The physiologic mechanisms of variable decelerations Robert H. Ball, MD, and Julian T. Parer, MD, PhD San Francisco, California OBJECTIVES: The purpose of this study was to determine the cause and physiologic consequences of variable decelerations. STUDY DESIGN: Previous studies of heart rate changes in human and experimental animals were critically reviewed with respect to the autonomic nervous system component, the cause of the increased vagal activity, and the role of cord compression or other stimuli in these heart rate changes. RESULTS: There is abundant evidence from experimental and human studies that variable decelerations can be reproduced by either cord compression or head compression. The vagal reflex produced is probably caused by a combination of chemoreflex (earlier in the deceleration) and baroreflex (later). The variable deceleration is accompanied by an acidosis, primarily respiratory, and probably hypoxemia. Cord compression results in decreased umbilical blood flow. CONCLUSIONS: Recent Doppler velocimetry studies suggest that even though the variable decelerations may be similar in duration and depth, the reduction of umbilical blood flow may be greater when the prime cause is cord compression than when the prime cause is vagal reflex from another source. (AM J OSSTET GVNECOL 1992;166:1683-9.)
Key words: Variable decelerations, cord occlusion, head compression, Doppler ultrasonography Variable decelerations of the fetal heart rate (FHR) are characterized by their abrupt onset and recovery, variable shape, and variable relationship to contractions. 1 In contrast to late decelerations they are considered less ominous because their association with fetal acidosis appears to be weaker. 2 There is general agreement that variable decelerations are produced by a vagal response of the fetus. However, the stimulus for the vagal activity is less certain. In this report we will examine the evidence for the mechanisms of variable decelerations and discuss the potential importance of different mechanisms in the management of the fetus during labor.
Umbilical cord occlusion Barcroft' in 1947 occluded the umbilical cord of the fetal goat, producing a deceleration of a type now considered as variable deceleration, which was abolished by vagotomy. Since then numerous investigators have performed studies involving occlusion of the umbilical cord or its component vessels in both experimental anFrom the University of California San Francisco, Department of Obstetrics, Gynecology and Reproductive Sciences, and the Cardiovascular Research Institute. Presented at the Fifty-eighth Annual Meeting of the Pacific Coast Obstetrical and Gynecological Society, Ashland, Oregon, September
9-12,1991. Reprint requests: Robert H. Ball, MD, University of California San Francisco, Department of Obstetrics, Gynecology and Reproductive Sciences, Box 0550, San Francisco, CA 94143-0550.
6/6/37491
imals and humans. The studies have been carried out on either anesthetized or chronically prepared animals, the occlusion has been partial or complete, and the duration of occlusion has varied from several seconds up to 1 hour. In a number of the studies fetal blood pressures and arterial blood gases have also been measured. In addition, the involvement of the autonomic nervous system has been examined. Heart rate changes during umbilical cord occlusion. As with the work of Barcroft, the majority of workers have reported an abrupt decrease in heart rate with cord occlusion. Studies have been carried out in fetal sheep"·9 goats,1O baboons, II and humans. 12• IS The duration of occlusion has ranged from 5 seconds to >60 minutes, and either umbilical arteries alone, umbilical veins alone, or the whole of the umbilical cord has been occluded. In some cases the occlusion is described as partial and in other cases as complete. The responses to umbilical arterial occlusion are similar, when reported, to those of umbilical cord occlusion of all vessels. The initial response is a decrease in heart rate, although some workers under some conditions have reported a tendency for fetal heart rate to return toward normal values. Thus Towell and Salvador lO noted that there was a bradycardia for the first 5 minutes of occlusion, but this reverted toward normal in the second 5 minutes of occlusion. Ball et al. 9 noted that the initial bradycardia returned toward a normal heart rate approximately 1o minutes after severe umbilical arterial occlusion producing an oxygen satura-
1683
1684 Ball and Parer
tion below about 10% in the fetal arterial blood. Goodlin and Lowe" noted a deceleration with subsequent acceleration with I to 2 minutes of occlusion in the premature human fetus. With respect to umbilical vein occlusion alone, most workers similarly report an abrupt drop in heart rate. However, there are some exceptions, and James et al." in the anesthetized (fluothane) fetal baboon noted that with partial mild' occlusion of the umbilical vein that there was a tachycardia, although with complete occlusion there was a bradycardia. Their occlusion was from 0.2 to 7.7 minutes and the time course of changes is not stated. Kunzel et al. S noted that there was only a drop in heart rate with umbilical venous occlusion when the fetal arterial oxygen saturation was 60%. Similar observations can be made in human newborns after a series of severe variable decelerations or prolonged bradycardias. Presumably in these latter cases there is reduced umbilical blood flow, allowing time for the sluggishly moving blood to equilibrate with maternal blood in the placenta. Although data on arterial blood gases during human cord occlusion are not available, there are some data on fetal scalp blood gases taken during variable deceleration complexes during labor. '6 There is a decrease in pH by the end of a variable deceleration, but this resolves over the next several minutes. The acidosis is primarily due to a concomitant elevation of carbon dioxide." The role of the autonomic nervous system in heart rate and blood pressure changes during umbilical cord occlusion. Since the initial work of Barcroft' showing that vagotomy altered the initial bradycardia produced by cord compression, thus demonstrating the role of the vagus in producing the bradycardia, a number of other workers have confirmed this either by surgical' or pharmacologic vagotomy.6. " Investigators have also concerned themselves with whether the initial bradycardia is caused by chemoreceptor influence (because of the hypoxemia) or whether the baroreflex is responsible because of the rapid increase in arterial blood pressure. Itskovitz et aI.' make a good case for the bradycardia during partial cord occlusion being initially due to the chemoreflex and later being due to the baroreflex, because there was a decrease in fetal oxygen content but no specific blood pressure elevation for about 30 seconds. In considering the mechanical changes during cord occlusion, in compression of the umbilical artery there will be an increase in resistance, and this will manifest
Volume 166 Number 6, Part I
as an increase in arterial blood pressure. Should there be only partial cord occlusion, with compression of the umbilical vein but not the artery, then there will be a decrease in venous return to the heart. This may result in a fetal hypoxemia that could then be responsible for a chemoreceptor-produced bradycardia. As noted above, however, it is impossible to know in any particular situation of partial cord occlusion what the differential arterial and venous umbilical blood flows are because this has not yet been measured. There has been some work on the role of the sympathetic nervous system in the cardiovascular changes produced by umbilical cord occlusion. De Haan et al. 6 used a-adrenergic blockade during umbilical cord occlusion and found that there was still an initial increased blood pressure but that this decreased subsequently. In umbilical cord occlusion in anesthetized baboons,James et al. II noted that with partial mild occlusion there was an increase in FHR, which was prevented by sympathetic blockade with propranolol (a p-adrenergic blocker) or phentolamine (an a-adrenergic blocker). From the evidence presented one could conclude that during the vagally-produced bradycardia there is simultaneous sympathetic chronotropic activity that is less and that this is masked by the intensity of the vagal bradycardia. Amnioinfusion and other clinical evidence of cord occlusion. There have been attempts to correlate umbilical cord position at delivery with the presence of variable decelerations in labor. In one study population,18 28% of fetuses were found at delivery to have a cord position that was felt to be susceptible to compression. However, there was no convincing evidence that the incidence of variable decelerations was greater in the group with "abnormal" cord position. The authors described a heart rate pattern consisting of accelerations and variable decelerations, and this finding was significantly associated with an abnormal cord position. They felt that these changes occurred with mild occlusion of the cord. Gabbe et al. 19 documented the association of oligohydramnios and variable decelerations in a chronic rhesus monkey preparation and felt this was caused by cord compression. When amniotic fluid volume was replaced through an indwelling catheter, the variable decelerations were virtually abolished. The logical extension of this work was to investigate whether replenishing amniotic fluid in the uterine cavity in humans would reduce the incidence of variable decelerations. In 1985 Miyazaki and Nevarez 20 reported on the results of a randomized trial of amnioinfusion in patients with variable decelerations in labor. With an infusion of 800 ml of normal saline solution over 40 to 60 minutes, variables were abolished in 51 % of patients compared with 4.2% that resolved spontaneously in the nontreat-
Mechanisms of variable decelerations
1685
ment group. Nageotte et al. 21 in the same year also found amnioinfusion in preterm fetuses to be successful. They randomized a group of patients with preterm premature rupture of membranes to treatment with amnioinfusion or control during labor. They documented a significantly lower incidence of variable decelerations in the treated group. The premise is that amnioinfusion provides renewed cushioning for the umbilical cord, protecting it from compression. Its apparent success provides indirect evidence for the role of cord compression in the origin of variable decelerations. The influence of fetal head compression
It has been known for over 150 years that compression of the fetal vertex appears to be associated with decelerations in heart rate. It has been speculated that compression of the brain leads to "activation of vagal centers. "22 In 1933 Rech 23 elicited fetal bradycardia that mimicked variable decelerations of varying severity by adjusting the force of the squeeze with his hand on the fetal vertex. Chung and Hon 2' refined this by placing hard pessaries in the vagina of women in advanced labor. In between contractions pressure was exerted on these pessaries, compressing the fetal vertex. This led to variable-type decelerations in 18 of 19 attempts. They felt this was similar to the pressure exerted by the dilating cervix. Further evidence in their minds for vagal stimulation by head compression was the consistent decelerations noted with the locking of forceps and initiation of traction. Paul et al. 25 studied this phenomenon in anesthetized fetal sheep. Manual compression of the fetal vertex was performed for 20 to 50 seconds. Carotid blood flow decreased markedly with the compression, and blood pressure increased as profound bradycardia ensued. The investigators believed that this apparent vagal response could be either baroreceptor or chemoreceptor mediated. Mann et al!6 used a metal clamp to apply pressure to fetal sheep heads. They found no consistent changes in heart rate, but cerebral blood flow and metabolism decreased as did cerebral oxygen consumption. The fact that variable decelerations cannot be effectively abolished by amnioinfusion in the second stage of labor supports the view that head compression is a mechanism for variable decelerations!' Kelly" speculated that bradycardia with head compression was an appropriate physiologic response. With increased diastole, according to him, blood flow through the foramen ovale and the ductus arteriosus changes so as to shunt more oxygenated umbilical venous blood to the brain. In contrast, Goodlin and Haesslein 27 suggested several additional mechanisms for variable decelerations, including fetal Valsalva maneuver and maternal activities
1686 Ball and Parer
June 1992 Am J Obslel Gynecol
such as voiding and seizures. They felt that some of the bradycardias may be "adult" reflexes that are inappropriate for the fetus and may result in reduced fetal blood flow and oxygenation.
a deceleration of heart rate
Key words: Variable decelerations, cord occlusion, head compression, Doppler ultrasonography Variable decelerations of the fetal heart rate (FHR) are characterized by their abrupt onset and recovery, variable shape, and variable relationship to contractions. 1 In contrast to late decelerations they are considered less ominous because their association with fetal acidosis appears to be weaker. 2 There is general agreement that variable decelerations are produced by a vagal response of the fetus. However, the stimulus for the vagal activity is less certain. In this report we will examine the evidence for the mechanisms of variable decelerations and discuss the potential importance of different mechanisms in the management of the fetus during labor.
Umbilical cord occlusion Barcroft' in 1947 occluded the umbilical cord of the fetal goat, producing a deceleration of a type now considered as variable deceleration, which was abolished by vagotomy. Since then numerous investigators have performed studies involving occlusion of the umbilical cord or its component vessels in both experimental anFrom the University of California San Francisco, Department of Obstetrics, Gynecology and Reproductive Sciences, and the Cardiovascular Research Institute. Presented at the Fifty-eighth Annual Meeting of the Pacific Coast Obstetrical and Gynecological Society, Ashland, Oregon, September
9-12,1991. Reprint requests: Robert H. Ball, MD, University of California San Francisco, Department of Obstetrics, Gynecology and Reproductive Sciences, Box 0550, San Francisco, CA 94143-0550.
6/6/37491
imals and humans. The studies have been carried out on either anesthetized or chronically prepared animals, the occlusion has been partial or complete, and the duration of occlusion has varied from several seconds up to 1 hour. In a number of the studies fetal blood pressures and arterial blood gases have also been measured. In addition, the involvement of the autonomic nervous system has been examined. Heart rate changes during umbilical cord occlusion. As with the work of Barcroft, the majority of workers have reported an abrupt decrease in heart rate with cord occlusion. Studies have been carried out in fetal sheep"·9 goats,1O baboons, II and humans. 12• IS The duration of occlusion has ranged from 5 seconds to >60 minutes, and either umbilical arteries alone, umbilical veins alone, or the whole of the umbilical cord has been occluded. In some cases the occlusion is described as partial and in other cases as complete. The responses to umbilical arterial occlusion are similar, when reported, to those of umbilical cord occlusion of all vessels. The initial response is a decrease in heart rate, although some workers under some conditions have reported a tendency for fetal heart rate to return toward normal values. Thus Towell and Salvador lO noted that there was a bradycardia for the first 5 minutes of occlusion, but this reverted toward normal in the second 5 minutes of occlusion. Ball et al. 9 noted that the initial bradycardia returned toward a normal heart rate approximately 1o minutes after severe umbilical arterial occlusion producing an oxygen satura-
1683
1684 Ball and Parer
tion below about 10% in the fetal arterial blood. Goodlin and Lowe" noted a deceleration with subsequent acceleration with I to 2 minutes of occlusion in the premature human fetus. With respect to umbilical vein occlusion alone, most workers similarly report an abrupt drop in heart rate. However, there are some exceptions, and James et al." in the anesthetized (fluothane) fetal baboon noted that with partial mild' occlusion of the umbilical vein that there was a tachycardia, although with complete occlusion there was a bradycardia. Their occlusion was from 0.2 to 7.7 minutes and the time course of changes is not stated. Kunzel et al. S noted that there was only a drop in heart rate with umbilical venous occlusion when the fetal arterial oxygen saturation was 60%. Similar observations can be made in human newborns after a series of severe variable decelerations or prolonged bradycardias. Presumably in these latter cases there is reduced umbilical blood flow, allowing time for the sluggishly moving blood to equilibrate with maternal blood in the placenta. Although data on arterial blood gases during human cord occlusion are not available, there are some data on fetal scalp blood gases taken during variable deceleration complexes during labor. '6 There is a decrease in pH by the end of a variable deceleration, but this resolves over the next several minutes. The acidosis is primarily due to a concomitant elevation of carbon dioxide." The role of the autonomic nervous system in heart rate and blood pressure changes during umbilical cord occlusion. Since the initial work of Barcroft' showing that vagotomy altered the initial bradycardia produced by cord compression, thus demonstrating the role of the vagus in producing the bradycardia, a number of other workers have confirmed this either by surgical' or pharmacologic vagotomy.6. " Investigators have also concerned themselves with whether the initial bradycardia is caused by chemoreceptor influence (because of the hypoxemia) or whether the baroreflex is responsible because of the rapid increase in arterial blood pressure. Itskovitz et aI.' make a good case for the bradycardia during partial cord occlusion being initially due to the chemoreflex and later being due to the baroreflex, because there was a decrease in fetal oxygen content but no specific blood pressure elevation for about 30 seconds. In considering the mechanical changes during cord occlusion, in compression of the umbilical artery there will be an increase in resistance, and this will manifest
Volume 166 Number 6, Part I
as an increase in arterial blood pressure. Should there be only partial cord occlusion, with compression of the umbilical vein but not the artery, then there will be a decrease in venous return to the heart. This may result in a fetal hypoxemia that could then be responsible for a chemoreceptor-produced bradycardia. As noted above, however, it is impossible to know in any particular situation of partial cord occlusion what the differential arterial and venous umbilical blood flows are because this has not yet been measured. There has been some work on the role of the sympathetic nervous system in the cardiovascular changes produced by umbilical cord occlusion. De Haan et al. 6 used a-adrenergic blockade during umbilical cord occlusion and found that there was still an initial increased blood pressure but that this decreased subsequently. In umbilical cord occlusion in anesthetized baboons,James et al. II noted that with partial mild occlusion there was an increase in FHR, which was prevented by sympathetic blockade with propranolol (a p-adrenergic blocker) or phentolamine (an a-adrenergic blocker). From the evidence presented one could conclude that during the vagally-produced bradycardia there is simultaneous sympathetic chronotropic activity that is less and that this is masked by the intensity of the vagal bradycardia. Amnioinfusion and other clinical evidence of cord occlusion. There have been attempts to correlate umbilical cord position at delivery with the presence of variable decelerations in labor. In one study population,18 28% of fetuses were found at delivery to have a cord position that was felt to be susceptible to compression. However, there was no convincing evidence that the incidence of variable decelerations was greater in the group with "abnormal" cord position. The authors described a heart rate pattern consisting of accelerations and variable decelerations, and this finding was significantly associated with an abnormal cord position. They felt that these changes occurred with mild occlusion of the cord. Gabbe et al. 19 documented the association of oligohydramnios and variable decelerations in a chronic rhesus monkey preparation and felt this was caused by cord compression. When amniotic fluid volume was replaced through an indwelling catheter, the variable decelerations were virtually abolished. The logical extension of this work was to investigate whether replenishing amniotic fluid in the uterine cavity in humans would reduce the incidence of variable decelerations. In 1985 Miyazaki and Nevarez 20 reported on the results of a randomized trial of amnioinfusion in patients with variable decelerations in labor. With an infusion of 800 ml of normal saline solution over 40 to 60 minutes, variables were abolished in 51 % of patients compared with 4.2% that resolved spontaneously in the nontreat-
Mechanisms of variable decelerations
1685
ment group. Nageotte et al. 21 in the same year also found amnioinfusion in preterm fetuses to be successful. They randomized a group of patients with preterm premature rupture of membranes to treatment with amnioinfusion or control during labor. They documented a significantly lower incidence of variable decelerations in the treated group. The premise is that amnioinfusion provides renewed cushioning for the umbilical cord, protecting it from compression. Its apparent success provides indirect evidence for the role of cord compression in the origin of variable decelerations. The influence of fetal head compression
It has been known for over 150 years that compression of the fetal vertex appears to be associated with decelerations in heart rate. It has been speculated that compression of the brain leads to "activation of vagal centers. "22 In 1933 Rech 23 elicited fetal bradycardia that mimicked variable decelerations of varying severity by adjusting the force of the squeeze with his hand on the fetal vertex. Chung and Hon 2' refined this by placing hard pessaries in the vagina of women in advanced labor. In between contractions pressure was exerted on these pessaries, compressing the fetal vertex. This led to variable-type decelerations in 18 of 19 attempts. They felt this was similar to the pressure exerted by the dilating cervix. Further evidence in their minds for vagal stimulation by head compression was the consistent decelerations noted with the locking of forceps and initiation of traction. Paul et al. 25 studied this phenomenon in anesthetized fetal sheep. Manual compression of the fetal vertex was performed for 20 to 50 seconds. Carotid blood flow decreased markedly with the compression, and blood pressure increased as profound bradycardia ensued. The investigators believed that this apparent vagal response could be either baroreceptor or chemoreceptor mediated. Mann et al!6 used a metal clamp to apply pressure to fetal sheep heads. They found no consistent changes in heart rate, but cerebral blood flow and metabolism decreased as did cerebral oxygen consumption. The fact that variable decelerations cannot be effectively abolished by amnioinfusion in the second stage of labor supports the view that head compression is a mechanism for variable decelerations!' Kelly" speculated that bradycardia with head compression was an appropriate physiologic response. With increased diastole, according to him, blood flow through the foramen ovale and the ductus arteriosus changes so as to shunt more oxygenated umbilical venous blood to the brain. In contrast, Goodlin and Haesslein 27 suggested several additional mechanisms for variable decelerations, including fetal Valsalva maneuver and maternal activities
1686 Ball and Parer
June 1992 Am J Obslel Gynecol
such as voiding and seizures. They felt that some of the bradycardias may be "adult" reflexes that are inappropriate for the fetus and may result in reduced fetal blood flow and oxygenation.
a deceleration of heart rate
