Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Somatosensory evoked potentials assess the efficacy of circumcision for premature ejaculation


To assess the efficacy and mechanism of circumcision in the treatment of premature ejaculation (PE) with redundant prepuce, we enrolled a total of 81 PE patients who received circumcision. The patients’ ejaculatory ability and sexual performances were evaluated before and after circumcision by using questionnaires (Intravaginal ejaculation latency time (IELT), Chinese Index of PE with 5 questions (CIPE-5) and International Index of Erectile function- 5 (IIEF-5)). Furthermore, somatosensory evoked potentials (SEPs) including dorsal nerve (DNSEP) and glans penis (GPSEP) of the patients were also measured. The mean IELTs of preoperation and post operation were 1.10±0.55 and 2.48±2.03 min, respectively (P<0.001). In addition, the geometric mean IELT after operation was 2.16 min, compared with the baseline 1.07 min before the operation, the fold increase of the IELT was 2.02. Compared with the uncircumcised status, scores of CIPE-5 showed a significant increase after circumcision (P<0.001). The mean latencies (and amplitudes) of GPSEP and DNSEP were 38.1±4.0 ms (3.0±1.9 uV) and 40.5±3.4 ms (2.8±1.6 uV) before circumcision, respectively; and 42.8±3.3 ms (2.8±1.6 uV) and 40.5±4.1 ms (2.4±1.2 uV) in the follow-up end point after circumcision. Only the latencies of GPSEP showed significant prolongation before and after circumcision (P<0.001). The ejaculation time improvement after circumcision is so small, and equal to placebo response, therefore it could not be interpreted as a therapeutic method in men with PE.


Premature ejaculation (PE) is the most common male sexual complaint with a prevalence rate of 20–30%.1 Although PE has been defined in different ways by professional organizations and individuals, it commonly comprises the following three essential criteria: (i) a short ejaculatory latency; (ii) a perceived lack of control or inability to delay ejaculation, both related to the broader construct of perceived self-efficacy; and (iii) distress and interpersonal difficulty to the individual and/or partner.2 On the basis of the clinical evidence, the International Society for Sexual Medicine (ISSM) proposed a unified definition of PE: it is a male sexual dysfunction characterized by ejaculation that always or nearly always occurs prior to or within about 1 min of vaginal penetration from the first sexual experience (lifelong premature ejaculation), or a clinically significant reduction in latency time, often to about 3 min or less (acquired premature ejaculation); the inability to delay ejaculation on all or nearly all vaginal penetrations; and negative personal consequences, such as distress, bother, frustration and/or the avoidance of sexual intimacy.3 Currently, several treatment options are available, such as behavior therapy, selective serotonin reuptake inhibitor, topical application of local anesthetics and surgery including male circumcision and selective resection of dorsal nerves of penis.4

Male circumcision is a common procedure, with the benefits of lower sexually transmitted infection and penile carcinoma risk.5 In contrast to the established medical benefits, there are continued controversies in the relationship between circumcision and PE.6, 7, 8, 9, 10, 11, 12 There is a hypothesis with conflicting outcome of studies that glans hypersensitivity and hyperexcitability is one of the PE pathomechanisms.13, 14 On the basis of the belief that circumcision could decrease glans sensitivity, it has been used by some clinicians to treat PE.7, 8, 12 However, other strong evidence acquired from two successive populations in five countries found no difference between circumcised and uncircumcised men.15

To evaluate whether the circumcision could improve PE patients’ ejaculatory functions, we took a comprehensive clinical study using Intravaginal ejaculation latency time (IELT) and Chinese Index of PE with 5 questions (CIPE-5).16 To our knowledge, we are also the first to use somatosensory evoked potentials (SEPs) to estimate the changes of glans sensitivity in pre-circumcised and post-circumcised patients. In addition, we investigated the relationship between IELT and SEP parameter changes after circumcision.

Materials and methods


In this clinical trial, we registered 81 patients with PE, which was defined as previously described by ISSM,3 and redundant foreskin attending to the Department of Andrology of our hospital from June 2013 to September 2014. The patients with the redundant prepuce were included, and their IELT was less than 3 min when they had a stable, heterosexual relationship with a sexual partner for at least 6 months. Exclusion criteria were the following: history or physical examination of any other sexual disorder or genital abnormalities, such as erectile dysfunction and chronic prostatitis; any chronic psychiatric or physical illness, such as seizure and depression; alcohol or drug abuse; systemic disease (diabetes, hypogonadism, hyperthyroidism and so on). We collected the general information and complete sexual history, physical examinations, IELT, CIPE-5 and International Index of Erectile function- 5 (IIEF-5)17 before circumcision and in the last following time. IELT was measured with a stopwatch for a 4-week baseline period during which patients were asked to have sexual intercourse at least four times. Each man recorded the IELT during sexual activity at four different times, both in a baseline and after the operation. In addition, SEPs were also performed at preoperation and the first year after surgery. This study was ethically approved by the Declaration of Helsinki and the institutional review boards of Nanjing Drum Tower Hospital (GLYY20130525). Informed and operation consent was signed by every patient before the trial.


All the operative procedures were performed by the same experienced surgeon specialized in Andrology in our hospital. Circumcision was performed on a standardized process under local nerve block anesthesia. We grasped the prepuce at the 6, 10 and 14 o’clock positions by three mosquito clamps, and pulled it over the balanus. Then a Kocher clamp was applied just over the glans, taking care to avoid clamping or damaging the glans. The prepuce was pulled a little tightly, and subsequently, the redundant foreskin was cut along the edges of the Kocher clamp. Bleeders were electrically coagulated. The skin and mucosal incisions were approximated using interrupted 4-0 Vicryl rapide and then pressure-bandaged. Patients were followed on days 2 and 7, 4 weeks and 1 year after circumcision. Patients were instructed to refrain from sexual intercourse for at least 4 weeks after the surgery. At the last follow-up visit (1 year after circumcision), IELT, CIPE, IIEF and SEPs of glans penis (GPSEP) and dorsal nerve (DPSEP) were evaluated again.

An electromyography and evoked potential equipment (Keypont 4, Alpine BioMed Aps, Copenhagen, Denmark) was used to measure GPSEP and DNSEP, with two different electrical stimulation places. To conduct the DNSEP, we put the stimuli rings on the subcoronal region and the distance between the anode and the cathode was approximately 2 cm. Electrical stimuli were applied to the penile shaft by ring electrodes with an intensity of the penile shaft sensory threshold value as described in our previous studies.18 In GPSEP, a pair of round surface electrodes were placed at the glans penis. We delivered the same electric power (10.0 mA) on the round surface electrodes among all the patients. Recording electrodes consisted of two needles. On the basis of international 10-20 electrode placement protocol, the active recording electrode was placed on the scalp 2 cm behind the Cz electroencephalographic and the reference needle electrode was fixed in the Fpz recording site.19 The right forearm was used for grounding. Responses were recorded on the skin with an impedance of <5 kΩ. The duration of stimuli was 1.0 ms and the frequency was 3 per second. Both the SEPs were performed twice, each time was conducted by averaging 200 cortical potentials, and then merged the effective image, which was to eliminate the test-to test variability. The latency was obtained form the origin of the stimulus to the first positive peak cerebral response and the amplitude was determined from the first positive peak to the first negative peak points.


The Kolmogorov–Smirnov test was used to test the normal distribution. Continuous variables normally distributed were expressed as mean±standard deviation and compared by comparative t-test. Pearson correlation was estimated for two continuous variables. According to method of Waldinger et al.,20 the geometric mean IELT was calculated by transforming each separate IELT measurement of each man into the natural logarithmic IELT outcomes; then, we calculated the mean of each four logarithm sets, and we got the mean ln (IELT) for every person; consequently, we acquired the geometric mean IELT outcomes of the group before and after circumcision by averaging the geometric mean IELT of each man. Meanwhile, the fold increase of the IELT was computed by the geometric mean IELT value after circumcision divided by the value at baseline.20 Statistical analyses were performed using SPSS V16.0 (SPSS, Chicago, IL, USA). A two-sided P<0.05 was regarded as statistically significant.


A total of 81 patients were enrolled in our study from June 2013 to September 2014. A detailed overview of the included patients’ characteristics was summarized in Table 1. The mean age of the studied population was 29.8±5.6 years (range, 22–54 years), and the mean (range) height was 171.8±5.4 (160–183) cm. Among the patients, 26 were acquired premature ejaculation, the other 55 were lifelong premature ejaculation. Before the circumcision, the mean IELT of the acquired premature ejaculation patients was 1.58±0.27 min, while that of the lifelong premature ejaculation patients was 0.88±0.50 min.

Table 1 Characteristics of the study population and differences in parameters between preoperation and post operation among IELT, CIPE, IIEF and SEPs

There were no severe side effects after circumcision, and all the patients fully recovered and could have sexual intercourse with their parteners 3 months after the surgery. There was no statistical difference in IIEF-5 scores of the patients between preoperation and post operation. Figure 1 shows the IELT distribution of 81 PE men who were treated with circumcision. The preoperation and post operation IELT distributions are both skewed. The mean IELTs of preoperation and post operation were 1.10±0.55 and 2.48±2.03 min, respectively. When compared with the preoperation IELT values, subjects achieved statistically significant prolongation of IELTs following circumcision (P<0.001). However, the the mean increase in post-circumcision ejaculation time was only 1.38 min. In addition, the geometric mean IELT after operation was 2.16 min, compared with the baseline 1.07 min before the operation; the fold increase of the IELT was 2.02 (Table 1). In other words, the 2.02-fold increase was 102% IELT increase through circumcision intervention. At the follow-up end point, patients who had undergone circumcision showed a significant increase over the preoperation statue in terms of CIPE-5 score (P<0.001, Table 1). Subsequently, to elucidate a detailed impact of CIPE-5 with circumcision, we compared the changes of each question in CIPE-5 before and after excision of redundant foreskin. As summarized in Figure 2, the scores of Q1 (2.2±0.9 vs 3.3±1.4), Q2 (1.4±0.7 vs 2.5±1.4), Q3 (1.4±0.5 vs 2.3±1.2) and Q4 (1.4±0.5 vs 2.2±1.2) were significantly higher after circumcision than preoperation (P<0.001 in all). Unfortunately, the score of Q5 (3.8±1.3 vs 4.1±1.2) denoting the satisfaction of sexual partner had no substantial improvement (P=0.35).

Figure 1

Intravaginal ejaculation latency time (IELT) distribution of 81 men. Before circumcision (a): Mean IELT=1.10 min; geometric mean IELT=1.07 min; median IELT=0.90 min. After circumcision (b): Mean IELT=2.48 min; geometric mean IELT=2.16 min; median IELT=1.95 min.

Figure 2

Comparison of each question of CIPE-5 between pre- and post-circumcision one by one. *P<0.001. Q1: How long from intromission to ejaculation? Q2: Can you prolong the intercourse time? Q3: Your sexual satisfaction? Q4: Your partner's sexual satisfaction? Q5: Do you feel anxiety, depression or stress in sexual activity?

Consistent with previous research, we obtained the M-shaped SEPs (GPSEP and DNSEP) in all the subjects. The typical GPSEP waveforms of one patient in both preoperation and post operation were presented in Figure 3. Moreover, we found that the latencies of SEPs were normally distributed, but the amplitudes were abnormally distributed. Figure 3 illustrated the intuitionistic distribution of latencies and amplitudes of SEPs in each subjects. As shown in Figure 4, the latencies of GPSEP in post operation seem to be longer than in preoperation. The mean latencies (and amplitudes) of GPSEP and DNSEP were 38.1±4.0 ms (3.0±1.9 uV) and 40.5±3.4 ms (2.8±1.6 uV) before circumcision, respectively; and 42.8±3.3 ms (2.8±1.6 uV) and 40.5±4.1ms (2.4±1.2 uV) in the follow-up end point after excision of redundant foreskin. Compared with preoperation prepuce statue, circumcision could prolong 2.4 ms of the mean latencies of GPSEP (P<0.001), while no significant change of DNSEP (P=0.992) at the last follow-up pointwas detected. Next, we accessed the amplitudes of SEPs in preoperation and post operation. There were no significant difference in the mean amplitudes of GPSEP (P=0.568) or DNSEP (P=0.118) in preoperation and post operation. The changes of the latencies of GPSEP were weakly positively correlated with the increment of IELT (r=0.357), yet it was not statistically meaningful (P=0.104). However, there was a strong and statistically meaningful positive correlation between the changes of GPSEP latencies and the CPIE-5 scores (r=0.449, P<0.001).

Figure 3

The result of somatosensory evoked potential of glans (GPSEP). Before circumcision (a); after circumcision (b).

Figure 4

The distribution of GPSEP and DNSEP. The x axis is latencies (ms) and the y axis is amplitudes (uV).


PE is a physiological or psychological disorder characterized by a shortened ejaculatory latency time, poor control over ejaculation, low satisfaction with sexual intercourse and distress regarding the condition and has plagued many males throughout the world. Male circumcision is the most common surgical procedure worldwide for medical or religious reasons. It has been hypothesized that prepuce was a risk factor of PE and circumcision could be a kind of treatment of PE because the intervention may result in decreased penile sensitivity to tactile and erogenous stimuli.

In fact, the effect of circumcision on ejaculation has been studied by some researchers. However, the results are still controversial. For the reason that circumcision could improve IELT, some authors recommended that it was a selective treatment of PE.7, 8 Contrary to the results of the aforementioned studies, Tang et al.21 and O’Hara et al.9 presented that the males became even more likely to have PE after circumcision. Meanwhile, some other researchers reported that circumcision did not significantly affect IELT.15 In our study, we evaluated the impact of male circumcision on the changes of IELT and CIPE-5 score, in conjunction with neurophysiological tests.

Our results indicated that IELT of the circumcised patients became a little longer after the first year following surgery. This is equal to saying that the ejaculation time of PE patients seems to improve after circumcision. However, it is necessary to mention that the average IELT of the patients with PE and prepuce is just only 2.48 min, which is still less than the international IELT critical value of PE, that is, 3 min. In addition, applying the geometric mean IELT, we found that the fold increase of the IELT was 2.02. This is similar to the fold increase of the placebo response (2.04).22 Therefore, based on the IELT improvement, the delaying effect is similar to a placebo response. Besides, we used CIPE-5 score to further assess the impact of circumcision on PE. It is consistent with the above result that the score of Question 1 denoting the IELT precisely is slightly increased at the end point of follow-up. The self-control ability in prolonging sexual intercourse involved in Question 2 is ameliorated after circumcision. Question 3 indicates sexual satisfaction of the patient himself, which is improved after circumcision. On one hand, sexual satisfaction is derived from the better ability in prolonging sexual intercourse. On the other hand, numerous studies have suggested that the self confidence obtained through circumcision may also increase their sexual satisfaction.17 Furthermore, the present study confirms previous findings that circumcision can enhance the satisfaction of sexual partners by analyzing Question 4. Riess et al.23 reported that circumcision status is an important factor for women’s sexual decision-making, including partner choice and condom use because they perceived circumcised men as cleaner, carrying fewer diseases and taking more time to reach ejaculation. Remarkably, Question 5 reflecting patient’s mood has not significantly changed between uncircumcised and circumcised status. This further indirectly supports for the fact that circumcision just has a placebo effect in the male. We have made the correlative analysis between changes in IELT and CPIE-5 score, and found a positive correlation. This demonstrates that IELT is the most important indicator for PE assessment.

In addition, we used DNSEP and GPSEP tests to explore the mechanism of the relationship between circumcision and PE. The dorsal nerve of the penis is composed of two distinct populations of axons, one to innervate the glans penis and the other to innervate the penile shaft and anterior urethra.24 DNSEP is the objective neurophysiological data that evaluate the complete, peripheral and central dorsal nerve afferent pathway of penis, while GPSEP is a modification in DNSEP by sending stimuli at the glans penis.13 Our results showed that the latencies of GPSEP were remarkably delayed, whereas the amplitudes had no differences after circumcision. Meanwhile, no significant changes were seen about the latencies and amplitudes of DNSEP between preoperation and post operation. As we know, the glans penis of the patient with redundant prepuce or capistration was hypersensitive and hyperexcitable to external stimuli. After circumcision, the sensitivity of glans penis increased first, then declined and finally reached a stable status over time, consequently, the keratinized glans penis can result in decreased sensitivity.25 However, circumcision could not remove most of the sensitive parts of penis, such as the penile shaft skin and anterior urethra. Therefore, the latencies of GPSEP rather than DNSEP observably prolonged after circumcision. Correspondingly, IELT only had a little longer. Interestingly, Zhang et al.26 performed selective resection of the dorsal nerve of penis rather than only circumcision for PE patients, and found that IELT significantly increased from 1.1±0.9 min to 3.8±3.1 min after surgery. This phenomenon confirms our above hypothesis.

According to the above results, we suggest that the ejaculation time improvement after circumcision is so small, and equal to placebo response, so it could not be interpreted as a therapeutic method in men with PE.

The major limitation is that we only assessed the SEPs of the PE patients before and after circumcision, and did not have the data of DNSEP and GPSEP of normal controls who have measured their IELT with a stopwatch in this study.


Circumcision can provide a little improvement to PE patient with redundant prepuce, but its effect is similar to the placebo response. The major mechanism is that exposing glans by excising the redundant prepuce decreases the sensitivity of penile glans after a period of time. Nevertheless, only circumcision could not be a therapeutic method for PE patients.


  1. 1

    Hatzimouratidis K, Amar E, Eardley I, Giuliano F, Hatzichristou D, Montorsi F et al. Guidelines on male sexual dysfunction: erectile dysfunction and premature ejaculation. Eur Urol 2010; 57: 804–814.

    Article  Google Scholar 

  2. 2

    McMahon CG, Althof SE, Waldinger MD, Porst H, Dean J, Sharlip ID et al. An evidence-based definition of lifelong premature ejaculation: report of the International Society for Sexual Medicine (ISSM) ad hoc committee for the definition of premature ejaculation. J Sex Med 2008; 5: 1590–1606.

    Article  Google Scholar 

  3. 3

    Serefoglu EC, McMahon CG, Waldinger MD, Althof SE, Shindel A, Adaikan G et al. An evidence-based unified definition of lifelong and acquired premature ejaculation: report of the second international society for sexual medicine ad hoc committee for the definition of premature ejaculation. Sex Med 2014; 2: 41–59.

    Article  Google Scholar 

  4. 4

    Serefoglu EC, Saitz TR . New insights on premature ejaculation: a review of definition, classification, prevalence and treatment. Asian J Androl 2012; 14: 822–829.

    Article  Google Scholar 

  5. 5

    Laumann EO, Masi CM, Zuckerman EW . Circumcision in the United States. Prevalence, prophylactic effects, and sexual practice. JAMA 1997; 277: 1052–1057.

    CAS  Article  Google Scholar 

  6. 6

    Alp BF, Uguz S, Malkoc E, Ates F, Dursun F, Okcelik S et al. Does circumcision have a relationship with ejaculation time? Premature ejaculation evaluated using new diagnostic tools. Int J Impot Res 2014; 26: 121–123.

    CAS  Article  Google Scholar 

  7. 7

    Senkul T, Iseri C, Sen B, Karademlr K, Saracoglu F, Erden D . Circumcision in adults:effect on sexual function. Urology 2004; 63: 155–158.

    Article  Google Scholar 

  8. 8

    Namavar MR, Robati B . Removal of foreskin remnants in circumcised adults for treatment of premature ejaculation. Urol Ann 2011; 3: 87–92.

    Article  Google Scholar 

  9. 9

    O’Hara K, O’Hara J . The effect of male circumcision on the sexual enjoyment of the female partner. BJU Int 1999; 83: 79–84.

    Article  Google Scholar 

  10. 10

    Payne K, Thaler L, Kukkonen T, Carrier S, Binik Y . Sensation and sexual arousal in circumcised and uncircumcised men. J Sex Med 2007; 4: 667–674.

    Article  Google Scholar 

  11. 11

    Krieger JN, Mehta SD, Bailey RC, Agot K, Ndinya-Achola JO, Parker C et al. Adult male circumcision: effects on sexual function and sexual satisfaction in Kisumu, Kenya. J Sex Med 2008; 5: 2610–2622.

    Article  Google Scholar 

  12. 12

    Zhang SJ, Zhao YM, Zheng SG, Xiao HW, He YS . [Correlation between premature ejaculation and redundant prepuce]. Zhonghua Nan Ke Xue 2006; 12: 225–227.

    CAS  PubMed  Google Scholar 

  13. 13

    Xin ZC, Choi YD, Rha KH, Choi HK . Somatosensory evoked potentials in patients with primary premature ejaculation. J Urol 1997; 158: 451–455.

    CAS  Article  Google Scholar 

  14. 14

    Xia JD, Zhou LH, Han YF, Chen Y, Wang R, Dai YT . A reassessment of penile sensory pathways and effects of prilocaine-lidocaine cream in primary premature ejaculation. Int J Impot Res 2014; 26: 186–190.

    Article  Google Scholar 

  15. 15

    Waldinger MD, Quinn P, Dilleen M, Mundayat R, Schweitzer DH, Boolell M . A multinational population survey of intravaginal ejaculation latency time. J Sex Med 2005; 4: 492–497.

    Article  Google Scholar 

  16. 16

    Yuan YM, Xin ZC, Jiang H, Guo YJ, Liu WJ, Tian L et al. Sexual function of premature ejaculation patients assayed with Chinese Index of Premature Ejaculation. Asian J Androl 2004; 6: 121–126.

    PubMed  Google Scholar 

  17. 17

    Rhoden EL, Teloken C, Sogari PR, Vargas Souto CA . The use of the simplified International Index of Erectile Function (IIEF-5) as a diagnostic tool to study the prevalence of erectile dysfunction. Int J Impot Res 2002; 14: 245–250.

    CAS  Article  Google Scholar 

  18. 18

    Xia JD, Han YF, Pan F, Zhou LH, Chen Y, Dai YT . Clinical characteristics and penile afferent neuronal function in patients with primary delayed ejaculation. Andrology 2013; 1: 787–792.

    Article  Google Scholar 

  19. 19

    Ramos-Arguelles F, Morales G, Egozeue S, Pabon RM, Alonso MT . Basic techniques of electroencephalography: principles and clinical applications. An Sist Sanit Navar 2009; 32: 69–82.

    Article  Google Scholar 

  20. 20

    Waldinger MD, Zwinderman AH, Olivier B, Schweitzer DH . Geometric mean IELT and premature ejaculation: appropriate statistics to avoid overestimation of treatment efficacy. J Sex Med 2008; 5: 492–499.

    Article  Google Scholar 

  21. 21

    Tang WS, Khoo EM . Prevalence and correlates of premature ejaculation in a primary care setting: a preliminary cross-sectional study. J Sex Med 2011; 8: 2071–2078.

    Article  Google Scholar 

  22. 22

    Dinsmore WW, Hackett G, Goldmeier D, Waldinger M, Dean J, Wright P et al. Topical eutectic mixture for premature ejaculation (TEMPE): a novel aerosol-delivery form of lidocaine-prilocaine for treating premature ejaculation. BJU Int 2007; 99: 369–375.

    CAS  Article  Google Scholar 

  23. 23

    Riess TH, Achieng MM, Bailey RC . Women's beliefs about male circumcision, HIV prevention, and sexual behaviors in Kisumu, Kenya. PloS One 2014; 9: e97748.

    Article  Google Scholar 

  24. 24

    Yang C, Bradley WE . Peripheral distribution of the human dorsal nerve of the penis. J Urol 1998; 159: 1912–1917.

    CAS  Article  Google Scholar 

  25. 25

    Fink KS, Carson CC, DeVellis RF . Adult circumcision outcomes study: effect on erectile function, penile sensitivity, sexual activity and satisfaction. J Urol 2002; 167: 2113–2116.

    Article  Google Scholar 

  26. 26

    Zhang GX, Yu LP, Bai WJ, Wang XF . Selective resection of dorsal nerves of penis for premature ejaculation. Int J Androl 2012; 35: 873–879.

    Article  Google Scholar 

Download references


The study was supported by the National Natural Science Foundation of China (Grant No. 81270694 and No. 81501245).

Author information



Corresponding authors

Correspondence to J-D Xia or Y-T Dai.

Ethics declarations

Competing interests

The authors declare no conflict of interest.



Chinese Index of Sexual Function for Premature Ejaculation with 5 questions (CIPE-5):16

Table 2

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Xia, JD., Jiang, HS., Zhu, LL. et al. Somatosensory evoked potentials assess the efficacy of circumcision for premature ejaculation. Int J Impot Res 28, 127–132 (2016).

Download citation

Further reading


Quick links