pediatrics 2008 karemaker 978 87

DOI: 10.1542/peds.2007-3409 2008;122;978Pediatrics

and Cobi J. HeijnenTersteeg-Kamperman, Wim Baerts, Sylvia Veen, Jannie F. Samsom, Frank van Bel

Rosa Karemaker, John M. Karemaker, Annemieke Kavelaars, MarijkeResponse of Children at School Age

Effects of Neonatal Dexamethasone Treatment on the Cardiovascular Stress

http://pediatrics.aappublications.org/content/122/5/978.full.html

located on the World Wide Web at: The online version of this article, along with updated information and services, is

of Pediatrics. All rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.Boulevard, Elk Grove Village, Illinois, 60007. Copyright © 2008 by the American Academy published, and trademarked by the American Academy of Pediatrics, 141 Northwest Pointpublication, it has been published continuously since 1948. PEDIATRICS is owned, PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

at Indonesia:AAP Sponsored on August 22, 2013pediatrics.aappublications.orgDownloaded from


http://pediatrics.aappublications.org/

ARTICLE

Effects of Neonatal Dexamethasone Treatment onthe Cardiovascular Stress Response of Children atSchool AgeRosa Karemaker, MDa,b, John M. Karemaker, PhDc, Annemieke Kavelaars, PhDa, Marijke Tersteeg-Kamperman, PhDa, Wim Baerts, MDd,

Sylvia Veen, MDe, Jannie F. Samsom, MDf†, Frank van Bel, MDb, Cobi J. Heijnen, PhDa

aLaboratory of Psychoneuroimmunology and bDepartment of Neonatology, University Medical Center Utrecht, Utrecht, Netherlands; cDepartment of Physiology,Academic Medical Centre, Amsterdam, Netherlands; dDepartment of Neonatology, Isala Clinics Zwolle, Zwolle, Netherlands; eDepartment of Neonatology, UniversityMedical Center Leiden, Leiden, Netherlands; fDepartment of Neonatology, Free University Medical Center, Amsterdam, Netherlands

The authors have indicated they have no financial relationships relevant to this article to disclose.

What’s Known on This Subject

There is growing concernwith respect to possible adverse long-term effects of neonataldexamethasone treatment of chronic lung disease of prematurity. Data from animalstudies have shown that neonatal dexamethasone treatment can have long-lastingadverse effects on the cardiovascular system.

What This Study Adds

We report for the first time that human neonatal dexamethasone treatment hasconsequences for the cardiovascular response to a stressful challenge at school age.In contrast, we did not observe any effect of neonatal hydrocortisone treatment.

ABSTRACT

OBJECTIVE. The goal was to investigate cardiovascular responses to a psychosocial stres-sor in school-aged, formerly premature boys and girls who had been treated neona-tally with dexamethasone or hydrocortisone because of chronic lung disease.

METHODS.We compared corticosteroid-treated, formerly preterm infants with formerlypreterm infants who had not been treated neonatally with corticosteroids (referencegroup). Children performed the Trier Social Stress Test for Children, which includesa public speaking task and a mental arithmetic task. Blood pressure was recordedcontinuously before, during, and after the stress test. Plasma norepinephrine levelswere determined before the test, directly after the stress task, and after recovery.

RESULTS.Overall, in response to stress, girls had significantly larger changes in systolicblood pressure and mean arterial pressure and in stroke volume and cardiac output,compared with boys. Boys exhibited larger total peripheral resistance responses,compared with girls. The hydrocortisone group did not differ significantly from thereference group in any of the outcome measures. However, dexamethasone-treatedchildren had smaller stress-induced increases in systolic and mean arterial bloodpressure than did hydrocortisone-treated children. In addition, the dexamethasonegroup showed smaller increases in stroke volume and blunted norepinephrine re-sponses to stress, compared with children in the reference group. Correction forgender did not affect these results.

CONCLUSIONS. The differences in cardiovascular stress responses between girls and boysare consistent with known gender differences in adult cardiovascular stress responses. Our data demonstrate thatneonatal treatment with dexamethasone has long-term consequences for the cardiovascular and noradrenergic stressresponses; at school age, the cardiovascular stress response was blunted in dexamethasone-treated children. Hydro-cortisone-treated children did not differ from the reference group, which suggests that hydrocortisone might be a safealternative to dexamethasone for treating chronic lung disease of prematurity. Pediatrics 2008;122:978–987

IN PAST DECADES, the use of neonatal corticosteroid therapy to treat or to prevent chronic lung disease (CLD) ofprematurity was common practice. A few human follow-up studies have since been published, showing adverse

long-term effects on behavior, neurologic and motor development, and hypothalamic-pituitary-adrenal responsive-ness and the immune response.1–7 The corticosteroid most widely used for treatment of CLD is dexamethasone.2,8

However, since the first study was published in the 1970s by Baden et al,9 our hospital has instead used hydrocor-tisone.6 Although corticosteroid therapy for newborns is now actively discouraged, neonatologists continue to use itif other options fail.1,10,11 Consequently, there still are and will be large numbers of children affected by the possiblelong-term effects of neonatal corticosteroid treatment.

www.pediatrics.org/cgi/doi/10.1542/peds.2007-3409

doi:10.1542/peds.2007-3409

†Deceased.

KeyWordscardiovascular, corticosteroids, follow-upstudies, neonatology, retrospective clinicalstudy

AbbreviationsCLD—chronic lung diseaseTSST-C—Trier Social Stress Test forChildren

Accepted for publication Feb 22, 2008

Address correspondence to Cobi J. Heijnen,PhD, Laboratory of Psychoneuroimmunology,University Medical Center Utrecht, OfficeKC03.068.0, Lundlaan 6, 3584 EA Utrecht,Netherlands. E-mail: [email protected]

PEDIATRICS (ISSN Numbers: Print, 0031-4005;Online, 1098-4275). Copyright © 2008 by theAmerican Academy of Pediatrics

978 KAREMAKER et al at Indonesia:AAP Sponsored on August 22, 2013pediatrics.aappublications.orgDownloaded from


The full scope of the lifelong effects of neonatal cor-ticosteroid treatment has mainly been evaluated in rats.The developmental stage of a newborn rat is comparableto that of the human fetus during the third trimester ofpregnancy.12,13 Recent studies showed that rats treatedwith dexamethasone on days 1, 2, and 3 of life had ashorter life span and died as a result of cardiomyopathyand renal failure.14 Moreover, neonatal dexamethasonetreatment of rats inhibits cardiomyocyte proliferationand is associated with increased cardiomyocyte volumeat adult age.15,16 Functionally, dexamethasone-treatedrats exhibit cardiac systolic dysfunction during their life-time.17

Acute and transient cardiovascular adverse effects ofneonatal dexamethasone treatment have been describedfor human patients, although conflicting data exist.17–19

However, little is known about possible long-term effectsof neonatal dexamethasone treatment on the humancardiovascular system. It is well known from humanresearch that an altered cardiovascular response to stressis an early predictor of an increased risk for cardiovas-cular disease in later life.20,21 Measurement of blood pres-sure and heart rate before, during, and after stress is avaluable method to investigate the cardiovascular stressresponse. Various noninvasive methods of blood pres-sure measurements have been applied, with the Fi-napres and Portapres systems (Finapres, Portapres, andModelflow are registered trademarks, presently mar-keted by Finapres Medical Systems, Amsterdam, Neth-erlands), based on the volume-clamp method describedby Penaz22 and developed by the Dutch TNO BiomedicalInstruments Group,23 being common, well-validated,equipment choices.24,25

The Trier Social Stress Test for Children (TSST-C)26 isa widely used inducer of psychosocial stress, consistingmainly of public speaking and mental arithmetic in frontof an audience. We reported recently on the activity ofthe immune system and the hormonal (corticotrophinand cortisol) responses to the TSST-C in the same co-hort.27

Previous animal studies showed pulmonary and sys-temic hypertension in adult rats treated neonatally withdexamethasone.14,27,28 Therefore, we hypothesized thatneonatal exposure of prematurely born children to cor-ticosteroids might be associated with increased bloodpressure at rest and an altered cardiovascular response tothe TSST-C, as determined at school age. To investigatethe long-term effects of neonatal corticosteroid treat-ment on cardiovascular responses to a psychosocialstressor, we studied school-aged, formerly prematureboys and girls who had been treated with dexametha-sone or hydrocortisone because of CLD.

METHODS

PatientsIn a retrospective, matched-cohort study, we comparedthe long-term effects of neonatal treatment with 2 dif-ferent corticosteroids (dexamethasone and hydrocorti-sone) and included a reference group consisting of for-merly premature infants who had not been treated with

steroids neonatally. As reported previously,27 the studypopulation consisted of prematurely born infants whowere admitted between December 1993 and July 1997to the NICUs of the University Medical Centre Utrecht,the Leiden University Medical Centre, the Free Univer-sity Medical Centre Amsterdam, and the Isala ClinicsZwolle in the Netherlands. The study was approved bythe medical ethics committee of the University MedicalCentre Utrecht and the scientific boards of the partici-pating hospitals. Written parental consent was alwaysobtained. The NICU of the University Medical CentreUtrecht used exclusively hydrocortisone therapy to re-duce CLD, in a treatment regimen starting with 5 mg/kgper day and tapering to 1 mg/kg per day over a 22-dayperiod, whereas the other hospitals used dexamethasonefor this purpose, starting with 0.5 mg/kg per day andtapering to 0.1 mg/kg per day over a 21-day period. In allcenters, glucocorticoid treatment was sometimes ex-tended or shortened, depending on the responses to andacute adverse effects of therapy. In all instances, treat-ment was used as a rescue modality, that is, when it wasimpossible to wean the infant from the ventilator, withprolonged dependence on supplementary continuousoxygen (fraction of inspired oxygen of �0.30), in theinitial phase of CLD. The average age at the start ofglucocorticoid treatment is presented in Table 1.

The decision to treat was always left at the discretionof the attending neonatologist. In addition to the 2groups treated with glucocorticoids neonatally, agroup of prematurely born infants not treated withglucocorticoids in the neonatal period was includedfor comparison.

Study GroupsEligibility for inclusion in one of the study groups were(1) surviving the neonatal period; (2) availability to par-ticipate in the study protocol, as indicated below; (3)neonatal cerebral ultrasound scans showing maximallygrade II periventricular hemorrhage, classified as de-scribed by Papile et al29; and (4) absence of major con-genital anomalies. Infants with periventricular leu-komalacia also were excluded. The hydrocortisoneand dexamethasone groups were constituted as fol-lows. The charts of all consecutively admitted, preterminfants born at �32 weeks of pregnancy in the partici-pating hospitals were reviewed systematically. The NICUof the University Medical Centre Utrecht used only hy-drocortisone, and 131 eligible infants were treated withhydrocortisone during the defined time period. Thedexamethasone group, which was recruited in a similarway from the 3 hospitals that used only dexamethasonefor reduction of CLD, consisted of 198 eligible infants.Children from the dexamethasone and hydrocortisonegroups were matched with respect to gestational age, birthweight, gender, year of birth, severity of infant respiratorydistress syndrome (determined according to clinicalsymptoms and the classification system described byGiedeon et al,30 that is, no, moderate, or severe respira-tory distress syndrome), and whether minor periven-tricular/intraventricular hemorrhage (grade I or II) waspresent. During the initial phase of recruitment, a total

PEDIATRICS Volume 122, Number 5, November 2008 979 at Indonesia:AAP Sponsored on August 22, 2013pediatrics.aappublications.orgDownloaded from


of 37 children eligible for the reference group, 29 chil-dren for the hydrocortisone group, and 28 children forthe dexamethasone group declined to participate. Ulti-mately, 52 individuals from the hydrocortisone groupcould be matched with 52 from the dexamethasonegroup.

Fifty-two participants in the reference group werealso recruited from the 4 participating hospitals (50%from the University Medical Centre Utrecht and theothers equally distributed over the other hospitals). Thereference group consisted of prematurely born infantswho had not been treated with glucocorticoids neona-tally and did not have periventricular leukomalacia, ma-jor periventricular/intraventricular hemorrhage (gradeIII or more), or any other major complications duringthe neonatal period. Although we tried to match thislatter group, with respect to gender, birth weight, andgestational age, with a dexamethasone/hydrocortisonecouple, this was not always possible (Table 1).

If a child did not understand the TSST-C and subse-quently did not show any signs of stress, defined as anincrease in either heart rate and/or blood pressure, theresults were excluded from the analysis (n � 3). Twogirls refused to participate. Technical problems for 17children resulted in 139 children having their bloodpressure recorded with a Portapres system throughoutthe experiment.

Preparation and Timing of the TSST-CParents were asked to ensure that their children did notconsume foods or medication that might interfere with

blood pressure regulation or norepinephrine reactivity.Twenty-four hours before the visit to the hospital, adietary list was provided for this purpose. On the testday, children first underwent a physical examination,including manual blood pressure measurements on botharms when seated (Table 2). Parents were asked to limitthe amount of fluids their children consumed at lunch-time to a maximum of 250 mL. Preparation for the stresstest started at 1:15 PM, at which time an antecubitalintravenous line for blood sampling was inserted (ie, 45minutes before rest period 1 started). The actual TSST-Cpublic speaking task was performed between 2:45PM

and 3:15 PM, under the direction of a researcher and aresearch nurse working together with a strict timeschedule. All children were tested by these same 2investigators.

Portapres ApplicationIn accordance with the original TNO Portapres proto-col,23 the finger cuff was applied to the middle phalanx ofthe third finger in a standard manner. To correct theblood pressure level for gravitational effects of handmovements, the Portapres height correction device wastaped to the child’s chest at heart level.31 The child wasseated dressed in a comfortable chair or wheelchair,with pillows to secure a stable upright sitting positionthroughout the experiment. The child’s biometric data(ie, length, weight, and age) were entered into the con-trol unit of the Portapres device and a test recording wasmade, with observation of the real-time blood pressuresignal on a laptop computer, to ensure stable recording.

TABLE 1 Perinatal and Neonatal Characteristics

Reference Group(N � 46)

Hydrocortisone Group(N � 48)

Dexamethasone Group(N � 45)

Gestational age, mean � SD, wk 28.6� 1.0 28.1� 1.5 27.8� 1.9a

Birth weight, mean � SD, g 1095� 193 1044� 208 972� 237Head circumference, mean � SD, cm 26.6� 2.4 26.3� 2.8 25.1� 2.0Gender, n (%)

Female 28 (61) 20 (42) 14 (31)Male 18 (39) 28 (58) 31 (69)a

Mode of delivery, n (%)Vaginal 18 (39) 17 (35) 20 (45)Cesarean section 28 (61) 31 (65) 24 (55)

Apgar score, median (range)At 1 min 6 (0–9) 6 (1–9) 5 (1–9)At 5 min 8 (1–10) 8 (4–10) 7 (3–10)

Prenatal steroid treatment, n (%) 27 (59) 38 (79) 32 (71)Infant respiratory distress syndrome, n (%)None 23 (50) 8 (17)b 8 (18)b

Grade I or II 15 (33) 13 (27) 15 (33)Grade III or IV 8 (17) 27 (56)b 22 (49)b

Postnatal age at start of steroid treatment,mean � SD, d

NA 14� 6 17� 6

Assisted ventilation duration, mean � SD, d 7.8� 8.1 16.5� 9.4b 25.7� 12.3b

Periventricular/intraventricular hemorrhage,n (%)

No 38 (83) 37 (77) 36 (80)Grade 1/2 8 (17) 11 (23) 9 (20)

NA indicates not applicable.a P � .05 versus reference group.b P � .01 versus reference group.



After the initial physiologic calibration32 made by Por-tapres, the blood pressure signal became constant andreliable. Portapres was switched off and the cuff was leftin place around the child’s finger. Throughout the ex-periment, if the child indicated that he or she felt chilly,then blankets and a hand-warming muff were provided,to increase comfort and the reliability of Portapres re-cordings.24,33 The physiologic calibration feature (the in-ternal self-control mechanism of Portapres, in whichpart of the pressure-volume measurement of the vascu-lar bed of the finger is repeated to correct for a possibleslow drift in blood pressure measurements, which re-sults in loss of information for 2 or 3 heart beats per70–90 beats) of Portapres was not switched off duringthe experiment.32

TSST-C ProcedureThis test was described in full by Buske-Kirschbaum etal.26 In brief, the test consists of a 30-minute relaxationperiod in front of a video with neutral contents, a 10-minute preparation period, a 5-minute public speakingtask, and a 5-minute mental arithmetic task (numbersubtraction). During the speaking task, children aregiven the beginning of a story by the investigator and areprompted to complete the story as excitingly as possiblein front of a “committee” judging the child’s perfor-mance. The difficulty level of the subsequent arithmetictask was adjusted to the mental development level of thetested child. After this 10-minute stress period, there wasa 10-minute debriefing period, during which the chil-dren were praised for their excellent performance, fol-lowed by another 45-minute video with neutral con-tents. The 2 videos shown were the same for allparticipants. During the second video, Portapres wasswitched on for the last 20 minutes. After this lastrecording, the finger cuff and height corrector were

removed, and the child was reunited with his or herparents.

Measurement of Norepinephrine Levels in PlasmaEDTA-treated blood for determination of plasma norepi-nephrine levels was collected on ice after rest period 1,immediately after the mental arithmetic task ended, andafter rest period 2 (Fig 1). Plasma was frozen rapidly at�80°C, and norepinephrine levels were determinedwith an enzyme-linked immunosorbent assay (DLD Di-agnostika, Hamburg, Germany).

Data AnalysisPortapres data were analyzed by using Modelflow soft-ware (FMS, Amsterdam, Netherlands),34 which renderedmeasurements of systolic, diastolic, and mean arterialpressure and heart rate, with subsequent pulse contouranalysis-based estimates of stroke volume, cardiac out-put, and total peripheral resistance.23 For every child,10 periods of 30-second duration25,35 were selected,whereby physiologic calibrations were excluded.32 The10 selected periods were representative of the differentphases of the TSST-C, that is, (1) rest period 1 withvideo, (2) beginning of instruction on the TSST-C, (3)end of story preparation phase, (4) start of telling thestory, (5) end of telling the story, (6) start of arithmetictask, (7) end of arithmetic task, (8) start of debriefingphase, (9) end of debriefing phase, and (10) rest period2 (Fig 1).

Because it is known that there is large interindividualvariability in blood pressure results measured with thePortapres system,25 the blood pressure values and Mod-elflow output data for each child were transformed intodata relative to the first rest period for each child bysetting the value measured in rest period 1 at 1.0. Theoutcome measures for the other 9 periods were recalcu-

TABLE 2 Patient and Parent Characteristics at Follow-up Age

Reference Group(N � 46)

Hydrocortisone Group(N � 48)

Dexamethasone Group(N � 45)

PatientAge, mean � SD, y 8.7� 0.6 8.8� 0.6 8.23� 0.8a

Weight, mean � SD, kg 29.3� 5.8 28.6� 6.9 26.2� 5.5Height, mean � SD, cm 132� 6 131� 7 129� 7Head circumference, mean � SD, cm 52.5� 3.1 53.0� 2.1 52.3� 2.5Blood pressure, mean � SD, mmHgSystolic 104� 13 103� 13 99� 12Diastolic 65� 11 66� 10 64� 11

Heart rate, mean � SD, beats per min 83.7� 10 83.0� 10 81.2� 11School performance, n (%)Mainstream education 41 (89) 43 (90) 32 (71)Special education 5 (11) 5 (10) 13 (29)a

ParentEducation of father, n (%)Academic 6 (13) 4 (8) 4 (9)College 17 (37) 21 (44) 23 (51)High school 18 (39) 22 (46) 17 (38)Elementary school only 5 (11) 1 (2) 1 (2)

Divorced, n (%) 7 (16) 8 (15) 4 (9)a P � .05 versus reference group.



lated as fractions relative to rest period 1. In this way,each child served as his or her own control, and datawere comparable between subjects. Only these rela-tive data were used to compare the outcomes of treat-ment groups. Because the experiments were con-ducted over a period of 2 years, we compared earlyand later measurements to validate internally ourmeasurements and their stability. We found no differ-ence between the early and later measurements (datanot shown), and we concluded that our measure-ments were comparable over the time it took to com-plete the inclusions for this study.

Portapres and norepinephrine data were analyzed byusing SPPS 13.0.1 (SPSS, Chicago, IL), with repeated-measurement analysis. Because all data were normalizedto the value at rest period 1 and all children returnedapproximately to this value in rest period 2 (data notshown), the relaxation periods before and after theTSST-C were not included in the final analysis, and therepeated measurements were analyzed over the 8 peri-ods that represented the actual TSST-C.

Outliers were excluded before SPSS analysis, andposthoc analysis was performed by using Bonferroni’scorrection, to compensate for multiple comparisons.Gender differences were analyzed by using the samemethods, and a group-gender interaction was investi-gated for all cardiovascular indices. This resulted, in ef-fect, in a 3 � 2 study design. All P values for treatmentgroup differences mentioned in the text were correctedfor gender, unless otherwise specified. A P value of �.05was considered statistically significant. Because it wasknown that factors such as BMI, social status, birthweight, and age at the time of testing were possibleconfounders of blood pressure, we tested whether therewere any significant covariates present to be included inour analysis (see Tables 1 and 2 for all covariates inves-

tigated). No significant covariates were identified in ourdatabase, however, and these variables were not in-cluded in data analyses.

RESULTS

PatientsAs described above, 139 of the 156 children included inthis study performed the TSST-C and had their bloodpressure recorded. Basic characteristics at birth and atfollow-up assessment are provided in Tables 1 and 2. Indata analysis, it seemed that, for 29 children, the record-ing was not stable during the entire time of the TSST-C,defined as a recorded period of �30 seconds betweenphysiologic calibrations.24 We could reliably analyze datafor 33 children in the reference group, 41 in the hydro-cortisone group, and 36 in the dexamethasone group.The characteristics of these 110 children did not differfrom those of the total group of 139 children who per-formed the TSST-C (Tables 1 and 2).

To exclude a genetic component in blood pressureresponses, we asked the parents of participating childrenabout the prevalence of cardiovascular disease, hyper-cholesterolemia, and diabetes mellitus in their first-de-gree relatives. There were no differences between treat-ment groups (data not shown). In the morning beforethe stress test, there were no group differences in base-line blood pressure (measured with classic manometry)or seated heart rate (Table 2).

Blood Pressure Response to the StressorIn response to the TSST-C, systolic, diastolic, and meanarterial pressure increased and reached maximal valuesat time point 4, during the speech task. At time point 9,after the 10-minute debriefing period, all blood pressurevalues had returned to the initial levels of the prepara-

FIGURE 1Schedule for the TSST-C, providing an overview of sam-pling periods for the different blood pressure measure-ments and norepinephrine (NE) sampling.



tion period (Fig 2, restricted to mean pressures for clar-ity). In view of the strong parallel between mean anddiastolic pressure outcome values, we limit the descrip-tion to systolic and mean pressure values only.

When we analyzed group differences in mean arterialand systolic blood pressure responses to the stressor, weobserved a lower response in dexamethasone-treatedversus hydrocortisone-treated children (mean arterial,P � .04; systolic, P � .04), whereas the response of thehydrocortisone group was similar to that of the referencegroup (P � 1.0), after correction for gender. The differ-ence between the dexamethasone group and the refer-ence group did not reach statistical significance (Fig 2).

Girls overall showed a higher systolic stress responsethan did boys (P � .026). Although the treatment group-gender interaction did not reach statistical significance(P � .089), we cannot exclude the possibility that peri-natal treatment contributed to this finding.

Cardiac Response to the StressorSimilar to the observed blood pressure responses, themaximal increase in heart rate during the TSST-C wasobserved at time point 4. Immediately after terminationof the stress task at time point 8, heart rate returned toinitial levels in all groups (Fig 3).

As depicted in Fig 3, the heart rate response to theTSST-C did not differ significantly between groups (P �.39). In addition, girls did not differ from boys in heartrate response (P � .133; data not shown).

The stress-induced change in stroke volume wassmaller in the dexamethasone group than in the refer-ence group (P � .012) (Fig 4) and tended to be lower inthe dexamethasone group than in the hydrocortisonegroup (P � .075) after correction for gender. Girls hadlarger stroke volume changes than boys (P � .001), andthere was no group-gender interaction for this parame-ter (P � .127). However, combined with slightly higherheart rate responses, these larger changes in stroke vol-ume resulted in greater stress-induced alterations in car-diac output for girls than for boys (P � .001) (Fig 5). Inaddition, there was a significant group-gender interac-

tion (P � .012). Nevertheless, we observed an overalldifference in cardiac output (P � .001, after correctionfor gender), with dexamethasone-treated children hav-ing significantly lower cardiac output responsivenessthan children in the other 2 groups (dexamethasoneversus reference, P � .002; dexamethasone versus hy-drocortisone, P � .001; hydrocortisone versus reference,P � 1.0, after correction for gender). Boys showed asignificantly larger total peripheral resistance increasethan did girls (P � .001) (Fig 6), but there was no effectof glucocorticoid treatment on the stress-induced changein total peripheral resistance (P � .699) and no group-gender interaction (P � .646) for this parameter (Fig 6).

Norepinephrine Response to the StressorRepeated-measures analysis of norepinephrine data re-vealed a significant effect of time (P � .0001) and asignificant group effect (P � .040). The norepinephrineresponses of children in the dexamethasone group weresignificantly lower than those of children in the refer-

FIGURE 2Relative mean arterial pressure (MAP) during the TSST-C. Error bars represent SEM. Timepoints on the x-axis correspond to the timepoints shown in Fig 1. REF indicates reference;HC, hydrocortisone; DEX, dexamethasone (dexamethasone, n� 36; hydrocortisone, n�41; reference, n � 33; dexamethasone versus hydrocortisone, P � .04; dexamethasoneversus reference, not significant; hydrocortisone versus reference, not significant).

FIGURE 3Relative heart rate response during the TSST-C. Error bars represent SEM. Time points onthe x-axis correspond to the time points shown in Fig 1. REF indicates reference; HC,hydrocortisone; DEX, dexamethasone (dexamethasone, n� 36; hydrocortisone, n� 41;reference, n � 33; no significant group effects).

FIGURE 4Relative stroke volume (SV) changes during the TSST-C. Error bars represent SEM. Timepoints on the x-axis correspond to the timepoints shown in Fig 1. REF indicates reference;HC, hydrocortisone; DEX, dexamethasone (dexamethasone, n� 36; hydrocortisone, n�41; reference, n� 33; dexamethasone versus reference, P� .012; dexamethasone versushydrocortisone, P � .075; reference versus hydrocortisone, not significant after correc-tion for gender).



ence group (P � .041), after correction for gender. Thedifference between the hydrocortisone and dexametha-sone groups did not reach statistical significance (P �.166). The hydrocortisone group did not differ from thereference group (Fig 7). There was no gender differencein norepinephrine response (P � .372) or group-genderinteraction (P � .22).

DISCUSSIONWe show here that neonatal dexamethasone treatmentfor CLD of prematurity has long-lasting consequencesfor the cardiovascular and norepinephrine responses topsychosocial stress at school age. The psychosocial stress-induced changes in systolic blood pressure, mean arterialpressure, and cardiac output in dexamethasone-treatedchildren were significantly smaller than those in hydro-cortisone-treated children. Moreover, the stress-inducedchanges in cardiac output and stroke volume were lowerin the dexamethasone-treated group than in the refer-ence group. Hydrocortisone-treated children did not dif-fer from the reference group in their cardiovascular re-sponses to psychosocial stress, in any of the parameterstested.

There are only a few publications on human cardio-vascular follow-up evaluations after neonatal dexameth-asone treatment. In 8 children treated neonatally withdexamethasone, no abnormalities were seen with echo-

cardiography at 8 years of age, and no signs of hyper-tension were present.36 In another study, no differencesin resting blood pressure for 68 dexamethasone-treatedchildren versus 74 nontreated children at 13 to 17 yearsof age were observed.37 In both groups, however, theblood pressure was �95th percentile for height and age,which was attributed to premature birth, irrespective ofneonatal steroid treatment.37 We also did not observegroup differences in blood pressure at rest, but bloodpressures were all within the reference range for height

FIGURE 6Relative changes in total peripheral resistance (TPR) during the TSST-C, treatment groupdifferences, and gender differences (A, girls, n� 62; B, boys, n� 77). Error bars representSEM. Time points on the x-axis correspond to the time points shown in Fig 1. REF indi-cates reference; HC, hydrocortisone; DEX, dexamethasone (gender, P� .001; group, P�.699; group-gender interaction, P � .646).

FIGURE 7Plasma levels of norepinephrine before, immediately after, and 60 minutes after theTSST-C. Error bars represent SEM. Time points on the x-axis correspond to the time pointsshown in Fig 1. REF indicates reference; HC, hydrocortisone; DEX, dexamethasone. *Sig-nificant difference between dexamethasone and reference groups, P � .05 (dexameth-asone, n� 36; hydrocortisone, n� 41; reference, n� 33; dexamethasone versus hydro-cortisone, P � .166; dexamethasone versus reference, P � .041; reference versushydrocortisone, not significant after correction for gender.

FIGURE 5Relative changes in cardiac output (CO) during the TSST-C, treatment group differences,and gender differences (A, girls, n� 62; B, boys, n� 77). Error bars represent SEM. Timepoints on the x-axis correspond to the timepoints shown in Fig 1. REF indicates reference;HC, hydrocortisone; DEX, dexamethasone (gender, P � .001; group, P � .001; group-gender interaction, P � .012; dexamethasone versus reference, P � .002; dexametha-sone versus hydrocortisone, P � .001; hydrocortisone versus reference, P � 1.0, aftercorrection for gender).



and age in our group of prematurely born children (Ta-ble 2). Although baseline heart rate and blood pressurewere normal, the stress-induced changes in blood pres-sure were affected by neonatal dexamethasone treat-ment, which suggests that the capacity of the cardiovas-cular system to adapt to changes in the environment isblunted in dexamethasone-treated children.

In previous animal studies from this department,long-lasting morphologic changes of the heart in dexa-methasone-treated animals, such as hypertrophy of car-diomyocytes and thickening of the left ventricular wall,were detected.38 Furthermore, neonatal dexamethasonetreatment resulted in decreased systolic function andreduced heart weight at 4 weeks of age.39 This decreasein systolic function mimics the reduced systolic andmean arterial pressure responses we observed in dexa-methasone-treated children. However, echocardiogra-phy performed at rest did not show a reduction in esti-mated heart weight or a difference in wall thicknessbetween dexamethasone-treated, hydrocortisone-treated,and reference children.15 Given the many similarities weobserved with neonatal dexamethasone treatment of ratsand humans, we cannot exclude the possibility that mor-phologic changes may develop later in life among thesechildren.

Investigation of gender-specific differences in the re-sponse to stress showed that girls reacted to the stressorwith larger changes in systolic blood pressure, strokevolume, and cardiac output, compared with boys. Thelatter parameter was the only one that showed a group-gender interaction, but the treatment effect was main-tained after correction for gender. Overall, boys showedhigher total peripheral resistance responses, indepen-dent of perinatal treatment.

It is known that adult men and women differ in stressresponses; women are considered to be predominantlycardiac responders, and men are considered to respondmainly through adaptation of their vascular resistance.40

Not much is known about gender differences at prepu-bertal ages, although hormones (mainly the differencein estrogen levels) are thought to be the main explana-tion for the differences seen in adults, because the gen-der differences in stress responses disappear after femalemenopause. The gender differences in stress responsesobserved in our study seemed to follow the adult stressresponse pattern,41 although the oldest child tested wasno more than 10 years of age. However, the gender-specific responses to stress do not explain the significantdifferences between dexamethasone-treated children onthe one hand and hydrocortisone-treated and referencechildren on the other; dexamethasone-treated childrenshowed an overall blunted cardiovascular stress re-sponse. Therefore, our data do not allow us to pinpointwhich gender is more at risk for developing cardiovas-cular dysfunction as a result of neonatal dexamethasonetreatment.

Physiologically, the normal short-term response to apsychological stressor starts in the hypothalamic para-ventricular nucleus, resulting in activation of the sym-pathetic nervous system. Epinephrine and norepineph-rine are released from the adrenal medulla and

postganglionic sympathetic nerve endings, respectively.Consequently, heart rate and cardiac contractility in-crease, with a decreased ejection time. In addition, givensufficient venous return, the inotropic �1-adrenergic ef-fect of epinephrine plus norepinephrine increases strokevolume and, in conjunction with the increased heartrate, increases cardiac output. The �1-adrenergic effect ofcatecholamines causes peripheral vasoconstriction, in-creasing blood pressure and total peripheral resistance,42

opposing the nitric oxide-dependent muscle vasodila-tion.43,44 The increase in mean arterial blood pressure inresponse to the stressor was significantly lower in dex-amethasone-treated children, compared with hydrocor-tisone-treated children. In addition, the combined in-crease in stroke volume and heart rate, and thuscardiac output, in reaction to the psychosocial stressorin the dexamethasone group was lower than that inthe hydrocortisone group. The hydrocortisone groupdid not differ from the reference group. The mostobvious explanation for our observations is that alower sympathetic stress response results in lowerplasma norepinephrine levels and consequently ablunted cardiovascular stress response. However, pre-vious research showed that interindividual variability invascular adrenergic responsiveness contributes to thebalance of factors that maintain normal blood pressurein individuals with differing levels of sympathetic neuralactivity.45 Whether a clear group difference in sympa-thetic drive or altered adrenergic sensitivity and/or nitricoxide-driven muscle vasodilation is responsible for thereduced cardiovascular response to stress in the dexam-ethasone-treated children remains to be elucidated.

An alternative explanation for the blunted cardiovas-cular stress response observed in dexamethasone-treatedchildren would be that these children perceived thestressor in a different way. However, as we reportedearlier, we also analyzed the response of the hypotha-lamic-pituitary-adrenal axis (corticotropin and cortisol)as part of the same study. Although the overall activityof this axis was lower in the dexamethasone group, wedid not observe group differences in stress-induced in-creases in corticotropin and cortisol levels.27 Therefore,we do not expect that differences in appraisal of thestressor underlie the observed blunted cardiovascularresponse to the stressor.

It remains an intriguing phenomenon that neonataltreatment with hydrocortisone does not show the long-term adverse effects that we observed with dexametha-sone treatment. Hydrocortisone and dexamethasone dif-fer in the dose used and in several genomic andnongenomic effects and vary in the binding specificityfor mineralocorticoid and glucocorticoid receptors. vander Heide-Jalving et al6 established that hydrocortisoneis as clinically effective as dexamethasone in mitigatingshort-term negative outcomes, such as ventilator depen-dence or the occurrence of CLD. This study providesoriginal long-term data obtained in a carefully plannedstudy. A weakness may be that it was not a randomized,controlled trial. Because there has been no randomized,long-term, follow-up trial of dexamethasone versus hy-drocortisone, we strongly encourage, on the basis of the



collective data now available, additional study of hydro-cortisone as a safer alternative to dexamethasone for thetreatment of CLD of prematurity.6,46 The similarities be-tween the long-term consequences of neonatal dexa-methasone treatment for rats and humans are a reasonfor serious concern, especially because we now knowthat, at adult age, dexamethasone-treated rats also dis-play dysregulation of the immune system and the neu-roendocrine system, severe nephropathy, and a signifi-cantly reduced life span.14,47–49

It will be of great importance to reexamine this cohortof formerly preterm infants, to investigate whether theobserved differences are sustained and/or develop intorisk factors for cardiovascular disease during the postpu-bertal period. Moreover, reinvestigation of this cohortmight allow an early intervention to minimize or toprevent possible cardiovascular risks of early neonataldexamethasone treatment.

ACKNOWLEDGMENTSThis study was supported by a grant from the CatharijneFoundation and the Dirkzwager-Assink Fund.

We thank Jitske Zijlstra and Miriam Maas for per-forming all of the laboratory analyses.

REFERENCES1. Shinwell ES, Karplus M, Bader D, et al. Neonatologists are

using much less dexamethasone. Arch Dis Child Fetal NeonatalEd. 2003;88(5):F432–F433

2. Barrington KJ. The adverse neuro-developmental effects ofpostnatal steroids in the preterm infant: a systematic review ofRCTs. BMC Pediatr. 2001;1(1):1

3. Halliday HL, Ehrenkranz RA, Doyle LW. Moderately early(7–14 days) postnatal corticosteroids for preventing chroniclung disease in preterm infants. Cochrane Database Syst Rev.2003;(1):CD001144

4. O’Shea TM, Kothadia JM, Klinepeter KL, et al. Randomizedplacebo-controlled trial of a 42-day tapering course of dexa-methasone to reduce the duration of ventilator dependency invery low birth weight infants: outcome of study participants at1-year adjusted age. Pediatrics. 1999;104(1):15–21

5. Yeh TF, Lin YJ, Lin HC, et al. Outcomes at school age afterpostnatal dexamethasone therapy for lung disease of prematu-rity. N Engl J Med. 2004;350(13):1304–1313

6. van der Heide-Jalving M, Kamphuis PJ, van der Laan MJ, et al.Short- and long-term effects of neonatal glucocorticoidtherapy: is hydrocortisone an alternative to dexamethasone?Acta Paediatr. 2003;92(7):827–835

7. Karemaker R, Heijnen CJ, Veen S, et al. Differences in behav-ioral outcome and motor development at school age afterneonatal treatment for chronic lung disease with dexametha-sone versus hydrocortisone. Pediatr Res. 2006;60(6):745–750

8. Thebaud B, Lacaze-Masmonteil T, Watterberg K. Postnatal glu-cocorticoids in very preterm infants: “the good, the bad, andthe ugly”? Pediatrics. 2001;107(2):413–415

9. Baden M, Bauer CR, Colle E, Klein G, Taeusch HW Jr, Stern L.A controlled trial of hydrocortisone therapy in infants withrespiratory distress syndrome. Pediatrics. 1972;50(4):526–534

10. American Academy of Pediatrics, Committee on Fetus andNewborn, Canadian Paediatric Society, Fetus and NewbornCommittee. Postnatal corticosteroids to treat or preventchronic lung disease in preterm infants. Pediatrics. 2002;109(2):330–338

11. Kaempf JW, Campbell B, Sklar RS, et al. Implementing poten-

tially better practices to improve neonatal outcomes after re-ducing postnatal dexamethasone use in infants born between501 and 1250 grams. Pediatrics. 2003;111(4). Available at:www.pediatrics.org/cgi/content/full/111/4/e534

12. Dobbing J, Sands J. Comparative aspects of the brain growthspurt. Early Hum Dev. 1979;3(1):79–83

13. Bohn M. Glucocorticoid induced teratologies of the nervoussystem. In: Yanai J, ed. Neurobehavioral Teratology. New York,NY: Elsevier; 1984:365–387

14. Kamphuis PJ, De Vries WB, Bakker JM, et al. Reduced lifeexpectancy in rats after neonatal dexamethasone treatment.Pediatr Res. 2007;61(1):72–76

15. de Vries WB, Karemaker R, Mooy NF, et al. Cardiovascularfollow-up at school age after perinatal glucocorticoid exposurein prematurely born children: perinatal glucocorticoid therapyand cardiovascular follow-up. Arch Pediatr Adolesc Med. 2008;162(8):738–744

16. Bal MP, de Vries WB, van Oosterhout MF, et al. Long-termcardiovascular effects of neonatal dexamethasone treatment:hemodynamic follow-up by left ventricular pressure-volumeloops in rats. J Appl Physiol. 2008;104(2):446–450

17. Werner JC, Sicard RE, Hansen TW, Solomon E, Cowett RM,Oh W. Hypertrophic cardiomyopathy associated with dexa-methasone therapy for bronchopulmonary dysplasia. J Pediatr.1992;120(2):286–291

18. Doyle LW, Davis PG, Morley CJ, McPhee A, Carlin JB. Low-dose dexamethasone facilitates extubation among chronicallyventilator-dependent infants: a multicenter, international, ran-domized, controlled trial. Pediatrics. 2006;117(1):75–83

19. Skelton R, Gill AB, Parsons JM. Cardiac effects of short coursedexamethasone in preterm infants. Arch Dis Child Fetal NeonatalEd. 1998;78(2):F133–F137

20. Lovallo WR. Cardiovascular reactivity: mechanisms and path-ways to cardiovascular disease. Int J Psychophysiol. 2005;58(2–3):119–132

21. Treiber FA, Kamarck T, Schneiderman N, Sheffield D, KapukuG, Taylor T. Cardiovascular reactivity and development of pre-clinical and clinical disease states. Psychosom Med. 2003;65(1):46–62

22. Penaz J. Photoelectric measurement of blood pressure, volumeand flow in the finger. In: Digest of the 10th International Confer-ence on Medical and Biological Engineering. Dresden, Germany:IFBME; 1973:104

23. Wesseling KH, Settels JJ, de Wit B. The measurement of con-tinuous finger arterial pressure noninvasively in stationarysubjects. In: Schmidt TH, Dembroski TM, Bluemchen G, eds.Biological and Psychological Factors in Cardiovascular Disease. Ber-lin, Germany: Springer-Verlag; 1986:355–375

24. Imholz BP, Wieling W, van Montfrans GA, Wesseling KH.Fifteen years experience with finger arterial pressuremonitoring: assessment of the technology. Cardiovasc Res. 1998;38(3):605–616

25. Tanaka H, Thulesius O, Yamaguchi H, Mino M, Konishi K.Continuous non-invasive finger blood pressure monitoring inchildren. Acta Paediatr. 1994;83(6):646–652

26. Buske-Kirschbaum A, Jobst S, Psych D, et al. Attenuated freecortisol response to psychosocial stress in children with atopicdermatitis. Psychosom Med. 1997;59(4):419–426

27. Karemaker R, Kavelaars A, ter Wolbeek M, et al. Neonataldexamethasone treatment for chronic lung disease of prema-turity alters the hypothalamus-pituitary-adrenal axis and im-mune system activity at school age. Pediatrics. 2008;121(4).Available at: www.pediatrics.org/cgi/content/full/121/4/e870

28. le Cras TD, Markham NE, Morris KG, Ahrens CR, McMurtry IF,Abman SH. Neonatal dexamethasone treatment increases therisk for pulmonary hypertension in adult rats. Am J PhysiolLung Cell Mol Physiol. 2000;278(4):L822–L829



29. Papile LA, Burstein J, Burstein R, Koffler H. Incidence andevolution of subependymal and intraventricular hemorrhage:a study of infants with birth weights less than 1 500 gm.J Pediatr. 1978;92(4):529–534

30. Giedion A, Haefliger H, Dangel P. Acute pulmonary X-raychanges in hyaline membrane disease treated with artificialventilation and positive end-expiratory pressure (PEP). PediatrRadiol. 1973;1(3):145–152

31. Wesseling KH. Finger arterial pressure measurement with Fi-napres. Z Kardiol. 1996;85(suppl 3):38–44

32. Wesseling KH, de Wit B, van der Hoeven GMA, van GoudoeverJ, Settels JJ. Physiocal, calibrating finger vascular physiologyfor Finapres. Homeostasis. 1995;36:67–82

33. Tanaka H, Thulesius O. Effect of temperature on finger arterypressure evaluated by volume clamp technique. Clin Physiol.1993;13(5):535–545

34. Wesseling KH, Jansen JR, Settels JJ, Schreuder JJ. Computa-tion of aortic flow from pressure in humans using a nonlinear,three-element model. J Appl Physiol. 1993;74(5):2566–2573

35. Imholz BP, Settels JJ, van der Meiracker AH, Wesseling KH,Wieling W. Non-invasive continuous finger blood pressuremeasurement during orthostatic stress compared to intra-arterial pressure. Cardiovasc Res. 1990;24(3):214–221

36. Mieskonen S, Eronen M, Malmberg LP, Turpeinen M, KariMA, Hallman M. Controlled trial of dexamethasone in neona-tal chronic lung disease: an 8-year follow-up of cardiopulmo-nary function and growth. Acta Paediatr. 2003;92(8):896–904

37. Jones RA. Randomized, controlled trial of dexamethasone inneonatal chronic lung disease: 13- to 17-year follow-up study,part II: respiratory status, growth, and blood pressure. Pediat-rics. 2005;116(2):379–384

38. de Vries WB, Bal MP, Homoet-van der Kraak P, et al. Suppres-sion of physiological cardiomyocyte proliferation in the rat pupafter neonatal glucocorticosteroid treatment. Basic Res Cardiol.2006;101(1):36–42

39. Bal MP, de Vries WB, van der Leij FR, et al. Neonatal glucocor-ticosteroid treatment causes systolic dysfunction and compen-

satory dilation in early life: studies in 4-week-old prepubertalrats. Pediatr Res. 2005;58(1):46–52

40. Allen MT, Stoney CM, Owens JF, Matthews KA. Hemody-namic adjustments to laboratory stress: the influence of genderand personality. Psychosom Med. 1993;55(6):505–517

41. Evans JM, Ziegler MG, Patwardhan AR, et al. Gender differ-ences in autonomic cardiovascular regulation: spectral, hor-monal, and hemodynamic indexes. J Appl Physiol. 2001;91(6):2611–2618

42. Marieb EN. Unit III: regulation and integration of the body. In:Fogel L, Brassert C, Earl W, eds. Human Anatomy and Physiology.4th ed. Menlo Park, CA: Benjamin/Cummings; 1998:505–511,608–609, 612–613, 679

43. Cardillo C, Kilcoyne CM, Cannon RO III, Panza JA. Impair-ment of the nitric oxide-mediated vasodilator response to men-tal stress in hypertensive but not in hypercholesterolemic pa-tients. J Am Coll Cardiol. 1998;32(5):1207–1213

44. Joyner MJ, Dietz NM. Sympathetic vasodilation in humanmuscle. Acta Physiol Scand. 2003;177(3):329–336

45. Charkoudian N, Joyner MJ, Sokolnicki LA, et al. Vascularadrenergic responsiveness is inversely related to tonic activityof sympathetic vasoconstrictor nerves in humans. J Physiol.2006;572(3):821–827

46. Rademaker KJ, Rijpert M, Uiterwaal CS, et al. Neonatal hydro-cortisone treatment related to 1H-MRS of the hippocampusand short-term memory at school age in preterm born chil-dren. Pediatr Res. 2006;59(2):309–313

47. de Vries WB, van der Leij FR, Bakker JM, et al. Alterations inadult rat heart after neonatal dexamethasone therapy. PediatrRes. 2002;52(6):900–906

48. Bakker JM, Kavelaars A, Kamphuis PJ, et al. Neonatal dexa-methasone treatment increases susceptibility to experimentalautoimmune disease in adult rats. J Immunol. 2000;165(10):5932–5937

49. Raff H. Neonatal dexamethasone therapy: short- and long-term consequences. Trends Endocrinol Metab. 2004;15(8):351–352



DOI: 10.1542/peds.2007-3409 2008;122;978Pediatrics

and Cobi J. HeijnenTersteeg-Kamperman, Wim Baerts, Sylvia Veen, Jannie F. Samsom, Frank van Bel

Rosa Karemaker, John M. Karemaker, Annemieke Kavelaars, MarijkeResponse of Children at School Age

Effects of Neonatal Dexamethasone Treatment on the Cardiovascular Stress

ServicesUpdated Information &

mlhttp://pediatrics.aappublications.org/content/122/5/978.full.htincluding high resolution figures, can be found at:

References

ml#ref-list-1http://pediatrics.aappublications.org/content/122/5/978.full.htat:This article cites 41 articles, 19 of which can be accessed free

Citations

ml#related-urlshttp://pediatrics.aappublications.org/content/122/5/978.full.htThis article has been cited by 2 HighWire-hosted articles:

Subspecialty Collections

ular_disorders_subhttp://pediatrics.aappublications.org/cgi/collection/cardiovascCardiovascular Disorders

y_subhttp://pediatrics.aappublications.org/cgi/collection/neonatologNeonatology

orn_infant_subhttp://pediatrics.aappublications.org/cgi/collection/fetus:newbFetus/Newborn Infantthe following collection(s):This article, along with others on similar topics, appears in

Permissions & Licensing

mlhttp://pediatrics.aappublications.org/site/misc/Permissions.xhttables) or in its entirety can be found online at: Information about reproducing this article in parts (figures,

Reprints http://pediatrics.aappublications.org/site/misc/reprints.xhtml

Information about ordering reprints can be found online:

rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.Grove Village, Illinois, 60007. Copyright © 2008 by the American Academy of Pediatrics. All and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elkpublication, it has been published continuously since 1948. PEDIATRICS is owned, published, PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

at Indonesia:AAP Sponsored on August 22, 2013pediatrics.aappublications.orgDownloaded from


http://pediatrics.aappublications.org/content/122/5/978.full.html#ref-list-1

http://pediatrics.aappublications.org/content/122/5/978.full.html#related-urls

http://pediatrics.aappublications.org/cgi/collection/fetus:newborn_infant_sub

http://pediatrics.aappublications.org/cgi/collection/neonatology_sub

http://pediatrics.aappublications.org/cgi/collection/cardiovascular_disorders_sub

http://pediatrics.aappublications.org/site/misc/Permissions.xhtml

http://pediatrics.aappublications.org/site/misc/reprints.xhtml


pediatrics 2008 karemaker 978 87

Documents