Polygraphic Studies of Kitten Development: Respiratory Rate and Variability during Sleep- Waking States

MICHAEL STEVENSON DENNIS J. McGINTY

Departments of Anatomy and Psychology University of California, Los Angela and

Veterans Administration Hospital Sepulveda, California

The developmental course of respiration rate and variability during sleep states and waking was measured in chronically prepared kittens. Kittens had higher respiration rates during active sleep (AS) as compared to quiet sleep (QS) at all ages, with rates declining developmentaUy in both sleep states. Compared to waking, respiration rate and variability were decreased during sleep. The decrease was greatest in the youngest animals and during QS. Minute-to-minute respiratory variability declined from 10 to 40 days of age for each state, whereas breath-to-breath variability declined during the same period only during QS. Respiratory variability was higher in AS than QS in older kittens. Heart rate and variability were found to be correlated with respiratory parameters only during QS. These observations support the hypothesis that the control and development of respiration during sleep is achieved by different processes in QS and AS.

In adult mammals, respiratory rates in slow wave sleep or quiet sleep (QS) are regular, whereas in REM sleep or active sleep (AS) they tend to be more variable (Jouvet, Michel, & Mounier, 1960; Snyder, Hobson, Morrison, & Goldfrank, 1964). Related studies of heart rate, and many other autonomic, somatic, and central nervous system parameters have led to the conception of QS as a quiescent state and AS as a phasically variable state (Jouvet, 1967; McGinty, Harper, & Fairbanks, 1974). Recently, interest in respiratory changes during sleep has been renewed by the description of a group of clinical disorders of children and adults characterized by sleep-related respiratory dysfunction (Guilleminault, Tilkian, & Dement, 1976). In addition, victims of the Sudden Infant Death Syndrome (SIDS) exhibit histological changes indicating a history of hypoxemia which is presumed to have resulted from sleep-related respiratory dysfunctions (Naeye, 1973, 1974). Infants at risk for SIDS

Reprint requests should be sent to Michael Stevenson, Ph.D., Neurophysiology Research, V. A. Hospital, Sepulveda, California 91343, U.S.A.

Received for publication 28 February 1977 Revised for publication 31 May 1977 Developmental Psycho biology, 1 1 ( 5 ) : 393403 ( 1 9 78) @ 1978 by John Wiley & Sons, Inc. 001 2-1630/78/0011-0393$01 .OO

394 STEVENSON AND McGINTY

exhibit altered respiratory patterns (Guilleminault, Peraita, Souquet, & Dement, 1975 ; Hoppenbrouwers, Hodgman, Harper, McCinty, & Sterman, 1977). The elucidation of sleep-related respiratory abnormalities in infants must be preceded by an analytical description of normal development. The present report describes development of respiratory rate and variability in the kitten.

Newborn human infants have higher and more variable respiration rates during AS when compared to QS (Bolton & Herman, 1974; Hathorn, 1974; Prechtl, Akiyama, Zinkm, & Kerr Grant, 1968). In a longitudinal study extending beyond the human newborn period, Hoppenbrouwers, Harper, Hodgman, Sterman, and McGinity (in press) showed that respiratory rate and variability significantly decreased between 1 week and 6 months of age. They also found that respiratory rate and variability were highest in AS at all ages tested.

In the kitten, Schwieler (1968) found that respiration rate decreases with age in resting animals, reaching the adult level by 6 weeks of age. However, no sleep state analysis was reported. Hoppenbrouwers (1974), using a partial longitudinal paradigm, also studied respiratory rate development from 1 to 6 weeks of age in kittens. In contrast to human infants, no significant respiratory rate differences were found between AS and QS at any of the ages studied. A significant drop in respiratory rate during sleep occurred between 4 and 6 weeks of age. Respiratory variability during AS was always greater than during QS at all ages studied, but no overall age trend was noted.

The objectives of this study were to investigate the development of respiratory rate and variability as a function of age, sleep state, gender, and birth weight, and to provide statistical descriptions of respiratory parameters from 12-hr polygraphic recordings. In order to approximate the kittens’ natural sleep environment, monitoring was carried out in the presence of littermates and mother.

Methods

Eighteen kittens (Felis dornesticu) were derived from a laboratory breeding colony. The kittens were obtained from 13 different females and weighed from 78-122 g at birth. Subjects were randomly distributed into 3 age groups (10-day, 20-day, and 40-day), each containing 3 males and 3 females. Subsequently, a female in the 10-day group was replaced by a male due to a poor respiratory recording.

Forty-eight to 72 hr prior to recording, the kittens were prepared with chronic electrodes under sodium pentobarbitol anesthesia (7-day kittens, 15 mg/kg; 17-day kittens, 20 mg/kg; 37-day kittens, 40 mg/kg) supplemented with halothane and topical lidocaine when needed. Two pairs of stainless steel machine screws (1.6 mm) with attached lead wires were threaded through the skull on each side of the midline over the sensorimotor cortex for EEG recording. Enamel insulated stainless steel wires (175 pm) with exposed 2 mm loops at the tip were slipped subcutaneously through a hypodermic needle (22 gauge) to a position immediately posterior to the eye for recording eye movement potentials (EOG) and into the dorsal neck muscles to record neck electromyogram (EMG). Electrocardiographic (EKG) recordings were obtained with ultraminiature gold plated fish hooks (No. 24), which were hooked into the intercostal muscles just above the sternum, 5-10 mm on either side of the midline. A

RESPIRATORY RATE AND VARIABILITY 395

2nd pair of hooks were placed at the same rostro-caudal level on the back, left and right of the vertebral column. Lead wires were threaded subcutaneously to the top of the head. Respiratory rate was measured by means of a .33-mm diameter, 100,000-ohm Veco (51A30) thermistor suspended in front of 1 naris. The thermistor and its lead wires were supported and protected by a stainless steel tube (19 gauge) that was fitted along the bridge of the nose. All lead wires were routed subcutaneously and collected on top of the head and soldered to an ultraminiature Winchester 14-pin male connector. The EEG electrodes and electrical connections were covered with Caulk Grip Cement and Teets Denture Material. Kittens were given Combiotic (20,000 units) immediately and 48 hr postoperatively. A topical antibiotic was also used.

Females and their litters (mean size, 4.8 kittens) were housed together in standard cages (91 x 76 x 76 cm) except for the duration of surgery and the 12-hr period of polygraphic recording. A 14-hr light/lO-hr dark cycle was maintained in the colony with the dark period beginning at 1900 hours. Food and water were available ad libitum.

Polygraphic recordings were carried out in a shielded wooden chamber (58 x 71 x 81 cm) with an observation window (61 x 41 cm). A shallow plastic box containing bedding material was provided for the mother and litter. The mother had access through an elevated opening to an adjacent cage (61 x 46 x 81 cm) containing food, water, resting sites, and a litter pan. Temperature was maintained at 25 ? 3"C, and room illumination was reduced on the same schedule as the maternal colony. During recordings, the kitten was attached to a light, flexible, shielded cable suspended from the top of the recording cubicle by an elastic material. This allowed the kitten unrestricted movement and nursing opportunities.

Continuous 12-hr polygraphic records were begun between 1900 and 2000 hours. The polygraphic signals, left and right EEG, EOG, EMG, ECG, and respiration, together with a Slow Code-Type B with a 30-sec interval, were recorded on a Grass Polygraph (model 78B) and simultaneously stored on a Sangamo analog tape recorder (model 4). An IRIG E time code with a 10-sec interval was recorded on tape only.

The polygraphic paper record was classified by state by identifying each successive minute as Waking (W), QS, AS,Mixed, or Unknown. Waking was identified by the presence of tonic and phasic EMG activity and eye movements at a rate of more than 2/min. Active sleep was also recognized by phasic EMG activity and eye movements, but was distinguished from W by the absence of tonic EMG activity and a much lower respiratory rate. Quiet sleep was identified by the absence of phasic EMG activity and eye movement (frequency less than 2/min). (See Fig. 1.) The onset and offset of W and AS epochs were specified at the time of the 1st and last eye movements or phasic EMG events. The kitten EEG does not change clearly with sleep state prior to 3 weeks of age and thus was not useful in state scoring 10- and 20-day kittens.

After states were identified as specified above, most minutes were readily classified accordingly. However, some minutes contained state transitions and consequently included 2 or more states. The rules for classification of minutes containing 2 or more states were: Any minute containing more than 24 sec of W, QS, or AS and less than 24 sec of the other sleep state and less than 12 sec W was classified as the dominant state. Any minute containing more then 12 sec and less than 24 sec W or more than 24 sec of both QS and AS was called Mixed. Minutes containing artifacts severe enough to preclude scoring were called Unknown.

396 STEVENSON AND MtGINTY

, . + L ’ P

Fig. 1. Examples of polygraphic recordings during QS, AS, and W from 110-, 20-, and 40day-old kittens. The polygraphic data shown are left and right EEG, EOG, EMG, EKG, and respiration (inspiration down). Each age period is represented by 1 kitten.

Respiratory rate and variability during sleep were calculated by a computer program that detected respiratory peaks and troughs and then computed the interval between them. Median intervals and interquartile ranges of interbreath intervals were printed for each minute of the 12-hr recording by a PDP 12 computer (Bannett, Mason, & Harper, 1976). A comparison with hand counted minutes by traned observers indicated a high reliability for data under 60 breaths/min. The W respiratory rates were too rapid for this program and thus were hand scored.

The rates were sampled from 3 periods out of the 12-hr recording: Hours 2-4, 5-7, and 8-10. Ten 1-min samples for each state (W, QS, and AS) were randomly chosen from each of the 3 periods for a total of 30 min for each state. Only uninterrupted state periods lasting a minimum of 1 min were used. Respiratory rates, minute-to-minute variability (standard deviations of 10-min samples), and breath-to- breath variability (interquartile ranges of interbreath intervals) during each state at 10, 20, and 40 days of age were analyzed statistically as described in the results below. (Minute-to-minute variability is a measure of the fluctuation of tonic respiratory rate during sustained periods. Breath-to-breath variability indicates the short-tcrm stability of the respiratory rate. The independence of these measures is indicated by differences in their developmental course as shown below.)

RESPIRATORY RATE AND VARIABILITY 397

In addition, the relationships between these respiratory data and the birth weight, body weight at time of recording, and heart rate were assessed. Heart rate and variability were calculated by a computer program that detected the R wave of the ECG signal. The median and interquartile range of R-R intervals for each minute were then measured and converted to beats/min (Harper, Hoppenbrouwers, Sterman, McGinty, & Hodgman, 1976).

Results

Respiratory Rate and Minute-to-Minute Variability

Kitten respiratory rate means and minute-to-minute variabilities (standard deviations) were subjected to 3 (age) x 3 (state) x 3 (sample time) analyses of variances with repeated measures on the state and sample time factors. No significant gender differences were noted for any state or age so this factor level was collapsed. As the between subject variances of the data tended to be proportional to the means, and the distributions positively skewed, the numbers were first transformed using a log(X f 1) transform (Winer, 1962). Specific mean comparisons were based on the Duncan Multiple Range Test.

Main rate effects for age (F = 15.75, df = 2/15, p < .01) and state (F = 217.91, df = 2/30, p < .OOl) were significant. Mean respiration rate significantly declined from 10 to 40 days of age for each state ( p < .OOl). All state comparisons at each age were significant ( p < .01). (See Fig. 2.) A significant drop in respiratory rate was noted 0, < .05) from the 1st sample period (Hours 2-4) to the 2nd sample period (Hours 5-7) across all states and ages.

Main minute-to-minute variability effects for age (F = 5.71, d f = 2/15, p < .05) and state (F = 122.39, df = 2/30, p < .001) were significant. Variability significantly declined from 10 to 40 days of age for each state (QS: p < ,001; AS: p < .05; and W: p < .01). Waking showed significantly more variability ( p < .001) compared to QS or AS at all ages. Minute-to-minute variability differences between AS and QS were found only at 40 days of age (AS > QS, p < .001). No sample time differences were noted.

Breath-to-Breath Variability

Interquartile ranges of end expiratory intervals were subjected to a 3 (age) x 2 (state) x 3 (sample time) analysis of variance during sleep with repeated measures on the state and sample time factors. The W ranges were too short to be measured practically and thus were not included in this analysis.

Main effects for sleep state (F = 43.45, df = 1/15, p < .OOl) and sample time (F = 6.21, df 2/30, p < .O l ) were significant. Mean breath-to-breath variability declined ( p < .05) from 10 to 40 days of age for QS. No developmental change was noted in AS. At 20 and 40 days of age AS showed higher variability than QS ( p < .Ol ) . (See Fig. 3. ) A significant increase in variability was found ( p < .01) from the 1st sample period (Hours 2 4 ) to the rest of the night across all states and ages.

398 STEVENSON AND McGINTY

I10

W k 3 z Zz 90 L1I W a W t ' 70 Z 0 c Q Ir a v, w 50 LT

Z Q W 2

3 0

...._ f-. .............

I I I 10 20 40

AGE IN DAYS

Fig. 2. Mean respiration rate/min with respect to age and sleep state. Each data point rcpresents 6 kittens over 3 sample periods. Ninety-five percent confidence limits for the means arc shown.

50

w c3 z 5 cc 0 I 40 G w cc 0 t I

m

; 30 cc z m

5 > 20

I I

J

Ib i0 4b AGE IN DAYS

Fig. 3 . Mean breath-to-breath variability in seconds with respect to age and sleep state. Each data point represents 6 kittens over 3 sample periods. Ninety-five percent confidence limits for the means are shown.

RESPIRATORY RATE AND VARIABILITY 399

Body Weight and Heart Rate

Partial Pearson Correlations computed between birth weight or body weight at time of recording and various respiratory measures revealed significant inverse relations (r = -.49, p < .05) between birth weight and respiratory rate during AS, and between body weight a t time of recording and respiratory minute-to-minute variability (AS: r = -.58; W: r = -.054; p < .05). Partial Pearson Correlations computed between heart and respiratory rate and variability measures revealed significant direct relations only for QS (Rate: r = .68, p < .01; Variability: r = .56, p < .05).

Discussion

Our results show that kittens have higher respiration rates during AS as compared to QS at all ages. A recent finding from our laboratory that respiratory rate in human infants declines developmentally (Hoppenbrouwers et al., in press) was also observed in kittens. Both breath-to-breath and minute-to-minute variability were lower in QS than AS in 40-day-old kittens, but were not different during these sleep states in 10-day-old kittens. Minute-to-minute respiratory variability declined from 10 to 40 days of age for each state, whereas breath-to-breath variability declined during the same period only for QS. These findings are summarized in Table 1.

Our findings do not agree with Hoppenbrouwers’ (1974) results in the kitten: she found no respiratory rate differences between AS and QS at any age between 1 to 6 weeks, a significant drop in respiratory rate only after 4 weeks of age, and no age trend in respiratory variability for either sleep state. The feature accounting for the discrepancies between the studies appears to be the large between-subject variance prior to 4 weeks of age in the Hoppenbrouwers experiment. In that study, kittens were monitored after removal from mother and littermates for periods of 2-4 hr, beginning

TABLE 1. Summary of Significant Differences in Respiration as a Function of Age arid State.

Age (in days)

Measure 10 20 40 ~~

Mean Rate Minute-to-Minute Variability Breath-to-Breath Variability QS = AS QS < AS QS < AS

QS < AS < W QS = AS < W

QS < AS < W QS = AS < W

QS < AS < W QS < A S < W

State

QS AS W

10 > 20 > 40 10 > 40 10 = 20 = 40

Mean Rate Minute-to-Minu te Variability Breath-to-Breath Variability 10 > 40

10 > 20 > 40 10 > 20 > 40

10 = 20 > 40 10 = 20 > 40

-

400 STEVENSON AND McCINTY

at varying times during the day. Summary statistics were derived from kitten groups whose age range spanned as much as 2 weeks. By contrast, our data were derived from uniform age groups, and from 12-hr recordings obtained at the same time of day. Thus, the recording of kittens in isolation, and different sampling, and data summary procedures may have yielded increased variability and precluded finding any significant age or state variations.

A comparison of waking with sleep states indicates a striking decrease in respiratory rate and variability during sleep. This decrease was greatest in the youngest animals and was greater during QS than AS. In other words, the relative depression in excitatory drive on respiration rate between W and sleep appears to be greatest in the young animals and in QS. However, the assumption that respiratory rate alone is an adequate datum for comparing state effects on ventilation is unwarranted. Orem, Netick, and Dement (1976) have shown in adult cats that a decrease in AS tidal volume may offset the increased respiration rate such that minute volume is reduced. This type of measurement must be extended to the kitten.

The increased respiratory rate in AS may reflect changes in the function of respiratory control systems. In the adult human the response to elevated inspired COz exhibits a lower threshold and more sensitive response curve in AS when compared with QS (Biilow & Ingvar, 1961), although a recent study in human infmts failed to find any significant state difference in CO, responsiveness (Fagenholtz, O’Connell, & Shannon, 1976). Moreover, AS may reduce susceptibility to hypoxic respiratory dysfunction. Baker and McGinty (1 976) have reported that respiratory d’epression in the kitten resulting from experimental hypoxia is expressed maximally in QS anti is reversed in AS. Studies of Ondine’s curse in infants have also shown respiratory depression in QS that is reversed in AS (Shannon, Marsland, Could, Callahan, Todres, & Dennis, 1976).

Many kinds of data support the concept that AS is a state characterized by nervous system excitation. During the final third of the gestational period in the fetal lamb (115-147 days), Dawes, Fox, Leduc, Liggins, and Richards (1972) have distinguished 3 types of behavior resembling QS, AS, and W. Only AS was accompanied by bursts of rapid irregular respiratory movements (fetal breathing), although spontaneous respiratory activity was present from 40 days gestation onwards and was not state related. In the adult cat, AS has been characterized as a state of nervous system arousal because of the greatly elevated rate of neuronal discharge in most brain sites, as well as augmented cerebral metabolism (McGinty el d., 1974). Thus, augmented respiratory rate during AS may be a result of elevated nonspecific excitation.

On the other hand, AS-related respiratory variability may itself represent a change in regulation. Apnea rates are higher during AS than QS in infants (Hoppenbrouwers, Harper, Hofmann, Sterman, & Hodgman, 1977). Sleep related respiratory dysfunction characterized by upper airway obstruction, as in the Pickwickian syndrome, is sometimes exacerbated in AS (Coccagna, Mantovani, Brignani, Parchi, & Lugaresi, 1972). Diminution of tonic somatic muscle tone (Pompeiano, 1067) supporting the upper airway may account for increased upper airway obstruction in this state. In the cat, intercostal EMG activity associated with breathing is depressed in AS (Parmeggiani & Sabattini, 1972) and breathing is accomplished by diaphragmatic activity alone in this state.

RESPIRATORY RATE AND VARIABILITY 401

The QS breath-to-breath variability declined between 10 and 40 days of age whereas AS variability was unchanged. This result may be an additional manifestation of differential respiratory control in QS and AS as reviewed above. That is, variability in the 2 states may represent 2 different processes that exhibit different developmental changes. The QS state itself undergoes very significant developmental changes: the overall proportion of sleep having the properties of QS increases rapidly, individual epochs of QS increase in duration, the intrusion of phasic somatic activities decreases, and cortical EEG patterns characterizing adult QS gradually emerge (McGinty, Stevenson, Hoppenbrouwers, Harper, Sterman, & Hodgman, 1977). These changes have been described as representing the coalescence of the QS state (Dreyfus-Brisac, 1970; Hoppenbrouwers & Sterman, 1975; Parmelee & Stern, 1972).

Heart rate and variability were found to be correlated with respiratory rate only during QS. The interaction between respiration and heart rate that characterizes sinus arrhythmia is also most clearly seen in Q S (Harper, Hoppenbrouwers, Sterman, Hodgman, & Bannett, 1975). These observations suggest that reflexes that mediate cardio-respiratory interaction may be more stable during Q S . "h i s interpretation is consistent with the report that spontaneous vagal nerve recordings show diminished activity in AS compared with QS (Leichnetz, 1972).

In summary, all of these observations suggest the hypothesis that the control of respiration is accomplished by different neuronal processes or systems in AS as compared with QS.

Notes

This research was supported by the National Institute of Child Health and Human Development Contract No. NO 1-HD-4-2810 and the Veterans Administration.

References

Baker, T., and McGinty, D. (1976). Hypoxic conditioning in kittens: Sleep state and cardiorespiratory responses. Sleep Res., 5 : 40.

Bannett, D., Mason, J., and Harper, R. M. (1976). A computer procedure for quantifying respiratory activity during sleep and waking. Sleep Res., 5 : 197.

Bolton, D. P. G., and Herman, S. (1974). Ventilation and sleep state in the new-born. J. Physiol. (Lond.), 240: 67-77.

Biilow, K., and Ingvar, D. H. (1961). Respiration and state of wakefulness in normals studied by spyrography, capnography and EEG. Acta. Physiol. Scand., 51: 230-238.

Coccagna, C., Mantovani, M., Brignani, F., Parchi, C., and Lugaresi, E. (1972). Continuous recording of the pulmonary and systemic arterial pressure during sleep in syndromes of hypersornnia with periodic breathing. Bull. Physiopathol. Respir. (Nancy), 8 : 1159-1 172.

Dawes, G. S., Fox, H. E., Leduc, B. M., Liggins, G. C., and Richards, R. T. (1972). Respiratory movements and rapid eye movement sleep in the fetal lamb. J. Physiol. (Lond.J, 220:

Dreyfus-Brisac, C. (1970). Ontogenesis of sleep in human prematures after 32 weeks of conceptual

Fagenholz, S. A., O'Connell, K., and Shannon, D. C. (1976). Chemoreceptor function and sleep state

119-143.

age. Dev. Psychobiol., 3: 91-121.

in apnea. Pediatrics, 58: 31-36.

402 STEVENSON AND McCINTY

GuiUeminault, C., Peraita, R., Souquet, M., and Dement, W. C. (1975). Apneas during slecp in infants: Possible relationship with sudden infant death syndrome. Science, I YO:

Guilleminault, C., Tikian, A., and Dement, W. C. (1976). The sleep apnea syndromes. Ann. Rev. Med., 27: 465484.

Harper, R. M., Hoppenbrouwers, T., Sterman, M. B., Hodgman, J., and Bannctt, D. (1975). Quantification of cardiorespiratory interaction during sleep and waking states in infants. SlceLJ Res., 4 : 128.

Harper, R. M . , Hoppenbrouwers, T., Sterman, M. B., McGinty, D. J., and Hodgman, J . (1976). Polygraphic studies of normal infants during the first six months of life. I. Heart rate and variability as a function of state. Pediat. Res., 10: 945-951.

Hathorn, M . K. S. (1974). The rate and depth of breathing in new-born infants in different sleep states. J. Physiol. (Lond.), 243: 101-113.

Hoppenbrouwers, T. (1974). The development of functional states in the kitten. Unpublished doctoral dissertation, University of California, Los Angeles.

Hoppenbrouwers, T., and Sterman, M . B. (1975). Development of state patterns in the kittcn. Exp. Neurol., 49: 822-838.

Hoppenbrouwers, T., Harper, R. M., Hodgman, J. E., Sterman, M. B., and McCinty, D. J. (in press). Polygraphic studies of normal infants during the first six months of l i l c . 11: Respiratory rate and variability as a function of state. Pediaf. Rcs.

Hoppenbrouwers, T., Harper, R. M., Hofmann, E., Sterman, M. B., and Hodgman, J. (1977). Developmental polygraphic studies of normal infants during the first six months of Life. 111: Incidence of apnea and periodic breathing. Pediatrics, 60: 418425.

Hoppenbrouwers, T., Hodgman, J. E., Harper, R. M., McGinty, D. J., and Sterman, M. B. (1977). Respiratory rates and apnea in infants at high and low risk for sudden infant death syndrome. Clin. Res., 25: 189A.

677-679.

Jouvet, M. (1967). Neurophysiology of the states of sleep. Physiol. Rev., 47: 117-177. Jouvet, M., Michel, F., and Mounier, D. (1960). Analyse electroencephalographique comparde du

sommeil physiologique chez le Chat et chez 1'Homme. Rev. Neurol., 103: 189-205. Leichnetz, G . R. (1972). Relationship of spontaneous vagal activity to wakefulness and sleep in the

cat. Exp. Neurol., 35: 194-210. McGinty, D. J . , Harper, R. M., and Fairbanks, M. K. (1974). Neuronal unit activity and the control

of sleep states. In E. D. Weitzman (Ed.), Advances in Sleep Research. New York: Spectrum Publications. Pp. 173-216.

McGinty, D. J., Stevenson, M., Hoppenbrouwers, T., Harper, R. M., Sterman, M. B.. and Hodginan, J. (1977). Polygraphic studies of kitten development: Sleep state patterns. Dev. Psychohid., 10: 455-469.

Naeye, R. L. (1973). Pulmonary arterial abnormalities in the sudden infant death syndrome. N. Engl. J. Med., 289: 1167-1 170.

Naeye, R. L. (1974). Hypoxemia and the sudden infant death syndrome. Science, 186: 837-838. Orem, J. , Netick, A., and Dement, W. D. (1976). Breathing during sleep and wakefulness in the cat.

Sleep Rex, 5 : 50. Parmeggiani, P. L., and Sabattini, L. (1972). Electromyographic aspects of postural, respiratory and

thermoregulatory mechanisms in sleeping cats. Electroencephalogr. Clin. Neurophysiol., 33:

Parmelee, A. H., and Stern, E. (1972). Development of states in infants. In C. D. Clcmente, D. P. Purpura, and F. E. Mayer (Eds.), Sleep and the Maturing Nervous System New York: Academic Press. Pp. 199-228.

Pompeiano, 0. (1967). The neurophysiological mechanisms of the postural and motor events during desynchronized sleep. In S. S. Kety, E. V. Evarts, and H. L. Williams (Eds.), Slriep and Altered States o f Consciousness. Baltimore: Williams and Wilkins. Pp. 35 1423.

Prechtl, H. F. R., Akiyama, Y., Zinkin, P., and Kerr Grant, D. (1968). Polygraphic studies of the full term newborn. I : Technical aspects, and qualitative analysis. In M. C. 0. Bax and R. C . MacKeith (Eds.), Studies in Infancy. Clinics in Developmental Medicine, No. 2% London: Heinemann. Pp. 1-25.

Schwieler, G. H. (1968). Respiratory regulation during postnatal development in cats and rabbits and some of its morphological substrate. Acta Physiol. Scand. [Suppl.], 304: 1-123.

1-13.

RESPIRATORY RATE AND VARIABILITY 403

Shannon, D. C., Marsland, D. W., Gould, J . B., Callahan, B., Todres, D., and Dennis, 3. (1976).

Snyder, F., Hobson, J. A., Morrison, D. R., and Goldfrank, F. (1964). Changes in respiration, heart

Winer, B. J. (1962). Statistical Principles in Experimental Design. New York: McGraw-Hill. Pp.

Central hypoventilation during quiet sleep in two infants. Pediatrics, 57: 342-346.

rate, and systolic blood pressure in human sleep. J. Appl. Physiol., 19: 417-422.

2 18-222.