association between gastroesophageal reflux and pathologic apneas in infants.pdf

REVIEW ARTICLE

Association between gastroesophageal reflux and

pathologic apneas in infants: a systematic review

MARIJE J. SMITS,* MICHIEL P. VAN WIJK,* MIRANDA W. LANGENDAM,† MARC A. BENNINGA* & MERIT M. TABBERS*

*Department of Pediatric Gastroenterology and Nutrition, Emma Children’s Hospital, Academic Medical Center, Amsterdam,

The Netherlands

†Dutch Cochrane Centre, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands

Key Messages

• There is insufficient evidence for an association between gastro esophageal reflux (GER) and apneas in infants.

High quality studies using uniform inclusion criteria, definitions according to accepted guidelines, and patient

relevant outcome measures are needed.

• The purpose of this systematic review was to determine whether an association between GER and apneas in

infants exists.

• PubMed, EMBASE, and Cochrane databases were searched for studies assessing the relationship between GER

and apneas in infants by means of simultaneous monitoring of GER and apneas.

• One of six included studies found an increase of apneic events after GER, the remaining 5 studies did not find an

association. Two studies assessed apnea followed by GER as well, but did not find sufficient evidence for an

association.

Abstract

Background In infants, apneas can be centrally med-

iated, obstructive or both and have been proposed to

be gastroesophageal reflux (GER) induced. Evidence

for this possible association has never been systemat-

ically reviewed. Purpose To perform a systematic

review using PubMed, EMBASE and Cochrane data-

bases to determine whether an association between

GER and apnea in infants exists. Studies with n ≥ 10

infants, aged <12 months, were included. GER had to

be studied by pH-metry or pH-impedancemetry. GER

episodes were defined as pH <4 for ≥5 s and/or a drop

of >50% of baseline in impedance signal in distal

channels. An apneic event was defined as a cessation

of breathing for >20 s, or ≥10 s with hypoxemia or

bradycardia. An epoch of ≤2 min was used to define

temporal relation between GER and apnea. Method-

ological quality of studies was assessed with Newcas-

tle Ottawa Scale (NOS). Of 1959 abstracts found, 6

articles met the inclusion criteria. All studies had poor

methodological quality. A total of 289 infants were

included. The temporal association of GER followed

by apnea was assessed in all studies, with epochs

varying from 10 s to 2 min. One study found an

increase of apneic events after GER, the remaining 5

studies did not find an association. Two studies

assessed apnea followed by GER as well, but did not

find sufficient evidence for association. This system-

atic review showed insufficient evidence for an

association between GER and apneas in infants. High

quality studies using uniform inclusion criteria,

Address for CorrespondenceMarije Smits, Department of Pediatric Gastroenterology andnutrition, Emma Children’s Hospital, Academic MedicalCenter, Room C2-312, Meibergdreef 9, Amsterdam 1105 AZ,The Netherlands.Tel: 0031-20-5665270;e-mail: [email protected]: 17 April 2014Accepted for publication: 27 June 2014

© 2014 John Wiley & Sons Ltd 1527

Neurogastroenterol Motil (2014) 26, 1527–1538 doi: 10.1111/nmo.12405

Neurogastroenterology & Motility

definitions according to accepted guidelines, and

patient relevant outcome measures are needed.

Keywords apnea, cardiorespiratory monitoring, gas-

tro esophageal reflux, infants, pH-impedance metry,

pH-metry.

Abbreviations: GERD, Gastroesophageal reflux dis-

ease; GER, Gastroesophageal reflux; LCR, Laryngeal

chemical reflexes; LES, Lower esophageal sphincter;

NOS, Newcastle Ottawa Scale; NTS, Nucleus tractus

solitarius; pH-MII, pH impedancemetry; RI, Reflux

index; TLESR, Transient lower esophageal sphincter

relaxation; RAAP, reflux associated apneas.

INTRODUCTION

Gastroesophageal reflux (GER) is the passivemovement

of gastric contents into the esophagus. This occurs

primarily during transient lower esophageal sphincter

relaxation (TLESR).1,2 TLESR is a relaxation of the lower

esophageal sphincter (LES) not preceded by a swallow

and is mediated by a vago-vagal pathway.3

Apnea is the cessation of respiratory airflow and

can be central (no respiratory effort), obstructive

(respiratory effort present but not resulting in airflow

due to upper airway obstruction) or mixed in nature

(no respiratory effort followed by obstruction-like

respiratory efforts).4,5 Clinically significant apnea lasts

>20 s, or for a period of ≥2 breaths accompanied by

significant desaturation, hypoxemia or bradycardia.4,6

The Nucleus Tractus Solitarius (NTS) in the brain-

stem plays a large role in pathogenesis of apneas by

expressing a paradoxical overriding inhibitory

response to signals from peripheral afferents. These

afferents are excited by, for example, hypoxia, hyper-

capnia, and stimulation (e.g. obstruction) of the upper

airways and may lead to respiratory pauses of variable

duration.7,8

Apneas, irrespective of their nature, have been

proposed to be GER induced. The persuasion that

GER and apnea are causally related is embedded in

guidelines for the treatment of GER disease (GERD)

and pathological apneas.9 It is hypothesized that GER-

induced esophageal distension as well as pharyngeal

penetration of GER can activate local vagal stretch and

chemical receptors.10,11 In animal models, distension

of the esophagus can indeed induce apnea and it is

established that laryngeal chemoreceptor reflexes

(LCR) can trigger central apneas in infants.12–14 Thus,

vagal afferents triggered by GER seem to be able to

induce apneas through pathways mediated by the brain

stem. However, a recent study found no effect of feed

thickening, which significantly reduced GER episodes

reaching the proximal esophagus, on the frequency and

duration of apnea episodes.15

Another hypothesis is that apneas cause GER

episodes.16 In a retrospective physiologic study, it has

been shown that all clinically relevant apneas (>20 s)

identified, were followed by a decrease in LES pressure.

Although only manometry was performed in this

study, and occurrence of GER episodes was not

assessed, this decrease in LES pressure putatively

allows GER to occur after apnea.17 This likelihood

further increases when abdominal pressure rises, e.g.,

during arousal and straining after apnea. Furthermore,

obstructive respiratory efforts and repetitive swallow-

ing are long known manifestations of previously

mentioned LCR.18 Apart from inducing apneas, LCR

also causes decrease in LES pressure. LCR can elicit

coughing (as occurs after arousal from apnea) which

increases abdominal pressure, and therefore the chance

of GER occurring in the presence of low LES

pressure.12

Finally, given the close proximity of the GER pattern

generator and breathing control centers in the brain

stem, it could be hypothesized that GER episodes and

apneas temporally relate as result of a common central

pathway.

A substantial amount of studies and non-systematic

reviews have attempted to pinpoint the relationship

between GER and apneas.16,19,20 However, their con-

clusions contradict each other. Therefore, we system-

atically reviewed evidence for an association between

GER and apneas, and vice versa, in infants.

METHODS

Search strategy

We searched Medline, Embase, Cochrane electronic database, andCochrane Controlled Trials Register for prospective studies andsystematic reviews from 1980 up to June 2014. The following keyterms were used: gastro(-o)esophageal reflux, apn(o)ea, ApparentLife Threatening Event, and infant. No language restriction wasapplied. Obtained reviews and articles were hand searched foradditional studies. The full search strategy is available from thecorresponding author.

Inclusion criteria

Two reviewers (MS, MvW) independently judged articles as shownin Figure 1. Full length articles were included if the study met ourinclusion criteria: (i) Prospective study or case–control study, (ii)n > 10 study subjects per patient group investigated with acorrected age <12 months and suspect for GER related symptoms,GERD, apnea and/or apparent life-threatening events (ALTE), (iii)Simultaneous GER investigation (pH-metry or pH-impedance

© 2014 John Wiley & Sons Ltd1528

M. J. Smits et al. Neurogastroenterology and Motility

(pH-MII) measurement and/or manometry) and apnea assessment(cardiorespiratory monitoring), (iv) Apneic events defined as acessation of breath >20 s or ≥10 s with significant desaturationand/or bradycardia, (v) GER events were defined as pH<4 for ≥5 sand/or a drop of >50% of baseline in impedance signal in the distaltwo channels, (vi) Epoch (time window) between GER and apneaor vice versa was defined ≤2 min, (vii) Study aim was to determineany relation between GER and apnea.

Quality assessment

Before data extraction, included full length articles were judgedby two assessors (MS and MvW) for quality, using the Newcas-tle Ottawa Scale (NOS). This scale is a validated tool for scoringthe methodological quality (risk of bias) of comparative obser-vational studies.21,22 As the NOS is aimed at comparativestudies and most included studies were expected to be patientseries, we adjusted the outcome question: ‘was follow up timelong enough for the outcome to occur?’ into ‘was study timelong enough to asses outcomes?’. Also, only in studies incor-porating two or more groups, the latter were divided into anexposed and non-exposed cohort. Definition of an exposedcohort was chosen as defined by authors (e.g. treatment/ notreatment or presence/absence of symptoms at start of study).Similarly, outcome definitions were chosen as defined byauthors (e.g., a decrease in GER, apnea and/or symptoms).Comparability of cohorts was only scored for studies with twoor more patient groups. Furthermore, we added three scoringitems that could be answered with ‘A’, ‘B’ or ‘C’. This modifiedversion of the NOS consists of 11 questions (Table 1). Depend-ing on the answer, questions are rated with one star per item,mapping the methodological quality of the article. There are nocutoff values for this scale, therefore no summary scores werecalculated. In general, the more stars, the higher the quality ofthe article (range 0–11 stars).

Data extraction

Two assessors (MS, MvW) independently performed structureddata extraction. The data extracted from each article includedauthor and year of enrollment, study setting, methods, type andnumber of subjects, method of GER and apnea assessment,potential follow-up, outcome measures, and results. If disagree-ment between the two reviewers existed, consensus was foundwhere possible, or a third reviewer (MT) made a final judgment.

RESULTS

Literature search and quality assessment

The search strategy generated 1959 titles. No system-

atic reviews were found. Based on title and abstract, 41

studies were selected to potentially meet inclusion

criteria (Fig. 1). After retrieving full text articles and

scoring in- and exclusion criteria, an additional 35

articles were excluded. Reasons for exclusion were: (i)

n < 10 patients included (n = 3). (ii) Definition of GER

and/or apnea did not meet the inclusion criteria or no

definitions were given (n = 20). (iii) Duration of epochs

used to define a temporal association between GER and

apnea, or vice versa, was unclear or >2 min (n = 9). (iv)

Retrospective study or non-systematic review (n = 2)

and (v) other (n = 1).

Finally, six studies could be included for our

systematic review.23–28 Of these, two had a case–

control design,24,25 the other four were patient series.

Data from 289 infants (five studies reported gestational

age: range 24–43 weeks, one reported postnatal age:

range 1–34 weeks23) were included. Studies were con-

ducted in tertiary centers in The United States of

America23,24,26–28 and Europe.25

The two assessors initially agreed on 56 of 66 scored

quality items, with a Cohens Kappa agreement of 0.77

(0.61–0.8 = substantial agreement). Agreement after

discussion was reached in 100% of cases. NOS scoring

items per article are described in Table 1. Overall

quality was considered to be poor. Outcome assess-

ment was blinded in none of the studies and only three

of six define outcome measures. Since only two of six

studies studied >1 patient group, not all NOS scoring

items were applicable for the other four studies.24,25

Due to heterogeneity in design, outcome measures,

and analysis, pooling of results was impossible. There-

fore, studies are discussed separately.

Patient demographics

All but one study included prematurely born infants

(Table 2).23 In general, included infants were suspected

of having apneas and/or GER related symptoms. One

study included a group of ‘symptomatic’ (apneas and/or

bradycardias present or GER after feed) and a group of

‘asymptomatic’ infants25 (no GER symptoms or apneas

present). One study assessed infants after near miss

sudden infant dead syndrome (SIDS).23

GER assessment

All studies using esophageal measurements had a

metric cutoff to determine the presence of GER (e.g.

a pH drop below 4 for 5–15 s; Table 3).23–27 The

study using pharyngeal pH-metry defined GER as

appearance of gastric contents into the mouth,

supported by pH changes measured with pharyngeal

recording.28

Apnea assessment

Ariagno et al. did not report the absolute number of

apneas but the number of apneas per hour (Apnea

Index; Table 3).24 Di Fiore et al. also did not report on

absolute numbers, but displayed results graphically in


Volume 26, Number 11, November 2014 Systematic review association infant GERD and apnea

the article.26 The other four studies reported on

overall numbers of apneas that may or may not be

classified into central, obstructive and/or mixed or

prolonged and short. For this review only prolonged,

presumed clinically significant, apneas were consid-

ered. De Ajuriaguerra et al. assessed both symptom-

atic and asymptomatic infants and found only one

asymptomatic infant to have apneas (mixed and

obstructive), all other apneas were recorded in symp-

tomatic infants.25

Table 1 Newcastle Ottowa Scale

Ariagno

et al.23Ariagno

et al.24de Ajuriaguerra

et al.25 Di Fiore 26 Di Fiore 27 Menon 28

Selection

(1) Representativeness of the exposed cohort C B* B* B* B* B*

(a) Truly representative of community*

(b) Somewhat representative of community*

(c) Selected group

(d) No description

(2) Selection of the non-exposed cohort na A* A* na na na

(a) Same community as exposed cohort*

(b) Drawn from a different source

(c) No description

(3) Ascertainment of exposure na A* D na na na

(a) Secure record (eg surgical records)*

(b) Structured interview*

(c) Written self report

(d) No description

(4) Outcome not present at start of study na A* na na na na

(a) Yes*

(b) No

Comparability

(1) Comparability of cohorts (design/analysis) na B* B* na na na

(a) Study controls for epoch between GER and Apnea

and vice versa*

(b) Study controls for GER and Apnea*

Outcome

(1) Assessment of outcome E E E E E E

(a) Independent blind assessment*

(b) Record linkage*

(c) Self report

(d) No description

(e) Independent not blinded assessment

(2) Study time long enough A* A* A* A* A* B

(a) Yes*

(b) No

(3) Adequacy of follow-up of cohorts A* A* A* A* A* A*

(a) Complete follow-up*

(b) Subjects lost to follow-up unlikely to introduce bias

(c) Follow-up rate <20% and no description of those lost

(d) No statement

Added items by authors

(1) Were in- and exclusion criteria specified? A* A* B A* A* B

(a) Yes*

(b) Only inclusion criteria specified

(c) No criteria specified

(2) Definitions of outcomes given? A* B B A* A* B

(a) Yes*

(b) No

(3) Was the outcome assessor blinded? C C C C C C

(a) Yes*

(b) No

(c) Unclear

Quality of articles using Newcastle Ottowa Scale including added items by the authors. Depending on the answer, questions are rated with a star.

There are no cutoff values for this scale. In general, more stars mean a higher methodological quality.



Association of GER and apnea

GER causing apnea (n = 6) Only one study found a

significant increase of apneic events after GER,28 the

other five studies did not find an association (Table 4).

Of the latter five, Ariagno et al. report one of 45 infants

to have an apneic event associated with GER (<2 min

following GER). In this infant, 24 of the total recorded

study apnea (n = 31) were recorded.23 De Ajuriaguerra

reported one apnea occurring <1 min after a GER

episode. In this study, no correlation between apnea

and GER was found in either the symptomatic infant

group (n = 14) or the entire study population (n = 20).

No further details for association in the six asymp-

tomatic infants were reported.25 The two studies by Di

Fiore et al., found neither apneas nor cardiorespiratory

events to be associated with GER, nor did GER

increase the duration of apnea episodes.26,27

Ariagno et al. (2001) reported reflux associated

apneas (RAAP) in 24 infants, i.e., the number of apneas

occurring within 1 min after a GER event, which were

found in 62% of infants. Cisapride treatment adminis-

tered in this study did not reduce RAAP significantly.

Authors stated that, although acid GER and mixed

apneas were reduced in the cisapride group, not enough

evidence was present to claim GER is causally related

to apnea, nor that treatment of GER related symptoms

will significantly improve management of infants with

persistent apneas.24

Apnea causing GER and assessment of common

etiology (n = 2) Di Fiore et al. assessed GER induced

by apneas. Nine percent of GER events were preceded

by a cardiorespiratory event.27 The other study from Di

Fiore et al. found 85% of apneas related to GER were

followed by arousal within 30 s and significant more

apneas of >10 s occurred before or during GER as

compared to apneas immediately after GER. Authors

concluded these findings suggest arousal is associated

with termination of GER, rather than apneas are able

to induce GER.26 None of the included studies specif-

ically searched for a common etiology of GER and

apnea.

Medical therapy in studies (n = 4) Three out of six

studies reported on medical therapy that infants

received during the study, whereas apnea therapy was

abated prior to the commencement of one study.25

Ariagno et al. (2001) found that 8 days of cisapride

therapy significantly reduced the number of GER

episodes of ≥5 min (p = 0.026), the RI (p = 0.017) and

overall mixed apneas compared to baseline, but it did

not reduce RAAP.24 The effect of administered raniti-

dine or metoclopramide on GER or apnea parameters

was not reported in the studies by Di Fiore et al.26,27

There was no significant difference in the number of

GER events (p = 0.34) or reflux index (RI, p = 0.94)

between infants with or without caffeine and/or the-

ophylline therapy.26,27 Accordingly, Ariagno et al.

(2001) stated that caffeine therapy had no effect on

GER parameters (no p-values reported).24

DISCUSSION

This systematic review shows insufficient evidence for

any association between GER and apneas in infants.

Figure 1 Flow chart of included studies.



Tab

le2

Patientdem

ograp

hics

Author

Subjects(n)

Age

Male,

n(%

)Inclusioncriteria

Exclusioncriteria

Med

icationduring

study

Recording

(hours)

Atbirth

(GA

weeks)

Atstudy

(weeks)

Ariagnoetal.23

45

-10�

8PNA

25(56%)

NearmissSID

Sterm

infants

Failure

tothrive

nr

17.95�

3.25

Vomiting

Recurren

tpneu

monia

Ariagnoetal.24

24

28.8

�3.1

35.6

�4GA

16(67%)

Hospitalized

infants

suspect

forGER,ap

nea

and/orfeed

ing

intolerance.

Neg

baselinepH

study(n

=12)*

Caffeine‘

n=12before

Cis

n=9afterCis

†

2reco

rdings

of15-16

PositivepH

study*

Surgeryduringtreatm

ent(n

=1)

8day

stherap

ywithCis

(0.10mg/kg/6hrs)

Rep

eatmeasu

remen

t>8day

s

afterstartCis

(n=4)

Nasal

O†su

ppl

n=12before

Cis

n=12afterCis

PSG/pH

studybefore

+afterCis

deAjuriaguerra

etal.25

20

31(27.5-36.5)

38.7

(73-43)PCA

11(55%)

Hospitalized

premature

born

infants

=/>

37weeksPCA

nr

Caffeineab

ated

>7day

sprior

tostudy

6.18

(3.95-6.80)

Symptomatic:ap

neasan

d/or

bradycardiasorregu

rgitation

afterfeed

(n=14)

Asymptomatic

infants

(n=6)

DiFiore

etal.26

119

28�

237�

4PCA

nr

Hospitalized

premature

infants

Congenital

malform

ations

Xan

thinen=26‡

12overnight

SymptomsofGER

<1500gat

birth

Nasal

O†su

ppl

n=40

Apneas,

bradycardia

and/or

desaturations

Ran

itidinean

d/or

Metoclopramiden=15

DiFiore

etal.27

71

29.4

�3

37.3

�2.6

GA

43(61%)

Hospitalized

premature

infants

<44weeksat

measu

remen

t§Congenital

malform

ations

Caffeinen=5¶

12overnight

Referredforcardiorespiratory

andGER

overnightmonitoring

Needforven

tilarsu

pport

Ran

itidinean

d/or

Metoclopramide

n=3

Men

onetal.28

10

28-40

1-4

PCA

nr

Hospitalized

infants

nr

nr

1.4

(n=6)

2-3

(n=4)

Postfeed

ingregu

rgitation

Idiopatic

prolongedap

nea/A

OP

*Refl

uxindex

within

2SD

ofmeanacco

rdingto

criteria

Van

den

plasetal.9†Noeffect

onGER

values.‡EffectonRInsp=0.94.§<34weeksGA

atbirth.¶ N

oeffect

onnumber

ofGER

even

ts(ns,

p=0.34).PNA,postnatal

age;

PCA,postco

nceptional

age;

GA,gestational

age;

SID

S,su

dden

infantdeath

syndrome;

PSG,polysomnograp

hy=cardiorespiratory

monitoring;

GER,gastro

esophageal

reflux;AOP,ap

nea

ofprematurity;Cis,cisapride;

nr,notreported

.



Tab

le3

GER

andap

nea

param

eters

Author

GER

Apnea

GER

test

GER

andGERD

defi

nition

Number

ofGER

episodes,

(%ofall)/R

IApnea

test

Apnea

defi

nition

No.Ofap

nea

orApnea

Index

pH

MII

GERD

TypeApnea

Duration

+ad

dition

Ariagno

etal.23

pH-m

etry

pH

<4≥1

5s

na

nr

n=356

Respeffort

ECG

ApC

>10s

n=341*

Ariagno

etal.24

pH-m

etry

pH

<4≥5

sna

RI>2SD

†Before

Cis/24h

(mean,SD)

After

Cis/24h

(mean,SD)

p-value

Respeffort

Nasal

airfl

ow

ECG

O2saturation

AIbefore

Cis

AIafterCis

p-value

n=91.7

�45.3

n=73.4

�65.3

0.068

Short

ApC

10–1

5s

0.62�

0.89

0.54�

0.68

0.4

RI16.6

�15.2

RI9.1

�8.4

0.017

LongApC

>15s

0.36�

0.70

0.41�

0.89

0.3

Obstructive

>10s+Resp

effort

0.09�

0.15

0.04�

0.14

0.2

ApM,ApC,

ApO

>3s+10s

0.5

�0.47

0.3

�0.36

0.026

de Ajuriaguerra

etal.25

pH-m

etry

pH

<4≥1

5s

na

RI>10.4%

n=134

Respeffort

Nasal

airfl

ow

Heart

rate

O†

saturation

ApC

>10s

n=26

GERD

in

symptomatic

infants:

n=10(71%)

ApO

+ApM

>10s+Resp

effort

n=113

GERD

in

asymptomatic

infants:

n=2(22%)

Isolated

Bradycardias

<80bpm

n=46

DiFiore

etal.26

pH-m

etry

pH

<4≥5s

na

RI (cutoff

nr)

n=6255

Respeffort

ECG

Heart

rate

O†saturation

ApC,ApO,

ApM

>10s

Noex

actnumbers‡

pH

<4per

12h

NoGER

episodes

/12h:n=53�

41

17�

14%

ApC,ApO

orApM

>15s

Noex

actnumbers‡

DiFiore

etal.27

pH-M

IIpH<4

+50%

fall

MIIbaseline

2distal

chan

nels

nr

n=1183(29%)

Respeffort

ECG

Heart

rate

O†saturation

ApC

>10s

n=1419

WA

pH>4+

DpH

>1

n=341(8%)

ApO

>10s

n=657

NA

pH>4

n=1709(41%)

ApM

>10s

n=42

Acidonly

pH

<4≥5s

+noMII

even

t

n=931(22%)

Bradycardia

<80bpm

n=456

O2 desaturation

<85%

n=10383

CR

even

tsAp+brady+

desaturation

n=12957

Men

on

etal.28

Pharyngeal

pH-m

etry

Visibilityof

GER

in

mouth,DpH

=su

pport

na

nr

n=44

Respeffort

Nasal

airfl

ow

ECG

Prolonged

>20s

n=100

Short

<20s+Brady

<100bpm

See

commen

t§

*In

46measu

remen

ts(1

subject

studiedtw

ice).†Asco

mpared

tostan

dardvalues

Refl

uxindex

acco

rdingto

Van

den

plasetal.33‡Apneas>

15san

d>10s30sbefore,afteran

dduringGER

inarticle

wereprovided

inahistogram

(noex

actfigu

res).§Notmatch

ingourap

nea

defi

nition,therefore

notdisplayed

here.

RI,refluxindex

(%oftimeesophagealpH<4);GER,gastroesophagea

lreflux;GERD,

GERdisease;WA,weakly

acid

GER;NA,non-acidGER;Cis,CisparideRespeffort

=respiratory

effort;Ap,ap

nea;ApC,central

apnea;ApO,obstructiveap

nea;ApM,mixed

apnea;AI,ap

nea

index

:

noofap

neasper

hour;

CR

even

t,cardiorespiratory

even

t;ECG,electrocardigram;DpH,ch

ange

inpH;brady,bradycardia;bpm,beats

per

minute;desat,desaturation;nr,

notreported

;na,

not

applicable;SD,stan

darddev

iation.



Most available studies addressing this topic are meth-

odologically inadequate in terms of sample size,

design, outcome measures, and analysis, or use tech-

niques that are no longer state of the art. Therefore, we

only included a small number of studies with clinically

relevant criteria for GER and apnea that met interna-

tionally accepted standards and had a clear definition

of temporal relation, with epochs of 2 min maxi-

mum.4,9,29 Even with these strict inclusion criteria,

data from the six included studies were too heteroge-

neous to be pooled.

In general, all included studies have methodological

limitations. Firstly, not all GER events are same in

terms of acidity, proximal extent, and volume and

could, therefore, trigger other neuroregulatory mecha-

nisms through different mechano- and chemorecep-

tors. Five of six included studies used pH-metry to

evaluate GER instead of pH-MII technology, which

enables differentiation between these different types of

GER.27 pH-metry, as compared to pH-MII measure-

ments, misses out on non-acid (pH >7) and weakly acid

(4< pH ≤7) GER episodes, while the latter have been

shown the predominant type of GER in infancy.30–32

Secondly, all but one study25 did not categorize the

apneas into central, obstructive or mixed for their

analysis. It might be that central and obstructive

apneas are differently related to GER, as immaturity

of the central nervous system and LCR, respectively,

are considered the most likely underlying mechanism

in infants.

Thirdly, administration of medication modulating

GER and apnea could have influenced results of the

included studies. This has been shown in several

pediatric and adult studies.33,34

Nevertheless, only three studies did report on the

effect of medication administered.24,26,27 Of those,

Cisapride is unavailable nowadays due to associated

cardiac side effects and Metoclopramide is no longer

Table 4 Association of GER and apnea

Author Epoch

GER followed by apnea Apnea followed by GER Association

GER-Apnea

according to

authors?Measure Outcome Measure Outcome

Ariagno et al.23 ≤2 min Aps after GER/ total Aps 1/341 (0.2%) - No

Aps during GER/ total Aps 31/341 (9.1%)

Ariagno et al.24 ≤1 min Infants with ≥1 Ap after

GER (RAAP)*14 (62%) - Not enough

evidence

de Ajuriaguerra

et al.25≤1 min Aps after GER/ total Aps 1/139 (0.7%) - No

ApsC during GER/total ApsC 2/26 (7.7%)

ApsM+ApsO during GER/

total ApsM+ApsO

11/113 (9.7%)

Brady after GER/ total Brady 8/46 (17.4%)

Brady during GER/ total Brady 1/46 (2.1%)

DiFiore et al.26 ≤30 s Aps of >10 s before, during or

after acid GER/ total GER

532/ 6255 (8.5%) Number of Aps >10 s

before, during or

after acid GER

More Aps before

+ during GER

compared to

during/after

GER (p = 0.01)†

No

Aps >15 s before, during or

after acid GER/ total GER

64/6255 (1%) Number of Aps >15 s

before, during or

after acid GER

Trend toward

more Aps before

+ during GER

compared to

after GER

Exact figures +significance nr

Infants with ≥1 GER episode

associated with Ap >10 s/

total infants

83/119 (69.7%)

Infants with ≥1 GER episode

associated with Ap >15 s/

total infants

27/119 (23%)

DiFiore et al.27 ≤30 s Aps after GER/all Aps 72/2118 (3.4%) GER after CR event 378/4164 (9%) No

CR events after GER/ all

CR events

375/12 957 (<3%)

Menon et al.28 10 s post-GER

periods

Aps in after-GER periods/

all Aps

8/100 (8%) - Yes

Mean freq Aps in post-GER

periods vs freq Aps during

GER free periods

p < 0.05†

*RAAP, reflux associated apnea, Cisapride treatment reduced RAAP (not significant s, p = 0.12). †No exact numbers or calculations reported. GER,

gastroesophageal reflux; Ap, apnea; Aps, Apneas; ApsC, central apneas; ApsO, obstructive apneas; ApsM, mixed apneas; CR event, cardiorespiratory

event (including all apnea, solitary bradycardia and solitary oxygen desaturations); nr, not reported.



advocated for infant GERD after being linked to tardive

dyskinesia.35,36 From the studies reporting on medica-

tion, apnea therapy (caffeine and/or theophylline)

appeared to have no effect on GER parameters, nor

did anti-GER medication seem to influence GER

related apnea, implying GER and apnea are not asso-

ciated. Still, considering the lack of data provided, no

firm conclusions can be drawn.

Finally, statistical analysis of the association

between GER and apneas was very poor across

included studies. None of the studies used a GER or

apnea specific association index, whereas mean values

and standard deviations were sparsely provided. This

makes interpretation of the data difficult. Although

current available association indices for GER-related

symptoms all have their limitations, it is indispensable

to prove the presence of association between GER and

apneas.37 New association indices specifically aimed at

the relation between GER and respiratory events might

proof suitable alternatives for future research. How-

ever, no golden standard exists yet.29,38

One study that showed a positive association

between GER and apnea, used pharyngeal pH-metry

to assess GER.28 Due to its position, the catheter is

unable to detect distal GER events not reaching the

pharynx, which can theoretically trigger apnea through

esophageal distension and subsequent vagal activation.

In addition, simultaneous recording time of GER and

apneas in the study of Menon et al. was very short,

with a mean of 1.4 h for six children and 2–3 h for the

remaining four infants. For clinical use, a minimum of

18 h recording of GER is commonly used to reliable

associate symptoms with GER.39 Recently, the influ-

ence of recording time on the association found was

clearly shown, confirming the need for prolonged

monitoring.40

In general, GER is universally present in infants,

with 50% of infants of 0–3 months having visible

regurgitation at a daily basis. Normative values of the

total amount of acid and weakly/non-acid GER not

reaching the mouth are unknown, but the amount of

distal acid GER episodes alone in healthy infants can

reach up to 56 episodes. Apneas are a relative rare

phenomenon in mature infants, especially after the

first days of life. With a large physiological amount of

GER episodes (further enhanced by posture, frequent

liquid feeds, and anatomical properties of the infants

upper gastrointestinal tract) and little apneas, associa-

tion can be found as a result of chance. Especially with

small sample sizes and with inadequate definitions of

GER, apneas and association indices. Moreover, even if

a causal relation exists, a reduction in apneas does not

necessarily reduce the number of GER events signifi-

cantly, because most GER events will occur simply

because all infants, including completely healthy

babies, experience GER events frequently. In addition

to the paucity of high quality evidence on this topic, a

recent study even states GER is associated with

interruptions of sleep in infants (arousals), which

implies nocturnal GER episodes are likely to amplify

respiratory reflexes.40

A better understanding of the association of GER and

respiratory mechanisms (i.e., apneas) in infants will

have great diagnostic and therapeutic consequences for

this group of patients and could potentially lead to

shortened hospital stay and a reduction in healthcare

costs. GERD is diagnosed in up to 50% infants admitted

for recurrent apneas and frequently accompanied by

costly and invasive diagnostics and even therapeutic

approaches like antireflux surgery.41–43 Thus, there is a

clear need well-powered studies.

One of the challenges when conducting research

addressing the association between GER and apnea in

symptomatic infants is the lack of consensus regarding

criteria which are needed to identify an infant as

‘symptomatic’. According to clinical guidelines, GERD

is defined as GER causing troublesome symptoms and

complications, with apneas mentioned as one of those

possible complications.9 Similar, only when the num-

ber of apneas per hour exceeds a threshold, a patient is

classified into one of the apneic sleeping disorder

syndromes (either central, obstructive or mixed).44

However, theoretically, every single GER episode

could trigger a significant apneic event and vice versa.

Therefore, when defining inclusion criteria, it is of

importance to realize that assessing the association

between GER and apnea is different from diagnosing

GERD or an apnea syndrome in clinical practice.

More generally, successful execution of randomized

clinical trials is often obstructed by difficulties of

recruiting research groups of sufficient size due to

ethical dilemmas both for medical ethical committees

and parents.

We suggest future studies assess premature and

mature infants separately, to distinguish the possible

different pathophysiological mechanisms of apnea of

prematurity and persistent apneas after reaching full

term. Furthermore, the use of state-of-the-art tech-

niques that include at least pH-MII for the detection of

all GER episodes is essential to find an association

between a normal and ubiquitous physiological phe-

nomenon as GER and a relatively rare phenomenon as

pathological apneas. Additional esophageal manometry

further enables assessment of predominant neurologi-

cal mechanism behind GER: TLESRs.2,45 For apnea, a

clear cut definition according to the latest standards



should be used, dividing apneas into central, obstruc-

tive, and mixed.4 Calculations on association of GER

and apneas must be based on best available association

indices and a well described epoch for association. To

sufficiently power studies, Barriga et al. demonstrated

that a prolonged measurement increases the possibility

to detect an association. Indeed, in recordings of

<12 hrs or with <70 GER episodes and <10 symptoms

(apneas) per 24 h, finding an association between GER

and symptoms is mathematically impossible.38 This

problem might be partially overcome by elongating the

measurements artificially with a mathematical

algorithm generated by computer systems. While a

recent study showed that GER distributions can indeed

be reproduced in silico, future research will need to

show if symptoms such as apneas can be predicted by

computer models in a similar way.46

In conclusion, despite substantial literature on the

association between GER and apnea in infants, this

systematic review identified only six articles that met

our inclusion criteria and were suitable for data

extraction. This indicates there is a lack of high quality

data addressing this topic. We found insufficient

evidence for an association between GER and apnea

in infants. Nevertheless, it might be that GER is

causally related to apnea in a selected group of infants.

To adequately identify if this group exists and what the

nature of this hypothesized GER-apnea association is,

there is a need for high quality, well-powered studies

using uniform inclusion criteria, definitions according

to accepted guidelines, state-of-the-art techniques,

reliable symptom association indices, and patient

relevant outcome measures.

ACKNOWLEDGMENTS

We would like to thank Dhr. A. Leenders of the AMC library(Medical Library, Academic Medical Centre, Amsterdam, TheNetherlands) for conducting the literature.

FUNDING

No funding was secured for this study.

DISCLOSURE

The authors have no financial relationships relevant to this articleto disclose. None of the authors received an honorarium, grant orother form of payment to produce the manuscript.

AUTHOR CONTRIBUTION

MJS wrote the first draft of the manuscript. All authors acceptresponsibility for the reported content and declare they all haveparticipated in the concept, design, draft, and revision of thismanuscript. All authors have approved this version of themanuscript; MJS conceptualized and designed the study, scoredall abstracts and full text articles, drafted the initial manuscript,and approved the final manuscript as submitted; MPWwas secondrater of abstracts and full text articles for systematic review,reviewed and revised the manuscript, and approved the finalmanuscript as submitted; MWL, MAB reviewed and revised themanuscript, and approved the final manuscript as submitted;MMT was deciding rater of full text articles for systematic review,senior author, reviewed and revised the manuscript, and approvedthe final manuscript as submitted.

REFERENCES

1 Omari TI, Barnett C, Snel A, Golds-worthy W, Haslam R, Davidson G,Kirubakaran C, Bakewell M et al.Mechanisms of gastroesophagealreflux in healthy premature infants.Gut 2002; 133: 650–4.

2 Omari TI, Barnett CP, Benninga MA,Lontis R, Goodchild L, Haslam RR,Dent J, Davidson GP. Mechanisms ofgastro-oesophageal reflux in pretermand term infants with reflux disease.J Pediatr 1998; 51: 475–9.

3 Boeckxstaens GE. The lower oesoph-ageal sphincter. NeurogastroenterolMotil 2005; 17(Suppl. 1): 13–21.

4 Berry RB, Budhiraja R, Gottlieb DJ,Gozal D, Iber C, Kapur VK, MarcusCL, Mehra R et al. Rules for scoringrespiratory events in sleep: update ofthe 2007 AASM Manual for the Scor-ing of Sleep and Associated Events.

Deliberations of the Sleep Apnea Def-initions Task Force of the AmericanAcademy of Sleep Medicine. JCSM2012; 8: 597–619.

5 Statement Consensus. National Insti-tutes of Health consensus develop-ment conference on infantile apneaand home monitoring. Pediatrics

1987; 79: 292–9.6 Slocum C, Hibbs AM, Martin RJ,

Orenstein SR. Infant apnea and gas-troesophageal reflux: a critical reviewand framework for further investiga-tion. Curr Gastroenterol Rep 2007; 9:219–24.

7 Martin RJ, Abu-Shaweesh JM. Con-trol of breathing and neonatal apnea.Biol Neonate 2005; 87: 288–95.

8 Abu-Shaweesh JM, Martin RJ. Neona-tal apnea: what’s new? Pediatr Pul-

monol 2008; 43: 937–44.9 Vandenplas Y, Rudolph CD, Di Lore-

nzo C, Hassall E, Liptak G, Mazur L,

Sondheimer J, Staiano A et al. Pediat-ric gastroesophageal reflux clinicalpractice guidelines: joint recommen-dations of the North American Soci-ety for Pediatric Gastroenterology,Hepatology, and Nutrition (NASP-GHAN) and the European Society forPediatric Gastroenterology, Hepatolo-gy, and Nutrition (ESPGHAN). J Pe-

diatr Gastroenterol Nutr 2009; 49:498–547.

10 Rao SV, Sant’Ambrogio FB, Sant’Am-brogio G. Respiratory reflexes evokedby tracheal distension. J Appl PhysiolRespir Environ Exerc Physiol 1981;50: 421–7.

11 Satpathy NK, Al-Sattar NA. Theeffects of acute oesophageal disten-sion on arterial blood pressure, E.C.G.and respiration in dog. Indian J Phys-

iol Pharmacol 1984; 28: 105–14.12 Thach BT. Maturation of cough and

other reflexes that protect the fetal



and neonatal airway. Pulm Pharma-col Ther 2007; 20: 365–70.

13 Page M, Jeffery HE. Airway protectionin sleeping infants in response topharyngeal fluid stimulation in thesupine position. Pediatr Res 1998; 44:691–8.

14 Pickens DL, Schefft G, Thach BT.Prolonged apnea associated withupper airway protective reflexes inapnea of prematurity. Am Rev Respir

Dis 1988; 137: 113–8.15 Corvaglia L, Spizzichino M, Aceti A,

Legnani E, Mariani E, Martini S,Battistini B, Faldella G. A thickenedformula does not reduce apneasrelated to gastroesophageal reflux inpreterm infants. Neonatology 2013;103: 98–102.

16 Poets CF, Brockmann PE. Myth: gas-troesophageal reflux is a pathologicalentity in the preterm infant. Semin

Fetal Neonatal Med 2011; 16: 259–63.17 Omari TI. Apnea-associated reduc-

tion in lower esophageal sphinctertone in premature infants. J Pediatr

2009; 154: 374–8.18 Davies AM, Koenig JS, Thach BT.

Upper airway chemoreflex responsesto saline and water in preterminfants. J Appl Physiol 1988; 64: 1–10.

19 Jeffery HE, Rahilly P, Read DJ. Multi-ple causes of asphyxia in infants athigh risk for sudden infant death.Arch Dis Child 1983; 58: 92–100.

20 Molloy EJ, Di Fiore JM, Martin RJ.Does gastroesophageal reflux causeapnea in preterm infants? Biol Neo-nate 2005; 87: 254–61.

21 Hartling L, Hamm M, Milne A, Van-dermeer B, Santaguida PL, Ansari M,Tsertsvadze A, Hempel S et al. Valid-ity and Inter-Rater Reliability Testingof Quality Assessment Instruments.Agency for Healthcare Research andQuality, 2012. Report No.: 12-EHC039-EF. Available at: http://www.ncbi.nlm.nih.gov/books/NBK92293/

22 Wells G, Shea B, O’Connel D, Peter-son J, Welch V, Losos M, Tugwell P.The Newcastle-Ottawa Scale (NOS)for assessing the quality of nonran-domized studies in meta-analyses.Available at: http://www.ohri.ca/pro-grams/clinical_epidemiology/oxford.asp (accessed 6 August 2011).

23 Ariagno RL, Guilleminault C, Bald-win R, Owen-Boeddiker M. Move-ment and gastroesophageal reflux inawake term infants with “near miss”SIDS, unrelated to apnea. J Pediatr

1982; 100: 1–4.

24 Ariagno RL, Kikkert MA, MirmiranM, Conrad C, Baldwin RB. Cisapridedecreases gastroesophageal reflux inpreterm infants. Pediatrics 2001; 107:1–9.

25 de Ajuriaguerra M, Radvanyi-BouvetMF, Huon C, Moriette G. Gastro-esophageal reflux and apnea in pre-maturely born infants duringwakefulness and sleep. Am J Dis

Child 1991; 145: 1132–6.26 Di Fiore JM, Arko M, Whitehouse M,

Kimball A, Martin RJ. Apnea is notprolonged by acid gastroesophagealreflux in preterm infants. Pediatrics

2005; 116: 1059–63.27 Di Fiore J, Arko M, Herynk B, Martin

R, Hibbs AM. Characterization ofcardiorespiratory events followinggastroesophageal reflux in preterminfants. J Perinatol 2010; 30: 683–7.

28 Menon AP, Schefft GL, Thach BT.Apnea associated with regurgitationin infants. J Pediatr 1985; 106: 625–9.

29 Glen DR, Murakami P, Nunez JS.Symptom index P-value and symptomsensitivity index P-value to determinesymptom association between apneaand reflux in premature infants atterm. Dis Esophagus 2012; 26: 549–56. doi:10.1111/dote.12001.

30 Mattioli G, Pini-Prato A, Gentilino V,Caponcelli E, Avanzini S, Parodi S,Rossi GA, Tuo P et al. Esophagealimpedance/pHmonitoring inpediatricpatients: preliminary experience with50 cases.DigDis Sci 2006; 51: 2341–7.

31 L�opez-Alonso M, Moya MJ, Cabo JA,Ribas J, del CarmenMac�ıas M, Silny J,Sifrim D. Twenty-four-hour esopha-geal impedance-pH monitoring inhealthy preterm neonates: rate andcharacteristics of acid, weakly acidic,and weakly alkaline gastroesophagealreflux.Pediatrics 2006; 118: e299–308.

32 Pilic D, Fr€ohlich T, N€oh F, Pappas A,Schmidt-Choudhury A, K€ohler H,Skopnik H, Wenzl TG. Detection ofgastroesophageal reflux in childrenusing combined multichannel intra-luminal impedance and pH measure-ment: data from the GermanPediatric Impedance Group. J Pediatr2011; 158: 650–654.e1.

33 Wasilewska J, Semeniuk J, CudowskaB, Klukowski M. Respiratoryresponse to proton pump inhibitortreatment in children with obstruc-tive sleep apnea syndrome and gas-troesophageal reflux disease. SleepMed 2012; 13: 824–30.

34 Wilshire CL, Salvador R, Sepesi B,Niebisch S, Watson TJ, Litle VR,

Peyre CG, Jones CE et al. Reflux-associated oxygen desaturations: use-fulness in diagnosing reflux-relatedrespiratory symptoms. J Gastrointest

Surg 2013; 17: 30–8.35 Parkman HP, Hasler WL, Fisher RS.

AmericanGastroenterologicalAssoci-ationtechnical reviewonthediagnosisandtreatmentofgastroparesis.Gastro-

enterology 2004; 127: 1592–622.36 Abell TL, Bernstein RK, Cutts T,

Farrugia G, Forster J, Hasler WL,McCallum RW, Olden KW et al.

Treatment of gastroparesis: a multi-disciplinary clinical review. Neuro-

gastroenterol Motil 2006; 18: 263–83.

37 Bredenoord A, Smout A. Associationbetween reflux and symptoms duringambulatory reflux monitoring: prosand cons of existing methods. Neuro-

gastroenterol Motil 2013; 25: 633–7.38 Barriga-Rivera A, Elena M, Moya MJ,

Lopez-AlonsoM. The binomial symp-tom index: toward an optimal methodfor the evaluation of symptom asso-ciation in gastroesophageal reflux.Neurogastroenterol Motil 2013; 25:664–9.

39 Vandenplas Y, Ashkenazi A, Belli D,BoigeN, Bouquet J, Cadranel S, CezardJP, Cucchiara S et al.Aproposition forthe diagnosis and treatment of gastro-oesophageal reflux disease in children:a report from a working group ongastro-oesophageal reflux disease. EurJ Pediatr 1993; 152: 704–11.

40 Barriga-Rivera A, Elena M, Moya MJ,L�opez-Alonso M. Monte Carlomethod for the evaluation of symp-tom association. Dis Esophagus 2012.doi: 10.1111/j.1442-2050.2012.01436.x [Epub ahead of print].

41 Machado R, Woodley FW, Skaggs B,Di Lorenzo C. Gastroesophagealreflux causing sleep interruptions ininfants. J Pediatr Gastroenterol Nutr

2013; 56: 431–5.42 Tieder JS, Cowan CA, Garrison MM,

Christakis DA. Variation in inpatientresource utilization and managementof apparent life-threatening events.J Pediatr 2008; 152: 629–35, 635.e1-2.

43 Doshi A, Bernard-Stover L, Kuelbs C,Castillo E, Stucky E. Apparent life-threatening event admissions andgastroesophageal reflux disease: thevalue of hospitalization. Pediatr

Emerg Care 2012; 28: 17–21.44 Koivusalo AI, Pakarinen MP, Wi-

kstr€om A, Rintala RJ. Assessmentand treatment of gastroesophagealreflux in healthy infants with apneic



episodes: a retrospective analysis of87 consecutive patients. Clin Pediatr

2011; 50: 1096–102.45 Marcus CL. Sleep-disordered breath-

ing in children. Am J Respir Crit CareMed 2001; 164: 16–30.

46 Mittal RK, Holloway RH, Penagini R,Blackshaw LA, Dent J. Transientlower esophageal sphincter relaxation.Gastroenterology 1995; 109: 601–10.

47 Barriga-Rivera A, Moya MJ, Elena M,Lopez-Alonso M. Inter-reflux and

bolus clearance times in non-patho-logic pediatric patients: data supportcomputational models. Dis Esopha-

gus 2014. doi: 10.1111/dote.12180[Epub ahead of print].



association between gastroesophageal reflux and pathologic apneas in infants.pdf

Documents