Effect of Human Milk Appetite Hormones, Macronutrients, and Infant Characteristics on Gastric

21 Pages·2016·0.83 MB·English

nutrients
Article
Effect of Human Milk Appetite Hormones,
Macronutrients, and Infant Characteristics on
Gastric Emptying and Breastfeeding Patterns of
Term Fully Breastfed Infants
ZoyaGridneva1,*,SambaviKugananthan1,2,AnnaR.Hepworth1,WanJ.Tie1,ChingT.Lai1,
LeighC.Ward3,PeterE.Hartmann1andDonnaT.Geddes1
1 SchoolofChemistryandBiochemistry,TheUniversityofWesternAustralia,Crawley,Perth,
WesternAustralia6009,Australia;[email protected](S.K.);
[email protected](A.R.H.);[email protected](W.J.T.);[email protected](C.T.L.);
[email protected](P.E.H.);[email protected](D.T.G.)
2 SchoolofAnatomy,PhysiologyandHumanBiology,TheUniversityofWesternAustralia,Crawley,Perth,
WesternAustralia6009,Australia
3 SchoolofChemistryandMolecularBiosciences,TheUniversityofQueensland,St.Lucia,Brisbane,
Queensland4072,Australia;[email protected]
* Correspondence:[email protected];Tel.:[phone]
Received:29September2016;Accepted:22December2016;Published:28December2016
Abstract: Humanmilk(HM)componentsinfluenceinfantfeedingpatternsandnutrientintake,yet
it is unclear how they influence gastric emptying (GE), a key component of appetite regulation.
This study analyzed GE of a single breastfeed, HM appetite hormones/macronutrients and
demographics/anthropometrics/bodycompositionoftermfullybreastfedinfants(n=41,2and/or
5mo). Stomachvolumes(SV)werecalculatedfrompre-/post-feedultrasoundscans,thenrepeatedly
untilthenextfeed. Feedvolume(FV)wasmeasuredbythetest-weighmethod. HMsampleswere
analyzedforadiponectin,leptin,fat,lactose,totalcarbohydrate,lysozyme,andtotal/whey/casein
protein.Linearregression/mixedeffectmodelswereusedtodetermineassociationsbetweenGE/feed
variablesandHMcomponents/infantanthropometrics/adiposity. HigherFVswereassociatedwith
faster(−0.07[−0.10, −0.03], p<0.001)GErate, higherpost-feedSVs(0.82[0.53, 1.12], p<0.001),
andlongerGEtimes(0.24[0.03,0.46],p=0.033). Higherwheyproteinconcentrationwasassociated
withhigherpost-feedSVs(4.99[0.84,9.13],p=0.023). LongerGEtimewasassociatedwithhigher
adiponectinconcentration(2.29[0.92,3.66],p=0.002)anddose(0.02[0.01,0.03],p=0.005),andlower
casein:wheyratio(−65.89[−107.13,−2.66],p=0.003). FVandHMcompositioninfluenceGEand
breastfeedingpatternsintermbreastfedinfants.
Keywords: human milk; term breastfed infants; gastric emptying; feeding frequency; ultrasound;
stomachvolumes;appetitehormones;macronutrients;feedvolume;anthropometrics;bodycomposition
1. Introduction
Breastfeedinganditslongerdurationareassociatedwithreducedrisksofdevelopingobesity
and other chronic non-communicable diseases later in life [1,2]. This unique protection could be
the result of many mechanisms associated with both nutritive and non-nutritive components of
human milk (HM) [3] as well as breastfeeding patterns and behaviour [4,5]. It has been shown
that HM has the pleiotropic role, providing immune and anti-inflammatory protection [6,7] and
endocrine, developmental, neural, and psychological benefits [2]. Non-nutritive HM components
suchashormones,growthfactors,neuropeptides,andanti-inflammatoryandimmune-modulating
Nutrients2017,9,15;doi:10.3390/nu9010015 www.mdpi.com/journal/nutrients
Nutrients2017,9,15 2of21
agentsinfluencethegrowth,development,andfunctionofthegastrointestinal(GI)tractduringearly
infancy [8], while some micronutrients act as nutritional antioxidants, improving GI functions [9];
however, there is much to be learned about the spectrum of HM programming agents, how their
patternschangethroughoutlactationperiod,andtheirshort-termeffectonthegastricemptying(GE)
rateofthebreastfedinfants.
GEisaprocessbywhichingestedfoodismechanicallyandchemicallypartiallybrokendownand
deliveredtotheduodenumatacontrolledrateforfurtherdigestionandabsorption[10,11]. Whilewell
studiedinthepretermpopulation[12–14],inhealthytermfullybreastfedinfantstheGErateandits
relationshipwithbreastfeedingpatternsarenotfullyunderstood.
GE rate and patterns are known to depend on the nature and macronutrient composition of
theingestedmeal. HMorformulaintheinfantstomachseparatesintotwophases,aliquidphase
consistingofwater,wheyproteins,lactose,etc.,andasemi-solidphaseconsistingofcurdformedby
caseinandlipids. Thesemi-solidphasetypicallyemptiesmoreslowlythantheliquidphase. Different
proportionsofthesephasesinpartexplainthedifferencebetweenGEpatternsofformula-fedand
breastfedinfants—linearandcurvilinear,respectively[12,15].
HMhasauniquecomposition,includingnutrients,growthfactors,immunefactors,andhormones.
Despite numerous investigations into the different effects of HM and formula, few components,
includingmajormacronutrients,havebeenstudiedinconnectionwiththeGEofbreastedterminfants.
FattyacidsprofilesarenotassociatedwithGErateinpreterminfants[16],whileinterminfants
morerapidGEhasbeenattributedtothefatandproteincomponentsoffeedswithsimilarlactose
concentrationandosmolality[17].
BothosmolalityandcarbohydratecontentareknowntoinfluencetherateofGEinadults[18],
butininfantsresultsaredependentonthetypeofcarbohydrate[19,20].
ProteinsfromdifferentHMfractionssuchaswheyandcaseinareresistanttoproteolysisinthe
infantstomach[21]andtheproteincontentofafoodhasalsobeenshowntoinfluenceappetiteand
itsregulation[22]. InfantformulagenerallyemptiesmoreslowlythanHMinterminfants;further,
formulaswithdifferentcasein:wheyproteinratioexhibitdifferentGErates,withcasein-predominant
formulasemptyingslowerthanwhey-predominantformulas[23]. Thusthecasein:wheyratioofHM
couldplayanimportantroleincontrollingGEinthebreastfedinfant.
HMlysozyme,alsopresentinwheyinarelativelyhighconcentration,catalyzesthehydrolysis
ofspecificbondsinGram-negativebacteriacellwallsandplaysmultiplerolesindigestivestrategy,
suchascontrollingthemicrobiomeinthestomachandspeedingupthedigestionofmicrobialprotein,
whichmayaffectgastricmotilityandGErate[24,25].
Thesatietyhormoneleptinandtheappetite-stimulatinghormoneadiponectinarealsopresent
inHM.Althoughnottransferredtotheinfantcirculationindirectmanner,levelsofHMleptinand
adiponectinfromHMhavebeenfoundtocorrelatewithlevelsofthesehormonesininfantserum[26,27]
andareknowntoaffectbothappetitecontrolandinfantbodycomposition(BC)[28,29],butareyettobe
investigatedinrelationtoGEintheterminfant. Inanimalmodels(rat,mouse),injectionofleptininto
thefourthventriclehasbeenshowntodelayGE[30]andoraladministrationreducedfoodintake[31].
Leptin in HM is by far the most studied appetite hormone, but predominantly in skim milk [32].
LeptinmeasuredinskimHMwasnotassociatedwithtimebetweenfeeds[33,34]orGE[34]interm
breastfedinfants,emphasizingtheneedforstudiesincludingwholemilkleptin,wherethelevelsof
leptinareshowntobehigher[32]. Adiponectinhasthehighestconcentrationofanyappetitehormone
inHM.Itispresentinabiologicallyactiveformthatisresistanttodigestion[35]. Intheanimalmodel
adiponectininhibitstension-sensitivegastricvagalafferentmechanosensitivity,modulatingsatiety
signalsinbothleanandobeseanimals,whilesimultaneouslyincreasingthemechanosensitivityof
mucosalgastricvagalafferentintheobesity-inducedmodel[36]. Inhumans,elevatedserumlevels
ofadiponectinareassociatedwithmorerapidGEindiabeticpatients[37]. Itisnotknownwhether
adiponectinlevelsimpactGEintheinfantandthiswarrantsfurtherinvestigation.
Nutrients2017,9,15 3of21
Thevolumeofmilktakenatasinglefeedvariesgreatlybothwithinandbetweeninfants[38].
ThismaybeaffectedbyHMcomposition,withgreaterbreastfeedingfrequencyassociatedwithlower
total 24-h protein intakes and higher lactose concentrations [39]. This suggests that the variations
in HM components between mothers may potentially influence GE rate and time, and therefore
feedingpatterns.
This study investigated the effects of HM appetite hormones (whole milk adiponectin and
leptin,skimmilkleptin)andmacronutrients(fat,totalcarbohydrates,lactose,oligosaccharides,total
protein, caseinandwheyprotein,lysozyme)onfeedingfrequencyandGE.Furtherexplorationof
infantdemographics,anthropometrics,andBCwascarriedouttodeterminerelationshipswithinfant
feedingandGE.
2. MaterialsandMethods
2.1. Participants
Lactatingmothersandtheirinfants(n=27)wererecruitedpredominantlythroughtheAustralian
BreastfeedingAssociation. Inclusioncriteriawere: healthysingletons,gestationalage≥37weeks,
fullybreastfedondemandatthepointofmeasurement. Exclusioncriteriawere: infanthealthissues
requiringmedicationthatcouldpotentiallyinfluenceGErate(e.g.,reflux),indicationsoflowmaternal
milkproductionorinfantgrowthissues. Allmothersprovidedwritteninformedconsenttoparticipate
inthestudy,whichwasapprovedbytheUniversityofWesternAustralia,HumanResearchEthics
Committee (RA/1/4253) and registered with the Australian New Zealand Clinical Trials Registry
(ACTRN12616000368437).
2.2. StudyDesign
ParticipantsarrivedatourlaboratoryatKingEdwardMemorialHospitalforWomen(Subiaco,
Perth,WA,Australia)inthemorning(09:30–11:30a.m.) toavoidcircadianinfluenceontheoutcomes,
andstayedfortwoconsecutivebreastfeedingsessions. Beforethefirstfeed(F1)infantswereweighed
andhadultrasoundstomachvolumesrecorded(pre-feedresidual,R1). Mothersexpressedapre-feed
sample(fore-milk)ofmilkfromthefeedingbreast/breastsandthenbreastfedtheirinfantsasusual.
ImmediatelyafterF1,infantstomachvolumesimagesandinfantweightsweretaken,andmothers
expressedapost-feed(hind-milk)milksample. Subsequentscansofthestomachwerescheduledat
15–20minintervals(althoughattendinginfants’needscausedsomevariation)untiltheinfantcued
forthenextfeed(F2),whenafinalstomachvolumeimmediatelybeforeF2wasmeasured(pre-feed
residual,R2).
To assess infant BC bioimpedance spectroscopy measurements were taken pre-feed, unless
impractical—thentheyweretakenpost-feed[40]. Ultrasoundskinfold,length,andheadcircumference
measurements were taken post-feed. This combination of two methods for measuring infant BC
wasusedtoensuresafe,non-invasiveandaccurateassessmentandtoavoidtheinherentlimitations
ofasingulartechnique[41]. Clothingwasremovedforthemeasurementsexceptforadrydiaper
andasinglet.
2.3. FeedingFrequency
Motherswereaskedhowfrequentlytheirinfantfeeds,andtheself-reportedtypicaltimebetween
thefeeds(e.g.,everythreehours)duringtheweekpriortothestudysessionwastakenasaproxy
measureoffeedingfrequency.
2.4. FeedVolumeMeasurement
Thevolumeofmilktransferredfromabreast/breastsbytheinfantwasdeterminedbyweighing
theinfantimmediatelybeforeandafterthebreastfeedusingelectronicscales(±2.0g,MedelaElectronic
BabyWeighScales,MedelaInc.,McHenry,IL,USA).Milkintake(g)wascalculatedbydeductingthe
Nutrients2017,9,15 4of21
Nutrients 2017, 9, 015  4 of 21 
initialweightfromthefinalweightoftheinfant[42]andwasconvertedtomL(feedvolume;FV)using
HMdeddeuncstiitnygo tfh1e .0in3itgia/lm wLei[g4h3t] .from the final weight of the infant [42] and was converted to mL (feed 
volume; FV) using HM density of 1.03 g/mL [43]. 
2.5. StomachMeasurementswithUltrasound
2.5. Stomach Measurements with Ultrasound 
Theinfant’sstomachwasscannedusingtheAplioXG(Toshiba,Tokyo,Japan)machine,witha
The infant’s stomach was scanned using the Aplio XG (Toshiba, Tokyo, Japan) machine, with a 
high-resolutionPVT-674BT(6MHz)transducerandParkerultrasonicgel(Fairfield,NJ,USA).Threeto
high‐resolution PVT‐674BT (6MHz) transducer and Parker ultrasonic gel (Fairfield, NJ, USA). Three 
nine(median[IQR]:5[5; 6])serialmeasurementsofinfantstomachsweretaken3to62minapart
to nine (median [IQR]: 5 [5; 6]) serial measurements of infant stomachs were taken 3 to 62 min apart 
(16±10). Scanswereperformedwiththeinfantinthesemi-supinepositionaccordingtothemethod
(16 ± 10). Scans were performed with the infant in the semi‐supine position according to the method 
validatedinpreterminfants[44]. Briefly,thesagittalandtransverseplanesofthestomachwereused
validated in preterm infants [44]. Briefly, the sagittal and transverse planes of the stomach were used 
tomeasurethelongitudinal(L),anterior-posterior(AP)andtransverse(T)diametersdirectlyfrom
to measure the longitudinal (L), anterior‐posterior (AP) and transverse (T) diameters directly from 
imagesontheultrasoundscreenusingelectroniccalipers(Figure1). Oneexperiencedsonographer
images on the ultrasound screen using electronic calipers (Figure 1). One experienced sonographer 
withgoodintra-andinterraterreliability[44]performedallofthemeasurements. Gastricvolume(mL)
with good intra‐ and interrater reliability [44] performed all of the measurements. Gastric volume 
was(mcaLl)c wulaast ecdalcfruolamtetdh feroambo tvhee mabeoavseu rmedeadsuiarmede dteiarsmuesteinrsg ufsoilnlogw foilnlogweiqnuga etiqounatfioorna fnore allnip eslloipidsaolidbaold  y:
body: 
Stomachvolume(mL)=L(mm)×AP(mm)×T(mm)×0.52. (1)
Stomach volume (mL) = L (mm) × AP (mm) × T (mm) × 0.52.  (1) 

Figure 1. Measurements of infant’s stomach with ultrasound. Ultrasound images of infant’s stomach: 
Figure 1. Measurements of infant’s stomach with ultrasound. Ultrasound images of infant’s
(a) transverse view with anterior‐posterior (AP) and transverse (T) diameter measurements; (b) 
stomach:(a)transverseviewwithanterior-posterior(AP)andtransverse(T)diametermeasurements;
longitudinal view with longitudinal (L) diameter (maximum length) measurement. Stomach volume 
(b)longitudinalviewwithlongitudinal(L)diameter(maximumlength)measurement.Stomachvolume
(mL) = longitudinal diameter (mm) × anterior‐posterior diameter (mm) × transverse diameter (mm) × 
(mL)=longitudinaldiameter(mm)×anterior-posteriordiameter(mm)×transversediameter(mm)×0.52.
0.52. 
2.6.2.M6. iMlkiSlka mSapmlepCle oClloelcleticotinon 
MoMthoetrhsehrsa nhda-nedx‐perxepsrseesdseodr pour mpupmedpsemd aslml(a1l–l 2(1m–L2 )mprLe)- panred‐ paonsdt- fpeoesdt‐mfeieldk smamilkp lseasminptloess eipnator ate
5-mseLppaoraltyep r5o‐mpyLl epnoelypplarostpiyclveniael sp(lDasitsipc ovsaiablsle (PDriosdpouscatbs,leA Pdreoladiudcets,,S AA,dAeluaisdtrea, lSiaA)., FAautsctoranlciae)n. trFaatti on
wascomnceeansturraetidon(b welaosw m)aenasdusreadm p(bleelsowwe) raenfdro szaemnpaltes− w20er◦eC frfoozrefnu ratht e−r20b i°oCch feomr ifcuarlthaenra blyiosicsh.emical 
analysis. 
2.7. BiochemicalAnalysis
2.7. Biochemical Analysis 
2.7.1. FatContent
2.7.1. Fat Content 
Percentagefatwasmeasuredinpre-andpost-feedsamplesimmediatelyaftersamplecollection
withthePcerrecaemntaatgoec fraitt mwaesth moeda[s4u5r]edu siinn pgrteh‐e aCndre paomsta‐tfoeecrdi tsaPmlupsldese vimicmee(Mdieadteellya aIfntecr., sMamcHpleen croyl,lIeLct,iUonS A).
Fatwcoitnhc tehnet rcarteiaomnaotfotchreit pmree-thaondd [p4o5]s tu-fseinedg mthiel kCsraemampaletosc(rgit/ PLl)uws adsevcaiclec u(lMateeddelfaro Imnc.t,h MeccrHeeanmryc,o InLt,e nt
oftUheSAm).i lFkast acmonpcleenst,rbaatisoend oof nthteh pereeq‐ uaantdio pno[s4t‐6f]e:ed milk samples (g/L) was calculated from the cream 
content of the milk samples, based on the equation [46]: 
Fat(g/L)=3.56+(5.917×creampercentage). (2)
Fat (g/L) = 3.56 + (5.917 × cream percentage).  (2) 
Fat concentration in the volume consumed by the infant was further calculated [47]:
Nutrients2017,9,15 5of21
Fatconcentrationinthevolumeconsumedbytheinfantwasfurthercalculated[47]:
Fat(g/L)=0.53×Fat +0.47×Fat . (3)
pre-feed post-feed
2.7.2. SamplePreparation
Prior to further analysis, all samples were thawed for two hours at room temperature (RT)
and aliquoted into 1.5-mL tubes (Sarstedt, Numbrecht, Germany). Components’ concentrations
were determined in both pre- and post-feed samples in case of adiponectin, skim and whole milk
leptin, fat, and lactose, and in pooled samples in case of total protein, casein, whey protein, total
carbohydrates,andlysozyme. Concentrationsofpre-andpost-feedsampleswereaveragedtoarrive
attheconcentrationusedforstatisticalanalyses. Wholemilkwasusedformeasuringwholemilk
adiponectin and leptin concentration. Milk samples were defatted (by centrifugation at RT in a
BeckmanMicrofuge11(AberdonEnterpriseInc.,ElkGroveVillage,IL,USA)at10,000×gfor10min
andremovingthefatlayerbyclippingitoffwiththetopofthetube[48])foranalysisofskimmilk
leptin,totalprotein,lysozyme,lactose,andtotalcarbohydratesconcentrations. Thestandardassays
wereadaptedforandcarriedoutusingaJANUSworkstation(PerkinElmer,Inc.,Waltham,MA,USA)
andmeasuredonEnSpire(PerkinElmer,Inc.,Waltham,MA,USA).
2.7.3. Leptin
Leptin concentration in HM was measured using the R & D Systems Human Leptin enzyme
linkedimmunosorbentassay(ELISA)DuoSetkit(Minneapolis,MN,USA)optimizedtomeasureleptin
insonicatedskimHM,aspreviouslydescribedbyCannonetal.[33]andfurthermodifiedtomeasure
leptininskimandwholeHMmilkasdescribedbyKugananthanetal.[32]. Recoveryofleptinwas
97.7%±9.7%(n=10)withadetectionlimitof0.05ng/mLandaninter-assayCVof<7.2%.
2.7.4. Adiponectin
AdiponectinconcentrationinwholemilkwasmeasuredusingtheBiovendorHumanAdiponectin
SandwichELISAkit(LifeTechnologies,Asheville,NC,USA).Adiponectinrecoverywas96.2%±3.2%
(n=10)withadetectionlimitof1ng/mLandaninter-assayCVof<2.5%.
2.7.5. Protein
Casein and whey proteins were separated by the method fully described by Kunz and
Lonnerdal[49], andKhanetal.[50]. Proteinconcentrations(totalproteinofskimHM,caseinand
wheyproteins)weremeasuredusingtheBradfordProteinAssayadaptedfromMitoulasetal.[51].
Recoveryofproteinwas100.6%±5.2%(n=5)withadetectionlimitof0.031g/Landaninter-assay
CVof7.8%(n=18). Casein:wheyratiowascalculatedasfollows:
Casein:wheyratio=caseinconcentration/wheyproteinconcentration. (4)
2.7.6. Lysozyme
Lysozymeconcentrationwasdeterminedusingamodifiedturbidimetricassay[52]. Henegg
whitelysozyme(EC3.2.1.17,Sigma,St. Louis,MA,USA)standards(range0.00075–0.0125g/L)and
skimmilksampleswerediluted10-foldwith0.1MofNa HPO /1.1mMofcitricacid(pH5.8)buffer.
2 4
Twenty-five microliters of standards or diluted skim milk samples were placed into the wells of
aplate(GreinerBio-One,Frickenhausen,Germany),175µLofMicrococcuslysodeiltikussuspension
(0.075% w/v, ATCC No. 4698, Sigma, St. Louis, MA, USA) was added into each well and plate
wasincubatedatRTfor1h. Theabsorbancewasmeasuredat450nm. Recoveryoflysozymewas
97.0%±5.0%(n=8)withadetectionlimitof0.007g/Landaninter-assayCVof13.0%(n=8).
Nutrients2017,9,15 6of21
2.7.7. Carbohydrates
Defattedmilkwasdeproteinizedwithtrichloroaceticacid[53]beforedehydrationbysulphuric
acid[54]. Thistechniquereliablyestimatesconcentrationsandcarboncontentformonosaccharides,
disaccharides,andpolysaccharides. TotalcarbohydrateswereanalyzedbyUV-spectrophotometry.
Recoveryoftotalcarbohydrateswas101.4%±2.1%(n=7)withadetectionlimitof0.007g/Landan
inter-assayCVof3.3%(n=7).
LactoseconcentrationwasmeasuredusingtheenzymaticspectrophotometricmethodofKuhn
and Lowenstein [55], adapted from Mitoulas et al. [51], with recovery of 98.2% ± 4.1% (n = 10),
detectionlimitof30mMandinter-assayCVof3.5%.
The human milk oligosaccharides (HMO) concentration (g) was calculated by deducting
concentrationoflactose(g)fromconcentrationoftotalcarbohydrates(g). Theglucoseandgalactose
werenotmeasuredoraccountedforastheirconcentrationsinHMaresmallandcomparableorless
thantheassayserrors[56].
2.8. HormoneandMacronutrientDose
Doses were defined as the amount of hormone/macronutrient ingested during a breastfeed
and calculated as average of the pre- and post-feed HM component concentration, multiplied
by the corresponding FV. When an infant fed from both breasts at the breastfeeding session,
hormone/macronutrient doses from these individual breastfeeds were calculated separately and
addedtogether.
2.9. Infants’AnthropometricsandBodyComposition
2.9.1. AnthropometricMeasurements
Infants’weightwasdeterminedbyweighingbeforebreastfeedingusingMedelaElectronicBaby
WeighScales(±2.0g;MedelaInc.,McHenry,IL,USA).Clothingwasremovedexceptforadrydiaper
andasinglet. Infantcrown-heellengthwasmeasuredoncetothenearest0.1cmusingnon-stretchtape
andheadpieceandfootpiece,bothappliedperpendiculartothehardsurface.Infantheadcircumference
wasmeasuredwithnon-stretchtape. InfantBMIwascalculatedaccordingtothefollowingformula:
BMI=Bodyweight(kg)/(Height(m))2. (5)
2.9.2. BodyCompositionwithBioelectricalImpedanceSpectroscopy
Infants’ whole body bioimpedance were measured using the Impedimed SFB7 bioelectrical
impedanceanalyzer(ImpediMed,Brisbane,Queensland,Australia)applyinganadultprotocol(wrist
to ankle) according to the manufacturer’s instructions and analyzed with settings customized for
eachinfantaccordingtoLingwoodetal.[57]andGridnevaetal.[41]. Valuesofresistance(ohm)at
frequencyof50kHz(R )weredeterminedfromthecurveofbestfit,averagedforanalysispurposes
50
andusedintheLingwoodetal. agematched(3and4.5moinfants)equationsforfat-freemass(FFM)
of2and5moinfantsrespectively[57]:
FFM3mo=1.458+0.498×W−0.197×S+0.067×L2/R (6)
50
FM4.5mo=2.203+0.334×W−0.361×S+0.185×L2/R , (7)
50
where L is body length (cm), R is resistance (Ω), S is sex (male = 1, female = 2) and W is infant
50
weight(kg).
%FMwascalculatedasfollows:
%FM=100(Weight(kg)−FFM(kg))/Weight(kg). (8)
Nutrients2017,9,15 7of21
2.9.3. BodyCompositionwithUltrasoundSkinfoldMeasurements
InfantultrasoundskinfoldmeasurementswerecarriedoutusingtheAplioXG(Toshiba,Tokyo,
Japan)ultrasoundmachine,PLT-1204BX14-8MHztransducerandsterilewater-basedParkerultrasonic
gel(Fairfield,NJ,USA).Singleultrasoundscansoffouranatomicalsites(biceps,subscapular,suprailiac,
andtriceps)wereperformedontheleftsideofthebodywithminimalcompression. Skinfoldthickness
(skinthicknessandtheskin–fatinterfacetofat–muscleinterfacedistance)wasmeasureddirectlyfrom
imagesonthescreenusingelectroniccalipers. Oneexperiencedsonographer(DG)withgoodintra-
andinterraterreliability[44]performedallofthemeasurements.
The doubled ultrasound skinfold thickness was used in Brook body density (d) age-matched
(3–18mo)equations[58]developedforskinfoldsmeasuredwithcalipers:
∑
Maled=1.1690−0.0788×log( SFT) (9)
∑
Femaled=1.2063−0.0999×log( SFT), (10)
wheredisinfantbodydensity(kg/L)and∑SFTisasumoffourskinfolds(mm).
Predictedbodydensitywasconvertedto%FMusingtheLohmanequation[59]:
%FM=100×(5.28/d−4.89), (11)
wheredistheinfantbodydensity(kg/L).
2.10. StatisticalAnalysis
StatisticalanalysiswasperformedinR2.9.0[60]forMacOSXusingadditionalpackagesnlme[61];
lattice[62],latticeextra[63],andcar[64];MASS[65],sfsmisc[66]andmultcomp[67]formixedeffects
modeling,datarepresentation,robustregression,andmultiplecomparisonsofmeans,respectively.
Descriptivestatisticsarereportedasmean±standarddeviation(SD)(range)ormedian(IQR)unless
otherwise stated; model parameters are presented as estimate ± standard error (SE), and, where
appropriate,anapproximate95%confidenceinterval(95%CI).
Measurementsmissingduetoinsufficientsamplevolume: skimmilkleptin,wholemilkleptin,
adiponectin,totalprotein,wheyandcaseinprotein,lactoseandtotalcarbohydrate(n=3);lysozyme
(n=5). Measurementsoffat(n=14)weremissingasaresultofeitherinsufficientsamplevolumes
orabsenceofseparatefeedvolumesfrombreastswherebothbreastwereofferedduringonefeed.
Alsomissingwerefeedingfrequencyasreportedbymothers(n=6),measurementsoflength,head
circumference,infantBMI,%FMmeasuredwithbioelectricalspectroscopy(n=4)and%FMmeasured
withultrasoundskinfolds(n=5).
GEtimewasdeterminedasthetimefromthestartofF1tothestartofF2andincludedthetime
betweentwofeedsandfeedduration. Feeddurationwasincludedasupto80%ofHMconsumedby
termhealthybreastfedinfantsinthefirst4–5min[68]. GEduringbreastfeedingwasdefinedasthe
volumeofmilktohaveleftthestomach,calculatedasthedifferencebetweentheimmediatepost-feed
stomachvolumesandthesumofR1andFV.
Duetothelackofterminfantgastric-emptyingstudiesfocusingonstomachvolume,nopower
calculation/sample size determination could be performed for this study. A goal of 20 infants at
each two and five months was selected with the expectation that this would be sufficient to show
overallpatterns. Whenavailable,infantswereincludedinbothsubsetstoallowforinvestigationof
longitudinalpatterns. Linearmixedeffectsmodelsallowustotreattheindividualfeedsasseparate,
withouthavingtoassumeindependence,whentheremaybecorrelationsbetweenfeedswithininfants.
Influences on GE rate were analyzed by first fitting a time curve to the sequential post-feed
stomach volumes using linear mixed effects models; as curves differed significantly within and
betweeninfants(p<0.001),randomtimecurveswerefittedtofeedswithininfants. Timeterms(linear,
squareroot)wereselectedasperthefractionalpolynomialmethodof[69];thismodelalsoconsidered
Nutrients2017,9,15 8of21
possibleconfoundingeffectsofFV(median-centred)andfeedduration(median-centred). Interaction
termsinvolvingthetimecurveindicatedchangesintheGErate;maineffectsindicatedoveralleffects
on post-feed stomach volumes but not the GE rate. The addition of one term to this base model
was used to investigate associations with (a) concentrations/doses of hormones/macronutrients;
(b)infantcharacteristics/anthropometrics/BC;(c)R1. WhethertheoveralleffectofHMcomponent
concentrationsdiffersbyfeedvolumewasinvestigatedbyincludinginteractionsbetweenFVand
concentrationmeasures. Modelsusingtheselectedtechniquedidnotconvergeforfatconcentration,
lysozymeconcentration,orlysozymedose. Omittingtherandomeffectoffeedwithininfantprovided
converging models, but no evidence of an association with fat or lysozyme was seen. Given the
complexity of linear mixed effects models used to analyze GE rate, no further adjustments were
performedandp<0.05wasconsideredtobestatisticallysignificant.
Associations between pre-feed residual stomach volumes, FV, immediate post-feed stomach
volumes,feedduration,feedingfrequencyandbothhormoneandmacronutrientconcentrationsand
doses,andinfantanthropometrics/BCparametersweretestedusingrobustlinearregression. Mixed
effectsmodelswereconsidered,butwerenotsignificantlybetter(p>0.1)Robustlinearregression
(rlm) was chosen so as to address heteroscedasticity in the data and points with high leverage in
themajorityofthepredictors; MM-estimation(M-estimationwithTukey’sbiweight,initializedby
aspecificS-estimator)accountingforappropriatecovariateswasused[65]. Approximatep-values
weredeterminedusingtheWaldtest. MultivariatemodelsaccountingforFVwereusedfortestingthe
relationshipwithFV-dependentpredictor(fatdoseandconcentration).
Possible age differences in HM components, infant characteristics, and GE/breastfeeding
parameterswereanalyzedwitheitherlinearmixedeffectsmodelsorrobustlinearregressionmodels;
modeltypewasdeterminedusinglikelihoodratiotests. Linearmixedeffectsmodelswereusedto
analyzerelationshipsofGEduringfeedtimewithHMcomponentsandinfantcharacteristics. R1,FV
andfeeddurationwerenotassociatedwithstomachvolumereductionduringthefeedtime,therefore
univariatemodelswererun. MultivariatelinearmixedeffectsmodelsaccountingforR1,FVandfeed
durationwereusedinanalysisofrelationshipsofimmediatepost-feedstomachvolumeswithHM
componentsandinfantcharacteristics.
Owingtothelargenumberofcomparisons,afalsediscoveryrateadjustment[70]wasperformed
onassociatedsubgroupingsofresultswithoneormorep-values<0.05. p-valueswereconsidered
to be significant at <0.011 for GE time, <0.031 for feeding frequency, <0.038 for R2, and <0.008 for
associationsbetweenHMcomponents’concentrations.
3. Results
3.1. Participants
Characteristics of the 27 participants (2 months (n = 20; longitudinal: 7 females, 7 males;
cross-sectional: 2females,4males);5months(n=21;longitudinal: 7females,7males;cross-sectional:
6females);overalln=41feeds)aredescribedinTable1. Atthestudysession,infantsfedfromone
(n=23)orboth(n=18)breasts.
3.2. InfluenceofInfantAge
Infantanthropometricsand%FMmeasuredwithbioimpedancespectroscopysignificantlydiffered
byinfantage(p<0.001),whilebreastfeedingandGEparametersdidnotchangesignificantly(p>0.067)
(Table1).
Lowerwheyproteinconcentration(5.51±0.96g/L5movs. 6.41±1.39g/L2mo,p=0.034)
andsubsequentlyahighercasein:wheyratio(0.32±0.145movs. 0.22±0.072mo,p=0.035)were
observedat5months. Allothermeasuredappetitehormonesandmacronutrientconcentrationsdid
notdiffersignificantlybyinfantage(p>0.053).
Nutrients2017,9,15 9of21
3.3. AnalyzedHumanMilkComponents
AppetitehormonesandmacronutrientconcentrationsanddosesperfeedarepresentedinTable2.
Higherskimmilkleptinconcentrationswereassociatedwithlowerwholemilkleptinconcentrations
(−0.25[−0.34,−0.16],p<0.001)andhigherproteinconcentrationswereassociatedwithhigherwhey
proteinconcentrations(0.68[0.41,0.95],p<0.001). HigherHMOconcentrationswereassociatedwith
highertotalcarbohydratesconcentrations(p<0.001)andlowerlactoseconcentrations(p<0.001).
Table1.Participantcharacteristicsexpressedasmean±SDandrange.
2moa 5mob Totalc
Characteristics
Mean±SD Range Mean±SD Range Mean±SD Range
Infantcharacteristics
Infantage(weeks) 9±1 6–10 22±1 18–23 16±7 6–23
Infantlength(cm) 57±2 53–61 65±2*** 62–69 61±4 53–69
Infantweight(kg) 5.3±0.8 4.2–6.3 7.2±1.0*** 5.8–9.5 6.3±1.3 4.2–9.5
InfantBMI 15.9±1.3 13.9–18.1 17.6±1.7*** 14.9–20.4 16.7±1.7 13.9–20.4
HC(cm) 39±1 37–42 43±2*** 40–46 41±2 37–46
FatMasswithBIS(%) 21.4±3.6 11.1–27.1 28.9±3.2*** 21.7–35.8 25.3±5.0 11.1–35.8
FatMasswithUS(%) 24.2±3.6 17.5–30.5 26.6±3.6 20.8–35.9 25.5±3.8 17.5–35.9
BF/GEcharacteristics
Feedvolume(mL) 86±34 35–140 85±33 36–180 86±33 35–180
SVafterfeed1(mL) 87±36 32–141 93±41 22–189 90±38 22–189
Feedduration(min) 28±14 11–72 20±8 6–37 24±12 6–72
SVreduction(mL)d 5±21 (−42)–33 4±26 (−57)–56 4±24 (−57)–56
GEtime(min)e 94±29 44–153 88±18 50–140 91±24 44–140
Residual1(mL) 6±12 0–50 11±19 0–62 9±16 0–62
Residual2(mL) 20±20 0–81 15±15 0–55 18±18 0–81
Feedingfrequency(h)f 2.3±0.7 1.0–4.0 2.7±0.8 1.5–4.0 2.5±0.7 1.0–4.0
Dataaremean±SDandranges. a n=20; b n=21. c n=41feeds. d Stomachvolumereductionduring
feedtimeiscalculatedasthedifferencebetweenthesumofresidual1andfeedvolumeandtheimmediate
stomachvolumeafterFeed1. e GEtimeisthetimefromthestartofFeed1tothestartofFeed2(time
betweenfeedsplusfeedduration).fFeedingfrequencyself-reportedbymothersastohowofteninfantfeeds
(e.g.,everythreehours). ***Indicatessignificantdifferences(p<0.001)betweentwo-andfive-month-old
infants.Abbreviations:BF—breastfeeding;BIS—bioimpedancespectroscopy;GE—gastricemptying;HC—head
circumference;SV—stomachvolume;US—ultrasound.
Table2.ConcentrationsanddosesofmeasuredHMhormonesandmacronutrients.
Concentration DosePerFeed
Components
Mean±SD Range Mean±SD Range
Adiponectin(ng/mL,ng) 10.02±4.08 6.18–22.58 868.62±491.32 238.60–2536.91
WMleptin(ng/mL,ng) 0.51±0.18 0.23–1.10 44.80±24.30 10.15–115.03
SMleptin(ng/mL,ng) 0.28±0.12 0.20–0.84 24.8±15.0 6.91–73.00
Totalprotein(g/L,g) 11.29±2.56 7.60–24.16 0.99±0.39 0.35–2.29
Casein(g/L,g) 1.54±0.53 0.69–3.45 0.14±0.07 0.04–0.29
Wheyprotein(g/L,g) 5.97±1.26 3.82–9.08 0.52±0.19 0.17–0.95
Casein:wheyratio 0.27±0.11 0.10–0.73 n/aa n/aa
Lysozyme(g/L,g) 0.14±0.12 0.05–0.48 0.01±0.01 0.003–0.030
TCH(g/L,g) 82.72±7.89 67.08–97.49 7.28±2.62 3.28–15.18
Lactose(g/L,g) 65.84±5.14 53.49–77.94 5.86±2.22 2.19–12.06
HMO(g/L,g) 16.88±9.89 (−10.86)b–35.77 1.42±0.94 (−1.09)b–3.78
Fat(g/L,g) 42.74±12.10 17.42–66.79 3.57±1.45 0.64–6.40
Dataaremean±SDandranges,n=41feeds.aCasein:wheyratiosfordosesarethesameasforconcentrations.
bNegativevaluesareseenforhumanmilkoligosaccharides(HMO)whenlactosemeasurementsarehigher
thantotalcarbohydrates.Abbreviations:SM—skimmilk;TCH—totalcarbohydrates;WM—wholemilk.
3.4. GastricEmptyingRate
The overall decreasing curvilinear pattern of GE (linear: 0.04 [−0.17, 0.24], p = 0.72; square
root: −10.5[−12.7, −8.2], p<0.001)isshowninFigure2. HigherFVswereassociatedwithfaster
Nutrients2017,9,15 10of21
Nutrients 2017, 9, 015  10 of 21 
< 0.001). No association was seen between feed duration and post‐feed stomach volume (−0.25 [−0.68, 
(−0.07[−0N.u1tr0ie,nt−s 2001.70, 39, ]0,15p  <0.001)GErate(Figure3)andhigheroverallpost-feedstom10 oafc 2h1  volumes
0.18], p = 0.23). 
(0.82[0.53,< 10..01021]),. Npo< as0s.o0c0ia1t)io.nN woasa sseseon cbieattwioeenn wfeeads dsuereatniobn eatnwd epeosnt‐ffeeeedd stdomuarcaht ivoonluamned (−p0.2o5s [t−-0fe.6e8d,  stomach
volume(−00.1.28]5, p[ −= 00.2.638). ,0.18],p=0.23).
150
)
L150
m
e (L)
mm
olume ( 100
mach vh volu100
c
oa
Stm 50
o
St 50
0
0
0 20 40 60 80 100 120
0 20 40 60 80 100 120
Time post-feed (min)
Time post-feed (min)  

Figure 2. Overall curvilinear pattern of gastric emptying (n = 41 feeds). The lines represent the overall 
Figure2.pOFaivtgteeurrrnae  l2ol.f c Ocuhvraevrnaiglllei ncsu einrav rsiltiponmaeatatrce phra nvttoeolrufnm goafe sg atasrs imtcriecea emsmupprettyydi inbngyg  (un(l nt=r a4=1so 4fue1endfdes )ie.m dTahsge). ilniTnghe. sBe roellipdnr eelissneenr etr etphprere eossveeennrtatsltl l hoecaol verall
pattern of changes in stomach volume as measured by ultrasound imaging. Bold line represents local 
patternofrecghraenssgioens simnosotothmera (cLhOvEoSSlu, smpaena =s 0m.9)e. aDsoutrteedd libnyesu rletprraessoeuntn cdonimfidaegnicne gin.tBerovladl. linerepresentslocal
regression smoother (LOESS, span = 0.9). Dotted lines represent confidence interval. 
regressionsmoother(LOESS,span=0.9).Dottedlinesrepresentconfidenceinterval.
MI: 35-55 mL MI: 65-85 mL MI: 90-110 mL MI: 115-180 mL
202000 MI: 35-55 mL MI: 65-85 mL MI: 90-110 mL MI: 115-180 mL
151500
L)L)
mm
e (e (
mm
mach volumach volu101000
StoSto 50
50
0
0
0 30 60 90 120 0 30 60 90 120 0 30 60 90 120 0 30 60 90 120
0 30 60 90 120 0 30 60 90 120 0 30 60 90 120 0 30 60 90 120
Time post-feed (min)
Time post-feed (min)  

Figure 3. Gastric emptying of individual feeds in term breastfed infants (n = 41 feeds). Feeds are 
Figure3.FgGigroauusrpetr ei3dc.  bGeyma mstpirlkitcy i neintmagkpeto y(MfiniIg)n  tdoof ii vlilnuiddstiurvaaitdel utfhaeele  efdeffesedcitsn  oifnt te htreem rfmeeb dbr vreeoaalusstmtffeeedsd;  ainipnfpafrnaotnxsi tm(sna (t=ne l4y=1 e q4feu1eadlf nse)ue. mdFbesee).rdss F aereed sare
groupedbgayrroemu piinelckdlu bidnyet dma kiilnek  (einMatcahIk) ept ao(MnielIll).u  tDos atirltlaau tspetortianhttees  tehrefefp eerfcefsteecontf to tfsh ttohemef afeeceehdd v vvooolullumummees sec;s aa;lpcaupplraoptxerdiom xfariotmemlay t ueeqlltuyraaselo qnuuunmadl bnerusm  bers
areincludaeirmde aiignnecsel;ua cdcoehndnp eicantni neegal .clihDn epas atlainnepkl .om Dienaatstsau rrpeemopirenentstss e rfnreotpmsret toshemen sta acsmthoem vfeoaeclduh.  mBvooeldlsu lcminaeelcs r uecplaarletcesuedlnattfser dloo mcfarlo urmeltg rruaelsstsroiaousnon udnidm  ages;
imsmaogoetsh; ecro n(LnOecEtSinS,g s lpianne s=  l0in.9k) .m  easurements from the same feed. Bold line represents local regression 
connectinglineslinkmeasurementsfromthesamefeed.Boldlinerepresentslocalregressionsmoother
smoother (LOESS, span = 0.9). 

(LOESS,span=0.9).

Immediatepost-feedstomachvolumeswerenotassociatedwithR1(p=0.91).
Afteraccountingfortimepost-feed,FV,andfeedduration,aspertheabovemodel,largerR1
volumes(0.55[0.24,0.86],p=0.003)andhigherwheyproteinconcentrations(4.99[0.84,9.13],p=0.023)
were associated with larger post-feed stomach volumes, while the casein:whey ratio (2.2 ± 0.88,
p=0.030)andlactoseconcentration(−0.04±0.02,p=0.037)modifiedtheGEcurvedependingon