Retrograde ejaculation and sexual dysfunction in men with diabetes mellitus: a prospective, controlled study



Jens Fedder, Laboratory of Reproductive Biology, Scientific Unit, Horsens Hospital, Sundvej 30, DK-8700, Horsens, Denmark. E-mail:


Retrograde ejaculation (RE) and erectile dysfunction may be caused by diabetes mellitus (DM), but the prevalence of RE among DM patients is unknown. A prospective, blinded case–control study comparing men with DM with matched controls according to RE and erectile dysfunction was performed. Twenty-seven men with DM matched the inclusion criteria and agreed to participate in the study, and of these 26 delivered an ejaculate. We were able to recruit 18 matching controls, and of these 16 delivered an ejaculate.

Nine of 26 men with DM who delivered an ejaculate had RE, whereas none of 16 controls had RE (p < 0.01). The mean duration of DM was longer for DM patients with RE (20 years) compared with DM patients in whom RE could not be demonstrated (13 years), but the difference was not statistically significant. RE could not be associated with BMI, waist circumference, blood pressure, Haemoglobin A1c (HgbA1c), high-density lipoprotein HDL cholesterol, triglycerides, fasting glucose, or s-testosterone. Diabetics suffering from RE more frequently exhibited erectile dysfunction compared with non-diabetics and diabetics without RE, and the last-mentioned group again more frequently than controls. These findings could not be explained by use of antihypertensive drugs.

Whereas none of the included control participants showed signs of abnormal ejaculation, every third man with DM exhibited retrograde ejaculation. It is important to be aware of this association, and that post-ejaculatory urine is routinely analysed from aspermic fertility clinic attendants and diabetics with low ejaculate volumes.


Ejaculatory disorders may be caused by organic or psychological factors (Kamischke & Nieschlag, 1999). Whereas anejaculation is defined as total failure of seminal emission into the posterior urethra (Kamischke & Nieschlag, 1999), retrograde ejaculation (RE) is a situation in which semen is indeed emitted, however, not through the penile part of the urethra, but into the urine bladder through an open internal vesical sphincter (bladder neck) during ejaculation. RE may be partial or complete; in the latter case the man shows aspermia and is obviously infertile. As RE is an infrequent cause of infertility (Yavetz et al., 1994), diagnoses of RE may be overlooked by fertility doctors. RE is owing to neuropathy, which may be caused by diabetes mellitus (DM), neurological disorders, or pharmaceuticals. Coronary heart disease, neuropathy, retinopathy, nephropathy and foot wounds in diabetic patients seem related to hyperglycaemia, which is frequent in dysregulated diabetes.

Spermatozoa have little chance to survive in unprepared urine because of high osmolarity and low pH (Aust et al., 2008). This effect seems immediate and irreversible (Aust et al., 2008), and electron microscopy studies have shown that the mechanism might be destruction of the plasma membrane allowing influx of extracellular fluid into the spermatozoa (Vijayakumar et al., 1986) compromising the fertilizing ability of the spermatozoa.

RE is often searched reversed by medical treatment using alpha-agonistic or anticholinergic and antihistaminic drugs. Azoospermic men in general are now a day often treated with spermatozoa retrieved from testis or epididymis (Kamischke & Nieschlag, 1999). However, in some cases of RE, antegrade ejac-ulation can be obtained by medical treatment (Yavetz et al., 1994), electro vibration applied to the penile surface or to the ventral surface of the ‘Glans penis’, or by trans-rectal electro ejaculation (Gerig et al., 1997). Moreover, if the sperm count is sufficiently high, spermatozoa for intrauterine or assisted reproduction may be harvested from post-coital urine after ejaculation (Shangold et al., 1990; Aust et al., 2008).

As little focus has been set hereto, we have found it relevant to evaluate the frequencies of retrograde ejaculation, aspermia (no semen ejaculated) and erectile dysfunction in men with DM compared with normal men.

Although fertility personnel is well aware that a patient presenting with retrograde ejaculation may be diabetic, little focus has been set to identify patients with RE based on diabetes diagnoses. This study is the first study, which focuses primarily on the frequency of RE in men with DM compared with controls and secondarily on associations between RE and diabetes regulation [Haemoglobin A1c (HgbA1c)], obesity, metabolic syndrome, diabetic complications and erectile dysfunction. Furthermore, we have searched to highlight the frequency of sexual dysfunction in diabetic men with and without RE compared with controls.

Materials and methods

Definition of RE, aspermia and azoospermia

Semen may pass into the urine bladder to variable extent, as RE can be partial or complete. Indeed, some spermatozoa may be present in the post-ejaculatory urine also in men without RE, as spermatozoa may be retrieved in the urethra. Since at this point, literature does not define clear-cut limits for the number of spermatozoa in post-ejaculatory urine from normally ejaculating men, we have chosen to define this limit as one million spermatozoa in total in the urine. Although others have found surprisingly large numbers of spermatozoa in post-ejaculatory urine from normal fertile men (Sigman et al., 2008), the control men in this study all had significantly less than a total of one million post-ejaculatory urine spermatozoa. Aspermia is a condition in which no ejaculate is produced by ejaculation, whereas azoospermia is a condition in which an ejaculate with no spermatozoa is ejaculated.

Recruitment of participants

Men with DM and controls were recruited from a Diabetes Biobank including nearly 5 000 diabetic patients and nearly 5 000 controls, all with Danish ethnicity. The biobank was constructed on a case–control basis including men and women in the ages 26–74 years. Patients with DM, type 1 and 2, were identified from a local diagnosis code register (Vejle County) and a central database at Vejle Hospital. Control persons, matching the men with DM in sex and age, were identified from the Danish Central Person Register, which is an EDP-based systematic registering of name, Birth County, current address and civil status of all Danish citizens since 1924. Participants were interviewed and they filled out a questionnaire according to lifestyle, medications and DM-related complications. In addition, anthropometric measures such as height, weight, waist- and hip circumference, fat per cent and blood pressure were registered. Plasma, serum and urine samples were collected for scientific purposes.

Participants were recruited among persons included in the Diabetes Biobank. For the subproject on RE only men 30–55 years of age were included. For practical reasons it was not possible to recruit men from the two geographically distant hospitals in which men were recruited to the Diabetes Biobank, but all participants examined in Horsens and Vejle Hospitals were asked to participate in this study.

Collection and analysis of ejaculates and urine samples

Men were recommended 3–7 days of abstinence before ejaculation. Immediately before ejaculation, the men urinated with the purpose that the urine bladder was as empty as possible. Immediately after voiding, the men masturbated and ejaculates were collected in 50-mL plastic syringes (Falcon, Meus, Piove di Sacco, Italy). Immediately following masturbation, a fresh urine sample was collected in a 50-mL plastic syringe. This had the consequence that in some cases where the post-ejaculatory urine had a volume of more than 50 mL, the total number of spermatozoa in the urine may have been underestimated. The participants were only asked to deliver one ejaculate and one post-ejaculatory urine sample to recruit as many participants as possible.

Urine samples were analysed immediately, and ejaculates were analysed after allowed liquefaction for approximately half an hour. Concentration and total number of spermatozoa in the ejaculate and urine were calculated, and sperm motility was determined using a Makler chamber (Sefi Medical Instruments Ltd., Haifa, Israel) and a Leica light microscope with objective ×400. Fat-% was determined with a Tanita Body Composition Analyser 418 (Tanita Europe BV, Amsterdam, the Netherlands).

Analysis of blood samples

Concentrations of serum glucose, triglycerides and C-reactive protein were measured on a Roche Modular System (Roche Diagnostics Deutchland GmbH, Mannheim, Germany). Haemoglobin A1C (HgbA1C) was measured on the Tosoh 7HPLC System (Tosoh Europe N.V., Tessenderlo, Belgium), and serum total testosterone was measured by mass spectrometry with a reference range of 8.4–30.0 mmol/L.

Questionnaire for self-evaluation of erectile function

As a result of a consensus of a large group of experts (NIH Consensus Conference, 1993), an International index of erectile function (IIEF), which is a self-report composite multidimensional measurement scale for male sexual function, has become available (Rosen et al., 1999). The IIEF-15, which has been used throughout this study, is a 15-item scale, and each item is rated from zero or one to five. A response of one for a question was considered the least functional and a response of five the most functional. The questions encompass erectile function (Q1–5 + Q15), intercourse satisfaction (Q6–8), orgasmic function (Q9–10), sexual desire (Q11–12) and overall satisfaction (Q13–14).

Study approval

The study has been approved by the Danish Data Protection Agency and the local Danish Scientific Ethics Committee (M-20070268) and registered at (NCT01743599).


Frequency of RE in diabetic men compared with controls were calculated with Fisher's exact test owing to the relatively small/modest number of persons included in the study. Mean, 95% confidence intervals and p-values of parametric clinical and biochemical data and the answer scores of questions on erectile dysfunction were calculated using an unequal variance t-test (Ruxton, 2006) and the bootstrap method (James & Bithell, 2000).


Descriptive variables

Of 4 898 diabetic men and controls, only 1 337 fulfilled the inclusion criteria for this study. For practical reasons, we only contacted men recruited to the database via Horsens Hospital. Of 71 diabetics and 70 controls, 53 (75%) and 38 (54%), respectively, accepted to receive written material about the project. Only about half of the informed men chose to participate in the study; the major part agreed to participate in the full project, but a few only participated with questionnaire because of former vasectomy (Fig. 1). At the end, 27 diabetic men and 18 control men were enrolled in the study.

Figure 1.

Scheme showing the recruitment of study participants in numbers. The Diabetes Biobank included nearly 5 000 men (diabetics and non-diabetics), but only a little over ¼ of these met the study inclusion criteria (men 30–55 years of age). Of these men all, who lived within manageable distance to Horsens Hospital (71 diabetics and 70 non-diabetics), were contacted by phone. At the end, 27 diabetic men and 18 controls participated.

Prevalence of retrograde ejaculation, aspermia and azoospermia in diabetic men

RE was demonstrated in nine of 26 (34.6%) of diabetic men, who delivered an ejaculate, whereas none of 16 controls exhibited RE (p < 0.01, Table 1). Of the nine men with RE, two had aspermia, and in each group of diabetic men and controls, one man had azoospermia. The ejaculate volume of the azoospermic control person was 4.0 mL and thus RE was considered unlikely. Conversely, the ejaculate volume of the diabetic man with azoospermia was only 0.7 mL and RE cannot be excluded; however, this man was included in the non-RE group according to the chosen definition. Ejaculate volumes were not significantly different between the groups (Table 2).

Table 1. Frequencies of retrograde ejaculation, aspermia, and azoospermia in diabetic men and controls
 Included participantsParticipants who delivered ejaculatesNumbers with azoospermiaNumbers with retrograde ejaculateNumbers with aspermia
  1. a

    p < 0.01 (Fishers exact test).

  2. b

    Each with retrograde ejaculation.

Diabetes mellitus272619a2b
Table 2. Factors which may increase the risk of retrograde ejaculation (RE) and impotence in patients with diabetes mellitus (DM). P1: Diabetic men with RE vs. diabetic men without RE. P2: Diabetic men with RE vs. controls without DM. p-values were calculated with Mann–Whitney test since the data were not normally distributed
FactorDiabetic men with RE Mean ± SD (95% CI; n)Diabetic men without RE Mean ± SD (95% CI; n)Controls (without RE) Mean ± SD (95% CI; n)p-values
  1. a

    HgbA1c: Hemoglobin A1C.

  2. b

    HDL cholesterol: High-density lipoprotein cholesterol.

  3. c

    CRP: C-reactive protein.

Age (years)48.6 ± 6.2 (43.8–53.3; 9)43.1 ± 8.0 (39.1–47.1; 18)46.4 ± 7.2 (42.8–49.9; 18)NSNS
Diabetes duration (years)20.1 ± 13.4 (7.8–32.5; 7)13.3 ± 10.3 (8.1–18.4; 18)NS
Insulin dependency (%)50% (n = 8)76% (n = 17)0%NS
Diabetic complication, eyes (laser)n = 0n = 2n = 0
Diabetic complication, kidneysn = 1n = 1n = 0
Diabetic complication, neuritisn = 1n = 1n = 0
Diabetic complication, foot woundsn = 1n = 0n = 0
β-Blockersn = 1n = 1n = 1
Thiazide diureticsn = 0n = 3n = 2
Calcium antagonistsn = 2n = 2n = 0
Body mass index (BMI)27.9 ± 4.7 (23.5–32.2; 7)29.0 ± 5.8 (26.1–31.9; 18)27.8 ± 6.3 (24.6–31.1; 17)NSNS
Blood pressure, systolic (mmHg)145 ± 20 (126–163; 7)144 ± 12 (138–150; 18)141 ± 14 (133–148; 15)NSNS
Blood pressure, diastolic (mmHg)82 ± 6 (76–87; 7)86 ± 7 (82–89; 18)87 ± 10 (82–92; 15)NSNS
Fasting glucose10.3 ± 4.3 (6.3–14.3; 7)11.4 ± 5.6 (8.6–14.1; 18)5.7 ± 0.5 (5.5–6.0; 17)NS0.03
HgbA1c (%)a7.5 ± 1.1 (6.5–8.5; 7)8.1 ± 1.6 (7.3–8.9; 18)5.4 ± 0.4 (5.2–5.6; 17)NS0.009
Fat-%23.3 ± 4.8 (18.8–27.8; 7)23.5 ± 6.7 (20.1–26.8; 18)23.5 ± 8.5 (19.1–27.8; 17)NSNS
Waist circumference101 ± 11 (91–111; 7)100 ± 14 (93–107; 18)99 ± 18 (90–109; 17)NSNS
HDL cholesterolb1.4 ± 0.5 (1.0–1.8; 7)1.3 ± 0.4 (1.1–1.5; 18)1.3 ± 0.3 (1.2–1.5; 17)NSNS
Triglycerides1.6 ± 1.2 (0.5–2.7; 7)1.8 ± 2.0 (0.8–2.7; 18)1.3 ± 0.6 (1.0–1.5; 17)NSNS
CRPc2.0 ± 2.0 (0.2–3.8; 7)2.0 ± 2.3 (0.9–3.2; 18)2.5 ± 2.5 (1.2–3.8; 17)NSNS
s-Testosterone (nmol/L)17.9 ± 5.6 (12.8–23.1; 7)16.1 ± 6.0 (13.1–19.1; 18)13.7 ± 4.6 (11.3–16.0; 17)NSNS
Ejaculate volume (mL)1.6 ± 1.6 (0.3–2.9; 9)2.6 ± 1.1 (1.8–3.4; 16)2.7 ± 1.8 (1.7–3.7; 15)NSNS

Diabetic key data and complications

The mean duration of DM was longer for DM patients with RE compared with DM patients in whom RE could not be demonstrated, this difference was not statistically significant (Table 2). RE was not significantly associated with insulin dependency or diabetic complications such as retinopathia, kidney disease, neuritis or foot wounds, however, as could be expected fasting glucose and HgbA1c were significantly elevated for the diabetic men (with and without RE) compared with controls (Table 2).

Obesity and metabolic factors

Fat-%, waist circumference, concentration of high-density lipoprotein (HDL) cholesterol and concentration of triglycerides were not significantly different between study and control groups (Table 2).


Seven of the nine diabetic men with RE filled out the questionnaire. These were all smokers or ex-smokers. Conversely, more than 1/3 of diabetic men without RE (7/18) and of controls (7/17), respectively, had never been smoking (p < 0.07). No relationship between RE, DM and alcohol intake could be demonstrated.

Self-evaluation of erectile function in relation to DM and retrograde ejaculation

Men with DM and RE scored significantly lower than controls (without DM) in 10 of 15 questions in the INH-15 self-evaluation test, reflecting a lower level of sexual satisfaction in the men with RE (Table 3). The sexual desire was not significantly lower for the men with RE (Q11–12), but they scored lower in questions reflecting erectile function (Q2–5 + Q15), and therefore also intercourse satisfaction (Q7–8), orgasmic function (Q9–10) and overall satisfaction (Q13) gave lower scores in the RE group. DM men with RE tended to score lower than DM men without RE, although the differences were not significant (Table 3).

Table 3. Erectile dysfunction in patients with diabetes mellitus (DM) with or without retrograde ejaculation (RE) compared to controls and evaluated using a validated self-evaluation scale. Mean score results were compared and p-values calculated with the unequal variance t-test. Questions all start with ‘Over the past 4 weeks’
QuestionsDM and RE  Mean ± SD (95%CI; n)DM but not RE  Mean ± SD (95%CI; n)Control  Mean ± SD (95%CI; n)p-value
  1. a

    Diabetic men with RE vs. diabetic men without RE.

  2. b

    Diabetic men with RE vs. controls without DM.

  3. SD: Standard deviation: CI: Confidence interval.

Number, total in the groups91818  
Q1. How often were you able to get an erection during sexual activity?4.8 ± 1.5 (3.6–5.9; 9)5.3 ± 1.6 (4.4–6.1; 17)5.9 ± 0.5 (5.7–6.1; 18)NSNS
Q2. When you had erections with sexual stimulation, how often were your erections hard enough for penetration?4.2 ± 1.6 (3.0–5.5; 9)5.2 ± 1.7 (4.4–6.1; 17)5.9 ± 0.3 (5.7–6.0; 18)NS0.02
Q3. When you attempted sexual intercourse, how often were you able to penetrate (enter) your partner?3.8 ± 2.3 (2.0–5.5; 9)4.7 ± 2.2 (3.6–5.8; 17)5.7 ± 1.2 (5.1–6.3; 18)NS0.04
Q4. During sexual intercourse, how often were you able to maintain your erection after you had penetrated (entered) your partner?3.7 ± 2.1 (2.1–5.3; 9)4.6 ± 2.2 (3.5–5.7; 17)5.7 ± 1.2 (5.1–6.3; 18)NS0.02
Q5. During sexual intercourse, how difficult was it to maintain your erection to completion of intercourse?3.8 ± 2.2 (2.1–5.5; 9)4.4 ± 2.3 (3.2–5.6; 17)5.7 ± 1.2 (5.1–6.3; 18)NS0.03
Q6. How many times have you attempted sexual intercourse?3.2 ± 1.8 (1.8–4.6; 9)3.3 ± 1.8 (2.4–4.2; 17)3.9 ± 1.1 (3.3–4.4; 17)NSNS
Q7. When you attempted sexual intercourse, how often was it satisfactory to you?3.2 ± 2.1 (1.6–4.8; 9)4.5 ± 2.3 (3.4–5.6; 18)5.6 ± 1.2 (5.0–6.2; 18)NS0.01
Q8. How much have you enjoyed sexual intercourse?3.1 ± 2.0 (1.5–4.8; 9)4.6 ± 2.1 (3.6–5.6; 18)5.3 ± 1.2 (4.7–5.9; 18)NS0.02
Q9. When you had sexual stimulation or intercourse, how often did you ejaculate?4.8 ± 1.4 (3.7–5.8; 9)5.5 ± 1.3 (4.9–6.2; 17)5.8 ± 0.4 (5.6–6.0; 18)NS0.05
Q10. When you had sexual stimulation or intercourse, how often did you have the feeling of orgasm or climax?4.4 ± 1.3 (3.4–5.5; 9)5.1 ± 1.6 (4.2–5.9; 18)5.9 ± 0.3 (5.7–6.1; 17)NS0.01
Q11. How often have you felt sexual desire?4.0 ± 1.1 (3.1–4.9; 9)4.3 ± 0.8 (3.9–4.8; 18)4.7 ± 0.8 (4.3–5.1; 16)NSNS
Q12. How would you rate your level of sexual desire?3.6 ± 0.7 (3.0–4.1; 9)3.7 ± 0.8 (3.3–4.1; 18)3.8 ± 0.7 (3.4–4.2; 17)NSNS
Q13. How satisfied have you been with your overall sex life?3.0 ± 1.4 (1.9–4.1; 9)3.8 ± 1.1 (3.3–4.3; 18)4.3 ± 1.0 (3.8–4.8; 18)NS0.03
Q14. How satisfied have you been with your sexual relationship with your partner?3.9 ± 0.9 (3.0–4.7; 7)4.1 ± 1.1 (3.5–4.6; 16)4.6 ± 0.6 (4.3–4.9; 17)NSNS
Q15. How do you rate your confidence that you could get and keep an erection?3.1 ± 1.4 (2.1–4.2; 9)4.2 ± 0.9 (3.8–4.7; 18)4.5 ± 0.5 (4.3–4.8; 17)0.050.01

Use of pharmaceuticals with impotence potential

β-Blockers, thiazides and calcium antagonists are all antihypertensive pharmaceuticals with the potential to affect the erectile function and thus bias the answers to the questionnaire. In each of the three groups (DM with RE, DM without RE and controls) one man was prescribed β-blockers, and while none of the diabetic men with RE was prescribed thiazide diuretics, three of the diabetic men without RE and two of the controls took thiazide diuretics. Calcium antagonists were prescribed two of the diabetic men with RE and two of the diabetic men without RE (Table 2).


This study demonstrates that retrograde ejaculation is a frequent condition in men with DM, and this cannot be explained by use of antihypertensive drugs. Although urine is toxic to spermatozoa (Aust et al., 2008), fertility treatment with rapidly purified spermatozoa retrieved from post-ejaculatory urine should be the first choice treatment to couples presenting with retrograde ejaculation, which is not medically curable.

No consensus as to defining retrograde ejaculation in form of number of spermatozoa in post-ejaculatory urine or per cent of ejaculated spermatozoa in urine (number of spermatozoa in post-ejaculatory urine over total spermatozoa retrieved from the ejaculate and the post-ejaculatory urine) exists. Indeed, such a definition would be difficult to implement as the fraction of antegrade ejaculated spermatozoa remaining in the urethra post ejaculation may vary between and within individuals. However, we found very few spermatozoa in the post-ejaculatory urine of those study participants defined as antegrade ejaculating.

Men suffering from RE had in average been diagnosed with DM for 20 years, whereas the group of diabetic men without RE in average had had the diagnosis for 13 years. Although this study lacks power to conclude that RE is a long-term effect of DM, such an association may be suggested.

The INH-15 self-evaluation scale used in the study reflects erectile function, intercourse satisfaction, orgasmic function, sexual desire and overall satisfaction. In all aspects, men with DM scored lower than controls and men with RE and DM scored lower than diabetic men without RE, although these differences were not statistically significant (Table 3). One explanation to this might be that diabetic men may have more extensive vascular atherosclerosis/calcifications and neuropathia, increasing the risk of impotence/erectile dysfunction (Creager et al., 2003). This might be clarified in future studies with CT scans. Although calcium channel blockers apparently have no negative influence on erectile function (Moulik & Hardy, 2003; Baumhäkel et al., 2011), other antihypertensive pharmaceuticals such as β-blockers or thiazide diuretics have been shown to increase the risk of impotence (Fovaeus et al., 1987). This, however, does not seem to explain the higher frequency of erectile dysfunction in the diabetic men in this study.

Despite the size of the Diabetes Database (including nearly 5 000 diabetic patients and 5 000 controls), great effort allowed us to enrol only 27 diabetic men and 18 controls. Approximately half of the persons in the Database were women and further, a considerable number were above their fertile age. Our target group was narrowed to persons 30–55 years of age, consequently excluding the major part of men in the database. Although the relative modest material constitutes a major weakness for the study, the selection of participants was performed based on geographical criteria and thus was unbiased. However, expanded studies are needed to confirm the present findings.

Most importantly, the study included a sufficient number of participants to conclude that retrograde ejaculation is frequent in men suffering from DM, and that retrograde ejaculating men are less sexually satisfied compared with antegrade ejaculating non-diabetic men. Fertility personnel should be aware that a yearlong diagnosis of DM and aspermia or very small ejaculate volumes calls for post-ejaculatory urine analysis.


The authors thank Professor Erik Parner, Department of Public Health, Section for Biostatistics, Aarhus University, for statistical assistance.


The authors declare no conflict of interests.

Author contributions

The study was designed by J.F. in cooperation with I.B. and A.H. All authors contributed to writing the manuscript based on a draft written by J.F. and M.D.K.