Central
Hypoventilation
During
Quiet
Sleep
in Two
Infants
Daniel C. Shannon, M.D., David W. Marsland, M.D., Jeffrey B. Gould, M.D., Barry
Callahan, I. David Todres, M.D., and Jane Dennis, Ph.D.
From the Joseph S.Barr Pediatric intensive Care Unit, Children ‘s Service, the Pulmonary Unit, Massachusetts General Hospital, the Division of Perinatal Medicine, Boston City Hospital, the Department of Pediatrics,
Harvard Medical School, and the Department of Pediatrics, Boston University School of Medicine, Boston,
Massachusetts
ABSTRACT. Expired ventilation (VE), tidal volume (VT), frequency (I), and alveolar PCO2 (PAC02) were examined in six normal infants at 41 to 52 weeks post-conceptional age and in two infants who were apneic at birth. Their response to breathing 5% carbon dioxide in air and to 100% oxygen in quiet sleep were compared to those in rapid eye movement (REM) sleep.
V in normal infants was 259 mI/kg/mm in REM and
200.2 ml/kg/min in quiet sleep with the difference being due to decreased carbon dioxide production and to decreased dead space. VK increased 34.4 ml/kg/min/mm Hg of PCO2 elevation with 5% carbon dioxide breathing during REM and was not significantly different during quiet sleep. During oxygen breathing VE fell by 32.7% at 30 seconds before increasing again.
In the affected infants, V and PACO9 during REM at 1 and 4 months were normal. At 1 month, during quiet sleep, each infant became apneic and PACO2 rose 9 and 8 mm Hg/mm respectively. At this time mechanical ventilation was begun. At 4 months, during quiet sleep, VF was 0.064 and 0.063 ml/ kg/mm at PACO2 of 66 mm Hg in each infant. The change was due entirely to a decrease in VT to 2.3 and 2.5 mI/kg. At this time 5% carbon dioxide breathing given during normal ventilation in REM produced an abrupt fall in VT to 2.0 and 2.2 mI/kg with no change in frequency.
Oxygen
breathing during REM at one month had no effect but at 4 months produced apnea requiring mechanical ventilation after one minute.The findings suggest that the ventilatory response to carbon dioxide is (1) important in initiation of extrauterine ventilation and (2) in sustaining ventilation particularly in quiet sleep. It is not necessary in sustaining ventilation awake or in REM sleep and it represents a balance between the stimulatory and depressant effects of carbon dioxide on the central nervous system. Pediatrics, 57:342-346, 1976,
APNEA, HYPOVENTILATION, SLEEP STATES, CHEMORECEPTORS.
Central
and
peripheral
chemoreceptors
regu-late
ventilation
in
an
attempt
to
maintain
adequate
oxygenation
and
acid-base
homeostasis.
Central
receptors
respond
to
changes
in
CSF
[H
+
I
determined
predominantly
by
changes
in
Pco2’
and
peripheral
receptors
respond
mainly
to
arterial
oxygen
levels
and
pH.24
Similar
reflexes
appear
to
be
responsible
for
regulating
ventila-tion
in
the
newborn,
although
hypoxemia
frequently
fails
to produce
an increase
in
ventila-tion.7
We
report
here
two
infants
who
demon-strated
profound
ventilatory
failure
in quiet
sleep
and
a markedly
abnormal
ventilatory
response
to
carbon
dioxide
breathing.
CASE REPORTS Case 1
A 3.2-kg (25th percentile), full-term female newborn was delivered by caesarian section from a 27-year-old, gravida 2, para 1, blood group A, Rh-positive mother. The indication for caesarian section was a previous caesarian section. No medications other than vitamins were taken prenatally, and none other than spinal anesthesia were used during the
(Received May 16; revision accepted for publication July 29, 1975.)
Supported in part by grant 5T01 HLO5 767 froln the
National Institutes of Health.
Read before the Pediatric Assembly, American Thoracic Society, Cincinnati, Ohio, May 1974.
ADDRESS FOR REPRINTS: (D.C.S.) Pediatric Intensive
TABLE I
DETERMINATION OF SLEEP STATES
PHYSIOLOGIC STUDIES
Each
infant
was
studied
at
1 and
4 months
of
age.
Tidal
volume
(VT)
and
respiratory
frequency
(f)
were
determined
by
temporal
integration
of the
signal
from
a
differential
pressure
transducer
connected
to the
expiratory
end
of a nasal
pneu-motachygraph
(1-cm
ID)
similar
to that
described
by
Rigatto
and
Brady.9
A
precisely
metered
constant
flow
of
compressed
air
was
provided
through
the
pneumotachygraph
to
eliminate
a
dead
space
effect;
the
electrical
signal
at this
flow
was
set
electronically
to zero
so that
expired
air
could
be measured
as a change
in flow.
A 100
ml/
mm
sample
of expired
air was
constantly
pumped
through
an
infra-red
carbon
dioxide
analyzer
from
one
nasal
adaptor
to measure
expired
carbon
dioxide.
Electroencephalogram
was
recorded
from
two
standard
leads
(FP
1-C3
and
C
3-01),
electro-oculogram
(EOG)
was
recorded
from
periorbital
leads,
electromyogram
(EMG)
was
recorded
from
a pair
of submental
electrodes
and,
body
move-Case 2
delivery. Fetal heart rate was regular at 148 beats per minute for the hour before and during delivery without episodes of bradycardia. The membranes were intact and amniotic fluid was clear.
At birth the infant did not breathe spontaneously. At 10 minutes, although pink when ventilated through a bag and mask, respiratory effort was poor. A sample of umbilical arterial blood obtained at 30 minutes of age yielded a partial pressure of carbon dioxide (Pco2) of 115 mm Hg and pH 6.98. Chest X-ray showed expansion of the right lung but failure of expansion of the left. Endotracheal intubation was per-formed. At age 5 hours transfer was made to the Pediatric Intensive Care Unit, Massachusetts General Hospital.
Physical examination revealed a pink infant, estimated gestation 38 weeks, receiving assisted ventilation. The rectal temperature was 37.5 C. Head circumference was 35.4 cm (25th percentile). There were no detectable abnormalities of the heart and abdomen. The urinalysis was normal. The hematocrit was 42%; the white cell count was 25,700/cu miii with 52% neutrophils, 30% lymphocytes, 17% mono-cytes, and 1% eosinophils. Analysis of cerebrospinal fluid yielded sugar and protein concentrations of 56 and 80 mg/ 100 ml respectively without pleocytosis. Analysis of serum revealed that the concentration of sodium was 137 mEq/ liter; potassium, 5.0 mEq/liter; and sugar 87, calcium 8.8, phosphorus 5.1, total bilirubin 4.0, and the blood urea nitrogen 8.0 mg/100 ml. The total protein was 5.0 gm/100
ml and the osmolarity 274 mOsm/liter. Serum immune
globulins were present in normal concentrati9n for age. Cultures of blood and CSF were negative and cultures of stool and tracheal aspirate were negative for bacterial pathogens. Amino aciduria was not found. A chest X-ray
demonstrated that both lung fields were now fully
expanded.
By the fourth day, the infant had normal muscle tone. Brain stem function as determined by “doll’s head” eye movement, pupillary responses, gag, and swallowing reflexes was intact. The infant also fixed and followed light. At this time the infant breathed spontaneously when awake or when rapid eye movements (REM) were observed but not when they were absent. Because this problem persisted, ventilation during sleep was maintained by a negative pressure device (Drinker respirator) both in the hospital and subsequently at home.
At 4 months she rolled over to both right and left, had moderate head lag when pulled to sit, grasped but did not transfer, and attempted to mouth an object in her hand. Overall muscle tone was slightly reduced; brain stem reflexes were again normal. She responded well with a social smile. Negative-pressure ventilation failed to maintain the infant at 6 months of age when she experienced her first upper respiratory tract infection and sustained a near fatal hypoxic episode. A tracheotomy was performed and a positive-pressure device (Emerson pediatric ventilator) was substi-tuted. At last follow-up she was 2 years old, walked, and spoke ten Portuguese words. Neurologic development by the Denver Developmental scale was normal for 18 months of age. Mechanical ventilation during sleep was still necessary because of persistent quiet sleep hypoventilation.
A 3.2-kg (25th percentile) male infant was born sponta-neously at 40 weeks gestation to a 19-year-old gravida 1, para 1 mother. Fetal heart rate during labor and delivery showed no unusual bradycardia either spontaneously or with uterine contractions. The infant breathed at birth and had Apgar
Measure REM Sleep#{176} Quiet Sleep#{176}
EEC LVF = 1 HVS, BS = 0
EOG REM+ =1 REM- =0
EMG o/phasic = 1 Tonic = 0
Movement + = 1 Startle = 0
Respiration Irregular = 1 Regular = 0
Total 5 0
#{176}LVF= low voltage fast; HVS high voltage slow; BS = burst suppression.
scores of eight at 1 minute and nine at 5 minutes. He required resuscitation with physical stimulation and ventila-tion with a bag and mask when cyanosis and hypotonia developed at 5 hours of age. At 12 hours of age intubation was necessary because of apnea. His subsequent course was almost identical to that of patient 1; the same laboratory investigations were also unrevealing. He was treated with a negative-pressure ventilator during sleep. Neurological eval-uation of brain stem reflexes and of infant automatisms did not reveal other evidence of dysfunction at 1 week or 3 months of age. This infant died of staphylococcal septicemia
at
6 months of age. The phrenic nerves and diaphragm wereintact.
The external architecture of brain, cranial nerves, and cerebral vessels was intact. Despite a careful light micro-scopic analysis of serial sections of brain from pons through cervical cord at 300js intervals, no pathologic abnormality could be found.ments
were
observed
and
recorded
by
a
time
marker.
Sleep
states
were
recorded
as shown
in
Table
I.
Sleep
state
scoring
was
performed
TABLE II
VEmmTIoN IN UNAFFECTED INFANTS BREATHING AIR
State ‘E (ml/min/kgJ VT (mi/kg)
f
Pco2REM 259.0 ± 30 6.4 ± 0.7 40.5 ± 5.2 35.2 ± 2.9
Quiet 200.2 ± 18 5.9 ± 0.5 34.2 ± 3.3 35.7 ± 2.4
Three
or
more
consecutive
30-second
epochs
with
a score
of 5 were
defined
as REM
state;
three
or
more
with
a score
of 0 were
defined
as quiet
state;
all
others
were
defined
as
“indeterminate
sleep
state.”
Data
were
recorded
during
spontaneous
air
breathing
awake,
in
REM
sleep,
and
in
quiet
sleep.
Carbon
dioxide
production
(c’co2)
was
measured
in
both
sleep
states
by
analyzing
the
carbon
dioxide
content
of a two-minute
expired
air
collection.
When
there
was
evidence
of REM
or of quiet
sleep
state
and
minute
ventilation
was
unchanged
for
the
previous
minute,
5%
carbon
dioxide
in
air
or
100%
oxygen
was
provided
through
the
nasal
pneumotachygraph.
The
steady
state
ventilatory
response
to
carbon
dioxide
breathing
(generally
at
three
minutes)
and
the
transient
response
to
oxygen
breathing
were
compared
to those
of six normal
infants
of 41 to 52
weeks
post-conceptional
age.
Studies,
in
which
sleep
state
changed
during
administration
of
a
test
gas,
were
excluded.
Breaths
in
which
an
alveolar
plateau
could
be
identified
were
selected
to
measure
PAco2
during
any
particular
test
period.
Arterial
values
for Pco2
were
not
obtained
on
these
infants
but,
using
similar
techniques,
Rigatto
and
Brady9
found
that
arterial
and
alveolar
Pco2
correlated
within
1 mm
Hg
on
the
average.
TABLE III
VENTILATION IN AFFECTED INFANTS BREATHING AIR
Case Sleep State
VT
(mi/mm/kg) (mi/kg)
f
Pco2
At 1 Month of Age
1 fREM Quiet 200 7.1 0 0.0 28 0 38 Increased 2 IREM 1,Quiet 190 5.3 0 0.0 36 0 43 Increased I fREM iQuiet
At 4 Months of Age
194 6.7 64 2.3 29 28 45 66 2 fREM “i#{231}uiet 205 7.3 63 2.5 28 25 45 66 RESULTS
Ventilation
(ml/min/kg
BTPS)
in
infants
was
259.0
(SE,
30)
in REM
sleep
at
a PAco2
of 35.7
mm
Hg
(Table
II).
Vco,
fell
from
6.13
(SE,
0.29)
in
REM
sleep
to
5.69
(SE,
0.28)
in
quiet
sleep
(P
=.05).
Values
obtained
for
PAco2
lie between
the
mean
arterial
Pco2
of newborns
(34
mm
Hg)
and
children
(37
mm
Hg).”
Ventilation
in
the
two
affected
infants
at
1
month
was
slightly
lower
in REM
sleep
(200
and
190
ml/kg/min
respectively)
and
was
absent
in
quiet
sleep
when
PAco2
rose
9 and
8 mm
Hg
per
minute
of apnea
respectively
(Table
III).
PAco2
was
measured
in the
first
breath
following
inter-ruption
of
apnea
with
positive-pressure
lung
inflation.
Ventilation
at 4 months
was
the
same
in
REM
sleep
(194
and
205
mi/kg/mm)
although
PAco2
was
mildly
elevated
to
45
mm
Hg
while
ventilation
in quiet
sleep
was
now
sustained
at 64
and
63
ml/kg/min
with
PAco2
of 66
mm
Hg
in
both
infants.
The
fall
in ventilation
in quiet
sleep
was
due
entirely
to a fall
in VT
to 2.3
and
2.5
ml/
kg
(Table
III).
Ventilation
during
5%
carbon
dioxide
breath-ing
in REM
sleep
in normal
infants
increased
34.4
ml/min/kg/mm
Hg
of Pco2
elevation,
a change
that
was
due
entirely
to
an
increase
in VT
from
6.4
to 14.5
ml/kg.
The
effect
of 5% carbon
dioxide
breathing
was
not
evaluated
at
1 month
of age
in
the
affected
infants.
However,
at 4 months
of age,
ventilation
in
REM
sleep
decreased
markedly
within
five
seconds
of
presenting
5%
carbon
dioxide
in
air
at
the
nasal
pneumotachygraph.
This
was
characterized
by
a fall
in VT
from
6.7
and
7.3
to 2.0
and
2.2
ml/kg
respectively
and
was
duplicated
repeatedly.
This
resulted
in
progres-sive
carbon
dioxide
retention
(Table
IV).
Ventilation
during
oxygen
breathing
in normal
infants
did
not
change
significantly
during
REM
sleep.
Ventilation
during
quiet
sleep
decreased
by
32.7%
of
the
resting
value
after
30
seconds.
Ventilation
during
oxygen
breathing
in
affected
infants
in REM
sleep
at
1 month
of age
did
not
change.
At 4 months
of age,
oxygen
breathing
was
TABLE
IV
VENTILATORY RESPONSE TO 5% CARBON DIoxIDE IN Am DURING REM SLEEP
Infants
Breathin g Air Breathi ng 5% Car bon Dio ride in Air
VT (mi/kg)
f
Pco,
(mm Hg) VT (mi/kg)f
Pco, (mm Hg)Unaffected 6.4 ± 0.7 40.5 ± 5.2 35.2 ± 2.9 14.5 ± 1.5 38.4 ± 4.9 44.2 ± 3.2
1 6.7 29.0 45.0 2.0 27.0 Progressive rise
2 7.3 28.0 45.0 2.2 29.0 Progressive rise
minute.
Bag
and
mask
ventilation
was
then
insti-tuted.
Thus,
ventilation
in
normal
infants
decreased
23% in quiet compared to REM sleep; in the two
affected
infants,
it
decreased
67%
and
69%
re-spectively
at 4 months
of age.
At
1 month
of age,
ventilation
in REM
sleep
was
sustained
at
compa-rable
levels
in normal
as well
as affected
infants
but
ceased
entirely
in
quiet
sleep
in
the
two
affected
infants.
Second,
while
VT
increased
227% in
the
normal
infant
breathing
5%
carbon
dioxide,
it
decreased
by
70%
in
each
affected
infant.
Third,
ventilation
decreased
transiently
during
oxygen
breathing
in
normal
infants
and
ceased
at
4
months
of
age
in
affected
infants.
Finally,
no pathologic
lesion
could
be found
in the
brain
and
cervical
cord
of one
infant
who
died.
The
remaining
affected
infant
requires
mechan-ical
ventilation
during
sleep
at
24
months
of
age.
DISCUSSION
Since
ventilation
was
normal
in
these
two
in-fants
when
awake
and
in REM
sleep,
it appears
that
a carbon
dioxide
ventilatory
response
is not
necessary in the infant at rest except during quiet
sleep. Normal ventilation while awake has also
been
noted
in two
other
infants12’13
and
in adults
with
Ondine’s
syndrome.
It
has
been
suggested
that
awake
ventilation
is maintained
either
by
a
hypoxemic
stimulus
or by
conscious
control.’4
In
these infants conscious control is unlikely and, at
1 month
of age,
hyperoxemic
ventilatory
depres-sion
could
not
be demonstrated.
This
suggests
that
neuronal
output
from
nonchemoreceptor
sources
is sufficient to drive ventilation in the awake and
REM
states
but
ceases
or becomes
inhibited
in the
quiet
sleep
state.
With
microelectrodes
im-planted
in
the
respiratory
center
of
cats,
Burns
observed
increased
electrical
discharges
simulta-neous with inspiratory neuron electrical activity
when
various
sensory
stimuli
were
applied
pe-ripherally.’5
Perhaps
similar
sensory
stimuli
help
to support
ventilation
in the
infant
born
without
chemoreceptor
control
when
awake
or
in
REM
sleep
but
cease
to
be
effective
in
quiet
sleep.
Similarly,
absence
of
response
to
these
stimuli
may
explain
the
decrease
in
ventilation
during
quiet
sleep
in normal
infants.
Ventilatory
depression
during
carbon
dioxide
breathing
in REM
sleep
was
striking
and
has
not
been
previously
recorded
in
central
alveolar
hypoventilation
syndromes.
The
onset
of
depres-sion
within
five
seconds
after
initiating
carbon
dioxide
breathing
suggests
that
it
was
reflex
in
origin.
It could
represent
either
a response
pecu-liar
to these
two
infants
or a normal
component
of
the
ventilatory
stimulus.
A similar
phenomenon
is
apparent
in
the
ventilatory
stimulus
to
hypox-emia,
i.e.,
there
is
a ventilatory
stimulus
with
hypoxemia
in the
mature
infant;
however,
venti-latory
depression
is
observed
in
infants
whose
carotid
body
mechanism
is
relatively
undevel-oped.7
Thus,
a
“flat”
carbon
dioxide
response
curve
could
represent
enough
of a chemoreceptor
stimulant
effect
to overcome
a ventilation
depres-sant
effect
of
carbon
dioxide.
This
depressant
effect
of
carbon
dioxide
could
be
explained
by
hyperpolarization
of
inspiratory
and
expiratory
neurons
as described
by Mitchell
and
Herbert’6
or
by decreased phrenic motor neuron
excitabili-ty.’7
Failure
of oxygen
breathing
to elicit
a
ventila-tory
response
in affected
infants
at 1 month
of age
implies
a
failure
of
carotid
chemoreceptor
re-sponse.
Such
a lack
of response
is well
described
in premature
infants
and
apparently
can
occur
in
term
infants
as well.
The
unique
association
between
alveolar
hypo-ventilation
and
a particular
state
of sleep
in these
infants
suggests
that
a lesion
in the
posterolateral
pontomedullary
tegmentum
similar
to
that
observed
by
Devereaux
et al.’8
might
be involved.
The
absence
of
such
a lesion
in
this
region
or
anywhere in the
brain
stem
in patient
2 suggests
that
a localized
failure
of development
of cells
or
Deonna
et
al.’2
have
recently
suggested
that
ventilatory
responsiveness
to carbon
dioxide
may
develop
with
time
since
their
infant
no
longer
requires
mechanical
ventilation.
However,
the
patient
reported
by
Mellins
et al.”
died
and
our
remaining
infant
requires
positive-pressure
venti-lation
during
sleep
at over
2 years
of age.
REFERENCES
1. Pappenheimer JR. Fend V, Heisy SR. et a!: Role of cerebral fluids in control of respiration as studied in an unanesthetized goat. Am
J
Physiol 208:436, 1965.2. Rebuck AS, Woodley WE: Ventilatory effects of
hypoxia and their dependence on PCO.
J
AppIPhysiol 38: 16, 1975.
3. Watt JG, Dumke PR, Comroe JH: Effects of inhalation of 100 percent and 14 percent oxygen upon
respira-tion of unanesthetized dogs before and after
chemoreceptor denervation. Am
J
Physiol 138:610, 1943.4. Heymans C, Bouckaert
JJ,
Dautreband L: Sinus caro-tidien et reflexes respiratoires: II. Influences respi-ratoires reflexes de l’acidose, de l’alcalose, de l’an-hydride carbonique, de I’ion hydrogene and de l’anoxemie. Arch Int Pharmacodyn 39:400, 1930. 5. Purves MJ: The effects of hypoxia in the newborn lambbefore and after denervation of the carotid chemo-receptors.
J
Physiol 185:60, 1966.6. Purves MJ: The respiratory response of the newborn lamb to inhaled CO with and without accompany-ing hypoxia.
J
Physiol 185:78, 1966.7. Rigatto H, Brady JP: Periodic breathing and apnea in pre-term infants: I. Evidence for hypoventilation
possibly due to central respiratory depression. Pedi-atrics 50:202, 1972.
8. Avery ME, Chernick V. Dutton RE, Permutt S: Ventila-tory response to inspired CO2 in infants and adults.
J
Appl Physiol 18:895, 1963.9. Rigatto H, Brady JP: A new nosepiece for measuring ventilation in preterm infants.
J
Appl Physiol 32:423, 1972.10. Anders T, Emde R, Parmelee A: A Manual of Standard-ized Terminology, Techniques and Criteria for Scoring of States of Sleep and Wakefulness in Newborn Infants. Los Angeles, U.C.L.A. Brain Information Service/ B.R.I. Publications Office,
1971.
11. Polgar G, Promadhat V: Pulmonary Function Testing in Children: Techniques and Standards. Philadelphia, WB Saunders, 1971, p 202.
12. Deonna T, Arezynska W, Torrado A: Congenital failure of automatic ventilation (Ondine’s curse).
J
Pediatr84:710, 1974.
13. Mellins RB, Balfour HH, Turino GM, et a!:
Failure
of
automatic control of ventilation (Ondine’s curse): Medicine 49:487, 1970.14. Strieder DJ, Baker WG, Baringer JR, Davis RL: Auto-matic respiratory failure. Arch Neurol 29:46,
1973.
15. Burns BD: The central control of respiratory move-ments. Br Med Bull 19:7, 1963.
16. Mitchell RS, Herbert DA: The effect of carbon dioxide on the membrane potential of medullary respira-tory neurons. Brain Res 75:345, 1974.
17. Gill PK, Kuns M: Properties of phrenic motoneurons.
J Physiol 168:258, 1963.
