do postnatal iron deficiency or maternal risk factors affect...

1
Do postnatal iron deficiency or maternal risk factors affect cogni6ve development in infants born moderate‐to‐late preterm? Infants born moderate‐to‐late preterm (32‐36 weeks gesta6on) are not closely followed a/er hospital discharge and have a greater risk for developing iron deficiency and neurodevelopmental delays. Individual differences in prenatal risk and neonatal risk factors may impact long‐term cogni=ve outcomes in these children. One individual difference that likely impacts cogni=ve development in infants born moderate‐to‐late preterm is postnatal iron levels. Iron deficiency is a common micronutrient deficiency that can cause altered cogni6ve development in children if not treated (ThorisdoBr, GunnarsdoBr, Palsson, Gretarsson, ThorsdoB, 2013). Iron is essen=al for neurodevelopmental processes such as myelina=on, neurotransmiKer produc=on, and development of the hippocampus; thus infants could suffer from memory problems as a result of low iron levels (Georgieff, 2011). Maternal risk factors during pregnancy, such as pre‐eclampsia and HELLP syndrome, both characterized by hypertension, have also been associated with lower performance by infants in various cogni6ve tasks (Kronenberg, Raz, and Sander, 2006). Few studies have been done specifically with moderate‐to‐late preterm infants and risk factors that could affect their cogni=ve development. Based on previous studies, we hypothesized that infants born moderate‐to‐late preterm would perform more poorly on measures of infant memory and general cogni=ve development, and that individual differences in maternal risk factor history (presence of hypertension) and postnatal iron levels (hemoglobin) would also predict infant’s performance. This research was supported by a University of Minnesota Graduate School Fellowship (Hodel), Doctoral Disserta6on Fellowship (Hodel), and Undergraduate Research Opportuni6es Program Grant (Sasson) and by the NIH under a Ruth L. Kircshtein Na6onal Research Service Award (T32HD007151) from the NICHD. Addi6onal support was provided by the University of Minnesota’s Ins6tute of Child Development, Center for Neurobehavioral Development, and Graduate Students of the College of Educa6on and Human Development. The authors thank the families of the par6cipants and members of Kathleen M. Thomas’ Cogni6ve Development & Neuroimaging Lab. Ba]ery of prefrontal‐dependent tasks for 9‐month old infants Iron Levels: PT and FT infants did not differ in HgB levels between 9‐12 months of age. Bayley: PT and FT infants did not differ on any of the Bayley subscales. Habitua6on Task: Of those who successfully habituated, both PT and FT groups showed a novelty preference at test (p < .01), although this effect was reduced in PT children (p < .04) and was related to gesta=onal age at birth across the en=re sample. Full‐Term Children (n=74) Preterm Children (n=70) Age at Test Μ = 9.10 months Μ = 10.21 months (9.00 months adjusted age) Sex 36 female, 38 male 33 female, 37 male Gesta,onal Age Μ = 39.84 weeks range = 37.43 – 42.71 weeks Μ = 35.34 weeks range = 32.00 – 36.86 weeks NICU/SCN stay? 9.5% of sample (< 24 hours) 68.6% of sample 9‐12 Month Hemoglobin M = 11.77 mg/dL range = 10.1 ‐ 13.9 mg/dL M = 11.78 mg/dL range = 8.7 ‐ 13.5 mg/dL Maternal Hypertension 0% 11.4% Infants born moderate‐to‐late preterm show subtle altera6ons in working memory at 9‐months corrected age. Both prenatal (maternal hypertension) and postnatal (low iron) risk factors also influence infants’ early cogni6ve development. ADDITIONAL MEASURES MEDICAL RECORD ABSTRACTION BAYLEY SCALES OF INFANT DEVELOPMENT (Bayley, 2006) Working Memory HABITUATION TASK Cogni-on is assessed by examining how infants react and learn about their surroundings Language is assessed by measuring the recogni=on of sounds and objects and produc=on of speech sounds and gestures Motor skills measured include both fine (e.g. infant’s grasp) and gross skills (e.g. crawling) Parents consented to release their child’s birth hospitaliza=on, well‐baby check records, and any iron‐related labs All records were abstracted for gesta-onal age and pregnancy and birth‐related risk factors; hemoglobin labs at birth and between 9‐12 months of age were collected Infant driven, sliding window design using NimS=m facial s=muli (based on Markant et al., 2014) Novelty detec=on and dishabitua=on are prefrontal dependent in young infants (Nakano et al., 2008) AKen=on geKer s=mulus (variable dura=on) Eye movements tracked and coded online Test phase – six trials of novel and familiar s=mulus with one face per trial Complete infant task ba=ery also includes eye‐ tracking measures of reversal learning, processing speed, and behavioral measures of inhibitory control (A not B task) and problem solving Complete parent ques-onnaire ba=ery also includes parent‐report measures of parental depression and life stress, paren=ng styles, infant temperament, and infant sleep and feeding paKerns ‐6000 ‐4000 ‐2000 0 2000 4000 6000 8000 30 32 34 36 38 40 42 Novelty Preference at Test (Novel ‐ Familiar Look Dura6on in ms) Gesta6onal Age at Birth (in weeks) Gesta6onal Age at Birth Predicts Magnitude of Novelty Preference r = .21, p < .03 0 500 1000 1500 2000 2500 PT FT Novelty Preference at Test (Novel ‐ Familiar Look Dura6on in ms) Preterm Infants Demonstrated a Smaller Novelty Prefence at Test Face s=mulus (maximum 15 seconds) Habitua-on criterion – average look dura=on over three consecu=ve trials <50% of previous trial set (maximum 21 trials) 8 9 10 11 12 13 14 15 6 8 10 12 14 Bayley Recep6ve Language Score 9‐12 Month Hemoglobin Level Higher Iron Levels Predict Be]er Recep6ve Language Abili6es ‐1500 ‐1000 ‐500 0 500 1000 1500 2000 Maternal hypertension No maternal hypertension Novelty Preference at Test (Novel – Familiar Look Dura6on in ms) Maternal Hypertension Predicts Poorer Infant Working Memory Performance ‐6000 ‐4000 ‐2000 0 2000 4000 6000 8000 10000 6 7 8 9 10 11 12 13 14 Novelty Preference at Test (Novel‐Familiar Look Dura6on in ms) 9‐12 Month Hemoglobin Level Iron Levels Were Unrelated to Infant Working Memory 0 2 4 6 8 10 12 14 16 Cogni=ve R.Language E.Language F.Motor G.Motor Bayley Subscale Score Maternal Hypertension Does Not Predict Infant Bayley Scores No hypertension Maternal hypertension Moderate‐to‐late preterm infants exhibited poorer working memory development at 9 months of age. Addi=onally, there was evidence that iron deficiency and maternal risk factors such as hypertension affect infant’s memory and cogni6ve development. Limita6ons: Many of the preterm infants in this study were supplemented with iron which likely contributed to their normal iron levels at 9‐12 months of age. Addi-onally, the lack of rela-onship between iron levels and memory performance may be specific to the task used. The habitua-on task measures working memory, which relies on the prefrontal cortex. Iron deficiency has been most strongly associated with hippocampal development, so other memory tasks could have yielded different results. Last, infants born to mothers with hypertension could have other risk factors driving their poorer performance. Future studies should con=nue to inves=gate the effects of iron deficiency on cogni=ve outcomes in infants, including those born moderate‐to‐late preterm. Moreover, these studies should inves=gate which pregnancy‐related risk factors predict children’s outcomes in order to ensure appropriate interven=ons take place to prevent cogni=ve developmental delays. Bayley: Infants with higher HgB levels had higher recep6ve language scores (r = .22, p < .03). Habitua6on Task: Iron levels were not related to novelty preference scores (r = ‐.04, p < .71). Bayley: Infants did not differ on any of the Bayley scales. There was a trend (p < .12) toward lower cogni=ve scores in the risk group. Habitua6on: The risk group failed to show a novelty preference at test (p < .62) and significantly differed from the non‐risk group in novelty preference magnitude (p < .03).

Upload: others

Post on 28-Jan-2020

4 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Do postnatal iron deficiency or maternal risk factors affect ...innovation.umn.edu/cdnlab/wp-content/uploads/sites/38/2018/08/... · moderate‐to‐late preterm would perform more

Dopostnatalirondeficiencyormaternalriskfactorsaffectcogni6vedevelopmentininfantsbornmoderate‐to‐latepreterm?

Infantsbornmoderate‐to‐latepreterm(32‐36weeksgesta6on)arenotcloselyfolloweda/erhospitaldischargeandhaveagreaterriskfordevelopingirondeficiencyandneurodevelopmentaldelays.Individualdifferencesinprenatalriskandneonatalriskfactorsmayimpactlong‐termcogni=veoutcomesinthesechildren.

Oneindividualdifferencethatlikelyimpactscogni=vedevelopmentininfantsbornmoderate‐to‐latepretermispostnatalironlevels.Irondeficiencyisacommonmicronutrientdeficiencythatcancausealteredcogni6vedevelopmentinchildrenifnottreated(ThorisdoBr,GunnarsdoBr,Palsson,Gretarsson,ThorsdoB,2013).Ironisessen=alforneurodevelopmentalprocessessuchasmyelina=on,neurotransmiKerproduc=on,anddevelopmentofthehippocampus;thusinfantscouldsufferfrommemoryproblemsasaresultoflowironlevels(Georgieff,2011).Maternalriskfactorsduringpregnancy,suchaspre‐eclampsiaandHELLPsyndrome,bothcharacterizedbyhypertension,havealsobeenassociatedwithlowerperformancebyinfantsinvariouscogni6vetasks(Kronenberg,Raz,andSander,2006).

Fewstudieshavebeendonespecificallywithmoderate‐to‐latepreterminfantsandriskfactorsthatcouldaffecttheircogni=vedevelopment.Basedonpreviousstudies,wehypothesizedthatinfantsbornmoderate‐to‐latepretermwouldperformmorepoorlyonmeasuresofinfantmemoryandgeneralcogni=vedevelopment,andthatindividualdifferencesinmaternalriskfactorhistory(presenceofhypertension)andpostnatalironlevels(hemoglobin)wouldalsopredictinfant’sperformance.

ThisresearchwassupportedbyaUniversityofMinnesotaGraduateSchoolFellowship(Hodel),DoctoralDisserta6onFellowship(Hodel),andUndergraduateResearchOpportuni6esProgramGrant(Sasson)andbytheNIHunderaRuthL.KircshteinNa6onalResearchServiceAward(T32HD007151)fromtheNICHD.Addi6onalsupportwasprovidedbytheUniversityofMinnesota’sIns6tuteofChildDevelopment,CenterforNeurobehavioralDevelopment,andGraduateStudentsoftheCollegeofEduca6onandHumanDevelopment.Theauthorsthankthefamiliesofthepar6cipantsandmembersofKathleenM.Thomas’Cogni6veDevelopment&NeuroimagingLab.

Ba]eryofprefrontal‐dependenttasksfor9‐montholdinfants

IronLevels:PTandFTinfantsdidnotdifferinHgBlevelsbetween9‐12monthsofage.

Bayley:PTandFTinfantsdidnotdifferonanyoftheBayleysubscales.

Habitua6onTask:Ofthosewhosuccessfullyhabituated,bothPTandFTgroupsshowedanoveltypreferenceattest(p<.01),althoughthiseffectwasreducedinPTchildren(p<.04)andwasrelatedtogesta=onalageatbirthacrosstheen=resample.

Full‐TermChildren(n=74) PretermChildren(n=70)

AgeatTestΜ=9.10months Μ=10.21months

(9.00monthsadjustedage)

Sex 36female,38male 33female,37male

Gesta,onalAge

Μ=39.84weeksrange=37.43–42.71weeks

Μ=35.34weeksrange=32.00–36.86weeks

NICU/SCNstay?

9.5%ofsample(<24hours) 68.6%ofsample

9‐12MonthHemoglobin

M=11.77mg/dLrange=10.1‐13.9mg/dL

M=11.78mg/dLrange=8.7‐13.5mg/dL

MaternalHypertension

0% 11.4%

Infantsbornmoderate‐to‐latepretermshowsubtlealtera6onsinworkingmemoryat9‐monthscorrectedage.Bothprenatal(maternalhypertension)andpostnatal(lowiron)riskfactorsalsoinfluenceinfants’earlycogni6vedevelopment.

ADDITIONALMEASURESMEDICALRECORDABSTRACTION

BAYLEYSCALESOFINFANTDEVELOPMENT

(Bayley,2006)

WorkingMemoryHABITUATIONTASK

• Cogni-onisassessedbyexamininghowinfantsreactandlearnabouttheirsurroundings• Languageisassessedbymeasuringtherecogni=onofsoundsandobjectsandproduc=onofspeechsoundsandgestures• Motorskillsmeasuredincludebothfine(e.g.infant’sgrasp)andgrossskills(e.g.crawling)

• Parentsconsentedtoreleasetheirchild’sbirthhospitaliza=on,well‐babycheckrecords,andanyiron‐relatedlabs

• Allrecordswereabstractedforgesta-onalageandpregnancyandbirth‐relatedriskfactors;hemoglobinlabsatbirthandbetween9‐12monthsofagewerecollected

• Infantdriven,slidingwindowdesignusingNimS=mfacials=muli(basedonMarkantetal.,2014)

• Noveltydetec=onanddishabitua=onareprefrontaldependentinyounginfants(Nakanoetal.,2008)

AKen=ongeKers=mulus(variabledura=on)

Eyemovementstrackedandcodedonline

Testphase–sixtrialsofnovelandfamiliars=muluswithonefacepertrial

• Completeinfanttaskba=eryalsoincludeseye‐trackingmeasuresofreversallearning,processingspeed,andbehavioralmeasuresofinhibitorycontrol(AnotBtask)andproblemsolving

• Completeparentques-onnaireba=eryalsoincludesparent‐reportmeasuresofparentaldepressionandlifestress,paren=ngstyles,infanttemperament,andinfantsleepandfeedingpaKerns

‐6000

‐4000

‐2000

0

2000

4000

6000

8000

30 32 34 36 38 40 42

NoveltyPreferenceatTest

(Novel‐FamiliarLoo

kDura6

oninm

s)

Gesta6onalAgeatBirth(inweeks)

Gesta6onalAgeatBirthPredictsMagnitudeofNoveltyPreference

r=.21,p<.03

0

500

1000

1500

2000

2500

PT FT

NoveltyPreferenceatTest

(Novel‐FamiliarLoo

kDura6

oninm

s)

PretermInfantsDemonstratedaSmallerNoveltyPrefenceatTest

Faces=mulus(maximum15seconds)

Habitua-oncriterion–averagelookdura=onoverthreeconsecu=vetrials<50%ofprevioustrialset(maximum21trials)

8

9

10

11

12

13

14

15

6 8 10 12 14

BayleyRecep

6veLan

guageScore

9‐12MonthHemoglobinLevel

HigherIronLevelsPredictBe]erRecep6veLanguageAbili6es

‐1500

‐1000

‐500

0

500

1000

1500

2000

Maternalhypertension Nomaternalhypertension

NoveltyPreferenceatTest

(Novel–Fam

iliarLoo

kDura6

oninm

s)

MaternalHypertensionPredictsPoorerInfantWorkingMemoryPerformance

‐6000

‐4000

‐2000

0

2000

4000

6000

8000

10000

6 7 8 9 10 11 12 13 14

NoveltyPreferenceatTest

(Novel‐Fam

iliarLoo

kDura6

oninm

s)

9‐12MonthHemoglobinLevel

IronLevelsWereUnrelatedtoInfantWorkingMemory

0

2

4

6

8

10

12

14

16

Cogni=ve R.Language E.Language F.Motor G.Motor

BayleySub

scaleScore

MaternalHypertensionDoesNotPredictInfantBayleyScores

Nohypertension Maternalhypertension

Moderate‐to‐latepreterminfantsexhibitedpoorerworkingmemorydevelopmentat9monthsofage.Addi=onally,therewasevidencethatirondeficiencyandmaternalriskfactorssuchashypertensionaffectinfant’smemoryandcogni6vedevelopment.

Limita6ons:Manyofthepreterminfantsinthisstudyweresupplementedwithironwhichlikelycontributedtotheirnormalironlevelsat9‐12monthsofage.Addi-onally,thelackofrela-onshipbetweenironlevelsandmemoryperformancemaybespecifictothetaskused.Thehabitua-ontaskmeasuresworkingmemory,whichreliesontheprefrontalcortex.Irondeficiencyhasbeenmoststronglyassociatedwithhippocampaldevelopment,soothermemorytaskscouldhaveyieldeddifferentresults.Last,infantsborntomotherswithhypertensioncouldhaveotherriskfactorsdrivingtheirpoorerperformance.

Futurestudiesshouldcon=nuetoinves=gatetheeffectsofirondeficiencyoncogni=veoutcomesininfants,includingthosebornmoderate‐to‐latepreterm.Moreover,thesestudiesshouldinves=gatewhichpregnancy‐relatedriskfactorspredictchildren’soutcomesinordertoensureappropriateinterven=onstakeplacetopreventcogni=vedevelopmentaldelays.

Bayley:InfantswithhigherHgBlevelshadhigherrecep6velanguagescores(r=.22,p<.03).

Habitua6onTask:Ironlevelswerenotrelatedtonoveltypreferencescores(r=‐.04,p<.71).

Bayley:InfantsdidnotdifferonanyoftheBayleyscales.Therewasatrend(p<.12)towardlowercogni=vescoresintheriskgroup.

Habitua6on:Theriskgroupfailedtoshowanoveltypreferenceattest(p<.62)andsignificantlydifferedfromthenon‐riskgroupinnoveltypreferencemagnitude(p<.03).