Evaluation of anal sphincter resistance and distensibility in healthy controls using EndoFLIP ©


Address for Correspondence
Barry P. McMahon MSc, PhD, Department of Medical Physics & Clinical Engineering, Adelaide & Meath Hospital, Tallaght, Dublin 24, Ireland.
Tel: +353 1 4145898; fax: +353 1 4142510;
e-mail: barry.mcmahon@tcd.ie


Background  Ano-rectal disorders are common in the general population. Although they are not life threatening conditions, they do represent a social stigma and a reduced quality of life for the sufferer. The underlying physiology of muscle function contributing to ano-rectal competence is complex and there is room for a much better understanding so that treatments can improve.

Methods  A cylindrically shaped, liquid filled bag (12 cm long), mounted on a catheter was inserted into the anus and positioned straddling the ano-rectal region in 20 healthy volunteers (10 females). Series of volume-controlled distensions (40 mL min−1 to 40 mL) were carried out and data on 16 CSA at 5 mm apart and bag pressure were recorded. Provocative tests using squeeze and cough at bag volumes of 20, 30, and 40 mL were carried out.

Key Results  Ramp distension of the anal canal showed that the opening pressure for females (mean, 11 mmHg) was higher than for males (mean, 5 mmHg) (< 0.001). Geometric profile of the anal canal at low distension volumes showed narrow bands at proximal and distal ends of the anal profile and shortening of a middle narrow zone at higher volumes. Inter-individual differences were observed in the behavior of the proximal end and the distal end of the anal profile during squeeze.

Conclusions & Inferences  This distensibility technique provides an important new way of studying the anal canal and hence may have a role in testing sphincter competence in patients with disorders.


Fecal incontinence (FI) is characterized by involuntary loss of rectal contents through the anal canal. It is a complex, challenging and multifaceted clinical problem.1 Between 7.1% and 9.5% of the adult population under age 70 and 15.3% of the population over age 70 suffer from fecal incontinence.2 However, 12% of adult women are known to have FI.3

The mechanisms of defecation and continence, regulated by voluntary and reflex actions, are complex and depend on the interactions of several components. The ano-rectal region consists of two sphincters. The internal anal sphincter (IAS) which is composed of smooth muscle extends more than 1 cm above the external anal sphincter (EAS). The EAS is composed of striated muscle and under conscious control. Function in this area is also affected by the pelvic floor muscles which maintain the ano-rectal angle.4 The IAS is persistently tonically contracted and participates in the act of defecation by reflex relaxation in response to rectal distension known as the recto-anal inhibitory reflex.5 The EAS with the puborectalis muscles are mainly responsible for the squeezing pressure as a response to voluntary effort, or induced by increased intra-abdominal pressure during coughing and sneezing.6,7

The sphincteric response to the movement of stools within the anus is not tested by the usual methods of evaluating of ano-rectal disorders.8 Currently, anal canal pressure measurements using ano-rectal manometry are the most common means of assessment of sphincter function.9 Low resting and squeeze pressures generally reflect weakness of the IAS and EAS respectively.10,11 Despite this, the relationship between the function of the two sphincters is difficult to understand because there is a large degree of anatomical and functional overlap between the IAS and the EAS. Both sphincters contribute to maintain continence.12 It is often the view that it is impossible to demonstrate two separate zones of contraction of the sphincters using manometry.13,14

Some studies have been looking at the concept of distending sphincter regions as a better measure of its performance.15,16 The functional lumen imaging probe (FLIP) is a novel technique which has the ability to provide real-time images of the function of human gastrointestinal sphincter during distension. For example FLIP has been shown to be capable of measuring esophago-gastric junction competence.17 The ano-rectal sphincter normally responds to the presence of stools filling the rectum and distending the anal canal. Therefore our hypothesis is that FLIP can simulate this distension and measure the resulting pressure and geometric changes in the anal sphincter directly in response to distension and measure its resistance to this distension.

The aim of this study was to identify geometric and pressure changes in the ano-rectal sphincter region by means of FLIP and to infer the biomechanical action of the region based on measured results.

Materials and Methods


Twenty-one healthy volunteers 36.5 ± 2.5 years (mean ± SEM) were studied, 11 female (two multi-parous) with no history of gastroenterology disorders. None of the subjects were taking any medication and had no medical conditions or previous surgery related to bowel disorders. Ethics was approved under protocol VN2003 and the studies were carried out at Aalborg Hospital in Denmark. Informed consent was signed by all participants before the procedure.

Distensibility measurements

Measurement of anal canal distensibility was made using the FLIP based on the technique described previously18 as shown in Fig. 1A. In brief, the FLIP catheter had a 3 mm outer diameter. A non-compliant bag, 12 cm long, which could be filled to a volume of 40 mL, was mounted on the distal end of the probe. The cylindrical bag was designed to be inflated to 2.5 cm without contributing to the resistance to inflation and therefore not contributing to the compliance measurement of the sphincter as the fill volume was always less than the max volume (40 mL) of the bag. The detectable cross-sectional area (CSA) range was 16–450 mm2.

Figure 1.

 (A) Illustration showing the functional lumen imaging probe in position straddling the anal canal in the ano-rectal region. (B) Diagram for volunteer protocol starting with three ramp distensions (0–40 mL) at 40 mL min−1. (C) Then three further step distensions at 20, 30, and 40 mL, at each step volume the volunteers were asked to perform two provocative maneuvers (squeeze and cough).

The bag contained 17 ring electrodes spaced 5 mm apart for impedance planimetry measurements. As the bag was filled with a specially formulated conductive solution, the impedance across each segment caused a measured voltage change which was calibrated to represent the radial CSA of the bag at that position. The probe also contained a solid state pressure transducer mounted towards the distal end of the bag 2 mm from the last electrode. This transducer was capable of measuring the pressure of the liquid-filled bag during distension.

Sixteen CSA measurements between adjacent electrodes and the data from the pressure transducer representative of the intra-bag pressure were sampled at 10 Hz and stored in the data acquisition system of the EndoFLIP®, (Crospon, Galway, Ireland). The data were displayed in real time as a 7.5 cm cylinder of CSAs along its length reflecting diameters estimated from the 16 measured CSAs. The probe and the pressure transducers were precalibrated during the manufacturing process and required no additional calibration before use (CSA resolution 0.8 mm2, accuracy ±0.8 mm2; intra-bag pressure resolution 0.1 mmHg, accuracy ±0.8 mmHg).

Study protocol

The subjects lay on a bed and were positioned on their left side, with hips and knee flexed for the procedure. Before using the system, any air was removed from the EndoFLIP assembly by using an automated purge sequence and the intra-bag pressure was zeroed using a function on the system. The FLIP probe was inserted rectally until the bag straddled the anal canal as illustrated in Fig. 1A. The FLIP bag was inflated to a small volume (5–10 mL) so that three distal measurements were seen on the FLIP display to be increasing in diameter indicating that they had reached the rectum and the probe was held in this position by an assistant’s hand. The bag was deflated and the subjects were allowed 5–10 min to become accustomed to the probe before starting the procedure.

Probe positioning

Once 10 mL volume was filled into the FLIP bag a profile of the narrow zone was observed on the visual display. The 16 CSAs provided estimated diameter measurement of the narrow zone and the wider regions at either end. The probe was adjusted so that the three most distal wider diameter measurements emerged from the narrow zone. This wider region called the distal border was defined as the point at which the difference between two adjacent distal diameters was greater than 5 mm. The wider region proximal to the narrow zone, called the proximal border, was also defined as the point at which the difference between two adjacent proximal diameters was greater than 5 mm. At volumes between 10 and 20 mL the wider regions at the proximal and distal narrow zone ends was determined to have been reached when the differences between two adjacent measured diameters was 5 mm. At lower volumes (below 10 mL) the change for the proximal and distal wider regions was seen when the measured diameter was more than 1 mm.

The CSA and the corresponding distension pressure measurements were recorded during three ramp distensions where the bag was filled with the conductive liquid at 40 mL min−1 to a volume of 40 mL (Fig. 1B) and then a series of provocative maneuvers were carried out with the bag filled in steps to 20, 30, and 40 mL respectively (Fig. 1C). The maneuvers were squeezing, where the subject was asked to tighten the ano-rectal region as much as they could for 10–20 s, and coughing, when they were asked to spontaneously cough. There was a 1-min relaxation period between each maneuver. EndoFLIP measurements were monitored in real-time to ensure proper bag placement by directly visualizing the functional image on the EndoFLIP system during all stages of the protocol. If the bag was observed to move, resulting in less than three measurements in the rectum, the probe was repositioned and the measurement was repeated. In all cases the two narrow zones, at low volumes, were clearly identified as being at either end of the anal canal.

Statistical analysis

Statistical analysis was performed using a software package PASW statistics 18.0 (SPSS Inc., Chicago, IL, USA). As a first step, numeric values were analyzed for the presence of normal distribution. In cases of normal distribution, values are stated as mean ± SEM. Distensibility data were analyzed by using ancova on mean pressure and CSAs. The maneuver data were analyzed using two-way anova and included all individuals. The paired t-test was used to calculate the probability of the increase in the narrow zone during squeezing and coughing maneuvers.


All subjects underwent the planned procedure and tolerated the probe well for the duration of the study. There were no adverse incidents. Healthy controls (HC) no. 15 was excluded despite a successful study as it was discovered that she had a problem with constipation which was only revealed after the study was completed.

Ramp distensions

Initially during ramp distensions there was some variability in pressure readings but once the volume in the bag reached 10 mL the pressure changes were consistent and smooth. Fig. 2 shows a scatter plot of pressure in the bag against the narrowest CSA measured in the anal profile for males and females. The pressure at which the CSA starts to increase is indicated as the opening pressure on the plot. The opening pressure in females was twice that obtained in males (11 vs 5; < 0.001). Once opening had occurred the rate of CSA increase at the narrowest point in the anal profile relative to the rate of increase in bag pressure was 0.086 for females and 0.077 for males. The distensibility index was calculated as the narrowest CSA divided by the intra-bag pressure. There was a statistically significant difference between the distensibility index between males and females (> 0.001).

Figure 2.

 Distensibility plot (pressure plotted against the narrowest cross-sectional area) for ramp distensions at 40 mL min−1 for the male and female healthy controls. The distensibility index was 0.086 for females and 0.077 for males. The opening pressure for females was higher than for males.

The pattern of geometric changes in the ano-rectal region during the ramp distensions was very consistent across all 20 subjects. This is indicated in Fig. 3 where four examples of the geometric patterns represented by the functional images are shown at bag volumes from 10 to 40 mL. In these diagrams the top of each image is towards the rectum and the bottom of each image is outside the anal verge. Dest is the estimated diameter from the measured CSA. All subjects had a pattern of two narrow regions at either end of the anal profile at a volume of 10 mL developing into a cylindrical channel at some point between 10 and 20 mL and then by the time the bag filled to 30 mL a single narrow region formed in the centre of the profile.

Figure 3.

 Still images of geometric changes in the ano-rectal region during distension for four volunteers. HC indicates healthy controls, the number indicates the control number, (m) indicates male and (f) indicates female respectively. The top of each image (the distal end) is towards the rectum and the bottom (proximal end) is outside the anal verge. The x-axis represents the distended profile diameter. The y-axis on the left side represents the position of sixteen sensing electrodes across the functional lumen imaging probe bag with electrode 1 located toward the anal verge. The y-axis on the right side represents the changing in bag diameter in millimeters (Dmm) compared with the change in color in the anal profiles as appeared on the EndoFLIP display. Three main regions shown with the annotations A, B, & C were formed during volume distension in the anal canal. A & B illustrate narrow regions formed at low volumes in the distal and proximal ends and C illustrates a distinct narrowing towards the centre of the color profile at higher volumes.

During anal canal ramp distension from 0 to 40 mL, the narrow zone length was decreasing gradually because of the relaxation of the distal end of the anal canal (Fig. 4). Narrow zone length at 20 mL was 32 mm (mean) decreased to 10 mm (mean) at 40 mL volume.

Figure 4.

 Narrow zone length plotted against anal bag volume of 20, 30, and 40 mL. The length of the anal canal decreased when the probe distension volume increased.

Squeeze maneuvers

The geometric pattern was observed to have subtle but distinct changes during ‘squeeze’ maneuvers when compared with the ‘at rest’ state. Fig. 5A shows changes in the profile with a volume of 20 mL in the bag between at rest (blue outline) and squeeze (red line) for a representative subject. The points at where the greatest changes occurred were at the two ends of the anal canal. The CSA measurement at which the greatest difference occurred at either end of the canal were selected. In this case the more proximal (δ prox) was located at measurement detector D6 and the more distal (δ distal) at detector D14. Fig. 5B shows these diameter estimates (Dest) represented as their true CSA values. This indicates that during the squeeze maneuver, which correlates with the increased pressure shown by the blue tracing, the proximal anal canal CSA (δ prox – red line) initially reduced dramatically but within 4 s it has relaxed again to reach close to its normal value. In the meantime the distal anal canal CSA decreased a little slower initially when squeeze occurred. It maintains this decrease for the duration of the squeeze and only begins to increase to its normal CSA value after the squeeze pressure has ended. Fig. 5C shows the CSAs between the δ prox and δ distal points. They were observed to increase in CSA while the δ points were decreasing. The narrow zone shortened when squeeze occurred at higher bag volumes. This pattern was found in all HCs. Fig. 5D shows the increase in average narrow zone length during step distension. The increase in narrow zone length was seen to get smaller as the bag volume increased. However at each step volume the increase in narrow zone length was much greater at the proximal end of the canal than at the distal end.

Figure 5.

 (A) The changes in the anal profile during squeezing at 20 mL in the functional lumen imaging probe (FLIP) bag between at rest (blue outline) and squeeze (red line) for one health control. This image was used to determine the anal canal borders. The changes at both borders are represented by δ prox (proximal border toward anal verge) and δ distal (distal border towards rectum) which are as a result of changes in the cross-sectional area (CSA) values at the relevant detectors. This figure suggests that the anal canal is narrower at rest than during squeezing and the distended narrow zone is longer during squeezing than at rest. (B) CSA for δ prox and δ distal and bag pressure for the same volunteer as shown in (A). (C) CSA measurements along the anal canal that lie between the δ prox and δ distal CSA measurements for the same volunteer. Fig. 4B,C use the same x-axis time scale and represent identical points in time. (D) Plot of increase in the narrow zone length at three volumes (20, 30, and 40 mL) showing total, proximal and distal changes during squeezing with P values.

Table 1 show that for the squeeze events the average pressure was 33 mmHg. The average squeeze time was 8.8 s. There was little or no difference between the duration that the distal squeeze occurred (Tdistal) and the squeeze time represented by the squeeze pressure (Tsqueeze). The proximal squeeze was much faster to increase to its normal value (Tprox) this was, on average 3.7 s shorter than Tsqueeze.

Table 1.   Total time of squeezing (Tsqueeze), time of distal squeezing (Tdistal) and time of proximal squeezing (Tproximal) in seconds within the FLIP bag during squeezing reported as median
  P peak T squeeze T prox T distal T squeezeTdistal T squeezeTprox
  1. FLIP, functional lumen imaging probe.

Mean (SEM)32.9 (5.1)8.8 (0.7)5.1 (0.4)9.2 (0.7)−0.4 (0.1)3.7 (0.7)

Cough maneuvers

The changes in the geometric pattern were more subtle during cough maneuvers compared to squeeze maneuvers. This is illustrated for one subject in Fig. 6A. However when we observe these changes in time (Fig. 6B) it can be clearly seen that both δ prox and δ distal reduce, in this case for the duration of the cough indicated by the pressure increase. In fact, in this example there is a double peak indicating a secondary cough 5 s after the first one and showing a repeat pattern of pressure increase and δ prox and δ distal decrease. The narrow zone length increased for all HCs with an approximate equal elongation in both proximal and distal ends in the anal profile as in Fig. 6C.

Figure 6.

 (A) The changes in the anal profile during coughing at 20 mL in the functional lumen imaging probe bag between at rest (blue outline) and cough (red line) for one healthy control. This image was used to determine the anal canal borders. The changes at both borders are represented by δ prox (proximal border toward anal verge) and δ distal (distal border towards rectum) which are shown as changes in the cross-sectional area (CSA) values at the relevant detectors. This figure suggests that the distended narrow zone was longer during coughing than at rest. (B) CSA measurements for δ prox and δ distal and bag pressure for the same volunteer as shown in (A). (C) Plot of increases in the distended narrow zone at three volumes (20, 30, and 40 mL) showing total, proximal and distal changes during coughing with the P values.

The pressure in the anal bag during cough was higher than during squeeze at different bag volumes (< 0.001) (Fig. 7). The pressure was significantly lower at 20 mL (= 0.005). However this change is not significantly different between 30 and 40 mL for either maneuver (= 0.985). There was no significant difference between the maneuvers for pressure (= 0.696).

Figure 7.

 Average pressure plotted against volume during squeezing and coughing (= 20). Anal canal pressure during coughing was higher than during squeezing at all three step volumes. Also the squeezing and coughing pressure at 20 mL was significantly lower than the pressure measured at 30 and 40 mL.


This project represents a novel application of a technique originally developed for use in the upper gastrointestinal tract. The study demonstrates that controlled volume distension produces a remarkably similar profile for a group of 20 volunteers in a rapid, safe and effective manner. This data represents the distensibility of the anal canal region in a straight forward manner. The effects of provoking squeeze and cough on the dynamics of the anal canal can be measured.

Physiological data

Positioning of the FLIP was important to define a common baseline among all 20 volunteers; it guaranteed that the same amount of volume was inserted distally towards the rectum area, so that the relaxation in the distal end of the anal canal during bag distension would be uniform for all volunteers. Also the physiological variation in the distended narrow zone length between gender would not affect the results obtained using this technique.

The distension of the anal canal using the FLIP bag reproduces the effect of the entry of the stool to the anal canal. The bag fills first in low-pressure zones, primarily outside the anal verge at both ends. Also the bag filling may to some degree induce recto-anal inhibitory reflexes.19,20

While the interaction of the muscle components in the ano-rectal region is quite complex, it is amazing to see how linear the relationship between bag pressure and the narrowest CSA is in the anal canal during distension. The results tell us that the pressure at which the anal canal opens is lower in males than in females. The opening pressure is a measure of the ano-rectal region’s response to initial distension. In the present study we defined the opening pressure at the most narrow CSA location in the anal canal. Resting pressure measured manometrically is different from bag pressure measured using FLIP which is normally higher in men than women. We suggest that males have a different response to distension than females. A larger group of subjects in future work will confirm or reject this finding.

The difference between the males and females distensibility index is clinically meaningless as it represents a very small changes in the minimum CSA (2.7 mm), while there is a large change in pressure (30 mmHg) .

The consistent opening pattern demonstrated in Fig. 3 suggests that the technique may be valid across a normal healthy population. In future studies it will be interesting to see if this opening point and pattern is different for disease groups. At low volume the distal end was tonically contracted. In most subjects there was also an observed contraction towards the proximal (toward anal verge) end of the profile. When the volume increased to between 15 and 25 mL a full relaxation was observed indicated by the absence of blue color on Fig. 3 which is representative of very small CSAs. At larger volumes (>25 mL) a narrow region formed in the centre of the profiles indicated by the blue region. This could be a semi-voluntary inhibition that ensures defecation is controlled. Previous studies indicated that anal canal contraction increases with the increase in the probe size and anal distension.21,22Gregersen et al. used the impedance planimetry technique and found that the distal part of the anal canal has lower compliance than the proximal part.23

Squeeze is a standard provocative test from manometry. With FLIP the information we get from squeeze represents changes in the geometric profile of the canal. We can see that squeezing does not have a uniform effect along the anal canal profile; two distinct narrow regions formed (δ distal) and (δ prox) which suggests two groups of muscles are involved in squeezing. Moreover these muscles are under voluntary control.24 It is more pronounced towards the proximal end close to the anal verge (Fig. 5A).

As shown in Fig. 5B the proximal end fatigued after 3–5 s. This also concurs with the averaged results for all in these study subjects. The distal end (δ distal) remained contracted for the all of the squeeze time represented by the pressure profile. A peak pressure was observed at the beginning of the squeeze when neither the proximal nor distal ends were fatigued. This work confirms that contraction during squeeze is not as a result of tightening throughout the anal canal. In effect the squeezing at both ends of the canal probably traps the liquid in the FLIP bag which accounts for the increase in CSA recordings at detectors D9, D10, D11, and D12 confirming the effects of the proximal end and the distal end. The liquid will take the path of least resistance and move to the part of the bag experiencing the least force, this may be a source of error and limits interpretation of the physiological significance of this observation.

Under normal circumstances the anal sphincter appears to remain competent without voluntary squeeze being required, but a voluntary squeeze acts as an emergency method of maintaining continence when required. This may explain why we observe a very low pressure needed to cause ano-rectal opening (Fig. 2).

Limitations of the current study

No specific questionnaire was used to evaluate if the subjects had functional bowel disorders. This is a limitation of the study as many presumed HCs with symptom criteria for functional bowel disorders will consider their bowel habits to be ‘normal’. This issue will be considered in future studies.

While every effort was made to ensure there was little or no lateral movement in the probe during this study, we accept that there may be some small amounts of movement particularly with high volume and high pressure in the bag. However the consistent pattern of the distensions as demonstrated in Fig. 3 suggests movement was small.

As there was 5 mm distance between the CSA measurements, anatomical variations in geometry between the measurements could not be determined but we suggest from previous studies in other sphincter regions and from testing that measuring at 5 mm intervals is quite representative.18

The pressure seemed not be stable within and between subjects at very low volumes in the bag. The likely explanation is that because the pressure sensor is located towards the distal end of the bag, when there is very low or no liquid volume in the bag it is possible that this sensor’s pressure reading would be more indicative of the immediate squeeze on it, than of the average force inside the bag. For this reason data is only reported at 10 mL or greater inside the bag.

The physiological interpretation of this experiment is hindered by the absence of any validation components of this technology with other systems such as high definition manometry or pelvic floor MRI. Hence no specific anal components can be determined using this technology but comparative studies in the future may add new insight.

The statistical analysis for the ramp distension test was limited for mean pressure and cross sectional area data, so the inter subject and intra subject variability was not calculated.


This study is the first of its kind to use a distending technique to identify changes in the ano-rectal sphincter. Manometry measures anal pressures and EndoFLIP measures anal distensibility. Further studies comparing manometry and EndoFLIP in healthy subjects and patients are necessary to identify the strengths and limitations of EndoFLIP vs manometry. The technique may have a role in testing sphincter competence in disease groups. With the correct set up it may be possible to visualize ano-rectal function rapidly without specialized training and in the treatment room in the hospital. Studies in patient groups such as those suffering constipation or fecal incontinence will provide further information to determine the practical use of the technique as an aid to the clinician.


The authors wish to acknowledge Anne Brokjær for her assistance in this work.


Aalborg Hospital for funding this project in part. A student bursary from Trinity College Dublin.


BMM and HG are minority shareholders in Crospon Ltd.

Author Contributions

MA Helped in designing the study research, performed the study, analyzed the results, and wrote the paper; HG analyzed the results and wrote the paper; AD reviewed the paper; BMM designed the study, analyzed the results, and wrote the paper.