effect ofgastro-oesophageal reflux oesophageal sphincter motility … · gut1993;34:904-910...

Gut 1993; 34:904-910

Effect of gastro-oesophageal reflux on upper

oesophageal sphincter motility in children

J Willing, G P Davidson, J Dent, I Cook

AbstractMotor events of the upper oesophagealsphincter associated with gastro-oesophagealreflux were evaluated in 53 symptomatic child-ren (median age 13 months) who were studiedrecumbent and unsedated. Children weredivided into four groups according to symp-toms, and then into two groups according tothe presence or absence of neurologicaldeficit. No grouping had basal upper oesopha-geal sphincter pressure that differed signific-andy from any other. Oesophageal distentiondue to gastro-oesophageal reflux, which wasrecognisable as oesophageal common cavityepisodes, was associated with augmentation ofmean basal upper oesophageal sphincter pres-sure from 36*5 (SD 18) mm Hg to 48 5 (18) mmHg (p<00001), irrespective ofwhether gastro-oesophageal reflux caused oesophageal acidi-fication. Abrupt relaxations of the upper oeso-phageal sphincter independent of swallowingand lasting up to three seconds occurredduring 54% of common cavity episodes. Fortynine per cent of these relaxations occurredwithin four seconds after the onset of disten-tion. The oesophageal distention caused bygastro-oesophageal reflux is a potent stimulusof transient upper oesophageal sphincterrelaxations in children. These relaxations are amore likely explanation for oesophagopharyn-geal reflux than defective basal upper oesopha-geal sphincter tone.(Gut 1993; 34: 904-910)

TABLE I Major presenting symptoms

No ofpatients

Without Withneurological neurological

Symptom deficit (n=42) deficit (n-=l l)

Vomiting 27 7Irritability 23 4Recurrent respiratory disease 3 3Failure to thrive 2 2Abdominal pain 4 1Apnoea 5 0Food refusal 7 1Swallowing difficulties 1 I

Some patients presented with more than one major symptom.

gastro-oesophageal reflux or oesophago-pharyngeal reflux. The capacity to monitor UOSpressure for prolonged periods has only recentlybeen developed. Initially, a sleeve sensor wasadapted for this purpose in adults45 and we havenow developed the UOS sleeve sensor so that it issmall enough for use in children.6

In this study we aimed to record for the firsttime patterns of UOS motility associated withoccurrence of gastro-oesophageal reflux inchildren referred for evaluation of suspecteddisorders of oesophageal motility includingpathological gastro-oesophageal reflux. Wesought to investigate the hypothesis thatepisodes of spontaneous gastro-oesophagealreflux cause reflex UOS relaxations, which maybe the mechanism of regurgitation rather than adeficiency of basal UOS tone.

Gastroenterology Unit,Adelaide Children'sHospital, 72 King WilliamRoad, Adelaide, SouthAustraliaJ WillingG P Davidson

Gastroenterology Unit,Royal Adelaide Hospital,North Terrace, Adelaide,South AustraliaJ DentI CookCorrespondence to:Dr G P Davidson,Gastroenterology Unit,Adelaide Children's Hospital,72 King William Road, NorthAdelaide, South Australia5006.

Accepted for publication20 October 1992

Oesophagopharyngeal reflux is a common prob-lem, especially in infants and can have seriousconsequences. Major complications that can

occur include calorie loss due to vomiting withresultant growth failure and respiratory diseasesecondary to aspiration of refluxate.The upper oesophageal sphincter (UOS) is

considered to be the major barrier preventingoesophagopharyngeal reflux. The concept thatoesophagopharyngeal reflux occurs because ofabnormally low basal UOS tone has dominatedthinking about its pathogenesis. ' There isminimal information about patterns of UOSpressure in children with problems believed to

be related to oesophagopharyngeal reflux.Neither Sondheimer2 nor Staiano et al3 foundUOS pressure to be any different in childrenwith gastro-oesophageal reflux compared withthose without. They examined the hypothesisthat basal UOS hypotonia caused oesophago-pharyngeal reflux in an indirect manner. Themethods used did not allow continuous monitor-ing of UOS pressure or examination of motor

events associated with occurrence of either

Patients and methods

PATIENTSFifty five children aged 2 to 81 months (median13 months) were enrolled in the study. Thepatients were referred to the gastroenterologyunit of the Adelaide Children's Hospital forevaluation of symptoms thought to be due togastro-oesophageal reflux or a feeding disorder.All children were fully assessed clinically by oneof us (GPD) or one other consultant pediatricgastroenterologist. The study protocol wasapproved by the ethics research committee ofAdelaide Children's Hospital. Informed parentalconsent was obtained before the study. Table Igives the classification ofmajor presenting symp-toms and incidence of neurological dysfunctionin the 53 children in whom technically satis-factory UOS recordings were obtained.Neurological dysfunction was defined by thepresence ofsymptoms and signs of cerebral palsyor developmental delay confirmed by the Denverdevelopmental screening test.'

904

on Novem

ber 9, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.34.7.904 on 1 July 1993. Dow

nloaded from

http://gut.bmj.com/

Effect ofgastro-oesophageal reflux on upper oesophageal sphincter motility in children

PROTOCOLDual sleeve manometric assemblies were usedfor concurrent monitoring of lower oesophagealsphincter (LOS) and UOS pressures. Assemblieswere passed transnasally without the aid ofsedation after a three hour fast. After positioningof a pH electrode and manometric assembly (seelater) the child was allowed to settle and was thenfed appropriately for age with formula or non-

acid food such as sandwiches and milk. Monitor-ing of spontaneous patterns of motility andoesophageal pH was started at the end of themeal and continued for four hours with thechildren recumbent and unsedated.

Figure 1: Manometricassemblies usedforrecording. Bars=sleevesensors; closed circles=sidehok sensors; opencircles=pH electrodes.

uOS

Sleeve

RECORDING TECHNIQUEWe have described the technique previously.6Briefly, one of two manometric assemblies was

used (Fig 1). These had different intersleevedistances to cope with the range of inter-sphincteric distances found in this age group.

They had an outer diameter of 3-5 mm except forthe proximal sleeve, which had an oval cross

section (3 5mmx 5 - 5 mm). This allowed consist-ent positioning in the UOS in the anterior or

posterior orientation, thus controlling for radialasymmetry of the pressure profile.4 Six side-holes, perfused at 0- 15 ml/min, monitoredgastric and oesophageal body pressures; one

sidehole perfused at 0O07 ml/min monitoredpharyngeal pressures. The two sleeves wereperfused at 03 ml/min. Distal oesophageal pHwas monitored concurrently with a miniatureintraluminal monopolar glass electrode (Micro-electrodes Inc, New Hampshire, USA, modelMI-506), and a skin electrode used as a reference(Micro-electrodes Inc, New Hampshire, USA,model MI-402).The manometric assembly was positioned

such that the side holes at each end of the LOSsleeve showed gastric and oesophageal bodypressure patterns whereas the side holes at eitherend of the UOS sleeve indicated pharyngeal and

LOSSleeve

*@O0o\\\\NN\\\\\\\\\\

* \*\\N \ \

10 5

l

cm

oesophageal body pressures (Figs 1, 2, 3). In twochildren the intersleeve distance was not compat-ible with simultaneous, technically satisfactoryrecording from the UOS and LOS. In thesechildren the assembly was positioned to meet theclinical requirement for adequate LOS mano-

metry. Consequently the data from the 53 child-ren with technically satisfactory UOS recordingswere analysed.

DATA ANALYSIS

Analysis ofmanometric tracingsManometric indicators of reflux - Abrupt,sustained increases in intra-oesophageal pressureto intragastric pressure known as common cavityepisodes, were used to recognise oesophagealdistension by reflux without reference to the pHrecording (Fig 2). Common cavity episodes havebeen described and evaluated in detail else-where,8 and have been shown by fluoroscopy tobe due to oesophageal body distention with gas.9

Common cavity episodes were only scored whenthe rise in intra-oesophageal pressure was

recorded in at least two oesophageal body mano-metric channels. Rises in basal oesophagealpressure due to straining or breath holding wereexcluded by recognition of similar rises in thegastric pressure tracing, whereas true oesopha-geal body common cavity episodes were notassociated with a comparable change of intra-gastric pressure (Fig 2).

Analysis of basal UOS pressure - For thepurposes of this study basal UOS pressure was

defined as the mean UOS pressure after editingof disturbances caused by strains or swallows. Astrain was defined as an abrupt rise in gastricpressure of 5 mm Hg or more above the baselinefor two or more seconds, with a temporallyassociated disturbance ofoesophageal body pres-sure. Swallowing was indicated by characteristicpharyngeal pressure waves in the most proximalsidehole (Figs 2, 3). The reference point fortiming of swallows was taken from the onset ofthis pressure wave. Basal UOS pressure was

referenced to basal end expiratory midoeso-phageal body pressure and analysed in two waysas outlined:

(1) Basal UOS pressure was determined forevery 12th minute throughout the monitoringperiod as described previously.6 Disturbancesinduced by swallowing were edited from theUOS pressure tracing by excluding pressuresfrom four seconds before to six seconds afterswallows. A visual mean of basal UOS pressurewas determined from each edited minute oftracing if there were more than 14 secondsremaining after editing. In view of our recentfinding that basal UOS pressure is influenced bythe child's state of arousal,6 this was assessedthroughout the monitoring period and noted as:

A - resting with eyes shut, B - resting with eyesopen, C - moving, comfortable, D - moving,uncomfortable, and E - crying. The child'sarousal classification was noted for each 12thminute value of UOS pressure and a mean valueobtained for arousal states A and B. Althougharousal states C to E were noted their data were

not analysed for this study.

20I

15

Sleeve sensor

* Side holes

o pH probe

I I I I I

905

on Novem


http://gut.bmj.com

/G


nloaded from

http://gut.bmj.com/

Willing, Davidson, Dent, Cook

Figure 2: Segment ofmanometric tracing showingpharynx, UOS, oesophagealbody, LOS, and gastricpressures. The bottom tracingis intraoesophagealpHrecorded 3 cm above thelower oesophageal sphincter.The first pressure spike in thepharyngeal tracing indicatesa normal swallow initiatinga normal oesophageal bodyperistaltic wavesimultaneous with LOSrelaxation. After thissequence LOS pressure isre-established for about 10seconds before there is atransient LOS relaxationthat is associated with anoesophageal body commoncavity episode andoesophageal acidification.

x mm Hgc 40 -

0

~00n

(_

0.0

0)0

80

0 -

87-

8: 1

8%

80 -

0ii80 _2 <<

80

80-

EI

O

-r7

IL 5]3

cavity episode was taken to indicate a period ofcommunication between the lumina of theoesophageal body and stomach. The percentageof time that such a pattern was seen duringcommon cavity episodes was recorded. Thechange of basal end expiratory intraoesophagealpressure produced by a common cavity episodewas measured from an oesophageal bodychannel. The extent of an oesophageal commoncavity was determined from the number ofsideholes proximal to the LOS that registered thecommon cavity episode.

Transient UOS relaxations - The tracings werescanned for UOS relaxations that occurredindependently of swallowing. This excluded anyrelaxation that had its onset within three secondsbefore or one second after the onset of a swallow.Figure 3 shows the analysis of transient UOSrelaxations. These were scored when UOS pres-sure dropped to less than 50% of the prevailingUOS pressure within one second without theoccurrence of swallowing or straining. In thecase of relaxations that occurred before commoncavity episodes, the reference basal UOS pres-sure was derived from the 10 seconds before theonset of the common cavity episode. For relaxa-tions that occurred during common cavityepisodes, the reference basal UOS pressure wasderived from the UOS pressure values during thecommon cavity. Transient UOS relaxations weretaken as having ended when UOS pressure roseabove 50% of the relaxation pressure.The nadir of transient UOS relaxations was

the most clearcut point and so was related to thetime of onset of the common cavity episode to thenearest second. Nadir pressure was referenced tothe prevailing midoesophageal body pressure inthat second and so took into account changes ofbasal oesophageal body pressure produced bycommon cavity episodes. The duration of thenadir of transient UOS relaxations was measuredto the nearest second.

L0 15 30 45 60 75

Time (s)

(2) UOS pressure around common cavityepisodes was analysed if there were no swallowsfor 13 seconds before and three seconds after theonset of each common cavity episode (Fig 3).Visual mean values of UOS pressure were takenfor each two second interval from 10 secondsbefore the onset of the common cavity episodeuntil one second before the first swallow after thecommon cavity episode, or the onset of the firstsecondary oesophageal peristaltic wave (Fig 3).An overall value of basal UOS pressure was

derived for the six seconds before and up to sixseconds after the onset of the common cavityepisode by averaging the individual two secondvalues. Mean values for pre- and post-commoncavity pressures were then derived for eachchild.

Measures of magnitude, duration, and extent ofoesophageal distention - Occurrence of a patternof intraoesophageal pressure oscillation identicalto that of intragastric pressure during a common

Analysis ofpH recordingAcid reflux was defined as a fall of oesophagealpH to four or less for four or more seconds. ThepH changes associated with each common cavityepisode were evaluated by taking the pH fiveseconds before and five seconds after the onset ofthe common cavity episode. The pH nadirduring the common cavity episode was alsonoted.

STATISTICAL ANALYSISStudent's paired t test was used to compare basalUOS pressure for the six seconds before thecommon cavity to the six seconds after the onsetof the common cavity. This was done by calculat-ing mean deltas ofUOS pressure for each subjectand comparing group mean differences beforeand after the common cavity.

Incidence of transient UOS relaxations amongsymptom groups and between neurologicaldivisions was compared by X2 analysis. Incidenceof transient UOS relaxations was compared withoesophageal acidification, change of oesophagealbody pressure, and extent of communication byX2 tests.

I I I I - - I

906

on Novem


http://gut.bmj.com

/G


nloaded from

http://gut.bmj.com/


Differences among nadir pressures oftransient UOS relaxations occurring before theonset of the common cavity, and the first,second, and third and subsequent transient UOSrelaxations after the onset of the common cavitywere analysed as a single factor experiment withadjustment for unequal sample size."'

All data are expressed as median and inter-quartile ranges unless otherwise stated.

Results

BASAL UOS PRESSURE INDEPENDENT OF REFLUX- I ables 11 and III show the mean values tor basal

UOS pressure according to arousal state, symp-tom classification, and presence or absence ofneurological deficit. No significant differenceswere evident among the different clinical groupsfor arousal states A or B. Mean basal UOSpressure was not significantly different in thepatients with neurological deficit compared withthe neurologically normal children. Consistentwith our previous report6 arousal state influencedUOS pressure irrespective of the clinical group-ing, with UOS pressure being significantlygreater in arousal state B in all cases (p<0-001).

5s

The difference in increase in UOS pressure forthose episodes with acid reflux and those withoutwas compared with one way ANOVA.A pxq factorial experiment with adjustment

for unequal cell frequencies was used for com-parison of basal UOS pressure among symptomgroups, the presence or absence of neurologicaldeficit, and the influence of arousal. An averageof six values was available for estimation of arepresentative cell observation.10TABLE II Relation between symptomatology and basal UOSpressure

Basal UOS pressure (mm Hg)

Arousal ArousalstateA stateB

Symptom group Mean (SD) Mean (SD)

Vomiting/failure to thrive 15-4(11-2) 23-3 (21-1)*(n=20) (n= 14)

Irritability/abdominal pain 16-2(7 8) 23-3 (16-7)*(n=24) (n=18)

Recurrent respiratory disease/ 20-0 (4-3) 26-0 (3 3)*apnoea (n=3) (n=3)

Food refusal/swallowing 12-0(5-6) 29-8 (19-0)*difficulties (n=4) (n=4)

*Significant differences between arousal classifications.For this analysis, patients were grouped without reference toneurological state and according to the most prominent symptom.Values were not available for both arousal states in every childbecause of variation in behaviour pattern. No significantdifferences were found between symptom groups within arousalclassifications.

TABLE iII Relation between neurological deficit and basalUOS pressure

Basal UOS pressure (mm Hg)

Arousal stateA Arousal state BMean (SD) Mean (SD)

Without deficit 16-1 (92) (n=40) 24-1 (18-7)* (n=31)Without deficit 10-7 (9-6) (n= 11) 23-4 (17-7)* (n=8)

*Significant differences between arousal classifications.Values were not available for both arousal groups in every childbecause of variations in behaviour pattern.

BASAL UOS PRESSURE ASSOCIATED WITH COMMONCAVITY EPISODESScreening of the tracings showed 369 commoncavity episodes. The median number identifiedper child was five, with a wide variation fromzero to 29. Of these, 102 episodes fitted thecriteria described in the methods section, whichwere designed to ensure that the effects ofswallowing or straining did not influence assess-ment of the effect of reflux on UOS pressures.These episodes gave a total of 551 two secondtime intervals before the onset of the commoncavity episodes, and 476 two second time inter-

80k

EE

a)

L-

enm

0i

a.

0

U)

* pH +veo pH -ve

60K

40

20K

-6 -4 -2 0 2 4 6

Time (s)Figure 4: Measurement ofUOS pressure in two secondintervalsfrom six seconds before the onset ofthe common cavityepisode to six seconds after the onset. Episodes associated withpH change are shown with filled squares, those with nopHchange are shown with open squares. Values are means (SD)deviations are shown.

mm Hg

X~~~~~~~~~~~~~~~~~~~~~~~I._-LI

* ~~~~~~~~~~~~~~~I.

A-- :-

x

C/)

0

30 -

60 -

30-

* ~~~~~~~~~~~~~I> 60 I10 .0I.0mO 30 - I

0 I

U, X0 0

d e f

Figure 3: Measurement oftransient UOS relaxations.The spikes in the pharynxpressure tracing aregenerated by swallows. a=50% ofUOS pressure,referenced to basal endexpiratory oesophageal bodypressure before the commoncavity onset; b=timing oftransient UOS relaxation;c=prevailing oesophagealbody pressure; d=excludedtime interval after aswallow; e= 10 second timeinterval preceding thecommon cavity onsetfromwhich UOS pressures weremeasured;f=interval afterthe common cavity onsetfromwhich UOS pressures weremeasured. This periodrangedfrom 2 to 42 seconds;g=excluded time intervalpreceding a swallow.

907

on Novem


http://gut.bmj.com

/G


nloaded from

http://gut.bmj.com/

Willing, Davidson, Dent, Cook

pressure before and during the common cavityepisode.

Overall, common cavity episodes caused a25% increase in mean UOS pressure from 36&5(SD 18 mm Hg to 48 5 (18) mm Hg, which washighly significant (p<0-0001). This effect didnot depend on oesophageal acidification as itoccurred to the same extent with both pH' andpH- common cavity episodes. As only threeepisodes ofacid reflux were seen in the absence ofa common cavity episode no statement could bemade about the effects of acidification alone.

L -a...-....... -. -. .--.I.-----

0 5 10 15 20 25

Time (s)

vals after the common cavity onset. The medianduration of the common cavities was eightseconds (interquartile range four to 10 seconds).The median rise in basal intraoesophageal pres-sure by common cavity episodes was six mm Hg(interquartile range 5-8 mm Hg).

Figure 4 shows the pattern of basal UOS

TRANSIENT UOS RELAXATIONSOne hundred and one transient UOS relaxationsoccurred around the time of 60 (54%) of the 112common cavity episodes. Figure 5 shows theirdistribution with time; 49% occurred in the firstfour seconds after the onset of the commoncavity episode and 34% were scattered over the5th to the 27th second after the onset of thecommon cavity. The remaining 17% occurredover the 10 seconds preceding the onset of thecommon cavity episode, being evenly distri-buted through this time. Most common cavityepisodes had a single transient UOS relaxation,but up to five were seen during a single episode.There were 65 first transient UOS relaxations, 24second transient UOS relaxations, and 12 thirdand subsequent transient UOS relaxations.No grouping of the patients according to

symptoms or the presence or absence of neuro-logical deficit showed any difference in incidenceof transient UOS relaxations.

Figure 6 shows the range of individual nadirpressures of transient UOS relaxations beforeand after the common cavity episode. The nadirpressures of the first transient UOS relaxationsthat occurred during the common cavityepisodes differed significantly from the firsttransient UOS relaxations that occurred beforethe common cavity (p<005). The duration ofthe nadir could only be graded into coarsedivisions of less than one second, one to twoseconds, or more than two seconds as the paperspeed at which the recordings were made did notallow greater accuracy. Eighty two per centlasted less than one second, 16% were one to twoseconds, and 2% were more than two seconds in

40

30

IEE

20

10

0

-10

Figure 6: Nadir pressure ofthe transient UOSrelaxations. Bars showmedian values.

3rd

40 -

30 -

I

I20 -

10o

0 *.....

0

0-

-10

1st 2nd

I00

00

It*0

is

0

1 0 I

t I

1 st 2nd 3rd

Before common cavity episode

III .11I '111

30 -

20 -

10 -

0

10 -

0

10 -

0 '

(/)C o

U) mc x

Lro

C C

C

400

G )

U)C C

.- .,p

VL-

Figure 5: Timing oftransientUOS relaxations withrespect to the onset ofthecommon cavity episodes.

. . . . m

908

I I,I v I T I v v v I IIIII

After common cavity episode

on Novem


http://gut.bmj.com

/G


nloaded from

http://gut.bmj.com/


TABLE IV Relation between increasing oesophageal bodypressure and triggering oftransient UOS relaxations

Oesophageal body pressure (mm Hg) 3-5 6-8 9-11 12-21No of episodes with relaxations 19 33 11 2No of episodes without relaxations 18 20 8 1

There was no significant difference between the numbers thattriggered transient UOS relaxations and those that did not.

TABLE V Relation between percentage oftime oesophagealbody pressure and gastric pressure are identical, and transientUOS relaxations

Length of time with communication (%) 0-49 50-99 100No of episodes with relaxations 19 33 11No of episodes without relaxations 18 20 8

There was no significant difference between the numbers thattriggered transient UOS relaxations and those that did not.

duration. There were no significant differencesin the mean duration of the nadirs of transientUOS relaxations according to their timingrelative to the common cavity onset among thesymptom groups.

Oesophageal acidification had no effect ontriggering of transient UOS relaxations as theproportions of acid negative (48 of 84) and acidpositive (14 of 22) episodes that triggeredtransient UOS relaxations did not differ signific-antly.

Neither of the measures of the extent ofoesophageal distention showed any correlationwith triggering of transient UOS relaxations.Table IV shows the range of increases in intra-oesophageal body pressure from the commoncavity episodes. Table V shows the percentage oftime the oesophageal and gastric tracingsindicated identical pressure recordings. In allbut one of the common cavity episodes thepressure increase was seen in all oesophagealbody monometric recording ports.

DiscussionThis report describes for the first time thechanges of UOS pressure that occur duringspontaneous gastro-oesophageal reflux inunsedated children. One major finding in thechildren studied, who had a range of clinicalproblems, was that transient UOS relaxationsoccurred during a proportion of episodes ofgastro-oesophageal reflux induced distention ofthe oesophagus. Another significant finding wasthat there was no difference in basal UOSpressure in children whose problems could beascribed to oesophagopharyngeal reflux whencompared with children in other groups. Thesefindings suggest that lowered basal UOS tone isunlikely to be a major cause of oesophago-pharyngeal reflux, but that transient relaxationsof the UOS in response to distention are a moreprobable cause of such reflux. Our analyses alsosupport further our previous finding that basalUOS pressure is highly labile and emphasise thatthe level of arousal must be recorded and con-trolled for any valid comparisons of basal UOSpressure.The scope ot the evaluations undertaken was

only possible through the use of perfused sleevemanometry of the UOS. The modifications ofthis approach for use in children have beenpresented previously.6 Earlier manometric

studies of the UOS in children2 3 used perfusedsidehole pull through methods with sedation, anapproach that only gives a few sample values ofUOS pressure and these are influenced by theeffects of sedation and stress. Our techniquepermits continuous recording from the upperoesophageal sphincter for periods of severalhours without resort to sedation. This approachenabled us to evaluate UOS function understandard conditions with adequate control forthe influence of level of arousal, and also pro-vided the opportunity to analyse events aroundthe time of occurrence of spontaneous episodesof gastro-oesophageal reflux.

There is a prevailing theory that basal hypo-tonia of the UOS is the cause of pathologicaloesophagopharyngeal reflux. The previouspaediatric studies of Sondheimer' and Staianoet al3 did not support this theory. The sleevesensor recordings of basal UOS pressure in thepresent study produced many hours of UOSpressure recording and allowed us to undertake asearching and carefully controlled evaluation ofthe possibility that UOS hypotonia is related toany particular symptom pattern in children. Thepatients were classified in two ways: firstly intofour groups based on their most prominentpresenting symptom, and secondly according tothe presence or absence of neurological deficit.The group of children with symptoms of vomit-ing (27 of 29) or failure to thrive (two of 29) werethe group whose problems were primarily due toregurgitation. When the level of arousal wastaken into account there was no significantdifference of UOS pressure in this group com-pared with any of the other three symptomgroups. The existence of a neurological deficitwas also shown to have no significant effect onbasal UOS pressure. We therefore support thefindings of Sondheimer2 and Staiano et al3 thatbasal hypotonia of the UOS is not a cause ofpathological oesophagopharyngeal reflux.Unfortunately, in the present study it was onlypossible to make comparisons among sympto-matic children, because ethical constraints pre-cluded measurement of UOS pressure fromchildren who were well.

Because of the perceived significance of aug-mentation of basal UOS pressure as a mechanismthat may prevent gastro-oesophageal reflux, weperformed an analysis of the effects of oesopha-geal distention and acidification on basal UOSpressure. Although there is general agreementthat distention of the oesophagus causes anincrease in basal UOS pressure,'" there is confu-sion as to the effects of oesophageal acidificationon basal UOS pressure.2 12 In the present studydistention of the oesophagus alone caused asignificant increase of 12 mm Hg, and concur-rent acidification had no added effect. Althoughthe post-reflux augmentation of UOS pressure isstatistically highly significant, in absolute termsit is a modest effect and we doubt that it is ofgreat importance for the prevention of oesopha-gopharyngeal reflux. There were only threeepisodes of acid gastro-oesophageal reflux with-out associated oesophageal distention, too few toallow us to comment on the effects of acidifica-tion alone.Our studies were designed to determine

909

on Novem


http://gut.bmj.com

/G


nloaded from

http://gut.bmj.com/

910 Willing, Davidson, Dent, Cook

whether transient UOS relaxations occur duringgastro-oesophageal reflux. Such relaxations areseen in healthy adults after abrupt oesophagealdistention with insufflated gas or gas gastro-oesophageal reflux. They are the basis for theaudible component of belching through oeso-phagopharyngeal passage of gas' and occurindependently of swallowing, having a some-what longer time course than the UOS relaxationinduced by swallowing. Our studies providedevidence of such relaxations during episodes ofspontaneous gastro-oesophageal reflux in child-ren. Transient UOS relaxations were clusteredpredominantly into the first few seconds ofoesophageal distention consistent with the trig-gering of transient UOS relaxations reported byKahrilas et al in adults.5 Acidification of theoesophagus in the presence of distention had noeffect on triggering of transient UOS relaxationsas seen by the proportions of common cavityepisodes with and without acidification that wereassociated with transient UOS relaxations. Thisconfirms a recent report by Vakil et all3 that therewas no correlation between oesophageal acidi-fication and occurrence of transient UOS relaxa-tions in normal volunteers or in patients withoesophagitis. There were no transient UOSrelaxations in the three episodes of acidificationof the oesophagus without associated commoncavity episodes, but this sample is too small toallow for any firm conclusions.The analysis of the magnitude of the increase

of basal intraoesophageal pressure produced bythe common cavity and the percentage of timethat the stomach was in pressure communicationwith the oesophageal lumen were attempts todetermine whether intensity of distention hadany bearing on the triggering of transient UOSrelaxations. Neither of these relatively crudemeasures suggest that there was any correlationof degree or duration of oesophageal distentionwith triggering of transient UOS relaxations.Our results should not be interpreted as exclud-ing this possibility as Kahrilas et al' did find sucha dose response relation in adults, but alsoconsiderable variation in thresholds for trigger-ing oftransient UOS relaxations among differentadult subjects. There is a possibility thattransient UOS relaxations may be unusually easyto trigger in some subjects and so predisposethem to oesophagopharyngeal reflux. Thistheory requires direct examination by correla-

tion of events after spontaneous episodes ofgastro-oesophageal reflux in which the volume ofrefluxate is monitored scintigraphically, or bystandardised testing of thresholds for transientUOS relaxations in infants by air insufflation.Our study has monitored UOS pressure in

unsedated children and established the occur-rence of transient UOS relaxations in children,which may contribute to oesophagopharyngealreflux. Basal UOS pressure is very labile,responding to the child's level of arousal and tooesophageal distention, but not to oesophagealacidification. The prevailing theory that oeso-phagopharyngeal reflux is due to lowered basalUOS tone has not been supported. Definitiveevaluation of the possible causes of oesophago-pharyngeal reflux awaits the recording of UOSpressure during well defined episodes of suchreflux. A smaller catheter to enable us to studyinfants in the first two months of life is eagerlyawaited, as this is the age group in which mostoesophagopharyngeal reflux occurs.

This work was supported by a project grant from the NationalHealth and Medical Research Council of Australia. We thank DrsWD A Ford, T Pouras, and D J Moore for referring their patientsto us.

1 Gerhardt DC, Castell DO, Winship DH, Shuck TJ.Esophageal dysfunction in esophageal regurgitation. Gastro-enterology 1980; 78: 893-7.

2 Sondheimer JM. Upper esophageal sphincter and pharyngo-esophageal motor function in infants with and withoutgastroesophageal reflux. Gastroenterology 1983; 85: 301-5.

3 Staiano A, Cucciara S, De Visia B, Andreotti MR, Auriccio S.Disorders of upper esophageal sphincter motility inchildren. J Pediatr Gastroenterol Nutr 1987; 6: 892-8.

4 Kahrilas PJ, Dent J, Dodds WJ, Hogan WJ, Arndorfer RC. Amethod for continuous monitoring of upper esophagealsphincter pressure. DigDis Sci 1987; 32: 121-8.

5 Kahrilas PJ, Dodds WJ, Dent J, Wyman JB, Hogan WJ,Arndorfer RC. Upper esophageal sphincter function duringbelching. Gastroenterology 1986; 91: 133-40.

6 Davidson GP, Dent J, Willing JK. Monitoring of upperoesophageal sphincter pressure. Gut 1991; 32: 607-11.

7 Frankenburg WK, Dodds JB. The Denver DevelopmentalScreening Test. J Pediatr 1967; 71: 181.

8 Dent J, Dodds WJ, Freidman RH, et al. Mechanism of gastro-oesophageal reflux in recumbent asymptomatic humansubjects. J Clin Invest 1980; 65: 256-67.

9 McNally EF, Kelly JE, Ingelfinger FJ. Mechanism of belch-ing: effects of gastric distention with air. Gastroenterology1964; 46: 254-9.

10 Weiner BJ. Statistical principles in experimental design. 2nd ed.Kogakusha: McGraw-Hill, 1962, 210-18 and 445-9.

11 Gerhardt DC, Shuck TJ, Bordeaux RA, Winship DH. Humanupper esophageal sphincter pressure, response to volume,osmotic and acid stimuli. Gastroenterology 1978; 75: 268-74.

12 Kahrilas PJ, Dodds WJ, Dent J, Haeberle B, Hogan WJ,Arndorfer RC. Effect of sleep, spontaneous gastro-oesophageal reflux and a meal on upper esophageal sphincterpressure in normal human volunteers. Gastroenterology 1987;92: 466-71.

13 Vakil N, Kahrilas PJ, Dodds W, Vanagunas A. Absence of anupper esophageal sphincter response to acid reflux. Am JGastroenterol 1989; 84: 606-10.

on Novem


http://gut.bmj.com

/G


nloaded from

http://gut.bmj.com/

effect ofgastro-oesophageal reflux oesophageal sphincter motility … · gut1993;34:904-910...

Documents