Autonomic regulation of preterm infants is enhanced by Family … · 2019. 6. 6. · Autonomic regulation of preterm infants is enhanced by Family Nu rture Intervention Article in

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/331734936

Autonomic regulation of preterm infants is enhanced by Family Nurture

Intervention

Article  in  Developmental Psychobiology · March 2019

DOI: 10.1002/dev.21841

CITATION

1READS

99

10 authors, including:

Some of the authors of this publication are also working on these related projects:

Mother-Child Study, Early Experience and Physiology Lab, Williams College View project

Nurture Science Program, Columbia University Medical Center View project

Stephen W Porges

Indiana University Bloomington

278 PUBLICATIONS   18,110 CITATIONS   

SEE PROFILE

Robert J Ludwig

Columbia University Irving Medical School

23 PUBLICATIONS   290 CITATIONS   

SEE PROFILE

Michael M Myers

Columbia University

124 PUBLICATIONS   2,450 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Amie Hane on 17 March 2019.

The user has requested enhancement of the downloaded file.

Developmental Psychobiology. 2019;1–11. wileyonlinelibrary.com/journal/dev  | 1© 2019 Wiley Periodicals, Inc.

Received:23July2018  |  Revised:10December2018  |  Accepted:23December2018DOI:10.1002/dev.21841

R E S E A R C H A R T I C L E

Autonomic regulation of preterm infants is enhanced by Family Nurture Intervention

Stephen W. Porges1,2 | Maria I. Davila2 | Gregory F. Lewis1,3 | Jacek Kolacz1 | Stephanie Okonmah‐Obazee2 | Amie Ashley Hane4 | Katie Y. Kwon5 | Robert J. Ludwig5 | Michael M. Myers5,6 | Martha G. Welch5,6,7

1TraumaticStressResearchConsortium,KinseyInstitute,IndianaUniversity,Bloomington,Indiana2DepartmentofPsychiatry,UniversityofNorthCarolinaatChapelHill,ChapelHill,NorthCarolina3IntelligentSystemsEngineering,IndianaUniversity,Bloomington,Indiana4DepartmentofPsychology,WilliamsCollege,Williamstown,Massachusetts5DepartmentofPediatrics,ColumbiaUniversityMedicalCenter,NewYork,NewYork6DepartmentofPsychiatry,ColumbiaUniversityMedicalCenter,NewYork,NewYork7DepartmentofAnatomyandCellBiology,ColumbiaUniversityMedicalCenter,NewYork,NewYork

CorrespondenceMarthaG.Welch,DepartmentofPediatrics,ColumbiaUniversityMedicalCenter,NewYork,NY.Email:[email protected]

Funding informationFundingwasprovidedbyTheEinhornFamilyCharitableTrust(MGW),TheFleurFairmanFamily(MGW),andMaryDexterStephenson(MGW).

AbstractPreterm infantshavematurationaldelays inseveralneurobehavioralsystems.ThisstudyassessestheimpactoftheFamilyNurtureIntervention(FNI)intheneonatalintensivecareunit(NICU)onthematurationofautonomicregulationofpretermin-fants.Preterminfantsbornat26–34weekspostmenstrualage(PMA)wereassignedto groups receiving either standard care (SC) or SC plus FNI, using a randomizedcontrolled trialdesign.At twocollection timepoints,approximately35weeksand41weeksPMA,electrocardiograms(ECG)weremonitoredforapproximately1hourduringsleep.Heartrateandrespiratorysinusarrhythmia(RSA)werequantifiedfromtheECG.Acrossthetwotimepoints,theFNIgroupexhibitedgreater increases inRSA(Cohen'sd=0.35)andslopebetweenRSAandheartrate,asameasureofvagalefficiency(Cohen'sd=0.62).TheseresultsdocumentthatFNIresultedinenhancedautonomicregulationconsistentwithgreatermaturationofcardiacfunction.Theseandpreviousfindingsstronglysuggestthatfacilitatingearlynurturing interactionsandemotionalconnectionbetweenpreterminfantsandtheirmothersisapractica-ble and effectivemeans of optimizing postnatal development in preterm infants.Interpretationoftheseautonomicfunctionresultsalsoenrichesourunderstandingof thepotential long‐termbeneficial outcomesof FNI bydrawinguponpolyvagaltheory,whichexplainshowautonomicstateprovidesaneurophysiologicalplatformforoptimalco‐regulationbetweeninfantandcaregiver,andbydrawinguponcalmingcycletheory,whichprovidesamodelforunderstandinghowrepeatedmother/infantcalming interactions positively condition autonomic state and reinforce approach,prosocialbehaviors.

K E Y W O R D S

calmingcycletheory,emotionalconnection,heartperiod,heartratevariability,polyvagaltheory,respiratorysinusarrhythmia,vagalefficiency

1  | INTRODUC TION

Prematurity is found in9%–12%ofbirths in theUS.Althoughad-vances in neonatalmedicinehavedecreasedmortality of pretermbirth, surviving infants are vulnerable to multiple morbidities,

including increased risk formedical problems (Barfield, 2018), at-tention deficits, cognitive and language delays, disturbed execu-tive function, autism spectrumdisorder (Agrawal, Rao, Bulsara,&Patole, 2018), anxiety and depression (Field, 2018), and cognitiveimpairment(Brydgesetal.,2018).Amongthecontributorstothese

2  |     PORGES Et al.

outcomes are the effects of physical and emotional separation ofinfantsandmothersduringtheneonatalintensivecareunit(NICU)stay(Korja,Latva,&Lehtonen,2012).

Considerable evidence suggests that after preterm birth neu-rodevelopmental interventionswith infants,mothers, or both canhavepositiveshort‐termeffects(Benzies,Magill‐Evans,Hayden,&Ballantyne,2013;Chertok,McCrone,Parker,&Leslie,2014;Moore,Bergman,Anderson,&Medley,2016), althoughmany studies lackdesignrigor (Ohlsson&Jacobs,2013)and long‐termprognosisforpreterm infants, as noted above, is still poor. Nonetheless, basicsciencestudiesprovidestrongevidencethatduringearly life,sen-sorystimulationembeddedinmother–infantnurturinginteractionsis critical for shaping neurodevelopment (Tang, Reeb‐Sutherland,Romeo,&McEwen,2014).

Thisresearch,alongwithan interventionpreviouslydevelopedbyWelch (1988),promptedthedesignofanovel in‐hospital inter-vention,theFamilyNurtureIntervention(FNI),whichaimstoemo-tionally connect infants andmothers during the traumatic effectsofearlyseparation.Theinterventionutilizedsometechniquesandprocedures commonly practiced. However, FNIwas singularly fo-cusedonanewmeasureablebehavioralconstruct,autonomicemo-tionalconnection(Haneetal.,2018;Welch,2016),asopposedtotheconventionalattachmentandbondingconstructs,whichemphasizepsychologicalconnection(Chambers,2017).

Theinitialrandomizedclinicaltrial(RCT),comparingFNIwithstandardcare(SC)demonstratedfeasibilityandsafetyoftheinter-vention (Welchetal.,2013),showedthatFNImothersexhibitedincreasedmaternalsensitivityduringnormalcaregivingactivitieswhen their infantswere approximately 36weeks postmenstrualage(PMA)(Haneetal.,2015),anddecreasedsymptomsofanxietyanddepressionwhen their infantswere4monthscorrectedage(Welchetal.,2016).At~41weeksPMA,FNIinfantshadrobustlyincreased high‐frequency electroencephalogram (EEG) power(Welch et al., 2014), altered EEG‐based functional connectivity(Myersetal.,2015),andadvancedmaturation(Welchetal.,2017).At18monthscorrectedage,FNIinfantshadimprovedscoresontheBayley‐III, fewerattentionproblemson theChildBehavioralCheck List, and decreased risk for socio‐emotional problems asassessedbytheModifiedChecklistforAutisminToddlers‐Revised(Welchetal.,2015).

In thecurrentstudy,maturationalchanges inautonomic reg-ulationwereevaluatedduring theperiodwhile the infantswerereceivingFNIorSC in theNICUor shortly thereafter.Polyvagaltheoryformsthetheoreticalbridge linkingtheattributeofemo-tionalconnection inFNIasa facilitatorofmoreoptimalmatura-tionofautonomicregulationandenhancedco‐regulation(Porges,1995,2007,2009,2015;Porges&Furman,2011).Polyvagalthe-oryhighlightstheontogenetictransitionsintheneuralpathwaysthat regulate autonomic state and provide an understanding ofhowautonomic stateprovidesaneurophysiologicalplatform fortheoptimalco‐regulationbetweeninfantandcaregiver.Thethe-oryarticulatesthemechanismsthroughwhichthematurationandmyelination of vagal pathways are involved in the regulation of

visceral state (e.g., heart rate) andhowvagalpathwaysarebidi-rectionallylinkedwiththebrainstemareascontrollingmanyoftheattributesofsocialengagement(e.g.,facialexpression,gaze,andvocalizations)throughwell‐definedvisceralefferentpathways inseveralcranialnerves.

Informed by the literature documenting thematuration of au-tonomic regulationduringgestation (Porges&Furman,2011), it ispossible to interpret maturational shifts observed during the in-terventionwith the samemetricsmonitored inother studieswithpreterm infants (DiPietro & Porges, 1991; Feldman & Eidelman,2003; Porges, 1992; Portales et al., 1997; Suess et al., 2000) andhealthy full‐term newborns (Porges, Doussard‐Roosevelt, Stifter,McClenny,&Riniolo,1999).Theabovecitedstudieshaveappliedavalidatedmethodofquantifyingacomponentofheartratevariabilityknownasrespiratorysinusarrhythmia(RSA)toindexcardiacvagaltone(Lewis,Furman,McCool,&Porges,2012;Porges,1985b).RSAisobservedasarhythmic increaseanddecrease inheartratethatoccursatthefrequencyofspontaneousbreathing.RSAisconveyedthroughmyelinatedvagalpathwaysoriginatinginthebrainstemareaknownasnucleusambiguus.Thisrespiratoryrhythmwasidentifiedmorethanacenturyagoin1910byHering(1910),whonotedthattheinhibitoryelectricaldischargethroughthevagustotheheart'spacemaker(i.e.,sino‐atrialnode)hadarespiratoryrhythm.Thus,thequantificationofRSAhasbecomeacontemporarymethodtoindexcardiacvagaltone.

Inthisstudy,weusetwomeasurestoevaluatethevagalmatura-tionandregulationoftheheart.Basedontheliterature(seeabove),RSAwasmonitored as an index of cardiac vagal tone to evaluatematuration,sinceRSAexhibitsadevelopmentalcurvesimilartohis-tologicalreportsofmyelinationofthevagus(seePorges&Furman,2011).InadditiontoRSA,wemeasuredtheslopebetweenthedy-namicandsynchronousshiftsinRSAandheartperiod,whichpro-videsameasureofregulationofheartratebyvagalpathways.Thiswerefertoas“vagalefficiency,”aconceptfirstintroducedasanex-tensionofpolyvagaltheoryin1999(Porgesetal.,1999).Accordingto polyvagal theory (Porges, 1995; Porges, Doussard‐Roosevelt,Portales,&Greenspan,1996),thevagalbrakeisconceptualizedasanadaptive neural physiologicalmechanism that fosters engagementand disengagement with the environment. The efficiency of thevagalbrakecanbeassessedbycalculatingtherelationshipbetweensynchronousmeasures of the amplitude of RSA and heart periodquantified insequential10sepochsduringeachsleepstate.As il-lustratedinFigure1,vagalefficiencyincreasesastheslopebetweenheartperiodandRSAbecomessteeper.IntheleftpanelofFigure1,foreachlogunitchangeinRSA,thereisachangeinheartperiodof9msec.IntherightpanelofFigure1,foreachlogunitchangeinRSAthereisachangeinheartperiodof19msec.Theseplotsillustratehowchangesincardiacvagaltone(measuredbyRSA)willdifferen-tiallyinfluenceheartperiod.

Since this novel measure has been documented to exhibitsleep‐statedifferencesinfulltermnewborns,quantificationinthisstudy provided an opportunity to evaluate the influence of FNI,which is aimed at establishment of emotional connection, on the

     |  3PORGES Et al.

maturational shiftof this variable and its ability todistinguishbe-tweensleepstates.

2  | METHOD

2.1 | Intervention procedures

Shortlyafterbirth,afterobtainingpermissionfromtheattendingphy-sician (average7days),a researchassistantapproached themother(andfather ifpresent) toexplain thestudyandobtainwrittencon-sent.TheconsentedfamilieswereassignedtoeitherFNIorSCusinga computer‐generated block randomization. Twin randomizationwasconductedatthelevelofthefamily.Motherswererandomizedandtwinswerejointlyassignedwiththeirmothertoasinglegroup.Afterconsent,mothersassignedtotheFNIgroupmetwithNurtureSpecialists,whoworkedwiththemother,infant,andfamilythrough-outthestudytofacilitateallaspectsoftheintervention.MothersandinfantsassignedtotheSCconditionreceivedthecarethatisstandardforinfantsadmittedtotheNICUparticipatinginthisstudy.

NurtureSpecialistsinitiatedthefirstFNIactivitiesapproximately1week after birth while the infant was still in the incubator. TheNurtureSpecialistswerenurses trainedby the staff at theNurtureSciencePrograminallaspectsoftheinterventionprocedures.Thefirstinterventionprocedureinvolvedengagingmothersandtheirinfantsin“calmingsessions.”Thisinvolvedmothers’communicatingdirectlyandemotionallywiththeirinfantsforaprolongedperiod,usually1hourormore,untiltheinfantandmotherwerenoticeablycalmer.FNImotherswerefacilitatedinmakingcontactwiththeirinfantsthroughtheportsoftheincubator,usingfirmandsustainedtouch.Eachmotherwaspro-videdtwosmallcottoncloths,onetowearinherbraandtheothertoplaceundertheheadofherinfant,whichtheywereencouragedtoex-changedaily.ThemotherswereledbytheNurtureSpecialisttospeak

directlytotheirinfantsinanemotionalmanner,includingexpressionoftheirupsetfeelingsabouttheearlybirth,theirinfant'sfragilecon-dition,andaboutthehardshipsposedtothepairbyNICUcare.Theywereaskedtospeakintheirnativelanguage,theemotionallanguagespokentothembytheirownmothersandfamily,whileestablishingeyecontact.Onaverage,theFNImothersengagedintheseactivitiesfor6hr/week(Welchetal.,2013).

When the infantswere sufficiently stable to leave the incuba-tor, FNImotherswere encouraged to engage in calming sessions,including skin‐to‐skin or non‐skin‐to‐skin holding, vocal soothing,emotionalexpression,andeye‐to‐eyecontact.Duringthesecalmingsessions,NurtureSpecialistsassistedmothersinholdingtheirinfantssafelyandsecurelyinanuprightpositionwhilethemotherwascom-fortablyseatedinarecliningchair.Onceinitiated, mothersengagedinthesemutualcalmingactivitiesforonehour,oraslongaspossibleduringeachcalming session.Themothers continued toworkwithNurtureSpecialistsonanindividualized,regularbasis(i.e.,anaverageof3.5times/weekforanaverageof6weeks)duringcalmingsessionsfocusedonemotionalcommunicationwithherinfant.

When family members were available to meet, NurtureSpecialistsengagedtheminsessionsthatfocusondiscussionsoftheimportanceofthecalmingsessionactivities.Thesefamilysessions,whichusuallyoccurredneartheendoftheNICUstay,weredesignedtohelpfamiliesunderstandtheuniqueemotionalneedsofapreterminfantandtoencouragethefamilyunitstocontinuesupportingthemotherandinfantinaregularcalmingcycleroutineafterdischargefromthehospital.

Giventheheterogeneityofthepreterminfantsrecruitedinthisstudy, therewerevariations in thedurationbetweenbirthandtheinitiationofFNI(median=7.0days,minimum–maximum=1–14days)andbetweentheinitiationofFNIandthefirstassessmentofauto-nomicregulation(median,13days;range4to35days).

F I G U R E 1  Exampleofvagalefficiencydefinedbytheslopeofsynchronouswithinstate10smeasuresofrespiratorysinusarrhythmia(RSA)andheartperiod.Thoughbothinfantshaveasimilarrangeofvagaltonevalues(measuredbyRSA),theinfantonthelefthasnotaslow‐vagalefficiency(slope=9.16),suggestingthatavagalbrakehasaweakeffectonheartperiod.Theinfantontherighthashigh‐vagalefficiency(slope=19.36),suggestingthatchangesincardiacvagaltonehaveastrongereffectondynamicchangesinheartperiod.Aswouldbeexpectedfromamoreefficientvagalbrake,theinfantwithhigher‐vagalefficiency(rightpanel)alsohasalonger‐meanheartperiod

4  |     PORGES Et al.

2.2 | Subjects

TheexperimentaldesignwasaRCTcomparingSCandFNIwithinthe NICU. The study was conducted in the level IV NICU atthe Morgan Stanley Children's Hospital of New York, ColumbiaUniversityMedicalCenter.Acompletedescriptionof theprotocolwaspublishedpreviously (Welchetal.,2012).Briefly,115familieswith150preterminfantsat26–34weeksPMAwereenrolledovera42‐monthperiodfromJanuary2009throughJuly2012.

Maturational changes in heart rate patternsweremeasured attwotimepoints.Thefirstor“early”measureoccurredbetween34and37weeksPMA(~35weeks)andthesecond,“near‐to‐term,”be-tween37and44weeksPMA(~41weeksPMA).DuetothefocusofthispaperonthedifferentialinfluencesofSCandFNIonmaturationoftheautonomicnervoussystem,onlyinfantswithquantifiableelec-trocardiogram(ECG)dataduringbothactiveandquietsleepstateswithinbothsessionswereanalyzed.Inthisstudy,heartratedataarepresentedfrom59infantswhofitthefollowingcriteria:(1)sufficientECGdatatoobtainreliablemeasures(i.e.,atleastfour30‐sepochs)duringeachsleepstatewithineachsessionand(2)hadbothearlyandnear‐to‐termassessmentswithanintervalof2weeksormore.

Oftheoriginal150subjectsenrolled,59providedsufficientECGdatatocalculatethevariablesevaluated intheanalyses.Todeter-mine if thereweredifferencesbetweensubjects includedandnotincluded,andifsowhethersuchdifferencesweresimilarinthetwotreatmentgroups,two‐wayANOVAswithtreatmentgroup(SCvs.FNI)andstudyinclusion(includedvs.notincluded)weretestedforseveralkey infantanddemographicvariables.Theseanalysesalsodeterminedifthereweresignificantinteractionsbetweentreatmentand inclusion groups. Analyses were conducted for GA at birth,birthweight, delivery mode (spontaneous vs. caesarian section),lengthofstayintheNICU,sex,motherandfatherages,andmothers’education.Theanalysesrevealedasmall,butsignificantdifferenceinGAatbirth(p=0.009)withtheincludedsubjectsbeingapproxi-mately0.5weeksolder.Thisdifferencewasparalleledbyagreaterbirthweight(p=0.035)andshorterlengthofstayintheNICUbytheinclusiongroup(p=0.014).However,thesedifferencesweresimilar

forbothSCandFNI.Therewerenosignificantdifferencesinmotherorfatheragesoreducation,asindexedbythematernaleducation.Analysis ofmother's age revealed a significant intervention groupby inclusiongroup interaction (p=0.045),withFNImothersbeing3.6yearsolder in thenot includedgroupbut0.9yearsyounger inthe included group. Overall, the analyses support the conclusionthat the group differences observed in the autonomic measureswerenotrelatedtoasystematicbiasincharacteristicsofinfantsintheinclusionandtreatmentgroups.Inclusionandexclusionsamplesdidnotdifferontimingofassessmentsoronkeyoutcomevariables(see TableS1).

Ofthe59infantsinthisstudy,28wereintheSCgroupand31intheFNIgroup.Table1providesinformationonbirthweight,gesta-tionalagebirth,ageattheearlytestsession,ageatthenear‐to‐termsession,andintervalbetweentestsessions.NotethattheFNIgroupwassignificantlyolderduringthesecondsession,whichcontributedtotheFNIgrouphavinggreaterdurationsbetweentestingsessions.Althoughnotsignificantlydifferent,thelowestbirthweightinfantswererandomlyassignedtotheFNIgroup.TheinclusionoftwinsinthisstudycontributedtothispotentialbiasandwasnotobservedinPMA.Thus,birthweightandthetimebetweentestsessionswereusedas covariates in theanalysesofvariancemodels.Groupsdidnot differ by infant sex (SC males=39.39%; FNI males=45.16%;χ2=0.04,df=1,p=0.85).

2.3 | ECG acquisition

ECGrecordingsweremadeinconjunctionwithanEEGstudy(Welchetal.,2014).Datawerecollectedbetween11a.m.and4p.m.within30minafteranormallyscheduledfeeding.TheECGsignalwasac-quired at 1kHz via electrodes placed on the infant's chestwith aBiopac MP 150 physiological acquisition system (Biopac Systems,Inc.,SantaBarbara,CA).ECGdatawerecollectedforapproximately1hour,whileresearchassistantscodedsleepstate(i.e.,quiet,active,indeterminate, awake, and crying) during each sequential minute.Sleep statewas coded using behavioral criteria previously demon-stratedtobeappropriateforpreterminfants(Stefanskietal.,1984).

TA B L E 1   Infantcharacteristics

Variable Group n Mean SD Min Max t df p

Birthweight(grams) Standardcare 28 1,605.18 368.31 940.00 2,605.00 1.18 57 0.244

FNI 31 1,493.94 357.54 792.00 2,380.00

GestationalAge(weeks)

Standardcare 28 31.62 1.86 26.43 34.71 0.92 57 0.364

FNI 31 31.16 1.99 27.14 34.86

Ageat“early”session(weeks)

Standardcare 28 34.85 0.73 33.57 36.57 0.79 57 0.432

FNI 31 34.70 0.74 33.57 36.43

Ageat“near‐to‐term”session(weeks)

Standardcare 28 40.01 1.58 37.43 42.29 −2.55 57 0.013

FNI 31 41.00 1.39 36.71 44.43

Intervalbetweensessions(weeks)

Standardcare 28 5.16 1.69 2.14 7.14 −2.71 57 0.009

FNI 31 6.30 1.53 2.00 10.00

FNI:FamilyNurtureIntervention;SD:standarddeviation.

     |  5PORGES Et al.

Due to limitations in representations of sleep states during the 1‐hourdatamonitoringsession,analyseswereconductedonlyondata collected during quiet and active (REM) sleep. Quiet sleep was assignedwhentheinfanthadregularbreathing,noeyemovements,and lowmuscle tone. Active sleepwas assignedwhen eyemove-mentswereobserved, respirationwas irregular,andbody twitcheswerefrequent.

Thedigitizeddatawereprocessedwithsoftwarethatdetectedthe inflection point of theR‐wave and the interval (i.e., heart pe-riod)betweensequentialR‐waves to thenearestms.Thesequen-tialheartperiodswerevisually inspectedandeditedofflinewithamodified version of CardioEdit+ software (Brain‐Body Center forPsychophysiology and Bioengineering, Department of Psychiatry,UniversityofNorthCarolinaatChapelHill). Editingconsistedoftwosequentialmethods.ThefirstmethodfocusedonthetimeseriesofheartperiodsandidentifiedatypicalheartperiodsduetomissedorfaultydetectionsofR‐waves.Inmostcases,atypicalbeatscouldbeeditedbyaddingadjacentperiodstogetherandthendividing,ordi-vidinganatypical long intervaldue toamissedR‐wavedetection.However, anomalies in ECG patterns were occasionally observedwhenveryshortheartperiodswerenotfollowedbycompensatorypauses.Anomalies,suchasanisolatedshortbeat,cannotbeeditedby integer arithmetic (e.g., adding anddividing adjacent intervals).Norcanthe isolatedshortbeatbeeditedbyremoving it fromthetimeseriesofheartperiods,sinceremovalwoulddisrupttheconti-nuityofthedataandviolateanecessaryassumptionforthetimesseries analysesused toquantify theoscillations inheart rate thatdefineRSA.Todealwiththesecases,asecondeditingroutinewasdevelopedthatindexeddatabytimeandnotsequentialbeats.Thisstrategyrequiredthattheeditedoutputofheartperiodsfromtheinitialeditingphasebetimesampled.Inthesecondphaseofediting,theheartperiod timeserieswas timesampledat5Hzandanom-alousvaluesof the time‐sampledheartperiodswere identifiedasdeviationsoutsidetherange75%to137.5%oftheprecedingsecondofdata.Whenanoutlierwasidentified,a linear interpolationcon-nected the brief segments of valid heart periods.Consistentwithreports that heart period exhibits a stronger linear relation withautonomiccontrolthanheartrate(Berntson,Cacioppo,&Quigley,1995;Quigley&Berntson,1996),allcardiacvariablesevaluatedinthisstudywerequantifiedfromheartperiodtimeseries.

2.4 | Data quantification

After the data were edited, a modified version of CardioBatch+calculatedheartperiod,theamplitudeofRSA,andRSAfrequency(see computationmethods below) in sequential epochs thatweresynchronizedwith the sleep‐state codes. CardioBatch+ calculatestheamplitudeofRSAwiththePorges–Bohrermethod(Lewisetal.,2012;Porges,1985a),whichemploysa time‐frequencymethodtoextractfromtheheartperiodtimeseriesafrequencyband‐limitedcomponentrepresentingRSA.Thisstrategymaximizedrepresenta-tionof physiological variables during short epochs (30s) of sleep,whentheinfant'ssleeppatternmaybeunstableandrepresentedby

transitorystateshifts.Inaddition,regressionanalysesbetweenRSAand heart period based on extremely short epochs (10s) enabledtheevaluationoftheefficiencyofvagalregulation(i.e.,mschangeinheartperiodperunitchangeinRSA,Porgesetal.,1999).

The Porges–Bohrer method involves the following sequentialsteps (Lewis et al., 2012; Porges, 1985a; Porges& Bohrer, 1990).First, when applied to newborn infants, heart period values aremeasured to the nearestms and resampled into equal time inter-vals every 200ms. Second, a detrending algorithm removes fromtheheartperiodtimesseriesthevarianceassociatedwithcomplexaperiodicbaselineshiftsandoscillationsslowerthanRSA.Thede-trending algorithm applies a moving polynomial filter (3rd order,21 point, 4.2s duration) to remove aperiodic baselines and slowoscillations. Third, the residual output from the moving polyno-mialisband‐passed,andtheheartperiodvarianceinthefrequencyband associatedwith spontaneous breathing in the newborn (i.e.,0.3–1.3Hz) is quantified. Fourth, to reduce distribution skewnessassociated with variance estimates, the band‐passed variance istransformed with a natural logarithm and reported in units of ln(ms)2asRSA.Inaddition,spectralanalysiswasperformedonthede-trendeddata(i.e.,timesampledevery200msanddetrendedwithathirdorder,21‐pointmovingpolynomial)todeterminethefrequency(i.e.,frequencyofpeakofthespectrum)ofRSA.

Consistent with other studies in which similar methods havebeenused(i.e.,MXedit,CardioBatch,andCardioBatch+)withneo-nates(Porges,1992;Porgesetal.,1999;Portalesetal.,1997;Suessetal.,2000),epochdurationsof30swereusedtoquantifyRSAandheartperiod.Anaverageofthewithinsleepstateepochvalueswereused in thestatisticalanalyses forheartperiod,RSA,and the fre-quencyofRSA.

SinceRSAprovidesavalidatedmetricofcardiacvagalefferenttonethroughpathwaysoriginatinginnucleusambiguus(Lewisetal.,2012),thisstudyevaluatedhowRSAwasrelatedtoheartperiodandhowthisrelationshipchangedwithmaturationandtheinterventionsadministered within the NICU (i.e., contrasting FNI with SC). BycalculatingtheslopeoftheregressionbetweentransitorylevelsofheartperiodandRSAoververyshortepochs(10s),anindexoftheefficiencyofvagalregulation(i.e.,vagalefficiency)ofheartratewasgenerated.Functionally,thismeasurequantifiedtheimpactofvagalefferentpathwaysonheartrate.

2.5 | Metrics

Thefollowingvariablesweregeneratedforeachinfantwithineachsleepstateandtestsession:

1. Heart Period: The average time between sequential heart pe-riods inms. Statistically, thismeasure is the reciprocal ofheartrate (i.e.,asheartperiodexpands induration,heart rateslows),but has better distributional features for parametric analyses.Heartperiodincreaseswithageandmaturationduetoincreasesin parasympathetic to sympathetic balance in autonomicregulation.

6  |     PORGES Et al.

2. RespiratorySinusArrhythmia (RSA):Theaverageof thenaturallog‐transformedepoch‐basedmeasuresoftheamplitudeofRSA.RSA,anindirectmeasureofparasympatheticregulationofheartrate(heartperiod),increaseswithageandmaturation.

3. RSAFrequency(RSAf):Theaverageofepoch‐basedmeasuresofpeakfrequencyofRSAderivedfromspectralanalysesconductedonthedetrendedtimesampled(i.e.,5Hz)heartperiodusingthePorges–Bohrer moving polynomial filter. The peak of the de-trendedspectrumprovidesanestimateofrespirationrate.Inpre-vious research, this method has been used as an estimate ofrespirationrateandcorrelationswithchestmovementapproach1.0(Denver,Reed,&Porges,2007).Respiratoryrateslowswithmaturationandage.

4. RSA/HeartPeriodcoupling(vagalefficiency):RSAandheartpe-riod were calculated in sequential 10s epochs and clusteredwithineachsleepstate.Theslopefromregressionanalysesbe-tweentheepoch‐basedheartperiodandRSAvalueswithineachsleep state during each session defined RSA‐heart period cou-plingasanindexoftheefficiencyofvagalregulationofheartrate(Porgesetal.,1999).Thismeasureofautonomicregulationwouldbeexpectedtoincreasewithageandmaturation.

2.6 | Design and analyses

Asnoted above, datawere collectedduring two sessions:The firstor “early” sessionwasat~35weeksPMAand the secondor “near‐to‐term”sessionwasat~41weeksPMA.AsdescribedinTable1,atthestartofthestudyinfantclinicalconditionsdidnotdifferbetweengroups.Withineachofthetwotestsessions,aftertheECGfilesweretransformedintoaheartperiodtimeseriesandedited,thedataweresegmented intoperiodsof activeandquiet sleep.Theheartperiodtimeserieswereprocessedtoquantifyepoch‐basedindicesofheartperiod,RSA,RSAf (i.e., respiration rate), and theefficiencyof vagalregulationofheartrate(slopeofthelinearregressionofRSAonheartperiod).

Sincetherewerenodifferencesinphysiologicalmeasuresasafunctionofsex,allanalyses reportedarecollapsedacrosssex.Todeal with unequal variances in a repeated measures design, thedata were analyzed with linear fixed effects models (Verbeke &Molenberghs, 2009) and were implemented using the restrictedmaximum likelihoodestimator in the SPSSMixed functionproce-dure.Theeffectsofsession (e.g.,~35weeksPMAand~41weeksPMA),sleepstate (activeandquiet),andtreatmentgroup (SCandFNI)were assessed formain effects and interactions. Sleep stateand sessionwere treated as repeatedmeasures. Birthweight anddurationoftheintervalbetweentestsessionswereincludedasco-variatesinallmodels.Thecovariancestructurewasestimatedusingheterogenouscompoundsymmetry,whichallowsfortheheterog-enous variable variances that may arise with repeated measures(Wolfinger,1996).Significant interactionswereprobedusingpair-wisecomparisonsderivedfromthemarginalmeans.EffectsizesforcategoricalpredictorsweremeasuredwithCohen'sd(Cohen,1988).

3  | RESULTS

3.1 | Respiratory sinus arrhythmia

Althoughthedurationbetweentestsessionswasasignificantco-variateF(1,56.5)=6.8,p=0.012,itdidnotinteractwithanyfactororinteraction.RSAexhibitedasignificantGroupbySessioninterac-tion,F (1, 161.95)=4.72,p=0.031.As illustrated in Figure 2, theincreaseinRSAfrom~35weeksPMAto~41weeksPMAwassignif-icantlygreaterintheFNIgroup(Cohen'sd=0.35).Inaddition,therewas a significant session effect, F (1, 161.95)=153.71, p < 0.001 (Cohen'sd=1.07). RSA increased from1.53 ln (ms2, SE=0.09) to2.45ln(ms2,SE=0.10).

3.2 | RSA frequency

Foreachcontinuoussegmentofdatawithineachsleepstate,spectralanalyseswereconductedtoidentifythepeakfrequencywithinthefrequencybandinwhichRSAwasdefined(i.e.,0.3–1.3Hz).Theaver-ageacrosssegmentsprovidedtheindexofRSAf.Analysesindicateda significant difference between quiet sleep (0.67Hz or approxi-mately40cycles/minute)andactivesleep(0.63Hzorapproximately38cycles/minute),F(1,164.62)=9.47,p=0.002(Cohen'sd=0.41).Therewereno significanteffectsofFNIgroupmembershipor itsinteractionwithsleepstateandsession(allp>0.05).

3.3 | Heart period

Asthepreterminfantsmatured,therewasasignificantslowingofheartratereflectedinanincreaseinheartperiodfrom~35weeksPMA to ~41weeks PMA, F (1, 134.7)=112.54, p<0.001. Theaverage heart period, combined across groups, increased from384.92ms (SE=2.87; 155.9bpm) during the early session to412.41ms (SE=3.63; 145.5bpm) during the near‐to‐term session(Cohen'sd=1.01).

Across sessions and groups, therewas a significant slowing ofheartratereflectedinlongerheartperiodsduringquietsleeprelativetoactivesleep,F (1,147.16)=10.88,p=0.001.Duringquietsleep,theaverageheartperiodwas402.91ms(SE=3.59;148.9bpm)com-paredto394.4ms(SE=2.91;152.1bpm)duringactivesleep(Cohen'sd=0.37).TherewerenosignificanteffectsofFNIgroupmembershiporitsinteractionwithsleepstateandsession(allp>0.05).

3.4 | Vagal efficiency

The slopeof the regressionbetweenchanges inheartperiodandRSAsignificantlyincreasedastheinfantsmaturedfrom~35weeksPMA to ~41weeks PMA F (1, 147.43)=50.61, p<0.001. Theslope increased from9.77 (SE=0.70) to16.33 (SE=0.99;Cohen'sd=0.95).At~35weeksPMA,thisindicatedthatachangeof1lnunitinRSAresultedinachangeinheartperiodoflessthan10ms,whileatapproximatelytermageachangeof1lnunitinRSAresultedinachangeofmorethan16ms.

     |  7PORGES Et al.

Across test sessions, sleep state influenced vagal efficiency.The slope was greater during active sleep (15.50, SE=0.96)than during quiet sleep (10.59, SE=0.74), F (1, 149.23)=28.83,p<0.001 (Cohen'sd=0.80) (Figure 3).Moreover, as illustrated inFigure4,therewasasignificantSessionbySleep‐Stateinteraction,F (1,150.29)=5.66,p=0.019.Althoughgreaterslopeswereobservedduringactivesleepduringbothsessions,therewasagreatermatu-rationalincreaseinslopeduringactivesleeprelativetoquietsleep(Cohen'sd=0.40).

AlthoughbothSCandFNI infantsexhibitedamaturational in-crease in the slope between RSA and heart period (i.e., vagal ef-ficiency), the groups differed in these age‐related changes. Asillustrated in Figure 4 the FNI group experienced a greater rateof increase in vagal efficiency. This was supported by a signifi-cant Session byGroup interaction,F (1, 147.43)=8.02,p = 0.005. Specifically,althoughatthe~35‐weektimepoint,theFNIgrouphadasignificantlylowerslope,bytermagetheFNIgrouphadcrossedoverthetrajectoryoftheSCgroup(Cohen'sd=0.62).

4  | DISCUSSION

FNI was designed to help reestablish the emotional connection(Haneetal.,2018)betweenmothersandtheirinfantsthatwasdis-ruptedbythepretermdeliveryandtheexperiencesofmotherandinfantwithintheNICU.FindingsthusfarhaveshownthatFNIledtogreatermaternalsensitivityduringroutinecaregivingbehavioralin-teractions(Haneetal.,2015),alteredinbrainactivityandfunctional

connectivityconsistentwithacceleratedbrainmaturation,particu-larlyinfrontalregions(Myersetal.,2015;Welchetal.,2017),low-eredsymptomsofmaternaldepressionpostdischarge(Welchetal.,2016),andimprovedneurobehavioraloutcomesto18monthsofage(Welchetal.,2015).Giventheemphasisoftheinterventionondyadicemotionalconnectionandthe linkagebetweenautonomicactivity,emotionregulation,andsocialengagement(Porges,2007,2009),wehypothesizedthatwithimprovedemotionalconnectionFNIinfantswouldexhibitsignsofamorematurevagalregulationsystemveryearlyinlife(Welch&Ludwig,2017a,2017b).Itisinterestingtonote,forinstance,thatinfantslaterdiagnosedwithautismdemonstratedslowerincreaseofRSAbetweenage1andage6,ascomparedwithnon‐autisticinfants(Sheinkopfetal.,2019).Futurestudieswillevalu-ate thecovariationof theseearly indicesofautonomicmaturationwiththeconcurrentmeasuresofEEGpreviouslyreported(seeabovecitations)andbehavioralandautonomicassessmentsobtainedfromthesubjectsaftertheyweredischargedfromtheNICU.

Collection of ECG signals during the EEG recording sessions atapproximately 35weeks PMA and again at approximately 41weeksallowed us to quantify RSA as a marker of these proposed effects.Consistentwithourhypothesis,resultsfromthecurrentanalysesshowthattheincreasedRSA,amarkerofcardiacvagaltoneandautonomicregulation,occurringbetweentheearly tothenear‐to‐termagetimepointsweresignificantlygreaterintheFNIgroup.Inaddition,theslopebetweenRSAandheartperiod,anindexofvagalregulationefficiency,wassignificantlysteeperintheinterventiongroup.Weinterpretbothofthesefindingsasevidenceofacceleratedmaturationofautonomiccon-trolinFNIinfants.HeartperiodandRSAfmeasures,reflectinglesswell‐definedvagalinfluence,werenotsensitivetointerventionassignment.

ThemechanismsthroughwhichFNIacceleratedthematurationofvagalregulationareasyetunclear.However,wehaveproposedthatPavlovianconditioningofautonomicstatesmayunderlietheseeffects.Accordingtocalmingcycletheory,thecalmingroutine,whichinvolvescommunicationofaffectduringcomforttouch,vocalsooth-ing,emotionalexpression,eye‐to‐eyecontact,andskin‐to‐skin (or

F I G U R E 2  Astheinfantsapproachedterm,increasesinrespiratorysinusarrhythmiaweresignificantlygreaterininfantsreceivingFamilyNurtureInterventioncomparedtothestandardcaregroup(p=0.031;Cohen'sd=0.38;errorbarsrepresentstandarderror)

F I G U R E 3  Astheinfantsapproachedterm,increasesinvagalefficiencyweresignificantlygreaterduringactivesleep(p = 0.019; Cohen'sd=0.40;errorbarsrepresentstandarderror)

8  |     PORGES Et al.

clothed)holding,facilitatedemotionalconnectionbetweenmotherand infant, which resulted in approach behaviors, and co‐condi-tioningofcalmingreflexesinbothmotherandinfant(Welch,2016;Welch&Ludwig,2017a,2017b).Inotherwords,themother/infantcalmingsessions,whichwerefacilitatedbyNurtureSpecialistsintheNICU,providedopportunitiestopairpositiveapproachbehaviorsofbothmotherandinfantwithahigh‐cardiacvagaltone,calmness,andregulatedautonomicstate.Thisrepeatedpairingovertimeresultsinaconditioningofanadaptivecardiaccalmingreflexuponcontact.

Thecalmingcycletheoryisconsistentwiththepolyvagaltheory,whichsuggeststhatrepetitionofthereciprocalbehaviorsfunctionasaneuralexerciseofthemother'sandinfant'sSocialEngagementSystems(Porges,2007,2009;Porges&Furman,2011).Byexercis-ingtheSocialEngagementSystem,thecranialnerves(i.e.,V,VII,IX,X,XI)involvedintheregulationofthestriatedmusclesofthefaceandhead(throughspecialvisceralefferentpathways),andtheheart(through themyelinatedvagalpathwaysoriginating in thenucleusambiguus)arerecruitedforsocialengagementandstateregulation.Thus, repeated engagement of these systems not only conditionsanadaptivecardiaccalmingreflexuponcontact,butalsoenhancesresilience by promotingmaturation of these systems during earlydevelopment.

Itisreasonabletoaskwhethermorerapiddevelopmentofauto-nomiccontrolisofbenefittotheinfant.Currently,theanswertothisquestionmustrelyoncorrelativeevidence.First, themorerapidlyincreasedRSAandvagalefficiencyseeninFNIinfantsweresimilartovaluesforRSAandvagalefficiencymeasuredinfullterminfantsinapriorstudy(Porgesetal.,1999).ThissuggeststhatSCpreterminfantslagintheirautonomicdevelopmentandthatFNIpromotednormalization in maturation in these systems. Second, as notedabove,FNIresultedinpositiveoutcomesacrossmultipledimensionswhentheinfantsreached18monthsofage.Thus,theaccelerationinautonomic regulationearly in lifepresagedpositiveoutcomesayearandahalflater.Again,thesedatasupporttheconclusionthat

accelerationofneuralregulationofautonomicfunction(RSA,vagalefficiency)andcentralnervousactivity(EEGpowerandcoherence)were likely early markers of positive effects of the intervention.Analyses examining whether these early physiologic markers arepredictive (i.e., correlated)with long‐term outcomes are currentlybeingevaluated.

AlthoughthedevelopmentaltrajectoriesofRSAandvagaleffi-ciencyweresteeperinFNIinfants,valuesforvagalefficiencywereactuallylowerinFNIinfantsattheearlytimepoint.ItisinterestingthatforsomebrainregionswenotedasimilareffectonEEGpower.Thatis,insomebrainregionsEEGpowerwaslowerinFNIinfantsattheearlytimepoint,butincreasedmorerapidlytotermage(Welchetal.,2014,2017).Itispossiblethattherandomassignmentproce-dureusedintheRCTactuallycreatedgroupsthathadfeaturesthatwerenotequivalentat the timeofenrollment.Alternatively, sincetheearlyassessmenttimepointat~35weeksPMAoccurredonav-erageabout5weeksafterthestartoftheintervention,it ispossi-blethatveryearly intheNICUstaytheFNIslowedinitialratesofdevelopmentbeforesubsequentlypromotingacceleration.ResultsfromanongoingmultisitereplicationtrialofFNI(Clinicaltrials.gov:NCT01439269)willhelpdistinguishthesealternatives.

TheresultsindicatethatRSAisselectivelyaffectedbyFNI,whileheartrateandrespiratoryrate(RSAf)werenot.ThelackofaneffectonRSAf,suggeststhattheintervention'seffectonRSA(i.e.,cardiacvagal tone) is likelyduetoadirectconsequenceofamaturationalchange inneural regulationof theheartvia thevagusandnot in-directlyduetochanges in respiratoryrate. Inaddition, the lackofaneffectonheartrate,suggeststhatthepositiveimpactofFNIonneuralmaturationofvagalpathwaysisnotdependentonachangeinbasalmetabolicrate.

Resultsreportedhere,asexpected,showedreliablesleepstate‐relateddifferencesinheartrateandRSA.Thesefindingssupporttheconclusionthatthemethodologiesused inthiscurrentstudywerevalid measures of these physiological markers of early autonomicfunction.Toourknowledge,thisisthefirststudytoshowthatthedevelopmentalslopeforvagalefficiencyisgreaterwheninfantsareinactive sleep.Thesestate‐relateddifferences inautonomic func-tion,evenpriortotermage,demonstratetheimportanceofmonitor-ingsleepstatewheninterpretingmeasuresofautonomicfunction.

Thereareseverallimitationstotheassessmentsmadeinthiscur-rentstudy.Theautonomicmeasureswereobtainedonlyduringsleep.Thereareclearadvantagestoacquiringphysiologicaldatawhenin-fantsareatrestwithminimalmovementdisturbances.However,itwouldbeimportanttodocumentiftheeffectsofFNIcanbedemon-stratedinawake‐infants.Inaddition,itwouldalsobeofinteresttomonitorthedynamicsofautonomicstatewhilebeingheldbytheirmothers. Based on our theoretical perspective, we would expectthatfollowingseveralweeksofemotionalengagementandconnec-tion,thevagaleffects,whichwehavereported,wouldbeevenmorepronounced during periods of maternal contact. This study alsodidnottestourhypothesisthatthechangesinvagalefficiencywereporthereafterFNIwere the resultofPavlovianconditioningorothermechanismsrelatedtoanassociationwithasafeandsocially/

F I G U R E 4  Astheinfantsapproachedterm,increasesinvagalefficiencyweresignificantlygreaterininfantsreceivingFamilyNurtureInterventioncomparedtothestandardcaregroup(p=0.005;Cohen'sd=0.62;errorbarsrepresentstandarderror)

     |  9PORGES Et al.

emotionally engaging parent.With only two points,we could notdocumentwhethertheeffectsoftheinterventionemergedindirectassociationwithinterventionsessions.Giventheheterogeneityandmultiplecomorbiditiesthatpreterminfantsexperience,itwouldnotbepossiblewiththeavailablesamplesizeandexperimentaldesigntocreatesubgroupsbasedonotherclinical features.Suchastudycould answer the potential role that other clinical conditions playin the sensitivity of the autonomic nervous system to FNI. Theseimportantquestionswouldrequirealargemultisitestudyevaluatingseveralhundredpreterminfantstobalancethesubjectsintosmallergroupsbasedonotherclinicalfeatures.

There are additional important aspects to our findings thatshouldbeconsideredfurther.First,whileourmeasureofvagalef-ficiencyinFNIsubjectsdoesincreasemorerapidlyfrom~35weeksPMAto~41weeksPMA,thevaluesat~35weeksarelowerinFNIwith no group difference apparent near‐to‐term age. The averageage at birth in these subjectswas ~30weeks PMA and the inter-vention began, on average, at ~1week after birth.All FNI infantshadreceivedatleastsomeinterventionpriortotheassessmentsat~35weeks.Thus,itmaybepossiblethatthereisabiphasicresponsetotheinterventionwithaninitialeffectofdecreasingvagaltoneandvagalefficiencyfollowedbyaperiodduringwhichtheinterventionpromotes more rapid increases. Although we would predict thatFNI infantswouldcontinueonthismorerapid increasetrajectory,we do not have repeated assessments during the early post‐dis-chargeperiodtoevaluatethishypothesis.Anotheralternativemightbe that in SC infants one of the effects of stress associatedwithpretermdeliveryandNICUhospitalization is that for someperiodof time after birth, these stressors acceleratematurationof para-sympathetic, andpossiblyother systems.But, forFNI infants, theeffectsofthesestressorsarebufferedbytheinterventionandthus,at~35weeksPMASCinfantshavehighervagaltoneandefficiencythanFNI.WithextensionoftheNICUstay,theeffectsoftheNICUstressors began to diminish, either due to “burn out” or perhapsadaptation. In FNI infants, the early buffering of effects of stresscontinueandarethenovertakenbypromotionofmaturationassoci-atedwiththebenefitsofphysiologicalandemotionconnectionme-diatedby the intervention.Thesehypotheses,whileplausible,willbedifficulttotest.But,inongoingreplicationtrials(NCT03267043;NCT02710474;NCT03442439), aswell asa trial inpreschool agechildren (NCT03442439), we are obtaining data about emotionalconnectionrepeatedly throughout theNICUstayusing theWelchEmotionalConnectionScreen,anewlydevelopedscreenerforemo-tionalconnection(Haneetal.,2018).Wehypothesizethatthetimecourse of increases in emotional connection will parallel changesinautonomicmeasuresandthus informtosomeextentthisunex-pectedpattern inphysiological change. Finally,we are aware thatacceleratedmaturation insomesystemsmaynotalwaysbebene-ficialastherecanbemismatchamongmultiplesystemcapabilities.Moreover,rapiddevelopmentalincreasesmaynotalwaysberelatedtorapidmaturationbutratherreflectongoingstress.Keepingthesealternativesinmindisofcoursecriticallyimportantwhenevaluatingandinterpretinganydevelopmentalintervention.

Inconclusion,thisstudysupportstheroleofFNIinenhancingthedevelopmentofautonomicregulationinpreterminfants.ThevaluesofRSAandvagalefficiencyobservedatnearterminFNIinfantsweresimilartopriorvaluesforRSAandvagalefficiencymeasuredinfullterminfants(Porgesetal.,1999).Thesefindingssuggest thatFNIpromotesnormalization inmaturation in thesesystems.Theresultspresentedherealsoenrichourunderstand-ingofhowFNImightmediate long‐termbeneficialoutcomesbydrawing upon both the polyvagal theory (Porges, 1995; Porgesetal.,1996)whichprovidesanunderstandingofhowautonomicstatesareaneurophysiologicalplatformforoptimalco‐regulationbetween infantandcaregiver,andcalmingcycle theory (Welch,2016;Welch& Ludwig, 2017b;Welch& Ludwig, 2017b)whichprovidesamodel forunderstandinghowrepeated, calming, co‐regulatory mother/infant interactions positively reinforce andcondition approach, prosocial autonomic states. By combiningtheseideas,anewtheoryofchangeforthelastingimpactofFNIemerges.Inthisway,facilitatingmother/infantphysiologicalco‐regulation and emotional connection in theNICU is a practica-bleandeffectivemeansofoptimizingpostnataldevelopmentofautonomicregulationandneurobehavioraloutcomes inpreterminfants.

CONFLIC T OF INTERE S T

Theauthorshavenoconflictsofinterest.

ORCID

Martha G. Welch https://orcid.org/0000‐0002‐8590‐1355

R E FE R E N C E S

Agrawal,S.,Rao,S.C.,Bulsara,M.K.,&Patole,S.K.(2018).Prevalenceof autism spectrum disorder in preterm infants: A meta‐analysis.Pediatrics,142,https://doi.org/10.1542/peds.2018‐0134

Barfield,W.D.(2018).Publichealthimplicationsofverypretermbirth.Clinics in Perinatology, 45, 565–577. https://doi.org/10.1016/j.clp.2018.05.007

Benzies, K. M., Magill‐Evans, J. E., Hayden, K. A., & Ballantyne, M.(2013). Key components of early intervention programs forpreterm infants and their parents: A systematic review andmeta‐analysis. BMC Pregnancy Childbirth, 13(Suppl. 1), S10. https://doi.org/10.1186/1471‐2393‐13‐S1‐S10

Berntson,G.G.,Cacioppo,J.T.,&Quigley,K.S.(1995).Themetricsofcardiacchronotropism:Biometricperspectives.Psychophysiology,32,162–171.https://doi.org/10.1111/j.1469‐8986.1995.tb03308.x

Brydges,C.R., Landes, J.K.,Reid,C. L.,Campbell,C.,French,N.,&Anderson,M. (2018).Cognitiveoutcomesinchildrenandadoles-centsbornverypreterm:Ameta‐analysis.Developmental Medicine and Child Neurology, 60, 452–468. https://doi.org/10.1111/dmcn.13685

Chambers,J. (2017).Theneurobiologyofattachment:Frominfancytoclinical outcomes. Psychodynamic Psychiatry,45, 542–563. https://doi.org/10.1521/pdps.2017.45.4.542

Chertok,I.R.,McCrone,S.,Parker,D.,&Leslie,N.(2014).Reviewofin-terventionstoreducestressamongmothersofinfantsintheNICU.

10  |     PORGES Et al.

Advances in Neonatal Care, 14, 30–37. https://doi.org/10.1097/ANC.0000000000000044

Denver, J. W., Reed, S. F., & Porges, S. W. (2007). Methodologicalissues in the quantification of respiratory sinus arrhythmia.Biological Psychology, 74, 286–294. https://doi.org/10.1016/j.biopsycho.2005.09.005

DiPietro,J.A.,&Porges,S.W. (1991).Vagal responsivenesstogavagefeedingas an indexofpretermstatus.Pediatric Research,29, 231–236.https://doi.org/10.1203/00006450‐199103000‐00003

Feldman, R., & Eidelman, A. I. (2003). Skin‐to‐skin contact (KangarooCare) accelerates autonomic and neurobehavioural maturation inpreterm infants. Developmental Medicine and Child Neurology, 45,274–281.https://doi.org/10.1111/j.1469‐8749.2003.tb00343.x

Field,T.(2018).Postnatalanxietyprevalence,predictorsandeffectsondevelopment:Anarrative review. Infant Behavior and Development,51,24–32.https://doi.org/10.1016/j.infbeh.2018.02.005

Hane, A. A., LaCoursiere, J. N.,Mitsuyama,M.,Wieman, S., Ludwig,R. J., Kwon, K. Y., …Welch, M. G. (2018). TheWelch EmotionalConnection Screen (WECS): Validation of a brief mother‐infantrelationalhealth screen.Acta Paediatrica, https://doi.org/10.1111/apa.14483

Hane,A.A.,Myers,M.M.,Hofer,M.A.,Ludwig,R.J.,Halperin,M.S.,Austin, J.,…Welch,M.G. (2015). Family nurture intervention im-provesthequalityofmaternalcaregiving in theneonatal intensivecare unit: Evidence from a randomized controlled trial. Journal of Developmental and Behavioral Pediatrics, 36, 188–196. https://doi.org/10.1097/dbp.0000000000000148

Hering,H.E.(1910).Functionaltestoftheheartvagiinman.Munch Med Wochenschr,1910,1931–1933.

Korja,R.,Latva,R.,&Lehtonen,L.(2012).Theeffectsofpretermbirthonmother‐infantinteractionandattachmentduringtheinfant'sfirsttwo years.Acta Obstetricia Et Gynecologica Scandinavica, 91, 164–173.https://doi.org/10.1111/j.1600‐0412.2011.01304.x

Lewis, G. F., Furman, S. A., McCool, M. F., & Porges, S. W. (2012).Statistical strategies to quantify respiratory sinus arrhythmia: Arecommonlyusedmetricsequivalent?Biological Psychology,89,349–364.https://doi.org/10.1016/j.biopsycho.2011.11.009

Moore,E.R.,Bergman,N.,Anderson,G.C.,&Medley,N.(2016).Earlyskin‐to‐skincontactformothersandtheirhealthynewborninfants.Cochrane Database Systematic Review, 11, CD003519. https://doi.org/10.1002/14651858.CD003519.pub4

Myers,M.M.,Grieve,P.G.,Stark,R.I.,Isler,J.R.,Hofer,M.A.,Yang,J.,…Welch,M.G. (2015).FamilyNurture Intervention inpreterm in-fantsaltersfrontalcorticalfunctionalconnectivityassessedbyEEGcoherence.Acta Paediatrica,104,670–677.https://doi.org/10.1111/apa.13007

Ohlsson,A.,& Jacobs,S.E. (2013).NIDCAP:A systematic reviewandmeta‐analysesofrandomizedcontrolledtrials.Pediatrics,131,e881–893.https://doi.org/10.1542/peds.2012‐2121

Porges, S. W. (1985a). US Patent No. 4,510,944. Washington, DC: USPatentandTrademarkOffice.

Porges,S.W.(1985b).Respiratory sinus arrhythmia: An index of vagal tone. NewYork,NY:PlenumPress.

Porges,S.W.(1992).Vagaltone:Aphysiologicmarkerofstressvulnera-bility.Pediatrics,90,498–504.

Porges,S.W.(1995).Cardiacvagaltone:Aphysiologicalindexofstress.Neuroscience and Biobehavioral Reviews, 19, 225–233. https://doi.org/10.1016/0149‐7634(94)00066‐A

Porges, S.W. (2007). Thepolyvagal perspective.Biological Psychology,74,116–143.https://doi.org/10.1016/j.biopsycho.2006.06.009

Porges,S.W.(2009).Thepolyvagaltheory:Newinsightsintoadaptivereactionsoftheautonomicnervoussystem.Cleveland Clinic Journal of Medicine,76(Suppl. 2), S86–90. https://doi.org/10.3949/ccjm.76.s2.17

Porges,S.W.(2015).Makingtheworldsafeforourchildren:Down‐regu-latingdefenseandup‐regulatingsocialengagementto'optimise'thehumanexperience.Children Australia,40,114.

Porges,S.W.,&Bohrer,R.E.(1990).Analysis of periodic processes in psy-chophysiological research.NewYork,NY:CambridgeUniversityPress.

Porges,S.W.,Doussard‐Roosevelt, J.A.,Portales,A.L.,&Greenspan,S. I. (1996). Infant regulation of the vagal "brake" predicts childbehavior problems: A psychobiological model of social behavior.Developmental Psychobiology,29,697–712.https://doi.org/10.1002/(SICI)1098‐2302(199612)29:8%3c697:AID‐DEV5%3e3.0.CO;2‐O

Porges, S. W., Doussard‐Roosevelt, J. A., Stifter, C. A., McClenny, B.D.,&Riniolo,T.C.(1999).Sleepstateandvagalregulationofheartperiodpatterns in thehumannewborn:Anextensionof thepoly-vagal theory.Psychophysiology,36, 14–21.https://doi.org/10.1017/S004857729997035X

Porges,S.W.,&Furman,S.A.(2011).Theearlydevelopmentoftheauto-nomicnervoussystemprovidesaneuralplatformforsocialbehavior:Apolyvagalperspective.Infant and Child Development,20,106–118.https://doi.org/10.1002/icd.688

Portales, A. L., Porges, S. W., Doussard‐Roosevelt, J. A., Abedin,M., Lopez, R., Young, M. A., … Baker, M. (1997). Vagal reg-ulation during bottle feeding in low‐birthweight neonates:Support for the gustatory‐vagal hypothesis. Developmental Psychobiology, 30, 225–233. https://doi.org/10.1002/(SICI)1098‐2302(199704)30:3<225:AID‐DEV5>3.0.CO;2‐R

Quigley, K. S., & Berntson, G. G. (1996). Autonomic interac-tions and chronotropic control of the heart: Heart period ver-sus heart rate. Psychophysiology, 33, 605–611. https://doi.org/10.1111/j.1469‐8986.1996.tb02438.x

Sheinkopf,S.J.,Levine,T.P.,McCormick,C.E.B.,Puggioni,G.,Conradt,E.,Lagasse,L.L.,&Lester,B.M.(2019).Developmentaltrajectoriesof autonomic functioning in autism from birth to early childhood.Biological Psychology,142,13–18.

Stefanski, M., Schulze, K., Bateman, D., Kairam, R., Pedley, T. A.,Masterson, J., & James, L. S. (1984). A scoring system for statesof sleep and wakefulness in term and preterm infants. Pediatric Research,18,58–62.

Suess,P.E.,Alpan,G.,Dulkerian,S.J.,Doussard‐Roosevelt,J.,Porges,S.W.,&Gewolb,I.H.(2000).Respiratorysinusarrhythmiaduringfeed-ing:Ameasureofvagalregulationofmetabolism,ingestion,anddiges-tioninpreterminfants.Developmental Medicine and Child Neurology,42,169–173.https://doi.org/10.1017/S001216220000030X

Tang,A.C.,Reeb‐Sutherland,B.C.,Romeo,R.D.,&McEwen,B.S.(2014).On the causesof early life experience effects: Evaluating the roleof mom. Frontiers in Neuroendocrinology, 35, 245–251. https://doi.org/10.1016/j.yfrne.2013.11.002

Verbeke,G.,&Molenberghs,G.(2009).Linearmixedmodelsforlongitu-dinaldata.Linear Mixed Models for Longitudinal Data,1–568.

Welch,M.G.(1988).Holding time: How to eliminate conflict, temper tan-trums, and sibling rivalry and raise happy, loving, sucessful children. New York,NY:SimonandSchuster.

Welch,M.G. (2016). Calming cycle theory: The role of visceral/auto-nomiclearninginearlymotherandinfant/childbehavioranddevel-opment.Acta Paediatrica,105,1266–1274.https://doi.org/10.1111/apa.13547

Welch,M.G.,Firestein,M.R.,Austin,J.,Hane,A.A.,Stark,R.I.,Hofer,M.A.,…Myers,M.M. (2015). FamilyNurture Intervention in theNeonatalIntensiveCareUnitimprovessocial‐relatedness,attention,andneurodevelopmentofpreterminfantsat18monthsinarandom-ized controlled trial. Journal of Child Psychology and Psychiatry,56,1202–1211.https://doi.org/10.1111/jcpp.12405

Welch,M.G.,Halperin,M.S.,Austin,J.,Stark,R.I.,Hofer,M.A.,Hane,A.A.,&Myers,M.M.(2016).Depressionandanxietysymptomsofmoth-ersofpreterminfantsaredecreasedat4monthscorrectedagewith

     |  11PORGES Et al.

FamilyNurtureInterventionintheNICU.Archives of Women's Mental Health,19,51–61.https://doi.org/10.1007/s00737‐015‐0502‐7

Welch,M.G.,Hofer,M.A.,Brunelli,S.A.,Stark,R.I.,Andrews,H.F.,Austin,J.,&Myers,M.M.(2012).Familynurtureintervention(FNI):MethodsandtreatmentprotocolofarandomizedcontrolledtrialintheNICU.BMC Pediatrics,12,14.https://doi.org/10.1186/1471‐2431‐12‐14

Welch, M. G., Hofer, M. A., Stark, R. I., Andrews, H. F., Austin, J.,Glickstein,S.B.,…Myers,M.M.(2013).RandomizedcontrolledtrialofFamilyNurtureInterventionintheNICU:Assessmentsoflengthof stay, feasibility and safety.BMC Pediatrics,13, 148. https://doi.org/10.1186/1471‐2431‐13‐148

Welch, M. G., & Ludwig, R. (2017a). Mother/Infant Emotional Communication Through the Lens of Visceral/Autonomic Learning. New York,NY:SpringerInternationalPublishing.

Welch,M. G., & Ludwig, R. J. (2017b). Calming cycle theory and theco‐regulation of oxytocin. Psychodynamic Psychiatry, 45, 519–540.https://doi.org/10.1521/pdps.2017.45.4.519

Welch,M.G.,Myers,M.M.,Grieve,P.G.,Isler,J.R.,Fifer,W.P.,Sahni,R.,…Stark,R.I.(2014).Electroencephalographicactivityofpreterminfants is increasedbyFamilyNurture Intervention:A randomizedcontrolledtrialintheNICU.Clinical Neurophysiology,125,675–684.https://doi.org/10.1016/j.clinph.2013.08.021

Welch,M.G.,Stark,R.I.,Grieve,P.G.,Ludwig,R.J.,Isler,J.R.,Barone,J.L.,&Myers,M.M.(2017).Familynurtureinterventioninpreterm

infantsincreasesearlydevelopmentofcorticalactivityandindepen-dence of regional power trajectories.Acta Paediatrica,106, 1952–1960.https://doi.org/10.1111/apa.14050

Wolfinger, R. D. (1996). Heterogeneous variance: Covariancestructures for repeated measures. Journal of Agricultural, Biological, and Environmental Statistics, 1, 205–230. https://doi.org/10.2307/1400366

SUPPORTING INFORMATION

Additional supporting information may be found online in theSupportingInformationsectionattheendofthearticle.

How to cite this article:PorgesSW,DavilaMI,LewisGF,etal.AutonomicregulationofpreterminfantsisenhancedbyFamilyNurtureIntervention.Developmental Psychobiology. 2019;00:1–11. https://doi.org/10.1002/dev.21841

View publication statsView publication stats