counterparts.^{12, 13, 14} Presumably, lower frequencies in RE men lead to shorter IELTs, because such men have had fewer opportunities to learn to anticipate and control their ejaculatory response, and/or because the normal inhibition of ejaculation caused by the male refractory period plays a diminished role.^{12, 13, 14} Finally, self-reported confidence in achieving an erection was included on the basis of previous research demonstrating that RE men who report erectile problems are less likely to benefit from the ejaculatory-inhibiting effects of clomipramine,⁴ presumably because of the mild antierectile effects this agent is known to impart.

Selection of final variables

Measures from these five domains were first intercorrelated to eliminate possible redundancy among the predictor variables, with elimination of one of a pair of variables that correlated at or above 0.50. This procedure reduced the number of domains to four, as arousal/excitement was correlated with ejaculation latency (r=0.69; P<0.01).

The four remaining variables were considered for inclusion as predictors of response or nonresponse to clomipramine. All measures except IELT were obtained prior to clomipramine treatment through questionnaire assessment, administered either as part of the screening procedure for inclusion in the study or following erotic stimulation presented in a psychophysiological laboratory session. IELT was determined from daily logs kept by the participant, and represented the average of the estimated times between intromission and ejaculation for all coital attempts, leading to ejaculation during a 17–20-day nontreatment phase.

Procedure

Data were collected as part of a double-blind, placebo-controlled investigation studying the effects of 25 mg clomipramine, taken as needed 4–6 h prior to intercourse.⁵ Clomipramine, also known as Anafranil^®, is a tricyclic antidepressant which, at clinical doses for depression, has been found to result in ejaculatory failure in about 25% of patients with psychological disorders.¹⁵ For purposes of this study, RE responders were identified as RE men whose ejaculatory latency increased during intercourse by an average of at least 60 s relative to placebo during the home-testing trials using 25 mg of clomipramine (responders' latency change=+4.9 min vs nonresponders' latency change=+0.4 min).

Analytical strategy

Logistic regression was used to determine the predictive utility of each of the four covariates on the outcome variable, response or nonresponse to clomipramine treatment. This analysis generates an exponential value for the slope (b) for each variable, thereby specifying the odds of one outcome to the other (e.g., response to nonresponse). The four predictor variables were entered using a backward elimination procedure, so as to identify the most parsimonious combination that would predict responsiveness to treatment. Backward elimination begins by entering all variables in the equation and removes them one at a time, based on the largest significance level. For this analysis, the significance of change, as determined by the likelihood ratio, upon the removal of each variable was used as the criterion. That is, any variable which, when removed, significantly reduced the fitness of the model (P<0.05) was retained in the model.

Results

Descriptive information

Two subjects were removed from the analysis because their IELTs (estimated prospectively) during placebo testing exceeded 150 s, and two others failed to provide complete sets of data. The final analyses were therefore carried out on 23 men. Table 1 provides descriptive information for responders and nonresponders on the four variables included in the regression analysis. In addition, differences resulting from comparison with t-tests are indicated.

Table 1 - Comparison of clomipramine responders and nonresponders on age and the four variables included in the analysis.

Full table

Logistic regression analysis

The four variables included in the initial analysis significantly predicted the responders and nonresponders (² (4)=21.5; P=0.001). Backward elimination resulted in the removal of one variable: ability to keep an erection. The remaining three variables—IELT, ejaculatory frequency, and sexual satisfaction—were retained in the model, with the removal of any one of these decreasing the overall predictive ability significantly (Table 2).

Table 2 - Results of logistic regression using backward entry, indicating removal of 'ability to keep erection,' a variable not meeting the 0.05 criterion for retention in the model.

Full table

Odds ratios (Table 3) indicated that men with higher IELTs, greater overall sexual satisfaction, and a higher frequency of ejaculation were more likely to respond to the antiejaculatory effects of 'as needed' clomipramine. Based on these covariates, 10 of 12 responders (83.3%) were predicted correctly, and nine of 11 nonresponders (81.8%) were predicted correctly.

Table 3 - Results of logistic regression, including odds ratios, for the final three variables retained in the model.

Full table

Discussion

This study identifies three variables that enable the prediction of responsiveness to 25 mg clomipramine taken 'as needed.' Specifically, men with low initial IELTs, particularly those under about 40 s, showed a poorer prognosis using this treatment strategy. Since IELTs were estimated rather than determined using the stopwatch method, this finding needs to be interpreted with caution. Nevertheless, the influence of this intuitively logical predictor—correlated with the participant's age and undoubtedly related to the severity of the problem—is perhaps less surprising than that of two other predictors of responsiveness, namely, overall sexual satisfaction and ejaculatory frequency. Specifically, men who initially rate their sexual satisfaction from 'average' to 'good' are eight times more likely to respond to as needed clomipramine than those who rate it as 'below average.' Interestingly, although 'satisfaction' ratings across RE men and sexually functional groups correlate fairly well with IELTs,^{4, 9} when ratings are restricted to a group of RE patients whose IELTs are less than 150 s, these two measures are independent of one another. Such findings suggest that a patient's subjective evaluation of his problem—as determined by an index of overall sexual satisfaction—may be as important, if not more so, to predicting a successful treatment outcome than simply knowing his initial IELT.

Finally, men who ejaculate twice a week or more, whether through intercourse or masturbation, are 19 times more likely to show substantial improvement under this regimen of clomipramine. Although differences in ejaculatory frequency between men with RE and sexually functional men have been reported previously,^{12, 13} we were surprised that this measure so strongly differentiated responders from nonresponders to as needed clomipramine treatment. Whether this effect is related to the inhibition of ejaculation due to a greater impact of refractory periods or to a habituation of arousal that might occur with more frequent sexual activity is not clear. Just as likely, this factor might stand as proxy for some other variable that changes along with sexual frequency; for example, higher accumulation of clomipramine in the system resulting from more frequent use during the treatment period, at least relative to nonresponders.

In conclusion, the results of this study suggest that three factors might routinely be investigated in RE men as part of the psychosexual interview, before proceeding with an as needed regimen of 25 mg clomipramine. Specifically, men who estimate an initial IELT greater than 60 s, who assess their sexual satisfaction as above average (5 or higher on a seven-point scale), and who engage in sexual activity several times or more each week, appear to be the better candidates for this type of treatment. Patients not meeting these criteria may be considered for 20 or 30 mg daily clomipramine administration.⁶ It is yet to be explored whether the three factors identified in this study might also effectively predict responsiveness to other ejaculation-retarding drugs such as the SSRIs.

References

1. Althof SE. Pharmacologic treatment of rapid ejaculation. Psychiatr Clin North Am 1995; 18: 85–94. | PubMed | ChemPort |
2. McMahon CG. Treatment of premature ejaculation with sertraline hydrochloride. Int J Impot Res 1998; 10: 181–184. | Article | PubMed | ISI | ChemPort |
3. Waldinger MD, Hengeveld MW, Zwinderman AH, Olivier B. Effect of SSRI antidepressants on ejaculation: a double-blind, randomized, placebo-controlled study with fluoxetine, fluvoxamine, paroxetine, and sertraline. J Clin Psychopharmacol 1998; 18: 274–281. | Article | PubMed | ISI | ChemPort |
4. Haensel SM, Rowland DL, Kallan K, Slob AK. Clomipramine and sexual function in men with premature ejaculation and controls. J Urol 1996; 56: 1310–1315. | Article |
5. Strassberg DS et al. Clomipramine in the treatment of rapid (premature) ejaculation. J Sex Marit Ther 1998; 25: 89–101. | Article |
6. Rowland DL, Brazao de Gouveia C, Slob AK. Effective daily treatment with clomipramine in men with premature ejaculation when 25 mg (as required) is not effective. Br J Urol 2001; 87: 357–360. | Article |
7. Rowland DL, Slob AK. Assessment of sexual health: premature ejaculation. (online); CD-ROM. Abstracts from Ovid Technologies, HaPI Item 209897 2000.
8. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4th edn., text revision. American Psychiatric Association: Washington, DC, 2000, 943pp.
9. Rowland DL, Cooper SE, Schneider M. Defining premature ejaculation for experimental and clinical investigations. Arch Sex Behav 2001; 30: 235–254. | Article | PubMed | ISI | ChemPort |
10. Rowland DL. Treatment of premature ejaculation: selecting outcomes to determine efficacy. Int Soc Sex Impot Res Newsbull 2003; 10: 26–27.
11. Kaplan HS. Premature Ejaculation: Overcoming Premature Ejaculation. Bruner/Mazel: New York, 1989.
12. Spiess WS, Geer J, O'Donohue WT. Premature ejaculation: investigation of factors in ejaculatory latency. J Abnorm Psychol 1984; 93: 242–245. | Article | PubMed | ChemPort |
13. Grenier G, Byers S. Operationalizing premature or rapid ejaculation. J Sex Res 2001; 38: 369–378. | ISI |
14. Masters WH, Johnson VE. Human Sexual Inadequacy. Little, Brown: Boston, 1970, pp 92–115.
15. Balon R, Yeragani VK, Pohl R, Ramesh C. Sexual dysfunction during antidepressant treatment. J Clin Psychiatry 1993; 54: 209–212. | PubMed | ChemPort |