Fetal and Neonatal Hand Movement.

Commentary on human fetal movement prior to the advent of imaging appears to have been based historically on myth and conjecture.[1] The application of ultrasound imaging to the human fetus in the 1970s permitted direct visualization of fetal movements in utero and stimulated the initiation of naturalistic studies of the fetus.[2] Observation of fetal movement has the potential for providing insights into the development of coordinated movement, which may be associated with neural development.[3-5] In addition, delineation of early human motor capabilities may corroborate the results of animal research,[6] marshaling interest in the study of early integration of human systems that promote more adaptive functioning.[7] Variables related to environmental effects[8-10] may be clarified and factors related to joint development[11] may be elucidated, possibly providing information for clinical application.[4]

The purpose of this article is to report on an exploration of human fetal and neonatal movement. Our goal was to determine (1) the duration and frequency of specific motor behaviors, (2) changes or trends in the expression of these behaviors during pregnancy, and (3) the manner in which these motor behaviors exist neonatally, as the environment and the child change.

Background

Fetal Movement

Qualitative and quantitative approaches have been used in the study of fetal movement by de Vries and colleagues in the Netherlands. Using ultrasound imaging, de Vries et al[12] focused on the first half of pregnancy. Roodenburg and colleagues[13] investigated the second half of pregnancy. Based on their observation of apparently typical spontaneous human movements, de Vries et al[14] recorded variations in "general movements" that they believed suggested abnormalities of the central nervous system. Following their lead, other researchers have selected, as a dependent variable, general movements or "gross movements involving the whole body. They may last from a few seconds to a minute.... They wax and wane in intensity, force, and speed, and their onset and end are gradual.... The movement is fluent and elegant and creates the impression of complexity and variability."[15(pp152-153)] The lack of definition specificity and long-term follow-up and of detailed methods and the reliability of descriptions have limited the usefulness of some of these studies. In only 2 studies--one on the development of head position[16] and one on handedness[17]--was fetal behavior described. Recently, the descriptor of "general movements" has been applied to the movement of preterm newborns.[18] The description of newborn movement is becoming more detailed with research.[19]

Sparling and Wilhelm[20] described spontaneous movements in fetuses from 12 to 35 weeks of gestation and recorded the characteristics of hand movement. Many movements appeared to be directed to a body part or the uterine wall. The hands of the fetuses moved with a variety of frequencies and apparent force. Joint ranges of motion changed throughout movements rather than remaining the same, as in floating. These movements suggested primary and secondary circular reactions[21] in which a movement is repeated, presumably because it has functional importance to the organism. Sparling and Wilhelm observed, for example, that early in fetal development, quick, progressively larger head flexion movements were repeated, resulting in a "somersault" that enabled the fetus to change position within the uterine cavity. In contrast, during later gestational periods, the fetuses' hands were directed to and manipulated body parts and features of the environment, such as the umbilical cord. Thus, later in pregnancy, the hands exhibited manipulative capability and suggested "intentionality," a term coined by Butterworth and Hopkins[22] to describe neonatal hand-to-mouth movement.

Other developmental tendencies in hand movement were noted in early observations[20] and are summarized in Table 1. In that study, movements such as thumb in mouth and bilateral leg extension against the uterine wall were considered by the authors as functionally important. The frequently observed leg extension against the uterine wall was believed by the authors to be a possible precursor to later participation in the birthing process. Validation of the importance of this movement has been noted in a similar movement of the chick readying itself for hatching.[23] Early movements of the arms appear to assist the fetus in identifying components of its environment. The hands can be observed to cross midline, with the palms "feeling" the uterine wall. The fetus' palms also mold to the occiput, grasp the umbilical cord, and appear to "reach" for the feet. Attributing function to any of these early movements, however, does not imply that the assigned function is preliminary to or necessary for the appearance of a spontaneous behavior.

Table 1. Developmental Motor Characteristics of Low-Risk Fetuses[20]

Gestational
Age (wk)      Description

8             Trunk flexion and extension
12            Isolated random-appearing movement of extremities
14            All movement patterns present; an increased
                frequency of movement that is more "organized"
                in appearance compared with movement at 12
                weeks; arms appear to "explore" while legs
                extend against uterine wall; arm crosses midline,
                extending palmar surface to opposite uterine wall
16            Decreased frequency of movement from 14 weeks,
                with pincer grasp, thumb in mouth
20            More bilateral movement (eg, legs extend together
                against uterine wall, arms flex, and hands are
                often held together near the face)
26-32         Independent movement of extremities to all parts of
                the uterus and specific body parts; no
                cephalocaudal development, but apparent
                distal-proximal development in extremities
37-38         Decreased frequency of movement; hands often
                molded to occiput, or dorsum of hand rests
                against uterine wall

Fetal movement has a wide range of expression. Although some low-risk fetuses appear to have a unique style of movement that is consistent over the gestational period, other low-risk fetuses have fairly wide fluctuations in duration and frequency of movement.[24] Studying the vagaries of movement of low-risk fetuses may provide a clue to understanding the motor behavior of fetuses with impairment.[25] Such studies are needed before movements can be seen as deviations from the norm and can be used for diagnostic purposes.[26]

Based on the literature and previous studies, we hypothesized that there would be extensive, apparently nonfunctional hand movements early in the fetal period. After the diminution of quantity of movement at 16 weeks of gestation noted previously,[27] which we posit was caused by preprogrammed neuronal cell death,[28-30] we would expect to see an increase in more specific movements such as hand to mouth, a movement considered to be functional for the fetus as well as the newborn.[22] The increase in these functional movements would suggest some development in motor behavior and an increased level of motor control by the fetus. The decrease in these behaviors might indicate a developmental regression in the behavior.

Method

The Committee on the Protection of the Rights of Human Subjects at the School of Medicine of the University of North Carolina at Chapel Hill requested a review of the safety of diagnostic versus therapeutic ultrasound prior to the approval of this study. Our review and earlier 3-year follow-up study[31] indicated no harmful effects of diagnostic ultrasound when used according to the American Institute of Ultrasound in Medicine regulations.[32] The committee approved the study, and all 25 low-risk pregnant women approached agreed to participate. One woman withdrew from the study at 26 weeks because of difficulty with her pregnancy and a preterm birth. Another woman was withdrawn from the study because of her difficulty in keeping clinic appointments. Two other women were ultimately eliminated from this study because of premature delivery. The fetuses of the remaining 21 women were the subjects of this study. The fetuses were considered to be medically within normal limits, as was the pregnancy, the neonates at birth, and the children at 12 months of age. For their participation, the mothers were given a copy of the ultrasound images and infant videotapes.

Subjects

The women who participated in this study were a convenience sample of low-risk pregnant women attending the University Hospitals' Obstetric Clinic for pregnancy assessment and follow-up. All women except one were Caucasian, and all women except one were married and living with their husbands. We believe that a single woman may be under more stress during her pregnancy and that this stress might affect the fetus and its movement. This was the first pregnancy for 9 of the 21 women. At the start of the study, 8 women had one living child, and 4 women had more than one child. The mean maternal age was 30.6 years (SD=5.1, range=21-40), and the mean paternal age was 31.7 years (SD=4.8, range=19-39). The 4-Factor Index of Social Status[33] was computed by averaging scores on grade level completed and occupation for the mother and father. Maternal education ranged from high-school completion to over 22 years of education ([bar]X=14.3, SD=2.6), and mean paternal education was 15.6 years (SD=4.4). Occupations ranged from category 3 (housewife) to category 9 (professional). The mean score of 42.3 (SD=12.1) gave an educational-occupational social status range of 2 to 5 ([bar]X=3.5) out of a possible 5 categories.

The designation of "low risk" for fetuses was made during the initial obstetric visit based on maternal medical history and health status and was checked at each subsequent visit by the obstetrician and the obstetrical nurse. The sex of the fetuses (12 male, 9 female) and their birth age, birth weight, length, head circumference, and Apgar scores at 1 and 5 minutes were recorded at birth. These data and the follow-up of the fetuses (Tab. 2) by examination of their medical records at a mean age of 12.3 months (SD=5.5, range=4-19) indicate that we had identified a low-risk sample of pregnant women whose pregnancies were within normal limits and whose infants were functioning within normal limits, according to physician report, within the first year after birth. There were no multiple births.

Table 2. Birth Data and Mean Time of Follow-up for 21 "Low-Risk" Fetuses

           Birth Age (wk)   Birth Weight (g)   Length (cm)

[Bar]  X        39.8             3,597             52.4
SD               0.9               506              2.5
Range        38.4-41.4        2,710-4,470       50-57.5

                Head             Apgar            Age at
           Circumference        Score at        Follow-up
                (cm)              (cm)             (mo)

[Bar]  X        34.9            8.1, 8.8           12.3
SD               1.4            1.1, 0.6            5.5
Range        33-37.5            5-9, 7-10           4-19

Instrumentation

Based on our extensive experience observing fetal movement clinically and in research, the following criteria were established for longitudinal measurement of movement: (1) Fetal position needed to be recorded prior to identifying extremity movement, (2) right and left hand movements should be scored separately to gain as much data as possible and describe asymmetries, and (3) the movement of one body part (eg, the hand) needed to be scored in order to collect longitudinal data beyond 18 to 20 weeks. After 18 to 20 weeks, the whole fetus cannot be seen on the imaging screen. The combined results of the scoring of the right and left hand movements are presented in this article. The 7 hand movements and a sample score sheet are shown in Table 3.

Table 3. Fetal and Neonatal Movement System and Sample Computer Printout Showing Hand Movement

Column
1. Position (in relation to gravity)
   0 - Cannot tell
   1 - Right side lying
   2 - Left side lying
   3 - Prone
   4 - Supine
   5 - Sitting (upright)
2. Right hand
   0 - Cannot tell
   1 - Hand to/at mouth (on lips or in mouth)
   2 - Hand to/at face/head (excluding lips)
   3 - Hand to/at trunk (from shoulders to hips)
   4 - Hand to/at knee/foot
   5 - Hand to/at uterine wall/mattress
   6 - Hand near mouth (in fluid/air)
   7 - Hand away from body (in fluid/air)
3. Left hand
   0 - Cannot tell
   1 - Hand to/at mouth (on lips, in mouth)
   2 - Hand to/at face/head (excluding lips)
   3 - Hand to/at trunk (from shoulders to hips)
   4 - Hand to/at knee/foot
   5 - Hand to/at uterine wall/mattress
   6 - Hand near mouth (in fluid/air)
   7 - Hand away from body (in fluid/air)

Data Set File: 37-14-JS
Description: Position, right and left hand. Time in hours,
minutes, seconds, frames.

Record   Time          Code    Description

1        00:00:04.14   Start
2        00:00:04.15   000     Cannot see
3        00:00:05.18   366     Prone, hands near mouth in fluid
4        00:00:06.09   362     Prone, right hand near mouth, left
                                 hand at head
5        00:00:10.12   302     Prone, cannot see hand, left hand
                                 at head
6        00:00:16.20   366     Prone, hands near mouth in fluid
7        00:01:36.04   377     Prone, hands away from body in fluid

A Phillips P-700 ultrasound imager(*) and a Panasonic AG 185 (X8 power zoom lens) videorecorder with tripod([dagger]) were used to obtain the fetal and neonatal data. A super-VHS tape deck and 38.1-cm (15-in) high-resolution monitor were used for viewing the videotapes. Data were input directly to a computer via a software package designed for start-stop and continuous scoring.([double dagger])

Procedure

Each woman was identified by the clinic nurse during her initial clinic visit at 12 weeks. Appointments were made for the first ultrasound imaging at 14 weeks according to maternal estimation of pregnancy duration. Biparietal diameter and femoral length were used to confirm maternal dates. The ultrasound images were taken at 14 weeks ([bar]X=14.0 weeks, SD=4 days, range=13.0-15.6 weeks), 20 weeks ([bar]X=20.0 weeks, SD=6 days, range=18.3-20.6 weeks), 26 weeks ([bar]X=26.1 weeks, SD=6 days, range=24.6-28.5 weeks), 32 weeks ([bar]X=31.7 weeks, SD=3 days, range=30.4-32.4), and 37 weeks ([bar]X=37.0 weeks, SD=4 days, range=36.2-38.1). Each woman was interviewed prior to ultrasound imaging to record whether there were any changes in work or family stress (ie, stressors that could possibly affect the movement of the fetus) that she may have perceived over the preceding 6 weeks. The third author obtained the images in mid-afternoon with the woman in a semi-recumbent position, with diminished lighting, consistent with clinical obstetrical imaging. People who were significant to each woman were encouraged to attend the imaging. One day after birth ([bar]X=39.9 weeks, SD=7 days, range=38.4-41.4 weeks), the newborns were videotaped in 3 randomly assigned positions: right and left side lying and supine. The prone position was not used because hand movement was constrained in that position. The camera was positioned above the infant so that the whole body could be viewed. The side-lying position was maintained by a bendable positioning aid (Bendy Bumper([sections])). An overhead heater maintained the unclothed child's temperature at 37 [degrees] to 38 [degrees] C. Super-VHS videotapes were copied with a digital time code. The mean duration of the videotaped images was recorded (Tab. 4). The percentages of fetal imaging film in which either the right or left hand, or both, could be visualized and scored were 83%, 83%, 91%, 86%, and 87% for the 5 imaging sessions, respectively. A mean of 74 minutes of imaging per subject (SD=1.8) was available for assessment. The total time that hands could be seen was determined and became the baseline for figuring the percentage of time (duration) that the hands were moving to and were at 1 of 7 locations. This percentage was calculated for each hand movement, for each fetus, at each age. The percentage of time available for scoring postnatally was 76% because the movement of the infant was not scored in Brazelton-designated states 1 (deep sleep) and 6 (crying).(34) The frequency was the number of times that a specific hand movement occurred within the total time that the hands could be seen clearly and thus scored accurately.

Table 4. Total Duration of Ultrasound Images and Percentage of Videotaped Images Scored at Each Gestational Age and at 1 Day After Birth

                  14 weeks    20 weeks     26 weeks

Duration (s)
  [bar]X          978         925          881
  SD              231         137          113
  Range           469-1,502   788-1,408    481-1,036

Percentage of
 image scored
  [bar]X          83          83           91
  SD              13          15            7
  Range           49-97       50-98        79-100

                  32 weeks    37 weeks     1 day

Duration (s)
  [bar]X          934         706          1,437
  SD              234         226          665
  Range           565-1,452   281-1,069    107-2,542

Percentage of
 image scored
  [bar]X          86          87           76
  SD              7           11           23
  Range           69-98       51-96        10-100

The scoring was the same for the fetal and neonatal movements. The position of the fetus was documented on the videotape by the obstetrician. In each position observed, start and stop times for hand movements were scored at 30 frames per second. An extensive scoring protocol (Appendix) was established to assist in accuracy of scoring.

Videotapes were viewed randomly and scored by the first author. Ten percent of the videotaped images, randomly selected from all of the images, was scored independently by the first 2 authors. Percentage of agreement was determined for the scores for the right and left hands combined. A kappa statistic was generated to control for chance agreement. Overall mean percentage of agreement was 84% (range=65%-95%). The overall mean kappa was .72 (range=.57-.95). To achieve this level of reliability on a 2-dimensional viewing of a 3-dimensional movement, extensive training with non-subject images over several months prior to scoring of any subject's videotapes was necessary. Agreement was similar for all movements.

Data Analysis

Two major approaches were taken to analyze the duration and frequency data: use of the nonparametric Friedman 2-way analysis of variance by ranks and a follow-up of age differences using the Wilcoxon signed-ranks test. Nonparametric tests were selected so that no assumptions were made about the measurement scale or the distribution of the data. The Friedman 2-way analysis of variance by ranks was used with each fetal subject acting as its own control. We believe that this approach increases precision because measurements across the time points are correlated with one another. With this nonparametric approach, 2 types of comparisons were made: (1) the row-mean-score statistic (chi square with number of repeated measurements minus 1 degree of freedom) identified overall differences in scores of motor responses at different gestational ages for each movement, and (2) the correlation statistic (chi square with 1 degree of freedom) accounted for the ordering of the repeated measurements, which was used to identify linear trends in the data. When overall differences but no trends were observed, pair-wise comparisons were examined (P=.01) to detect differences. When trends were observed, pair-wise comparisons were used (P=.05) to indicate any possible support for the trend. To determine the association between fetal and neonatal measurements, Spearman rank correlations and a regression were computed. Statistical results are presented in terms of (1) overall differences, (2) trends, and (3) differences over the 6 time periods for data on durations and data on frequencies. Figures 1 through 6 show results of durations only.

[Figures 1-6 ILLUSTRATION OMITTED]

Results

The movement of the fetuses' hands appeared to be directed to specific body parts and uterine locations and was variable. Because of the variability, medians were used to describe the duration and frequency of behaviors. To indicate the large variability in the data, the maximum percentage obtained by one subject on one behavior is depicted for each age category in Figures 1 through 6. Medians and ranges have been used in cases of fetal data variability.[12] These maximum scores were distributed randomly over the 21 subjects, with no outliers. For essentially all behaviors at all ages, the range in percentages of duration extended from 0% to the maximum percentage shown for one subject. Movement durations are depicted in Figures 1 through 6. At each gestational age and at neonatal day 1, the median percentage of duration of the 6 movements for the right and left hands combined is shown. To permit scoring of all movements, we scored 7 categories of movement, including the "hand to/at the trunk" movement. Few instances of the "hand to/at the trunk" movement were observed, so analysis was not conducted on this movement.

Prenatal and Postnatal Duration of Movements

There was a difference between prenatal and postnatal duration of movements over the 6 time periods. The percentages on which these results are based are shown in Figures 1 through 6.

Overall differences. For all movements (P [is less than] .038) except the "hand to/at uterine wall/mattress" movement, overall differences existed (Figs. 1-5).

Trends. Decreasing linear trends, from 14 weeks through the postnatal period, were noted in the following movements: "hand near mouth" (P [is less than] .005) (Fig. 2), "hand to/at knee/foot" (P [is less than] .006) (Fig. 4), and "hand away from body (in fluid)" (P [is less than] .012) (Fig. 5). Through the postnatal period, an increasing linear trend was noted for the "hand to/at uterine wall/mattress" movement (P [is less than] .043) (Fig. 6).

Pair-wise comparisons. There was an increase in the "hand to/at mouth" movement postnatally (Fig. 1) when compared with each gestational week. Postnatal "hand to/at face/head" movement scores were not different from prenatal "hand to/at face/head" movement scores (Fig. 3). In contrast, the-"hand to/at knee/foot" movement showed a decrease postnatally when compared with all gestational ages (Fig. 4).

Prenatal Duration of Movement Alone

Overall differences. There were no overall differences in prenatal duration of movement for the "hand to/at knee/foot" and "hand away from body (in fluid)" movements (Figs. 4 and 5).

Trends. A linear decrease (P [is less than] .001) existed over the gestational period in the "hand to/at mouth" movement (decrease in median duration of movement from 3 to 0, expressed as percentage of time observed) (Fig. 1) and the "hand near mouth" movement (decrease in median duration of movement from 21 to 3, expressed as percentage of time observed) (Fig. 2). No linear trends existed in the following movements: "hand away from body (in fluid)" (Fig. 5), "hand to/at uterine wall/ mattress" (Fig. 6), and "hand at/to knee/foot" (Fig. 4).

Pair-wise comparisons. Comparison of behaviors indicated duration of movement in the "hand to/at face/ head" movement (Fig. 3) decreased (P [is less than] .002) between 14 and 32 weeks and increased (P [is less than] .008) between 32 and 37 weeks. The "hand to/at mouth" and "hand near mouth" movement comparisons were similar, with the 37-week data less than the 14-week data (P [is less than] .003) and the 20-week data (P [is less than] .035) (Figs. 1, 2). For the "hand to/at mouth" movement, the 37-, 32-, and 26-week data were also different from the 14-week data (P [is less than] .025) (Fig. 1).

Although there were small median percentages for the "hand to/at knee/foot" movement over the gestational period, there was an extremely wide range of this movement at 32 weeks and even at 37 weeks of gestation (Fig. 4). Observed decreases in duration of movement in the "hand away from body (in fluid)" movement at 26 weeks were not significant due, in part, to 3 subjects whose durations were maximum for this movement (Fig. 5). For these same subjects, durations at 26 weeks for the "hand to/at uterine wall/mattress" movement were minimum.

Prediction

In an attempt to predict postnatal measurements, Spearman rank correlations resulted in the "hand to/at face/ head" movement showing the only "promising" association. A regression analysis indicated that scores at 32 weeks of gestation were the best predictors of postnatal scores (P=.026).

Prenatal and Postnatal Frequency of Movement

With the postnatal data included, there were overall differences in the frequency data in all 6 movements (Tab. 5). Few trends, however, were noted. The only frequency trend confirmed the decreasing duration trend for the "hand to/at knee/foot" movement from 14 weeks of gestation to postnatal day 1. Follow-up pair-wise comparisons among the time periods showed the influence of the postnatal data on the analysis. The prenatal data, therefore, were evaluated without the postnatal data.

Table 5. Median Frequencies and Ranges (in Parentheses) of 7 Hand Movements at 5 Gestational Ages and at Postnatal Day 1

Hand Movement                        14 weeks    20 weeks

Hand to/at mouth                      3 (0-20)    2 (0-12)
Hand near mouth                      15 (0-12)   12 (0-27)
Hand to/at face/head                 13 (2-26)   10 (0-32)
Hand to/at knee/foot                  4 (0-16)    3 (0-11)
Hand away from both (in fluid/air)   21 (5-39)   17 (1-38)
Hand to/at uterine wall/mattress      5 (0-30)    9 (1-24)
Hand to/at trunk                      2 (0-12)    2 (0-7)

Hand Movement                        26 weeks    32 weeks

Hand to/at mouth                      0 (0-2)    0 (0-7)
Hand near mouth                       2 (0-27)   2 (0-32)
Hand to/at face/head                  3 (0-12)   2 (0-14)
Hand to/at knee/foot                  2 (0-12)   2 (0-21)
Hand away from both (in fluid/air)    5 (0-29)   5 (0-20)
Hand to/at uterine wall/mattress      3 (0-20)   2 (0-18)
Hand to/at trunk                      0 (0-65)   0 (0-11)

Hand Movement                        37 weeks      1 day

Hand to/at mouth                      0 (0-2)    20 (2-56)
Hand near mouth                       2 (0-11)   29 (7-109)
Hand to/at face/head                  3 (0-12)   32 (6-81)
Hand to/at knee/foot                  1 (0-8)     0 (0-5)
Hand away from both (in fluid/air)    2 (0-14)   43 (3-57)
Hand to/at uterine wall/mattress      2 (0-13)   28 (0-76)
Hand to/at trunk                      1 (0-5)    16 (2-60)

Prenatal Frequency of Movement Alone

Excluding the postnatal data, there were overall differences (P [is less than] .003) in all movements throughout gestation except for the "hand to/at knee/foot" movement. A decreasing linear trend (P [is less than] .001) was noted in all frequency data throughout the gestational period. In general, there were high ranks at 14 weeks, decreasing to 37 weeks of gestation (Tab. 5).

Discussion

The aim of this study was to enhance understanding of human movement by addressing an area of development that has been neglected in the physical therapy literature. The beginnings of human movement, which we are now able to observe naturalistically, provide a window for our study of functional movement, the linkage of fetal to newborn movement, and the future delineation of aberrant movement.

Hand Directedness

At the level of analysis of this study, fetal hand behaviors were described in terms of movement directed to and at a body part. The apparent goal orientation of fetal movement from 14 weeks of gestation led us to assume a functional importance for the movements. Whether the modifier of "function" was incorrectly used in this study or applied to the wrong behaviors is unclear. A developing sensorimotor system, however, apparently was being observed. Movement of the hand occurs around the mouth with frequent subsequent sucking, and movement of the hand to specific body parts occurs with subsequent molding of the hand around the body part. Movement of the hand to the uterine wall can be observed, with subsequent flattening and sliding of the palm against the uterine wall. Fingering, grasping, and manipulation of the umbilical cord occurs. Fetal grasping of the umbilical cord can cause variable heart rate decelerations[35] and thus may be a behavior critical to assess in future studies. The functional appearance of much of fetal movement suggests the potential early relationship of sensorimotor status and environmental conditions that constrain or enhance movement (ie, affordances).[36-38]

These directed paths of movement are reminiscent of "contact paths" described in the ontogeny of facial grooming in mice,[39] a functional activity achieved by modifying environmental permissive conditions. Future intervention by physical therapists as well as obstetricians with fetuses and neonates may be better contemplated after identifying functional correlates of complex behavioral repertoires and recording environmental characteristics that permit their occurrence.

Variability

The wide ranges of scores of the movement of these low-risk fetuses was anticipated. Some authors[12,13] have noted this variability in young developing fetuses. Variability is a characteristic of "low-skill" movement,[40] a category into which low-risk fetal movement certainly falls. This variability may well be modified by a myriad of as yet unexplored exogenous as well as endogenous factors. We interpreted the variability that we observed in our study as a dynamic characteristic of human movement. Decreased variability, presenting as stereotypical behavior, may be observed in high-risk fetuses or those fetuses with a problem with sensorimotor adaptation.[43] Indeed, the movement of fetuses with neural tube defects has been observed as lacking in variability. In addition to our observations of variability, our results suggest that there was an emerging underlying structure. The combination of variations in variability and changes in movement during gestation might provide data to assist in prenatal diagnosis.

Postnatal Period Versus Prenatal Period

Differences in the movement scores occurred between the prenatal and postnatal periods. We assumed that these differences are based on physiological and environmental factors known to occur over the transition from the fetal period to the neonatal period. Neural maturation over the brief span between 37 weeks of gestation and 1 day after birth could not be sufficient to be responsible for these changes. The data show an increase in the "hand to/at mouth" and "hand near mouth" movements postnatally, a decrease in the "hand to/at knee/foot" and "hand away from body (in air)" movements, and an increase in the "hand to/at uterine wall/mattress" movement. These results are compatible with neonates' participation in control of the regulation of their behavioral state[36] and with the increased effect of gravity after birth. It is possible that the differences noted between the fetuses and the neonates could be used diagnostically (ie, if no differences were noted, delay or other developmental problems might be suggested). Because of the multitude of changes occurring over the birth transition, this interpretation is highly speculative at present.

Measurement of changes in movement across this transition have been difficult in the past because of a lack of an assessment tool that could compare the same behaviors in 2 totally different environments. The differences noted in our study suggest that we are measuring the effects of differing environmental factors that encourage or restrain movements. At present, extensive analysis of specific movements across this transition may provide spurious results. Research emphasis may best be placed on further analysis of the uterine environment, an ecological approach not unfamiliar to dynamical systems research on movement.[41]

Trends

Our hypothesis that the "hand to/at mouth" movement would be performed more frequently as development progressed was not supported. Instead, there was a linear decrease in the occurrence of this movement throughout gestation, with the hypothesized increase being observed only neonatally. The "hand to/at mouth" movement appeared in the newborn where it could be interpreted as a more functional activity assisting in neonatal modification of behavioral state and oral feeding. This type of developmental trend may be depicting the adaptedness of the organism to changing environmental and intrinsic challenges. The linear decrease in the occurrence of the "hand to/at mouth" and "hand near mouth" movements during gestation in our study is difficult to compare with the findings of other studies because of methodological differences. Roodenburg et al[13] and de Vries et al[14] identified a decrease in general movements over the gestational period, a result that was consistent in our data with only the "hand to/at mouth" and "hand near mouth" movements.

Other trends (U-shaped, inverted U-shaped) are apparent in our data. The decrease in the occurrence of the "hand to/at face/head" movement at 32 weeks and the "hand near mouth" movement at 37 weeks recalls regression curves described in the human infant,[25] in which a movement occurs, disappears, or is of short duration and later reappears, often in a more advanced pattern. At 32 weeks, the "hand away from body (in fluid)" movement was observed more often than the other hand movements. At 37 weeks, the "hand to/at face/head" movement was observed more often than the other hand movements. Only neonatally were the "hand to/at mouth" and "hand near mouth" movements observed. Although the fetus is able to move the hands to or at the mouth and to other specific locations prenatally, only in the postnatal period may this movement pattern be considered "intentional," as Butterworth and Hopkins[22] have suggested. The-tendency for the "hand to/at knee/ foot" and "hand away from body (in fluid)" movements to peak at 32 weeks may reflect some tactile proclivity in the mid-gestational fetus, gravitational constraint in the newborn, and neural maturation. The potential relationship of hand movements and neural status was not a direct focus of this study. Changes in these fetal behaviors, however, suggest that a developmental process could have been observed and that underlying that process would be neural, as well as musculoskeletal and physiological, maturation.

Frequency data indicated that declines occurred over the course of gestation, with increases in frequency occurring neonatally, suggesting possible space limitations as gestation progressed. This finding was not true for the "hand to/at knee/foot" movement data, which showed a decreasing trend through day 1, indicating the difficulty postnatally of antigravity movements of large body parts, as suggested by Thelen and Ulrich's work.[25]

Another difference was the decrease in frequency of the "hand to/at face/head" movement at 37 weeks, coinciding with an increase in duration of the "hand to/at face/head" movement during this period. This decrease in frequency at 37 weeks might be attributed to increasing energy conservation in preparation for birth, as well as to decreasing uterine space.

Limitations

Scoring based on a 3-dimensional view of a 3-dimensional movement threatens the credibility of the data. The achievement of good reliability was based on extensive training. Although the observers' scores were reliable, it is possible that-they were not scoring the true behaviors or that the behaviors were being modified by the procedures (eg, the imaging). Three-dimensional viewing is currently being used to assess patients with breast cancer, but it has not been commonly available to assess the movement of fetuses. The future availability of this technology may assist observers in more valid scoring.

The number of subjects in this study was relatively small. Additions to this database might permit further analysis, such as analysis of differences between male and female fetuses. Achieving significance in some movements using nonparametric statistics, specifying the actual amount of data used for analysis, and using a variety of subjects over the course of pregnancy with follow-up do not lessen the need for a larger subject pool.

We did not assess fetal behavioral state directly, and we scored neonatal movements only in behavioral states 2 through 5.[34] Although the initial linkage among the behavioral state variables of movement, heart rate, and eye movements may appear earlier (ie, at 20 weeks[42,43]), Swartjes et al[44] determined that "true behavioral states" were not found until 32 weeks of gestation. An attempt was made to consider variability in behavioral state factors in the following ways: by scoring the movements of over 20 subjects, by performing ultrasound imaging at least 2 hours after a meal so that glucose levels would not affect the movement, and by assessing fetuses at approximately the same time of day. Diurnal variation in fetal movement has been established[45] and could affect the duration and frequency of movement.

Concerns about the safety of ultrasound imaging have limited fetal research in this country. Human subject review committees and granting agencies have questioned the risk-benefit ratio of the use of this technology. In our study, identification of additional subjects, rather than increasing observation time, was used to obtain data for analysis. Recent increases in the amount of imaging time allowed by federal granting agents[42] are welcomed and supported by long-term studies on the safety of ultrasound imaging.[31] Most of the studies conducted in the Netherlands have used 1-hour imaging segments for analysis.

The results of studies of nutritional and other environmental factors vary. Based on research describing social influences on pregnancy,[46,47] there should be consideration of the potential effect on fetal movement of the environment beyond the uterus. Differences in lifestyle, nutrition, family, work, and stress levels may help to explain the large variance that all investigators have observed in movement duration and frequency. The demographics of this sample describe a wide variety of occupational and educational experience. These influences need to be disentangled in future studies, as their components might affect motor development.

Future Directions

Investigation of fetal sensorimotor development may detect behavioral patterns characteristic of impairment[4,48-50] and may increase our understanding of disorders such as cerebral palsy. Current data suggest that 60% of neurodevelopmental disabilities are caused during the prenatal period.[51] Within this category, cerebral palsy has been reported to be the most "life-limiting" of childhood disabilities. Kuban and Leviton have stated that "efforts to prevent cerebral palsy will require a focus on factors and events during pregnancy, including those that predispose the mother and fetus to preterm delivery."[52(p193)] Identification of changes in movement patterns over gestation may assist in explicating the role of some of these factors and events.

To address such critical questions, appropriate assessments will be necessitated. As the category of "general movements" is being further specified, so too may the movement patterns that we studied. Theoretical applications, such as those expressed by Sporns and Edelman,[53] will be important guides in this process. Further behavioral analysis of the human fetus may provide needed information about approaches and equipment that to date have not been fully tested.

Conclusion

Movement of low-risk fetuses appears to be nonrandom and is variable across the gestational period. Major differences in movement occur between fetuses and neonates. Developmental trends using conservative levels of significance in nonparametric analyses were identified in the movement of 21 low-risk fetal subjects. A linear trajectory decreasing over the pregnancy was noted in the "hand to/at mouth" and "hand near mouth" movements. A regression occurred in the "hand to/at face/head" movement pattern observed early in gestation, followed by a diminution in the recorded duration of this movement pattern and then a reappearance of this movement pattern neonatally. Other movement patterns, such as the "hand to/at knee/foot" movement pattern, were not observed early in gestation, appeared during midpregnancy, and then decreased across the perinatal period, apparently awaiting physiological stability and environmental support before further expression. The reciprocal relationship between motor and sensory development is suggested by the apparent goal-oriented and interactive characteristic of these movements.

Acknowledgments

We are appreciative of the statistical advice and assistance of Dr Gary Koch and Wendy Greene, Department of Biostatistics, School of Public Health, University of North Carolina at Chapel Hill.

(*) Phillips Medical Equipment, 710 Bridgeport Ave, Shelton, CT 06484.

([dagger]) Panasonic, 1 Panasonic Way, Secaucus, NJ 07094.

([double dagger]) Observational Coding System (OCS), Triangle Research Collaborative Inc, PO Box 12167, Research Triangle Park, NC 27709.

([sections]) Children's Medical Ventures Inc, 541 Main St S, South Weymouth, MA 02190.

References

[1] Harrison MR, Golbus MS, Filly RA. The Unborn Patient: Prenatal Diagnosis and Treatment. New York, NY: Grune & Stratton; 1984:1-10.

[2] Reinold E. Clinical value of fetal spontaneous movements in early pregnancy. J Perinatol Med. 1973;1:65-69.

[3] Okado N, Kojima T. Onogeny of the central nervous system: neurogenesis, fiber connection, synaptogenesis, and myelination in the spinal cord. In: Prechtl HFR, ed. Continuity of Neural Functions From Prenatal to Postnatal Life. Philadelphia, Pa: JB Lippincott Co; 1984: 31-45.

[4] Koyanagi T, Horimoto N, Maeda H, et al. Abnormal behavioral patterns in the human fetus at term: correlation with lesion sites in the central nervous system after birth. J Child Neurol. 1993;8:19-26.

[5] Grillner S. Neural networks for vertebrate locomotion. Sci Am. Jan 1996;274:64-69.

[6] Bekoff A. Development of motor behavior in chick embryos. In: Lecanuet JP, Fifer WP, Krasnegor NA, Smotherman WP, eds. Fetal Development: A Psychobiological Perspective. Hillsdale, NJ: Lawrence Erlbaum Associates Inc; 1995:191-204.

[7] Krasnegor NA, Blass EM, Hofer MA, Smotherman WP. Perinatal Development: A Psychobiological Perspective. New York, NY: Academic Press Inc; 1987.

[8] Riegger-Krugh C, Blair A, Sparling JW. Assessment of fetal knee angular velocity as a possible method to determine the effect of prenatal exposure to cocaine. Physical and Occupational Therapy in Pediatrics. 1996;16(1/2):173-186.

[9] Sival DA. Studies on fetal motor behaviour in normal and complicated pregnancies. Early Hum Dev. 1993;34:13-20.

[10] Smotherman WP, Robinson SR. Tracing developmental trajectories into the prenatal period. In: Lecanuet JP, Fifer WP, Krasnegor NA, Smotherman WP, eds. Fetal Development: A Psychobiological Perspective. Hillsdale, NJ: Lawrence Erlbaum Associates Inc; 1995:15-32.

[11] Reigger-Krugh C. Relationship of mechanical and movement factors to prenatal musculoskeletal development. Physical and Occupational Therapy in Pediatrics. 1993;12(2/3):19-37.

[12] de Vries JIP, Visser GHA, Prechtl HFR. The emergence of fetal behaviour, II: quantitative aspects. Early Hum Dev. 1985;12:99-120.

[13] Roodenburg PJ, Wladimiroff JW, van Es A, Prechtl HFR. Classification and quantitative aspects of fetal movements during the second half of normal pregnancy. Early Hum Dev. 1991;25:19-35.

[14] de Vries JIP, Laurini RN, Visser GHA. Abnormal motor behaviour and developmental postmortem findings in a fetus with Fanconi anaemia. Early Hum Dev. 1994;36:137-142.

[15] Prechtl HFR. Qualitative changes of spontaneous movements in fetus and preterm infant are a marker of neurological dysfunction. Early Hum Dev. 1990;23:151-158.

[16] Ververs IAP, de Vries JIP, van Geijn HP, Hopkins B. Prenatal head position from 12-38 weeks, I: developmental aspects. Early Hum Dev. 1994;39:83-91.

[17] Hepper PG, Shahidullah S, White R. Handedness in the human fetus. Neuropsychologia. 1991;29:1107-1111.

[18] Geerdink JJ, Hopkins B. Effects of birthweight status and gestational age on the quality of general movements in preterm newborns. Biol Neonate. 1993;63:215-224.

[19] Hadders-Algra M, Klip-Van den Nieuwendijk AWJ, Martijn A, van Eykern LA. Assessment of general movements: towards a better understanding of a sensitive method to evaluate brain function in young infants. Dev Med Child Neurol. 1997;39:88-98.

[20] Sparling JW, Wilhelm IJ. Quantitative measurement of fetal movement: Fetal-Posture and Movement Assessment (F-PAM). Physical and Occupational Therapy in Pediatrics. 1993;12(2/3):97-114.

[21] Piaget J. The Origins of Intelligence in Children. New York, NY: International Universities Press; 1952.

[22] Butterworth G, Hopkins B. Hand-mouth coordination in the newborn baby. British Journal of Developmental Psychology. 1988;6:303-314.

[23] Bekoff A, Kauer JA. Neural control of hatching: fate of the pattern generator for the leg movements of hatching in posthatching chicks. J Neurosci. 1984;4:2659-2666.

[24] de Vries JIP, Visser GHA, Prechtl HFR. The emergence of fetal behaviour, III: individual differences and consistencies. Early Hum Dev. 1988;16:85-103.

[25] Thelen E, Ulrich BD. Hidden skills: a dynamic systems analysis of treadmill stepping during the first year. Monogr Soc Res Child Dev. 1991;56:1-98.

[26] Reiss RE, Foy PM, Mendiratta V, et al. Ease and accuracy of evaluation of fetal hands during obstetrical ultrasonography: a prospective study. J Ultrasound Med. 1995;14:813-820.

[27] Long T, McCusker S, Ruble J, Sparling JW. Periods of activity and inactivity in the 12- to 16-week fetus. Physical and Occupational Therapy in Pediatrics. 1993;12 (2/3):163-179.

[28] Pittman R, Oppenheim RW. Neuromuscular blockade increases motoneurone survival during normal cell death in the chick embryo. Nature. 1978;271:364-366.

[29] Forger NG, Breedlove M. Motoneuronal death during human fetal development. J Comp Neurol. 1987;264:118-122.

[30] Oppenheim RW, Nunez R. Electrical stimulation of hindlimb increases neuronal cell death in chick embryo. Nature. 1982;295:57-59.

[31] Tucker LB, Gentry WR, Thomas EA, Sparling JW. Ultrasound safety: a descriptive study of the potential effects of early imaging. In: Sparling JW, ed. Concepts in Fetal Movement Research. Binghamton, NY: The Haworth Press Inc; 1993:77-95.

[32] American Institute of Ultrasound in Medicine. Bioeffects considerations for the safety of diagnostic ultrasound. J Ultrasound Med. 1988;7:S1-S38.

[33] Hollingshead AB. The Four-Factor Index of Social Position. New Haven, Conn: Privately Published; 1975.

[34] Brazelton TB. The Neonatal Behavioral Assessment Scale. 2nd ed. Philadelphia, Pa: JB Lippincott Co; 1984. Clinics in Developmental Medicine; no. 88.

[35] Petrikovsky BM, Kaplan GP. Fetal grasping of the umbilical cord causing variable fetal heart rate decelerations. J Clin Ultrasound. 1993;21:642-644.

[36] Thelen E, Corbetta D, Kamm K, et al. The transition to reaching: mapping intention and intrinsic dynamics. Child Dev. 1993;64:1058-1098.

[37] vonHofsten C. The organization of arm and hand movements in the neonate. In: von Euler C, ed. The Neurobiology of Early Infant Behavior. London, England: Macmillan Ltd; 1989:129-142.

[38] Gibson JJ. The Ecological Approach to Visual Perception. Boston, Mass: Houghton Mifflin Co; 1979.

[39] Golani I, Fentress JC. Early ontogeny of face grooming in mice. Dev Psychobiol. 1985;18:529-544.

[40] Newell KM, vanEmmerik REA, Sprague RL. Stereotypy and variability. In: Newell KM, Corcos DM, eds. Variability and Motor Control. Champaign, Ill: Human Kinetics Inc; 1993:475-496.

[41] Riccio GE. Information in movement variability about the qualitative dynamics of posture and orientation. In: Newell KM, Corcos DM, eds. Variability and Motor Control. Champaign, Ill: Human Kinetics Inc; 1993:317-357.

[42] DiPietro JA, Hodgson DM, Costigan KA, et al. Development of fetal movement: fetal heart rate coupling from 20 weeks through term. Early Hum Dev. 1996;44:139-151.

[43] Drogtrop AP, Ubels R, Nijhuis JG. The association between fetal body movements, eye movements, and heart rate patterns in pregnancies between 25 and 30 weeks of gestation. Early Hum Dev. 1990;23: 67-73.

[44] Swartjes JM, van Geijn HP, Mantel R, et al. Coincidence of behavioural state parameters in the human fetus at three gestational ages. Early Hum Dev. 1990;23:75-83.

[45] Nasello-Paterson C, Natale R, Connors G. Ultrasonic evaluation of fetal body movements over twenty-four hours in the human fetus at twenty-four to twenty-eight weeks' gestation. Am J Obstet Gynecol. 1988; 158:312-316.

[46] Groome LJ, Swiber MJ, Bentz LS, et al. Maternal anxiety during pregnancy: effect on fetal behavior at 38 to 40 weeks of gestation. J Dev Behav Pediatr. 1995;16:391-396.

[47] Clarke AS, Wittwer DJ, Abbott DH, Schneider ML. Long--term effects of prenatal stress on HPA axis activity in juvenile rhesus monkeys. Dev Psychobiol. 1994;27:257-269.

[48] Bejar R, Wozniak P, Allard M, et al. Antenatal origin of neurologic damage in newborn infants, I: preterm infants. Am J Obstet Gynecol. 1988;159:357-363.

[49] Horimoto N, Koyanagi T, Maeda H, et al. Can brain impairment be detected by in utero behavioural patterns? Arch Dis Child. 1993;69:3-8.

[50] Nelson KB, Ellenberg JH. Antecedents of cerebral palsy: multivariate analysis of risk. N Engl J Med. 1986;315:81-86.

[51] Lamb B, Lang R. Aetiology of cerebral palsy. Br J Obstet Gynaecol. 1992;99:176-178.

[52] Kuban KCK, Leviton A. Cerebral palsy. N Engl J Med. 1994;330: 188-195.

[53] Sporns O, Edelman GM. Solving Bernstein's problem: a proposal for the development of coordinated movement by selection. Child Dev. 1993;64:960-981.

Appendix. Scoring Protocol

Computer Printout -- directions for use of Observational Coding System (OCS) are in the manual from Triangle Research Collaborative Inc.

Column 1: Position -- it is critical to first designate the hand being viewed. Check on Dr Chescheir's designation of view and follow through with that until a change is noted. If left side lying, then left hand will probably not be observed easily and may not move as much as right hand.

Columns 2 and 3 are designed to determine whether the hand is in fluid, interacting with the environment (uterine wall, umbilical cord, mattress) or directed to other body parts. Position is the preliminary essential for scoring right and left hands.

Movement Patterns

1. Score "hand to/at mouth" movement (1) when thumb or fingers are in mouth or touching lips, or hand is touching immediate perioralregion. Contact of mouth is of primary importance. If hand touches mouth and at the same time touches uterine wall, for example, score "1" for mouth contact. The dorsum of hand may be on the uterine wall, with fingers at the mouth. When hand touches 2 things, however, score what is touched by fingers or the palm of the hand.

2. Score "hand to/at face/head" movement (2) when hand is in contact with the face or head but outside of the perioral region described for "hand to/at mouth" movement. Note the hand often molds to the occiput, so observe carefully.

3. Score "hand to/at trunk" movement (3) when there is hand contact anywhere on the body from shoulders to hips, including the genital areas. This movement may initially appear as "hand away from body (in fluid)" movement, but fingers are contacting genitals.

4. Score "hand to/at knee/foot" movement (4) when hand is in contact with lower extremity. This movement is almost always contact with knee, with hand molded to knee, or with foot, where fingers are in contact with toes and sometimes ankle.

5. Score "hand to/at uterine wall/mattress" movement (5) when hand contacts placenta or uterine wall and, in the case of the newborn, when hand is in contact with the mattress. The volar surface of the hand is usually the hand part that is contact with the uterine wall.

6. Score "hand near mouth" movement (6) when fingers are in fluid between nose and shoulders/nipples or between both shoulders. Hands must be at or below eyes, within ears, and less than a hand away from mouth to be scored.

7. Score "hand away from body (in fluid)" movement (7) when hand is not in contact with uterine wall, placenta, or body part, nor is it in small area near mouth.

When neither hand can be seen adequately to score a hand behavior, score "0."

Common Problems

1. Any time image of the hand is lost to view, but no change in position is observed when image returns within 4 seconds, continue coding as before loss of image.

2. Contact with body part is characterized by no obvious fluid between part and hand. Mark code when hand contacts new part.

3. To score any behavior, it must occur for 15-30 frames/second. Any movements quicker than 15 frames/second, therefore, are eliminated from a change in score or are scored "0."

4. When 2 hands are present, follow 1 hand only during a movement segment, then repeat the observation and scoring for the other hand.

5. Must see body part (ie, hand, face, trunk, legs) to score hands, unless clearly identified just before or just after segment to be scored. In case of doubt, score fluid rather than contact. For example, when in doubt, assign a score of "6" rather than "5."

Newborn Scoring

1. Film. neonate 2-3 hours after feeding. Best time is at 7 AM.

2. Do not score state 1 (deep sleep) or state 6 (vigorous crying). Can gently awaken neonate, or provide calming through voice, touch, and cuddling.

3. Any contact with face predominates over contact with trunk or mattress.

4. Arms appear to be "locked" into a square area around the face. Score "6" when hand is below nose, between ears, and above nipples.

5. Remember that the hand contact determines the score. However, if arm is moving, rating a score of "7," and momentarily rests on the body, do not assign a score of "3" but a score of "7." Hand must make contact for 15 frames before changing scores.

Using computer program, score hand behaviors each time a new position, movement, or posture is observed. Cannot score posture if cannot score columns 1 and either 2 or 3.

Column 1--Body position of fetus

Column 2--Right hand

Column 3--Left hand

Scoring for postures:

0--Cannot see or identify

1--Hand clasp (hand to hand). The following definition from the Assessment of Preterm Individual Behavior[9] may be helpful in scoring accurately: "The infant grasps his own hand or clutches his hands midline to his own body. The hands each may be closed, but they hold onto each other or actively press against each other. Interdigitation of fingers of one hand with those of the other hand is a subcategory of hand clasp."(a) The hand contact can occur from below elbow to include hand. If forearms or hands are crossed, assume they are touching and score "1."

2--Hand grasp. Als described as "the hand opens and closes."(a) We suggest that there may be some thumb opposition, as in grasping the umbilical cord or a body part. This is to differentiate grasp from clasp or hand to umbilical cord. Code "2" only for grasp using individual hand. When hand to hand, score "1" even though it may be a hand grasp.

3--Hand to cord. Dorsum or ventral hand surface touches umbilical cord. If grasp of umbilical cord occurs, score "2" first, then immediately score "3" to obtain the duration of the grasp.

4--Finger splay. Finger extension and abduction. Als described as "the infant's hands open, and the fingers are extended and separated from each other.(a)

5--Wrist drop. Hand flexes at wrist more than 30o. Do not score if dorsum of hand is pushing against uterine wall Or body part. This posture is seen in fluid.

6--Fist. Full flexion of fingers.

7--Finger posturing. Individual finger or thumb movements.

(a) Als H. Manual for the Naturalistic Observation of Newborn Behavior (Preterm and Full-Term Infants), 1984. Available from author at Children's Hospital, Boston, Mass.

This study was supported in part by a grant (MCJ000149) to Dr Sparling from the Maternal and Child Health Bureau, Health Resources and Services Administration, US Department of Health and Human Services.

This study was approved by the Committee on the Protection of the Rights of Human Subjects, School of Medicine, University of North Carolina at Chapel Hill.

This article was submitted November 10, 1997, and was accepted August 17, 1998.

JW Sparling, PhD, PT, OT, was Associate Professor of Human Movement and Project Director of the Maternal and Child Health Postgraduate Training Grant at the University of North Carolina at Chapel Hill at the time this study was conducted. Address all correspondence to Dr Sparling at 1444 Center Grove Church Rd, Moncure, NC 27559 (USA) ([e-mail]).

J Van Tol, PhD, is Research Assistant, Amsterdam, the Netherlands. At the time of this study, she was a Postdoctoral Fellow with Dr Sparling.

NC Chescheir, MD, is Associate Professor of Obstetrics, School of Medicine, University of North Carolina at Chapel Hill.