Determinants of ejaculatory dysfunction in a community-based longitudinal study

Authors


Gert R. Dohle, Department of Urology, Erasmus MC Rotterdam, PO Box 2040, 3000 CA Rotterdam, the Netherlands.
e-mail: g.r.dohle@erasmusmc.nl

Abstract

Associate Editor

Michael G. Wyllie

Editorial Board

Ian Eardley, UK

Jean Fourcroy, USA

Sidney Glina, Brazil

Julia Heiman, USA

Chris McMahon, Australia

Bob Millar, UK

Alvaro Morales, Canada

Michael Perelman, USA

Marcel Waldinger, Netherlands

OBJECTIVE

To analyse the incidence of ejaculatory dysfunction (EJD) and its associated bother, and to determine which factors predispose to incident EJD.

SUBJECTS AND METHODS

Men aged 50–78 years, registered in the general practices in Krimpen a/d Ijssel, the Netherlands, were recruited. Men were excluded if they had a history of prostatectomy, carcinoma of the bladder or prostate and neurogenic bladder disease. A baseline study and three follow-up assessments (I-III), all with questionnaires, i.e. the Benign Prostatic Hyperplasia impact index, International Prostate Symptom Score, International Continence Society (ICS)male sex questionnaire, and additional measurements, e.g. prostate volume, prostate specific antigen, were made at a mean of 2.2-year intervals. We assessed the objective variables of EJD as the ability to ejaculate, ejaculatory volume, painful ejaculation, and their associated bother (information extracted from the ICSmale sex questionnaire).

RESULTS

At baseline 671 of 1661 (40.4%) men already had EJD; the cumulative incidence of EJD was 16.5%, 24.7% and 33.1% after follow-up I, II and III, respectively. The mean percentage of men who were bothered with reduced ejaculatory volume or painful ejaculation was 18.3% and 40.6%, respectively. Multivariate analysis showed age, Sickness Impact Profile ‘social’ (questions on social impairment) and erectile dysfunction to be predisposing factors of EJD (P < 0.05 and R2 = 0.048). When EJD was defined as a significantly reduced ejaculatory volume or anejaculation only, age and previous transurethral resection of the prostate (TURP) were determinants of EJD (P < 0.05 and R2 = 0.083). Of the men who had TURP after the follow-up, 51.6% already had EJD at baseline.

CONCLUSIONS

The cumulative incidence of EJD after 6.5 years of follow-up was significant (33.1%) and EJD was bothersome, especially in men with painful ejaculation. Determinants of EJD were age, social impairment and erectile dysfunction. Predisposing factors of significant reduction of ejaculatory volume and anejaculation were age and TURP, although 51.6% of men already had EJD before TURP. Age appears to be the most significant predisposing factor of EJD.

Abbreviations
EJD

ejaculatory dysfunction

ED

erectile dysfunction

BPH-II

BPH Impact Index

SIP

Sickness Impact Profile

SFI

Sexual Function Inventory.

INTRODUCTION

Disorders of ejaculation (EJD) comprise a heterogeneous group of dysfunctions of either organic or functional origin. They can be classified as anejaculation, delayed ejaculation, retrograde ejaculation, premature ejaculation or painful ejaculation [1]. Potential risk factors for EJD are age, LUTS, neurological disease, diabetes mellitus, pharmacological treatment and surgery on the prostate. Previous cross-sectional studies reported correlations between EJD and factors such as age, LUTS and TURP; to confirm whether these variables predispose to the development of EJD, longitudinal data are required.

The Krimpen Study is a multidisciplinary community-based longitudinal study of the prevalence of LUTS and sexual dysfunction, and its sequelae, in the Dutch population. Data subsumed in this report, analysing the longitudinal patterns of EJD, were extracted from this large study. The aim of the present study was to analyse the incidence of EJD and its associated bother, and to determine factors that result in the development of EJD. Furthermore, we analysed whether EJD was more prevalent in men with LUTS and if there is a relation between TURP and EJD.

SUBJECTS AND METHODS

From August 1995 to January 1998, men aged 50–78 years were recruited from a total of 3924 men registered in the general practices in Krimpen a/d Ijssel, a suburb near Rotterdam. The design of the study was described previously [2]. Men were excluded if they had a history of prostatectomy, carcinoma of the bladder or prostate, neurogenic bladder disease or advice not to participate from their GP.

The design included baseline data and information obtained at three follow-up assessments (I-III) after 2.1, 4.2 and 6.5 years, respectively. Before every follow-up the GP reassessed the eligibility of the participants. At baseline and at every follow-up a self-administered questionnaire and a 3-day urinary frequency-volume chart were collected, and measurements were made at a health centre and urology outpatient department.

The questionnaires used included the IPSS, the BPH Impact Index (BPH-II) and sexuality-related questions of the ICSmale sex questionnaire. Health status was determined using the Sickness Impact Profile (SIP) and the Inventory of Subjective Health. Four categories of the SIP we selected were: ‘emotions, feelings and sensations’ (‘SIP emotions’), ‘leisure pastimes and recreation (‘SIP recreation’), ‘social interaction’ (‘SIP social’) and ‘usual daily work’ (‘SIP work’). Questions on medical conditions and sociodemographic factors were also included.

The questionnaires were completed at the health centre, and data recorded on current medications used. Blood pressure, height and body weight were measured. Urine was analysed by the dipstick method, to exclude lower UTI. At the urology outpatient department a DRE, TRUS and uroflowmetry were performed. The postvoid residual urine volume (using TRUS) and serum PSA level were measured.

In 1996 Donovan et al.[3] assessed the validity of the ICSmale questionnaire, by interviews with patients and urologists, testing hypotheses within sub-studies, and relating the questionnaire to frequency-volume diaries and uroflowmetry. Reliability was assessed by measures of internal consistency and a test-retest analysis. The questionnaire was confirmed to be valid and reliable.

To calculate the cumulative incidence of EJD and its associated bother, answers to the four questions of the ICSmale questionnaire were analysed (Appendix). Men who reported having anejaculation, any reduction in ejaculatory volume and/or painful ejaculation on one or several occasions were considered to have EJD. To correct for possible over-reporting of EJD, a second analysis was used in men with EJD, which was redefined as a significant reduction of ejaculatory volume or anejaculation only. As questions on premature ejaculation were not part of the ICSmale questionnaire, no data on this item were available.

At baseline men were considered not to have EJD if they reported no pain or discomfort during ejaculation, or any reduction in ejaculatory volume. The factors of men developing EJD after follow-up on one or several occasions were the object of study. We used univariate and multivariate logistic regression analysis to assess these factors, considering P < 0.05 to be statistically significant.

RESULTS

At baseline, of 3924 men aged 50–78 years registered in all general practices in Krimpen a/d Ijssel, 3398 were invited to participate. Of the 1688 respondents, baseline data on EJD were complete in 1661. At 2.1, 4.2 and 6.5 years of follow-up, data were collected from 1186, 875 and 809 men, respectively (Fig. 1).

Figure 1.


The flow chart of the Krimpen study.

At baseline, 671 of 1661 (40.4%) men already had EJD; after 2.1 years of follow-up, 16.5% developed EJD. Bother was reported by 13.8% of men with reduced volume and by 27.3% of men with painful ejaculation. The cumulative incidence of EJD after 4.2 years of follow-up was 24.7%, with bother reported by 21.4% of men with a reduction in volume and by 54.5% of men with pain. In all, 328 men (33.1%) had EJD after 6.5 years of follow-up. The cumulative incidence of men who were bothered by either reduced volume or painful ejaculation was 18.3% and 40.6%, respectively. Few men reported painful ejaculation (cumulative incidence, 15 of 990 at follow-up III). Also, we stratified the associated bother of reduced ejaculatory volume and anejaculation by age, but there was no significant difference between the age groups (< 60 years, 20.8% bother; 60–69 years, 15.9% bother; >70 years, 21.1% bother) and the overall bother was 18.4%.

An overview of all the possible determinants of EJD analysed is outlined in Table 1. After multivariate logistic regression analysis we identified age, SIP ‘social’ and erectile dysfunction (ED) as predisposing factors of EJD (P < 0.05 and R2 = 0.048, Table 1).

Table 1. 
Univariate and multivariate logistic regression analysis of determinants of men with no EJD at baseline, who developed EJD after three follow-up assessments
CharacteristicPExp (β)R2
  • *

    Reduced or significantly reduced ejaculatory volume, anejaculation, pain/discomfort during ejaculation;

  • †Significantly reduced ejaculatory volume or anejaculation only.

Univariate*
Age, years<0.0011.0490.026
Body mass index, kg/m20.5661.014 
Body weight, kg0.9571.000 
Height, m0.3140.342 
Alcohol consumption, units/day0.5720.906 
Prostate volume, mL0.2211.006 
PSA level, ng/mL0.0830.949 
Blood pressure, mmHg
 systolic0.8450.999 
 diastolic0.5440.995 
Treatment for or presence of (yes vs no):
 Hypertension0.6900.925 
 Diabetes0.8180.915 
 Cardiac symptoms0.5361.239 
 Nitrite in urine0.6290.820 
 Leukocytes in urine0.2270.329 
 UTI0.4822.025 
Marital status0.7241.119 
(widower/divorced vs married/living together)
Educational level0.8100.949 
(primary education only vs more than primary education)
SIP (points)
 social0.0131.0200.009
 emotional0.2231.009 
 work0.2341.013 
 recreation0.0791.008 
IPSS, points0.0031.0440.012
BPH-II, points0.0171.1590.008
Smoking habits (smokers vs non-smokers)0.7751.045 
Period of abstinence, years0.9771.002 
ED (yes vs no)<0.0011.9170.019
Vasectomy (yes vs no)0.1451.248 
Multivariate*  0.048
Age, years<0.0011.047 
SIP social (0–100 points)0.0131.021 
ED (yes vs no)0.0021.741 
Multivariate  0.083
Age, years0.0011.128 
History of TURP (yes vs no)0.04419.365 

To correct for possible over-reporting of EJD, a second analysis was done for men with EJD, which was redefined as a significant reduction of ejaculatory volume or anejaculation only. Multivariate analysis showed age and a history of TURP to be determinants of EJD (P < 0.05 and R2 = 0.083, Table 1). Of 31 men who had TURP during the follow-up, 16 already had EJD at baseline. In only four of these 16 men was this possibly related to the use of α1-adrenoceptor antagonists.

DISCUSSION

Sexual dysfunction is highly prevalent in men with LUTS; as a result of the ageing of the population and liberalisation of attitudes toward sexuality, issues of sexual dysfunction, amongst which is EJD, are reported more frequently. Identifying risk factors for EJD might result in the development of better management strategies and improve the counselling of men with ejaculatory problems. To determine these risk factors in men with LUTS, longitudinal data from community studies are required.

In the present study the prevalence of EJD at baseline was 40.4% and there was a significant increase in incidence of 33.1% after 6.5 years of follow-up. In concordance with other studies, age appeared to be the most important predisposing factor in the development of EJD. At baseline the significant risk factors for EJD were age, ED, LUTS and a history of TURP [4]. Interestingly, contrary to other cross-sectional reports, we found no relationship with LUTS after multivariate analysis. Probably the coincidence of LUTS and EJD is mainly due to a simultaneous effect of age on both LUTS and EJD.

Reduced ejaculatory volume and pain or discomfort on ejaculation are two distinct entities that probably have a different origin. Furthermore, a reduction in ejaculatory volume is a subjective observation by the men, and which was not quantified objectively. To correct for possible over-reporting of EJD, a second analysis was used in men with EJD, which was redefined as a significant reduction of ejaculatory volume or anejaculation only. Multivariate analysis showed age and a history of TURP to be determinants of EJD (P < 0.05 and R2 = 0.083). Despite analysing a large variety of possible determinants of EJD, and correction for over-reporting, we identified only a few significant predisposing factors for the development of EJD, with weak correlations.

Two other longitudinal studies focused on sexual interest and sexual activity during ageing [5,6]. One of these found that levels of sexuality remained more stable than previously suggested in cross-sectional studies [5]. However, Verwoerdt et al.[6] found patterns of sustained activity and interest relatively typical for men aged 60–70 years, decreasing activity and interest for men aged 70–80 years, and continuously absent activity and interest beyond that age.

In a recent longitudinal analysis of changes in sexual function over a 9-year follow-up, a broad spectrum of sexual function variables were analysed, including sexual intercourse, erection frequency, sexual desire, ejaculation with masturbation, satisfaction with sex, and difficulty with orgasm. For ejaculation, only the frequency of ejaculation with masturbation was recorded, which showed no change between baseline and follow-up. For all other variables there were statistically significant changes strongly associated with age [7].

The Sexual Function Inventory (SFI) was developed to measure male sexual function; it incorporates questions on delayed ejaculation and decrease in ejaculatory volume. Cross-sectional analysis of the age-specific mean (sd) on ejaculatory function by O’Leary et al[8,9], based on the SFI (normal function = 8), identified a decline in ejaculatory function with age, from 7.4 (1.4) for men in their fifth decade to 3.6 (3.2) for men aged ≥70 years.

A cross-sectional study by Sak et al.[10] assessed 1420 men for prostate symptoms and described a correlation between the variables of urinary function and EJD, also based on the SFI. EJD correlated significantly with age (P < 0.001), IPSS (P < 0.01), maximum urinary flow rate (P < 0.01) and the IPSS quality-of-life question (P < 0.05), but after multivariate analysis only age remained a significant determinant (P < 0.001).

Tubaro et al.[11] found sexual dysfunction in men with LUTS to be significantly associated with urinary symptoms, in particular urine loss. The odds ratio calculated from multivariate logistic regression for the contribution of urine leakage to a reduction or absence of ejaculate was 1.91. By contrast with the present findings, where associated bother was independent of age, a reduced volume of ejaculate was reported by >55% but was only perceived as a problem by a minority of patients (33.4%), who tended to be younger. Age was significantly associated with a reduction or absence of ejaculate. A reduced urinary stream (odds ratio 1.13), sedentary lifestyle (2.17) and concomitant cardiovascular disease (a higher mean score for ICSmale) were more likely to also result in a lack or reduction of ejaculate.

Similar results were reported by Frankel et al.[12], who compared a community population (423 men aged ≥40 years, in the UK) to a clinic sample (1271 men aged ≥45 years, in 12 countries) with LUTS. A reduction in ejaculation was more common and associated bother was higher in the clinic population (62.3% vs 47.0%, and 47% vs 24%, respectively). The strongest and most significant associations were with storage symptoms, particularly incontinence, with odds ratios of >2.0. Also, there was a clear relationship between reduced ejaculation and age.

Vallancien et al.[13] analysed sexual function in 1274 European men with LUTS, using the Danish Prostate Symptom Score sex questionnaire. Reduced ejaculation and pain/discomfort on ejaculation was reported by 63% and 23% of patients, respectively, of whom 82% and 91% were bothered. They also identified reduced ejaculation to be strongly related to age (42% aged < 60 years, 82%≥70 years), as well as LUTS severity (55% mild, 68% severe) and previous BPH surgery.

In the study by Rosen et al.[14] of 11 114 men, 46% reported a reduced amount of ejaculation and 5% reported anejaculation. The prevalence increased significantly with age and severity of LUTS, and bother was unrelated to age.

Several longitudinal studies of the semen quality in men reported no reduction in volume over time. However, in those studies the men were relatively young [15–17]. Cross-sectional studies found a reduced sperm output in ageing men [18,19].

The present multivariate longitudinal regression analysis showed that a history of TURP is a risk factor for the development of reduced ejaculatory volume and anejaculation. However, of the men who had TURP after the follow-up, 51.6% already had EJD at baseline. In only four of 16 men was this possibly related to the use of α1-adrenoceptor antagonists.

Retrograde ejaculation after TURP, which has been reported to occur in 65–70% of patients, depends on the degree of bladder neck resection [20]. Some authors suggested that antegrade ejaculation can be maintained in 80% of patients if >1 cm of the supramontanal prostate is preserved. If they are well-informed, patients are willing to accept the slight voiding symptoms that remain, compensated by the maintenance of normal sexuality [21].

Arai et al.[22] prospectively determined the impact of standard TURP, transurethral microwave thermotherapy, interstitial laser coagulation of the prostate and transurethral needle ablation on quality of life and sexual function. Ejaculation loss or severe decrease in ejaculate volume was reported by 48.6% of patients who had TURP. There was a highly significant (P < 0.001) association of ejaculatory dysfunction with an adverse impact on sexual activity.

Severe bother relating to the loss of ejaculation after TURP was reported [23]. A careful interview with patients complaining of sexual dysfunction after TURP showed the false impression by patients that EJD is completely or partly equal to ED [24]. Thus counselling on the risk of retrograde ejaculation after TURP could diminish the associated bother.

In conclusion, EJD is highly prevalent in ageing men with LUTS. At baseline, 40.4% of men had EJD, and the cumulative incidence of EJD after 7 years of follow-up is significant (33.1%); EJD is bothersome, especially in men with painful ejaculation. Associated bother is independent of age, older men also being bothered. The determinants of EJD were age, social impairment and ED; LUTS appeared not to be correlated. Predisposing factors of a significant reduction of ejaculatory volume and anejaculation are age and a history of TURP. Of the men who had TURP after follow-up, half already had EJD at baseline, and in only four of 16 men was this possibly related to the use of α1-adrenoceptor antagonists. Overall, age appears to be the most significant predisposing factor for EJD.

CONFLICT OF INTEREST

None declared.

Appendix

  • Items of the ICSmale sex questionnaire used to define EJD.

  • 1) Do you have pain or discomfort during ejaculation?
    • No

    • Yes, slight pain/discomfort

    • Yes, moderate pain/discomfort

    • Yes, severe pain/discomfort

  • 2How much of a problem is this for you?
    • Not a problem

    • A bit of a problem

    • Quite a problem

    • A serious problem

  • 3Do you have an ejaculation of semen?
    • Yes, normal quantity

    • Yes, reduced quantity

    • Yes, significantly reduced quantity

    • No ejaculation

  • 4How much of a problem is this for you?
    • Not a problem

    • A bit of a problem

    • Quite a problem

    • A serious problem

Ancillary