Loop analysis of the anal sphincter complex in fecal incontinent patients using functional luminal imaging probe

Abstract

Cardiac loops have been used extensively to study myocardial function. With changes in cardiac pump function, loops are shifted to the right or left. Functional luminal imaging probe (FLIP) recordings allow for loop analysis of the anal sphincter and puborectalis muscle (PRM) function. The goal was to characterize anal sphincter area-pressure/tension loop changes in fecal incontinence (FI) patients. Fourteen healthy subjects and 14 patients with FI were studied. A custom-designed FLIP was placed in the vagina and then in the anal canal, and deflated in 20-ml steps, from 90 to 30 ml. At each volume, subjects performed maximal voluntary squeezes. Area-pressure (AP) and area-tension (AT) loops were generated for each squeeze cycle. Three-dimensional ultrasound imaging of the anal sphincter and PRM were obtained to determine the relationship between anal sphincter muscle damage and loop movements. With the increase in bag volume, AP loops and AT loops shifted to the right and upward in normal subjects (both anal and vaginal). The shift to the right was greater, and the upward movement was less in FI patients. The difference in the location of AP loops and AT loops was statistically significant at volumes of 50 ml to 90 ml (P < 0.05). A similar pattern was found in the vaginal loops. There is a significant relationship between the damage to the anal sphincter and PRM, and loop location of FI patients. We propose AP and AT loops as novel ways to assess the anal sphincter and PRM function. Such loops can be generated by real-time measurement of pressure and area within the anal canal.

NEW & NOTEWORTHY We describe the use of area-pressure (AP) and area-tension (AT)-loop analysis of the anal sphincters and puborectalis muscles in normal subjects and fecal incontinent patients using the functional luminal imaging probe (FLIP). There are differences in the magnitude of the displacement of the loops with increase in the FLIP bag volume between normal subjects and patients with fecal incontinence. The latter group shifts more to the right in AP and AT space.

INTRODUCTION

In a recent large survey of residents in the United States, 14% of subjects reported fecal incontinence (FI) in the past, and 33% within the preceding 7 days (18). The prevalence of FI is age-related; it is more prevalent among individuals with inflammatory bowel disease, celiac disease, irritable bowel syndrome, or diabetes. The etiology of FI is multifactorial; stool consistency, rectal reservoir function, and anal sphincter muscle function play important roles in its genesis (2, 27). However, anal sphincter dysfunction is the most important factor in the development of FI (1, 28). Severe FI symptoms are more often observed in female patients; more than 80% of patients in the FI treatment clinical trials are women (7, 11, 12). The latter is most likely because women are susceptible to obstetric injuries to the anal sphincter and pelvic floor muscles. Ultrasound and magnetic resonance imaging studies show that 20–35% of women develop damage to the anal sphincters and pelvic floor muscles following vaginal childbirth (3, 31).

Anal manometry has been in use for more than 50 years to test the strength of the anal sphincter muscles. Many types of manometry systems have been used over the years; high-resolution anal manometry (HRAM) is the current gold standard to diagnose anal sphincter strength (16). Manometry probes measure anal sphincter strength under isometric conditions (no change in muscle length), i.e., with a fixed diameter and noncollapsible probe. More recently, the functional luminal imaging probe (FLIP) has been used to measure the anal sphincter strength in normal subjects and patients with FI (6, 30). Unlike manometry probes, FLIP records concentric contraction of the anal sphincter muscle, i.e., a decrease in muscle length with contraction. Anal canal distensibility, the parameter used in the FLIP studies, is the ratio of luminal cross-sectional area to bag pressure; it is higher in FI patients compared with normal subjects.

We recently reported that the FLIP allows one to study the length-tension function of the anal sphincter muscle (32). Cardiologists have used cardiac loops to evaluate the function of myocardium for many decades (5). A cardiac loop represents changes in the luminal cross-sectional area (CSA) and pressure or tension during a cardiac cycle (systole and diastole). Using cardiac loops, one can assess how changes in the luminal CSA (left ventricular volume in the case of heart) cause difference in the muscle contractility and workload of the myocardium (length-tension property or the Starling Curve). In normal myocardium, cardiac loops are shifted to the right and upward, with an increase in the ventricular CSA. Damage to the myocardium results in loop shift to the right. FLIP data lend itself to loop analysis, which has never been performed in the assessment of anal sphincter function.

The goal of our study was to use loop analysis to determine the anal sphincter function in healthy subjects and patients with FI. Furthermore, we assessed the relationship between dysfunction of anal sphincter loops in patients with FI and damage to the anal sphincter and puborectalis muscle (PRM). The latter was assessed using three dimensional-ultrasound imaging of the anal sphincter and PRM.

MATERIALS AND METHODS

Fourteen nulliparous women (age 34 ± 13 yr) with no history of pelvic floor dysfunction and 14 women with FI symptoms (age 62 ± 15 yr) of greater than 3-mo duration were recruited for this study (Table 1). The protocol for the study was approved by the Human Research Protection Program at University of California, San Diego, and all subjects signed a written informed consent before their participation in the study. All participants completed the FI questionnaire (or FIQLS), the Pelvic Floor Distress Inventory-20 (or PFDI-20), and questionnaires related to severity of FI (FISI and Wexner questionnaires).

Table 1.

Patient demographics in all patients with fecal incontinence

Values	Patients	Controls	P
Age	62 ± 15	34 ± 13	<0.0001
Female sex	100%	100%
Body mass index, kg/m2	23.9 ± 4.0	20.0 ± 1.9	0.0028
Number of vaginal deliveries	1.8 ± 0.8
Birth trauma	12/14 (86%)
Forceps or instrument-assisted deliveries	6/14 (42.86%)
Tear or episiotomy	8/14 (57.14%)
FISI	28.07 ± 13.53
Wexner	11.64 ± 4.31
PFDI-20	89.95 ± 40.95
POPDI-6	19.34 ± 14.85
CRADI-8	45.31 ± 15.92
UDI-6	25.30 ± 23.88
PFIQ-7	77.76 ± 55.61
Duration of FI, yr	5.5 ± 4.63; n = 8
Past surgery
Hysterectomy	5/14 (35.71%)
Hemorrhoid surgery	2/14 (14.29%)
Prolapse repair	1/14 (7.14%)
Anal fistula repair	1/14 (7.14%)
Vaginal laser	1/14 (7.14%)

Values are expressed as means ± SD; n = 14 patients. CRADI, Colorectal-Anal Distress Inventory-8; FI, fecal incontinence; FISI, Fecal Incontinence Severity Index PFDI, Pelvic Floor Distress Inventory-20; PFIQ, Pelvic Floor Impact Questionnaire; POPDI-6, Pelvic Organ Prolapse Distress Inventory 6; UDI, Urogenital Distress Inventory.

The FLIP measurements and three-dimensional-ultrasound imaging (3D-US) were performed with the participants in the lithotomy position. A custom-designed FLIP that incorporated a 30-mm diameter bag when fully distended was used for these studies. Initially, bench studies were performed to determine the maximal volume used in the bag before in vivo use, which was the volume at which bag pressure starts to increase. For the 30-mm bag used in our studies, the maximal volume was 90 ml, and this was the maximal volume used for the in vivo studies. Our earlier studies (10, 13, 22) show that the vaginal high-pressure zone is related to the PRM function, which was the rationale for FLIP studies of the vaginal canal. The FLIP bag was first placed in the vagina and distended to maximal volume and then deflated in steps of 20 ml starting from 90 to 30 ml. Participants were asked to perform maximum voluntary squeezes for 10 s, three times at each bag volume with the cue to contract the anal sphincter/pelvic floor muscle, as if they were trying to stop the flow of urine or passing of gas. The above procedure was repeated with the FLIP bag in the anal canal (Fig. 1).

Fig. 1.

A sample subject undergoing the functional luminal imaging probe (FLIP) stepped volume distension protocol, starting from the vaginal canal and proceeding to the anal canal. The subject was asked to perform a voluntary contraction following a 30-s rest period. The pressure change for this recording is shown as the red waveform, and the surface topograph represents the change in balloon diameter. The changing colors from dark blue to yellow illustrate increasing diameter. The subject’s squeeze can be identified from the peaks of the red line, and the corresponding reduction in diameter of the anal canal and color change toward dark blue on the diameter topograph.

To determine reproducibility of the loops analysis, five patients with FI were studied twice using the same FLIP testing protocol, one week apart. Reproducibility analysis was carried out for the anal canal area-pressure loops.

3D-US imaging of the anal sphincters [internal anal sphincter (IAS) and external anal sphincter (EAS)] and PRM was performed using the HD11-US system (Philips Medical Systems, Bothell, WA), as described earlier (17, 33). The 3- to 9-MHz transducer was placed in the labial fourchette for imaging the anal canal and pelvic floor hiatus. The transducer was directed in the posterior and cranial directions for imaging the anal canal and PRM, respectively. The 3D-US volumes were captured at rest and during a sustained squeeze.

Data analysis.

US image analysis was performed offline using Phillips Q-laboratory-5.0 software to assess the presence and severity of damage using a scoring system, described previously (15). The damage to each of the PRM hemi-slings was scored on a scale of 0–2, with a score of 0 = no damage, score 1 = damage to <50% of the PRM length, and score 2 = damage to >50% of the PRM length. The scores from two hemi-slings were added for a maximal score of 4. Pelvic floor hiatus length was measured in the plane of the least pelvic floor dimension, i.e., a line connecting the lower edge of the pubic symphysis to the anorectal angle. Damage to the EAS and IAS was determined individually on each 1-mm cross-sectional slice of the anal canal, along the entire length of the anal canal. Score 0 = no damage, score 1 = damage to ≤25% of length, score 2 = damage between 25 and 50% of length, score 3 = damage between 50 and 75% of length, and score 4 = damage greater than 75% of length. The senior investigator (R. K. Mittal), who was blinded to the FLIP data, performed the US image scoring.

The FLIP data of the entire study in each subject were exported in text file format; it contained 16 dynamic diameter and pressure measurements. The area-pressure (AP) and area-tension (AT) loops were generated from the above recordings. Each loop consisted of a continuous time signal comprising transition from the baseline to peak pressure and back to baseline. For each loop, the channel with the minimum area was used for the AP and AT loops. At each time point of the loop, tension was calculated (pressure × radius) in Newton/meter using custom-developed software. Only those distension volumes that produced luminal diameters larger than the FLIP probe diameter (8.3 mm) were used for the tension calculations.

Statistical analysis.

Data are presented as means ± SD, unless otherwise stated. Normality was tested using the Shapiro-Wilk test. A mixed analysis of variance (ANOVA) with post hoc Bonferroni correction was carried out on the data for mean comparisons, between and within subjects. Statistical significance was defined as P < 0.05. All calculations were performed in MATLAB 2017b (MathWorks, Natick, MA).

The relationship between anatomy (damage to the muscles) and function (AP-loops) was assessed by first estimating the normal group mean in area-pressure space, using nonparametric probability density estimation of the normal group area-pressure loop centroids. The latter allows robust generalization for unseen data, incorporating uncertainty into the estimated sample mean rather than a simple averaging of the samples. Next, the distance between each of the patient loop centroids was compared with the peak area-pressure coordinate of the estimated normal mean value in area-pressure space. Finally, the resulting vector of distances were correlated with the US findings. The analysis sought to answer the question of whether patients with greater FI symptoms would have higher distances from the normal group mean. As to assessing reproducibility, which here refers to variability in the time interval between the repeated measurements (assuming observer, instrument used, instrument calibration, environment [temperature, humidity, etc., remain the same], one-way ANOVA is used to calculate the within-subject standard deviation (SD). The reproducibility limit (R) is also reported as (1.96√2 × SR) (i.e., 95% confidence interval). Note, smaller SD values indicate closer agreement between the loops.

Optimal volume selection for discriminating among groups (FI vs. normal).

Using the area-pressure space coordinates of the loop centroids (of all subjects) as predictor variables, a binary classification problem was defined, and a logistic regression model was fitted for computing the Receiver-Operating-Characteristic (ROC) and Precision-Recall (PR) curves. The probability estimates of the logistic regression model were used as scores. Two evaluation metrics, precision-recall area under the curve (PR-AUC), and receiver-operating-characteristic area under the curve (ROC-AUC) were used to evaluate classifier performance.

RESULTS

The FLIP bag pressure and tension increased with the increase in the bag volume in normal subjects (P < 0.05), (Figs. 1–7). Above was true for both the anal and vaginal measurements. Example of the squeeze cycle loops from a normal subject and a FI patient for the bag volumes of 50, 70, and 90 ml are shown in Figs. 2 and 3. Each loop represents changes in the pressure and CSA from rest to the peak of muscle contraction and return to rest with relaxation. With contraction, the pressure increases and area decreases, and both return to rest with relaxation. With increasing bag volume, the anal as well as vaginal loops are shifted to the right (increase in CSA) and upward (increase in pressure) in normal subjects (Figs. 3, 4, and 6). Anal and vaginal area-tension (AT) loops follow the same pattern as area pressure (AP) loops in normal subjects, i.e., with an increase in bag volume, there is a shift of loops to the right and upward (Figs. 3, 5, and 7). Figures 5 and 7 show the relationship between anal AT loops and vaginal AT loops at bag volumes of 50, 70, and 90 ml in normal subjects. Similar to the AP loops, the AT loops also move to the right (P < 0.05) and upward in normal subjects.

Fig. 7.

Effect of functional luminal imaging probe (FLIP) bag volume on the cross-sectional area (CSA)-tension anal and vaginal loops in normal subjects (A and B) and fecal incontinence (FI) patients (C and D). Note, the shift of the loops to the right and upwards in both groups.

Fig. 2.

Squeeze cycles (loops) of a subject at volumes 50–90 ml: normal subjects (A) and fecal incontinence (FI) patients (B). Note the higher diameter values in the FI patient compared with the normal subject.

Fig. 3.

Three-dimensional-ultrasound (US) images of the anal sphincters and puborectalis muscle (PRM) in a normal subjects (A) and fecal incontinence (FI) patients (B), along with the anal and vaginal loops at bag volumes of 50, 70, and 90 ml. Patient has extensive damage to the anal sphincter and PRM. Note the shift of the loops to the right and upward with increase in bag volume in both normal subjects and FI patients. The shifts are greater in patients compared with normal subjects.

Fig. 4.

Cross-sectional area (CSA)-pressure loops for volumes 50, 70, and 90 ml in the anal canal (A–C), and vaginal canal (D–F). Note the shift of area-pressure (AP) loops to the right and upward with an increase in the balloon volume in normal subjects and patients. Also note the differences in the shift between the two groups

Fig. 6.

Effect of functional luminal imaging probe (FLIP) bag volume on the cross-sectional area (CSA)-pressure anal and vaginal loops in normal subjects (A and B) and fecal incontinence (FI) patients (C and D). Note, the shift of the loops to the right and upwards in both groups.

Fig. 5.

Cross-sectional area (CSA)-tension loops for volumes of 50, 70, and 90 ml in the anal canal (A–C) and vaginal canal (D–F). Note the shift of the loops to the right and upward in normal subjects with the increase in the bag volume. Also note the difference in the shift between the two groups. See results for details.

In patients with FI, similar to normal subjects, with increasing bag volume there is increase in the bag area and pressure, which results in a shift of the AP loops to the right and upward (anal as well as vaginal) (Figs. 3, 4, and 6). However, compared with normal subjects, FI patients produce a greater shift of the loops to the right (P < 0.01), and the upward shift is less (P = 0.08). Similar to AP loops, the AT loops show a greater loop movement to the right in the FI patients compared with normal subjects. On the other hand, unlike AP loops, the AT loops show greater movement in the upward direction in FI patients compared with normal subjects (Figs. 5 and 7).

Each loop contains CSA, pressure, and tension values at rest (baseline) and during the contraction cycle. We compared the above parameters at rest and at the peak of contraction between the two groups at the corresponding bag volumes. The anal canal pressures at rest were not significantly different between the two groups, (P = 0.33). However, the bag pressure was higher at the peak squeeze in normal subjects compared with FI patients, even though this did not reach significance (P = 0.08). On the other hand, the CSA was significantly higher at rest, as well as at the peak of contraction in the FI patients, P = 0.004 and P = 0.006, respectively. The above pattern was also observed for the vaginal loops at squeeze (P = 0.02). The anal tension was not different between the two groups at squeeze (P = 0.26); however, it tended to be higher in the FI patients compared with normal subjects at baseline (P = 0.075).

Correlation of ultrasound findings with CSA-pressure loops.

3D-US imaging of the anal sphincter and PRM revealed varying degrees of damage to the IAS, EAS, and PRM in the FI patients. The distance of the AP-loop centroid of each patient to the mean centroid of all normal AP-loops is shown in Fig. 8B. The relationship between degree of damage to the IAS, EAS, and PRM to loop dysfunction was assessed using correlation analysis. In the anal canal, the IAS had the highest significant correlation with the distance from the normal mean at 50 ml (r = 0.51, P value = 0.03) and 70 ml (r = 0.47, P value = 0.04). At 50 ml, the EAS (r = 0.42, P = 0.072) and puborectalis muscles (r = 0.44, P = 0.074) showed a similar pattern. In the vaginal canal, only the puborectalis muscles correlated positively with the distance from the normal vaginal mean, both at 50 ml (r = 0.63, P = 0.003) and 70 ml (r = 0.774, P = 0.004).

Fig. 8.

Centroid distances in normal and patients. A: spread of centroids, where green denotes normal, and pink denotes fecal incontinence (FI) in the area-pressure plane. Nonparametric kernel density estimation (KDE) shows the probability density estimate in the normal group. B: distance of the FI patient loop centroid to the center of the KDE panel. Injury severity scores of the external anal sphincter (EAS), internal anal sphincter (IAS), and the PRM were determined from the ultrasound (US) images and were correlated to the distances of the patient AP-loop centroids from the normal AP-loop centroid mean (red star in B).

Optimal volume selection for discriminating between groups.

For the anal canal, the 50-ml volume produced the highest AUC values (i.e., better classification performance) for discriminating between FI patients and healthy controls (0.942 ROC-AUC and 0.802 PR-AUC). In the vaginal canal, the same volume (i.e., 50 ml) also produced the highest values for discrimination between the FI patients and controls (0.857 ROC-AUC and 0.801 PR-AUC).

Loop reproducibility.

To assess reproducibility, the peak squeeze pressure point in the anal area-pressure loops was chosen as a sample landmark in each of the five patients. As there were three loops (squeeze cycles) at each volume, the loop with the peak squeeze pressure value closest to the median of three loops was selected as the representative loop for each volume (Fig. 9). The results revealed a reproducibility (SD) of 23.69 [reproducibility limit (R) of 3.49] for 50 ml, R = 12.89 (R = 9.95) for 70 ml, and SD = 8.46 (R = 8.06) for 90 ml. (Note, to have the reproducibility results in the same unit as pressure (i.e., mmHg), the square root of the SD values was used, as variation values are squared).

Fig. 9.

Cross-sectional area pressure loops in four fecal incontinence (FI) patients at 50, 70, and 90 ml. Red and blue tracings are from the two testing sessions, 1 wk apart.

DISCUSSION

We describe the area-pressure (AP) and area-tension (AT) loop analysis of the anal sphincters and PRM in normal subjects and patients with FI. Our findings are that with an increase in the FLIP bag volume, the AP and AT loops shift to the right and upward in normal subjects, which represents the length-tension property of the normal anal sphincters and PRM. There are differences in the magnitude of the shift of loops, with an increase in the bag volume between normal subjects and FI patients. These differences in the shift of loops represent dysfunctional muscle. We found a significant correlation between the severity of damage to anal sphincters and PRM with the magnitude of loop shift in the FI patients.

Anal sphincter function has been assessed using many methodologies. A three balloon system (Schuster balloon) was one of the earliest methods described (29). It was replaced by infusion manometry (21), which employed either side holes or a sleeve sensor to record pressure. Solid-state pressure sensor catheters were also used by many laboratories. Closely spaced, solid-state pressure sensors and color pressure topography, also called high-resolution anal manometry, is the current gold standard in the assessment of anal sphincter function. Three-dimensional, high-definition anal manometry (HDAM), another novel measurement system, provides information on the asymmetry of pressure in the anal canal (23). All of the above systems measure pressure using fixed diameter, noncollapsible probes (4-mm probe for HRM and 10 mm for 3D-HDAM) and, hence, measure isometric contraction of the IAS (pressure at rest) and EAS (peak pressure) at one muscle length. Another approach to measure the anal sphincter and PRM muscle function is to study their length-tension property (Starling Curve, best known in myocardial function assessment), which we reported earlier using different size probes in the anal (20) and vaginal canal (14). Experimental studies in the laboratory usually measure muscle strength by recording the force generated at various muscle lengths, i.e., the length-tension property.

The relationship between changes in muscle length and muscle function can be assessed by loop analysis, which has been used extensively in the cardiac literature. A cardiac loop describes changes in the ventricular CSA and pressure/tension during a cardiac cycle (systole and diastole) (5). One can study the length-tension function, as well as damage to myocardium using loop analysis (4). In normal subjects, the cardiac loops are shifted to the right and upward with the increase in the ventricular volume (length-tension property) and are shifted to the right in damaged myocardium. Gregersen, the inventor of FLIP, suggested the feasibility of using FLIP for loop analysis (8, 9). For the first time, we describe the anal sphincter and PRM function loops in normal subjects and patients with FI. Increases in FLIP bag volume result in distension of the anal canal and, hence, increases in the anal sphincter muscle length. In an earlier study, we found a close relationship between the length-tension measured by FLIP and noncollapsible probes of different diameters (32). In the present study, we use loop analysis using FLIP methodology in normal subjects and FI patients. Interestingly, the shifts in loops that we observed in the anal sphincters of normal subjects are similar to those reported in the myocardium (i.e., shifts to the right and upward with an increase in muscle length/volume in normal subjects). The above finding is not surprising because, similar to myocardium, the anal sphincter and PRM operates at short sarcomere length or the ascending limb of the Frank Starling curve under in vivo conditions (14, 20, 25). Furthermore, and most interestingly, we observed that the shift in the anal sphincter and PRM loops in FI patients is similar to what is seen in the setting of damaged myocardium.

Several studies prove that the anal sphincter assessment using FLIP methodology can effectively distinguish anal sphincter function between FI patients and normal subjects. In all of those studies, investigators measured anal canal distensibility, which is a ratio of the bag CSA at the level of least dimension to bag pressure, and they found it to be higher in FI patients compared with normal subjects (6, 30). Distensibility is generally used to describe the material property, and it generally implies passive function. We believe loop assessment is a better analytical approach than distensibility because it allows one to assess the function at rest and during contraction, as well as the length-tension property. It is important to study muscle function at various lengths because we believe it reveals what we call the “muscle reserve function.” We observed that patients with FI can generate equivalent or greater amounts of muscle tension than normal subjects, both in the anal and vaginal canal. We were initially surprised at the above finding; however, inspection of the loops revealed that even though FI patients can generate comparable muscle tension to normal subjects, they achieve it at a significantly higher muscle length. The ability of muscle to achieve greater tension at longer length may be regarded as the “muscle reserve function.” We propose that the “muscle reserve function” may be tapped to improve muscle function in FI patients. Our earlier work shows that plication of the EAS muscle in rabbit increases sarcomere (muscle) length and increases anal canal pressure (24), which is sustained for 6 mo (26). Along those lines, we believe that if the FLIP methodology shows gain in tension with an increase in muscle length in an individual patient, plication of the sphincter and PRM is likely to improve the anal canal pressure and possibly restore continence function. On the other hand, if in an individual patient, there is no increase in muscle tension with an increase in the muscle length, plication of the sphincter and PRM is unlikely to improve sphincter pressure and continence function.

3D-US imaging of the anal sphincter and PRM, now being used by many groups, is a powerful tool to assess damage to the IAS, EAS, and PRM. We have previously described the technique, scoring system, and relationship between damage seen on the US images and function assessed by the length-tension function measured using fixed-diameter probes, which record isometric muscle contraction (15). In the present study, we found significant correlation between damage to the IAS and PRM seen on the 3D-US images and anal/vaginal loops. The relationship between loop dysfunction and muscle damage was stronger for the IAS and PRM than for the EAS. We suspect that the reason for the above finding is that US imaging is not ideal to determine the EAS morphology and muscle damage. The reason for the above is the unique “purse string” morphology of the EAS muscle that is not assessed by the US imaging, because it assumes a circular shape of the EAS muscle, which is actually not the case (19, 34). Transverse perinea and bulbospongiosus muscles are part of the EAS muscle and not assessed in the US images. More recently, we described magnetic resonance diffusion tensor imaging (MR-DTI) methodology to assess the EAS and other superficial muscles of the perineum (34). Studies are needed to investigate structure function assessment of the EAS muscle using MR-DTI technique.

Whereas our current study describes an important new way to assess the anal sphincter function, there are limitations to be considered. First, our sample size is relatively small, although the differences between patients and controls are compelling, even in this small sample size. Second, our groups are not age-matched. We tested age as a confounding variable and failed to reject the null hypothesis that there was an age effect between the two groups. Furthermore, in an earlier study of age and parity-matched asymptomatic women and FI patients, we found that age alone cannot explain anal dysfunction in FI patients (15). Interestingly, our groups were also different on the basis of body mass index (BMI) in the earlier studies. On the other hand, in the current study, the BMI of our FI patients was within the normal range (mean of 23.9). Future studies should include those with higher BMI to determine the impact of obesity on loop analysis. The patients in this study had FI for more than two years and, in some cases, even longer time periods. It is unclear whether the loop analysis would differ on the basis of the duration of FI symptoms. Furthermore, it would also be of interest to perform loop analysis in patients who have undergone posterior compartment repairs, mesh placement, or other surgical procedures.

In summary, FLIP can assess the length-tension property of the anal sphincters and PRM. Loop analysis, which is used extensively in the cardiac literature, is a novel way to study the length-tension function and dysfunction of the anal sphincter muscles. We found differences in the anal sphincter loops between normal subjects and patients with FI that are suggestive of damage to the muscle as the reason for anal sphincter dysfunction. We propose that loop analysis of FLIP data is an appropriate way to measure muscle function of the anal sphincters and PRM and may allow for measurement of functional outcomes of interventions, including predicting whether length adjustment of the sphincter and PRM may improve anal continence function.

GRANTS

This work was supported by National Institutes of Health Grant 1R01HD-088688.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

R.K.M. and L.J.T. conceived and designed research; R.K.M., J.S., and L.J.T. performed experiments; A.Z., R.K.M., D.C.K., J.S., G.B., and L.J.T. analyzed data; A.Z., R.K.M., and L.J.T. interpreted results of experiments; A.Z. prepared figures; A.Z., R.K.M., and L.J.T. drafted manuscript; A.Z., R.K.M., D.C.K., J.S., and L.J.T. edited and revised manuscript; A.Z., R.K.M., D.C.K., J.S., and L.J.T. approved final version of manuscript.