Relationship between premature ejaculation and genetic polymorphisms of the dopamine transporter gene (SLC6A3)


Mohammad Reza Safarinejad, PO Box 19395-1849, Tehran, Iran.


What’s known on the subject? and What does the study add?

The process of ejaculation is highly influenced by genetic and neurobiological factors. Central dopaminergic drugs facilitate ejaculation. Genetic variation in SLC6A3 may alter the expression of dopamine transporter gene (DAT). This study evaluated the associations between genetic polymorphisms of the DAT and PE.

A statistically significant association was observed between the presence of 9-repeat allele and 9/10 genotype and PE. The presence of 7-repeat allele had protective effect against PE.


• To investigate the possible relationships between premature ejaculation (PE) polymorphisms in the dopamine transporter (DAT) gene (SLC6A3, DAT1), which has a polymorphic 40 base pair (40 bp) variable number of tandem repeats (VNTR) sequence in the 3′-untranslated region (3′ VNTR).


• Cohorts of 270 Iranian men with PE and 266 age-matched healthy Iranian subjects were genotyped for the DAT1-VNTR polymorphism.


• The 10-repeat allele frequencies were similar in the control (90.2%) and patient groups (88.5%) (P = 0.8).

• A statistically significant association was observed between the presence of the nine-repeat allele and PE (chi-squared test = 4.346, odds ratio [OR] = 2.46, 95% confidence interval [CI] = 1.57–3.15, P = 0.026).

• The frequencies of the 9/10 genotype were also significantly higher in the PE patients than in normal controls (chi-squared test = 4.466, OR = 2.47, 95% CI = 1.52–3.21, P = 0.028). The presence of the seven-repeat allele had a protective effect against PE (chi-squared test = 2.324, OR = 0.62, 95% CI = 0.47–0.89, P = 0.034).


• The findings of the present study suggest that DAT1-VNTR polymorphisms resulting in higher dopamine concentrations were associated with vulnerability to PE.

• Further studies are needed to replicate these results and to evaluate the role of inconsistency in the DAT genes and how this affects the development of PE.


3′ untranslated region



9R allele

nine-repeat allele

10R allele

10-repeat allele




dopamine transporter gene


follicle-stimulating hormone


γ-aminobutyric acid


intravaginal ejaculatory latency time


luteinizing hormone


odds ratio


odds ratios


premature ejaculation




selective serotonin reuptake inhibitors


variable number tandem repeat.


Premature ejaculation (PE) is the most prevalent male sexual dysfunction, affecting up to 31% of men aged 18–59 years [1]. Although some psychological, behavioural and biological aetiologies have been proposed, the exact pathophysiology of PE has not yet been clearly understood. The process of ejaculation is greatly influenced by genetic [2] and neurobiological factors [3]. Familial predisposition to PE was first reported by Schapiro in 1943 [4]. Waldinger et al. [5] reported that 10 of 14 first-degree relatives of men with lifelong PE also experienced PE. As a result, the aetiology of PE is based on a genetic predisposition. More recently it has been shown that, men who carry the serotonin transporter gene-linked polymorphic region S/S, LG/LG or S/LG genotype have a higher risk of PE [6,7]. The most common pharmacotherapeutic agent currently used to treat PE is ‘off-label’ administration of selective serotonin reuptake inhibitors (SSRIs) such as paroxetine, fluoxetine, sertraline, citalopram and escitalopram [8–11]. The genotype of serotonin transporter contributes in unique ways to the outcome of PE treatment with SSRIs [12]. The positive response rates are significantly better for the LA/LA genotype of the serotonin transporter gene-linked polymorphic region than for S-allele carriers [12].

Many neurotransmitters are involved in the process of ejaculation, including 5-hydroxytryptamine (5-HT [serotonin]), oxytocin, noradrenaline, nitric oxide (NO), γ-aminobutyric acid (GABA) [13] and dopamine (DA) [14]. The role of the serotonergic system in genital organs is inhibitory, whereas DA is excitatory in genital organs and antagonizes the serotonergic system [15]. Central dopaminergic drugs facilitate ejaculation [16]. Thioridazine [17] and chlorprothixene [18] delayed ejaculation by blocking central DA receptors.

The DA transporter (DAT) is a presynaptic plasma membrane protein expressed solely in DA-synthesizing neurones [19]. It clears DA released in the synapses, by rapid reuptake of DA into the presynaptic nerve terminals [20]. Therefore, it plays a critical role in ending DA neurotransmission. The gene for the DAT SLC6A3 is located on chromosome 5 at p15.3, and contains 15 exons spanning about 50 kb [21]. Genetic variation in SLC6A3 could alter the expression of DAT gene (DAT1) (22). A 40-bp polymorphic variable number tandem repeat (VNTR) element is found in the 3′ untranslated region (3′ UTR) of the DAT1 and can vary from three to 12 repeats. The two most common alleles are the nine-repeat (9R) and the 10-repeat (10R) alleles. The common 10R allele is associated with an increased expression of the DAT [22,23], but the 9R allele has been associated with decreased levels of DAT expression [21]. Homozygosity for the 10R allele at DAT1 causes overactive DAT density in the brain. In other words, polymorphism of VNTR directly alters the DAT activity.

The human VNTR polymorphism has been extensively studied in some neurological and psychiatric disorders [24], and the variable number of VNTR polymorphism has been proposed to be associated with normal personal traits or psychoneurological disorders [25].

The present study aimed to assess the associations between genetic polymorphisms of the DAT gene and PE in a large Iranian sample. We hypothesized for the first time that genetic variants of the R9 allele would confer susceptibility to PE.


A total of 288 unrelated married men (mean age 31.4 ± 10.4 years) with PE were recruited through local advertisements for screening. Patients were recruited from a specialized sexual dysfunction clinic with a highly selective patient group. All patients had a diagnosis of lifelong PE (measured by stopwatch). Lifelong PE was defined as an intravaginal ejaculatory latency time (IELT) of less than 1 min in more than 90% of episodes. None of the patients had undergone previous PE treatment. The control group consisted of 281 healthy men (mean age 32.2 ± 11.2 years) (IELT > 1 min as measured by stopwatch) drawn from the same population. Study participants were employed (84%) and most had completed their high school education (83%). All participants underwent screening procedures, including a detailed medical history and physical and neurological examination. Baseline laboratory studies included a complete blood count, biochemical profile, sex hormones (luteinizing hormone [LH], follicle-stimulating hormone [FSH], testosterone, prolactin [PRL]), thyroid function studies and urine analysis. According to the Structured Clinical Interview for the DSM-IV [26], none of the participants had any neurological and current or lifetime axis I or axis II psychiatric disorders. Subjects with a family history of neurological or psychiatric disorder in first-degree relatives identified by history-taking were also excluded. Concerning medical history, whenever useful, we used further information obtained from family members or from medical records.

There was no history of alcohol abuse, illicit drug use, psychotropic or neuroleptic medications or drugs known to affect the brain DA system (such as bupropion) in either the patients or controls. The use of illicit drugs was also further screened by means of urine testing for illicit opioids, amphetamines, methamphetamines, benzodiazepines, cannabinoids and their derivatives.

The study was conducted with the approval of the ethics committee in accordance with the Helsinki Declaration. Of all PE patients and controls, 270 and 266, respectively, gave their informed consent and fulfilled the enrolment criteria. The two groups were not significantly different with regard to age, handedness or educational level.


Each subject provided 5 mL of EDTA blood, and genomic DNA was extracted from frozen blood using a standard salting-out procedure [27]. The VNTR polymorphism in DAT1 was genotyped using PCR method of Vandenbergh et al. [21] but with slight modifications [28]. The PCR reaction was carried out under standard conditions. In brief, 500 ng of DNA was amplified in 50 µL of buffer containing 1.5 mm MgCl2, 5 µL 10 × Tris-buffered saline, 1.0 U Taq polymerase and 0.2 mm each of deoxynucleotide triphosphates and both oligonucleotide primers, (DATVNTRF, 5′-TGT GGT GTA GGG AAC GGC CTG AG-3′; DATVNTRR, 5′-CTT CCT GGA GGT CAC GGC TCA AGG-3′). After the initial 5 min denaturation at 94 °C, the reactants were subjected to 35 thermal cycles consisting of 94 °C for 45 s, 67 °C for 60 s, 72 °C for 30 s, followed by 74 °C for 10 min for the final extension. The PCR products were electrophoresed on a 2% agarose gel, and visualized by ethidium bromide staining. Molecular weights of 320, 360, 400, 440, 480 and 520 bp corresponded with the six-, seven-, eight-, nine-, 10- and 11-copy alleles, respectively (Fig. 1).

Figure 1.

Analysis of the human DAT1 genetic polymorphisms for the 3′-VNTR genotypes by PCR and agarose gel electrophoresis. Lanes: M, 100 bp DNA ladder; 1, 9/9; 2, 6/10; 3, 7/7; 4, 7/9; 5, 9/10; 6, 10/11; 7, 7/10; 8, 10/10; 9, 8/10 genotype.


A series of logistic regression models were calculated, with the diagnosis of PE as the dependent variable. Frequencies of genotype and alleles of the DAT1-VNTR polymorphism between PE patients and control subjects were compared by chi-squared test. Association between VNTR polymorphism of the DAT1 and PE was determined with Pearson’s chi-squared statistics. To allocate comparisons of the different relationships determined, independently of the size of the sample, odds ratios (ORs) are also calculated here with their 95% CIs. Hardy–Weinberg equilibrium was calculated based on allele frequencies by the goodness-of-fit chi-squared test.

A two-tailed P < 0.05 was considered to indicate statistical significance. Statistical tests were computed with Statistical Package for the Social Science (SPSS) for Windows, release 14.0 (SPSS Inc., Chicago, IL, USA).


Demographic and clinical characteristics for patients with PE and controls are presented in Table 1. The allelic and genotypic distributions of the VNTR marker were compared between the patient and control groups (Table 2). The 10/10 genotype was the most common (77.8 and 80.8%, respectively), followed by the 9/10 (15.9 and 10.9%, respectively) genotype in both the patients with PE and controls. The genotype frequencies were consistent with Hardy–Weinberg equilibrium in control group. Regarding the VNTR polymorphism of the DAT1, 210 (77.8%) and 215 (80.8%) patients with PE and controls, respectively, were homozygous for the 10R allele (P = 0.7), one (0.4%) and one (0.4%) were homozygous for the 9R allele (P = 0.8), and one (0.4%) and one (0.4%) were homozygous for the R7 allele (P = 0.8). Evidence of association between PE and the VNTR marker was found under the allele-wise or genotype-wise model. The allele frequencies for DAT1 are shown in Table 2. Carriers of 9R polymorphism had significantly higher PE risk (Table 2). The 9R allele was more common in the patient group than in the control group in both allelic analysis (chi-squared test = 4.346, OR = 2.46, 95% CI = 1.57–3.15, P = 0.026) and genotypic analysis (chi-squared test = 4.466, OR = 2.47, 95% CI = 1.52–3.21, P = 0.028). This effect was observed only for individuals with the 9/10 genotype. Compared with the normal control group, the frequencies of the 7/10 genotype were significantly lower in the patient group (chi-squared test = 4.546, OR = 0.72, 95% CI = 0.49–0.91, P = 0.028). When subjects were grouped based on the presence of the 7R variant of VNTR, there was a significant association between the 10R allele and PE when the 7R allele was absent (P = 0.28). Patients with PE had statistically significant lower 7R alleles than control subjects. This effect was only observed for heterozygotic individuals with an OR of 0.62 (95% CI = 0.47–0.86), suggesting that 7R is protective factor against PE. No association was observed between 6R, 8R and 11R variants of VNTR polymorphism and PE.

Table 1.  Demographic and clinical characteristics of study and control groups
Variables valuePatients (n = 270)Controls (n = 266) [N (%) or mean (range)]P value
  1. TSH, thyroid-stimulating hormone; T4, thyroxine; T3, triiodothyronine; NS, not significant.

Age, years31.4 (20–43)32.2 (21–44)NS
Weight, kg78.4 (63–94)77.2 (62–93)NS
Marriage duration, years8.4 (2–14)8.2 (2–14)NS
Duration of PE, years8.4 (2–14)0 (0) 
IELT,s18 (6–29)187 (145–249) 
Education level   
 None0 (0)0 (0)NS
 Primary school15 (5.6)14 (5.3)NS
 High school157 (58.1)159 (59.8)NS
 Graduate98 (36.3)93 (35.0)NS
Occupational status   
 Employed227 (84.1)223 (83.8)NS
 Unemployed43 (15.9)43 (16.2)NS
 LH, mIU/mL8.1 ± 2.78.7 ± 2.6NS
 FSH, mIU/mL9.4 ± 2.49.6 ± 2.8NS
 Testosterone, nmol/L18.2 ± 4.117.9 ± 4.3NS
 PRL, pmol/L 386 ± 65 389 ± 71NS
 TSH, mIU/mL1.22 ± 0.51.21 ± 0.4NS
 Free T4, ng/dL1.2 ± 0.31.2 ± 0.4NS
 Free T3, ng/dL0.3 ± 0.10.4 ± 0.1NS
Table 2.  Distribution of the VNTR polymorphism of the DAT1 and allele frequency in patients with PE and control subjects
VNTR copyPE patients (n = 270)Controls (n = 266)Chi-squaredOR95% CIP-value
Total alleles        


In the present study, we tested for an association between DAT1 gene polymorphisms and PE. We found a strong association between the polymorphism in 9R and 7R genes and PE. There is no similar study for comparison. The present study showed a significant association between the DAT1 10R allele and PE only when the 7R allele was absent. Individuals with this genotype have low concentrations of DA because of increased DAT activity due to the presence of 10R allele as well as a reduced response of the 7R receptor to DA [29]. The 9R alleles have been associated with lower levels of DAT expression [21,22]. According to the type of genotype, individuals can be categorized into three groups: those with low concentrations of DA (7R allele + without the 9R allele), those with moderate concentrations of DA (the 7R allele + the 9R, or without the 7R allele + without the 9R allele), and those with high concentrations of dopamine (without the 7R allele + the 9R allele). Adrenergic, dopaminergic and serotonergic systems affect male sexual function [30]. Central dopaminergic or anti-serotonergic drugs facilitate ejaculation [16]. In addition, DA antagonizes the serotonergic system. Therefore, men with high concentrations of DA might be genetically predisposed to develop PE. Relationships between different DAT1 gene polymorphisms and a wide spectrum of neurological and psychiatric disorders, such as Parkinson’s disease, affective disorders, schizophrenia, drug abuse and attention deficit hyperactivity disorder, have already been reported [31]. The 10R allele is the most frequent among the DAT1-VNTR polymorphism [21]. The frequencies of the 10R allele vary among populations, from 52% in Greeks to 100% in South Americans [32]. The prevalence of the genotypes and allele frequencies examined in the present study (9R and 10R) was highly consistent with several reports from Asian countries [33–35]. Studies on Asian populations have shown that more than 90% of the Japanese, Mongolian and Korean populations have the 10R allele [33–35]. In the present study, the patients with PE were more likely to have the 9/10 genotype and 9R allele than the control group. It has been reported that genetic [2] and neurobiological factors [3] determine the degree of the IELT. This means that PE is present throughout life and becomes apparent upon beginning sexual activity [14]. Also a possible familial genetic vulnerability to PE has been reported in previous studies [36].

The present study has several strengths. Firstly, the findings were not due to sample stratification, because the control group was ethnically matched, and the allele frequencies were similar to the findings of previous studies in Asian populations. Secondly, the sample size of patients with PE, especially those with the 9R allele, was sufficiently large and therefore can be considered representative of all men with PE in Iran. Finally, the control group was age-matched and therefore the results are not likely be caused by a confounding risk factor other than the trait of PE.

In conclusion, to the best of our knowledge, this is the first study to evaluate the relationship between the DAT1-VNTR polymorphism and PE. The results of the present study suggest the possibility of an association between the DAT1 9R and 7R alleles and the development of PE. It would be interesting to investigate other possible functional variants of the DAT1-VNTR polymorphism in PE. Further studies are needed to identify the precise role of DAT1 gene in the pathophysiology of PE.


We would like to thank the subjects who graciously participated in this study. We would also like to thank Saba Safarinejad for her help in statistical analysis and acknowledge the efforts of Nayyer Shafiei in preparing this manuscript.


None declared.