Clinical and Investigative Medicine

Adaptation of the rectal wall to distension in children with constipation

Michel Bouchoucha, MD, PhD
Jean-Phillipe Jais, MD, PhD
Pierre Arhan, MD, PhD
Paul Landais, MD, PhD
Denis Pellerin, MD

Clin Invest Med 1997; 20(1): 35-40.

[résumé]

Dr. Bouchoucha is with the Laboratoire de physiologie digestive, Hôpital Laënnec; Drs. Jais and Landais are with the Laboratoire de biostatistique et d'informatique médicale, Hôpital Necker­Enfants Malades; and Drs. Arhan and Pellerin are with the Clinique chirurgicale infantile, Hôpital Necker­Enfants Malades, Paris, France.

(Original manuscript submitted Mar. 27, 1996; received in revised form Oct. 24, 1996; accepted Oct. 25, 1996)

Reprint request to: Dr. Michel Bouchoucha, Laboratoire de physiologie digestive, Hôpital Laënnec, 42 rue de Sèvres F-75007 Paris, France; fax 33 1 4439-6922

Contents

Abstract

Objective: To determine whether there is significant change in the tension­length relationship of the digestive smooth-muscle mechanical function in constipation in children and to examine a new method of analysis of the mechanical properties of the rectal wall.

Design: Case­control study.

Participants: Thirty children with constipation and 30 control children who did not have constipation.

Intervention: Rapid distension of the rectal wall by inflation of a rectal balloon with air.

Main outcome measures: The in vivo rectal pressure­volume P[t,V] curve was determined according to the quasilinear viscoelasticity law. The recorded pressure was defined as the product of 2 functions: the elastic response P0[V] and the reduced relaxation function G[t], a normalized function of time such that P[t,V] = P0V] × G[t]. Analysis of variance with repeated measures and modelling (linear for P0[V] and exponential plus constant term for G[t]) were used for data analysis.

Results: The quasilinear viscoelastic law can be applied to the in vivo determination of the mechanical properties of the rectal wall in controls and in children with constipation. The elastic response was similar in the 2 groups. The reduced relaxation function was significantly different between the 2 groups, with the absence of an asymptotic value in the group with constipation (p < 0.01).

Conclusion: Distension of the rectal wall with the use of an air-inflated balloon, with this type of interpretation according to the viscoelastic law, is useful in the analysis of chronic constipation.

[Table of contents]

Résumé

Objectif : Déterminer si le rapport tension­longueur de la fonction mécanique des muscles lisses de l'appareil digestif change pour la peine chez les enfants constipés et examiner une nouvelle méthode d'analyse des propriétés mécaniques de la paroi rectale.

Conception : Étude cas­témoin.

Participants : Trente enfants constipés et 30 enfants-témoins non constipés.

Intervention : Distension rapide de la paroi rectale par inflation d'un ballonnet rectal avec de l'air.

Principales mesures des résultats : On a établi la courbe pression­volume P[t, V] rectale in vivo selon la loi de la viscoélasticité quasi linéaire. La pression consignée a été définie comme le produit de 2 fonctions : la réponse élastique Po[V] et la fonction de relaxation réduite G[t], une fonction normalisée du temps telle que P[t,V] = P0[V] × G[t]. Pour analyser les données, on a analysé les écarts, effectué des mesures répétées et établi des modèles (linéaires pour P0[V] et exponentiels plus constante pour G[t]).

Résultats : Il est possible d'appliquer la loi de la viscoélasticité quasi linéaire à la détermination in vivo des caractéristiques mécaniques de la paroi rectale chez des enfants­témoins et des enfants constipés. La réaction élastique a été la même chez les sujets des 2 groupes. La réduction de la fonction de relaxation a été très différente entre les 2 groupes et l'on a constaté l'absence de valeurs asymptotiques chez les sujets constipés (p < 0,01).

Conclusion : Avec ce type d'interprétation conforme à la loi de la viscoélasticité, la distension de la paroi rectale au moyen d'un ballonnet gonflé à l'air est utile dans l'analyse de la constipation chronique.

[Table of contents]

Introduction

The tension­length relationship is an important feature of digestive smooth-muscle mechanical function. The in vivo isometric mechanical behaviour of the human rectum has been investigated in patients with Hirschsprung disease,[1] ulcerative colitis,[2] ischemic fecal incontinence[3] and megarectum.[4] Viscoelastic properties of the rectum differ in health and disease; they can be studied to distinguish constipation caused by outlet obstruction -- resulting from a hypotonic or hypertonic rectum -- from simple anal achalasia.[5] In previous studies, the accommodation properties of the rectum were investigated after rapid distension with air[1­4] or slow distension with water.[5,6]Both techniques of distension permit the study of the elastic properties of the rectal wall, but only the first allows the investigation of its viscoelastic properties.

In a previous study,[7] we proved that the law of quasilinear viscoelasticity[8] can be used to determine the viscoelastic properties of the rectal wall with the use of distending volumes of 20 mL or more. The human rectum, when stretched by various volumes injected into a spherical balloon, behaves according to the quasilinear viscoelastic law P[t,V] = P0 [V] × G[t], where P[t,V] is the pressure at time t after distension of V mL, V is the distending volume, P0[V] is the elastic response of the rectal wall for the distension of V mL, and G[t] is the reduced relaxation function, a normalized function of time.

In this study, we wished to show how this method can demonstrate functional motor abnormalities of the rectum in patients with constipation. These abnormalities are not merely the changes passively induced by outlet obstruction due to spastic pelvic floor syndrome (i.e., a striated muscle problem), but could result from smooth-muscle disturbances of the rectal wall.

The aim of this study was to determine the 2 functions P0[V] and G[t] in children with constipation and control children who did not have constipation to discover any significant differences in these functions in the 2 samples.

[Table of contents]

Material and methods

Patients

Two samples of children were studied: 30 children with constipation and 30 control children without constipation. The control subjects were 6 months to 14 years of age and had been admitted to hospital for disorders unrelated to the gastrointestinal tract. They had normal bowel habits and no anorectal abnormalities. They were studied either before surgery or on the day of discharge. Results from these subjects were presented in a previous study involving a classic analysis.[1]

The patients with constipation were 1 year and 6 months to 15 years of age. They had complained of constipation and defecated less than 3 times per week[9,10] while eating a diet containing more than 0.20 g of dietary fibre per kilogram. An extensive clinical evaluation was performed to exclude known causes of constipation. The evaluation was carried out with the same criteria used in previously published studies.[9,11] In particular, the recto-anal inhibitory reflex was present in all patients, thus eliminating any patients with Hirschsprung disease.[12]

Rheological studies

Viscoelastic properties of the rectal wall were studied by a previously described method.[13] The subjects were investigated in the morning. All were lying in bed. Some were sedated because of restlessness (with diazepam 0.15 mg/kg given intramuscularly). Sedation did not influence the validity of the results.[10] A four-channel probe was used (Ets MARQUAT, Zone Industrielle La Haie Griselle, Boissy Saint Leger, France). Two channels were connected to a latex spherical balloon fixed at the tip of the probe and placed into the rectum. The other channels were connected to ring-shaped balloons placed 4 cm below the distending balloon and into the anal canal. The rectum was then distended by inflating the rectal balloon with 20, 30, 40 and 50 mL of air, in random order, in less than 0.5 seconds; distension was maintained for 40 seconds. Pressure was measured through 1 channel connected to the rectal balloon, recorded by transducers (Statham P23 DC amplifiers) and displayed on a graphic recorder. These studies were repeated in vitro, in ambient air, with similar volumes of distension. Data from in vitro curves were subtracted from the in vivo curves to obtain the actual rectal contribution to measured pressure. Data were recorded at a paper speed of 25 mm per second during the first 10 seconds and at 2 mm per second after that. Measurements were taken every 0.2 seconds for the first 4 seconds, every second until the 10th second, and at 15, 20, 30 and 40 seconds.

Normalization of experimental curves and determination of the passive state

As in our previous study,[7,14] the recorded pressure was decomposed into 2 functions by calculating the elastic response P0[V] as the first measured point (P[0,V]) and the reduced relaxation function G[t], a normalized curve, as: G[t] = P[t,V]P[0,V], where P[0,V] was the measured stress for the distending volume V at time 0.

The resting pressure was then defined as the inferior envelope of the reduced relaxation function. The activity was suppressed with a simple algorithm of decrease of the curve, as defined previously.[7,14]

Modelling

The functions P0[V] and G[t] were modelled according to previously described results: a linear regression was performed for P0[V][14] and an exponential regression was performed for G[t]15 according to the equations P0[V] = a + b × V and G[t] = alpha exp[ßt] + gamma.

Data analysis

Data were compared by analysis of variance (ANOVA) with repeated measures.[16] The design of ANOVA included 1 subject factor and 2 within factors (volume of distension, with 4 levels, and time, with 30 measures). The 2 groups were compared by adding a trial factor (the group) to the previous ANOVA design. Posthoc comparisons were made with the Student­Newman­Keuls test. The analysis was conducted with SAS software (SAS Institute Inc., Cary, NC).

[Table of contents]

Results

In the first part of the analysis, the reduced relaxation function (normalized properties of the rectal wall), G[t], was compared in subjects with constipation and controls for the curves corresponding to 20 to 50 mL of distension. Children with constipation had a higher mean normalized pressure than controls (0.705, standard deviation [SD] 0.119 v. 0.571, SD 0.036; p < 0.0002), and there was a significant difference in the shape of the curves between the 2 groups (Fig. 1, p < 0.0001).

Nevertheless, for each group, the mean level of each of the 4 curves (20 to 50 mL) did not differ significantly (i.e., there was no curve effect).

The linear modelling of the elastic response of the rectal wall P0[V] (r = 0.97 for controls and r = 0.95 for children with constipation) showed no significant difference for either coefficient a (52.8 mm Hg in controls v. 45.6 mm Hg in children with constipation) or b (0.21, SD 0.04 mL-1 in controls v. 0.37, SD 0.08 mL-1 in children with constipation). As seen in Fig. 2, the curves were similar except in the case of the lowest volume (20 mL), for which P0[V] was higher in controls than in patients with constipation.

In contrast, the exponential modelling of G[t] (r = 0.97 for controls and r = 0.98 for children with constipation) showed significant differences in coefficients alpha (0.96, SD 0.20 v. 0.56, SD 0.02 in children with constipation, p < 0.01), ß (-0.0195, SD 0.002 Hz in controls v. -0.155, SD 0.009 Hz in children with constipation, p < 0.01) and gamma (0.497, SD 0.003 in controls v. 0, SD 0.531 in children with constipation, p < 0.01). The significantly different decrease in the reduced relaxation function G[t] in the 2 groups (Fig. 1) was related to:

[Table of contents]

Discussion

This study shows that the measurement of the length­tension relationship, so useful for in vitro measurement,[17] is feasible in in vivo investigations. It provides physiological information (Fig. 1) and the elastic properties of the rectal wall (Fig. 2) as the tissue relaxes. The use of the quasilinear viscoelastic law allows us to separate the elastic properties, which were not modified in constipated subjects, from relaxation, which was strongly altered in constipation. This type of investigation could be useful in the management of defecatory disorders, especially when the underlying cause is a disturbance of anorectal function.[18] The method permits the investigation of mechanical rectal-wall properties independently of the rectal volume and shows that the viscoelastic properties of the rectal wall are modified in chronic constipation. Hence, a large rectum in a patient with constipation may be due to 1 of 2 causes: a spastic pelvic floor or an inelastic rectum.

In previous studies, viscoelastic properties of the rectal wall were analysed with the use of a linear model.[13,14] Similar models were proposed for the bladder relaxation curve.[19] These models were based on (1) the analysis of the relaxation function into the sum of exponential functions and (2) the identification of each exponent with the rate constant of a relaxation mechanism. However, the representation of empirical data by a sum of exponential functions is not unique in practice.[20] In this study, we used more simplistic modes: the immediate elastic response, P0[V], was fitted according a linear function and the relaxation function, G[t], according to a mono-exponential function.

To determine the step-stretch law of digestive segments for in vitro studies in a resting state, the spontaneous activity of the tested segment must be suppressed. However, the suppressed state depends on the method of suppression (administering epinephrine or calcium-free EDTA solution, or lowering the temperature less than 20°C) and is not unique.21 The use of drugs in the determination of the viscoelastic properties of biological materials did not allow us to distinguish the pharmacological effects of the drug used from the viscoelastic properties of the biological material.[22] Therefore, these methods are impossible to use in vivo. We proposed a new method that does not involve the use of any pharmacological or physical agents for the determination of the resting state.[15] The resting stress is defined as the inferior envelope of the experimental stress­time curve. The principal advantages of this method are that it can be reproduced and that it can distinguish between the effects of the drug on the mechanical properties (passive or active) of the tested material and the step-stretch response. This method is therefore helpful in clinical investigations.[7] Some of us have used this method to measure the active properties of the sigmoid colon in intraluminal distension.[23] However, in that study, we found that contractions, analysed previously as reactive pressure peaks,[13] varied not only from patient to patient but also for each individual patient; their incidence increased with the volume of distension and their amplitude was slight after normalization of the curve.

Tests of anorectal function have made a major contribution to the understanding of normal and abnormal lower intestine function.[24] The rectum acts as a compliant reservoir for intestinal contents when feces are propelled into it. A previous study has shown a considerable variation of rectal volume in response to distension with a constant preset pressure.[25] Since the variation correlated neither with sex nor with anthropometric data, it was interpreted as reflecting physiological[25] or pathological[26] variations in rectal compliance. Our method of interpretation of the normalized curves limits this factor and decreases the variation due to the variability of the elastic response.[26]

The use of the quasilinear viscoelasticity law for the interpretation of rectal adaptation to distension could have important clinical implications for the exploration of patients with inflammatory bowel disease, fecal incontinence or constipation. In patients with constipation with outlet obstruction,[11] the proposed analysis allows separation of rectal accommodation, G[t], and the immediate elastic response to distension, P0[V].

Normally, the rectum is sensitive to pressure or distension, which in extreme cases may produce pain in some patients. Until now, little practical information has been derived from the study of the viscoelastic properties of the rectum. These properties are nevertheless important, because the recto-anal inhibitory reflex is elicited during rectal accommodation to sustained distension.[5,26] However, the rate of accommodation is reduced in Hirschsprung disease.[1]The method used in this study permits a description of normal and abnormal rectal adaptation to distension and its analysis in terms of a few elementary variables, but it is inadequate to separate the neural and muscular factors involved in rectal adaptation.

In human subjects, clinical as well as pathological evidence has proved the paramount importance of rectal sensation in the physiology of defecation.[25,27] Impaired rectal sensation appears to be a major pathological factor in terminal constipation, encopresis[25] and incontinence.[18] Rectal sensitivity is mediated through the pelvic nerves, but stretch receptors in the puborectalis or pelvic floor muscles are sufficient to transmit the normal urge to defecate.[18] In our study, the volume of distension was lower than or similar to the first volume required to create the first perception of rectal filling. Therefore, the similarity of reduced curves leads us to think that the muscular properties of the rectal wall, rather than the nervous receptors, are involved in the rectal adaptation to distension. This method could therefore be helpful for analysis of rectal properties in chronic colonic pseudo-obstruction[4] or impaired rectal sensation in fecal incontinence[28] or irritable bowel syndrome.[29]

Conclusions

The following conclusions can be drawn from this study.
  1. In vivo adaptation of the rectal wall to distension is similar to that demonstrated in previous in vitro studies.
  2. The quasilinear viscoelasticity law is applicable to the values of pressure, after determination of the passive state.
  3. Children with constipation have different rectal adaptation to distension than control children.

[Table of contents]

References

  1. Arhan P, Devroede G, Danis K, Dornic C, Faverdin C, Persoz B, et al. Viscoelastic properties of the rectal wall in Hirschsprung's disease. J Clin Invest 1978;62:82-7.
  2. Denis P, Colin R, Galmiche J-P, Geoffroy Y, Hecketsweiler P, Lefrançois R, et al. Elastic properties of the rectal wall in normal adults and patients with ulcerative colitis. Gastroenterology 1979;77:45-8.
  3. Devroede G, Vobecky S, Masse S, Arhan P, Leger C, Duguay C, et al. Ischemic fecal incontinence and rectal angina. Gastroenterology 1982;83:970-80.
  4. Loening-Baucke VA, Anuras S, Mitros FA. Change in colorectal function in patients with chronic colonic pseudo-obstruction. Dig Dis Sci 1987;32:1104-12.
  5. Verduron A, Devroede G, Bouchoucha M, Arhan P, Schang J-C, Poisson J, et al. Megarectum. Dig Dis Sci 1988;33:1164-74.
  6. Devroede G, Hemond M. Anorectal manometry: small balloon tube. In: Smith LE, editor. Practical guide to anorectal testing. New York, Tokyo: Igaku-Shoin Medical Publisher, 1990:55-64.
  7. Bouchoucha M, Jais J-P, Arhan P, Landais P, Faverdin C, Pellerin D. Application of the quasilinear viscoelastic law to the in vivo determination of the viscoelastic properties of the rectal wall in normal subjects. Clin Invest Med 1994;17:107-14.
  8. Fung YC. Stress-strain history relations of soft tissues in simple elongation. In: Fung YC, Perronne N, Anliker M, editors. Biomechanics: its foundations and objectives. Englewood Cliffs (NJ): Prentice-Hall, 1972:181-208.
  9. Martelli H, Devroede G, Arhan P, Duguay C, Dornic C, Faverdin C. Some parameters of the large bowel motility in normal man. Gastroenterology 1978;75:612-8.
  10. Arhan P, Devroede G, Jehannin B, Faverdin C, Revillon Y, Lefevre D, et al. Idiopathic disorders of fecal continence in children. Pediatrics 1983;71:774-9.
  11. Martelli H, Devroede G, Arhan P, Duguay C. Mechanisms of idiopathic constipation: outlet obstruction. Gastroenterology 1978;75:623-31.
  12. Tobon F, Reid NCRW, Talbert JL, Schuster MM. Non-surgical test for the diagnosis of Hirschsprung's disease. N Engl J Med 1968;278:188-94.
  13. Arhan P, Faverdin C, Persoz B, Devroede G, Dubois F, Dornic C, et al. Relationship between viscoelastic properties of the rectal wall and anal pressure in man. J Appl Physiol 1976;41:677-82.
  14. Bouchoucha M, Arhan P, Oddou C, Héro M, Persoz B, Pellerin D. Methode d'étude in vitro des propriétés rhéologiques passives de l'intestin vivant. J Biophysique Biomécanique 1987;11:155-60.
  15. Bouchoucha M, Denis P, Arhan P, Faverdin C, Héro M, Devroede G, et al. Morphology and rheology of the rectum in patients with chronic idiopathic constipation. Dis Colon Rectum 1989;32:788-92.
  16. Winner BJ. Statistical principles in experimental design. 2nd ed. New York: McGraw Hill, 1971:298-378.
  17. Meiss RA. Mechanical properties of gastrointestinal smooth muscle. In: Handbook of physiology. Vol I, part I. Bethesda (MD): American Physiological Society, 1989:273-330.
  18. Caruana BJ, Wald A, Hinds JP, Eidelman BH. Anorectal sensory and motor function in neurogenic fecal incontinence. Gastroenterology 1991;100:465-70.
  19. Kondo A, Susset JG. Physical properties of the urinary detrusor muscle. J Biomech 1973;6:141-51.
  20. Riggs DS. The mathematical approach to physiological problems. Baltimore: Williams & Wilkins, 1970:120-67.
  21. Fung YC. Mechanics of soft tissue. In: Skalak R and Chien S, editors. Handbook of bioengineering. New York: McGraw Hill, 1987:1.1-1.12.
  22. Bouchoucha M, Arhan P, Devroede G, Bortolussi M, Héro M, Barbier J-P, et al. Action du trimethyl-3, 3-N-hexene-lactame (THL) sur les propriétés mécaniques des fibres longitudinales du colon de rat. Étude in vitro. Biomed Pharmacother 1987;41:457-62.
  23. Chevalier T, Arhan P, Bouchoucha M, Devroede G, Mignon M, Bonfils S, et al. Étude de la motricité sigmoïdienne stimulée par la distension luminale. Modification de la réponse motrice dans la constipation par ralentissement du transit colique gauche. Gastroenterol Clin Biol 1989;13:245-9.
  24. O'Kelly TJ, Mortensen NJM. Tests of anorectal function. Br J Surg 1992;79:988-9.
  25. Akervall S, Fasth S, Nordgren S, Oresland T, Hulten L. Rectal reservoir and sensory function studied by graded isobaric distension in normal man. Gut 1989;30:496-502.
  26. Rao SSC, Read NW, Davison PA, Bannister JJ, Holdsworth CD. Anorectal sensivity and responses to rectal distension in patients with ulcerative colitis. Gastroenterology 1987;93:1270-5.
  27. Collet L, Meunier P, Duclaux R, Chery-Croze S, Falipou P. Cerebral evoked potentials after endorectal mechanical stimulation in humans. Am J Physiol 1988;193:G477-G482.
  28. Sun WM, Read NW, Miner PB. Relation between rectal sensation and anal function in normal subjects and patients with fecal incontinence. Gut 1990;31:1056-61.
  29. Mertz H, Naliboff B, Munakata J, Niazi N, Mayer EA. Altered rectal perception is a biological marker of patients with irritable bowel syndrome. Gastroenterology 1995;109:40-52.
| CIM: February 1997 / MCE : février 1997 |
CMA Webspinners / >