- Email: [email protected]
Fetal habituation to vibroacoustic stimulation in relation to fetal states and fetal heart rate parameters
Early Human Development 61 (2001) 135–145 www.elsevier.com / locate / earlhumdev
Fetal habituation to vibroacoustic stimulation in relation to fetal states and fetal heart rate parameters a,
b
Cathelijne F. van Heteren *, P. Focco Boekkooi , Henk W. Jongsma a , Jan G. Nijhuis c a
Department of Obstetrics and Gynaecology, University Medical Centre Nijmegen, PO Box 9101, 6500 HB Nijmegen, The Netherlands b Department of Obstetrics and Gynaecology, St Elisabeth Hospital, Tilburg, The Netherlands c Department of Obstetrics and Gynaecology, University Hospital Maastricht, Maastricht, The Netherlands
Received 18 August 2000; received in revised form 20 November 2000; accepted 21 November 2000
Abstract Objectives: Fetal habituation to repeated stimulation has the potential to become a tool in the assessment of fetal condition and of the function of the fetal central nervous system (CNS). However, the influence of fetal quiescence and activity on habituation remains to be clarified. We studied habituation and the influence of fetal state and fetal heart rate (FHR) parameters on habituation in healthy term fetuses. Subjects and method: We studied habituation in 37 healthy fetuses in two tests with an interval of 10 min. The vibroacoustic stimuli were applied to the maternal abdomen above the fetal legs for a period of 1 s every 30 s. A fetal trunk movement within 1 s after stimulation was defined as a positive response. Habituation rate is defined as the number of stimuli applied before an observed non-response to four consecutive stimuli. The FHR patterns (FHRP) of the 10 min observation period before and after the tests were visually classified. Fetal states were defined according to the FHRP. Baseline FHR, FHR variability and the number of accelerations were calculated in a subgroup of 25 fetuses. Results: Of the 32 fetuses that responded normally during the first test, 26 habituated and six had persistent
*Corresponding author. Tel.: 1 31-243-616-801; fax: 1 31-243-541-194. E-mail address: [email protected] (C.F. van Heteren). 0378-3782 / 01 / $ – see front matter 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S0378-3782( 00 )00130-4
136
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
responses. The median habituation rate decreased significantly in the second test (P 5 0.001). There was no difference in habituation rate between fetuses that where initially in a quiet state and those in an active state. The FHR parameters before the first test and the difference between these FHR parameters before and after the test did not correlate with the habituation rate. Conclusions: Although the majority of healthy fetuses was able to habituate, the interfetal variability in habituation performance is such that testing of habituation seems not to be a sensitive tool for the assessment of the fetal CNS. This variability is neither the result of differences in fetal state nor of the various FHR parameters before testing, nor of the difference in change of FHR parameters arising from stimulation. 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Fetal habituation; Vibroacoustic stimulation; Fetal behavioural states; Fetal heart rate
1. Introduction Habituation is the decrease in, and ultimate cessation of, response to repeated stimulation. Habituation can be seen as a basic form of learning that requires an intact central nervous system (CNS) [1]. If habituation represents CNS functioning, then conditions affecting the CNS would be expected to affect habituation. This has been reported in studies on adults with various disorders of the CNS [2–6]. Several studies on habituation in normal fetuses have shown that fetuses are able to habituate to repeated stimuli [7–11]. Studies in compromised or chromosomally abnormal fetuses show that the habituation pattern of these fetuses is different from that of normal fetuses [12–14]. Therefore, fetal habituation might identify a fetus with poor fetal condition or impaired CNS functioning. Although fetal habituation has been studied widely [10,13–15], its clinical relevance remains unclear. This can be explained by the variety of methodologies that have been used to test fetal habituation — there still exists no standardised test. Neither is it clear whether an abnormal habituation pattern is the result of fetal distress, impaired functioning of the CNS, or of a difference in environmental or fetal physiological factors, such as fetal behavioural states. Behavioural states have been found to affect the magnitude of the fetal response and the time between the stimulus and the response [16–18]. Although one study has demonstrated that behavioural states did not significantly influence the habituation pattern to vibroacoustic stimulation [15], another study has shown that the decrement in response is caused by a behavioural state transition rather than by true habituation [19]. It remains unclear whether, and to what extent, fetal behavioural states influence habituation. Because of the potential clinical relevance of fetal habituation, it is essential to know whether fetal behavioural states and state transition affect habituation. This study was designed to test fetal habituation in healthy term fetuses and to study the influence of fetal states (fetal quiescence and activity), state transition and FHR parameters on habituation.
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
137
2. Subjects and method
2.1. Subjects Thirty-seven healthy pregnant women were recruited from a low-risk obstetric population (midwives’ practice) at the University Medical Centre Nijmegen in the Netherlands. The study was approved by the hospital’s ethical committee and all participants gave their written informed consent. Inclusion criteria were: (1) gestational age between 37 and 40 completed weeks (pregnancies were dated using the last menstrual period or by ultrasound when dates were uncertain); (2) the absence of maternal medical or obstetric complications; (3) single fetus without apparent structural anomalies and with an expected birth weight above the 10th percentile; (4) normal amniotic fluid volume as assessed by ultrasound; (5) no maternal use of alcohol, drugs or medication other than vitamins and / or iron.
2.2. Study design All habituation tests were done under the same conditions (the mother was not allowed to smoke, drink coffee or eat for 3 h before testing), in the same room, and by the same examiner between 4.00 p.m. and 7.00 p.m. The women were placed in a semi-recumbent position. The stimuli were produced by a fetal vibroacoustic stimulator (Corometrics model 146, Wallingford, CT, USA; audible sound 20 to 9000 Hz, vibrations between 67 and 83 Hz, sound level 74 dB at 1 m in air). Stimuli of 1 s duration were repeatedly applied to the maternal abdomen above the fetal legs every 30 s. The fetal trunk was displayed in a parasagittal plane by a real-time ultrasound scanner (Hitachi model EUB-525, Tokyo, Japan). The timing of the stimuli was chosen to occur when the fetus was not moving. A general movement of the fetal trunk within 1 s of application of the stimulus was defined as a positive response. Response decrement was noted as a changing from a more intense to a less intense response pattern with successive trials. A lack of response to four consecutive stimuli indicated habituation. We allowed a maximum number of 24 stimulus applications in each test. However, a minimum of four additional stimuli would be necessary to show habituation if the fetus responded to the 21st stimulus. We therefore stopped stimulating if a fetus still responded to the 21st stimulus. The habituation rate was defined as the number of stimuli applied before a fetus stopped responding. The procedure was repeated 10 min after the first test to investigate whether the fetus habituated more rapidly during the second series of stimuli. The FHR was recorded continuously with a cardiotocograph (Sonicaid FM7, Oxford, UK) at a paper speed of 3 cm / min, starting 10 min prior to the tests and continuing for 10 min afterwards. The outcome of each pregnancy was examined for birth weight, gestational age at delivery, Apgar scores, umbilical artery blood pH, and the presence of neonatal complications. The infants were followed up 3 months after birth by telephone interviews with one of the parents.
138
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
2.3. Data analysis In 37 patients, the FHR recordings of the 10 min observation periods before and after the first test were visually classified into four FHR patterns (FHRP) in accordance with Nijhuis et al. [20]. FHRP A is defined as a stable heart rate with little variability and isolated accelerations. FHRP B has a wider oscillation bandwidth than FHRP A with frequent accelerations. FHRP C is a stable pattern with a wider oscillation bandwidth than FHRP A and without accelerations. FHRP D is unstable, with large and long-lasting accelerations, which are frequently fused into a sustained tachycardia. As the incidence of FHRP C and D is relatively low, these episodes were considered to be an FHRP corresponding to an active state. Finally, the FHR recordings were divided into an FHRP corresponding to a quiet state (FHRP A) and into an FHRP corresponding to an active state (FHRP B, C and D). In our study, determination of whether the fetus was in a quiet state or an active state was based on analysis of FHRP alone, because such a distinction is possible in normal fetuses after 35 weeks gestation [21]. Computerized analysis of the FHR of a subgroup of 25 fetuses was carried out using a Sonicaid System 8002 (Oxford Sonicaid Ltd, Abingdon, UK) [22]. Baseline FHR, FHR variability (long-term variation and short-term variation) and the number of accelerations were calculated for the observation periods before and after the first test. Post-acquisition analysis of any fraction of the recording down to a minimum of 10 min is possible. The system reduces the data over 3.75 s (1 / 16 min epochs). The baseline FHR is defined as the mean rate averaged over all periods of low variation. If no low variation is present, it is derived from a statistical analysis. A baseline is fitted to the FHR trace and long-term variation is subsequently measured as the range of mean pulse intervals around the baseline, calculated minute by minute, excluding data from decelerations and errors. The system averages data over the total recording time in milliseconds (ms). Short-term variation is calculated in ms as the average of sequential 1 / 16th min pulse interval differences, after exclusion of decelerations and errors. Accelerations are defined as periods with an amplitude of more than 15 beats / min (bpm) above the baseline FHR and lasting for more than 15 s. The system treats signal loss of 30% or more as a high loss, which was also an exclusion criterion for our study. Fetuses that responded abnormally to the stimuli, in that alternation of responses and non-responses were observed, were excluded from further analysis because it was unclear whether habituation was present or not. Data were compared using Wilcoxon matched-pairs signed-ranks test and Mann–Whitney U test, and correlated using Spearman rank correlation coefficients, with a two-tailed P-value of , 0.05 considered significant. 3. Results
3.1. Habituation and the influence of fetal heart rate patterns The data of five fetuses were excluded because they responded irregularly in the first test, which prevented interpretation of the test. These fetuses did not differ from
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
139
the other 32 fetuses in gender, presentation, birth weight, maternal abdominal wall thickness, amniotic fluid volume, gestational age at testing and delivery, and neonatal outcome. In the remaining 32 fetuses in which regular responses could be observed, the median gestational age at time of testing was 38 2 / 7 (range 37 0 / 7 – 39 5 / 7 ) weeks, the median gestational age at delivery 40 3 / 7 (range 38 5 / 7 – 42 3 / 7 ) weeks, the median birth weight 3385 (range 2660–4070) g. Seven mothers were delivered by cesarean section for different reasons, but not for fetal distress. All infants, 14 males and 18 females, were in good health at birth, with birth weights above the 10th percentile for gestational age, a 5-min Apgar score $ 8 and an umbilical artery pH . 7.10. Follow-up after 3 months revealed no serious abnormalities in any infant. Of the 32 fetuses, 26 habituated in the first test. Twenty of them habituated more rapidly or did not respond at all in the second test. In the second test, only two fetuses habituated more slowly and in four fetuses habituation could not be determined as they alternated between response and non-response. Of the 32 fetuses, six failed to habituate within 21 stimuli (four even showed no response decrement) in the first test, three of them also failed to habituate in the second test and three habituated (after 15, 13 and three stimuli, respectively). The median habituation rate decreased significantly in the second test (9.5 vs. 2, P 5 0.001), meaning that fetuses habituated more rapidly during the second test. (Fig. 1) We classified the FHRP of the recordings prior to the first test into FHRP A, corresponding to a quiet state and into an FHRP non-A, corresponding to an active state. Of the remaining 31 recordings (the data of five fetuses were excluded and one recording was uninterpretable owing to high signal loss), FHRP A was determined in 14 FHR recordings, and a FHRP corresponding to an active state in 17. There was no significant difference in the median habituation rate of fetuses that were initially in a quiet state and of fetuses in an active state (Fig. 2).
Fig. 1. Habituation rate in test 1 and test 2 after 10 min (n 5 32). Data are combined in a box and whisker plot (median, 25th and 75th percentiles, maximum and minimum are shown). A habituation rate of 23 represents a persistent response. *P 5 0.001.
140
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
Fig. 2. Habituation rate in fetuses who were initially in a quiet state (n 5 14) and in an active state (n 5 17).
The 10-min FHR recordings after the first test were also visually classified. Of the remaining 31 recordings, FHRP A was observed in only three, and an FHRP corresponding to an active state in 28. Two fetuses remained in a quiet state after the test and one showed a state transition from active to quiet. From the 28 fetuses that showed an FHRP consistent with an active state after the first test, 12 were in a quiet state prior to testing. The median habituation rate of these 12 fetuses did not differ from the habituation rate of the 16 fetuses that remained in an active state after testing.
3.2. Habituation and the influence of baseline FHR, FHR variability and number of accelerations Computer analysis of the FHR was performed in a subgroup of 25 fetuses. It was not possible to determine habituation in four fetuses because of alternate responses and non-responses, so these FHR recordings were excluded from statistical analysis. One recording was uninterpretable as a result of high signal loss. Of the remaining 20 fetuses, only the baseline FHR increased significantly after the first test (P 5 0.001). The long-term variation, short-term variation and number of accelerations did not change significantly after the first test. (Table 1) There is no correlation between the habituation rate of the first test and baseline FHR, long-term variation, short-term variation and number of accelerations of the observation period prior to this test. Neither is there a correlation between the habituation rate of the first test and the difference in baseline FHR between periods prior and after this test, nor to the difference in long-term variation, short-term variation and number of accelerations between the observation periods prior to and after this test (Table 2, Figs. 3 and 4).
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
141
Table 1 FHR parameters before and after test 1 (n 5 20)a
Basal FHR (bpm) LTV (ms) STV (ms) Number of accelerations ( . 15 bpm, . 15 s)
Before test 1
After test 1
P
133 (127.5–137) 50.5 (28–83) 10.1 (8–13.3) 2 (1–3.5)
138 (134.5–147) 59 (40–79.5) 11.2 (9–13.1) 2.5 (0.5–4)
0.001 NS NS NS
a
Data are given in medians and interquartile ranges and compared using Wilcoxon matched-pairs signed-ranks test. LTV, long-term variation; STV, short-term variation; NS, not significant.
Table 2 Correlation coefficients of habituation rate and FHR parameters of observation period before test 1 and of habituation rate and difference between FHR parameters before and after test 1 (n 5 20)a D before and after test 1
Before test 1
Basal FHR (bpm) LTV (ms) STV (ms) Number of accelerations
r
P
r
P
2 0.06 2 0.2 2 0.2 2 0.13
0.8 0.3 0.3 0.6
2 0.2 0.2 0.2 0.14
0.4 0.3 0.3 0.6
a
Data are correlated using Spearman rank correlation coefficient. LTV, long-term variation; STV, short-term variation.
Fig. 3. Habituation rate of test 1 and the baseline FHR prior to test 1 (n 5 20).
142
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
Fig. 4. Habituation rate of test 1 and the difference in baseline FHR before and after test 1 (n 5 20).
4. Discussion In this study we investigated the fetal habituation to repeated stimulation in healthy term fetuses and studied the influence of fetal quiescence and activity, state transitions and FHR parameters on habituation. The type of stimulus we used induced an immediate fetal movement response regardless of the fetal state. By using this stimulator and observing the immediate movement response to repeated stimulation it is easy for a clinician to distinguish between a response and a non-response and to determine fetal habituation. The method we used was not adequate to determine the presence of habituation in all fetuses. Five fetuses (13.5%) responded irregularly to the repeated stimuli. These fetuses exhibited alternate responses and non-responses, which made determination of habituation impossible. The habituation rate of the first test varied greatly in our study, with six fetuses failing to habituate and six fetuses habituating within five stimuli. All fetuses did well after birth. To confirm that the observed response decrement was the result of habituation rather than motor fatigue or sensory adaptation, we repeated the test within 10 min of the first test. Repeated stimulation may physically exhaust the fetus or it may tire the auditory or tactile receptor, which results in receptor adaptation. In true habituation, the original stimulus, when presented again, initially elicits a response but this response decreases more rapidly than when first presented. In our study the habituation rate decreased significantly in the second test. This can be explained by the fact that fetuses recognize the stimuli and habituate more rapidly [23]. Nonetheless, three fetuses continued to respond in the first test as in the second test and two fetuses habituated more slowly in the second test. The behavioural states of neonates and fetuses have been found to affect the response to a single vibroacoustic or sound stimulus. Neonatal responses to various stimuli vary with behavioural state, with the neonate being less responsive during quiet sleep than during an active period [24,25]. Weiner et al. [16] examined the motor and FHR
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
143
response to a door buzzer in term fetuses. The FHR response time and movement response time (time between the stimulus and the onset of the response) was significantly longer during fetal quiescence. In contrast, Jensen [18] described that FHR responses to sound stimulation were mainly observed in fetuses with low baseline FHR levels prior to stimulation, in other words in fetuses in a quiet state. A non-response to this stimulus was observed mainly in fetuses in an active period. If behavioural states affect the response to a single stimulus, one might think that it also influences the habituation of responses to repeated stimuli. Although one study has demonstrated that fetal states did not significantly influence the habituation pattern to vibroacoustic stimulation [15], another study has shown that the decrement in response is caused by a behavioural state transition rather than by true habituation [19]. Groome et al. [17] observed the movement response magnitude to a fixed number of eight vibroacoustic stimuli in fetuses between 34 and 40 weeks gestational age. They showed that decrement of the movement response was not influenced by the variability of the FHR prior to testing. Shalev et al. [15] also examined the influence of fetal states, according to FHR pattern and fetal movements, on habituation to vibroacoustic stimuli. They found that fetuses in a quiet state habituated more rapidly (mean rate of five stimuli) than fetuses in an active state (mean rate of eight stimuli). We found that the median habituation rate of fetuses that were initially in an active state, according to visual analysis of the FHRP, did not differ from that of fetuses in a quiet state. Our findings that the baseline FHR, long-term variation, short-term variation and accelerations prior to the habituation test do not correlate with the habituation rate, support the conclusion that fetal habituation is not affected by the FHRP prior to testing. Repeated stimulation alters the behavioural state in term fetuses. This could suggest that fetal habituation is induced by a state transition from a quiet to an active state [17,19]. Hutt et al. [19] were unable to demonstrate habituation of the movement response to sound stimulation in 10 neonates. Response decrement was observed only when there was a state transition from an active to a quiet state, whereas transition from a quiet to an active state was accompanied by a recovery of response. This suggests that response decrement might not be the result of habituation but rather of a state transition. Although almost all fetuses showed an active FHRP after the first test, our results show that the habituation rate varied greatly. We cannot conclude that a state transition from an active to quiet or from a quiet to active state induced by repeated stimulation influenced the habituation rate. The difference between FHR parameters before and after the first test did not correlate with the habituation rate. From these results we conclude that a change in FHRP and FHR parameters resulting from stimulation did not affect the habituation rate. Although several methodologies have been used, all studies of normal fetuses conclude that fetuses are able to habituate to repeated stimulation [7–11]. Studies in compromised or chromosomally abnormal fetuses reveal that the habituation pattern of these fetuses is different from that of normal fetuses [12–14]. For example, fetuses with Down’s syndrome habituated more slowly than unaffected fetuses [14]. Leader et al. studied fetal habituation to stimuli produced by an electric toothbrush in compromised fetuses (small for gestational age, meconium-stained amniotic fluid
144
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
or decreased growth velocity of the fetal biparietal diameter) [13]. They found that the compromised fetuses habituated more rapidly (requiring 1–9 stimuli) or more slowly ( . 50 stimuli) compared with normal fetuses. In our study of normal fetuses, the habituation rate in the first test varied greatly. Six fetuses failed to habituate within 21 stimuli and six habituated within five stimuli. Although we followed another protocol with a different stimulator, and a maximum of 21 stimulus applications, we cannot conclude that fetuses that failed to habituate or habituated more rapidly showed fetal distress or adverse outcome. Therefore, there are apparently normal fetuses that fail to habituate to repeated stimulation or that habituate more rapidly than other fetuses. There are even apparently normal fetuses in which it is impossible to determine habituation due to irregular responses. From this study we conclude that the habituation rate varies greatly in healthy term fetuses. Given the variability of the habituation rate and the small difference between the habituation rate in fetuses in a quiet state and fetuses in an active state, the influence of fetal states and FHR can be ignored. And although habituation can be clearly observed in the majority of term fetuses, the interfetal variability is such that identification of an abnormal fetus will be very difficult. This means that the clinical relevance of fetal habituation has still to be elucidated. Therefore, further studies to investigate whether or not fetal habituation can be used for the assessment of fetal condition and for the assessment of fetal CNS in high-risk fetuses are in progress.
Acknowledgements This study was financially supported by the ZorgOnderzoek Netherlands (grant 28-3054) The Hague, The Netherlands and the Dutch Brain Foundation, The Hague, The Netherlands.
References [1] Jeffrey WE, Cohen LS. Habituation in the human infant. Adv Child Dev Behav 1971;6:63–97. [2] Gandhavadi B. Glabellar reflex habituation in mentally retarded adults. J Ment Defic Res 1982;26:271–8. [3] Gandhavadi B, Melvin JL. Electrical blink reflex habituation in mentally retarded adults. J Ment Defic Res 1985;29:49–54. [4] Hutt SJ, Hutt C. Hyperactivity in a group of epileptic (and some non-epileptic) brain-damaged children. Epilepsia 1964;5:334–51. [5] Lader MH, Wing L. Physiological measures in agitated and retarded depressed patients. J Psychiatr Res 1969;7:89–100. [6] Schafer EW, Peeke HV. Down syndrome individuals fail to habituate cortical evoked potentials. Am J Ment Defic 1982;87:332–7. [7] Madison LS, Madison JK, Adubato SA. Infant behavior and development in relation to fetal movement and habituation. Child Dev 1986;57:1475–82. [8] Groome LJ, Gotlieb SJ, Neely CL, Waters MD. Developmental trends in fetal habituation to vibroacoustic stimulation. Am J Perinatol 1993;10:46–9.
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
145
[9] Shalev E, Benett MJ, Megory E, Wallace RM, Zuckerman H. Fetal habituation to repeated sound stimulation. Isr J Med Sci 1989;25:77–80. [10] Leader LR, Baillie P, Martin B, Vermeulen E. The assessment and significance of habituation to a repeated stimulus by the human fetus. Early Hum Dev 1982;7:211–9. [11] Kuhlman KA, Burns KA, Depp R, Sabbagha RE. Ultrasonic imaging of normal fetal response to external vibratory acoustic stimulation. Am J Obstet Gynecol 1988;158:47–51. [12] Leader LR, Baillie P. The changes in fetal habituation patterns due to a decrease in inspired maternal oxygen. Br J Obstet Gynaecol 1988;95:664–8. [13] Leader LR, Baillie P, Martin B, Vermeulen E. Fetal habituation in high-risk pregnancies. Br J Obstet Gynaecol 1982;89:441–6. [14] Hepper PG, Shahidullah S. Habituation in normal and Down’s syndrome fetuses. Q J Exp Psychol B 1992;44:305–17. [15] Shalev E, Weiner E, Serr DM. Fetal habituation to sound stimulus in various behavioral states. Gynecol Obstet Invest 1990;29:115–7. [16] Weiner E, Serr DM, Shalev E. Fetal motorical and heart response to sound stimulus in different behavioral states. Gynecol Obstet Invest 1989;28:141–3. [17] Groome LJ, Singh KP, Burgard SL, Neely CL, Deason MA. Motor responsivity during habituation testing of normal human fetuses. J Perinat Med 1995;23:159–66. [18] Jensen OH. Fetal heart rate response to controlled sound stimuli during the third trimester of normal pregnancy. Acta Obstet Gynecol Scand 1984;63:193–7. [19] Hutt C, Bernuth H, Lenard HG, Hutt SJ, Prechtl HF. Habituation in relation to state in the human neonate. Nature 1968;220:618–20. [20] Nijhuis JG, Prechtl HF, Martin Jr. CB, Bots RS. Are there behavioural states in the human fetus? Early Hum Dev 1982;6:177–95. [21] Pillai M, James D. The development of fetal heart rate patterns during normal pregnancy. Obstet Gynecol 1990;76:812–6. [22] Dawes GS, Lobb M, Moulden M, Redman CW, Wheeler T. Antenatal cardiotocogram quality and interpretation using computers. Br J Obstet Gynaecol 1992;99:791–7. [23] Heteren van CF, Boekkooi PF, Jongsma HW, Nijhuis JG. Fetal learning and memory. Lancet 2000;356:1169–70. [24] Hutt SJ, Hutt C, Lenard HG, Bernuth H, Muntjewerff WJ. Auditory responsivity in the human neonate. Nature 1968;218:888–90. [25] Prechtl HF. The behavioural states of the newborn infant (a review). Brain Res 1974;76:185–212.
Fetal habituation to vibroacoustic stimulation in relation to fetal states and fetal heart rate parameters a,
b
Cathelijne F. van Heteren *, P. Focco Boekkooi , Henk W. Jongsma a , Jan G. Nijhuis c a
Department of Obstetrics and Gynaecology, University Medical Centre Nijmegen, PO Box 9101, 6500 HB Nijmegen, The Netherlands b Department of Obstetrics and Gynaecology, St Elisabeth Hospital, Tilburg, The Netherlands c Department of Obstetrics and Gynaecology, University Hospital Maastricht, Maastricht, The Netherlands
Received 18 August 2000; received in revised form 20 November 2000; accepted 21 November 2000
Abstract Objectives: Fetal habituation to repeated stimulation has the potential to become a tool in the assessment of fetal condition and of the function of the fetal central nervous system (CNS). However, the influence of fetal quiescence and activity on habituation remains to be clarified. We studied habituation and the influence of fetal state and fetal heart rate (FHR) parameters on habituation in healthy term fetuses. Subjects and method: We studied habituation in 37 healthy fetuses in two tests with an interval of 10 min. The vibroacoustic stimuli were applied to the maternal abdomen above the fetal legs for a period of 1 s every 30 s. A fetal trunk movement within 1 s after stimulation was defined as a positive response. Habituation rate is defined as the number of stimuli applied before an observed non-response to four consecutive stimuli. The FHR patterns (FHRP) of the 10 min observation period before and after the tests were visually classified. Fetal states were defined according to the FHRP. Baseline FHR, FHR variability and the number of accelerations were calculated in a subgroup of 25 fetuses. Results: Of the 32 fetuses that responded normally during the first test, 26 habituated and six had persistent
*Corresponding author. Tel.: 1 31-243-616-801; fax: 1 31-243-541-194. E-mail address: [email protected] (C.F. van Heteren). 0378-3782 / 01 / $ – see front matter 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S0378-3782( 00 )00130-4
136
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
responses. The median habituation rate decreased significantly in the second test (P 5 0.001). There was no difference in habituation rate between fetuses that where initially in a quiet state and those in an active state. The FHR parameters before the first test and the difference between these FHR parameters before and after the test did not correlate with the habituation rate. Conclusions: Although the majority of healthy fetuses was able to habituate, the interfetal variability in habituation performance is such that testing of habituation seems not to be a sensitive tool for the assessment of the fetal CNS. This variability is neither the result of differences in fetal state nor of the various FHR parameters before testing, nor of the difference in change of FHR parameters arising from stimulation. 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Fetal habituation; Vibroacoustic stimulation; Fetal behavioural states; Fetal heart rate
1. Introduction Habituation is the decrease in, and ultimate cessation of, response to repeated stimulation. Habituation can be seen as a basic form of learning that requires an intact central nervous system (CNS) [1]. If habituation represents CNS functioning, then conditions affecting the CNS would be expected to affect habituation. This has been reported in studies on adults with various disorders of the CNS [2–6]. Several studies on habituation in normal fetuses have shown that fetuses are able to habituate to repeated stimuli [7–11]. Studies in compromised or chromosomally abnormal fetuses show that the habituation pattern of these fetuses is different from that of normal fetuses [12–14]. Therefore, fetal habituation might identify a fetus with poor fetal condition or impaired CNS functioning. Although fetal habituation has been studied widely [10,13–15], its clinical relevance remains unclear. This can be explained by the variety of methodologies that have been used to test fetal habituation — there still exists no standardised test. Neither is it clear whether an abnormal habituation pattern is the result of fetal distress, impaired functioning of the CNS, or of a difference in environmental or fetal physiological factors, such as fetal behavioural states. Behavioural states have been found to affect the magnitude of the fetal response and the time between the stimulus and the response [16–18]. Although one study has demonstrated that behavioural states did not significantly influence the habituation pattern to vibroacoustic stimulation [15], another study has shown that the decrement in response is caused by a behavioural state transition rather than by true habituation [19]. It remains unclear whether, and to what extent, fetal behavioural states influence habituation. Because of the potential clinical relevance of fetal habituation, it is essential to know whether fetal behavioural states and state transition affect habituation. This study was designed to test fetal habituation in healthy term fetuses and to study the influence of fetal states (fetal quiescence and activity), state transition and FHR parameters on habituation.
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
137
2. Subjects and method
2.1. Subjects Thirty-seven healthy pregnant women were recruited from a low-risk obstetric population (midwives’ practice) at the University Medical Centre Nijmegen in the Netherlands. The study was approved by the hospital’s ethical committee and all participants gave their written informed consent. Inclusion criteria were: (1) gestational age between 37 and 40 completed weeks (pregnancies were dated using the last menstrual period or by ultrasound when dates were uncertain); (2) the absence of maternal medical or obstetric complications; (3) single fetus without apparent structural anomalies and with an expected birth weight above the 10th percentile; (4) normal amniotic fluid volume as assessed by ultrasound; (5) no maternal use of alcohol, drugs or medication other than vitamins and / or iron.
2.2. Study design All habituation tests were done under the same conditions (the mother was not allowed to smoke, drink coffee or eat for 3 h before testing), in the same room, and by the same examiner between 4.00 p.m. and 7.00 p.m. The women were placed in a semi-recumbent position. The stimuli were produced by a fetal vibroacoustic stimulator (Corometrics model 146, Wallingford, CT, USA; audible sound 20 to 9000 Hz, vibrations between 67 and 83 Hz, sound level 74 dB at 1 m in air). Stimuli of 1 s duration were repeatedly applied to the maternal abdomen above the fetal legs every 30 s. The fetal trunk was displayed in a parasagittal plane by a real-time ultrasound scanner (Hitachi model EUB-525, Tokyo, Japan). The timing of the stimuli was chosen to occur when the fetus was not moving. A general movement of the fetal trunk within 1 s of application of the stimulus was defined as a positive response. Response decrement was noted as a changing from a more intense to a less intense response pattern with successive trials. A lack of response to four consecutive stimuli indicated habituation. We allowed a maximum number of 24 stimulus applications in each test. However, a minimum of four additional stimuli would be necessary to show habituation if the fetus responded to the 21st stimulus. We therefore stopped stimulating if a fetus still responded to the 21st stimulus. The habituation rate was defined as the number of stimuli applied before a fetus stopped responding. The procedure was repeated 10 min after the first test to investigate whether the fetus habituated more rapidly during the second series of stimuli. The FHR was recorded continuously with a cardiotocograph (Sonicaid FM7, Oxford, UK) at a paper speed of 3 cm / min, starting 10 min prior to the tests and continuing for 10 min afterwards. The outcome of each pregnancy was examined for birth weight, gestational age at delivery, Apgar scores, umbilical artery blood pH, and the presence of neonatal complications. The infants were followed up 3 months after birth by telephone interviews with one of the parents.
138
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
2.3. Data analysis In 37 patients, the FHR recordings of the 10 min observation periods before and after the first test were visually classified into four FHR patterns (FHRP) in accordance with Nijhuis et al. [20]. FHRP A is defined as a stable heart rate with little variability and isolated accelerations. FHRP B has a wider oscillation bandwidth than FHRP A with frequent accelerations. FHRP C is a stable pattern with a wider oscillation bandwidth than FHRP A and without accelerations. FHRP D is unstable, with large and long-lasting accelerations, which are frequently fused into a sustained tachycardia. As the incidence of FHRP C and D is relatively low, these episodes were considered to be an FHRP corresponding to an active state. Finally, the FHR recordings were divided into an FHRP corresponding to a quiet state (FHRP A) and into an FHRP corresponding to an active state (FHRP B, C and D). In our study, determination of whether the fetus was in a quiet state or an active state was based on analysis of FHRP alone, because such a distinction is possible in normal fetuses after 35 weeks gestation [21]. Computerized analysis of the FHR of a subgroup of 25 fetuses was carried out using a Sonicaid System 8002 (Oxford Sonicaid Ltd, Abingdon, UK) [22]. Baseline FHR, FHR variability (long-term variation and short-term variation) and the number of accelerations were calculated for the observation periods before and after the first test. Post-acquisition analysis of any fraction of the recording down to a minimum of 10 min is possible. The system reduces the data over 3.75 s (1 / 16 min epochs). The baseline FHR is defined as the mean rate averaged over all periods of low variation. If no low variation is present, it is derived from a statistical analysis. A baseline is fitted to the FHR trace and long-term variation is subsequently measured as the range of mean pulse intervals around the baseline, calculated minute by minute, excluding data from decelerations and errors. The system averages data over the total recording time in milliseconds (ms). Short-term variation is calculated in ms as the average of sequential 1 / 16th min pulse interval differences, after exclusion of decelerations and errors. Accelerations are defined as periods with an amplitude of more than 15 beats / min (bpm) above the baseline FHR and lasting for more than 15 s. The system treats signal loss of 30% or more as a high loss, which was also an exclusion criterion for our study. Fetuses that responded abnormally to the stimuli, in that alternation of responses and non-responses were observed, were excluded from further analysis because it was unclear whether habituation was present or not. Data were compared using Wilcoxon matched-pairs signed-ranks test and Mann–Whitney U test, and correlated using Spearman rank correlation coefficients, with a two-tailed P-value of , 0.05 considered significant. 3. Results
3.1. Habituation and the influence of fetal heart rate patterns The data of five fetuses were excluded because they responded irregularly in the first test, which prevented interpretation of the test. These fetuses did not differ from
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
139
the other 32 fetuses in gender, presentation, birth weight, maternal abdominal wall thickness, amniotic fluid volume, gestational age at testing and delivery, and neonatal outcome. In the remaining 32 fetuses in which regular responses could be observed, the median gestational age at time of testing was 38 2 / 7 (range 37 0 / 7 – 39 5 / 7 ) weeks, the median gestational age at delivery 40 3 / 7 (range 38 5 / 7 – 42 3 / 7 ) weeks, the median birth weight 3385 (range 2660–4070) g. Seven mothers were delivered by cesarean section for different reasons, but not for fetal distress. All infants, 14 males and 18 females, were in good health at birth, with birth weights above the 10th percentile for gestational age, a 5-min Apgar score $ 8 and an umbilical artery pH . 7.10. Follow-up after 3 months revealed no serious abnormalities in any infant. Of the 32 fetuses, 26 habituated in the first test. Twenty of them habituated more rapidly or did not respond at all in the second test. In the second test, only two fetuses habituated more slowly and in four fetuses habituation could not be determined as they alternated between response and non-response. Of the 32 fetuses, six failed to habituate within 21 stimuli (four even showed no response decrement) in the first test, three of them also failed to habituate in the second test and three habituated (after 15, 13 and three stimuli, respectively). The median habituation rate decreased significantly in the second test (9.5 vs. 2, P 5 0.001), meaning that fetuses habituated more rapidly during the second test. (Fig. 1) We classified the FHRP of the recordings prior to the first test into FHRP A, corresponding to a quiet state and into an FHRP non-A, corresponding to an active state. Of the remaining 31 recordings (the data of five fetuses were excluded and one recording was uninterpretable owing to high signal loss), FHRP A was determined in 14 FHR recordings, and a FHRP corresponding to an active state in 17. There was no significant difference in the median habituation rate of fetuses that were initially in a quiet state and of fetuses in an active state (Fig. 2).
Fig. 1. Habituation rate in test 1 and test 2 after 10 min (n 5 32). Data are combined in a box and whisker plot (median, 25th and 75th percentiles, maximum and minimum are shown). A habituation rate of 23 represents a persistent response. *P 5 0.001.
140
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
Fig. 2. Habituation rate in fetuses who were initially in a quiet state (n 5 14) and in an active state (n 5 17).
The 10-min FHR recordings after the first test were also visually classified. Of the remaining 31 recordings, FHRP A was observed in only three, and an FHRP corresponding to an active state in 28. Two fetuses remained in a quiet state after the test and one showed a state transition from active to quiet. From the 28 fetuses that showed an FHRP consistent with an active state after the first test, 12 were in a quiet state prior to testing. The median habituation rate of these 12 fetuses did not differ from the habituation rate of the 16 fetuses that remained in an active state after testing.
3.2. Habituation and the influence of baseline FHR, FHR variability and number of accelerations Computer analysis of the FHR was performed in a subgroup of 25 fetuses. It was not possible to determine habituation in four fetuses because of alternate responses and non-responses, so these FHR recordings were excluded from statistical analysis. One recording was uninterpretable as a result of high signal loss. Of the remaining 20 fetuses, only the baseline FHR increased significantly after the first test (P 5 0.001). The long-term variation, short-term variation and number of accelerations did not change significantly after the first test. (Table 1) There is no correlation between the habituation rate of the first test and baseline FHR, long-term variation, short-term variation and number of accelerations of the observation period prior to this test. Neither is there a correlation between the habituation rate of the first test and the difference in baseline FHR between periods prior and after this test, nor to the difference in long-term variation, short-term variation and number of accelerations between the observation periods prior to and after this test (Table 2, Figs. 3 and 4).
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
141
Table 1 FHR parameters before and after test 1 (n 5 20)a
Basal FHR (bpm) LTV (ms) STV (ms) Number of accelerations ( . 15 bpm, . 15 s)
Before test 1
After test 1
P
133 (127.5–137) 50.5 (28–83) 10.1 (8–13.3) 2 (1–3.5)
138 (134.5–147) 59 (40–79.5) 11.2 (9–13.1) 2.5 (0.5–4)
0.001 NS NS NS
a
Data are given in medians and interquartile ranges and compared using Wilcoxon matched-pairs signed-ranks test. LTV, long-term variation; STV, short-term variation; NS, not significant.
Table 2 Correlation coefficients of habituation rate and FHR parameters of observation period before test 1 and of habituation rate and difference between FHR parameters before and after test 1 (n 5 20)a D before and after test 1
Before test 1
Basal FHR (bpm) LTV (ms) STV (ms) Number of accelerations
r
P
r
P
2 0.06 2 0.2 2 0.2 2 0.13
0.8 0.3 0.3 0.6
2 0.2 0.2 0.2 0.14
0.4 0.3 0.3 0.6
a
Data are correlated using Spearman rank correlation coefficient. LTV, long-term variation; STV, short-term variation.
Fig. 3. Habituation rate of test 1 and the baseline FHR prior to test 1 (n 5 20).
142
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
Fig. 4. Habituation rate of test 1 and the difference in baseline FHR before and after test 1 (n 5 20).
4. Discussion In this study we investigated the fetal habituation to repeated stimulation in healthy term fetuses and studied the influence of fetal quiescence and activity, state transitions and FHR parameters on habituation. The type of stimulus we used induced an immediate fetal movement response regardless of the fetal state. By using this stimulator and observing the immediate movement response to repeated stimulation it is easy for a clinician to distinguish between a response and a non-response and to determine fetal habituation. The method we used was not adequate to determine the presence of habituation in all fetuses. Five fetuses (13.5%) responded irregularly to the repeated stimuli. These fetuses exhibited alternate responses and non-responses, which made determination of habituation impossible. The habituation rate of the first test varied greatly in our study, with six fetuses failing to habituate and six fetuses habituating within five stimuli. All fetuses did well after birth. To confirm that the observed response decrement was the result of habituation rather than motor fatigue or sensory adaptation, we repeated the test within 10 min of the first test. Repeated stimulation may physically exhaust the fetus or it may tire the auditory or tactile receptor, which results in receptor adaptation. In true habituation, the original stimulus, when presented again, initially elicits a response but this response decreases more rapidly than when first presented. In our study the habituation rate decreased significantly in the second test. This can be explained by the fact that fetuses recognize the stimuli and habituate more rapidly [23]. Nonetheless, three fetuses continued to respond in the first test as in the second test and two fetuses habituated more slowly in the second test. The behavioural states of neonates and fetuses have been found to affect the response to a single vibroacoustic or sound stimulus. Neonatal responses to various stimuli vary with behavioural state, with the neonate being less responsive during quiet sleep than during an active period [24,25]. Weiner et al. [16] examined the motor and FHR
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
143
response to a door buzzer in term fetuses. The FHR response time and movement response time (time between the stimulus and the onset of the response) was significantly longer during fetal quiescence. In contrast, Jensen [18] described that FHR responses to sound stimulation were mainly observed in fetuses with low baseline FHR levels prior to stimulation, in other words in fetuses in a quiet state. A non-response to this stimulus was observed mainly in fetuses in an active period. If behavioural states affect the response to a single stimulus, one might think that it also influences the habituation of responses to repeated stimuli. Although one study has demonstrated that fetal states did not significantly influence the habituation pattern to vibroacoustic stimulation [15], another study has shown that the decrement in response is caused by a behavioural state transition rather than by true habituation [19]. Groome et al. [17] observed the movement response magnitude to a fixed number of eight vibroacoustic stimuli in fetuses between 34 and 40 weeks gestational age. They showed that decrement of the movement response was not influenced by the variability of the FHR prior to testing. Shalev et al. [15] also examined the influence of fetal states, according to FHR pattern and fetal movements, on habituation to vibroacoustic stimuli. They found that fetuses in a quiet state habituated more rapidly (mean rate of five stimuli) than fetuses in an active state (mean rate of eight stimuli). We found that the median habituation rate of fetuses that were initially in an active state, according to visual analysis of the FHRP, did not differ from that of fetuses in a quiet state. Our findings that the baseline FHR, long-term variation, short-term variation and accelerations prior to the habituation test do not correlate with the habituation rate, support the conclusion that fetal habituation is not affected by the FHRP prior to testing. Repeated stimulation alters the behavioural state in term fetuses. This could suggest that fetal habituation is induced by a state transition from a quiet to an active state [17,19]. Hutt et al. [19] were unable to demonstrate habituation of the movement response to sound stimulation in 10 neonates. Response decrement was observed only when there was a state transition from an active to a quiet state, whereas transition from a quiet to an active state was accompanied by a recovery of response. This suggests that response decrement might not be the result of habituation but rather of a state transition. Although almost all fetuses showed an active FHRP after the first test, our results show that the habituation rate varied greatly. We cannot conclude that a state transition from an active to quiet or from a quiet to active state induced by repeated stimulation influenced the habituation rate. The difference between FHR parameters before and after the first test did not correlate with the habituation rate. From these results we conclude that a change in FHRP and FHR parameters resulting from stimulation did not affect the habituation rate. Although several methodologies have been used, all studies of normal fetuses conclude that fetuses are able to habituate to repeated stimulation [7–11]. Studies in compromised or chromosomally abnormal fetuses reveal that the habituation pattern of these fetuses is different from that of normal fetuses [12–14]. For example, fetuses with Down’s syndrome habituated more slowly than unaffected fetuses [14]. Leader et al. studied fetal habituation to stimuli produced by an electric toothbrush in compromised fetuses (small for gestational age, meconium-stained amniotic fluid
144
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
or decreased growth velocity of the fetal biparietal diameter) [13]. They found that the compromised fetuses habituated more rapidly (requiring 1–9 stimuli) or more slowly ( . 50 stimuli) compared with normal fetuses. In our study of normal fetuses, the habituation rate in the first test varied greatly. Six fetuses failed to habituate within 21 stimuli and six habituated within five stimuli. Although we followed another protocol with a different stimulator, and a maximum of 21 stimulus applications, we cannot conclude that fetuses that failed to habituate or habituated more rapidly showed fetal distress or adverse outcome. Therefore, there are apparently normal fetuses that fail to habituate to repeated stimulation or that habituate more rapidly than other fetuses. There are even apparently normal fetuses in which it is impossible to determine habituation due to irregular responses. From this study we conclude that the habituation rate varies greatly in healthy term fetuses. Given the variability of the habituation rate and the small difference between the habituation rate in fetuses in a quiet state and fetuses in an active state, the influence of fetal states and FHR can be ignored. And although habituation can be clearly observed in the majority of term fetuses, the interfetal variability is such that identification of an abnormal fetus will be very difficult. This means that the clinical relevance of fetal habituation has still to be elucidated. Therefore, further studies to investigate whether or not fetal habituation can be used for the assessment of fetal condition and for the assessment of fetal CNS in high-risk fetuses are in progress.
Acknowledgements This study was financially supported by the ZorgOnderzoek Netherlands (grant 28-3054) The Hague, The Netherlands and the Dutch Brain Foundation, The Hague, The Netherlands.
References [1] Jeffrey WE, Cohen LS. Habituation in the human infant. Adv Child Dev Behav 1971;6:63–97. [2] Gandhavadi B. Glabellar reflex habituation in mentally retarded adults. J Ment Defic Res 1982;26:271–8. [3] Gandhavadi B, Melvin JL. Electrical blink reflex habituation in mentally retarded adults. J Ment Defic Res 1985;29:49–54. [4] Hutt SJ, Hutt C. Hyperactivity in a group of epileptic (and some non-epileptic) brain-damaged children. Epilepsia 1964;5:334–51. [5] Lader MH, Wing L. Physiological measures in agitated and retarded depressed patients. J Psychiatr Res 1969;7:89–100. [6] Schafer EW, Peeke HV. Down syndrome individuals fail to habituate cortical evoked potentials. Am J Ment Defic 1982;87:332–7. [7] Madison LS, Madison JK, Adubato SA. Infant behavior and development in relation to fetal movement and habituation. Child Dev 1986;57:1475–82. [8] Groome LJ, Gotlieb SJ, Neely CL, Waters MD. Developmental trends in fetal habituation to vibroacoustic stimulation. Am J Perinatol 1993;10:46–9.
C.F. van Heteren et al. / Early Human Development 61 (2001) 135 – 145
145
[9] Shalev E, Benett MJ, Megory E, Wallace RM, Zuckerman H. Fetal habituation to repeated sound stimulation. Isr J Med Sci 1989;25:77–80. [10] Leader LR, Baillie P, Martin B, Vermeulen E. The assessment and significance of habituation to a repeated stimulus by the human fetus. Early Hum Dev 1982;7:211–9. [11] Kuhlman KA, Burns KA, Depp R, Sabbagha RE. Ultrasonic imaging of normal fetal response to external vibratory acoustic stimulation. Am J Obstet Gynecol 1988;158:47–51. [12] Leader LR, Baillie P. The changes in fetal habituation patterns due to a decrease in inspired maternal oxygen. Br J Obstet Gynaecol 1988;95:664–8. [13] Leader LR, Baillie P, Martin B, Vermeulen E. Fetal habituation in high-risk pregnancies. Br J Obstet Gynaecol 1982;89:441–6. [14] Hepper PG, Shahidullah S. Habituation in normal and Down’s syndrome fetuses. Q J Exp Psychol B 1992;44:305–17. [15] Shalev E, Weiner E, Serr DM. Fetal habituation to sound stimulus in various behavioral states. Gynecol Obstet Invest 1990;29:115–7. [16] Weiner E, Serr DM, Shalev E. Fetal motorical and heart response to sound stimulus in different behavioral states. Gynecol Obstet Invest 1989;28:141–3. [17] Groome LJ, Singh KP, Burgard SL, Neely CL, Deason MA. Motor responsivity during habituation testing of normal human fetuses. J Perinat Med 1995;23:159–66. [18] Jensen OH. Fetal heart rate response to controlled sound stimuli during the third trimester of normal pregnancy. Acta Obstet Gynecol Scand 1984;63:193–7. [19] Hutt C, Bernuth H, Lenard HG, Hutt SJ, Prechtl HF. Habituation in relation to state in the human neonate. Nature 1968;220:618–20. [20] Nijhuis JG, Prechtl HF, Martin Jr. CB, Bots RS. Are there behavioural states in the human fetus? Early Hum Dev 1982;6:177–95. [21] Pillai M, James D. The development of fetal heart rate patterns during normal pregnancy. Obstet Gynecol 1990;76:812–6. [22] Dawes GS, Lobb M, Moulden M, Redman CW, Wheeler T. Antenatal cardiotocogram quality and interpretation using computers. Br J Obstet Gynaecol 1992;99:791–7. [23] Heteren van CF, Boekkooi PF, Jongsma HW, Nijhuis JG. Fetal learning and memory. Lancet 2000;356:1169–70. [24] Hutt SJ, Hutt C, Lenard HG, Bernuth H, Muntjewerff WJ. Auditory responsivity in the human neonate. Nature 1968;218:888–90. [25] Prechtl HF. The behavioural states of the newborn infant (a review). Brain Res 1974;76:185–212.