Brainstem Finding in SIDS
by Henry F. Krous, MD
Evidence continues to accumulate that SIDS is due to a defect in the neural control of
respiration during sleep. Apnea is a normal physiologic event occurring regularly during
sleep, but, if not interrupted, could lead to death. Within the last two decades, it was
thought that SIDS might be a result of prolonged or recurrent sleep apnea. However, newer
data suggest that failure of mechanisms terminating apnea, i.e., initiating respiration,
are important in SIDS. Arousal stimulates respiration after the onset of sleep apnea, and
therefore has become a focus of intense investigation.
Before addressing exciting new research findings, it may be helpful to review briefly
respiration. Simply stated, respiration is the delivery of oxygen (O2) to and carbon
dioxide (CO2) from the organs; this process enables cells to carry out their metabolic
functions, such as muscle contraction and nerve impulse transmission. Respiration is
regulated by a complex series of events involving the central and peripheral nervous
systems , the lungs, the muscles of the chest, neck and diaphragm, and chemoreceptors
(specialized tissues which detect imbalances in the blood gases (O2, and CO2), and acid
(H+). Tight regulation of blood gas and acid levels must be maintained in order for the
body's organs to function normally; responses to abnormal levels requires their
recognition (sensory or afferent part of a loop) and appropriate response (motor or
efferent part of a loop). If, for example, blood CO2 and H+ levels increase above the
normal range, as can occur with vigorous exercise, then the body responds by increasing
the rate and depth of respiration. The chemical and neural recognition of abnormal blood
gas or acid levels is a sensory or afferent function, while contraction of the muscles of
respiration is a motor or efferent function. The afferent sensory information is
communicated to the (efferent) muscles of respiration by a series of electrochemical
reactions wherein specialized molecules known as neurotransmitters interact with their
receptor sites located in tissues of the respiratory system. Acetylcholine (Ach) is one of
the neurotransmitters acting in the complex sensory and motor pathways of respiration.
The arcuate nucleus (ARC) located on the ventral surface of the brainstem may be an
important site which senses abnormal levels of CO2 and H+ in the blood. Therefore, Dr.
Hannah Kinney and her colleagues hypothesized that there may be an abnormality in this
brainstem nucleus in SIDS. To test their hypothesis they analyzed brainstem tissues for
the presence of Ach binding sites from 77 infants, including 45 SIDS cases, 14 infants who
died acutely of known causes, and 18 infants who died of longer illnesses characterized by
chronic or recurrent low blood oxygen levels (hypoxemia) (Science 1995;269:1446- 1450).
The SIDS and acute control cases were well matched for gestational and postconceptual age,
birth weights, Apgar scores, postnatal course and postmortem intervals. They found on
average fewer binding sites for Ach in the arcuate nucleus in SIDS compared to the acute
control group. This exciting discovery: (1) suggests that some infants are at increased
risk of SIDS since their responses to increased blood CO2 and H+ levels may be reduced,
(2) appears to support the hypothesis that the arousal response to abnormal blood gas
levels and acidosis is defective in some cases of SIDS, (3) may help explain the higher
incidence of SIDS with the prone sleep position, which could promote abnormally high blood
CO2 levels as a result of either upper airway obstruction or rebreathing CO2 exhaled into
the microenvironment around the baby's nose and mouth, (4) is consistent with other
studies suggesting that infants dying of SIDS have subtle delays in the development of
their central and peripheral nervous systems, and (5) provides direction for investigation
of other related anatomic sites.
In this study SIDS and acute controls differed statistically as groups, yet there was
overlapping of individual SIDS cases and acute controls. In other words, some of the acute
controls had diminished binding while some of the SIDS cases had normal values. These
observations add further evidence that SIDS is a complex event and that subsets of infants
have subtly different abnormalities predisposing them to this tragedy.
Presently SIDS is diagnosed after other causes of sudden infant death are excluded. The
results of the study of Kinney et al holds out the possibility that pathologists will make
a diagnosis of SIDS on positive rather than negative criteria.
It must be emphasized that clinical tests to identify this abnormality are presently
not available. Further, Kinney et al suggest that the decreased binding "may be an
insensitive marker of a more fundamental mechanism, rather than the cause of
cardiorespiratory dysfunction...[and simply may be] associated with an increase in the
probability of death from SIDS".
Finally, it is worth noting that tissues SIDS cases from San Diego County were included
in this study by Dr. Kinney and her colleagues. This was possible because of the landmark
California legislation authored by Senator Daniel Boatright, the support of many SIDS
parents, the gracious cooperation of Dr. Brian Blackbourne and his staff at the Medical
Examiners Office and the San Diego SIDS Research Project. In many other states, this kind
of important work is still not possible. The San Diego SIDS Research Project will continue
to work with and support Dr. Kinney and other investigators in order that a fuller
understanding of SIDS can be achieved.
Reprinted from the San Diego Guild For Infant Survivals newsletter Reaching Out
The Sudden Infant Death Syndrome
Alliance Media Alert
Phipps Cohe, National Public Affairs Director
September 6, 1995
Findings of Kinney Study Released: Cell
Defect May Put Some Babies at Higher Risk of SIDS
The September 8th issue of Science reports the findings of a multi-center study led by
Hannah Kinney, MD (Harvard/Boston). Entitled, "Decreased Muscarinic Receptor Binding
in the Arcuate Nucleus in Sudden Infant Death Syndrome," the study identifies an
abnormality in specific brain cells of many babies who died of Sudden Infant Death
According to the report, the research team analyzed brain tissue from group of babies
who died of SIDS and compared it with samples from infants who died from known causes.
Many of the SIDS victims had a neurochemical abnormality in a region of the brain that is
thought to play a major role in heart and long control- the "arcuate nucleus."
The report further suggests that the cell defect in the SIDS group involves faulty binding
of specific receptors that aid the flow of nerve impulses between cells. Less binding of
these receptors, as found in the SIDS group, may hamper normal respiratory activity.
The authors recognize that the cell defect alone is probably not the sole cause of
SIDS, but may contribute to a larger respiratory impairment. As Dr. Kinney explains,
"It may be that in normal babies, the nervous system helps infants detect breathing
difficulties and arouses them so they can mount protective reflexes, but SIDS babies with
this defect are unable to perform these protective reflexes." Dr. Kinney also
suggests that the brainstem defect may be intertwined with other factors, such as prone
sleeping or smoke exposure, which could work in concert to make an infant more vulnerable
The SIDS Alliance is proud to have provided funding to help support Dr. Kinney's
important brainstem research. The SIDS Center for Excellence at Children's
Hospital/Harvard Medical School was established by the SIDS Alliance several years ago to
encourage multidisciplinary and collaborative investigation among SIDS scientists.
The reference is:
Kinney HC, Filiano JJ, Sleeper LA, Mandell F, Valdes-Dapena M, and Frost White W:
Decreased muscarinic receptor binding in the arcuate nucleus in sudden infant death
syndrome. Science Vol 269:1446-1450, September, 1995.
Hope this is helpful. I would imagine that public libraries carry the journal Science.
If you get the article and have questions about it, I would be happy to try to answer
John L. Carroll, M.D.
The Johns Hopkins Children's Center
I think a note of caution is in order. There has been a great deal of
work on brainstem pathology in SIDS and the results are, at best,
conflicting, and at worst confusing and confounding.
Below I've posted 3 abstracts describing a few recent publications with
some comments after each one. The main ideas I'd like to convey here are
1) that this is very difficult area to study and 2) that one should
approach all such reports with a great deal of caution. Generally, in
science, one doesn't accept anything as 'true' and nothing as even
'probably true' until it's been shown by multiple investigators. Even
better, multiple investigators using different techniques and getting the
The first example below calls into question a classic SIDS pathological
finding... (you may want to focus on the last 2 sentences of the
AU Pamphlett R. Treloar L.
TI ASTROCYTES IN THE HYPOGLOSSAL NUCLEI OF SUDDEN INFANT DEATH SYNDROME
(SIDS) INFANTS - A QUANTITATIVE STUDY.
SO Neuropathology &Applied Neurobiology. 22(2):136-143, 1996 Apr.
AB It has been suggested that brain stem hypoxia or ischaemia underlies
the sudden infant death syndrome (SIDS)I but previous reports of
astrocytosis in the brain stems of SIDS infants have been
contradictory. A volumetric quantitative technique was, therefore,
developed to compare astrocyte numbers and sizes in the hypoglossal nuclei
of SIDS and control infants. In 12 SIDS and eight control infants,
serial sagittal sections were taken through the hypoglossal nucleus and
every tenth section was stained for glial fibrillary acidic protein.
Astrocytes were counted in the central 4% of a grid stepped throughout the
hypoglossal nucleus, and the heights of 100 astrocyte nuclei were measured
with a microcator, Astrocyte number, corrected for section thickness and
nuclear height, was divided by the volume of the hypoglossal nucleus to
calculate astrocyte density, Numbers of astrocytes did not differ
significantly between SIDS (mean number 44 7291 so 12 096) and control
(mean number 46 562, so 11000) infants, Astrocyte nuclear height did not
differ significantly between groups (SIDS: mean height 3.98 mu m, so
0.22: control: mean height 3.84 mu m, so 0.31). Astrocyte density was
similar in SIDS (mean density 24 378 astrocytes/mm(3), so 6155) and
control (mean density 23 978 astrocytes/mm(3), so 4031) infants, No
quantitative evidence of astrocytosis was found in the hypoglossal nuclei
of SIDS infants, This implies that SIDS infants die without
previous episodes of hypoxia/ischaemia severe enough to damage the brain
stem. [References: 35]
Comment: Brainstem gliosis (an abnormal proliferation of a neuronal cell
type called 'glial cells') is a classic finding in SIDS and is said to
indicate evidence that the infant experienced low oxygen levels before
death. Now, an investigator comes along who claims to use a better
technique and finds no difference between SIDS infants and controls
(presumably controls not exposed to low oxygen levels - this point needs
verifying). The point is that, in 1996, a classic finding is now being
questioned. Not only, but a deeply rooted classic concept in the current
thinking about SIDS pathophysiology is being challenged - that SIDS
infants experience episodes of hypoxia/ischemia (low oxygen levels) prior
to death sufficient to damage the brainstem. This is the nature of
science. More studies will now attempt to refute or confirm these
findings. That's how science works.
Here is another one looking at brainstem areas involved in cardiac, upper
airway, and respiratory control... (again, focus on last 2 sentences)
AU Lamont P. Murray N. Halliday G. Hilton J. Pamphlett R.
TI BRAIN STEM NUCLEI IN SUDDEN INFANT DEATH SYNDROME (SIDS) -
VOLUMES, NEURONAL NUMBERS AND POSITIONS.
SO Neuropathology &Applied Neurobiology. 21(3):262-268, 1995 Jun.
AB It has been suggested that the defect underlying the sudden infant
death syndrome (SIDS) lies in brain stem nuclei involved in cardiac
and respiratory function. However, most studies have not used rigorous
quantitative techniques to assess brain stem nuclear volumes and
neuronal numbers. We have measured the volume, neuronal numbers and
position of brain stem nuclei in 11 SIDS and 11 aged-matched control
infants. Using serial sagittal sections, nuclei involved in maintaining
airway patency (hypoglossal, ambiguus and retroambiguus), heart rate
(dorsal vagal) and generation of respiratory rhythm (ambiguus and dorsal
vagal) were studied. No significant differences were found in nuclear
volume increase with age, total neuronal number or nuclear position
between SIDS and control cases. These findings support the hypothesis
that the nervous system in SIDS may be normal until the final event
that kills these infants. [References: 27]
Comment: Here again, a challenge to traditional notions that there may be
something wrong with the central nervous system in SIDS infants. There
are several good lessons here. First, the conclusion of this article goes
way to far, it over-interprets the findings. It may be true that they
investigated several brainstem regions (nuclei) involved in respiratory,
cardiac, and upper airway control, but the brainstem is much more
complicated than this and cardiorespiratory control involves many more
brainstem areas than the ones they looked at. They could have missed
something abnormal in other areas. Secondly, they looked at nuclear (cell
group) volume, position, and cell number. It may well be that cells
(neurons) can be dysfunctional without being anatomically abnormal.
Finally, the way it is written the conclusion sounds strong. However,
basically, they used a limited technique to look at limited areas and they
didn't find anything. It is important to beware of conclusions that seem
to say more than they actually do.
Finally... here's one that agrees *and* disagrees...
AU Okusky JR. Kozuki DE. Norman MG.
TI SUDDEN INFANT DEATH SYNDROME - POSTNATAL CHANGES IN THE VOLUMES OF
THE PONS, MEDULLA AND CERVICAL SPINAL CORD.
SO Journal of Neuropathology &Experimental Neurology. 54(4):570-580, 1995
AB The brainstem and cervical spinal cord were sampled from 45 cases of
sudden infant death syndrome (SIDS), from 17 control cases without
neurological disease, and from three negative control cases with abnormal
growth of the central nervous system (36-98 postconceptional weeks).
Morphometric analyses were performed on serial Nissl sections to determine
the total volumes of the pons, nucleus pontis, medulla and cervical spinal
cord. Normal development was characterized by a linear increase in the
volumes of these regions during the first postnatal year. Regression
analysis revealed that in SIDS cases the rates of increase in the volumes
of the pens and nucleus pontis were significantly greater than in controls
(56% and 83%, respectively), while growth rates did not differ
significantly for the medulla and cervical spinal cord. By direct
comparison, there was a significant increase in the mean volumes of the
pons (33%), nucleus pontis (38%) and medulla (19%) in SIDS cases when
compared to controls. No evidence of excessive edema or gliosis was noted
in the brainstem by light and electron microscopy to account for the
increased volumes. Subtle morphological abnormalities in brainstem
neurons from SIDS cases, including an increased size of Nissl bodies in
the cytoplasm of large motor neurons and the presence of paranucleolar
coiled bodies, were consistent with an increased synthesis and transport of
ribosomal RNA, an increased synthesis of cellular proteins and neuronal
hypertrophy. [References: 50]
Comment: Here's one that seems to agree with the first paper that there
is no evidence of increased gliosis in SIDS cases compared to controls.
However, they *disagree* with the second one (above) and do find increased
volumes (size) of certain brainstem regions involved in cardiorespiratory
control. So... this is science. Now many more papers will have to
examine these questions over the next few years. What we finally get out
of all this as 'probable truth' will really only come from putting
together many many studies over the next few years and trying to see where
they all agree.
No doubt many of you are aware of how science works. However, for those
who aren't fully aware of how science produces "knowledge" I thought these
abstracts provided a nice illustration of how tricky it can be to sift
through the scientific literature on SIDS. As a general rule - don't
believe any scientific paper's conclusions until they have been confirmed
by at least 1 independent researcher or group.
Hope this is helpful.
John L. Carroll, M.D.
The Johns Hopkins Children's Center
My thanks to Doctor John Carroll, who gave me a lead where to
find the abstract on brainstem research in SIDS from Montreal.
It was presented at the Society for Pediatric Research Annual
Meeting in the first week of May, 1996. It reads as follows:
NEUROPATHOLOGICAL EVIDENCE FOR PREVIOUS HYPOXIA IN A POPULATION
OF INFANTS DYING FROM SIDS. K.A. Waters and A. Cote. McGill
University, Montreal Childrens Hospital, Montreal, PQ, Canada.
It is postulated that respiratory control abnormalities lead
to neuropathological evidence of previous hypoxia in a proportion
of infants who die from SIDS. The aim of our study was to
determine whether the pattern of recent vs chronic histological
changes in the brain were different in infants dying from SIDS as
compared to infants dying suddenly and unexepectedly from other
causes. All autopsies for sudden infant death from one pediatric
center included histological examination of the brain by one
neuropathologist. After review of the history, the death scene,
and the autopsy, a diagnosis of SIDS was made in 129 (67% male)
of the 149 cases from 1987-1994; in 20 cases (55% male) autopsy
revealed another cause for death. Cases of prolonged survival
(1-3 days) following resuscitation were excluded from analysis.
In 27% of SIDS and 60% of non-SIDS (P<0.005), there was evidence
for a recent anoxic event
from which the infant survived.
Gliosis, which may indicate more long standing hypoxia, was found
in 75% of SIDS victims and 30% of non-SIDS (P<0.001). infants
dying from SIDS with these
neuropathological findings tended to
be older at the time of their death (median 109 vs 69 days,
P<0.02). we conclude that the pattern of hypoxic changes that
precedes death from SIDS is
suggestive of long-standing
abnormalities, and differs significantly from infants dying
suddenly from other causes.
REFERENCE: Pediatric Research, 30(4): 392A, 1996.
COMMENT: If you analyze this carefully using the guidelines
suggested by Doctor John Carroll's excelent recent E-Mail, you
will see that there are some assumptions being made here which
affect the validity of the conclusions:
1) Doctors Waters and Cote are careful inviestigators, so I
assume the study was rigorusly performed.
2) The details of the controls is lacking. This remains one of
the most difficult areas with pathologic research in SIDS. If
SIDS is themost common cause of death in infants, who do you find
to compare them with. There are not many infants dying suddently
who were not previously ill.
3) Doctor Hannah Kinney has shown convincingly that the methods
one uses to sample the brainstem may affect your results. This
issue is not addressed.
4) The relationship between gliosis and chronic low oxygen is
not as solid as this abstract would imply. Gliosis is
non-specific, and may be cause by other things.
The above comments are not meant to discredit the study. As
Doctor Carroll pointed out, one needs to consider it carefully,
and not accept it all at face value. Suffice it to say, no
scientific study is perfect, and each must be viewed in the
context of the entire scientific inquiry. This study would seem
to line upon the side of investigators who believe that SIDS
victims may have had some chronic hypoxia prior to their death.
Because many studies address this hypothesis from different
angles, and because they continue to contradict each other, this
whole question remains one of the most controversial in SIDS
Childrens Hospital Los Angeles