- Email: [email protected]
From epidemiology to physiology and pathology: apnea and arousal deficient theories in sudden infant death syndrome (SIDS)—with particular reference to hypoxic brainstem gliosis
Forensic Science International 130S (2002) S21–S29
From epidemiology to physiology and pathology: apnea and arousal deficient theories in sudden infant death syndrome (SIDS)—with particular reference to hypoxic brainstem gliosis Toshiko Sawaguchia,*, Patricia Francob, Ineko Katoc, Satoru Shimizud, Hazim Kadhimb, Jose Groswasserb, Martine Sottiauxb, Hajime Togaric, Makio Kobayashie, Sachio Takashimaf, Hiroshi Nishidag, Akiko Sawaguchia, Andre Kahnb a
Department of Legal Medicine, School of Medicine, Tokyo Women’s Medical University, 8-1 Kawada-cho, Shinjuku-ku, Tokyo 162-8666, Japan b Pediatric Children’s Hospital Reine Fabiola, Free University of Brussels, Avenue J.J. Crocq 15, 1020 Brussells, Belgium c Department of Pediatrics, Nagoya City University, 1 Aza Kawasumi, Mizuho-cho, Mizuho-ku, Nagoya, Aichi 467-8601, Japan d Department of Hygiene and Public Health, School of Medicine, Tokyo Women’s Medical University, 8-1 Kawada-cho, Shinjuku-ku, Tokyo 162-8666, Japan e Department of Pathology, School of Medicine, Tokyo Women’s Medical University, 8-1 Kawada-cho, Shinjuku-ku, Tokyo 162-8666, Japan f National Institute of Neuroscience, National Center of Neurology and Psychiatry, 4-1-1 Ogawahigashi, Kodaira, Tokyo 187, Japan g Maternal and Perinatal Center, School of Medicine, Tokyo Women’s Medical University, 8-1 Kawada-cho, Shinjuku-ku, Tokyo 162-8666, Japan
Abstract Among 27,000 infants studied prospectively to characterize their sleep–wake behavior, 38 infants died under 6 months of age. They included 26 cases of sudden infant death syndrome (SIDS). Five infants who died from congenital cardiac abnormalities, two from infected pulmonary dysplasia, two from septic shock with multi-organ failure, one during a prolonged seizure, one from a prolonged neonatal hypoxemia, one from meningitis with brain infarction. All the infants had been recorded during one night in a pediatric sleep laboratory some 3–12 weeks before death. The frequency and duration of sleep apneas were analyzed. The infants’ brain stem material was collected and immunohistochemistry of glial fibrillary acidic protein (GFAP) was carried out. The density of GFAP-positive reactive astrocytes was measured in the cardiorespiratory and arousal pathway. Akaike information criterion statistics (AIC) were calculated to elucidate the relationship between the epidemiological data on sleep position, the physiological data and the pathological data in SIDS victims. The duration of obstructive apnea was the most significant variable to differentiate between SIDS victims and control infants. In conclusion, the present study sustains the possibility of an organic fragility within the arousal pathway in SIDS victims with repetitive sleep apneas. # 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Sudden infant death syndrome (SIDS); Brainstem; Gliosis; Apnea; Arousal
1. Introduction Studies of infants who died of sudden infant death syndrome (SIDS) reported the presence of subtle gliosis * Corresponding author. Tel.: þ81-3-52-697300; fax: þ81-3-52-697300. E-mail address: [email protected] (T. Sawaguchi).
within the reticular formation of the medulla [1]. Quantitative analysis of reactive astrocyte counts showed no difference between infants who died from SIDS and control subjects [2–6]. Alternatively, the density of astrocytes was shown to be higher in the brainstem of SIDS victims than in that of control infants at the level of the hilum of the inferior olive, the ventrolateral tegmental region and the nucleus tractus solitarius [7,8]. Other
0379-0738/02/$ – see front matter # 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 7 9 - 0 7 3 8 ( 0 2 ) 0 0 1 3 5 - 4
S22
T. Sawaguchi et al. / Forensic Science International 130S (2002) S21–S29
studies [9–12] reported brainstem gliosis in SIDS victims, at the level of the medullary reticular formation [11] and in the pontine reticular formation and pontine nuclei [12]. The present study was conducted to compare postmortem findings in SIDS victims with infants’ physiological sleep apnea characteristics and epidemiological data on sleep position.
2. Materials and methods 2.1. Physiological analyses 2.1.1. Subjects Over 20 years in various Belgian pediatric sleep laboratories 27,000 infants were studied prospectively to determine sleep–wake characteristics. The families were invited to join the study when leaving the maternity unit. The infants entered the study if they were born at term after a normal gestation, and if there was no family or personal history of apnea, apparent life threatening event (ALTE) or SIDS. At the time of recording, the infants were aged between 2 and 27 weeks, were healthy, and were receiving no medication. Following the sleep recording, 38 infants died suddenly. Following postmortem examinations and death scene investigations, 27 of the 38 infants were considered as SIDS victims [19]; five died from congenital cardiac abnormalities; two from infected pulmonary dysplasia; two from septic shock; one from a prolonged seizure; one from prolonged neonatal hypoxemia. The main characteristics of the dead infants are reported in Table 1. The maximum delay between the estimated time of death and the postmortem examination was 24 h. 2.1.2. Polygraphic monitoring For each infant, 8 h sleep studies were conducted overnight in a sleep laboratory, following standard recording techniques [13,14,20,21]. The recordings were done in a quiet and darkened room, at an ambient temperature ranging between 20 and 23 8C. All infants slept supine, without restraints. Recording started around 21:00 h. The infants were observed continuously during recording. They were fed on demand. Their behavior and any nursing intervention were charted. No infant had a pacifier during the recorded sessions. The following variables were recorded simultaneously: two scalp electroencephalogram with central and occipital leads, two electrooculograms and electrocardiogram. Thoracic respiratory movements were measured by impedance and airflow with thermistors taped under both nostrils and on the side of the mouth. Oxygen saturation was recorded continuously by a transcutaneous sensor (Nellcor, USA). Gross body movements were measured using an actigram placed on one arm. The data was collected on a computerized infant sleep recorder (Alice recording system III, Healthdyne, USA).
2.1.3. Data interpretation Based on the polygraphic recordings, sleep stages and sleep apneas were scored according to standard definitions [13,14,20–23]. Apneas were scored when they lasted 3 s or longer. They were classified as central apnea when flat tracings were obtained simultaneously from the strain gauges and thermistors. Periodic breathing was defined as the succession of more than two central apneas separated from each other by less than 20 s. Obstructive apneas were defined as continuous deflections from the strain gauges, with a flat tracing recorded from the thermistors. A mixed apnea was defined as a central apnea directly followed by an obstructive episode. Mixed apneas were scored together with the obstructive episodes. The frequency of obstructive apneas was measured by dividing the total number of apnea by the total sleep (in min) and multiplying by 60. The type, frequency (number per h of sleep) and duration (in s) of sleep apneas were computed. The recordings were analyzed visually by two independent scorers without knowledge of subjects’ age or sex to ensure reliability. Discrepancies were discussed and the agreed upon score was computed for analysis. 2.2. Pathological analyses 2.2.1. Subjects analyzed A total of 48 paraffin blocks were collected from the brain stems of the 38 infants: 7 blocks from the midbrain, 22 from the pons, and 19 from the medulla oblongata. The delay between death and postmortem examination was within 24 h. 2.2.2. Standard histological brain examination Hematoxylin-eosin (HE) staining was carried out as the standard histological brain staining. 2.2.3. Immunohistochemical examination Immunohistochemical studies using anti-glial fibrillary acidic protein (GFAP, Fig. 1) antibodies (DAKO, USA) was carried out and visualized by the HRP-DAB method using 4 mm thick sections. Five overlapped microscopic fields at 40 magnification were selected from several sites and the number of GFAP-positive reactive astrocytes was visually counted. Measurements were carried out in the following sites: periaqueductal gray matter, substantia nigra (pars compacta and pars reticulata), red nucleus and dorsal raphe nucleus of the midbrain; reticular formation, superior central nucleus, pontine nucleus, locus cerruleus and nucleus raphe magnus of the pons; dorsal vagus nucleus, solitary nucleus, reticular formation, nucleus raphe obscurus, nucleus raphe magnus, nucleus medulla oblongata centralis, olive nucleus and nucleus ambiguus of the medulla oblongata. For quantification, the superior central nucleus and nucleus raphe magnus of the pons, as well as the nucleus raphe obscurus and nucleus raphe magnus of the medulla oblongata were enclosed.
T. Sawaguchi et al. / Forensic Science International 130S (2002) S21–S29
S23
Table 1 General details of each case Case no.
Sex
Gestational age (weeks)
Postneonatal age (weeks)
Cause of death
SIDS cases 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
F F M F F M M F M M F M M M M F F M M M M M M F F M
40 38 40 41 37 40 39 40 39 38 40 39 40 38 40 36 40 38 39 37 35 40 40 40 37 31
16 13 12 18 19 16 12 14 21 10 11 22 40 36 10 19 3 4 6 18 6 22 31 31 20 20
SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS
Control cases 1 2 3 4 5 6 7 8 9 10 11 12
M M F M M M F M M F M F
39 40 40 33 40 37 37 39 40 37 39 40
24 11 6 21 14 4 5 6 7 7 9 9
Meningitis and brain infarction Pneumonia Mypcarditis Varisella Cardiopathy with pulmonary hypertension Syndrome of Opitz Hepatitis and general infection Bronchopneumopathy Myocarditis Infanticide Bronchopneumopathy Bronchopneumopathy
The pathological measurements were done twice by the same pathologist and quantitized data with a large standard deviation were recounted or rejected. 2.2.4. Data analysis The scorers of the sleep recordings as well as the pathologist did not known the causes of the infants’ death. Akaike information criterion statistics (AIC) [24,25] was calculated using each infant’s data including epidemiological data on sleep position, physiological data on sleep apnea, pathological data on the density of GFAP-positive reactive astrocytes as follows:
(1) to be SIDS or non-SIDS as response variables in each case with single, two or three explanatory variables (see Tables 2–5); (2) to be in a prone position or not as response variables in each case with single, two or three explanatory variables (see Tables 6–9).
2.3. Ethical issues The study was approved by the Ethical Committee of the University Children’s Hospital and was performed in
S24
T. Sawaguchi et al. / Forensic Science International 130S (2002) S21–S29
Fig. 1. Immunohistochemistry of glial fibrillary acidic protein in the brain of a SIDS victim.
accordance with the ethical standards prescribed by the Declaration of Helsinki.
3. Results Results of the calculation of AIC are shown in Tables 2–5 to determine being SIDS and Tables 6–9 to clarify the prone position. In Table 2, to determine the criteria for being SIDS or non-SIDS, the duration of obstructive apnea was the most significant as a single explanatory variable, followed by the frequency of central apnea. In Table 3, to determine being SIDS or non-SIDS, the gliosis in nucleus ambiguus in the medulla oblongata and superior central nucleus and nucleus magnus in the pons was the most significant, followed by the giosis in the nucleus ambiguus within the medulla oblongata and dorsal raphe nucleus of the midbrain. In Table 4, to determine being SIDS or non-SIDS, the gliosis in nucleus ambiguus in the medulla oblongata and superior central nucleus and nucleus magnus in pons and
dorsal raphe nucleus in the midbrain was the most significant finding. In Table 5, to determine being SIDS or non-SIDS, the gliosis in the nucleus ambiguus with the medulla oblongata and superior central nucleus and nucleus magnus in pons and dorsal raphe nucleus in the midbrain was the most significant finding. In Table 6, to determine being in the prone position, the duration of obstructive apnea was the most significant finding. In Table 7, to determine being in the prone position, the gliosis in the dorsal raphe nucleus of the midbrain and solitary nucleus in medulla oblongata was the most significant finding. In Table 8, to determine being in the prone position, the gliosis in the dorsal raphe in the midbrain and solitary nucleus, nucleus raphe obscurus and nucleus raphe magnus in medulla oblongata was the most significant, followed by the gliosis in dorsal raphe in the midbrain, solitary nucleus in medulla oblongata and being SIDS or non-SIDS. In Table 9, to determine being in the prone position or not, the gliosis in dorsal raphe in the midbrain was the largest subset of explanatory variables.
Table 2 List of single explanatory variables in the calculation of AIC when the response variable is to be SIDS or non-SIDS No.
Explanatory variable
Number of categories of explanatory variable
AIC
Difference of AIC
Weight
1 2
OAPD CAPF
5 7
100.64 82.91
0.00 17.73
1.00000 0.00014
OAPD: duration of obstructive apnea (s); CAPF: frequency of central apnea (no. per h).
T. Sawaguchi et al. / Forensic Science International 130S (2002) S21–S29
S25
Table 3 List of the combination of two explanatory variables in the calculation of AIC when the response variable is to be SIDS or non-SIDS No.
Explanatory variable
Number of categories of explanatory variable
AIC
Difference of AIC
Weight
1 2 3 4 5 6 7 8 9 10 11 12
MOAN, PMR MOAN, MBDR PMR, MBDR MOAN, PAR MOAN, MBPG MOAN, MOLR MOAN, MODNV MOAN, POSITION MOAN, MOMR MOAN, MOSN MOAN, OAPD MOAN, MOAR
6 9 6 12 6 6 6 6 9 6 15 9
15.59 9.16 9.01 5.00 4.00 2.70 2.70 1.05 2.25 2.52 4.46 7.79
0.00 6.43 6.58 10.59 11.59 12.89 12.89 14.54 17.84 18.11 20.05 23.38
1.00000 0.04024 0.03718 0.00502 0.00304 0.00159 0.00159 0.00069 0.00013 0.00012 0.00004 0.00001
MBPG: periaqueductal gray in midbrain; MBDR: dorsal raphe nucleus in midbrain; PAR: reticular formation in pons; PMR: superior central nucleus and nucleus raphe magnus in pons; MOAR: reticular formation in medulla oblongata; MODNV: dorsal raphe nucleus in medulla oblongata; MOSN: solitary nucleus in medulla oblongata; MOMR: nucleus raphe obscuus, nucleus raphe magnus in medulla oblongata; MOAN: nucleus ambiguus in medulla oblongata; OAPD: duration of obstructive apnea (s); POSITION: in prone position or not.
4. Discussion Tables 2–5 show the contribution of each factor for being SIDS or not. Tables 6–9 show the contribution of each factor for being in the prone position or not. Compared between Tables 2 and 6, the duration of obstructive apnea was the most significant explanatory variable for being SIDS. This finding is reminiscent of previous reports of the potential importance of obstructive sleep apneas in SIDS [13,14]. In this study, the duration of obstructive apnea seemed to give more direct and stronger effect than the frequency of obstructive apnea as the hypoxic load. Both the effects of sleep position in Table 2 and of
‘‘CLASS (SIDS or non-SIDS)’’ in Table 6 were without weight. This result is not consistent with the general epidemiological knowledge of SIDS that shows the importance of the prone sleep position as a major risk factor of SIDS [15,16]. Comparing Tables 3 and 7, a slight difference was seen: the density of GFAP-positive reactive astrocytes in nucleus ambiguus in medulla oblongata played a major role with the combination of those in raphe nucleus in the midbrain and in pons in Table 3 and the density of GFAPpositive reactive astrocytes in the superior central nucleus. In addition, the nucleus raphe magnus in pons played a major role with the combination of other cardiorespiratory nucleus such as the solitary nucleus in the midbrain shown in Table 7.
Table 4 List of the combination of three explanatory variables in the calculation of AIC when the response variable is to be SIDS or non-SIDS No.
Explanatory variable
Number of categories of explanatory variable
AIC
Difference of AIC
Weight
1 2 3 4 5 6 7 8 9 10 11
MOAN, PMR, MBDR MOAN, PMR, MBPG MOAN, PMR, MOLR MOAN, PMR, MODNV MOAN, PMR, MOMR MOAN, PMR, MOSN MOAN, MBDR, MBPG PMR, MBDR, MBPG MOAN, PMR, PAR MOAN, PMR, POSITION MOAN, PMR, MOAR
18 12 12 12 18 12 18 12 24 12 18
24.87 12.29 10.70 10.70 6.02 5.76 4.72 4.57 4.44 2.88 0.47
0.00 12.58 14.17 14.17 18.85 19.11 20.15 20.30 20.43 21.99 24.40
1.00000 0.00186 0.00084 0.00084 0.00008 0.00007 0.00004 0.00004 0.00004 0.00002 0.00001
MBPG: periaqueductal gray in midbrain; MBDR: dorsal raphe nucleus in midbrain; PAR: reticular formation in pons; PMR: superior central nucleus and nucleus raphe magnus in pons; MOAR: reticular formation in medulla oblongata; MODNV: dorsal raphe nucleus in medulla oblongata; MOSN: solitary nucleus in medulla oblongata; MOMR: nucleus raphe obscuus, nucleus raphe magnus in medulla oblongata; MOAN: nucleus ambiguus in medulla oblongata; OAPD: duration of obstructive apnea (s); POSITION: in prone position or not.
S26
T. Sawaguchi et al. / Forensic Science International 130S (2002) S21–S29
Table 5 List of subsets of explanatory variables in the calculation of AIC when the response variable is to be SIDS or non-SIDS No.
Explanatory variable
Number of categories of explanatory variable
AIC
Difference of AIC
Weight
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
MOAN, PMR, MBDR MOAN, PMR, MBDR, MBPG MOAN, PMR, MBDR, MOLR MOAN, PMR, MBDR, MODNV MOAN, PMR, MBDR, MOMR MOAN, PMR, MBDR, MOSN MOAN, PMR MOAN, PMR, MBDR, PAR MOAN, PMR, MBPG MOAN, PMR, MBDR, MOAR MOAN, PMR, MOLR MOAN, PMR, MODNV MOAN, PMR, MBDR, POSITION MOAN, MBDR PMR, MBDR MOAN PMR MOAN, PMR, MOMR MOAN, PMR, MOSN MOAN, PAR MOAN, MBDR, MBPG PMR, MBDR, MBPG MOAN, PMR, PAR MOAN, MBPG MOAN, PMR, POSITION MOAN, MOLR, MOAN, MODNV MOAN, POSITION
18 36 36 36 54 36 6 72 12 54 12 12 36 9 6 3 2 18 12 12 18 12 24 6 6 6 6 6, 6
24.87 20.41 19.97 19.97 17.11 16.84 15.59 13.67 12.29 11.54 10.70 10.70 9.39 9.16 9.01 7.57 7.41 6.02 5.76 5.00 4.72 4.57 4.44 4.00 2.88 2.70 2.70 1.05, 0.77
0.00 4.46 4.90 4.90 7.76 8.03 9.28 11.20 12.58 13.33 14.17 14.17 15.48 15.71 15.86 17.30 17.46 18.85 19.11 19.87 20.15 20.30 20.43 20.87 21.99 22.17 22.17 23.82, 24.10
29 30
MOSN MOAN
2 18
1.00000 0.10752 0.08615 0.08615 0.02063 0.01805 0.00966 0.00370 0.00186 0.00128 0.00084 0.00084 0.00044 0.00039 0.00036 0.00017 0.00016 0.00008 0.00007 0.00005 0.00004 0.00004 0.00004 0.00003 0.00002 0.00002 0.00002 0.00001, 0.00001 0.00001 0.00001
0.77 0.47
24.10 24.40
MBPG: periaqueductal gray in midbrain; MBDR: dorsal raphe nucleus in midbrain; PAR: reticular formation in pons; PMR: superior central nucleus and nucleus raphe magnus in pons; MOAR: reticular formation in medulla oblongata; MODNV: dorsal raphe nucleus in medulla oblongata; MOSN: solitary nucleus in medulla oblongata; MOMR: nucleus raphe obscuus, nucleus raphe magnus in medulla oblongata; MOAN: nucleus ambiguus in medulla oblongata; POSITION: in prone position or not.
This tendency was also shown in the comparison between Tables 4 and 8 and between Tables 5 and 9. Gliosis in the arousal pathway, including raphe nucleus, may be related with SIDS. Tables 6–9 show that gliosis in periaqueductal gray matter in the midbrain was involved with a higher score than in Tables 2–5. Many projections come from periaqueductal gray matter in the midbrain to raphe nucleus and some of them refer to visceral alerting responses [17]. There are
major inputs to the pedunculopontine tegmental nucleus from the dorsal raphe and from the periaqueductal gray, that may all be important in the arousal process [18]. Each region of the arousal pathway associated in SIDS (Tables 2–5) and in a prone position (Tables 6–9) overlapped but were not completely similar. The following significant observations were made from comparisons between Tables 3–5 and Tables 7–9. The combination between being prone and the
Table 6 List of single explanatory variables in the calculation of AIC when the response variable is to be in prone position or not No.
Explanatory variable
Number of categories of explanatory variable
AIC
Difference of AIC
Weight
1
OAPD
7
274.68
0.00
1.00000
OAPD: duration of obstructive apnea (s).
T. Sawaguchi et al. / Forensic Science International 130S (2002) S21–S29
S27
Table 7 List of the combination of two explanatory variables in the calculation of AIC when the response variable is to be in prone position or not No.
Explanatory variable
Number of categories of explanatory variable
AIC
Difference of AIC
Weight
1 2 3 4 5 6 7 8 9 10
MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR,
10 10 6 14 4 10 14 18 14 14
4.22 1.47 1.05 0.58 0.43 0.59 1.28 3.98 4.39 5.30
0.00 2.75 3.17 3.64 4.65 4.81 5.51 8.21 8.61 9.52
1.00000 0.25253 0.20465 0.16173 0.09026 0.09026 0.06373 0.01652 0.01350 0.00855
MOSN MOLR MOAN MOMR CLASS MODNV MBPG PAR MOAR OAPD
MBPG: periaqueductal gray in midbrain; MBDR: dorsal raphe nucleus in midbrain; PAR: reticular formation in pons; PMR: superior central nucleus and nucleus raphe magnus in pons; MOAR: reticular formation in medulla oblongata; MODNV: dorsal raphe nucleus in medulla oblongata; MOSN: solitary nucleus in medulla oblongata; MOMR: nucleus raphe obscuus, nucleus raphe magnus in medulla oblongata; MOAN: nucleus ambiguus in medulla oblongata; CLASS: being SIDS or non-SIDS.
gliosis in the nucleus ambiguus in the medulla oblongata and superior central nucleus, nucleus raphe magnus and dorsal raphe in the midbrain is associated with being SIDS or non-SIDS as shown in Tables 2–5. The combination between the fact being SIDS or non-SIDS and gliosis in dorsal raphe in the midbrain and solitary nucleus in medulla oblongata is associated with being in a prone position or not. Gliosis only in dorsal raphe in the midbrain was common in both the groups. Therefore, dorsal raphe in the midbrain seemed to be the target cite in the pathophysiology in SIDS.
In addition, Tables 6–9 show the combination of the duration of obstructive apnea and gliosis in dorsal raphe in the midbrain is associated with being in a prone position. Tables 2–5 show the combination of the duration of obstructive apnea and gliosis in nucleus ambiguus in medulla oblongata is associated with being SIDS or non-SIDS. Duration of the obstructive apneas may be associated with the incidence of SIDS and infants’ sleep position through the linkage of particular sites in the brainstem. In Table 8, to determine being in a prone position, the combination of being SIDS, the density of GFAP-positive
Table 8 List of combination of three explanatory variables in the calculation of AIC when the response variable is to be in prone position or not No.
Explanatory variable
1 2 3 4 5 6 7 8 9 10 11 12 13 14
MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR,
MOSN, MOSN, MOSN, MOSN, MOSN, MOSN, MOSN, MOSN, MOSN, MOSN, MOSN, MOSN, MOSN, MOSN,
MOMR CLASS MBPG MOLR MOAN PAR OAPD MODNV PMR MOAR MOAR OAPF CAPD PeriodBR
Number of categories of explanatory variable
AIC
Difference of AIC
Weight
70 20 70 50 30 90 70 50 100 80 70 100 80 110
3.97 4.75 5.29 6.30 6.99 8.22 12.96 14.66 16.22 21.21 22.28 23.26 24.23 35.37
0.00 0.78 1.32 2.33 3.02 4.25 8.99 10.69 12.25 17.24 18.31 19.29 20.26 31.40
1.00000 0.67658 0.51634 0.31168 0.22123 0.11954 0.01116 0.00477 0.00219 0.00018 0.00011 0.00006 0.00004 0.00000
MBPG: periaqueductal gray in midbrain; MBDR: dorsal raphe nucleus in midbrain; PAR: reticular formation in pons; PMR: superior central nucleus and nucleus raphe magnus in pons; MOAR: reticular formation in medulla oblongata; MODNV: dorsal raphe nucleus in medulla oblongata; MOSN: solitary nucleus in medulla oblongata; MOMR: nucleus raphe obscuus, nucleus raphe magnus in medulla oblongata; MOAN: nucleus ambiguus in medulla oblongata; CAPD: duration of obstructive apnea (s); CLASS: being SIDS or non-SIDS; PeriodBR: periodic breathing.
S28
T. Sawaguchi et al. / Forensic Science International 130S (2002) S21–S29
Table 9 List of the summary of subsets of explanatory variables in the calculation of AIC when the response variable is to be in prone position or not No.
Explanatory variable
Number of categories of explanatory variable
AIC
Difference of AIC
Weight
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
MBDR MBDR, MBPG MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, CLASS MBDR, MBDR, OAPD MBDR, MBDR, MBDR, MBDR, MOSN MBDR, MOLR MBDR,
2 10 7 10 6 14 4 10 14 2 70 18 7 14 20 70 14 5 50 5 30
8.46 4.22 2.39 1.47 1.05 0.58 0.43 0.59 1.28 3.79 3.97 3.98 4.21 4.39 4.75 5.29 5.30 5.71 6.30 6.44 6.99
0.00 4.24 6.07 6.99 7.41 7.88 8.89 9.05 9.75 12.25 12.43 12.45 12.68 12.85 13.22 13.76 13.76 14,17 14.77 14.91 15.45
1.00000 0.12001 0.04796 0.03031 0.02456 0.01941 0.01173 0.01083 0.00765 0.00219 0.00199 0.00198 0.00177 0.00162 0.00135 0.00103 0.00103 0.00084 0.00062 0.00058 0.00044
MOSN MOLR MOAN MOMR CLASS MODNV MBPG MOSN, MOMR PAR, MOAR MOSN, CLASS MOSN, MBPG OAPD MOSN, MOLR MOSN, MOAN
MBPG: periaqueductal gray in midbrain; MBDR: dorsal raphe nucleus in midbrain; PAR: reticular formation in pons; MOAR: reticular formation in medulla oblongata; MODNV: dorsal raphe nucleus in medulla oblongata; MOSN: solitary nucleus in medulla oblongata; MOMR: nucleus raphe obscuus, nucleus raphe magnus in medulla oblongata; MOAN: nucleus ambiguus in medulla oblongata; OAPD: duration of obstructive apnea (s); CLASS: being SIDS or non-SIDS.
reactive astrocytes in the dorsal raphe of the midbrain and in the solitary nucleus of medulla oblongata was possible using its AIC score. However, the number of cases with sleep position information was so small in this study that this association might not be significant. Finally, comparisons of Tables 3 and 7, and between Tables 4 and 8 reveal the following findings. The effect of SIDS or non-SIDS in prone position in Tables 7 and 8 was stronger than the effect of prone position in SIDS in Tables 3 and 4 in terms of AIC scores and weights. In conclusion, the arousal pathway in infants seems to be fragile in the situation with hypoxic loads during the prone position. In addition, not only abnormality of the arousal phenomenon, but also some organic changes involving cardiorespiratory sites may contribute to SIDS.
Acknowledgements The authors sincerely thank Prof. C. De Prez for the kind help extended to us at her laboratory. This study was supported by Health Sciences Research Grants for Research on Children and Families from the Japanese Ministry of Health and Welfare, Satake Takako Award and Itoe Okamoto Award from Tokyo Women’s Medical University.
References [1] R.L. Naeye, Brain stem and adrenal abnormalities in the sudden infant death syndrome, Am. J. Clin. Pathol. 66 (1976) 526–539. [2] M. Oehmichen, P. Linke, K. Zilles, K.S. Saternus, Reactive astrocytes and macrophages in the brain stem of SIDS victims? Eleven age- and sex-matched SIDS and control cases, Clin. Neuropathol. 8 (1989) 276–283. [3] E. Reske-Nielsen, M. Gregersen, E. Lund, Astrocytes in the postnatal central nervous system, Acta Pathol. Microbiol. Immunol. Scand. 95 (1987) 347–356. [4] J. Pearson, L. Brandeis, Normal aspects of morphometry of brainstem astrocytes, carotid bodies, and ganglia in SIDS, in: J.T. Tildon, L.M. Roeder, A. Steinschneider (Eds.), Sudden Infant Death Syndrome, Academic Press, New York, 1983, pp. 115–122. [5] M.W. Ambler, C. Neave, W.Q. Sturner, Sudden and unexpected death in infancy and childhood: neuropathologic findings, Am. J. Forensic Med. Pathol. 2 (1981) 23–30. [6] H.C. Kinney, J.J. Filiano, Brainstem research in sudden infant death syndrome, J. Neuropathol. Exp. Neurol. 15 (1988) 240– 250. [7] H.C. Kinney, P.C. Burger, F.E. Harrell, R.P. Hudson, Reactive gliosis in the medulla oblongata of victims of the sudden infant death syndrome, Pediatrics 72 (1983) 181–187. [8] J.J. Filiano, H.C. Kinney, Sudden infant death syndrome and brainstem research, Pediatr. Ann. 24 (1995) 379–383.
T. Sawaguchi et al. / Forensic Science International 130S (2002) S21–S29 [9] S. Takashima, D. Armstrong, L.E. Becker, C. Bryan, Cerebral hypoperfusion in the sudden infant death syndrome? Brain stem gliosis and vasculature, Ann. Neurol. 4 (1978) 257– 262. [10] L.E. Becker, Neural maturation delay as a link in the chain of events leading to SIDS, Can. J. Neurol. Sci. 17 (1990) 361– 371. [11] G.C. Summers, V.C. Parker, The brainstem in the sudden infant death syndrome: a postmortem survey, Am. J. Forensic Med. Pathol. 2 (1981) 23–30. [12] H. Yamanouchi, S. Takashima, L.E. Becker, Correlation of astrogliosis and substance P immunoreactivity in the brainstem of victims of sudden infant death syndrome, Neuropediatrics 24 (1993) 200–203. [13] A. Kahn, J. Groswasser, E. Rebuffat, M. Sottiaux, D. Blum, M. Foester, P. Franco, A. Bochner, M. Alexander, A. Bachy, P. Richard, M. Verghote, D. LePolain, J.L. Wayenberg, Sleep and cardiorespiratory characteristics of infant victims of sudden death: a prospective case-control study, Sleep 15 (1992) 287–292. [14] A. Kahn, D. Blum, E. Rebuffat, M. Sottiaux, J. Levitt, A. Bochner, M. Alexander, J. Groswasser, M.F. Muller, Polysomnographic studies of infants who subsequently died of sudden infant death syndrome, Pediatrics 82 (1988) 721– 727. [15] A.C. Engelbert, G.A. de Jonge, Choice of sleeping position for infants: possible association with cot death, Arch. Dis. Child 65 (1990) 462–467.
S29
[16] AAP Task Force on Infant Positioning and SIDS: Positioning and SIDS, Pediatrics 89 (1992) 1120–1126. [17] R. Nieuwenhuys, J. Voogd, V.C. Huijen, The Human Central Nervous System Synopsis and Atlas, Springer, Berlin, 1998, p. 311, p. 319. [18] W.W. Blessing, The Lower Brainstem and Bodily Homeostasis, Oxford University Press, New York, 1997, p. 288. [19] M. Willinger, S. James, C. Catz, Defining the sudden infant death syndrome (SIDS), Pediatr. Pathol. 11 (1991) 677–684. [20] C. Guilleminault, M. Souquet, Sleep states and related pathology, in: R. Korobkin, C. Guilleminault (Eds.), Advances in Perinatal Neurology, Vol. 1, Spectrum Publications, New York, 1979, pp. 225–247. [21] C. Guilleminault, M. Souquet, R.L. Ariagno, R. Korobkin, F.B. Simmons, Five cases of near-miss sudden infant death syndrome and development of obstructive apnea syndrome, Pediatrics 73 (1984) 71–78. [22] D.H. Kelly, H. Golub, D. Carley, D.C. Shannon, Pneumograms in infants who subsequently died of sudden infant death syndrome, J. Pediatr. 109 (1986) 249–254. [23] D.H. Kelly, D.C. Shannon, Periodic breathing in infants with near-miss sudden infant death syndrome, Pediatrics 63 (1979) 355–360. [24] Y. Sakamoto, M. Ishiguro, G. Kitagawa, Akaike Information Criterion Statistics 3-222, D. Reidel Publishing Company, Dordrecht, 1986. [25] Y. Sakamoto, Categorical Data Analysis by AIC 1-161, Kluwer Academic Publishers, Dordrecht, 1991.
From epidemiology to physiology and pathology: apnea and arousal deficient theories in sudden infant death syndrome (SIDS)—with particular reference to hypoxic brainstem gliosis Toshiko Sawaguchia,*, Patricia Francob, Ineko Katoc, Satoru Shimizud, Hazim Kadhimb, Jose Groswasserb, Martine Sottiauxb, Hajime Togaric, Makio Kobayashie, Sachio Takashimaf, Hiroshi Nishidag, Akiko Sawaguchia, Andre Kahnb a
Department of Legal Medicine, School of Medicine, Tokyo Women’s Medical University, 8-1 Kawada-cho, Shinjuku-ku, Tokyo 162-8666, Japan b Pediatric Children’s Hospital Reine Fabiola, Free University of Brussels, Avenue J.J. Crocq 15, 1020 Brussells, Belgium c Department of Pediatrics, Nagoya City University, 1 Aza Kawasumi, Mizuho-cho, Mizuho-ku, Nagoya, Aichi 467-8601, Japan d Department of Hygiene and Public Health, School of Medicine, Tokyo Women’s Medical University, 8-1 Kawada-cho, Shinjuku-ku, Tokyo 162-8666, Japan e Department of Pathology, School of Medicine, Tokyo Women’s Medical University, 8-1 Kawada-cho, Shinjuku-ku, Tokyo 162-8666, Japan f National Institute of Neuroscience, National Center of Neurology and Psychiatry, 4-1-1 Ogawahigashi, Kodaira, Tokyo 187, Japan g Maternal and Perinatal Center, School of Medicine, Tokyo Women’s Medical University, 8-1 Kawada-cho, Shinjuku-ku, Tokyo 162-8666, Japan
Abstract Among 27,000 infants studied prospectively to characterize their sleep–wake behavior, 38 infants died under 6 months of age. They included 26 cases of sudden infant death syndrome (SIDS). Five infants who died from congenital cardiac abnormalities, two from infected pulmonary dysplasia, two from septic shock with multi-organ failure, one during a prolonged seizure, one from a prolonged neonatal hypoxemia, one from meningitis with brain infarction. All the infants had been recorded during one night in a pediatric sleep laboratory some 3–12 weeks before death. The frequency and duration of sleep apneas were analyzed. The infants’ brain stem material was collected and immunohistochemistry of glial fibrillary acidic protein (GFAP) was carried out. The density of GFAP-positive reactive astrocytes was measured in the cardiorespiratory and arousal pathway. Akaike information criterion statistics (AIC) were calculated to elucidate the relationship between the epidemiological data on sleep position, the physiological data and the pathological data in SIDS victims. The duration of obstructive apnea was the most significant variable to differentiate between SIDS victims and control infants. In conclusion, the present study sustains the possibility of an organic fragility within the arousal pathway in SIDS victims with repetitive sleep apneas. # 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Sudden infant death syndrome (SIDS); Brainstem; Gliosis; Apnea; Arousal
1. Introduction Studies of infants who died of sudden infant death syndrome (SIDS) reported the presence of subtle gliosis * Corresponding author. Tel.: þ81-3-52-697300; fax: þ81-3-52-697300. E-mail address: [email protected] (T. Sawaguchi).
within the reticular formation of the medulla [1]. Quantitative analysis of reactive astrocyte counts showed no difference between infants who died from SIDS and control subjects [2–6]. Alternatively, the density of astrocytes was shown to be higher in the brainstem of SIDS victims than in that of control infants at the level of the hilum of the inferior olive, the ventrolateral tegmental region and the nucleus tractus solitarius [7,8]. Other
0379-0738/02/$ – see front matter # 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 7 9 - 0 7 3 8 ( 0 2 ) 0 0 1 3 5 - 4
S22
T. Sawaguchi et al. / Forensic Science International 130S (2002) S21–S29
studies [9–12] reported brainstem gliosis in SIDS victims, at the level of the medullary reticular formation [11] and in the pontine reticular formation and pontine nuclei [12]. The present study was conducted to compare postmortem findings in SIDS victims with infants’ physiological sleep apnea characteristics and epidemiological data on sleep position.
2. Materials and methods 2.1. Physiological analyses 2.1.1. Subjects Over 20 years in various Belgian pediatric sleep laboratories 27,000 infants were studied prospectively to determine sleep–wake characteristics. The families were invited to join the study when leaving the maternity unit. The infants entered the study if they were born at term after a normal gestation, and if there was no family or personal history of apnea, apparent life threatening event (ALTE) or SIDS. At the time of recording, the infants were aged between 2 and 27 weeks, were healthy, and were receiving no medication. Following the sleep recording, 38 infants died suddenly. Following postmortem examinations and death scene investigations, 27 of the 38 infants were considered as SIDS victims [19]; five died from congenital cardiac abnormalities; two from infected pulmonary dysplasia; two from septic shock; one from a prolonged seizure; one from prolonged neonatal hypoxemia. The main characteristics of the dead infants are reported in Table 1. The maximum delay between the estimated time of death and the postmortem examination was 24 h. 2.1.2. Polygraphic monitoring For each infant, 8 h sleep studies were conducted overnight in a sleep laboratory, following standard recording techniques [13,14,20,21]. The recordings were done in a quiet and darkened room, at an ambient temperature ranging between 20 and 23 8C. All infants slept supine, without restraints. Recording started around 21:00 h. The infants were observed continuously during recording. They were fed on demand. Their behavior and any nursing intervention were charted. No infant had a pacifier during the recorded sessions. The following variables were recorded simultaneously: two scalp electroencephalogram with central and occipital leads, two electrooculograms and electrocardiogram. Thoracic respiratory movements were measured by impedance and airflow with thermistors taped under both nostrils and on the side of the mouth. Oxygen saturation was recorded continuously by a transcutaneous sensor (Nellcor, USA). Gross body movements were measured using an actigram placed on one arm. The data was collected on a computerized infant sleep recorder (Alice recording system III, Healthdyne, USA).
2.1.3. Data interpretation Based on the polygraphic recordings, sleep stages and sleep apneas were scored according to standard definitions [13,14,20–23]. Apneas were scored when they lasted 3 s or longer. They were classified as central apnea when flat tracings were obtained simultaneously from the strain gauges and thermistors. Periodic breathing was defined as the succession of more than two central apneas separated from each other by less than 20 s. Obstructive apneas were defined as continuous deflections from the strain gauges, with a flat tracing recorded from the thermistors. A mixed apnea was defined as a central apnea directly followed by an obstructive episode. Mixed apneas were scored together with the obstructive episodes. The frequency of obstructive apneas was measured by dividing the total number of apnea by the total sleep (in min) and multiplying by 60. The type, frequency (number per h of sleep) and duration (in s) of sleep apneas were computed. The recordings were analyzed visually by two independent scorers without knowledge of subjects’ age or sex to ensure reliability. Discrepancies were discussed and the agreed upon score was computed for analysis. 2.2. Pathological analyses 2.2.1. Subjects analyzed A total of 48 paraffin blocks were collected from the brain stems of the 38 infants: 7 blocks from the midbrain, 22 from the pons, and 19 from the medulla oblongata. The delay between death and postmortem examination was within 24 h. 2.2.2. Standard histological brain examination Hematoxylin-eosin (HE) staining was carried out as the standard histological brain staining. 2.2.3. Immunohistochemical examination Immunohistochemical studies using anti-glial fibrillary acidic protein (GFAP, Fig. 1) antibodies (DAKO, USA) was carried out and visualized by the HRP-DAB method using 4 mm thick sections. Five overlapped microscopic fields at 40 magnification were selected from several sites and the number of GFAP-positive reactive astrocytes was visually counted. Measurements were carried out in the following sites: periaqueductal gray matter, substantia nigra (pars compacta and pars reticulata), red nucleus and dorsal raphe nucleus of the midbrain; reticular formation, superior central nucleus, pontine nucleus, locus cerruleus and nucleus raphe magnus of the pons; dorsal vagus nucleus, solitary nucleus, reticular formation, nucleus raphe obscurus, nucleus raphe magnus, nucleus medulla oblongata centralis, olive nucleus and nucleus ambiguus of the medulla oblongata. For quantification, the superior central nucleus and nucleus raphe magnus of the pons, as well as the nucleus raphe obscurus and nucleus raphe magnus of the medulla oblongata were enclosed.
T. Sawaguchi et al. / Forensic Science International 130S (2002) S21–S29
S23
Table 1 General details of each case Case no.
Sex
Gestational age (weeks)
Postneonatal age (weeks)
Cause of death
SIDS cases 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
F F M F F M M F M M F M M M M F F M M M M M M F F M
40 38 40 41 37 40 39 40 39 38 40 39 40 38 40 36 40 38 39 37 35 40 40 40 37 31
16 13 12 18 19 16 12 14 21 10 11 22 40 36 10 19 3 4 6 18 6 22 31 31 20 20
SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS SIDS
Control cases 1 2 3 4 5 6 7 8 9 10 11 12
M M F M M M F M M F M F
39 40 40 33 40 37 37 39 40 37 39 40
24 11 6 21 14 4 5 6 7 7 9 9
Meningitis and brain infarction Pneumonia Mypcarditis Varisella Cardiopathy with pulmonary hypertension Syndrome of Opitz Hepatitis and general infection Bronchopneumopathy Myocarditis Infanticide Bronchopneumopathy Bronchopneumopathy
The pathological measurements were done twice by the same pathologist and quantitized data with a large standard deviation were recounted or rejected. 2.2.4. Data analysis The scorers of the sleep recordings as well as the pathologist did not known the causes of the infants’ death. Akaike information criterion statistics (AIC) [24,25] was calculated using each infant’s data including epidemiological data on sleep position, physiological data on sleep apnea, pathological data on the density of GFAP-positive reactive astrocytes as follows:
(1) to be SIDS or non-SIDS as response variables in each case with single, two or three explanatory variables (see Tables 2–5); (2) to be in a prone position or not as response variables in each case with single, two or three explanatory variables (see Tables 6–9).
2.3. Ethical issues The study was approved by the Ethical Committee of the University Children’s Hospital and was performed in
S24
T. Sawaguchi et al. / Forensic Science International 130S (2002) S21–S29
Fig. 1. Immunohistochemistry of glial fibrillary acidic protein in the brain of a SIDS victim.
accordance with the ethical standards prescribed by the Declaration of Helsinki.
3. Results Results of the calculation of AIC are shown in Tables 2–5 to determine being SIDS and Tables 6–9 to clarify the prone position. In Table 2, to determine the criteria for being SIDS or non-SIDS, the duration of obstructive apnea was the most significant as a single explanatory variable, followed by the frequency of central apnea. In Table 3, to determine being SIDS or non-SIDS, the gliosis in nucleus ambiguus in the medulla oblongata and superior central nucleus and nucleus magnus in the pons was the most significant, followed by the giosis in the nucleus ambiguus within the medulla oblongata and dorsal raphe nucleus of the midbrain. In Table 4, to determine being SIDS or non-SIDS, the gliosis in nucleus ambiguus in the medulla oblongata and superior central nucleus and nucleus magnus in pons and
dorsal raphe nucleus in the midbrain was the most significant finding. In Table 5, to determine being SIDS or non-SIDS, the gliosis in the nucleus ambiguus with the medulla oblongata and superior central nucleus and nucleus magnus in pons and dorsal raphe nucleus in the midbrain was the most significant finding. In Table 6, to determine being in the prone position, the duration of obstructive apnea was the most significant finding. In Table 7, to determine being in the prone position, the gliosis in the dorsal raphe nucleus of the midbrain and solitary nucleus in medulla oblongata was the most significant finding. In Table 8, to determine being in the prone position, the gliosis in the dorsal raphe in the midbrain and solitary nucleus, nucleus raphe obscurus and nucleus raphe magnus in medulla oblongata was the most significant, followed by the gliosis in dorsal raphe in the midbrain, solitary nucleus in medulla oblongata and being SIDS or non-SIDS. In Table 9, to determine being in the prone position or not, the gliosis in dorsal raphe in the midbrain was the largest subset of explanatory variables.
Table 2 List of single explanatory variables in the calculation of AIC when the response variable is to be SIDS or non-SIDS No.
Explanatory variable
Number of categories of explanatory variable
AIC
Difference of AIC
Weight
1 2
OAPD CAPF
5 7
100.64 82.91
0.00 17.73
1.00000 0.00014
OAPD: duration of obstructive apnea (s); CAPF: frequency of central apnea (no. per h).
T. Sawaguchi et al. / Forensic Science International 130S (2002) S21–S29
S25
Table 3 List of the combination of two explanatory variables in the calculation of AIC when the response variable is to be SIDS or non-SIDS No.
Explanatory variable
Number of categories of explanatory variable
AIC
Difference of AIC
Weight
1 2 3 4 5 6 7 8 9 10 11 12
MOAN, PMR MOAN, MBDR PMR, MBDR MOAN, PAR MOAN, MBPG MOAN, MOLR MOAN, MODNV MOAN, POSITION MOAN, MOMR MOAN, MOSN MOAN, OAPD MOAN, MOAR
6 9 6 12 6 6 6 6 9 6 15 9
15.59 9.16 9.01 5.00 4.00 2.70 2.70 1.05 2.25 2.52 4.46 7.79
0.00 6.43 6.58 10.59 11.59 12.89 12.89 14.54 17.84 18.11 20.05 23.38
1.00000 0.04024 0.03718 0.00502 0.00304 0.00159 0.00159 0.00069 0.00013 0.00012 0.00004 0.00001
MBPG: periaqueductal gray in midbrain; MBDR: dorsal raphe nucleus in midbrain; PAR: reticular formation in pons; PMR: superior central nucleus and nucleus raphe magnus in pons; MOAR: reticular formation in medulla oblongata; MODNV: dorsal raphe nucleus in medulla oblongata; MOSN: solitary nucleus in medulla oblongata; MOMR: nucleus raphe obscuus, nucleus raphe magnus in medulla oblongata; MOAN: nucleus ambiguus in medulla oblongata; OAPD: duration of obstructive apnea (s); POSITION: in prone position or not.
4. Discussion Tables 2–5 show the contribution of each factor for being SIDS or not. Tables 6–9 show the contribution of each factor for being in the prone position or not. Compared between Tables 2 and 6, the duration of obstructive apnea was the most significant explanatory variable for being SIDS. This finding is reminiscent of previous reports of the potential importance of obstructive sleep apneas in SIDS [13,14]. In this study, the duration of obstructive apnea seemed to give more direct and stronger effect than the frequency of obstructive apnea as the hypoxic load. Both the effects of sleep position in Table 2 and of
‘‘CLASS (SIDS or non-SIDS)’’ in Table 6 were without weight. This result is not consistent with the general epidemiological knowledge of SIDS that shows the importance of the prone sleep position as a major risk factor of SIDS [15,16]. Comparing Tables 3 and 7, a slight difference was seen: the density of GFAP-positive reactive astrocytes in nucleus ambiguus in medulla oblongata played a major role with the combination of those in raphe nucleus in the midbrain and in pons in Table 3 and the density of GFAPpositive reactive astrocytes in the superior central nucleus. In addition, the nucleus raphe magnus in pons played a major role with the combination of other cardiorespiratory nucleus such as the solitary nucleus in the midbrain shown in Table 7.
Table 4 List of the combination of three explanatory variables in the calculation of AIC when the response variable is to be SIDS or non-SIDS No.
Explanatory variable
Number of categories of explanatory variable
AIC
Difference of AIC
Weight
1 2 3 4 5 6 7 8 9 10 11
MOAN, PMR, MBDR MOAN, PMR, MBPG MOAN, PMR, MOLR MOAN, PMR, MODNV MOAN, PMR, MOMR MOAN, PMR, MOSN MOAN, MBDR, MBPG PMR, MBDR, MBPG MOAN, PMR, PAR MOAN, PMR, POSITION MOAN, PMR, MOAR
18 12 12 12 18 12 18 12 24 12 18
24.87 12.29 10.70 10.70 6.02 5.76 4.72 4.57 4.44 2.88 0.47
0.00 12.58 14.17 14.17 18.85 19.11 20.15 20.30 20.43 21.99 24.40
1.00000 0.00186 0.00084 0.00084 0.00008 0.00007 0.00004 0.00004 0.00004 0.00002 0.00001
MBPG: periaqueductal gray in midbrain; MBDR: dorsal raphe nucleus in midbrain; PAR: reticular formation in pons; PMR: superior central nucleus and nucleus raphe magnus in pons; MOAR: reticular formation in medulla oblongata; MODNV: dorsal raphe nucleus in medulla oblongata; MOSN: solitary nucleus in medulla oblongata; MOMR: nucleus raphe obscuus, nucleus raphe magnus in medulla oblongata; MOAN: nucleus ambiguus in medulla oblongata; OAPD: duration of obstructive apnea (s); POSITION: in prone position or not.
S26
T. Sawaguchi et al. / Forensic Science International 130S (2002) S21–S29
Table 5 List of subsets of explanatory variables in the calculation of AIC when the response variable is to be SIDS or non-SIDS No.
Explanatory variable
Number of categories of explanatory variable
AIC
Difference of AIC
Weight
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
MOAN, PMR, MBDR MOAN, PMR, MBDR, MBPG MOAN, PMR, MBDR, MOLR MOAN, PMR, MBDR, MODNV MOAN, PMR, MBDR, MOMR MOAN, PMR, MBDR, MOSN MOAN, PMR MOAN, PMR, MBDR, PAR MOAN, PMR, MBPG MOAN, PMR, MBDR, MOAR MOAN, PMR, MOLR MOAN, PMR, MODNV MOAN, PMR, MBDR, POSITION MOAN, MBDR PMR, MBDR MOAN PMR MOAN, PMR, MOMR MOAN, PMR, MOSN MOAN, PAR MOAN, MBDR, MBPG PMR, MBDR, MBPG MOAN, PMR, PAR MOAN, MBPG MOAN, PMR, POSITION MOAN, MOLR, MOAN, MODNV MOAN, POSITION
18 36 36 36 54 36 6 72 12 54 12 12 36 9 6 3 2 18 12 12 18 12 24 6 6 6 6 6, 6
24.87 20.41 19.97 19.97 17.11 16.84 15.59 13.67 12.29 11.54 10.70 10.70 9.39 9.16 9.01 7.57 7.41 6.02 5.76 5.00 4.72 4.57 4.44 4.00 2.88 2.70 2.70 1.05, 0.77
0.00 4.46 4.90 4.90 7.76 8.03 9.28 11.20 12.58 13.33 14.17 14.17 15.48 15.71 15.86 17.30 17.46 18.85 19.11 19.87 20.15 20.30 20.43 20.87 21.99 22.17 22.17 23.82, 24.10
29 30
MOSN MOAN
2 18
1.00000 0.10752 0.08615 0.08615 0.02063 0.01805 0.00966 0.00370 0.00186 0.00128 0.00084 0.00084 0.00044 0.00039 0.00036 0.00017 0.00016 0.00008 0.00007 0.00005 0.00004 0.00004 0.00004 0.00003 0.00002 0.00002 0.00002 0.00001, 0.00001 0.00001 0.00001
0.77 0.47
24.10 24.40
MBPG: periaqueductal gray in midbrain; MBDR: dorsal raphe nucleus in midbrain; PAR: reticular formation in pons; PMR: superior central nucleus and nucleus raphe magnus in pons; MOAR: reticular formation in medulla oblongata; MODNV: dorsal raphe nucleus in medulla oblongata; MOSN: solitary nucleus in medulla oblongata; MOMR: nucleus raphe obscuus, nucleus raphe magnus in medulla oblongata; MOAN: nucleus ambiguus in medulla oblongata; POSITION: in prone position or not.
This tendency was also shown in the comparison between Tables 4 and 8 and between Tables 5 and 9. Gliosis in the arousal pathway, including raphe nucleus, may be related with SIDS. Tables 6–9 show that gliosis in periaqueductal gray matter in the midbrain was involved with a higher score than in Tables 2–5. Many projections come from periaqueductal gray matter in the midbrain to raphe nucleus and some of them refer to visceral alerting responses [17]. There are
major inputs to the pedunculopontine tegmental nucleus from the dorsal raphe and from the periaqueductal gray, that may all be important in the arousal process [18]. Each region of the arousal pathway associated in SIDS (Tables 2–5) and in a prone position (Tables 6–9) overlapped but were not completely similar. The following significant observations were made from comparisons between Tables 3–5 and Tables 7–9. The combination between being prone and the
Table 6 List of single explanatory variables in the calculation of AIC when the response variable is to be in prone position or not No.
Explanatory variable
Number of categories of explanatory variable
AIC
Difference of AIC
Weight
1
OAPD
7
274.68
0.00
1.00000
OAPD: duration of obstructive apnea (s).
T. Sawaguchi et al. / Forensic Science International 130S (2002) S21–S29
S27
Table 7 List of the combination of two explanatory variables in the calculation of AIC when the response variable is to be in prone position or not No.
Explanatory variable
Number of categories of explanatory variable
AIC
Difference of AIC
Weight
1 2 3 4 5 6 7 8 9 10
MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR,
10 10 6 14 4 10 14 18 14 14
4.22 1.47 1.05 0.58 0.43 0.59 1.28 3.98 4.39 5.30
0.00 2.75 3.17 3.64 4.65 4.81 5.51 8.21 8.61 9.52
1.00000 0.25253 0.20465 0.16173 0.09026 0.09026 0.06373 0.01652 0.01350 0.00855
MOSN MOLR MOAN MOMR CLASS MODNV MBPG PAR MOAR OAPD
MBPG: periaqueductal gray in midbrain; MBDR: dorsal raphe nucleus in midbrain; PAR: reticular formation in pons; PMR: superior central nucleus and nucleus raphe magnus in pons; MOAR: reticular formation in medulla oblongata; MODNV: dorsal raphe nucleus in medulla oblongata; MOSN: solitary nucleus in medulla oblongata; MOMR: nucleus raphe obscuus, nucleus raphe magnus in medulla oblongata; MOAN: nucleus ambiguus in medulla oblongata; CLASS: being SIDS or non-SIDS.
gliosis in the nucleus ambiguus in the medulla oblongata and superior central nucleus, nucleus raphe magnus and dorsal raphe in the midbrain is associated with being SIDS or non-SIDS as shown in Tables 2–5. The combination between the fact being SIDS or non-SIDS and gliosis in dorsal raphe in the midbrain and solitary nucleus in medulla oblongata is associated with being in a prone position or not. Gliosis only in dorsal raphe in the midbrain was common in both the groups. Therefore, dorsal raphe in the midbrain seemed to be the target cite in the pathophysiology in SIDS.
In addition, Tables 6–9 show the combination of the duration of obstructive apnea and gliosis in dorsal raphe in the midbrain is associated with being in a prone position. Tables 2–5 show the combination of the duration of obstructive apnea and gliosis in nucleus ambiguus in medulla oblongata is associated with being SIDS or non-SIDS. Duration of the obstructive apneas may be associated with the incidence of SIDS and infants’ sleep position through the linkage of particular sites in the brainstem. In Table 8, to determine being in a prone position, the combination of being SIDS, the density of GFAP-positive
Table 8 List of combination of three explanatory variables in the calculation of AIC when the response variable is to be in prone position or not No.
Explanatory variable
1 2 3 4 5 6 7 8 9 10 11 12 13 14
MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, MBDR,
MOSN, MOSN, MOSN, MOSN, MOSN, MOSN, MOSN, MOSN, MOSN, MOSN, MOSN, MOSN, MOSN, MOSN,
MOMR CLASS MBPG MOLR MOAN PAR OAPD MODNV PMR MOAR MOAR OAPF CAPD PeriodBR
Number of categories of explanatory variable
AIC
Difference of AIC
Weight
70 20 70 50 30 90 70 50 100 80 70 100 80 110
3.97 4.75 5.29 6.30 6.99 8.22 12.96 14.66 16.22 21.21 22.28 23.26 24.23 35.37
0.00 0.78 1.32 2.33 3.02 4.25 8.99 10.69 12.25 17.24 18.31 19.29 20.26 31.40
1.00000 0.67658 0.51634 0.31168 0.22123 0.11954 0.01116 0.00477 0.00219 0.00018 0.00011 0.00006 0.00004 0.00000
MBPG: periaqueductal gray in midbrain; MBDR: dorsal raphe nucleus in midbrain; PAR: reticular formation in pons; PMR: superior central nucleus and nucleus raphe magnus in pons; MOAR: reticular formation in medulla oblongata; MODNV: dorsal raphe nucleus in medulla oblongata; MOSN: solitary nucleus in medulla oblongata; MOMR: nucleus raphe obscuus, nucleus raphe magnus in medulla oblongata; MOAN: nucleus ambiguus in medulla oblongata; CAPD: duration of obstructive apnea (s); CLASS: being SIDS or non-SIDS; PeriodBR: periodic breathing.
S28
T. Sawaguchi et al. / Forensic Science International 130S (2002) S21–S29
Table 9 List of the summary of subsets of explanatory variables in the calculation of AIC when the response variable is to be in prone position or not No.
Explanatory variable
Number of categories of explanatory variable
AIC
Difference of AIC
Weight
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
MBDR MBDR, MBPG MBDR, MBDR, MBDR, MBDR, MBDR, MBDR, CLASS MBDR, MBDR, OAPD MBDR, MBDR, MBDR, MBDR, MOSN MBDR, MOLR MBDR,
2 10 7 10 6 14 4 10 14 2 70 18 7 14 20 70 14 5 50 5 30
8.46 4.22 2.39 1.47 1.05 0.58 0.43 0.59 1.28 3.79 3.97 3.98 4.21 4.39 4.75 5.29 5.30 5.71 6.30 6.44 6.99
0.00 4.24 6.07 6.99 7.41 7.88 8.89 9.05 9.75 12.25 12.43 12.45 12.68 12.85 13.22 13.76 13.76 14,17 14.77 14.91 15.45
1.00000 0.12001 0.04796 0.03031 0.02456 0.01941 0.01173 0.01083 0.00765 0.00219 0.00199 0.00198 0.00177 0.00162 0.00135 0.00103 0.00103 0.00084 0.00062 0.00058 0.00044
MOSN MOLR MOAN MOMR CLASS MODNV MBPG MOSN, MOMR PAR, MOAR MOSN, CLASS MOSN, MBPG OAPD MOSN, MOLR MOSN, MOAN
MBPG: periaqueductal gray in midbrain; MBDR: dorsal raphe nucleus in midbrain; PAR: reticular formation in pons; MOAR: reticular formation in medulla oblongata; MODNV: dorsal raphe nucleus in medulla oblongata; MOSN: solitary nucleus in medulla oblongata; MOMR: nucleus raphe obscuus, nucleus raphe magnus in medulla oblongata; MOAN: nucleus ambiguus in medulla oblongata; OAPD: duration of obstructive apnea (s); CLASS: being SIDS or non-SIDS.
reactive astrocytes in the dorsal raphe of the midbrain and in the solitary nucleus of medulla oblongata was possible using its AIC score. However, the number of cases with sleep position information was so small in this study that this association might not be significant. Finally, comparisons of Tables 3 and 7, and between Tables 4 and 8 reveal the following findings. The effect of SIDS or non-SIDS in prone position in Tables 7 and 8 was stronger than the effect of prone position in SIDS in Tables 3 and 4 in terms of AIC scores and weights. In conclusion, the arousal pathway in infants seems to be fragile in the situation with hypoxic loads during the prone position. In addition, not only abnormality of the arousal phenomenon, but also some organic changes involving cardiorespiratory sites may contribute to SIDS.
Acknowledgements The authors sincerely thank Prof. C. De Prez for the kind help extended to us at her laboratory. This study was supported by Health Sciences Research Grants for Research on Children and Families from the Japanese Ministry of Health and Welfare, Satake Takako Award and Itoe Okamoto Award from Tokyo Women’s Medical University.
References [1] R.L. Naeye, Brain stem and adrenal abnormalities in the sudden infant death syndrome, Am. J. Clin. Pathol. 66 (1976) 526–539. [2] M. Oehmichen, P. Linke, K. Zilles, K.S. Saternus, Reactive astrocytes and macrophages in the brain stem of SIDS victims? Eleven age- and sex-matched SIDS and control cases, Clin. Neuropathol. 8 (1989) 276–283. [3] E. Reske-Nielsen, M. Gregersen, E. Lund, Astrocytes in the postnatal central nervous system, Acta Pathol. Microbiol. Immunol. Scand. 95 (1987) 347–356. [4] J. Pearson, L. Brandeis, Normal aspects of morphometry of brainstem astrocytes, carotid bodies, and ganglia in SIDS, in: J.T. Tildon, L.M. Roeder, A. Steinschneider (Eds.), Sudden Infant Death Syndrome, Academic Press, New York, 1983, pp. 115–122. [5] M.W. Ambler, C. Neave, W.Q. Sturner, Sudden and unexpected death in infancy and childhood: neuropathologic findings, Am. J. Forensic Med. Pathol. 2 (1981) 23–30. [6] H.C. Kinney, J.J. Filiano, Brainstem research in sudden infant death syndrome, J. Neuropathol. Exp. Neurol. 15 (1988) 240– 250. [7] H.C. Kinney, P.C. Burger, F.E. Harrell, R.P. Hudson, Reactive gliosis in the medulla oblongata of victims of the sudden infant death syndrome, Pediatrics 72 (1983) 181–187. [8] J.J. Filiano, H.C. Kinney, Sudden infant death syndrome and brainstem research, Pediatr. Ann. 24 (1995) 379–383.
T. Sawaguchi et al. / Forensic Science International 130S (2002) S21–S29 [9] S. Takashima, D. Armstrong, L.E. Becker, C. Bryan, Cerebral hypoperfusion in the sudden infant death syndrome? Brain stem gliosis and vasculature, Ann. Neurol. 4 (1978) 257– 262. [10] L.E. Becker, Neural maturation delay as a link in the chain of events leading to SIDS, Can. J. Neurol. Sci. 17 (1990) 361– 371. [11] G.C. Summers, V.C. Parker, The brainstem in the sudden infant death syndrome: a postmortem survey, Am. J. Forensic Med. Pathol. 2 (1981) 23–30. [12] H. Yamanouchi, S. Takashima, L.E. Becker, Correlation of astrogliosis and substance P immunoreactivity in the brainstem of victims of sudden infant death syndrome, Neuropediatrics 24 (1993) 200–203. [13] A. Kahn, J. Groswasser, E. Rebuffat, M. Sottiaux, D. Blum, M. Foester, P. Franco, A. Bochner, M. Alexander, A. Bachy, P. Richard, M. Verghote, D. LePolain, J.L. Wayenberg, Sleep and cardiorespiratory characteristics of infant victims of sudden death: a prospective case-control study, Sleep 15 (1992) 287–292. [14] A. Kahn, D. Blum, E. Rebuffat, M. Sottiaux, J. Levitt, A. Bochner, M. Alexander, J. Groswasser, M.F. Muller, Polysomnographic studies of infants who subsequently died of sudden infant death syndrome, Pediatrics 82 (1988) 721– 727. [15] A.C. Engelbert, G.A. de Jonge, Choice of sleeping position for infants: possible association with cot death, Arch. Dis. Child 65 (1990) 462–467.
S29
[16] AAP Task Force on Infant Positioning and SIDS: Positioning and SIDS, Pediatrics 89 (1992) 1120–1126. [17] R. Nieuwenhuys, J. Voogd, V.C. Huijen, The Human Central Nervous System Synopsis and Atlas, Springer, Berlin, 1998, p. 311, p. 319. [18] W.W. Blessing, The Lower Brainstem and Bodily Homeostasis, Oxford University Press, New York, 1997, p. 288. [19] M. Willinger, S. James, C. Catz, Defining the sudden infant death syndrome (SIDS), Pediatr. Pathol. 11 (1991) 677–684. [20] C. Guilleminault, M. Souquet, Sleep states and related pathology, in: R. Korobkin, C. Guilleminault (Eds.), Advances in Perinatal Neurology, Vol. 1, Spectrum Publications, New York, 1979, pp. 225–247. [21] C. Guilleminault, M. Souquet, R.L. Ariagno, R. Korobkin, F.B. Simmons, Five cases of near-miss sudden infant death syndrome and development of obstructive apnea syndrome, Pediatrics 73 (1984) 71–78. [22] D.H. Kelly, H. Golub, D. Carley, D.C. Shannon, Pneumograms in infants who subsequently died of sudden infant death syndrome, J. Pediatr. 109 (1986) 249–254. [23] D.H. Kelly, D.C. Shannon, Periodic breathing in infants with near-miss sudden infant death syndrome, Pediatrics 63 (1979) 355–360. [24] Y. Sakamoto, M. Ishiguro, G. Kitagawa, Akaike Information Criterion Statistics 3-222, D. Reidel Publishing Company, Dordrecht, 1986. [25] Y. Sakamoto, Categorical Data Analysis by AIC 1-161, Kluwer Academic Publishers, Dordrecht, 1991.