Nasopharyngitis and Prolonged Sleep Apnea

(1)

ARTICLES

Nasopharyngitis

and Prolonged

Sleep

Apnea

Alfred Steinschneider, M.D., Ph.D.

From the Department of Pediatrics, State University of New York, Upstate Medical Center, Syracuse

ABSTRACT. The effect of nasopharyngitis on the

simulta-neous occurrence of prolonged sleep apnea ( 20 seconds in duration) was studied in 26 infants managed at home on an apnea monitor. During the observation period, these in-fants had a total of 69 illnesses which appeared to represent an upper respiratory tract inflammatory process. In general, the daily frequency of prolonged apneic episodes was sig-nificantly greater during nasopharyngitis when compared to comparable time intervals immediately prior to and fol-lowing the illness. In addition, there was a decrease in the frequency of apneic episodes with increasing postnatal age until the episodes finally ceased to occur during the illness-related intervals. Apneic episodes ceased to occur at an ear-her age for the before- and after-illness intervals than for the time interval during which there were clinical symp-toms. Thus, it would appear that infants go through an age-related phase wherein prolonged apnea occurs during naso-pharyngitis but not when free of illness.

The implications of these results for the management of infants having prolonged sleep apnea are discussed. In view of the hypothesis that prolonged sleep apnea is part of the physiological process resulting in the sudden infant death syndrome, these results also provide for the prediction that infants who suddenly die in association with nasopharyn-gitis would do so, in general, at a later age than those who succumb when free of an upper respiratory tract infiamma-tory process. Pediatrics, 56:967-971, 1975, PROLONGED AP-NEA, SLEEP, NASOPHARYNGITIS, SUDDEN DEATH.

The frequency and severity of apneic and cyanotic episodes in the neonatal age period have prompted a number of investigative

ef-forts.’’ Illingworth,2 reporting on experience at

the Jessop Hospital for Women (Sheffield) from

1949 to 1956, noted that approximately 1.5% of all liveborn babies had such episodes and, fur-thermore, that 60% of these infants had been

born prematurely. Consistent with this general

observation, Daily, Klaus, and Meyer7 observed

that about 25% of pre-term infants had one or more prolonged apneic episodes during the first two weeks of life. These episodes are important not only because of their association with other pathologic processes, such as pulmonary, central nervous system, and metabolic disorders, but also because they may be severe enough to result in death. In addition, concern has been expressed that prolonged apneic episodes might, in

surviv-ors, lead to central nervous system damage.9 While most studies of this phenomenon have focused on the first month of life, Stevens’#{176} and

nshn’ have documented that pro-longed apneic and cyanotic episodes can occur in older infants otherwise free of a serious illness. Of the nine infants described in their two reports, six

died of the sudden infant death syndrome (SIDS). This evidence of the serious potential of apneic episodes requires that physicians give greater consideration to their etiology, medical manage-ment, and to the elucidation of those variables which alter the frequency and severity of such

occurrences.

Upper respiratory tract inflammatory disease is one condition which appears to alter the fre-quency of prolonged apneic episodes. All of the patients described by Stevens’#{176}had clinical signs of at least an upper respiratory tract infec-tion when the severe apneic episodes were first

(Received April 7; revision accepted for publication June 2, 1975.)

Supported in part by grant 1-R01-HD07460 from the Na-tional Institute of Child Health and Human Development.

ADDRESS FOR REPRINTS: Department of Pediatrics, State University of New York Upstate Medical Center, 750

(2)

observed. Steinschneider’1 also presented data indicating that, in general, the occurrence of nasopharyngitis (NP) was accompanied by an in-crease in the number of prolonged sleep apneic episodes, although this effect became less appar-ent with age. Since this association is based on a limited number of observations, the present study was undertaken to reexamine the relationship be-tween NP and the simultaneous occurrence of prolonged sleep apneic episodes.

SUBJECTS AND METHODS

Data were obtained from 26 infants managed at home on an impedance-type apnea monitor (Air Shields). Prior to employing the monitor, all infants were observed for a complete nap in a sleep laboratory and a continuous recording was obtained of respiration, cardiac rate, and eye movements. Nine infants had been referred for evaluation because of prolonged sleep apnea and transient cyanosis noted in the hospital or at home. The remaining were participants in one of several on-going laboratory sleep studies: normal full-term infants (No. = 8); low-birthweight

in-fants (No. = 3); and neonatal siblings born into a family where there had been a previous instance of SIDS (No. = 6). The latter three groups of in-fants (No. = 17) were placed on apnea monitors

at home solely because the laboratory data were interpreted as indicating high risk for prolonged sleep apnea. The subjects in the present study consisted of 9 boys and 17 girls, 25 Caucasians and 1 Black, and 17 normal full-term and 9 low-birthweight

(

2,500 gm) infants. The median age of infants when first placed on the home

monitor was 28 days. Although the age range was

3 to 223 days, the monitor had been employed initially by age 77 days in all but one infant. The

monitor was discontinued by 257 days of age in one half of the babies and in all by 491 days.

In addition to being shown how to operate the monitor (set to alarm when apnea persisted for 20 seconds), parents were instructed to record on a form sheet the data and time the alarm sounded, the condition of the baby when first seen after the alarm sounded (behavioral state, respiratory activity, skin color), and the status of the monitor (absolute impedance, sensitivity). In addition, they were asked to keep a daily log in-dicating any major change in the baby’s activity and the presence of any illness, as well as its de-scription. Parents were informed that this infor-mation was being requested for the purpose of determining if a relationship existed between the occurrence of prolonged sleep apnea and any change in the baby’s routine or environment.

During the period of instruction, extreme care was taken to avoid suggesting any specific rela-tionships.

A review of the form sheets revealed a total of 69 illnesses of at least three consecutive days’ du-ration (maximum duration, 18 days), described by the parents as a “cold,” cough, or “runny” or stuffed nose. Such an illness was presumed to rep-resent an upper respiratory tract inflammatory process and will be referred to as NP. During the period of observation, nine infants had one NP, eight infants had two, three infants had three, one infant had four, four infants had six, and one infant had seven.

The average daily frequency of prolonged ap-nea was determined during each NP as well as during a comparable interval of time immedi-ately prior to and following the NP. The before and after values were averaged to obtain a single rate (pre and post [PP]) for comparison with the NP rate and as a measure of the daily frequency when the infant was otherwise asymptomatic. The data were examined initially by considering all apneic episodes of 20 seconds or longer. This included a large number of apparently seif-termi-nating episodes since by the time the parent ar-rived at the crib-side the infant was seen to be breathing and the alarm had ceased. To exclude the possibility that the apnea monitor transiently failed to register shallow respirations (thereby leading to a false alarm), a second analysis was performed which included only those episodes in which the infant was found to be apneic and the parents applied external stimulation (e.g. shak-ing).

RESULTS

Over a period representing 1,713 subject days,

parents reported that the alarm was triggered 1,537 times while the infant was sleeping. In 50% of the episodes (769), the apnea persisted and stimulation was provided.

Apneic Episodes of 20 Seconds or Longer

Prolonged apnea occurred in 19 of the 26 in-fants during 47 of 69 illness-related periods (NP and PP interval). In 42 of these periods, the dai-ly frequency during NP exceeded the PP rate, whereas in 5 the NP daily rate was less (Sign test,22 p < .01). The average difference (NP minus PP) in daily rates (1.08) was significantly different from zero (t = 4.56;

df=

46) at less than

the 0.01 level of probability.

In order to determine if there existed an age-relatedness to the PP occurrence of prolonged apnea, as well as to compare the influence of NP

at Viet Nam:AAP Sponsored on September 8, 2020

www.aappublications.org/news

(3)

in the presence and absence of PP apneic epi-sodes, the 69 illness-related periods were divided into three apnea patterns: (1) prolonged apnea occurring during the PP interval; (2) prolonged apnea occurring only during the NP interval; and

(

3) no prolonged apnea occurring either during the PP or NP intervals. Table I presents the aver-age daily frequency for the PP and NP intervals and the average postnatal age for each of the three patterns. Examination of this table reveals a temporal sequence to the occurrence of pro-longed apnea in which pattern 1 occurred, on the average, at the earliest age (122.3 days) and pat-tern 3 at the latest (177.7 days). The age data were subject to a single classification analysis of variance23 and the resultant F ratio (4.53) was sig-nificant at less than the 0.05 level (2/66

df).

It can also be seen from Table I that for pat-tern 1 the average daily rate of prolonged apneic

episodes was almost twice as great during NP

(2.41 episodes/day) as during the PP interval (1.26 episodes/day). Of additional interest, the average daily rate during NP for pattern 2 was significantly less (P < .01) than that for pattern 1.

Apneic Episodes Receiving Stimulation

The 769 apneic episodes which persisted and for which physical stimulation was administered occurred in 13 infants and represented 30 ill-nesses. In 23 illness-related periods, the NP daily rate exceeded the PP daily rate, whereas in the remaining 7 the converse was obtained (Sign test,

P < .01). On the average, the daily rate during NP (1.48) was greater than its comparison PP in-terval (0.71). The f-test in this more stringent analysis also revealed a significant association

(t3.33; df29; P < .01).

Table II contains the average daily frequencies and age of the three apnea patterns. Similar to the results obtained when the less stringent cri-teria were used, apneic episodes occurred dur-ing NP at twice the PP rate (pattern 1); the daily frequency of apnea during NP was statistically greater (P < .01) for pattern 1 than for pattern 2; and there was a significant rise in the average

ages at which patterns 1, 2, and 3, respectively,

occurred (F= 6.65; df= 2/66; P < .01).

DISCUSSION

The results of this study provide substantial support for observations that the rate of occur-rence of prolonged sleep apnea tends to increase during an upper respiratory tract inflammatory illness.10’11 Although the data do not bear on the question of the manner in which NP may have its effect, several possibilities should be considered.

TABLE I

No. OF ILLNESS-RELATED PERIODS OF PROLONGED SLEEP

APNEA (20 SECONDS)

Daily Mean Frequency No. of Age

Periods (days) PP NP

122.3

iTTr

154.7 . . . 0.78

177.7 . . . ...

Apnea Pattern

1. Apnea PP 38

2. No apnea PP, apnea NP 9 3. No apnea PP or NP 22

Many studies have demonstrated the intimate in-terrelationship between the respiratory tract and respiratory control.’2’13 It is therefore possible that sensory input to the respiratory centers is modified by the inflammatory process to produce a significant change in respiratory control. The results are also consistent with the hypothesis of-fered by Shaw14 and Swift and Emery15 that the nasal obstruction accompanying NP reflexly in-duces apnea. It is also reasonable to consider the possibility that NP has its effect through an mdi-rect route, by producing increased irritability, sleep deprivation, and an alteration in the char-acteristics of sleep, thereby compromising res-piratory control.16

Of considerable practical and theoretical sig-nificance is the additional observation that the

apparent effect of NP is age-related. By exam-ining the relationship between age and the pro-longed apnea patterns, several important conclu-sions can be generated. First, age is associated with a decrease in the frequency of prolonged apneic episodes. Furthermore, infants become apnea free during the before and after intervals at an earlier age than for NP. Thus, predisposed infants at an early age usually will have episodes in both the before and after illness intervals, with even more frequent episodes during the NP inter-val. These infants then pass through a phase wherein episodes still occur in conjunction with NP but not in the neighboring time intervals. At a still later age, the infants cease having pro-longed sleep apneic episodes during any of the illness-related intervals.

TABLE II

No. OF ILLNESS-RELATED PERIODS OF PROLONGED SLEEP

APNEA RECEIVING STIMULATION

Apnea Pattern No. of Periods Mean Age (&zys) Daily Fre”uenc’ PP NP

1. ApneaPP 25 106.8 0.85 1.73

2. No apnea PP, apnea NP 5 134.6 . . . 0.21

(4)

These results underscore one of the difficulties encountered and, at the same time, suggest a po-tential improvement in the management of in-fants with prolonged sleep apnea. An important decision facing the physician concerns a determi-nation of when the infant is no longer at “risk” of developing prolonged apneic episodes. Phrased operationally, the physician must decide when it

is safe to discontinue the apnea monitor. One

ap-proach to this problem is the discontinuance of the monitor when no episodes occur for an arbi-trarily fixed period of time. However, the sugges-tion that predisposed infants will continue to have episodes during NP even after they no longer occur when the infant is free of apparent illness, raises serious question about the validity of this approach in determining risk status. It would certainly be more helpful in deciding when to discontinue the monitor if the infant also failed to have an episode during NP. Support for this latter approach is obtained from 12 of the in-fants in this study who had been observed on the apnea monitor for periods ranging from 19 to 277 days (median duration, 60 days) after the first NP during which no prolonged sleep apnea occurred in either the PP or NP intervals. The total pa-tient follow-up days was 1,207. Throughout this entire period, there were only seven apparently prolonged apneic episodes and in each instance the infant was noted to be breathing spontane-ously when a parent arrived at the crib. How-ever, since NP might not occur for some time, the development of other means for evaluating risk status would be of considerable value.

Epidemiological as well as histologic studies have demonstrated consistently the importance of an upper respiratory tract inflammatory pro-cess in the etiology of the SIDS.’7’18 The inci-dence of SIDS is both greatest during the winter months and positively associated with the occur-rence of respiratory illnesses in the community. Furthermore, the infants who succumb to SIDS frequently have histologic evidence of an upper respiratory tract inflammatory process.

These observations have provided the basis for several theories of causation. One that has been investigated to a considerable extent suggests that SIDS results from an overwhelming viremia. Although many viruses have been isolated and identified in SIDS infants, they have been local-ized primarily to superficial areas of the respira-tory and gastrointestinal tracts.19’2#{176}The results from these studies suggest that respiratory

dis-ease might have its effect on the occurrence of SIDS not as a consequence of an overwhelming viremia but, rather, as a trigger for the etiologic

mechanism. Consistent with this latter formula-tion and the results of the present study is the causal hypothesis proposed by Steinschneider that prolonged sleep apnea is part of the physio-logic process resulting in SIDS.11

Investigators consistently have observed that the peak incidence of SIDS occurs within the second to third month of life.17’18 This frequency distribution includes all SIDS victims and con-tains infants both with and without concurrent upper respiratory tract disease. In view of the present findings, one might predict on the basis of the sleep apnea hypothesis that infants who die during NP would be older, in general, than those who are free of a respiratory illness when they succumb to SIDS. Furthermore, the incidence of SIDS in the former group of infants would be re-lated to season whereas the incidence in the lat-ter group would not be.

Except for the report of Fedrick,21 a review of the literature failed to reveal data on the simulta-neous examination of the age of death, seasonal effects and the occurrence of a respiratory illness.

Consistent with the results of others, Fedrick noted a highly significant seasonal effect with al-most twice as many deaths occurring between November and April as between May and Octo-ber. However, the effect of season was observed in those infants who died after 12 weeks of age and not in the younger infants. This observation, in conjunction with the additional finding that the seasonal effect was also more marked in males, led Fedrick to conclude that there were

actually “. . . two distinct types of sudden

unex-pected infant death: that occurring mainly to in-fants under 12 weeks where there is little sex dif-ference or seasonal or secular variation, and that where the infant is aged 12 weeks or more.”2’ This conclusion in regard to the interactive ef-fects of season and postnatal age upon the mci-dence of SIDS is compatible with the prediction derived from the sleep apnea theoretical model and the results of the present study. More direct efforts at examining the age distribution of SIDS by considering infants with a minor respiratory illness separately from those free of illness are clearly indicated. Such studies would provide for a better understanding of SIDS and assist in eval-uating the hypothesis that prolonged sleep apnea

is part of the physiologic process resulting in SIDS.

REFERENCES

1. Blystad W: Blood gas determinations on premature in-fants: III. Investigations on premature infants with recurrent attacks of apnea. Acta Paediatr, 45:211, 1956.

(5)

2. Illingworth RS: Cyanotic attacks on newborn infants. Arch Dis Child 32:328, 1957.

3. Miller HC, Behrle FC, Smull NW: Severe apnea and ir-regular respiratory rhythms among premature

in-fants. Pediatrics 23:676, 1959.

4. Silverman WA: Dunham’s Premature Infants. New York, Harper & Row Inc, 1964.

5. Burnard ED, Grauaug A: Dyspnoea and apnoea in the newborn: Some results of investigation. Med J

Aust 1:445, 1965.

6. Reid DHS, Mitchell RG: Recurrent neonatal apnea.

Lancet 1:786, 1966.

7. Daily WJR, Klaus M, Meyer HBP: Apnea in premature

infants: Monitoring, incidence, heart rate changes and an effect of environmental temperature. Pedi-atrics 43:510, 1969.

8. Girling DH: Changes in heart rate, blood pressure, and pulse pressure during apnoeic attacks in newborn babies. Arch Dis Child 47:405, 1972.

9. Bacola E, Behrle FC, de Schweinitz L, et al: Perinatal and environmental factors in late ne#{252}rogenic se-quelae: I. Infants having birth weights under 1,500 grams. Am J Dis Child 112:359, 1966.

10. Stevens LH: Sudden unexplained death in infancy:

Ob-servations on a natural mechanism of adoption of the face down position. Am J Dis Child 110:243, 1965.

11. Steinschneider A: Prolonged apnea and the sudden in-fant death syndrome: Clinical and laboratory ob-servations. Pediatrics 50:646, 1972.

12. James JEA, Daly MdeB: Nasal reflexes. Proc R Soc Med

62:1287, 1969.

13. Widdicombe JG: Respiratory reflexes. In, Fenn WO, Rahn H (eds): Handbook of Physiology: Section 3. Respiration. Washington, American Physiological Society, 1964, vol 1, pp 585-630.

14. Shaw EB: Nasal obstruction theories as related to the sudden infant death syndrome. Read before the NICHHD Research Planning Workshop *7, Feb-mary 1974.

15. Swift PGF, Emery JL: Clinical observations on re-sponse to nasal occlusion in infancy. Arch Dis Child 48:947, 1973.

16. McGinty DJ, Harper RM: Sleep physiology and SIDS:

Animal and human studies. In, Robinson RR (ed): SIDS 1974. Canadian Foundation for the Study of

Infant Deaths, 1974, pp 201-229.

17. Valdes-Dapena MA: Sudden and unexpected death in infancy: A review of the world literature 1954-1966. Pediatrics 39:123, 1967.

18. Beckwith JB: The sudden infant death syndrome. Curs

Probi Pediatr 3:1, 1973.

19. Ray CG, Beckwith JB, Hebestreit NM, Bergman AB: Studies of the sudden infant death syndrome in King County Washington: I. The role of viruses. JAMA 211:619, 1970.

20. Urquhart GED, Grist NR: Virological studies of sud-den, unexplained infant deaths in Glasgow 1967-70. J Clin Pathol 25:443, 1972.

21. Fedrick J: Sudden unexpected death in infants in the Oxford record linkage area: An analysis with

re-spect to time and place. Br J Prey Soc Med 27:217, 1973.

22. Siegel S: Nonparametric Statistics for the Behavioral Sciences. New York, McGraw-Hill Book Co mc, 1956.

23. Snedecor GW: Statistical Methods. Ames, Iowa, Iowa State College Press, 1956.

ACKNOWLEDGMENT

(6)

1975;56;967

Pediatrics

Alfred Steinschneider