Motor Activity and Apnea in Preterm Infants Is There a Causal Relationship?1.2
OOMMEN P. MATHEW, CECIL K. THOPPIL, and MARY BELAN
Introduction
Apnea is a frequent and important clinical problem in preterm infants, its incidence being inverselyrelated to the gestational age (1, 2). It is generally believed that all the episodes of apnea occur during sleep. Recovery from obstructive apnea in adults is generally associated with arousal from sleep. Although the role of arousal in the termination of apnea in premature infants has not been investigated, the intervention of gentle and, occasionally, vigorous stimulation to terminate apnea is based on the premise that increased afferent stimuli induce arousal and resumption of breathing. Occurrence of apnea during wakefulness in preterm infants, unassociated with feeding, has been documented (3), but the frequency and characteristics of apnea that occur during wakefulness or immediately after arousals have not been established. For that matter, even the recent editions of the textbooks of neonatology or reviews on apnea of prematurity do not even mention the occurrence of apnea while awake (2, 4-6). The aim of the present study was to investigate this phenomenon in a group of apneic preterm infants. Methods A group of 10 preterm infants with clinical apnea form the basis of this report. They had a mean birth weight (± SD) of 1,237 ± 223 g and a gestational age of 31.1 ± 1.1 wk. These infants werestudied at a postnatal age of 10.2 ± 7.2 days. Six had recovered from respiratory illness (five transient tachypnea and one respiratory distress syndrome), and none of the infants was receiving supplemental oxygen at the time of the study. Five infants were receiving theophylline and one infant was receiving both theophylline and doxapram at the time of the study. These infants werepart of a study investigating the role of arousal in the resolution of apnea, which was approved by the institutional review board. Informed parental consent was obtained for all infants prior to the study, and no sedation was used in any of these infants. 842
SUMMARY It is generally believed that all apneic episodes In preterm infants occur during sleep. Because occurrence of apnea during wakefulness has been documented in these infants, we Investigated its frequency and characteristics In 10 premature Infants using slmultsneous polygraphic and video recordings. Behavioral arousal with motor activity preceded the onset of apnea In 202 episodes. Approximately 60% of episodes began within 15 s of arousal. Whereas most of the episodes were short and asymptomatic, hypoxia and/or bradycardia developed during 18apneic spells, and these episodes accounted for a third of all apneic episodes that resulted In bradycardia or hypoxia in these infants studied. Essentially, all movement-related apneas (17of 18) began within 15 s of arousal; motor activity continued throughout apnea In 13, whereas apnea resolved after cessation of motor activity in the remaining five. The main finding of the present study is that movementrelated apnea is far more frequent than previously recogniZed. This finding Is important In the clinical management of preterm Infants with apnea. AM REV RESPIR DIS 1991; 144:842-844
Each infant was videotaped and polygraphically monitored for 1 to 2 h after feeding. The infants were studied in a quiet room while lying supine under a radiant warmer. The skin temperature was servo-controlled at 36 to 36.50 C. Movements of the chest and abdomen weremonitored by inductive plethysmography (Respitrace'";Ardsley, NY), oxygen saturation and heart rate were measured by transcutaneous pulse oximetry (Nellcor Inc., Hayward, CA), and EKG using standard EKG electrodes and nasal airflow with a nasal screen flowmeter were monitored as previously described (7). Oral breathing was monitored by the sampling catheter of the CO 2 analyzer placed in front of the mouth. These parameters weredisplayed on an oscilloscope and recorded on a Gould eightchannel recorder (Gould Instruments, Cleveland, OH). The analog signals wereprocessed through multiplexors (Vettor Co, Rebersberg, PA) and stored on the HI-FI channels of the same videotapes in which the video images were being recorded. Auditory observations such as crying and other vocalizations as well as arousals induced by extraneous stimuli were noted on the polygraph paper by the investigators at the time of the study. The apneic episodes in the preterm infants were identified from the polygraphic recording by the absence of airflow for 5 s or greater. Apneas Iasting e 15 s were termed long apneas, whereas those lasting < 15 s were termed short apneas. Apneas were classified as central if the abdominal and chest wall movements were absent, obstructive if respiratory efforts persisted, and mixed if a combination of central and obstructive components was present. Bradycardia was defined
as a heart rate ~ 100 beats/min, and R-R interval from the EKG channel formed the basis of these calculations. Hypoxia was defined as an O 2 saturation of ~ 800/0. In order to eliminate falsely low pulse oximeter readings during movement, heart rate recorded from the pulse oximeter was compared with the heart rate derived from the EKG channel. If the heart rate obtained from the pulse oximeter was lower than the heart rate from the EKG by more than 5 beats/min, values of oxygen saturation observed during that period werenot included for further analysis. Bradycardia and hypoxia were considered to be associated with apneas if they occurred during the apneic period or within 5 s of the termination of the apnea. Playback of the tape allowed us to viewthe behavior of the infant during each episode. If movement was observed at apnea onset, the event was further reviewed to determine whether the criteria for behavioral arousal was met prior to the onset Of apnea as described before (8). If behavioral arousal criteria were not met, then the episode was coded as sleeprelated apnea. Arousal was defined as the occurrence of cry, opening of eyes (> 5 s), or
(Received in originalform September 14, 1990and in revised form February 5, 1991) 1 From the Departments of Pediatrics and Physiology & Biophysics, University of Texas Medical Branch, Galveston, Texas. 2 Correspondence and requests for reprints should be addressed to Oommen P. Mathew, M.D., Department of Pediatrics, University of TexasMedical Branch, Galveston, TX 77550.
MOVEMENT-RELATED APNEA IN PREMATURE INFANTS
843 %
% r-..
o :§
o
40
r---
% o
~ 30
Q)
c;
c 20
Q.
30
~
r---
o 20
~
Ql
C
«
o :;:; 20 o
:g
E
E o
o Q. E
-
1t 10 o
Q.
«
o
0-
10
E >,
>,
if)
(5
5-15
15-30
}30
Seconds
Fig. 1. Time interval between beginning of arousal and apnea onset.
concurrent occurrence of sustained movements (> 5 s) of head and neck, trunk, and extremities. Typically, these activities consisted of flexion or extension of the head and neck, slow writhing movement of the trunk, and flexion-extension movements of the extremities. These activities were similar to the "squirming" movements described by AbuOsba and coworkers (3). The relationship between beginning of the motor activity, apnea onset, and resolution were determined in all but nine episodes.
10
(f)
O'---..L-----'------'--------'-------'---'---5-10 10-15 2.15 Apnea Duration (sec) Fig. 2. Relationship between apnea duration and development of symptoms; 24% of long apnea (~ 15 s) became symptomatic, whereas only 7% of short apnea « 15 s) developed symptoms.
18 episodes motor activity continued throughout the apnea, whereas in the remaining five episodes apnea resolved after the cessation of motor activity. Discussion
To our knowledge, this is the first study in which apnea during wakefulness has been quantified. The main finding of the present study is that movement-related Results apnea is far more frequent than previA total of 352 apneic events were noted ously recognized. In the population studin these 10preterm infants. One hundred- ied, a third of all apneic episodes that fifty episodes of apnea occurred during resulted in hypoxia (~ 80% oxygen satusleep. In the remaining 202 episodes ration) or bradycardia (~ 1oo/min) ocarousal preceded the onset of apnea; mo- curred after arousal. Gross motor activtor activity continued during these epi- ity can occur during certain sleep states sodes, at least until apnea began. After (REM sleep), but it is brief and not susarousal breathhold breathing pattern was tained unless arousal occurs. An episode typically seen prior to apnea onset. Al- lasting 1.5 s or longer of increased EMG most all apneic episodes weremixed (n = activity with a coincidental decrease of 142) or obstructive (n = 53); only a few at least two sleep stages is considered were central (n = 7). Most of the apneas arousal in adults (9, 10). The pattern of were short: 181 were less than 15 s, and sustained motor activity for greater than only 21 were> 15 s, Onset of apnea oc- 5 s described here is indicative of arouscurred within 15 s of arousal in greater al. Although most episodes of arousals than 60070 of these episodes (with 41 % were transient, in some episodes motor of episodes beginning within 5 s); only activity continued, leading to wakefulin less than 20% did it occur after 30 s ness. It must be noted that the onset of apnea often occurred early in the course (figure 1). The vast majority of episodes associat- of motor activity; in fact, approximately ed with arousal resolved without produc- 40% of apnea occurred within 5 s of ing any significant cardiorespiratory se- arousal. Association between arousal quelae. However, hypoxia or bradycardia and apnea onset does not necessarily developed during 18of these episodes in imply causality, but it certainly raises six infants; this accounts for 7% of short the question. Apnea in preterm infants is believed apneas and 24% of long apneas (figure 2). Approximately 12.9% apnea that be- to occur exclusively during sleep; in fact, gan within 15s of arousal became symp- it is reported to occur more frequently tomatic, whereas only 1.4% of episodes in active or rapid eye movement sleep that began after 15 s produced any car- than in quiet sleep (11). There is no mendiopulmonary sequelae (figure 3). In tion of apnea occurring while awake in other words, essentially all symptomatic preterm infants in any of the standard apnea associated with arousal (17 of 18) textbooks of neonatology (4-6). Morebegan within 15 s of arousal, with 11 over, because arousal from sleep is beepisodes beginning within 5 s. In 13 of lieved to play an important role in the
,-,
OOMMEN P. MATHEW, CECIL K. THOPPIL, and MARY BELAN
Introduction
Apnea is a frequent and important clinical problem in preterm infants, its incidence being inverselyrelated to the gestational age (1, 2). It is generally believed that all the episodes of apnea occur during sleep. Recovery from obstructive apnea in adults is generally associated with arousal from sleep. Although the role of arousal in the termination of apnea in premature infants has not been investigated, the intervention of gentle and, occasionally, vigorous stimulation to terminate apnea is based on the premise that increased afferent stimuli induce arousal and resumption of breathing. Occurrence of apnea during wakefulness in preterm infants, unassociated with feeding, has been documented (3), but the frequency and characteristics of apnea that occur during wakefulness or immediately after arousals have not been established. For that matter, even the recent editions of the textbooks of neonatology or reviews on apnea of prematurity do not even mention the occurrence of apnea while awake (2, 4-6). The aim of the present study was to investigate this phenomenon in a group of apneic preterm infants. Methods A group of 10 preterm infants with clinical apnea form the basis of this report. They had a mean birth weight (± SD) of 1,237 ± 223 g and a gestational age of 31.1 ± 1.1 wk. These infants werestudied at a postnatal age of 10.2 ± 7.2 days. Six had recovered from respiratory illness (five transient tachypnea and one respiratory distress syndrome), and none of the infants was receiving supplemental oxygen at the time of the study. Five infants were receiving theophylline and one infant was receiving both theophylline and doxapram at the time of the study. These infants werepart of a study investigating the role of arousal in the resolution of apnea, which was approved by the institutional review board. Informed parental consent was obtained for all infants prior to the study, and no sedation was used in any of these infants. 842
SUMMARY It is generally believed that all apneic episodes In preterm infants occur during sleep. Because occurrence of apnea during wakefulness has been documented in these infants, we Investigated its frequency and characteristics In 10 premature Infants using slmultsneous polygraphic and video recordings. Behavioral arousal with motor activity preceded the onset of apnea In 202 episodes. Approximately 60% of episodes began within 15 s of arousal. Whereas most of the episodes were short and asymptomatic, hypoxia and/or bradycardia developed during 18apneic spells, and these episodes accounted for a third of all apneic episodes that resulted In bradycardia or hypoxia in these infants studied. Essentially, all movement-related apneas (17of 18) began within 15 s of arousal; motor activity continued throughout apnea In 13, whereas apnea resolved after cessation of motor activity in the remaining five. The main finding of the present study is that movementrelated apnea is far more frequent than previously recogniZed. This finding Is important In the clinical management of preterm Infants with apnea. AM REV RESPIR DIS 1991; 144:842-844
Each infant was videotaped and polygraphically monitored for 1 to 2 h after feeding. The infants were studied in a quiet room while lying supine under a radiant warmer. The skin temperature was servo-controlled at 36 to 36.50 C. Movements of the chest and abdomen weremonitored by inductive plethysmography (Respitrace'";Ardsley, NY), oxygen saturation and heart rate were measured by transcutaneous pulse oximetry (Nellcor Inc., Hayward, CA), and EKG using standard EKG electrodes and nasal airflow with a nasal screen flowmeter were monitored as previously described (7). Oral breathing was monitored by the sampling catheter of the CO 2 analyzer placed in front of the mouth. These parameters weredisplayed on an oscilloscope and recorded on a Gould eightchannel recorder (Gould Instruments, Cleveland, OH). The analog signals wereprocessed through multiplexors (Vettor Co, Rebersberg, PA) and stored on the HI-FI channels of the same videotapes in which the video images were being recorded. Auditory observations such as crying and other vocalizations as well as arousals induced by extraneous stimuli were noted on the polygraph paper by the investigators at the time of the study. The apneic episodes in the preterm infants were identified from the polygraphic recording by the absence of airflow for 5 s or greater. Apneas Iasting e 15 s were termed long apneas, whereas those lasting < 15 s were termed short apneas. Apneas were classified as central if the abdominal and chest wall movements were absent, obstructive if respiratory efforts persisted, and mixed if a combination of central and obstructive components was present. Bradycardia was defined
as a heart rate ~ 100 beats/min, and R-R interval from the EKG channel formed the basis of these calculations. Hypoxia was defined as an O 2 saturation of ~ 800/0. In order to eliminate falsely low pulse oximeter readings during movement, heart rate recorded from the pulse oximeter was compared with the heart rate derived from the EKG channel. If the heart rate obtained from the pulse oximeter was lower than the heart rate from the EKG by more than 5 beats/min, values of oxygen saturation observed during that period werenot included for further analysis. Bradycardia and hypoxia were considered to be associated with apneas if they occurred during the apneic period or within 5 s of the termination of the apnea. Playback of the tape allowed us to viewthe behavior of the infant during each episode. If movement was observed at apnea onset, the event was further reviewed to determine whether the criteria for behavioral arousal was met prior to the onset Of apnea as described before (8). If behavioral arousal criteria were not met, then the episode was coded as sleeprelated apnea. Arousal was defined as the occurrence of cry, opening of eyes (> 5 s), or
(Received in originalform September 14, 1990and in revised form February 5, 1991) 1 From the Departments of Pediatrics and Physiology & Biophysics, University of Texas Medical Branch, Galveston, Texas. 2 Correspondence and requests for reprints should be addressed to Oommen P. Mathew, M.D., Department of Pediatrics, University of TexasMedical Branch, Galveston, TX 77550.
MOVEMENT-RELATED APNEA IN PREMATURE INFANTS
843 %
% r-..
o :§
o
40
r---
% o
~ 30
Q)
c;
c 20
Q.
30
~
r---
o 20
~
Ql
C
«
o :;:; 20 o
:g
E
E o
o Q. E
-
1t 10 o
Q.
«
o
0-
10
E >,
>,
if)
(5
5-15
15-30
}30
Seconds
Fig. 1. Time interval between beginning of arousal and apnea onset.
concurrent occurrence of sustained movements (> 5 s) of head and neck, trunk, and extremities. Typically, these activities consisted of flexion or extension of the head and neck, slow writhing movement of the trunk, and flexion-extension movements of the extremities. These activities were similar to the "squirming" movements described by AbuOsba and coworkers (3). The relationship between beginning of the motor activity, apnea onset, and resolution were determined in all but nine episodes.
10
(f)
O'---..L-----'------'--------'-------'---'---5-10 10-15 2.15 Apnea Duration (sec) Fig. 2. Relationship between apnea duration and development of symptoms; 24% of long apnea (~ 15 s) became symptomatic, whereas only 7% of short apnea « 15 s) developed symptoms.
18 episodes motor activity continued throughout the apnea, whereas in the remaining five episodes apnea resolved after the cessation of motor activity. Discussion
To our knowledge, this is the first study in which apnea during wakefulness has been quantified. The main finding of the present study is that movement-related Results apnea is far more frequent than previA total of 352 apneic events were noted ously recognized. In the population studin these 10preterm infants. One hundred- ied, a third of all apneic episodes that fifty episodes of apnea occurred during resulted in hypoxia (~ 80% oxygen satusleep. In the remaining 202 episodes ration) or bradycardia (~ 1oo/min) ocarousal preceded the onset of apnea; mo- curred after arousal. Gross motor activtor activity continued during these epi- ity can occur during certain sleep states sodes, at least until apnea began. After (REM sleep), but it is brief and not susarousal breathhold breathing pattern was tained unless arousal occurs. An episode typically seen prior to apnea onset. Al- lasting 1.5 s or longer of increased EMG most all apneic episodes weremixed (n = activity with a coincidental decrease of 142) or obstructive (n = 53); only a few at least two sleep stages is considered were central (n = 7). Most of the apneas arousal in adults (9, 10). The pattern of were short: 181 were less than 15 s, and sustained motor activity for greater than only 21 were> 15 s, Onset of apnea oc- 5 s described here is indicative of arouscurred within 15 s of arousal in greater al. Although most episodes of arousals than 60070 of these episodes (with 41 % were transient, in some episodes motor of episodes beginning within 5 s); only activity continued, leading to wakefulin less than 20% did it occur after 30 s ness. It must be noted that the onset of apnea often occurred early in the course (figure 1). The vast majority of episodes associat- of motor activity; in fact, approximately ed with arousal resolved without produc- 40% of apnea occurred within 5 s of ing any significant cardiorespiratory se- arousal. Association between arousal quelae. However, hypoxia or bradycardia and apnea onset does not necessarily developed during 18of these episodes in imply causality, but it certainly raises six infants; this accounts for 7% of short the question. Apnea in preterm infants is believed apneas and 24% of long apneas (figure 2). Approximately 12.9% apnea that be- to occur exclusively during sleep; in fact, gan within 15s of arousal became symp- it is reported to occur more frequently tomatic, whereas only 1.4% of episodes in active or rapid eye movement sleep that began after 15 s produced any car- than in quiet sleep (11). There is no mendiopulmonary sequelae (figure 3). In tion of apnea occurring while awake in other words, essentially all symptomatic preterm infants in any of the standard apnea associated with arousal (17 of 18) textbooks of neonatology (4-6). Morebegan within 15 s of arousal, with 11 over, because arousal from sleep is beepisodes beginning within 5 s. In 13 of lieved to play an important role in the
,-,
