Secondary Logo

Journal Logo

Original Research

Innervation of the Pelvic Floor Muscles: A Reappraisal for the Levator Ani Nerve

Wallner, Christian1; Maas, Cornelis P.3,4; Dabhoiwala, Noshir F.2; Lamers, Wouter H.1; DeRuiter, Marco C.3

Author Information
doi: 10.1097/01.AOG.0000228510.08019.77
  • Free

The innervation of the pelvic floor remains controversial. Many well-regarded medical texts and review articles state that the pelvic floor muscles are innervated both by the pudendal nerve and by direct branches of the third and fourth sacral motor nerve roots.1 Such direct branches are more prone to damage during childbirth or surgical interventions. The effect of a pudendal nerve blockade on pelvic floor muscle function was recently reported in this journal.2 The study tested whether the pudendal nerve does indeed innervate the levator ani muscle. A blockade with 10 mL lidocaine of the pudendal nerve of nulliparous asymptomatic women without anal or urinary incontinence decreased intravaginal pressure, increased the length of the urogenital hiatus, and decreased electromyography activity of the puborectalis muscle. The authors therefore concluded that these results provide strong evidence that the pudendal nerve innervates the levator ani muscle.2 We would like to challenge their conclusions. Direct branches from S3 to S5, which lie on the visceral side of the pelvic floor and which are completely independent of the pudendal nerve, have been unambiguously documented.3 In anatomical textbooks this direct innervation is termed the “nerve to the levator ani” or “levator ani nerve.”4,5 We hypothesized that the anatomical relationship between the levator ani nerve and the pudendal nerve in the area of the ischial spine is so close that a pudendal nerve blockade affects both the pudendal nerve and the levator ani nerve. We, therefore, investigated the topography of the levator ani nerve from the perspective of obstetricians and gynecologists. We measured the distance between the levator ani nerve and pudendal nerve in the area of the ischial spine and positioned a mock pudendal nerve block to assess whether such a block can also affect the levator ani nerve. To verify our anatomical conclusions, we also studied (immuno-)histochemically stained sections of human fetal pelves.


Ten pelves of female cadavers without signs of pelvic surgery were dissected. The dissections were completed in a timeframe of 3 months, September to November 2005. The women were all postmenopausal. The cadavers were preserved by injection into the femoral artery of embalming fluid, which consisted of a mixture of formaldehyde, ethanol, glycerin, phenol, K2SO4, Na2SO4, NaHCO3, NaNO3, and Na2SO3. All pelves were transected midsagittally. The distal rectum was carefully detached from the pelvic floor to expose the sacral plexus, the pudendal nerve, and the levator ani nerve. The pudendal nerve was also dissected by a gluteal approach. The topographical relationship between the pudendal nerve and the levator ani nerve in the area of the ischial spine was systematically studied. A mock transvaginal “pudendal blockade” was positioned in the specimens by an experienced clinician (C.P.M.) following guidelines of standard clinical practice.2 The needle was inserted transvaginally through the pelvic floor approximately 8 mm medial to the ischial spine to a depth of 1 cm. No anesthetic fluid was injected. After dissection, the seven measurements shown in Figure 1 were taken (results are presented as median and 95% confidence intervals).

Fig. 1.Wallner. Levator Ani Nerve. Obstet Gynecol 2006.

To further help characterize the anatomical relationship of the LAN and the pudendal nerve, we also studied (immuno-)histochemically stained sections of human fetal pelves. The fetuses were obtained after legal abortion, and the study was approved by the medical-ethical committee of the Leiden University Medical Center. Two female fetal pelves (female, 14 and 19 weeks of gestation) were stained for the presence of striated muscle tissue, using a monoclonal antibody directed against myosin heavy chain (clone A4.1025; Upstate Cell Signaling Solutions, Charlottesville, VA), and for the presence of nerve tissue, using a polyclonal antibody directed against neurofilament 68 kD (AB1983; Chemicon International, Temecula, CA), as well as with hematoxylin-azophloxin. To verify the data obtained from these two fetuses, we also undertook investigations on existing fetal pelvic sections from the collections of the departments of Anatomy & Embryology, Leiden University Medical Center, and from the Academic Medical Center, Amsterdam. These fetuses (11 females and 7 males, ranging from 65 mm crown-rump length [CRL, 10 weeks of gestation] to 260 mm CRL [27 weeks of gestation]) were stained with hematoxylin and either azophloxin or eosin. Male fetuses were used to show that the levator ani nerve is present in both sexes. Three-dimensional reconstructions were prepared using Amira 3.0 (TGS Template Graphics Software, San Diego, CA; available at:


The levator ani nerve originated from sacral foramina S3 and/or S4. Its trajectory is on the visceral surface of the coccygeal and levator ani muscles and was situated 45 mm (95% confidence interval [CI] 40–49 mm) lateral to the midsagittal plane at the level of the ischial spine, 45 mm (95% CI 35–50 mm) lateral to the tip of the coccyx, and 8 mm (95% CI 3–20 mm) medio-caudal to the ischial spine. The main branch of the levator ani nerve entered the levator ani muscle at the level of the ischial spine in four pelves. In six pelves it continued its course further on the visceral surface of the muscle in the direction of the pubic symphysis to enter the muscle in its puborectalis portion 48 mm (95% CI 36–60 mm) distal to the point of passage of the ischial spine and 40 mm (95% CI 18–60 mm) lateral to the midsagittal plane (Fig. 2). In all cases the levator ani nerve was covered by the pelvic parietal fascia.

Fig. 2.Wallner. Levator Ani Nerve. Obstet Gynecol 2006.

The pudendal nerve, by contrast, coursed behind the sacrospinous ligament and the overlying coccygeal muscle before passing around the ischial spine. In the area of the ischial spine, the distance between the pudendal nerve and levator ani nerve (positioned inferior and superior to the levator ani muscle, respectively) was 6 mm (95% CI 4–11 mm). The tip of the needle, inserted during the mock pudendal nerve blockade maneuver, effectively reached the pudendal nerve in all specimens. The distance between the levator ani nerve and the passage of the needle through the levator ani muscle was 5 mm (95% CI 1–8 mm). In the fetuses the close topographic relationship of the levator ani nerve and the pudendal nerve as found in the adult specimens could be confirmed and elegantly demonstrated by using nerve and striated muscle–specific immunohistochemistry (Fig. 3).

Fig. 3.Wallner. Levator Ani Nerve. Obstet Gynecol 2006.

The levator ani nerve, originating from the third and/ or fourth sacral foramen and innervating the levator ani muscle on its visceral side, was demonstrable in female as well as male fetal pelves (Fig. 4A). Figure 4B shows the course of the nerves in a female fetal pelvis in a three-dimensional reconstruction.

Fig. 4.Wallner. Levator Ani Nerve. Obstet Gynecol 2006.


Our study provides strong and objective evidence that both the levator ani nerve and the pudendal nerve can be simultaneously blocked by a transvaginal “pudendal nerve blockade.” We have documented that the levator ani nerve lies only 8 mm medio-caudal from the ischial spine and that the shortest distance between the pudendal nerve and the levator ani nerve at the level of the ischial spine is only 6 mm. This close relationship is also demonstrable in the fetal sections. Because the 95% CI of the distance between the penetrating needle and the levator ani nerve ranges from 1 mm to 8 mm, while the injection of 10 mL of fluid theoretically forms a sphere with a radius of approximately 13 mm, a pudendal blockade with 10 mL anesthetic that does not affect the levator ani nerve is virtually impossible. These data throw considerable doubt on an earlier report2 that innervation of the levator ani muscle by the pudendal nerve can be deduced from a transvaginal lidocaine blockage of the nerve near the ischial spine. Clearly, the physiological parameters, measured clinically in that study, demonstrated that the pudendal nerve was blocked. In agreement, the tip of the needle in the mock transvaginal “pudendal nerve block” that we performed effectively reached the pudendal nerve in all tries. Guaderrama et al2 in their discussion also mentioned that the transvaginal “pudendal nerve block” might affect both the direct branches of sacral nerve roots S3 and S4 and the pudendal nerve. However, they ruled out the possibility of double blockade by explaining that the amount of lidocaine is too small to affect the direct branches and the pudendal nerve simultaneously.

In contrast, our anatomical data reveal that the topographical relation of the levator ani nerve to the pudendal nerve is so close at the point of injection that even the small amount of lidocaine that is used reaches both nerves. Moreover, literature data show the distribution area of anesthetics administered through pudendal nerve blockade to be large. A study in which spread of anesthetics during transvaginal pudendal nerve blockade was visualized with X-rays showed that anesthetics spread widely. The X-rays showed diffusion retrogradely up to the sacral roots of origin of the pudendal nerve.6

A limitation of our analysis could be that we studied cadavers as opposed to live subjects. One might argue that the age of our cadavers may have influenced the thickness of the pelvic floor muscles, because the pelvic floor in the elderly is frequently less functional, and that our assessment of the distance between the levator ani nerve and the pudendal nerve in cadavers is therefore an underestimate. Because the pelvic floor, both in the fetus and the adult, consists of one layer of muscle bundles only, we think that the possible differences in thickness in the pelvic floor are marginal. Also, although embalming of cadavers can cause distortion of anatomical relations, this distortion is probably minimal in the area of the ischial spine and the sacrospinous ligament because these are firm structures that resist the pressures created in embalming.

A second argument against our hypothesis could be that the levator ani muscle functions as an effective diaphragm and divides the space above and below the pelvic floor into separate compartments, which inhibits diffusion of lidocaine to the levator ani nerve. Only a study in which the diffusion pattern of the anesthetic is assessed could clarify whether the levator ani muscle does indeed prevent spreading of the anesthetic to the levator ani nerve.

Our findings confirm and extend earlier findings. Several studies argue that direct sacral branches are responsible for innervation of the levator ani muscle on its visceral side and that the pudendal nerve innervates the external anal sphincter only.3,7,8 This pattern is also seen in other primates9–11 and in rats.12 Furthermore, neurectomy of the levator ani nerve or the pudendal nerve in rats12 and squirrel monkeys9 shows that only denervation of the levator ani nerve affects the levator ani muscle, whereas denervation of the pudendal nerve affects the external anal sphincter. Neurostimulation of the levator ani nerve or the pudendal nerve in humans also shows muscle-specific effects on the levator ani and the external anal sphincter, respectively.7,13 Together, these studies convincingly demonstrate that the levator ani muscle and the external anal sphincter are innervated by different nerves.

The clinical implications of the innervation of the pelvic floor are evident: denervation of the pelvic floor muscles and the accompanying muscle dysfunction could cause urinary and/or anal incontinence, and pelvic organ prolapse. The concept of neural damage due to stretching and pressure during complicated vaginal childbirth or pelvic surgical procedures is more easily conceivable with the nerve to the levator ani muscle being located on the superior, visceral side of the pelvic floor. In contrast, the pudendal nerve is better protected in these situations because it lies on the inferior side of the pelvic floor, within Alcock’s canal. The same considerations apply to transvaginal sacrospinous ligament fixation for vault prolapse. One or more permanent stitches are put through the ligament by placing them 1 cm medial to the ischial spine.14 As can be concluded from our morphological findings, this procedure holds a risk of levator ani nerve entrapment or disruption, either immediately or as a delayed response to surgery. Documentation of recurrence rates of vaginal vault prolapse after sacrospinous ligament fixation is poor due to lack of follow-up studies with standardized parameters14,15 but is repeatedly mentioned in the literature as a clinical problem associated with this surgical procedure.3,14,15 Pelvic floor muscle denervation due to disruption of the levator ani nerve could explain these recurrent prolapses.

Given the clinical impact of a denervation of the levator ani muscles, all obstetricians, gynecologists, and pelvic surgeons should be aware of the levator ani nerve and its clinical anatomy. Although we fully realize that the functional assessment of the levator ani nerve is not trivial, our data clearly show that a pudendal blockade with an anesthetic is not the proper approach because it fails to distinguish between the pudendal and the levator ani nerves.


1. De Lancey JOL. Surgical anatomy of the female pelvis. In: Rock JA, Thompson JD, editors. Te Linde’s operative gynecology. 8th ed. Philadelphia (PA): Lippincott-Raven; 1997. p. 63–94.
2. Guaderrama N, Liu J, Nager CW, Pretorius DH, Sheean G, Kassab G, et al. Evidence for the innervation of pelvic floor muscles by the pudendal nerve. Obstet Gynecol 2005;106:774–81.
3. Barber MD, Bremer RE, Thor KB, Dolber PC, Kuehl TJ, Coates KW. Innervation of the female levator ani muscles. Am J Obstet Gynecol 2002;187:64–71.
4. Netter FH. Atlas of human anatomy. 2nd ed. East Hanover (NJ): Novartis; 1997.
5. Drake RL, Vogl W, Mitchell AWM. Gray’s anatomy for students. Philadelphia (PA): Elsevier; 2005.
6. Kobak AJ, Sadove MS, Mazeros WT. Anatomic studies of transvaginal regional anesthesia: roentgenographic visualization of neural pathways. Obstet Gynecol 1962;19:302–10.
7. Matzel KE, Schmidt RA, Tanagho EA. Neuroanatomy of the striated muscular anal continence mechanism: implications for the use of neurostimulation. Dis Colon Rectum 1990;33:666–73.
8. Juenemann KP, Lue TF, Schmidt RA, Tanagho EA. Clinical significance of sacral and pudendal nerve anatomy. J Urol 1988;139:74–80.
9. Pierce LM, Reyes M, Thor KB, Dolber PC, Bremer RE, Kuehl TJ, et al. Innervation of the levator ani muscles in the female squirrel monkey. Am J Obstet Gynecol 2003;188:1141–7.
10. Akita K, Sakamoto H, Sato T. Muscles of the pelvic outlet in the rhesus-monkey (Macaca mulatta) with special reference to nerve supply. Anat Rec 1995;241:273–83.
11. Sato K, Sato T. An anatomical investigation on the muscles of the pelvic outlet in crab-eating monkey (Macaca irus) and Japanese monkey (Macaca fuscata) with special reference to their nerve supply. Anat Anz 1987;163:425–38.
12. Bremer RE, Barber MD, Coates KW, Dolber PC, Thor KB. Innervation of the levator ani and coccygeus muscles of the female rat. Anat Rec A Discov Mol Cell Evol Biol 2003;275:1031–41.
13. Percy JP, Neill ME, Swash M, Parks AG. Electrophysiological study of motor nerve supply of pelvic floor. Lancet 1981;1:16–7.
14. Beer M, Kuhn A. Surgical techniques for vault prolapse: a review of the literature. Eur J Obstet Gynecol Reprod Biol 2005;119:144–55.
15. Sze EHM, Karram MM. Transvaginal repair of vault prolapse: a review. Obstet Gynecol 1997;89:466–75.
© 2006 The American College of Obstetricians and Gynecologists