Arousal responses to hypertension effect of sinoaortic denervation
in lambs:
ROSEMARY S. C. HORNE, NICOLE D. DE PREU, PHILIP J. BERGER, AND ADRIAN M. WALKER Centre for Early Human Development, Monash Medical Centre, Mqnash University, Clayton, Victoria 3168, Australia
HORNE, ROSEMARY S. C., NICOLE D. DE PREU, PHILIP J. BERGER, AND ADRIAN M. WALKER. Arousal responses to hypertension in lambs: effect of sinoaortic denervation. Am. J. Physiol. 260 (Heart Circ. Physiol. 29): H1283-H1289, 1991.Newborn lambs were subjected to hypertensive stimuli of lmin duration to examine features of hypertension-induced arousal from sleep. Reflex mechanisms involved were studied by performing the same tests after sinoaortic denervation (SAD). In intact lambs, hypertension increased the probability of arousal from both quiet sleep (QS) and rapid-eye-movement (REM) sleep. Hypertension resulted in arousal in 51% (QS) and 50% (REM) of tests. Arousal time was significantly longer in REM (29.3 t 0.9 s, mean k SE) than in QS (22.6 t 0.6 s, P < 0.01). Arterial oxygen saturation (SO& and partial pressure of oxygen (PO& measured at the point of arousal, or after 60 s if arousal failed to occur, were unchanged from control values. After SAD hypertension did not increase the probability of arousal. Arousals significantly decreased (P < 0.001) to 31% (QS) and 10% (REM). These findings indicate that acute hypertension, mediated via arterial baroreceptors, is a potent stimulus for arousal. In intact lambs, the- arousal probability increased and arousal time decreased with increasing stimulus strength (l-30 mmHg), but the arousal time difference between QS and REM remained constant. Consideration of these findings in terms of a simple baroreflex threshold model suggests that the slower response in REM sleep is explained by slower neural processes after the achievement of a critical arousal input rather than by a higher threshold for baroreceptor input in this state. sleep; baroreceptors
FROM SLEEP in response to respiratory stimuli has recently come under intensive investigation, particularly with the recognition that arousal may have important survival advantages (3, 23) and with suggestions that failure to arouse from sleep during an asphyxial episode may be one cause of the sudden infant death syndrome (11, 12). Consequently it is now well-known that respiratory stimuli such as hypoxia (5, 7, 21, 24), hypercapnia (22, 27), airway occlusion (1, 4, 13), and laryngeal stimulation (26) induce arousal from sleep. However, little attention has been given to the possibility that circulatory changes initiate arousal. Arousal responses to circulatory stimuli need to be understood, particularly as many respiratory challenges produce concomitant cardiovascular changes. Acute hypoxemia, for example, may be associated with pronounced hypotension or hypertension, depending on the severity of the
AROUSAL
0363-6135/91
$1.50 Copyright
hypoxic insult (16). Furthermore, systemic arterial blood pressure increases before arousal from sleep during rapidly developing hypoxemia (8). In a previous study (14), it has been shown that acute hypotension, mediated by arterial baroreceptors, affects arousal from sleep. Similarly, it has been reported that acute hypertension elicits arousal from sleep in newborn lambs (9), but a role for baroreceptors in mediating this response has not yet been established. This study, therefore, examined the effects of acute hypertension and sought evidence for baroreceptor involvement in the arousal response by contrasting intact and sinoaorticdenervated (SAD) animals. Additionally, by analyzing the effects of stimulus strength on the arousal response, we examined the alternative possibilities that an increased arousal threshold (4, 5, 15, 19, 21, 26, 27) or a neural latency or inhibition in the arousal process (7) explains the prolonged delay before arousal that is typical of the rapid-eye-movement (REM) state. METHODS
Experimental procedure. Experiments were conducted on seven newborn lambs that had been instrumented under general anesthesia (l-3% halothane) in utero as described previously (14); five were between 117 and 125 days gestation, whereas one was at 134 days and another was at 136 days gestation. An inflatable constrictor cuff of internal diameter of 8 mm (Hazen Everett) was placed around the descending aorta distal to the ductus arteriosus, with care being taken not to damage the adjacent recurrent laryngeal, vagus, a.nd phrenic nerves as well as the branches of the thoracic sympathetic nerves. A saline-filled balloon-ended catheter was placed into the pleural cavity for measurement of intrapleural pressure, and electrocardiogram (ECG) electrodes were sewn onto the chest wall beneath the skin. A small incision was made in the scalp, and bilateral silver electrocorticogram (ECoG) electrodes were implanted on the dura mater over the parietal cortex. Another small skin incision was made in the midline of the neck, and electromyogram (EMG) electrodes were implanted into the rhomboideus muscle. Incisions were closed, and the catheters and leads were coiled and secured in “pockets” formed by apposing adjacent skin folds with continuous silk sutures ensuring that delivery was unimpeded at term. The fetus was then returned to the uterus, and the uterus and the abdomen
0 1991 the American
Physiological
Society
H1283
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
H1284
ACUTE
HYPERTENSION
AND
AROUSAL
were closed. The ewe was treated with Z-ml intramuscular doses of antibiotics: 500 mg of dihydrostreptomycin sulfate and 500 mg of procaine penicillin (Streptopen, Glaxovet, Australia) for 7 days followed by (in mg) 470 benethamine penicillin, 250 procaine penicillin, and 300 benzylpenicillin (Triplopen, Glaxovet, Australia) every 4th day until delivery. All lambs were delivered naturally between 142 and 151 days gestation (mean 145 days) with birth weights of 3.3-4.3 kg (mean 3.8 kg). Within 2-36 h after birth, catheters (Tygon, 0.9 mm OD) were inserted into the left carotid artery and jugular vein under local anesthesia (Xylocaine 1%). Catheters were checked daily for patency and removal of any clots with an aseptic technique and were refilled with heparin-saline (1,000 U/ml). Lambs received Streptopen (1 ml im) for the first 7 days after delivery and Triplopen (1 ml im) every fourth day thereafter. Studies commenced after a minimum 2-h recovery period from catheter insertion. Carotid artery, jugular venous, and intrapleural pressure catheters were attached to pressure transducers (Hewlett-Packard 1280C) zeroed at midthoracic level and connected to carrier amplifiers (Hewlett-Packard 8805B). Arterial pressure was recorded in pulsatile form and averaged electronically (Baker Institute Signal Processor). ECoG and EMG electrodes were connected to alternating-current (AC) amplifiers (Neomedix NT ll4), and. the signals were filtered through continuously variable filter units (Neurolog NL 125). The ECoG signal was processed with a band pass of 0.8-50 Hz, and the EMG signal was processed with a band pass of 80 Hz-2 kHz. The ECG was preamplified (Hewlett-Packard Bioelectric Amplifier 881lA), and the signal was used to derive instantaneous heart rate (Neomedix NT 122). All signals were displayed on an eight-channel chart recorder (HewlettPackard 7758A). Lambs were comfortably suspended in a net hammock so that their hooves just touched the floor. Behavioral state was determined by observation of the animal’s behavior (eye movements, facial and ear twitching, respiratory and heart rate pattern, and head and body movements) complemented with ECoG and EMG recordings. The state was defined as quiet sleep (QS) when for at least 1 min the ECoG showed high-voltage (80100 ,uV peak-to-peak) predominantly slow waves (1-3 Hz) in association with regular quiet breathing, closed eyes, no REM, few body or limb movements, and regular heart rate and blood pressure. REM sleep was signified by low voltage (IO-20 PV peak-to-peak) predominantly fast waves on the ECoG, accompanied by irregular breathing, REM under closed or slightly open lids, twitching movements of the face, ears, and limbs, and variable heart rate and blood pressure. We distinguished phasic and tonic components of the REM state by observation and reference to the heart rate and blood pressure records; we confined our study to tonic REM. Phasic REM was characterized by transient episodes of tachycardia and hypertension, occurring in association with jerking movements of limbs and body, clearly superimposed on the more stable heart rate and blood pressure baseline of tonic REM. The quiet awake (QA) state was identified when low voltage fast waves on the
IN
NEWBORN
LAMBS
ECoG occurred in the presence of open eyes, quiet regular respiration, and few body or limb movements. Acute hypertension was produced by inflating the constrictor cuff around the dorsal aorta. Cuff inflations were not imposed until the animal had been in a defined behavioral state for at least 1 min. Tests were not repeated until after blood pressure and heart rate had returned to control levels (minimum time between stimuli was 2 min). During sleep, inflations were maintained until arousal occurred or, in the absence of arousal, for 1 min. During QA, inflations were maintained for 30 s to allow quantification of the baroreflex. Recording sessions comprised a l-h control period in which there were no cuff inflations to allow definition of the spontaneous probability of arousal and a l-h test period during which cuff inflations were performed. The order of control and test periods was reversed on sequential days. Behavioral arousal from sleep to wakefulness was judged to occur when the lamb opened its eyes and moved its head, accompanied by a change in the ECoG and onset of neck muscle EMG activity (Fig. 1). Blood samples were taken for arterial and venous blood gas measurements (Corning 168 pH/blood gas analyzer) and hemoglobin and oxygen saturation determinations (Radiometer OSMZ Hemoximeter). Samples were collected before and during constrictor cuff inflations in each behavioral state. After 12-25 days of study, five of the lambs underwent a bilateral SAD under halothane-oxygen anesthesia (l3% halothane) as previously described (14). Studies of these animals recommenced after 24- to 48-h postoperative recovery. Data analysis. Control values for mean arterial pres140 AP mmHg 40 140 MAP
mmHg
r
40 1 250 HR
----L
-
bpm 0 050
PP mmHg -50 500 ECoG
EMG
pv
pv
FIG. 1. Illustration of arousal responses to acute newborn lamb. AP, arterial pressure; MAP, mean HR, heart rate; PP, intrapleural pressure; ECoG, and EMG, nuchal electromyogram. Hypertension, tion of a constrictor cuff placed around descending caused an arousal (arrow) as signified by a change of neck muscle EMG activity associated with raising arousal time is significantly longer in rapid-eye-movement than in quiet sleep (QS).
hypertension in a arterial pressure; electrocorticogram; produced by inflaaorta (thick bar), of ECoG and onset of head. Note that (REM) sleep
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
ACUTE
HYPERTENSION
AND
sure (MAP), heart rate (HR), and respiration rate were measured from the chart record 10 s before the onset of the hypertensive stimulus. Hypertension values were measured 10 s after the onset of the stimulus as blood pressure rose rapidly and reached a stable level within two to three heart beats. Baroreflex gain was calculated as the change in HR for a given change in MAP. Values were normalized by expressing the changes between preand poststimulus values as a percentage change from the basal level. The effect of behavioral state (QA, QS, REM) on cardiorespiratory variables was examined by multiway analysis of variance (ANOVA) based on regression analysis (SPSS-X, SPSS, Chicago, IL). In all cases the between-animal differences were first taken into account to remove any bias toward highly sampled animals, and results are expressed as mean values for the “average” animal. Regression was followed by two-way ANOVA, and, if a significant F value was obtained, paired Student’s t tests were performed to identify the states that differed. Arousal time was defined as the time delay between the onset of constrictor cuff inflation and arousal from sleep. Arousal times were compared between sleep states with a one-way ANOVA. The probability of arousal during hypertension in each sleep state was calculated as the percentage of tests in which arousal occurred within 60 s of imposing the stimulus. The probability of spontaneous arousal was also determined for each sleep state during the control period, calculated as the percentage of spontaneous arousals during each minute of the sleep epoch (14). Not all balloon inflations resulted in arousal within 60 s. These were deemed nonarousals, and the percentage of nonarousals was calculated for each sleep state. Differences in probability of arousal between 1) control and test situations, 2) sleep states, and 3) intact and SAD lambs were analyzed by stratified Chi-square analysis, as were percent arousals in each sleep state. In the stratified analysis, observed and expected frequencies were calculated for each animal and pooled for calculation of Chi-square to avoid any bias toward highly sampled animals. To investigate the effect of stimulus strength on arousal, data in each sleep state were grouped in lommHg MAP bins. The effect on arousal time was tested by l-way ANOVA followed by a Newman-Keuls Studentized Range Statistic to test which bins of MAP differed from each other. The effect of stimulus strength on probability of arousal was tested by Chi-square analysis. RESULTS
Intact Lambs Seven lambs were studied on a total of 35 different days ranging from the day of birth to 20 days of age. A total of 92 inflations was given in QS, 129 in tonic REM sleep, and 80 in the QA state. Blood gas data (pooled for QA, QS, and REM) in intact animals (Table 1) are within published normal limits for our laboratory (2, 14) and other laboratories (9, 25).
AROUSAL
IN
NEWBORN
H1285
LAMBS
1. Baseline arterial blood gas values (QS, &A, and REM) in intact and SAD newborn lambs TABLE
so,, % PH
Pco~, mmHg PO,, mmHg HCO:%, mmol/l TCOz, mmol/l BE, mmol/l
Intact
SAD
94.1t1.0 7.37t0.01 42.0tl.0 88.9t3.3 23.5t0.6 24.7t0.6 -0.2t0.6
79.3tZ.O" 7.33t0.05'" 53.1t1.2'" 71.4t2.8* 27.3t0.4* 28.9t0.4"" 2.7t0.4'"
Values are means * SE; n no. of measurements. QS, quiet sleep; REM, rapid eye movement; ‘&A, quiet awake; TCOZ, total carbon dioxide; BE, base excess. * P < 0.001, Intact (n = 22) vs. sinoaorticdenervated (SAD; n = 24).
QS
FIG. 2. Probability of arousal from sleep in intact newborn lambs. Control probabilities (open bars) were obtained from periods of sleep in which no hypertensive tests were given. Control probability was calculated as percentage of spontaneous arousals from sleep during each minute of sleep epoch. Hypertension probability (solid bars), calculated as percentage of tests in which arousal occurred within 1 min of imposing hypertensive stimulus, is significantly increased above control probability in both sleep states. For both control and hypertension, P < 0.001.
Hemodynamics. Control values for mean arterial pressure were not significantly different among the three behavioral states (Table 2). Although control values for heart rate were significantly lower in REM than in QS or QA (P < O.Ol), heart rate decreased after cuff inflation by a similar value in all states. Baroreflex gain was not significantly different between behavioral states (QA: 1.16 t 0.07; QS: 1.57 t 0.07; and REM: 1.49 t 0.06). Sleep and arousal. The median duration of sleep epochs was 2-3 min for QS and 3-4 min for REM. As most sleep epochs were 4 min or less, the study was confined to this period. Hypertension increased the probability of arousal from both QS (Chi-square = 13.8, P < 0.001) and REM (Chi-square = 35.9, P < 0.001, Fig. 2). The time delay between the onset of the hypertensive test and the point of arousal (arousal time) was significantly longer in REM (29.3 t 0.9 s) than in QS (22.6 t 0.6 s, P < 0.01). Inflations produced an arousal in 51% of tests in QS and 50% of tests in REM; these values are not significantly different. Age had no effect on arousal time in either sleep state, nor did it affect arousal probability in QS [weeh 1(51%); weeks II and III (52%)]. However, arousal probability increased in REM (P < 0.02) after the first week of life [week 1(43%); weeks II and III (66%)]. The time the lam.b had been in a particular sleep state had no effect on arousa1 tim .e in either sleep state. Arousal
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
H1286 TABLE
ACUTE
2. Cardiorespiratory
HYPERTENSION
measurements
Respiration rate, breaths/min Control Hypertension Change ~ Values
are means
t SE for average
(80)
87.9t0.8 +15.2t0.4
(80) (80)
NS P < 0.01 P < 0.01
203t2 165t3 -37tl
(80) (80) (80)
62.7k2.3 68.7t2.5 +4.6t0.6
(38) (38) (38)
g
60
Ii -
40
0 L
s
20
0
m 0 E
0
P < 0.05 P c 0.01 NS
196zk2 162t2 -33t1
(92) (92) (92)
P < 0.001 P < 0.001 NS
NS P < 0.01 NS
64.0tl.7
59.8tl.9
(69) (65)
-2.8k0.5
(65)
NS NS NS
0 0
LI MAP
QS vs. REM
+12.9&0.4
0
10
lambs
NS NS NS
0
a *O 1
8
LAMBS
REM
71.6t0.6 84.4t0.7
+12.8&0.3 185k2
146t2 -38tl
&A vs. REM
(129) (129) (129)
NS P c 0.01 P < 0.01
(129) (129) (129)
P < 0.001 P < 0.001 NS
(95) (89) (89)
60.0tl.4
60.9tl.6 -0.3kO.4
NS P < 0.05 P < 0.01
no. of tests in parentheses.
100
is
NEWBORN
(92) (92) (92)
71.6kO.7 84.5t0.8
probability decreased in REM (P < 0.02) after lambs had been asleep for longer than 3 min but was unaffected by the length of the QS epoch. To investigate the possibility that baroreflex gain and arousability were correlated, baroreflex gain (combined for QS and REM) was plotted against percent arousal for individual animals on each experimental day. However, no relationship between arousability and gain could be demonstrated (r = -0.1643). Probability of arousal in QS increased with increasing stimulus strength from 39% at l-10 mmHg to 100% at 21-30 mmHg (P < 0.001); arousal probability in REM also tended to increase from 54% at l-10 mmHg to 77% at 21-20 mmHg, but this did not quite reach significance (Fig. 3). Arousal time decreased with increasing stimulus strength in both sleep states. When the arousal time at l-10 mmHg was compared with that at 21-30 mmHg there was asignificant decrease in both QS (P < O.Oz> and REM (P < 0.05) (Fig. 4). Respiratory changes. Control values for respiration rate were not different between states (Table 2). Respiration rate was affected variably by hypertension with rates declining in QS, remaining unchanged in REM, and increasing in QA. In four lambs on eight experimental days, blood gases (n = 26) were taken at the point of arousal or after 60 s if arousal failed to occur. Samples collected in QS and REM were pooled. Arterial So2 and Po2 values, measured
a
IN
QS
QA vs. QS
72.8t0.8
animal;
AROUSAL
in seven intact newborn
&A Arterial pressure, mmHg Control Hypertension Change Heart rate, beats/min Control Hypertension
AND
20
30
(mmHg)
FIG. 3. Illustrates effect of stimulus strength (change in mean arterial pressure; AMAP) on probability of arousal in QS (open circles) and REM sleep (closed circles). Probability of arousal increased with increasing stimulus strength in both sleep states.
40
T
vz
in lambs:
ROSEMARY S. C. HORNE, NICOLE D. DE PREU, PHILIP J. BERGER, AND ADRIAN M. WALKER Centre for Early Human Development, Monash Medical Centre, Mqnash University, Clayton, Victoria 3168, Australia
HORNE, ROSEMARY S. C., NICOLE D. DE PREU, PHILIP J. BERGER, AND ADRIAN M. WALKER. Arousal responses to hypertension in lambs: effect of sinoaortic denervation. Am. J. Physiol. 260 (Heart Circ. Physiol. 29): H1283-H1289, 1991.Newborn lambs were subjected to hypertensive stimuli of lmin duration to examine features of hypertension-induced arousal from sleep. Reflex mechanisms involved were studied by performing the same tests after sinoaortic denervation (SAD). In intact lambs, hypertension increased the probability of arousal from both quiet sleep (QS) and rapid-eye-movement (REM) sleep. Hypertension resulted in arousal in 51% (QS) and 50% (REM) of tests. Arousal time was significantly longer in REM (29.3 t 0.9 s, mean k SE) than in QS (22.6 t 0.6 s, P < 0.01). Arterial oxygen saturation (SO& and partial pressure of oxygen (PO& measured at the point of arousal, or after 60 s if arousal failed to occur, were unchanged from control values. After SAD hypertension did not increase the probability of arousal. Arousals significantly decreased (P < 0.001) to 31% (QS) and 10% (REM). These findings indicate that acute hypertension, mediated via arterial baroreceptors, is a potent stimulus for arousal. In intact lambs, the- arousal probability increased and arousal time decreased with increasing stimulus strength (l-30 mmHg), but the arousal time difference between QS and REM remained constant. Consideration of these findings in terms of a simple baroreflex threshold model suggests that the slower response in REM sleep is explained by slower neural processes after the achievement of a critical arousal input rather than by a higher threshold for baroreceptor input in this state. sleep; baroreceptors
FROM SLEEP in response to respiratory stimuli has recently come under intensive investigation, particularly with the recognition that arousal may have important survival advantages (3, 23) and with suggestions that failure to arouse from sleep during an asphyxial episode may be one cause of the sudden infant death syndrome (11, 12). Consequently it is now well-known that respiratory stimuli such as hypoxia (5, 7, 21, 24), hypercapnia (22, 27), airway occlusion (1, 4, 13), and laryngeal stimulation (26) induce arousal from sleep. However, little attention has been given to the possibility that circulatory changes initiate arousal. Arousal responses to circulatory stimuli need to be understood, particularly as many respiratory challenges produce concomitant cardiovascular changes. Acute hypoxemia, for example, may be associated with pronounced hypotension or hypertension, depending on the severity of the
AROUSAL
0363-6135/91
$1.50 Copyright
hypoxic insult (16). Furthermore, systemic arterial blood pressure increases before arousal from sleep during rapidly developing hypoxemia (8). In a previous study (14), it has been shown that acute hypotension, mediated by arterial baroreceptors, affects arousal from sleep. Similarly, it has been reported that acute hypertension elicits arousal from sleep in newborn lambs (9), but a role for baroreceptors in mediating this response has not yet been established. This study, therefore, examined the effects of acute hypertension and sought evidence for baroreceptor involvement in the arousal response by contrasting intact and sinoaorticdenervated (SAD) animals. Additionally, by analyzing the effects of stimulus strength on the arousal response, we examined the alternative possibilities that an increased arousal threshold (4, 5, 15, 19, 21, 26, 27) or a neural latency or inhibition in the arousal process (7) explains the prolonged delay before arousal that is typical of the rapid-eye-movement (REM) state. METHODS
Experimental procedure. Experiments were conducted on seven newborn lambs that had been instrumented under general anesthesia (l-3% halothane) in utero as described previously (14); five were between 117 and 125 days gestation, whereas one was at 134 days and another was at 136 days gestation. An inflatable constrictor cuff of internal diameter of 8 mm (Hazen Everett) was placed around the descending aorta distal to the ductus arteriosus, with care being taken not to damage the adjacent recurrent laryngeal, vagus, a.nd phrenic nerves as well as the branches of the thoracic sympathetic nerves. A saline-filled balloon-ended catheter was placed into the pleural cavity for measurement of intrapleural pressure, and electrocardiogram (ECG) electrodes were sewn onto the chest wall beneath the skin. A small incision was made in the scalp, and bilateral silver electrocorticogram (ECoG) electrodes were implanted on the dura mater over the parietal cortex. Another small skin incision was made in the midline of the neck, and electromyogram (EMG) electrodes were implanted into the rhomboideus muscle. Incisions were closed, and the catheters and leads were coiled and secured in “pockets” formed by apposing adjacent skin folds with continuous silk sutures ensuring that delivery was unimpeded at term. The fetus was then returned to the uterus, and the uterus and the abdomen
0 1991 the American
Physiological
Society
H1283
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
H1284
ACUTE
HYPERTENSION
AND
AROUSAL
were closed. The ewe was treated with Z-ml intramuscular doses of antibiotics: 500 mg of dihydrostreptomycin sulfate and 500 mg of procaine penicillin (Streptopen, Glaxovet, Australia) for 7 days followed by (in mg) 470 benethamine penicillin, 250 procaine penicillin, and 300 benzylpenicillin (Triplopen, Glaxovet, Australia) every 4th day until delivery. All lambs were delivered naturally between 142 and 151 days gestation (mean 145 days) with birth weights of 3.3-4.3 kg (mean 3.8 kg). Within 2-36 h after birth, catheters (Tygon, 0.9 mm OD) were inserted into the left carotid artery and jugular vein under local anesthesia (Xylocaine 1%). Catheters were checked daily for patency and removal of any clots with an aseptic technique and were refilled with heparin-saline (1,000 U/ml). Lambs received Streptopen (1 ml im) for the first 7 days after delivery and Triplopen (1 ml im) every fourth day thereafter. Studies commenced after a minimum 2-h recovery period from catheter insertion. Carotid artery, jugular venous, and intrapleural pressure catheters were attached to pressure transducers (Hewlett-Packard 1280C) zeroed at midthoracic level and connected to carrier amplifiers (Hewlett-Packard 8805B). Arterial pressure was recorded in pulsatile form and averaged electronically (Baker Institute Signal Processor). ECoG and EMG electrodes were connected to alternating-current (AC) amplifiers (Neomedix NT ll4), and. the signals were filtered through continuously variable filter units (Neurolog NL 125). The ECoG signal was processed with a band pass of 0.8-50 Hz, and the EMG signal was processed with a band pass of 80 Hz-2 kHz. The ECG was preamplified (Hewlett-Packard Bioelectric Amplifier 881lA), and the signal was used to derive instantaneous heart rate (Neomedix NT 122). All signals were displayed on an eight-channel chart recorder (HewlettPackard 7758A). Lambs were comfortably suspended in a net hammock so that their hooves just touched the floor. Behavioral state was determined by observation of the animal’s behavior (eye movements, facial and ear twitching, respiratory and heart rate pattern, and head and body movements) complemented with ECoG and EMG recordings. The state was defined as quiet sleep (QS) when for at least 1 min the ECoG showed high-voltage (80100 ,uV peak-to-peak) predominantly slow waves (1-3 Hz) in association with regular quiet breathing, closed eyes, no REM, few body or limb movements, and regular heart rate and blood pressure. REM sleep was signified by low voltage (IO-20 PV peak-to-peak) predominantly fast waves on the ECoG, accompanied by irregular breathing, REM under closed or slightly open lids, twitching movements of the face, ears, and limbs, and variable heart rate and blood pressure. We distinguished phasic and tonic components of the REM state by observation and reference to the heart rate and blood pressure records; we confined our study to tonic REM. Phasic REM was characterized by transient episodes of tachycardia and hypertension, occurring in association with jerking movements of limbs and body, clearly superimposed on the more stable heart rate and blood pressure baseline of tonic REM. The quiet awake (QA) state was identified when low voltage fast waves on the
IN
NEWBORN
LAMBS
ECoG occurred in the presence of open eyes, quiet regular respiration, and few body or limb movements. Acute hypertension was produced by inflating the constrictor cuff around the dorsal aorta. Cuff inflations were not imposed until the animal had been in a defined behavioral state for at least 1 min. Tests were not repeated until after blood pressure and heart rate had returned to control levels (minimum time between stimuli was 2 min). During sleep, inflations were maintained until arousal occurred or, in the absence of arousal, for 1 min. During QA, inflations were maintained for 30 s to allow quantification of the baroreflex. Recording sessions comprised a l-h control period in which there were no cuff inflations to allow definition of the spontaneous probability of arousal and a l-h test period during which cuff inflations were performed. The order of control and test periods was reversed on sequential days. Behavioral arousal from sleep to wakefulness was judged to occur when the lamb opened its eyes and moved its head, accompanied by a change in the ECoG and onset of neck muscle EMG activity (Fig. 1). Blood samples were taken for arterial and venous blood gas measurements (Corning 168 pH/blood gas analyzer) and hemoglobin and oxygen saturation determinations (Radiometer OSMZ Hemoximeter). Samples were collected before and during constrictor cuff inflations in each behavioral state. After 12-25 days of study, five of the lambs underwent a bilateral SAD under halothane-oxygen anesthesia (l3% halothane) as previously described (14). Studies of these animals recommenced after 24- to 48-h postoperative recovery. Data analysis. Control values for mean arterial pres140 AP mmHg 40 140 MAP
mmHg
r
40 1 250 HR
----L
-
bpm 0 050
PP mmHg -50 500 ECoG
EMG
pv
pv
FIG. 1. Illustration of arousal responses to acute newborn lamb. AP, arterial pressure; MAP, mean HR, heart rate; PP, intrapleural pressure; ECoG, and EMG, nuchal electromyogram. Hypertension, tion of a constrictor cuff placed around descending caused an arousal (arrow) as signified by a change of neck muscle EMG activity associated with raising arousal time is significantly longer in rapid-eye-movement than in quiet sleep (QS).
hypertension in a arterial pressure; electrocorticogram; produced by inflaaorta (thick bar), of ECoG and onset of head. Note that (REM) sleep
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
ACUTE
HYPERTENSION
AND
sure (MAP), heart rate (HR), and respiration rate were measured from the chart record 10 s before the onset of the hypertensive stimulus. Hypertension values were measured 10 s after the onset of the stimulus as blood pressure rose rapidly and reached a stable level within two to three heart beats. Baroreflex gain was calculated as the change in HR for a given change in MAP. Values were normalized by expressing the changes between preand poststimulus values as a percentage change from the basal level. The effect of behavioral state (QA, QS, REM) on cardiorespiratory variables was examined by multiway analysis of variance (ANOVA) based on regression analysis (SPSS-X, SPSS, Chicago, IL). In all cases the between-animal differences were first taken into account to remove any bias toward highly sampled animals, and results are expressed as mean values for the “average” animal. Regression was followed by two-way ANOVA, and, if a significant F value was obtained, paired Student’s t tests were performed to identify the states that differed. Arousal time was defined as the time delay between the onset of constrictor cuff inflation and arousal from sleep. Arousal times were compared between sleep states with a one-way ANOVA. The probability of arousal during hypertension in each sleep state was calculated as the percentage of tests in which arousal occurred within 60 s of imposing the stimulus. The probability of spontaneous arousal was also determined for each sleep state during the control period, calculated as the percentage of spontaneous arousals during each minute of the sleep epoch (14). Not all balloon inflations resulted in arousal within 60 s. These were deemed nonarousals, and the percentage of nonarousals was calculated for each sleep state. Differences in probability of arousal between 1) control and test situations, 2) sleep states, and 3) intact and SAD lambs were analyzed by stratified Chi-square analysis, as were percent arousals in each sleep state. In the stratified analysis, observed and expected frequencies were calculated for each animal and pooled for calculation of Chi-square to avoid any bias toward highly sampled animals. To investigate the effect of stimulus strength on arousal, data in each sleep state were grouped in lommHg MAP bins. The effect on arousal time was tested by l-way ANOVA followed by a Newman-Keuls Studentized Range Statistic to test which bins of MAP differed from each other. The effect of stimulus strength on probability of arousal was tested by Chi-square analysis. RESULTS
Intact Lambs Seven lambs were studied on a total of 35 different days ranging from the day of birth to 20 days of age. A total of 92 inflations was given in QS, 129 in tonic REM sleep, and 80 in the QA state. Blood gas data (pooled for QA, QS, and REM) in intact animals (Table 1) are within published normal limits for our laboratory (2, 14) and other laboratories (9, 25).
AROUSAL
IN
NEWBORN
H1285
LAMBS
1. Baseline arterial blood gas values (QS, &A, and REM) in intact and SAD newborn lambs TABLE
so,, % PH
Pco~, mmHg PO,, mmHg HCO:%, mmol/l TCOz, mmol/l BE, mmol/l
Intact
SAD
94.1t1.0 7.37t0.01 42.0tl.0 88.9t3.3 23.5t0.6 24.7t0.6 -0.2t0.6
79.3tZ.O" 7.33t0.05'" 53.1t1.2'" 71.4t2.8* 27.3t0.4* 28.9t0.4"" 2.7t0.4'"
Values are means * SE; n no. of measurements. QS, quiet sleep; REM, rapid eye movement; ‘&A, quiet awake; TCOZ, total carbon dioxide; BE, base excess. * P < 0.001, Intact (n = 22) vs. sinoaorticdenervated (SAD; n = 24).
QS
FIG. 2. Probability of arousal from sleep in intact newborn lambs. Control probabilities (open bars) were obtained from periods of sleep in which no hypertensive tests were given. Control probability was calculated as percentage of spontaneous arousals from sleep during each minute of sleep epoch. Hypertension probability (solid bars), calculated as percentage of tests in which arousal occurred within 1 min of imposing hypertensive stimulus, is significantly increased above control probability in both sleep states. For both control and hypertension, P < 0.001.
Hemodynamics. Control values for mean arterial pressure were not significantly different among the three behavioral states (Table 2). Although control values for heart rate were significantly lower in REM than in QS or QA (P < O.Ol), heart rate decreased after cuff inflation by a similar value in all states. Baroreflex gain was not significantly different between behavioral states (QA: 1.16 t 0.07; QS: 1.57 t 0.07; and REM: 1.49 t 0.06). Sleep and arousal. The median duration of sleep epochs was 2-3 min for QS and 3-4 min for REM. As most sleep epochs were 4 min or less, the study was confined to this period. Hypertension increased the probability of arousal from both QS (Chi-square = 13.8, P < 0.001) and REM (Chi-square = 35.9, P < 0.001, Fig. 2). The time delay between the onset of the hypertensive test and the point of arousal (arousal time) was significantly longer in REM (29.3 t 0.9 s) than in QS (22.6 t 0.6 s, P < 0.01). Inflations produced an arousal in 51% of tests in QS and 50% of tests in REM; these values are not significantly different. Age had no effect on arousal time in either sleep state, nor did it affect arousal probability in QS [weeh 1(51%); weeks II and III (52%)]. However, arousal probability increased in REM (P < 0.02) after the first week of life [week 1(43%); weeks II and III (66%)]. The time the lam.b had been in a particular sleep state had no effect on arousa1 tim .e in either sleep state. Arousal
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
H1286 TABLE
ACUTE
2. Cardiorespiratory
HYPERTENSION
measurements
Respiration rate, breaths/min Control Hypertension Change ~ Values
are means
t SE for average
(80)
87.9t0.8 +15.2t0.4
(80) (80)
NS P < 0.01 P < 0.01
203t2 165t3 -37tl
(80) (80) (80)
62.7k2.3 68.7t2.5 +4.6t0.6
(38) (38) (38)
g
60
Ii -
40
0 L
s
20
0
m 0 E
0
P < 0.05 P c 0.01 NS
196zk2 162t2 -33t1
(92) (92) (92)
P < 0.001 P < 0.001 NS
NS P < 0.01 NS
64.0tl.7
59.8tl.9
(69) (65)
-2.8k0.5
(65)
NS NS NS
0 0
LI MAP
QS vs. REM
+12.9&0.4
0
10
lambs
NS NS NS
0
a *O 1
8
LAMBS
REM
71.6t0.6 84.4t0.7
+12.8&0.3 185k2
146t2 -38tl
&A vs. REM
(129) (129) (129)
NS P c 0.01 P < 0.01
(129) (129) (129)
P < 0.001 P < 0.001 NS
(95) (89) (89)
60.0tl.4
60.9tl.6 -0.3kO.4
NS P < 0.05 P < 0.01
no. of tests in parentheses.
100
is
NEWBORN
(92) (92) (92)
71.6kO.7 84.5t0.8
probability decreased in REM (P < 0.02) after lambs had been asleep for longer than 3 min but was unaffected by the length of the QS epoch. To investigate the possibility that baroreflex gain and arousability were correlated, baroreflex gain (combined for QS and REM) was plotted against percent arousal for individual animals on each experimental day. However, no relationship between arousability and gain could be demonstrated (r = -0.1643). Probability of arousal in QS increased with increasing stimulus strength from 39% at l-10 mmHg to 100% at 21-30 mmHg (P < 0.001); arousal probability in REM also tended to increase from 54% at l-10 mmHg to 77% at 21-20 mmHg, but this did not quite reach significance (Fig. 3). Arousal time decreased with increasing stimulus strength in both sleep states. When the arousal time at l-10 mmHg was compared with that at 21-30 mmHg there was asignificant decrease in both QS (P < O.Oz> and REM (P < 0.05) (Fig. 4). Respiratory changes. Control values for respiration rate were not different between states (Table 2). Respiration rate was affected variably by hypertension with rates declining in QS, remaining unchanged in REM, and increasing in QA. In four lambs on eight experimental days, blood gases (n = 26) were taken at the point of arousal or after 60 s if arousal failed to occur. Samples collected in QS and REM were pooled. Arterial So2 and Po2 values, measured
a
IN
QS
QA vs. QS
72.8t0.8
animal;
AROUSAL
in seven intact newborn
&A Arterial pressure, mmHg Control Hypertension Change Heart rate, beats/min Control Hypertension
AND
20
30
(mmHg)
FIG. 3. Illustrates effect of stimulus strength (change in mean arterial pressure; AMAP) on probability of arousal in QS (open circles) and REM sleep (closed circles). Probability of arousal increased with increasing stimulus strength in both sleep states.
40
T
vz
