Fetal and Neonatal Growth and Development Workshop 

December 15-16, 2002
Baltimore Marriott Waterfront
Baltimore, MD

This meeting was held in conjunction with the National Children’s Study, which is led by a consortium of federal agency partners: the U.S. Department of Health and Human Services (including the National Institute of Child Health and Human Development [NICHD] and the National Institute of Environmental Health Sciences [NIEHS], two parts of the National Institutes of Health, and the Centers for Disease Control and Prevention [CDC]) and the U.S. Environmental Protection Agency (EPA).

Introduction

Moderators for the workshop were Donald Mattison, M.D., NICHD, NIH, DHHS; Catherine Spong, M.D., NICHD, NIH, DHHS; and Adolfo Correa, M.D., Ph.D., M.P.H., CDC, DHHS. Dr. Mattison provided an overview of the National Children’s Study (Study). Dr. Spong explained that the purpose of the workshop was to obtain expert input about measures (what, when, and how) related to fetal and neonatal growth and development and about new and emerging technologies. The workshop was divided into two workgroups:

• Fetal growth and development
• Neonatal and infant growth.

Fetal Growth and Development Workgroup

George Saade, M.D., University of Texas Medical Branch, led the discussion about what is known regarding evaluation of fetal growth and outcomes.

Growth Potential Models

Dr. Saade introduced Radek Bukowski, M.D., M.P.H., who discussed the Gardosi growth potential model, which uses individualized or customized growth charts as an alternative to using population norms for birth weights. The customized chart is generated by using six independent factors that can predict normal growth, including maternal height and weight, parity, ethnicity, and gender and gestational age of the fetus. An optimal birth weight is calculated and divided into 280 increments. This model has been shown to be the most precise method for predicting abnormal outcomes, including stillbirth and neonatal death. The Gardosi model has many advantages; however, it relies on weight as a single parameter and no U.S. norms currently exist. This model has not been used in relation to long-term outcomes, such as development of cardiovascular disease in adulthood. Dr. Bukowski noted that a large trial such as the Study could generate U.S. norms for birth weight.

Dr. Bukowski responded to questions and comments about the Gardosi model. He said that the questions related to three basic issues:

• The model’s factors (parity, ethnicity, and height) for predicting optimal weight
• How prediction of growth potential is tied to long-term outcomes
• Other factors, particularly maternal weight gain.

Dr. Saade then discussed the Rossavik growth model, noting that it was a complex mathematical formula that requires two measurements in the second trimester and assumes normal growth previously. Various measurements can be used (for example, abdominal circumference or femur length), and it is possible to combine all measurements into a prenatal growth assessment score. The fetal growth model has not been validated regarding outcomes; however, the neonatal model has been validated.

Richard Berkowitz, M.D., M.P.H., Mount Sinai School of Medicine, discussed three-dimensional (3-D) ultrasound technology, which collects volume data that are easily stored. This type of imaging does have limitations, and higher resolution and better software performance are needed. Good scanning technique is also very important. Dr. Berkowitz emphasized that the technology allows a wealth of data to be easily obtained, stored, and transmitted. He added that volume measures are more accurate. The technology allows for the storage of raw data, which can be analyzed in different ways in the future. He said that the issue is not whether to use the technology, but how to use it wisely.

Dr. Berkowitz said that the safety of ultrasound has been extensively studied and the use of diagnostic ultrasound at currently used frequencies has never been associated with independently reproducible biological effects. There have been reports of adverse effects, but these have not been confirmed in multiple studies.

Dr. Saade commented that Doppler ultrasound is useful regarding miscarriage, intrauterine fetal death, and preeclampsia. He also noted that measurements on both sides of the placenta (uterine and umbilical arteries) are needed. Dr. Saade asked the group whether or not Doppler should be included in the Study, and the consensus was yes.

Technology

Christian Macedonia, M.D., LTC, MC, National Naval Medical Center, made a presentation and led a discussion about technological methods to image the pregnant uterus in order to store information for later analysis. The discussion focused on availability, feasibility, reproducibility, and cost. Dr. Macedonia’s presentation, "Moving Pictures: Connecting Images to Clinical Data," stressed the need to collect 3-D ultrasound data correctly from the start and discussed the RADIUS study that revealed poor technique to be a major problem.

Dr. Macedonia said that MRI is definitely useful for some subsets, although there is less information on MRI safety for the fetus than for ultrasound safety. Limitations of MRI include cost and size of equipment, inconvenience, effects of motion and positioning, lack of experience by technicians, and time for high-resolution image acquisition. He discussed imaging and data standards, including Digital Image Communication in Medicine (DICOM) standards for image transmission, and Health Level 7 (HL7), a health informatics data interchange standard used in electronic medical records transmission. Dr. Macedonia noted that protecting patient privacy as required by the Health Insurance Portability and Accountability Act of 1996 (HIPAA) is a major issue.

Dr. Macedonia said that obtaining a full 3-D of a fetus depends on gestational age. Up to 17 weeks, one can image the entire fetus in about 10 seconds. Older fetuses need multiple measurements because the transducers aren’t large enough. A lower speed is necessary for high resolutions.

William P. Fifer, Ph.D., Columbia University, discussed heart rate variability as a useful measure of fetal neurodevelopment. Methods to evaluate fetal heart rate are:

• Fetal ECGs
• Fetal magnetoencephalography, which provides high-quality tracing and measures heart activity, with good results from 28 weeks
• Fetal cardiotocography, which measures early markers of neurological integrity.
• Computerized actocardiography, which measures motor activity along with heart rate.

Janet DiPietro, Ph.D., Johns Hopkins University, described her research protocol that includes 50 minutes of fetal monitoring to collect data on movement and heart rate. Dr. DiPietro noted that 28-32 weeks is the "magic" age for the neurological integration in the fetus. Her research is now following up on children at age 2 and 3 to determine how prenatal measures correlate with outcomes. Dr. DiPietro cautioned that it would be necessary to have at least 30 minutes’ duration of fetal monitoring for an adequate result, whether on paper or digitized. To plot a trajectory, it would be necessary to have three data points. It would be best to monitor before and after the 28-32-week transition period.

Summary

Dr. Spong reviewed the goal of the workshop, the various methodologies that were presented, and key points of the discussion during the day. Suggestions from participants included:

• 3-D sonograms during pregnancy in first trimester, at 18-20 weeks, and two sonograms in the third trimester
• Volumetric data from the 3-D sonograms to be stored for future analysis
• Fetal heart rate recordings to be stored for future analysis
• MRI, Doppler, and ECG in a subset of Study participants.

Recommendations for pilot studies included:

• Evaluation of patient acceptance of 3-D sonograms
• Development of a definition of fetal growth restriction (FGR) for core hypotheses.

Neonatal and Infant Growth and Development Workgroup

Jan Friedman, M.D., Ph.D., University of British Columbia, provided an overview, and Kenneth Lyon Jones, M.D., University of California, San Diego provided some historical context about neonatal and infant growth and development.

Growth and Measurement

• Parameter/Outcome.  Participants discussed measures of growth that should be done during the neonatal and infant periods:
  • The data collection system should ideally reflect previously established, national standards of normal growth for neonates and infants.
  • Birth weight and recumbent birth length
  • Head circumference, including the BPD/OCF ratio, to capture important variations in morphology
  • Body composition, including skinfold measures and circumferential measures, at the subscapular/suprailiac, the mid-upper arm/triceps, waist, and hip
  • Body proportion, using circumferential measures for both the thigh and calf, and measurement of the lower leg length
  • Parental height and weight for both parents.
  A number of pilot studies were suggested to evaluate the utility and reliability of various other body measurements. These measurements include the palpebral fissure length/orbital size; inner/outer (which has to be calculated) canthal ratio; philtrum length; ear length; length of the foot; and hand length (using the third digit as a reference point).
• Examiners.  Two people will be needed when using the measuring board. With the proper training, one of these can be a parent, especially if measurements are done during a home visit.
• Methods.  The Study will need to use standard measures of birth length. Spreading calipers are needed for head and other circumferential measures. Other methods and instruments discussed included a knemometer to measure lower leg length, three-dimensional (3-D) and ultrasound, MRI for organ volume, and laser holography for obtaining 3-D tomographic medical data to measure body morphology.
• Time Period.  Because hospital/birth location and setting are so variable, it was decided that the first encounter could be 7-10 days of life (DOL), and could successfully capture all or nearly all of the measures needed. A second encounter would occur at 3 to 4 months DOL.
• Settings.  Depending on what equipment is required and how portable it is, these measurements could be done during home visits. This could be facilitated by the use of mobile vans equipped with ultrasound equipment.
• Challenges.  Increasingly, mothers and neonates are leaving the hospital or birthing center as soon as possible. This practice creates challenges for Study measurements. Also, there are dramatic variations from one setting to another in terms of procedures and instrumentation. An additional challenge is the lack of a valid set of national data for neonatal growth parameters. Workshop participants identified some other practical challenges:
  • Circumstances can make measuring difficult, such as preterm birth or congenital malformations.
  • Some common measurements are unreliable. For example, the waist/hip ratio is more reliable than the chest circumference because of respiration.
  • Some equipment is expensive, for example, a knemometer would be useful, but costs about $5,000.
  • Some equipment, such as some types of calipers, is difficult to use on infants.
• Pilot Studies.  A number of pilot studies will be required to determine the reliability of measures and to discover additional measures. Some of those identified were the possible use of ultrasound for body composition; how circumferential measurements can adduce a measure of fat mass; what are the various "gold standards" that should be incorporated or at least used for reference; how various approaches compare in capturing certain parameters, such as MRI, ultrasound, and skinfold measurements; how organ size might be captured; and which subsamples might be useful and appropriate to target.

Neurodevelopment

• Parameter/Outcome. 
  • Preterm birth is a fundamental category that will drive which measures are studied.
  • Measures should include the labor and delivery experience, neonatal complications (infection, neonatal seizures, chronic lung disease), and mortality.
  • Neurological measures needed are muscle tone, postural control, visual fixation, and reflexes (pathological, primitive).
  • Later infant/child development assessment should include motor, cognitive, and deviant and/or delayed language development, and social/emotional development.
  • Hearing screenings are needed, using either the otoacoustic emissions (OAE) or the auditory brain stem response (ABR) criteria.
  • Visual testing usually is done after the neonatal period.
  Prenatal risk factors include socioeconomic status (SES), parental education, multiple gestations, and physical events (for example, neonatal infections, such as meningitis, sepsis; CLD; bilateral IP cysts; obstetrical complications; and neonatal PPHN and ECMO). One way of approaching such a large number of potential risk factors and confounders is to identify specific subgroups for analysis/study.
• Examiners.  Depending on which measures are decided upon, trained non-physicians could do some neonatal exams.
• Methods.  Participants suggested the following methods:
  • The Amiel-Tison examination protocol-a test of infant motor performance administered beginning at 4 months of age
  • Barry Lester’s standardized neonatal developmental study (NICHD, 2002)
  • Videotapes should be used for data capture.
  • The ITSEA/BITSEA parent report
  • Sleep states can be probed with "bed/mattress" studies to collect data on REM and non-REM/sleep.
  • Evaluate habituation, visual attention, consolability, and responsiveness.
• Time Period.  The first exam for full-term birth infants should be done sometime between 1-7 DOL, as well as for premature infants whose PCA is greater than 38 weeks.
• Challenges.  There must be a clear understanding of the difference between outcome measures, predictors, and confounders. The broad rubric of risk factors are those measurable phenomena associated with an increased likelihood for disability, though are not necessarily causal. Multiple risk factors can be additive. Autism is an organic social/emotional/cognitive disorder of the brain, around which there is more controversy than data. This disorder could be classified with other disorders of higher cortical function, but such children will require different subsequent measures.
• Pilot Studies.  A number of the tests and targets mentioned earlier need to be validated in pilot studies. For the neonate, there is a need to distinguish between neurodevelopmental and behavioral outcomes and measures. Once this is done, pilot studies will be needed to determine the best instruments. As many of these functions develop in stages, pilot studies need to pinpoint when is the best time to administer tests such as:
  • Visual assessment, teller cards, and visual preference
  • Developmental milestones of speech/language
  • Fine and gross motor skills
  • Habituation, adaptation, language, and other social/emotional parameters.

Structural Birth Defects (Cardiac Problems)

• Parameter/Outcome.  Because they are so often "silent" and yet correlate with many issues of long-term health and survival, cardiac defects provide an important opportunity for Study. It is necessary to have a physical exam to test for heart rate, respiratory rate, and blood pressure in right arm and both legs. There also should be an evaluation for mild, moderate, and severe congenital heart disease (CHD); murmurs; peripheral pulses; organomegaly; and evidence of syndromes. If feasible, tests should also be conducted for vascular reactivity, for evidence of rhythms and disturbances, and myopathies.
• Examiners.  Technicians who routinely administer echocardiograms and ultrasound will be required.
• Methods.  All medical records should be available (including autopsy in the event of death). A physical exam will be supplemented by the necessary technology-an echocardiogram (at a minimum), a pulse oximeter, a test for vascular reactivity, an ascending and descending aorta Doppler (which predicts hypertension later), and an image of the heart. There should also be a functional MRI of at least the kidneys.
• Time Period.  The first test should be done soon after birth (1-2 days DOL). Sometime between 4 and 6 months, a comprehensive evaluation is needed. An echocardiogram and a functional image are needed at 6 months.
• Challenges.  A number of issues arise in trying to balance the potential payoff with the difficulty of obtaining certain information.
  • Would it be an imbalanced use of available resources to try to identify mild defects?
  • Is it worth the payoff to identify bicuspid aortic valve or coarctation of the aorta?
  • How can the Study take advantage of newer technology and yet keep the measurements consistent throughout the sample, at many different sites with different in-house resources and equipment?
  • How are pregnancy terminations to be identified?
• Pilot Studies.  Pilot studies are needed to determine:
  • The best vascular reactivity measurements
  • Three-dimensional imaging, including the new "body stocking" with multiple electrodes to detect irregular rhythms.

Other Structural Malformations

• Parameter/Outcome.
  • A standard physical examination to pick up major external birth defects
  • Medical record review and various diagnostic studies
  • Photographs to evaluate external anomalies
  • Imaging studies to diagnose internal malformations
  • Blood tests to establish a reliable biologic specimen database for biomarkers, including DNA and cord blood
  • Newer imaging technology, such as holography, could provide an image from which various measurements could be derived in a standard fashion.
• Examiners.  A pediatrician is required for most of these tests. Depending on the test, the identified subgroup and the circumstances, other trained examiners will become involved, such as a clinical dysmorphologist. It is important to ensure quality control of the data.
• Methods.  The Study should standardize the use of photography and videography. Medical records must be obtained and reviewed. Autopsy for all fetal deaths, stillbirths, and infant deaths is imperative.
• Time Period.  Twelve months is an ideal time for most of these evaluations. For many major structural anomalies, surgical alteration may be an issue. Minor anomalies may not be ascertained well in the young infant.
• Settings.  The primary care physician’s office or a mobile unit affiliated with tertiary care centers could provide most of the tests.
• Challenges.  General health care must be integrated with this aspect of the Study. Challenges include inter-examiner reliability, examiner expertise, timing issues, resource availability, and cost. Whether or not to include examinations to detect minor anomalies, both internal and external, needs further discussion.
• Pilot Studies.  A number of issues of reconciling methods, measures, and approaches need to be clarified under the heading of inter-rater reliability. New methods, such as holography, that might replace conventional methods should be evaluated.

Facilitators and Presenters

Marilee C. Allen, M.D., Department of Neonatology, Johns Hopkins University Medical Institutions
Richard L. Berkowitz, M.D., M.P.H., Department of Obstetrics, Gynecology, and Reproductive Science, Mount Sinai School of Medicine
Radek Bukowski, M.D., Ph.D., Obstetrics and Gynecology, University of Texas Medical Branch
William P. Fifer, Ph.D., Pediatrics and Perinatology, Columbia University
Jan M. Friedman, M.D., Ph.D., Department of Medical Genetics, University of British Columbia
Maureen Hack, MBchB, Department of Pediatrics, Case Western Reserve University
Charlotte A. Hobbs, M.D., Ph.D., Arkansas Center for Birth Defects, University of Arkansas
Kenneth Lyon Jones, M.D., Department of Pediatrics, University of California, San Diego
Michelle Lampl, Ph.D., Department of Anthropology, Emory University
Christian Macedonia, M.D., LTC, MC, Women and Children’s Health Services, National Naval Medical Center
Cynthia A. Moore, M.D., Ph.D., National Center on Birth Defects and Developmental Disabilities, CDC, DHHS
George R. Saade, M.D., Department of Obstetrics and Gynecology, University of Texas Medical Branch

Other Participants

Antonio Barraza
Charles R. Bauer, M.D.
Richard L. Berkowitz, M.D., M.P.H.
Ira M. Bernstein, M.D.
Robert Bradley, Ph.D.
Amy Branum, M.S.P.H.
Molly S. Bray, Ph.D.
Kent R. Carlson, Ph.D.
Laura E. Caulfield, Ph.D.
Robert E. Chapin, Ph.D.
Robert L. Chevalier, M.D.
Edward B. Clark, M.D.
Adolfo Correa, M.D., Ph.D., M.P.H.
Janet DiPietro, Ph.D.
William P. Fifer, Ph.D.
Matthew W. Gillman, M.D., S.M.
Lynn Goldman, M.D.
Gilman Grave, M.D.
Doris B. Haire
Gary D. Hankins, M.D.
Phil Heard
Mary L. Hediger, Ph.D.
Deborah G. Hirtz, M.D.
Henry S. Kahn, M.D.
Mark A. Klebanoff, M.D., M.P.H.
Robert J. Kuczmarski, Dr.P.H., R.D.
William K. Kyle, M.S.E.E., M.B.A.
Monica Longo, M.D.
Barbara Luke, M.P.H., Sc.D.
Christian Macedonia, M.D., LTC, MC
Donald Mattison, M.D.
Francis John Meaney, Ph.D.
Bryan T. Oshiro, M.D.
Kenneth C. Schoendorf, M.D., M.P.H.
Kathi E. Shea
Eleanor Smith-Khuri, M.D.
Barbara C. Sonies, Ph.D.
Catherine Y. Spong, M.D.
Shumei Sun, Ph.D.
Kent L. Thornburg, Ph.D
Kevin C. Vigilante, M.D., M.P.H.
Robert M. Ward, M.D.
Desmond Williams, M.D., Ph.D.
Marian Willinger, Ph.D.
Marshalyn Yeargin-Allsopp, M.D.
Barry Zuckerman, M.D.