Effects of sex steroids on the positive estrogen feedback mechanism in intact women and castrate men.

We studied the effects of prolonged testosterone treatment on ovulatory function and positive estrogen feedback in women and of prolonged estrogen priming on gonadotropin feedback in castrate men. An estrogen provocation test was carried out in 4 groups of transsexual subjects: 12 female transsexuals in their early follicular phase (days 3-5; group 1A), 8 females who had been treated with Depo-testosterone (T) for 3-6 months (group 1B), 11 men who had been castrated 3 months previously (group 2A), and 4 male castrates treated with oral estrogen for 3 months starting 3 months after castration (group 2B). The estrogen provocation test consisted of 3 GnRH tests (100 micrograms) carried out immediately before (0 h) and 44 and 92 h after an im injection of estradiol valerate (10 mg). Responses to the estrogen provocation test in women with normal menstrual cycles (group 1A) were typically female. After initial suppression at 44 h, a LH surge (positive feedback) occurred at 92 h. Pituitary responsiveness, however, was amplified both at 44 and 92 h. Prolonged T priming of women in group 1B did not inhibit the estrogen-induced LH surge, nor was the amplitude of the surge blunted. Removal of androgens and other testicular factors (group 2A) did not result in the appearance of an estrogen-induced LH surge. On the other hand, prolonged estrogen priming in male castrates (group 2B) resulted in activation of the positive feedback mechanism; a LH surge in response to the estrogen provocation occurred. The results of the present study imply that 1) contrary to an earlier suggestion, testosterone does not block or blunt the LH surge, indicating that it is probably not responsible for suppressing the LH surge in normal men; 2) testosterone can cause ovulatory failure without suppressing the LH surge in women; and 3) prolonged estrogen priming may be involved in activation of the positive feedback mechanism in humans.

Differential regulation of cholesterol side-chain cleavage (P450scc) and aromatase (P450arom) enzyme mRNA expression by gonadotrophins and cyclic AMP in human granulosa cells.

The co-ordinated biosynthesis of progesterone and oestradiol in the human ovary is critical for reproductive cyclicity and eventual pregnancy. The crucial regulatory enzymes for progesterone and oestradiol biosynthesis in granulosa cells are the cholesterol side-chain cleavage (P450scc) and aromatase (P450arom) enzymes respectively. We utilized the cDNA sequences encoding P450arom and P450scc to examine the roles of FSH and LH, and their intracellular second messenger, cyclic AMP (cAMP), in regulating steroidogenic gene expression. Mature granulosa cells (aspirated before the onset of the endogenous LH surge) and granulosa lutein cells (obtained after an ovulatory dose of human chorionic gonadotrophin) were cultured for 4 days with FSH, LH or dibutyryl cAMP (dbcAMP). After the period of culture, total RNA was extracted from granulosa cells and Northern analyses were performed utilizing 32P-labelled cDNAs encoding P450arom and P450scc. Spent culture media were analysed for steroid and cAMP content. Both FSH and LH strongly stimulated P450arom mRNA expression and oestradiol production in mature granulosa cells. On the other hand, P450scc mRNA expression and progesterone biosynthesis were weakly induced by FSH; maximal synthesis occurred only in the presence of LH. With both gonadotrophins at equivalent concentrations, LH generated a 30-fold higher level of cAMP than FSH. Furthermore, the differential effects of FSH and LH on P450 mRNA expression were reproduced by the presence of low and high concentrations of dbcAMP respectively. LH (and high levels of dbcAMP) increased P450arom mRNA expression in mature granulosa cells but inhibited its accumulation in granulosa lutein cells. In contrast, it stimulated P450scc mRNA expression and progesterone synthesis in both mature granulosa and granulosa lutein cells. Therefore, FSH/low cAMP levels stimulated P450arom gene expression and oestradiol production, while LH/high cAMP levels maximally induced P450scc gene expression and function, in a development-related manner consistent with steroid production in vivo. These findings support the hypothesis that one set of genes (like P450arom) in human granulosa cells is regulated by FSH/low cAMP levels and another (like P450scc) by LH/high cAMP levels.

Cord plasma alpha-fetoprotein values and neonatal jaundice.

Umbilical cord plasma alpha-fetoprotein (AFP) values were determined in 127 infants with hyperbilirubinemia (56 glucose-6-phosphate dehydrogenase (G-6-PD) deficient and 71 G-6-PD normal) and 136 control subjects (73 G-6-PD deficient and 63 G-6-PD normal). The mean alpha-fetoprotein value of 173 +/- 35.2 (SD) mg/L for the group of infants with hyperbilirubinemia was significantly greater than that (122 +/- 21.7 mg/L) for the control infants (P less than .001). G-6-PD status and sex did not significantly affect the alpha-fetoprotein values. Using an alpha-fetoprotein level of 130 mg/L as a "cut-off" value, the incidence of false-positive results was 25.5% and the incidence of false-negative results was 11.8%. This test can be used as a screening procedure to detect infants at high risk for hyperbilirubinemia.

Plasminogen activators t-PA, u-PA and its inhibitor (PAI) in normal males and females.

We determined the plasma levels of tissue plasminogen activator (t-PA), plasminogen activator inhibitor (PAI) activity and their antigen levels including urokinase plasminogen activator (u-PA) in 33 male and 27 female normal subjects. Males had mean t-PA activity of 0.50 iu/ml which was significantly lower (p less than 0.01) than the females 0.64 iu/ml. Males had higher (p less than 0.001) mean PAI activity (15.5 AU/ml) as compared to females 10.3 AU/ml. The respective mean levels of t-PA and PAI antigen were significantly higher (p less than 0.01) in males (8.1 ng/ml and 17.6 ng/ml) than in females (6.2 ng/ml and 12.1 ng/ml). The mean u-PA level in males was 1.54 ng/ml which was significantly higher (p less than 0.01) than in females with 1.02 ng/ml. In post-venous occlusion studies, females had a greater mean response of 8.6 fold in t-PA activity as compared to males with a mean of 4.5 fold increase. The mean t-PA antigen response in males was 2.0 fold increase as compared to 2.6 fold increase in the females. No significant responses were seen in both sexes in either PAI activity or antigen levels when compared with the resting state. In zymography studies, free t-PA, its inhibitor complexes and u-PA were demonstrated in the euglobulin fractions of stored plasma. This study demonstrates that significant differences in t-PA, u-PA and PAI exist between male and female subjects which should be taken into account when determining their levels in clinical conditions.

Quantitative comparison of prolactin production by decidua compacta and spongiosa from the first trimester elective termination of pregnancy.

Prolactin (PRL) producing capacity was studied in explants of decidua compacta and decidua spongiosa obtained from 41 patients undergoing termination of pregnancy at gestation 6 to 12 weeks. In vitro PRL producing capacity, expressed as mIU/g protein, of the decidua compacta was significantly higher (P < 0.05) than those of decidua spongiosa. Production of PRL increased with gestation from 6 to 12 weeks with a more rapid rate at the later gestation. The pattern of increase fitted significantly (P < 0.0001) to the exponential model for both decidua compacta and decidua spongiosa. The exponential regression equations for decidua compacta and decidua spongiosa were (ln y = 4.25 + 0.19x) and (ln y = 2.80 + 0.31x) respectively. Hence, although both decidua compacta and decidua spongiosa had a similar pattern of increase in PRL production, the rate of increment was significantly greater in decidua spongiosa than in decidua compacta. These findings suggest that separating decidua compacta and decidua spongiosa of the first trimester would reduce the heterogeneity of decidual tissue and offer a new approach to the studies of the synthesis, release and regulation of PRL production by human decidua.

Partial androgen insensitivity and correlations with the predicted three dimensional structure of the androgen receptor ligand-binding domain.

Genetic defects of the human androgen receptor (AR) can cause a wide spectrum of androgen insensitivity syndromes (AIS) ranging from phenotypic females in those with complete AIS; ambiguous genitalia in partial AIS; to male infertility in minimal AIS. The majority of these defects are due to point mutations resulting in amino acid substitutions. It is however unclear why certain mutations result in partial AIS, whereas others in the same exon cause the complete syndrome. We present a case of partial AIS due to a point mutation affecting codon 758 of the AR ligand-binding domain (LBD) that changed the sense of the codon from asparagine to threonine (N758T). The mutant receptor displayed normal binding affinity to DHT but abnormal dissociation kinetics in both patient's fibroblasts and transfected COS-7 cells. The mutant AR was thermolabile, and resulted in approximately 50% reduction in receptor transactivation capacity when examined with a reporter gene incorporating an androgen-response-element. Although the 3-D structure of AR LBD is not known, the homologous region in a member of the steroid receptor superfamily, retinoid-X receptor (RXR-alpha), has been crystallized, allowing comparison of aligned amino-acid sequences of RXR-alpha and AR. The mutation, N758T, lies in a predicted linker region between the fifth alpha-helix (H5) and the first beta-strand (S1). Generally, mutations leading to partial AIS tend to cluster in the predicted linker regions located between the structural helices of the AR LBD. Most strikingly, the predicted linker regions contain over 70% of the mutant ARs associated with prostate cancer in the LBD. The occurrence of mutations associated with both partial AIS and prostate cancer in the same predicted linker regions, suggest that this clustering is not coincidental and that the predicted linker regions are likely to have important, but subtle, roles in defining androgen binding and ligand specificity.

Effect of growth hormone therapy in men with severe idiopathic oligozoospermia.

Some studies have suggested that growth hormone (GH) may enhance folliculogenesis in women, and similarly may enhance spermatogenesis in men with hypogonadotrophic hypogonadism. In this prospective open-controlled pilot study, we investigated the effect of daily subcutaneous GH for 5 months in 12 endocrinologically normal men with severe idiopathic oligozoospermia (< 10 million/ml). All the men had normal karyotype and endocrine tests, including a GH response of > 20,000 mU/l to insulin hypoglycaemia. Nine men with similar sperm counts acted as controls. During treatment, each patient was examined monthly, asked for side effects and had glycosylated haemoglobin, glucose and blood counts monitored. Five semen samples were obtained in the 4 months before treatment, two samples per month during treatment and three samples after stopping treatment. The mean insulin-like growth factor I (IGF-I) was normal before treatment and 1 month after ending treatment, at 206 and 182 micrograms/l, respectively, but increased significantly during treatment to 444 micrograms/l (p < 0.0001, ANOVA). The mean (SD) sperm counts were 2.6 (2.5), 2.5 (3.7) and 2.3 (2.1) million/ml before, during and after GH treatment, respectively, and did not show any statistically significant differences (ANOVA). We conclude that GH does not increase or decrease sperm counts in men with severe idiopathic oligozoospermia.

Luteinizing hormone surge in adult female rats after androgen priming.

The effects of androgens alone and in combination with oestrogens on luteinizing hormone secretion in adult female rats were studied after 1, 2 and 4 months of priming. Each adult female rat was implanted with a Silastic capsule filled with testosterone or dihydrotestosterone alone or together with another similar capsule of oestradiol. An oestrogen challenge test was carried out in each rat after priming. Blood samples were collected before androgen priming. During the oestrogen challenge test, blood samples were collected before (0 h or day 0) and at 12, 24, 36 and 48 h after the intramuscular injection of 50 µg oestradiol vaierate for rats primed for 1 month, or daily for 5 days (days 1 to 5) for other rats. An oestrogen challenge test was also performed in two groups of untreated male and female rats which served as controls. The results indicated that long-term androgen priming had resulted in the attenuation or total blockade of the luteinizing hormone surge which normally occurs during an oestrogen challenge test in adult female rats. This suggests that androgens may be important modulators of the luteinizing hormone surge mechanism in rats. Furthermore, the fact that the luteinizing hormone surge can be blocked by androgen priming in adult female rats suggests that the sexual dimorphic response to oestrogen feedback is not immutably imprinted in the hypothalamic-pituitary axis as was originally proposed.

Maternal-child hepatitis B virus transmission in Singapore.

A study of maternal-infant transmission of hepatitis B virus (HBV) was conducted in Singapore between June 1980 and June 1982. HBsAg carrier rate was highest among Chinese (6.2%) followed by Malay (2.3%) and Indian (0.6%) mothers. The presence of HBeAg in maternal sera correlated well with high titre HBsAg (p = 7.34 X 10(-5)). Overall HBV transmission occurred in 27/56 (48.2%) infants from carrier mothers. The majority of the transmission was perinatal. There was a very strong correlation between transmission and HBeAg status of the mother (p = 1.85 X 10(-9); odds ratio = 68.44) and to a lesser extent with high titre HBsAg (p = 0.002; odds ratio = 6.38). A strong negative correlation was seen between transmission and anti-HBeAg (p = 8.19 X 10(-7); odds ratio = 0.04). At one year 19 (70.4%) infants were still HBsAg positive while seven (25.9%) lost the antigenemia and acquired anti-HBsAg and one developed HBsAg after one year. It could be calculated that perinatal HBV transmission contributed about 18% to the total pool of HBsAg positive infants of one year of age.

Cocultures in human assisted reproduction. Support of embryos in vitro and their specificity.

Assisted reproductive techniques (ART) have contributed tremendously to alleviating infertility in childless couples. However, the "take-home baby" rate for in vitro fertilization (IVF) has been much below the rate for normal fertile couples and for the other tubal procedures such as gamete intrafallopian transfer (GIFT) and tubal embryo transfer (TET). It was therefore speculated that the tubal environment might be playing an important role towards the viability of the embryo and that, in IVF, 4-6 cell stage embryos that were usually replaced directly into the uterus might not be receiving the beneficial effects of the tubal environment. Pronuclear stage human and mouse embryos were thus cocultured with reproductive and other somatic cell types to evaluate the quality of the various cleaving stages and the percentage yield of blastocysts. Human ampullary cell cultures were established in 6 to 7 days and kept alive through one menstrual cycle. Cleavage to the compacted and cavitating stages was achieved in 78% and 69%, respectively, of human embryos cocultured in 24-48 hour human ampullary subcultures as compared to 50% and 33%, respectively, for embryos grown in culture medium alone. The percentages of expanded blastocysts and hatching stages in such cocultures were not significantly different from those of controls. However, human ampullary and cumulus and mouse ampullary and muscle fibroblast cultures supported cleavage and development of mouse embryos up to the hatching stages as well as or better than controls. It thus appears that there is no absolute specificity for a reproductive tract source of cocultured cells and that the beneficial effects of ampullary and cumulus cells are not species-specific. The coculture system also appears to overcome the embryonic blocks of the human and mouse, which may be mediated via growth factors that activate the embryonic genome.

The origin and distribution of cortical granules in human oocytes with reference to Golgi, nucleolar, and microfilament activity.

The origin and distribution of cortical granules were investigated in human preovulatory oocytes at various phases of maturation. Twenty-five oocytes obtained from unstimulated small antral follicles and from stimulated large antral and mature follicles were examined by transmission electron microscopy. Ovarian stimulation in women was accomplished by administering Clomid followed by hMG or hCG or both. Small antral follicle oocytes were dissected from ovarian biopsies, while the other oocytes were recovered by laparoscopy. Some oocytes were allowed to mature in Ham's F-10 or Whittingham's T-6 media before routine fixation in glutaraldehyde/osmium. Cortical granules originate from typical, hypertrophic Golgi complexes during early maturation and continue till its completion. Evidently there are two waves of cortical granule synthesis, the first more prolific than the second. The first occurred in small antral follicle oocytes, when there was a peak in Golgi activity, and the second was observed at the germinal vesicle stage, particularly at the onset of resumption of meiosis. Golgi complexes became progressively scarce as oocytes completed first maturation. Golgi membranes were also involved in the formation of lysosomes. A well-defined band of microfilaments was detected in small antral follicle oocytes which seemed to prevent the cortical granules, organized in a single layer, from migrating to the periphery. This band gradually became disorganized at the germinal vesicle stage as oocytes resumed meiosis, when cortical granules were apparently migrating to the surface. Metaphase I and mature oocytes had one to three discontinuous layers of cortical granules beneath the oolemma. The general organization of oocytes was also investigated and the roles of the nucleolus and endoplasmic reticulum in relation to Golgi activity and cell secretion were discussed.

Reliability of intrauterine pressure measurements.

The reliability of intrauterine pressure measurements was studied by placing two catheters in the same uterus in women in labor. Nineteen parturients were studied, five with two Intran I catheters in different pockets of amniotic fluid, five with two Intran II catheters in different pockets, and nine with two Intran II catheters in the same pocket of amniotic fluid. Among 1429 contractions studied, the catheters showed differences of pressure of less than 5 mmHg in 86% of contractions and of less than 10 mmHg in 94%. When cumulative uterine activity was calculated for each labor, the difference in the total active pressures shown by the two catheters was less than 5% in 17 cases and 5% or more in two cases. These differences are unlikely to be of any importance in the management of labor.

Premature rupture of membranes in nulliparas at term with unfavorable cervices: a double-blind randomized trial of prostaglandin and placebo.

OBJECTIVE: To determine whether the use of a prostaglandin (PG) E2 3-mg pessary followed by a delay of 12 hours before stimulation of labor with oxytocin improves obstetric outcome compared with the use of a placebo pessary. METHODS: One hundred fifty-five nulliparas at term with poor cervical scores (modified Bishop score below 6 of 10) and premature rupture of membranes (PROM) were recruited for this double-blind, placebo-controlled randomized trial. On admission to the study, either a PGE2 pessary or an identical-appearing placebo pessary was inserted into the posterior fornix. If labor did not start in the next 12 hours or if symptoms and signs of infection were evident, labor was induced with oxytocin infusion. Assignment was unblinded at the end of the study, and details of the labor and maternal and neonatal outcome in women who received a PG pessary were compared with those who received a placebo pessary. RESULTS: Women receiving a PG pessary were significantly less likely to require stimulation of labor at the end of 12 hours than were those given a placebo pessary (37 versus 58%, P = .002). The mean time between admission to study and delivery was significantly shorter in the PG group compared with the placebo group (15 versus 19 hours, P = .01). The rate of cesarean delivery was not statistically different in the two groups (13.9% with PG versus 15.8% with placebo). CONCLUSION: In nulliparas with poor cervical scores who present with PROM at term and no evidence of infection or obstetric complications, use of a PGE2 pessary resulted in more women establishing labor earlier, with a resultant reduction in the admission-to-delivery interval, compared with the use of a placebo pessary. The cesarean delivery rates in the two groups were similar, and there were no significant differences in neonatal outcome.

Augmentation of labor: does internal tocography result in better obstetric outcome than external tocography?

Labor was augmented in 250 patients for slow progress. These women were randomized to have uterine contractions recorded by either an external tocotransducer or an intrauterine catheter. Oxytocin was titrated to achieve an optimal frequency of contractions of six to seven every 15 minutes in each group; additional information on active contraction area profiles was available for those patients who had an intrauterine catheter. There was no statistically significant difference between the groups in the length of the post-augmentation period. The mean maximum dose of oxytocin was 11.1 mU/minute in the external tocography group and 11.0 mU/minute in the internal tocography group. Evidence of uterine hyperstimulation requiring temporary reduction of the oxytocin dose occurred in 19 and 20.2% of the patients in the external and internal tocography groups, respectively. Cesarean delivery was necessary in 12.6 and 16.9% of patients in the external and internal tocography groups, respectively, which is a nonsignificant difference. The incidence of low Apgar scores in the neonates and admission to neonatal intensive care was similar in both groups. The incidence of poor Apgar scores was not different between those who had transient hyperstimulation and those who had no hyperstimulation. In the management of augmented labor, monitoring of uterine contractions by intrauterine pressure catheters did not confer any advantage over tocography by external transducers.

Evaluation of maternal perception of sound-provoked fetal movement as a test of antenatal fetal health.

Maternal perception of sound-provoked fetal movement was correlated with the results of nonstress cardiotocography in 1097 women with obstetric or medical antenatal risk factors. Ninety-two percent of the mothers felt fetal movements with the stimulus; all but three had a reactive non-stress test (NST). These three women were taking multiple antihypertensive drugs and were less than 33 weeks' gestation. Of 88 patients with no maternal perception of sound-provoked fetal movement, ten had nonreactive NSTs. Ultrasound confirmed the absence of fetal movement to the stimulus. The outcome in nine of these ten cases suggested some evidence of fetal compromise. Maternal perception of sound-provoked fetal movement correlated well with the results of the NST; the sensitivity (76.9%), specificity (92.8%), and negative predictive value (99.7%) were all high, although the positive predictive value was only 11.4%. Maternal perception of sound-provoked fetal movement may suffice as an inexpensive and simple method of evaluating antenatal fetal well-being in risk situations. When the mother is doubtful or does not feel the sound-provoked fetal movement, NST is indicated to evaluate the fetal health.

Direct fetal therapy for hydrops secondary to congenital atrioventricular heart block.

BACKGROUND: The outcome for the fetus with hydrops fetalis secondary to complete congenital heart block is almost uniformly poor. Transplacental fetal therapy with inotropic drugs may be unreliable in the hydropic fetus. We describe our experience of direct fetal therapy with digoxin and furosemide in three cases. CASES: In the first two cases, fetal hydrops secondary to congenital heart block was detected at 33 and 31 weeks. Direct fetal therapy with digoxin and furosemide was given and the mothers received digoxin concurrently. The neonates were born normally at 35 weeks and 33 weeks; one required ventricular pacing, but both are currently alive and well. In the third case, fetal hydrops was detected at 24 weeks, when furosemide was used alone for direct fetal therapy. The mother developed chorioamnionitis at 29 weeks and was delivered by cesarean; the infant died of cardiac failure after 48 hours. CONCLUSION: Initial direct fetal therapy with digoxin and furosemide followed by transplacental treatment with digoxin and direct fetal therapy with diuretics might improve the perinatal salvage of hydropic fetuses with congenital heart block. Treatment must be individualized because development of infection and preterm labor might defeat the objective.

Umbilical artery Doppler velocimetry as a labor admission test.

Doppler ultrasound of the umbilical artery flow velocity waveform was studied prospectively as an admission test at the labor ward. Recordings were made in 575 women in various stages of labor before, during, and after uterine contractions, and evaluated in relation to intrapartum and fetal outcome variables. No association was found between abnormal flow velocity waveforms and cord complications, meconium-stained amniotic fluid, or abnormal fetal heart rate tracing, nor was there any association with operative delivery for fetal distress or low Apgar scores at 1 and 5 minutes. Small for gestational age fetuses had significantly more abnormal flow velocity waveforms than appropriate for gestational age fetuses, and so had those with umbilical artery acidemia compared with those with normal pH. The results indicate that Doppler recording of the umbilical artery flow velocity waveform as an admission test at the labor ward is not a good predictor of fetal distress in an unselected population.

Does prostaglandin confer significant advantage over oxytocin infusion for nulliparas with pre-labor rupture of membranes at term?

Ninety-four nulliparous women with a poor cervical score (less than 6) who had premature rupture of membranes at term were randomized by sealed envelope into two groups. One group received immediate stimulation of labor with oxytocin infusion. The second group received two prostaglandin E2 (PGE2) 3-mg pessaries 4 hours apart, followed by oxytocin infusion, if necessary. The interval between initiation of therapy to onset of labor was significantly longer in the PG group, but the length of labor was similar in both groups. The maximum dose of oxytocin needed was significantly higher in the oxytocin group. The cesarean delivery rate in the oxytocin group was 14.9%, compared with 19.1% in the PG group (not significantly different). All seven cesareans in the oxytocin group and seven of nine in the PG group were for failed stimulation of labor. Neonatal Apgar scores at 1 and 5 minutes and admission to the neonatal intensive care unit were similar in the two groups. The incidence of maternal and neonatal infection was small and was not different in the two groups. The use of PGE2 3-mg pessaries 4 hours apart, followed by oxytocin infusion if necessary, did not confer any benefit over the use of intravenous oxytocin in obstetric or neonatal outcome when both agents were started a few hours after admission.

No evidence of hearing loss due to fetal acoustic stimulation test.

The fetal acoustic stimulation test is used increasingly for fetal assessment. To evaluate the possibility of acoustic trauma, 465 children who had been exposed to vibroacoustic stimulation in utero were screened at 4 years of age for high-frequency hearing loss. A hand-held audiometer with test tones at 25 dB for 1000 and 4000 Hz was used because responses at this level indicate hearing within normal limits for middle and high frequencies. Thirty-one children failed the test. Failure to respond was followed by inspection of the ear canal for wax, tympanometry, and reflex measurement to assess the tympanic membrane, middle ear, and eustachian tube. Hearing was retested across the full frequency range using conventional audiometric technique, and referral for otolaryngologic examination and treatment was made if necessary. The causes of hearing loss were impacted wax in 12, current or recent upper respiratory tract infection with eustachian tube dysfunction and middle ear effusion in 15, and unresolved middle ear effusion after treatment with antibiotics for otitis media in two. Profound bilateral sensorineural hearing loss of unknown origin was found in one and slight bilateral gently sloping hearing loss in another. Retesting of all children with conductive hearing loss indicated that hearing had returned to normal after treatment. None of the children showed evidence of hearing loss.

Vibroacoustic stimulation and intrauterine sound pressure levels.

The sound pressure level in amniotic fluid generated by vibroacoustic stimulation, assessed with a hydrophone placed close to the fetal head, was studied in 16 subjects. The mean recorded sound pressure level was 115 dB and the highest level was 129 dB. The range of the background noise was 63.5-80.5 dB. There was no obvious relationship between the distance from the stimulator to the hydrophone and the intrauterine sound pressure level. Although sound pressure levels are high, they are probably reduced before reaching the cochlea of the fetus because of the surrounding amniotic fluid and the fluid in the middle ear.