- Email: [email protected]
Smaller mandibular size in infants with a history of an apparent life-threatening event
SMALLER MANDIBULAR SIZE IN INFANTS WITH A HISTORY OF AN APPARENT LIFE-THREATENING EVENT MARY H. HORN, RN, MS, RRT, DANIEL D. KINNAMON, MS, NALTON FERRARO, DMD, MD, MARTHA A.Q. CURLEY, RN, PHD, FAAN
AND
Objectives To examine small mandibular size and preference for a hand-to-chin posture as salient characteristics in infants with a history of an apparent life-threatening event (ALTE).
Study design This was a prospective case-control study of term infants, from birth to 6 months of age, admitted post-ALTE and matched 1:2 with healthy control infants (age within 2 weeks and weight within 0.5 kg). Infants with confirmed gastroesophageal reflux and congenital anomalies, including severe micrognathia, were excluded. Results Infants with a history of an ALTE (n ⴝ 25) were matched to 47 healthy controls. Infants with a history of an ALTE had mandibular indices (larger index indicates a smaller mandible) that were 3.8 mm greater on the left side (95% CI: 2.0-5.6, P < .001) and 4.2 mm greater on the right side (95% CI: 2.7-5.6, P < .001) adjusting for length and non-white race. Controlling for matching and length, a 1-mm increase in the average mandibular index increased the odds of an ALTE by 62% (OR ⴝ 1.62, 95% CI: 1.22-2.44, P < .001). Conclusions Smaller mandibular size was associated with ALTE, suggesting airway obstruction as a potential cause of ALTE. The association of this characteristic with ALTE also offers the potential for prospective quantification of ALTE risk. (J Pediatr 2006;149:499-504)
n apparent life-threatening event (ALTE) is a frightening episode with some combination of apnea, color change, loss of muscle tone, and choking or gagging.1 An ALTE occurs in approximately 6% of all term infants between 8 and 14 weeks of age. After an ALTE, the well-appearing infant is usually brought to the emergency department and then admitted to the hospital for observation using cardiac and pulse oximeter oxygen saturation (SpO2) monitoring for at least 24 hours and a full diagnostic evaluation. Before hospital discharge, parents are required to demonstrate basic skills in infant cardiopulmonary resuscitation (CPR) so that they can intervene if the ALTE reoccurs. After teaching CPR skills to numerous parents of infants with a history of an ALTE, the principal investigator (MH) recognized that this population of infants appeared to possess a similar physical characteristic: a mandible that appears smaller than normal. It was also noted that these infants frequently assume a similar body posture; specifically, a hand-to-chin position that appears to create a jaw thrust. Furthermore, ALTEs are often precipitated by a position that creates posterior displacement of the small jaw; for example, snuggling a breast-fed infant, using an infant swing or car seat with the head portion of the chair tilted back, or holding the infant in a supine position without completely supporting the head. From Surgical Nursing, Critical Care and Cardiovascular Nursing, and Oral and MaxAlthough the specific cause is unknown, a small mandible has not been linked to illofacial Surgery, Children’s Hospital BosALTEs. Further elucidation of the pathophysiology of ALTEs my help prevent some of ton; and the Division of Pediatric Clinical 2,3 the significant neurological deficits experienced by some infants post-ALTE. The Research, the Department of Pediatrics, the University of Miami Miller School of purpose of the study is to evaluate the role of smaller mandibular size and the preference Medicine, Florida. of hand-to-chin posture in infants presenting after an ALTE. Supported by the Center for Nursing, Chil-
A
METHODS Subjects This is a prospective case-control study of term infants, from birth to 6 months of age, admitted to Children’s Hospital Boston post-ALTE over a year’s time and matched ALTE CPR GN N
Apparent life-threatening event Cardiopulmonary resuscitation Gnathion Nasion
OBI OSA SIDS
Otobasion inferius Obstructive sleep apnea Sudden infant death syndrome
dren’s Hospital, Boston; The Thoracic Foundation. Submitted for publication Aug 24, 2005; last revision received Mar 27, 2006; accepted Jun 9, 2006. Reprint requests: Mary Horn; 10S; Children’s Hospital Boston; 300 Longwood Ave; Boston, MA 02115. E-mail: mary.horn@ childrens.harvard.edu. 0022-3476/$ - see front matter Copyright © 2006 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2006.06.018
499
Figure. The otobasion inferius (OBI) is located at the lower border of ear attachment to the face. Soft tissue nasion (N) is the point of maximum concavity between nose and forehead in midsagittal plane. Soft tissue gnathion (GN) is located in the midsagittal plane of the mandible, where the anterior curve in the outline of the chin merges into the body of the mandible. (Drawings courtesy of Hanan Althuwaini and Pam Dodds.)
1:2 with a convenience sample of healthy control infants (age within 2 weeks and weight within 0.5 kg). All inpatient infant units were screened for potential candidates Monday through Friday. Patients younger than 36 weeks postconceptual age, those diagnosed with gastroesophageal reflux or with a congenital anomaly, including micrognathia, were excluded. To avoid misdiagnosis and potential conflict of interest, parents were approached for consent only after the ALTE diagnosis was confirmed by an attending physician after a diagnostic workup and after pre-discharge CPR training was complete. As patients with a history of an ALTE were enrolled, healthy control infants were recruited from a well-baby clinic, a pediatric ambulatory practice, or an inpatient unit. Control patients recruited from the wards were hospitalized for nonpulmonary minor medical or surgical problems. The study was approved by the Committee on Clinical Investigation.
Mandibular Measurements Analysis of mandibular size was based on two mandibular indices calculated from four separate jaw measurements made by a single investigator (MH) and confirmed by a separate observer. Each measurement required investigator and observer agreement. Measurements were right- and left-side otobasion inferius (OBI) to soft tissue nasion (N) and OBI to soft tissue gnathion (GN) in the supine position with head midline.4 As illustrated in the Figure, the OBI is located at the lower border of ear attachment to the face. Soft tissue N is the point of maximum concavity between nose and forehead in the midsagittal plane. Soft tissue GN is located in the midsagittal plane of the mandible, where the anterior curve in the outline of the chin merges into the body of the mandible. The mandibular index for a given side (right or left) is defined as the difference in the OBI-N and OBI-GN measurements, with a larger mandibular index indicating a smaller mandibular size. For example, when the OBI-N measures 93 mm and the OBI-GN measures 80 mm, the mandibular index is 13 500
Horn et al
mm. The supine position was chosen because of its ease of access and because it is the recommended sleep/rest position for infants. Landmarks were first mapped on the patient’s skin using a washable nontoxic marking pen. Measurements were then obtained with the head in midline position using a narrow, washable plastic-covered cloth millimeter tape measure fitting over hard tissue. The tape measure was calibrated against a paper tape measure to control for stretching of the plastic tape with repeated use over time. All measurements were obtained in a quiet infant without a pacifier.
Parent Interview Patient demographic information and past medical history were obtained through chart review and were then confirmed during the parent interview. Parents of infants with a history of an ALTE were asked 10 questions to help them systematically describe the ALTE event; specifically, the infant’s activity preceding the event, the relationship of the event to feeding, and the location, body movements, airway control, breathing pattern, heart rate, color change, positioning, and length of the event. Both groups of parents were queried about family history of apnea, obstructive sleep apnea (OSA), snoring, and the infant’s preferences regarding handchin position at rest and body position at rest. Statistical Analysis Characteristics of the two groups were compared using exact conditional logistic regression stratified by matched group for dichotomous variables5 and linear regression with a fixed effect to control for matching for continuous variables. Linear regression models for mandibular index with a fixed effect to control for matching were fit separately on the left side and the right side to determine whether sex, non-white race, or length significantly affected the mandibular index or confounded the estimate of the relationship of mandibular The Journal of Pediatrics • October 2006
Table I. Infant characteristics according to ALTE status ALTE (n ⴝ 25)
Variable
Control (N ⴝ 47)
P value*
25 (53) 16 (35)
0.51 1.00
# (%) Female Non-white†
16 (64) 8 (32)
Age (days) Weight (kg) Length (cm)‡ Left mandibular index (mm) Right mandibular index (mm)
Mean ⫾ SD (Min, Max) 56 ⫾ 50 (3, 180) 52 ⫾ 48 (4, 192) 4.7 ⫾ 1.2 (2.7, 7.2) 4.6 ⫾ 1.3 (2.4, 7.5) 56.1 ⫾ 5.1 (48.0, 68.5) 56.0 ⫾ 5.7 (46.0, 71.0) 17.8 ⫾ 3.3 (10.0, 25.0) 14.3 ⫾ 3.3 (10.0, 22.5) 17.4 ⫾ 3.0 (10.0, 20.0) 13.8 ⫾ 3.8 (5.0, 25.0)
0.56 0.93 0.51 ⬍.001 ⬍.001
*Exact conditional logistic regression for dichotomous variables and linear regression with a fixed effect to control for matching for continuous variables. †n ⫽ 46 for the control group. ‡n ⫽ 45 for the control group.
index to history of an ALTE. Generalized additive models were used to assess potential nonlinearities in the relationship between length and mandibular index, and a quadratic relationship was included in the final model for both sides. A hypothesized difference in the relationship between mandibular index and history of an ALTE between case-control groups in which all members were ⬍3 months of age and the remaining groups was tested with an interaction term. Risk factors for ALTE were then explored using exact conditional logistic regression stratified by matched group. All P values were calculated using the exact conditional scores test, and point estimates and 95% confidence intervals for the odds ratios were obtained from the exact conditional probability density function.6 Univariate effects of clinically important risk factors including sex; non-white race; length; parental and sibling history of apnea, OSA, or snoring; multiple infant preferences for resting position; parental report of preference for hand-to-chin posture; and the average of left and right mandibular indices were examined. Backward selection at the 5% significance level with assessment of confounding was used to select a multivariate model from the significant univariate risk factors. Nonlinearities of continuous variables were assessed by adding tertiles of the continuous variable to the model including the linear term. A hypothesized difference in the relationship of the odds of an ALTE to mandibular index between case-control groups with all members ⬍3 months of age and the remaining groups was tested with an interaction term. Statistical analyses were performed using SAS/STAT software, Version 9.1.3 of the SAS System for Windows (SAS Institute Inc., Cary, NC). All P values were two-sided and considered significant if ⬍.05.
RESULTS Twenty-five infants with a confirmed ALTE diagnosis of unknown etiology were matched to 50 healthy control infants. (Table I). Three control patients were eliminated because their matches were not exact or because their measurement data were incomplete. Patients with ALTE were almost 8 weeks of age, 64% female, and predominately white
(68%; 4% black/non-Hispanic, 20% Hispanic, 8% other races). Parents of infants with ALTE described the following: 44% of the events occurred while the infant was sleeping; 44% were not related to feedings; and 36% of events occurred in the parent’s arms. Fifty-two percent of parents reported infant movement during the event most often (36%) involving head extension. Forty-eight percent reported that the infant was choking during the event, whereas 48% of the infants were reported to be attempting respiratory efforts. Thirty-two percent were noted to have a heart rate. Ninety-two percent experienced a color change most often described as blue (60%) or pale (40%) and most often surrounding the mouth (56%). ALTEs occurred in the following positions: 44% supine position, 36% sitting, 8% prone, 4% side-lying, and 8% other (held on parent’s shoulder). Events were reported to last a median of 30 seconds (range: 2-240). Based on left and right mandibular indices, infants who experienced ALTEs had a smaller mandibular size (Table I). The mandibular index of infants who experienced ALTEs was greater by 3.4 mm on the left side (95% CI: 1.6-5.3, P ⬍ .001) and 3.4 mm on the right side (95% CI: 1.7-5.1, P ⬍ .001) when adjusting for matching alone. Adjusting for matching, length, and non-white race, infants who experienced ALTEs showed mandibular indices that were 3.8 mm greater on the left side (95% CI: 2.0-5.6, P ⬍ .001) and 4.2 mm greater on the right side (95% CI: 2.7-5.6, P ⬍ .001). These results were not significantly influenced by age or sex. Table II presents significant risk factors for the occurrence of an ALTE in univariate analyses. A preference for hand-to-chin posture, parental history of snoring, and a larger average mandibular index were all significant risk factors for the occurrence of an ALTE (P ⱕ .006 for each variable). Sex; non-white race; length; parental history of apnea or OSA; sibling history of apnea, OSA, or snoring; and infant preferences for multiple sleeping positions were not associated with the occurrence of an ALTE. Backward selection yielded a model containing only average mandibular index. Preference for hand-to-chin pos-
Smaller Mandibular Size In Infants With A History Of An Apparent Life-Threatening Event
501
Table II. Significant univariate risk factors for an ALTE Variable
ALTE (n ⴝ 25)
Control (n ⴝ 47)
OR (95% CI)*
P value*
# (%) Preference for hand-to-chin posture†‡ Parental history of snoring‡
Average of left and right mandibular indices (mm)
17 (68) 13 (52)
13 (29) 8 (18)
4.77 (1.47, 20.26) 5.50 (1.43, 31.08)
Mean ⫾ SD (Min, Max) 17.6 ⫾ 2.4 (12.5, 22.5) 14.1 ⫾ 3.3 (7.5, 23.8)
1.36 (1.14, 1.68)
.004 .006 ⬍.001
*Exact conditional logistic regression of occurrence of an ALTE on a given risk factor. †Preference for hand-to-chin posture includes parent report of infant frequently assuming a position that includes hand(s) under or to side of chin. ‡n ⫽ 45 for the control group.
ture and parental history of snoring were not independent risk factors. Adding length to the model increased the coefficient on average mandibular index by 58%, and so length was included in the final model as a confounder. There was no evidence of nonlinearities in the effects of length (P ⫽ .57) or average mandibular index (P ⫽ 1.00). The effect of mandibular index on the odds of ALTE was not significantly different in matched groups in which all members were ⬍3 months of age compared with the remaining groups (interaction P ⫽ 1.00). Thus, in the final model including length and average mandibular index, a 1-mm increase in the average mandibular index increased the odds of an ALTE by 62% (OR ⫽ 1.62, 95% CI: 1.22-2.44, P ⬍ .001). For the adjusted difference in average mandibular index of about 4 mm between infants with a history of ALTE and controls observed in this study, this would translate to an increase of almost 7 times (OR ⫽ 6.87, 95% CI: 2.19-35.21) in the odds of an ALTE.
DISCUSSION Our results support the clinical observation that infants diagnosed with ALTE have smaller mandibles. Specifically, a smaller mandibular size increased an infant’s odds of experiencing an ALTE. Infant preference for hand-to-chin posture was not an independent risk factor in multivariate analysis. Our results must be considered within the context of a large number of clinical conditions that can manifest as an ALTE. These can include, but are not limited to, sleep apnea, apnea of prematurity, gastroesophageal reflux, cardiac arrhythmias, sepsis, and seizures as well as neurological, metabolic, or respiratory problems.7-9 In our study, all patients were diagnosed with an ALTE by their attending physician after consideration of their history, physical exam, and diagnostic workup.10 Our patients who had an ALTE underwent a wide variety of diagnostic tests (Table III). The goal of this investigation was to support the clinical observation that smaller mandibular size and hand-to-chin posture are linked to ALTE in at least some of the affected infants. Clearly, an infant with an ALTE attributed to a medical diagnosis such as gastroesophageal reflux may have, in addition, a smaller mandible. It is certainly plausible that mandibular hypoplasia may contribute to momentary or intermittent obstruction of the upper airway in some ALTE cases. At birth, the infant’s jaw is almost horizontal and flexible, and it can be easily 502
Horn et al
shifted backward.11 An obstructive event, especially with cyanosis and change of consciousness, would frighten a parent and result in an admission to the hospital for an ALTE workup. Although ALTE and obstructive sleep apnea (OSA) are not the same, there can certainly be some overlap in the underlying physiology that triggers both. Mandibular hypoplasia is a well-documented etiology of OSA in some patients. For example, in adults, the length of the mandibular plane and posterior airspace are inversely related to symptom levels in patients with OSA.9 Similarly, children with OSA show the same skeletal and soft tissue configurations as adults with sleep apnea.12 Recently, Arens et al13 used magnetic resonance imaging as a means of identifying anatomical factors consistent with OSA in school-aged patients (3-7 years of age). Their data suggest a positive correlation between increased adenoid and tonsil volume and an apneic index. A number of anatomical abnormalities developed in infancy may contribute to sleep disordered breathing in otherwise normal children 3 to 13 years of age with nasal obstruction. In addition, a large tongue can partially occlude the airway.11,14 Tonkin et al noted that an infant’s airway is vulnerable to unbalanced external forces that tend to narrow the airway.11 For example, Stark and Thach showed that mandibular pressure from a face mask can cause backward displacement of the jaw leading to airway obstruction.15 Muscle tone, neuromuscular control, and soft-tissue function are critical factors that control the airway during sleep.16 Finkelstein et al17 found that infants as young as 3 weeks of age have OSA with nasal resistance that, in combination, may be related to later abnormal facial development. One of their findings in these infants was shortened mandibular length. Furthermore, sudden infant death syndrome (SIDS) and ALTE are defined by many as distinct entities, but the relationship between the two cannot be ignored. There is evidence that some victims of SIDS had had ALTEs before the terminal event. Although the relationship between ALTEs and SIDS has not been systematically defined, there is a positive history of an ALTE in 5% to 10% of SIDS victims.2,3 Tonkin18 hypothesized that the cause of SIDS was a vulnerable airway at the oropharyngeal level between the soft palate and the base of the skull. Airway occlusion at this level might occur during muscle relaxation (which occurs The Journal of Pediatrics • October 2006
Table III. Clinical workup of ALTE Screening exam Electrocardiogram Chest x-ray Hematocrit Lumbar puncture Electroencephalogram pH probe Sleep study Head CT Scan Barium swallow Echocardiogram Upper GI series Head MRI Pneumogram Spinal ultrasound Bronchoscopy Formal consults
Frequency
Patient ID number
23 18 15 7 7 6 5 5 2 2 2 2 1 1 1 18
1,3,4,5,7,8,9,10,11,12,13,18,19,21,22,23,25,27,29,30,31,32,33 1,3,4,5,8,9,11,12,13,18,19,20,22,27,29,30,32,33 1,3,4,5,8,10,11,12,13,18,19,20,27,29,30 1,3,12,13,18,19,20 1,4,5,12,31,32,33 4,5,7,12,19,20 4,5,7,12,20 1,5,10,31,33 3,4 5,18 18,19 4,5 19 22 4 Neurology: 1, 4 Pulmonary: 4 Gastroenterology: 1,3,4,5,7,19,29 Otolaryngology: 1,4 Lactation specialist: 1,2,20,22 Cardiology: 18,19
Abbreviations: CT, computed tomography; GI, gastrointestinal; MRI, Magnetic resonance imaging.
during rapid eye movement [REM] sleep) facilitated by a hypermobile mandible or an enlarged tongue with a strong backward sucking action developed by artificial feeding of the infant. A recent study showed that upper airway size changes in supine position during magnetic resonance imaging and during tidal breathing in children with obstructive sleep apnea.19,20 Byard and Kennedy21 noted that SIDS may occur in infants with mandibular hypoplasia without a history of oxygen desaturation. They recommended that jaw size be systematically measured in all postmortem exams of SIDS infants. In addition, the observation of a small mandible in some victims of SIDS must be considered seriously. At this point in our understanding of SIDS and ALTEs, both appear to have many possible causes. The hand-to-chin positioning results were interesting. Parents were questioned about their infants’ hand position during sleep. Parents were quite specific as to whether the infant spent the majority of sleep time with the hand positioned under the chin or whether the hand and arms were in an open posture away from the chin. The parents’ observations are the basis for the hand-to-chin positioning finding. The differences between the ALTE and control groups in observed posture indicate the potential importance of this observation. One possible explanation for the posturing is a trial-and-error phenomenon that results in clinical relief of symptoms. All this can take place even in an infant and below the level of conscious decision. One of the authors (NFF) has treated a 4-year-old-boy with obstructive sleep apnea who slept in a sitting position on the floor with his chin supported
against the wall. This was most likely not the result of an understanding of the upper airway anatomy but the discovery of a position that provided relief. Although a preference for the hand-to-chin posture is associated with higher odds of ALTE, this positioning may not create an obstruction but rather may serve as a marker for infants with airway obstruction and illustrate how they compensate. In this study, parents were not asked about handchin location during the event but were asked about the infant’s usual preference. The infant may not have been able to assume this posture at that time of the ALTE, and therefore had an ALTE because of other cumulative factors. A limitation of this study is its case-control design. Because of the design, incidence rates of ALTE for a given combination of risk factors could not be estimated. Group assignment was not blinded, and thus facial measurements, though confirmed by a second observer, could have been subject to observer bias. After enrollment and measurement, neither group was followed for subsequent problems. Patients with ALTE may have developed an exclusion criterion, and control patients may have presented with an ALTE. A large, prospective cohort study through infancy would be necessary to address these issues. In summary, smaller mandibular size was associated with ALTE, suggesting airway obstruction as a potential cause of ALTE. The association of this characteristic with ALTE offers the potential for prospective quantification of ALTE risk. It is our hope that an assessment of this salient aspect of an infant’s appearance be evaluated in a well-baby exam.7,8,18
Smaller Mandibular Size In Infants With A History Of An Apparent Life-Threatening Event
503
We thank David Wypij, PhD, for statistical consultation, and Susan Longobucco-Hynes, Linda Dempsey, Julia Mancuso Perkins, Rosella Mancuso, Kathy Nygren Austin, and Herminia Shermont for their collegial support.
REFERENCES 1. National Institutes of Health Consensus Development Conference on Infantile Apnea and Home Monitoring, Sept 29 to Oct 1, 1986. Pediatrics 1987;79:292-9. 2. Beal S. Apparent life threatening events with serious sequelae in infants and young children. J Paediatr Child Health 1992;28:151-5. 3. Hunt CE. Sudden infant death syndrome and other causes of infant mortality: diagnosis, mechanisms, and risk for recurrence in siblings. Am J Respir Crit Care Med 2001;164:346-57. 4. Farkas LG. Athropometry of the Head and Face. 2nd ed. New York: Raven Press; 1994. 5. Hirji F, Mehta C, Patel N. Exact inference for matched case-control studies. Biometrics 1988;44:803-14. 6. SAS Institute. SAS 9.1.3 Help and Documentation. Cary, NC: SAS Institute Inc.; 2002-2004. 7. Brooks JG. Apparent life-threatening events and apnea of infancy. Clin Perinatol 1992;19:809-38. 8. Davies F, Gupta R. Apparent life threatening events in infants presenting to an emergency department. Emerg Med J 2002;19:11-6. 9. King B. Pediatric apnea. eMed J 2001;2:1-9. 10. Brand DA, Altman RL, Purtill K, Edwards KS. Yield of diagnostic testing in infants who have had an apparent life-threatening event. Pediatrics 2005;115:885-93. 11. Tonkin SL, Gunn TR, Bennet L, Vogel SA, Gunn AJ. A review of the
504
Horn et al
anatomy of the upper airway in early infancy and its possible relevance to SIDS. Early Hum Dev 2002;66:107-21. 12. See CC, Newman LJ, Berezin S, Glassman MS, Medow MS, Dozor AJ, et al. Gastroesophageal reflux-induced hypoxemia in infants with apparent life-threatening event(s). Am J Dis Child 1989;143:951-4. 13. Arens R, McDonough JM, Costarino AT, Mahboubi S, Tayag-Kier CE, Maislin G, et al. Magnetic resonance imaging of the upper airway structure of children with obstructive sleep apnea syndrome. Am J Respir Crit Care Med 2001;164:698-703. 14. Siebert JR, Haas JE. Enlargement of the tongue in sudden infant death syndrome. Pediatr Pathol 1991;11:813-26. 15. Stark AR, Thach BT. Mechanisms of airway obstruction leading to apnea in newborn infants. J Pediatr 1976;89:982-5. 16. Ferraro N. Craniofacial development and the airway during sleep. In: Loughlim G, Marcus C, eds. Sleep and Breathing in Children a Developmental Approach. New York: Mercel Dekker Inc; 2000:293-309. 17. Finkelstein Y, Wexler D, Horowitz E, Berger G, Nachmani A, Shapiro-Feinberg M, et al. Frontal and lateral cephalometry in patients with sleep-disordered breathing. Laryngoscope 2001;111(4 Pt 1):634-41. 18. Tonkin S. Sudden infant death syndrome: hypothesis of causation. Pediatrics 1975;55:650-61. 19. Arens R, Sin S, McDonough JM, Palmer JM, Dominguez T, Meyer H, et al. Changes in upper airway size during tidal breathing in children with obstructive sleep apnea syndrome. Am J Respir Crit Care Med 2005;171:1298-304. 20. Tsuiki S, Almeida FR, Bhalla PS, Lowe AA, Fleetham JA. Supinedependent changes in upper airway size in awake obstructive sleep apnea patients. Sleep Breath 2003;7:43-50. 21. Byard RW, Kennedy JD. Diagnostic difficulties in cases of sudden death in infants with mandibular hypoplasia. Am J Forensic Med Pathol 1996;17:255-9.
The Journal of Pediatrics • October 2006
AND
Objectives To examine small mandibular size and preference for a hand-to-chin posture as salient characteristics in infants with a history of an apparent life-threatening event (ALTE).
Study design This was a prospective case-control study of term infants, from birth to 6 months of age, admitted post-ALTE and matched 1:2 with healthy control infants (age within 2 weeks and weight within 0.5 kg). Infants with confirmed gastroesophageal reflux and congenital anomalies, including severe micrognathia, were excluded. Results Infants with a history of an ALTE (n ⴝ 25) were matched to 47 healthy controls. Infants with a history of an ALTE had mandibular indices (larger index indicates a smaller mandible) that were 3.8 mm greater on the left side (95% CI: 2.0-5.6, P < .001) and 4.2 mm greater on the right side (95% CI: 2.7-5.6, P < .001) adjusting for length and non-white race. Controlling for matching and length, a 1-mm increase in the average mandibular index increased the odds of an ALTE by 62% (OR ⴝ 1.62, 95% CI: 1.22-2.44, P < .001). Conclusions Smaller mandibular size was associated with ALTE, suggesting airway obstruction as a potential cause of ALTE. The association of this characteristic with ALTE also offers the potential for prospective quantification of ALTE risk. (J Pediatr 2006;149:499-504)
n apparent life-threatening event (ALTE) is a frightening episode with some combination of apnea, color change, loss of muscle tone, and choking or gagging.1 An ALTE occurs in approximately 6% of all term infants between 8 and 14 weeks of age. After an ALTE, the well-appearing infant is usually brought to the emergency department and then admitted to the hospital for observation using cardiac and pulse oximeter oxygen saturation (SpO2) monitoring for at least 24 hours and a full diagnostic evaluation. Before hospital discharge, parents are required to demonstrate basic skills in infant cardiopulmonary resuscitation (CPR) so that they can intervene if the ALTE reoccurs. After teaching CPR skills to numerous parents of infants with a history of an ALTE, the principal investigator (MH) recognized that this population of infants appeared to possess a similar physical characteristic: a mandible that appears smaller than normal. It was also noted that these infants frequently assume a similar body posture; specifically, a hand-to-chin position that appears to create a jaw thrust. Furthermore, ALTEs are often precipitated by a position that creates posterior displacement of the small jaw; for example, snuggling a breast-fed infant, using an infant swing or car seat with the head portion of the chair tilted back, or holding the infant in a supine position without completely supporting the head. From Surgical Nursing, Critical Care and Cardiovascular Nursing, and Oral and MaxAlthough the specific cause is unknown, a small mandible has not been linked to illofacial Surgery, Children’s Hospital BosALTEs. Further elucidation of the pathophysiology of ALTEs my help prevent some of ton; and the Division of Pediatric Clinical 2,3 the significant neurological deficits experienced by some infants post-ALTE. The Research, the Department of Pediatrics, the University of Miami Miller School of purpose of the study is to evaluate the role of smaller mandibular size and the preference Medicine, Florida. of hand-to-chin posture in infants presenting after an ALTE. Supported by the Center for Nursing, Chil-
A
METHODS Subjects This is a prospective case-control study of term infants, from birth to 6 months of age, admitted to Children’s Hospital Boston post-ALTE over a year’s time and matched ALTE CPR GN N
Apparent life-threatening event Cardiopulmonary resuscitation Gnathion Nasion
OBI OSA SIDS
Otobasion inferius Obstructive sleep apnea Sudden infant death syndrome
dren’s Hospital, Boston; The Thoracic Foundation. Submitted for publication Aug 24, 2005; last revision received Mar 27, 2006; accepted Jun 9, 2006. Reprint requests: Mary Horn; 10S; Children’s Hospital Boston; 300 Longwood Ave; Boston, MA 02115. E-mail: mary.horn@ childrens.harvard.edu. 0022-3476/$ - see front matter Copyright © 2006 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2006.06.018
499
Figure. The otobasion inferius (OBI) is located at the lower border of ear attachment to the face. Soft tissue nasion (N) is the point of maximum concavity between nose and forehead in midsagittal plane. Soft tissue gnathion (GN) is located in the midsagittal plane of the mandible, where the anterior curve in the outline of the chin merges into the body of the mandible. (Drawings courtesy of Hanan Althuwaini and Pam Dodds.)
1:2 with a convenience sample of healthy control infants (age within 2 weeks and weight within 0.5 kg). All inpatient infant units were screened for potential candidates Monday through Friday. Patients younger than 36 weeks postconceptual age, those diagnosed with gastroesophageal reflux or with a congenital anomaly, including micrognathia, were excluded. To avoid misdiagnosis and potential conflict of interest, parents were approached for consent only after the ALTE diagnosis was confirmed by an attending physician after a diagnostic workup and after pre-discharge CPR training was complete. As patients with a history of an ALTE were enrolled, healthy control infants were recruited from a well-baby clinic, a pediatric ambulatory practice, or an inpatient unit. Control patients recruited from the wards were hospitalized for nonpulmonary minor medical or surgical problems. The study was approved by the Committee on Clinical Investigation.
Mandibular Measurements Analysis of mandibular size was based on two mandibular indices calculated from four separate jaw measurements made by a single investigator (MH) and confirmed by a separate observer. Each measurement required investigator and observer agreement. Measurements were right- and left-side otobasion inferius (OBI) to soft tissue nasion (N) and OBI to soft tissue gnathion (GN) in the supine position with head midline.4 As illustrated in the Figure, the OBI is located at the lower border of ear attachment to the face. Soft tissue N is the point of maximum concavity between nose and forehead in the midsagittal plane. Soft tissue GN is located in the midsagittal plane of the mandible, where the anterior curve in the outline of the chin merges into the body of the mandible. The mandibular index for a given side (right or left) is defined as the difference in the OBI-N and OBI-GN measurements, with a larger mandibular index indicating a smaller mandibular size. For example, when the OBI-N measures 93 mm and the OBI-GN measures 80 mm, the mandibular index is 13 500
Horn et al
mm. The supine position was chosen because of its ease of access and because it is the recommended sleep/rest position for infants. Landmarks were first mapped on the patient’s skin using a washable nontoxic marking pen. Measurements were then obtained with the head in midline position using a narrow, washable plastic-covered cloth millimeter tape measure fitting over hard tissue. The tape measure was calibrated against a paper tape measure to control for stretching of the plastic tape with repeated use over time. All measurements were obtained in a quiet infant without a pacifier.
Parent Interview Patient demographic information and past medical history were obtained through chart review and were then confirmed during the parent interview. Parents of infants with a history of an ALTE were asked 10 questions to help them systematically describe the ALTE event; specifically, the infant’s activity preceding the event, the relationship of the event to feeding, and the location, body movements, airway control, breathing pattern, heart rate, color change, positioning, and length of the event. Both groups of parents were queried about family history of apnea, obstructive sleep apnea (OSA), snoring, and the infant’s preferences regarding handchin position at rest and body position at rest. Statistical Analysis Characteristics of the two groups were compared using exact conditional logistic regression stratified by matched group for dichotomous variables5 and linear regression with a fixed effect to control for matching for continuous variables. Linear regression models for mandibular index with a fixed effect to control for matching were fit separately on the left side and the right side to determine whether sex, non-white race, or length significantly affected the mandibular index or confounded the estimate of the relationship of mandibular The Journal of Pediatrics • October 2006
Table I. Infant characteristics according to ALTE status ALTE (n ⴝ 25)
Variable
Control (N ⴝ 47)
P value*
25 (53) 16 (35)
0.51 1.00
# (%) Female Non-white†
16 (64) 8 (32)
Age (days) Weight (kg) Length (cm)‡ Left mandibular index (mm) Right mandibular index (mm)
Mean ⫾ SD (Min, Max) 56 ⫾ 50 (3, 180) 52 ⫾ 48 (4, 192) 4.7 ⫾ 1.2 (2.7, 7.2) 4.6 ⫾ 1.3 (2.4, 7.5) 56.1 ⫾ 5.1 (48.0, 68.5) 56.0 ⫾ 5.7 (46.0, 71.0) 17.8 ⫾ 3.3 (10.0, 25.0) 14.3 ⫾ 3.3 (10.0, 22.5) 17.4 ⫾ 3.0 (10.0, 20.0) 13.8 ⫾ 3.8 (5.0, 25.0)
0.56 0.93 0.51 ⬍.001 ⬍.001
*Exact conditional logistic regression for dichotomous variables and linear regression with a fixed effect to control for matching for continuous variables. †n ⫽ 46 for the control group. ‡n ⫽ 45 for the control group.
index to history of an ALTE. Generalized additive models were used to assess potential nonlinearities in the relationship between length and mandibular index, and a quadratic relationship was included in the final model for both sides. A hypothesized difference in the relationship between mandibular index and history of an ALTE between case-control groups in which all members were ⬍3 months of age and the remaining groups was tested with an interaction term. Risk factors for ALTE were then explored using exact conditional logistic regression stratified by matched group. All P values were calculated using the exact conditional scores test, and point estimates and 95% confidence intervals for the odds ratios were obtained from the exact conditional probability density function.6 Univariate effects of clinically important risk factors including sex; non-white race; length; parental and sibling history of apnea, OSA, or snoring; multiple infant preferences for resting position; parental report of preference for hand-to-chin posture; and the average of left and right mandibular indices were examined. Backward selection at the 5% significance level with assessment of confounding was used to select a multivariate model from the significant univariate risk factors. Nonlinearities of continuous variables were assessed by adding tertiles of the continuous variable to the model including the linear term. A hypothesized difference in the relationship of the odds of an ALTE to mandibular index between case-control groups with all members ⬍3 months of age and the remaining groups was tested with an interaction term. Statistical analyses were performed using SAS/STAT software, Version 9.1.3 of the SAS System for Windows (SAS Institute Inc., Cary, NC). All P values were two-sided and considered significant if ⬍.05.
RESULTS Twenty-five infants with a confirmed ALTE diagnosis of unknown etiology were matched to 50 healthy control infants. (Table I). Three control patients were eliminated because their matches were not exact or because their measurement data were incomplete. Patients with ALTE were almost 8 weeks of age, 64% female, and predominately white
(68%; 4% black/non-Hispanic, 20% Hispanic, 8% other races). Parents of infants with ALTE described the following: 44% of the events occurred while the infant was sleeping; 44% were not related to feedings; and 36% of events occurred in the parent’s arms. Fifty-two percent of parents reported infant movement during the event most often (36%) involving head extension. Forty-eight percent reported that the infant was choking during the event, whereas 48% of the infants were reported to be attempting respiratory efforts. Thirty-two percent were noted to have a heart rate. Ninety-two percent experienced a color change most often described as blue (60%) or pale (40%) and most often surrounding the mouth (56%). ALTEs occurred in the following positions: 44% supine position, 36% sitting, 8% prone, 4% side-lying, and 8% other (held on parent’s shoulder). Events were reported to last a median of 30 seconds (range: 2-240). Based on left and right mandibular indices, infants who experienced ALTEs had a smaller mandibular size (Table I). The mandibular index of infants who experienced ALTEs was greater by 3.4 mm on the left side (95% CI: 1.6-5.3, P ⬍ .001) and 3.4 mm on the right side (95% CI: 1.7-5.1, P ⬍ .001) when adjusting for matching alone. Adjusting for matching, length, and non-white race, infants who experienced ALTEs showed mandibular indices that were 3.8 mm greater on the left side (95% CI: 2.0-5.6, P ⬍ .001) and 4.2 mm greater on the right side (95% CI: 2.7-5.6, P ⬍ .001). These results were not significantly influenced by age or sex. Table II presents significant risk factors for the occurrence of an ALTE in univariate analyses. A preference for hand-to-chin posture, parental history of snoring, and a larger average mandibular index were all significant risk factors for the occurrence of an ALTE (P ⱕ .006 for each variable). Sex; non-white race; length; parental history of apnea or OSA; sibling history of apnea, OSA, or snoring; and infant preferences for multiple sleeping positions were not associated with the occurrence of an ALTE. Backward selection yielded a model containing only average mandibular index. Preference for hand-to-chin pos-
Smaller Mandibular Size In Infants With A History Of An Apparent Life-Threatening Event
501
Table II. Significant univariate risk factors for an ALTE Variable
ALTE (n ⴝ 25)
Control (n ⴝ 47)
OR (95% CI)*
P value*
# (%) Preference for hand-to-chin posture†‡ Parental history of snoring‡
Average of left and right mandibular indices (mm)
17 (68) 13 (52)
13 (29) 8 (18)
4.77 (1.47, 20.26) 5.50 (1.43, 31.08)
Mean ⫾ SD (Min, Max) 17.6 ⫾ 2.4 (12.5, 22.5) 14.1 ⫾ 3.3 (7.5, 23.8)
1.36 (1.14, 1.68)
.004 .006 ⬍.001
*Exact conditional logistic regression of occurrence of an ALTE on a given risk factor. †Preference for hand-to-chin posture includes parent report of infant frequently assuming a position that includes hand(s) under or to side of chin. ‡n ⫽ 45 for the control group.
ture and parental history of snoring were not independent risk factors. Adding length to the model increased the coefficient on average mandibular index by 58%, and so length was included in the final model as a confounder. There was no evidence of nonlinearities in the effects of length (P ⫽ .57) or average mandibular index (P ⫽ 1.00). The effect of mandibular index on the odds of ALTE was not significantly different in matched groups in which all members were ⬍3 months of age compared with the remaining groups (interaction P ⫽ 1.00). Thus, in the final model including length and average mandibular index, a 1-mm increase in the average mandibular index increased the odds of an ALTE by 62% (OR ⫽ 1.62, 95% CI: 1.22-2.44, P ⬍ .001). For the adjusted difference in average mandibular index of about 4 mm between infants with a history of ALTE and controls observed in this study, this would translate to an increase of almost 7 times (OR ⫽ 6.87, 95% CI: 2.19-35.21) in the odds of an ALTE.
DISCUSSION Our results support the clinical observation that infants diagnosed with ALTE have smaller mandibles. Specifically, a smaller mandibular size increased an infant’s odds of experiencing an ALTE. Infant preference for hand-to-chin posture was not an independent risk factor in multivariate analysis. Our results must be considered within the context of a large number of clinical conditions that can manifest as an ALTE. These can include, but are not limited to, sleep apnea, apnea of prematurity, gastroesophageal reflux, cardiac arrhythmias, sepsis, and seizures as well as neurological, metabolic, or respiratory problems.7-9 In our study, all patients were diagnosed with an ALTE by their attending physician after consideration of their history, physical exam, and diagnostic workup.10 Our patients who had an ALTE underwent a wide variety of diagnostic tests (Table III). The goal of this investigation was to support the clinical observation that smaller mandibular size and hand-to-chin posture are linked to ALTE in at least some of the affected infants. Clearly, an infant with an ALTE attributed to a medical diagnosis such as gastroesophageal reflux may have, in addition, a smaller mandible. It is certainly plausible that mandibular hypoplasia may contribute to momentary or intermittent obstruction of the upper airway in some ALTE cases. At birth, the infant’s jaw is almost horizontal and flexible, and it can be easily 502
Horn et al
shifted backward.11 An obstructive event, especially with cyanosis and change of consciousness, would frighten a parent and result in an admission to the hospital for an ALTE workup. Although ALTE and obstructive sleep apnea (OSA) are not the same, there can certainly be some overlap in the underlying physiology that triggers both. Mandibular hypoplasia is a well-documented etiology of OSA in some patients. For example, in adults, the length of the mandibular plane and posterior airspace are inversely related to symptom levels in patients with OSA.9 Similarly, children with OSA show the same skeletal and soft tissue configurations as adults with sleep apnea.12 Recently, Arens et al13 used magnetic resonance imaging as a means of identifying anatomical factors consistent with OSA in school-aged patients (3-7 years of age). Their data suggest a positive correlation between increased adenoid and tonsil volume and an apneic index. A number of anatomical abnormalities developed in infancy may contribute to sleep disordered breathing in otherwise normal children 3 to 13 years of age with nasal obstruction. In addition, a large tongue can partially occlude the airway.11,14 Tonkin et al noted that an infant’s airway is vulnerable to unbalanced external forces that tend to narrow the airway.11 For example, Stark and Thach showed that mandibular pressure from a face mask can cause backward displacement of the jaw leading to airway obstruction.15 Muscle tone, neuromuscular control, and soft-tissue function are critical factors that control the airway during sleep.16 Finkelstein et al17 found that infants as young as 3 weeks of age have OSA with nasal resistance that, in combination, may be related to later abnormal facial development. One of their findings in these infants was shortened mandibular length. Furthermore, sudden infant death syndrome (SIDS) and ALTE are defined by many as distinct entities, but the relationship between the two cannot be ignored. There is evidence that some victims of SIDS had had ALTEs before the terminal event. Although the relationship between ALTEs and SIDS has not been systematically defined, there is a positive history of an ALTE in 5% to 10% of SIDS victims.2,3 Tonkin18 hypothesized that the cause of SIDS was a vulnerable airway at the oropharyngeal level between the soft palate and the base of the skull. Airway occlusion at this level might occur during muscle relaxation (which occurs The Journal of Pediatrics • October 2006
Table III. Clinical workup of ALTE Screening exam Electrocardiogram Chest x-ray Hematocrit Lumbar puncture Electroencephalogram pH probe Sleep study Head CT Scan Barium swallow Echocardiogram Upper GI series Head MRI Pneumogram Spinal ultrasound Bronchoscopy Formal consults
Frequency
Patient ID number
23 18 15 7 7 6 5 5 2 2 2 2 1 1 1 18
1,3,4,5,7,8,9,10,11,12,13,18,19,21,22,23,25,27,29,30,31,32,33 1,3,4,5,8,9,11,12,13,18,19,20,22,27,29,30,32,33 1,3,4,5,8,10,11,12,13,18,19,20,27,29,30 1,3,12,13,18,19,20 1,4,5,12,31,32,33 4,5,7,12,19,20 4,5,7,12,20 1,5,10,31,33 3,4 5,18 18,19 4,5 19 22 4 Neurology: 1, 4 Pulmonary: 4 Gastroenterology: 1,3,4,5,7,19,29 Otolaryngology: 1,4 Lactation specialist: 1,2,20,22 Cardiology: 18,19
Abbreviations: CT, computed tomography; GI, gastrointestinal; MRI, Magnetic resonance imaging.
during rapid eye movement [REM] sleep) facilitated by a hypermobile mandible or an enlarged tongue with a strong backward sucking action developed by artificial feeding of the infant. A recent study showed that upper airway size changes in supine position during magnetic resonance imaging and during tidal breathing in children with obstructive sleep apnea.19,20 Byard and Kennedy21 noted that SIDS may occur in infants with mandibular hypoplasia without a history of oxygen desaturation. They recommended that jaw size be systematically measured in all postmortem exams of SIDS infants. In addition, the observation of a small mandible in some victims of SIDS must be considered seriously. At this point in our understanding of SIDS and ALTEs, both appear to have many possible causes. The hand-to-chin positioning results were interesting. Parents were questioned about their infants’ hand position during sleep. Parents were quite specific as to whether the infant spent the majority of sleep time with the hand positioned under the chin or whether the hand and arms were in an open posture away from the chin. The parents’ observations are the basis for the hand-to-chin positioning finding. The differences between the ALTE and control groups in observed posture indicate the potential importance of this observation. One possible explanation for the posturing is a trial-and-error phenomenon that results in clinical relief of symptoms. All this can take place even in an infant and below the level of conscious decision. One of the authors (NFF) has treated a 4-year-old-boy with obstructive sleep apnea who slept in a sitting position on the floor with his chin supported
against the wall. This was most likely not the result of an understanding of the upper airway anatomy but the discovery of a position that provided relief. Although a preference for the hand-to-chin posture is associated with higher odds of ALTE, this positioning may not create an obstruction but rather may serve as a marker for infants with airway obstruction and illustrate how they compensate. In this study, parents were not asked about handchin location during the event but were asked about the infant’s usual preference. The infant may not have been able to assume this posture at that time of the ALTE, and therefore had an ALTE because of other cumulative factors. A limitation of this study is its case-control design. Because of the design, incidence rates of ALTE for a given combination of risk factors could not be estimated. Group assignment was not blinded, and thus facial measurements, though confirmed by a second observer, could have been subject to observer bias. After enrollment and measurement, neither group was followed for subsequent problems. Patients with ALTE may have developed an exclusion criterion, and control patients may have presented with an ALTE. A large, prospective cohort study through infancy would be necessary to address these issues. In summary, smaller mandibular size was associated with ALTE, suggesting airway obstruction as a potential cause of ALTE. The association of this characteristic with ALTE offers the potential for prospective quantification of ALTE risk. It is our hope that an assessment of this salient aspect of an infant’s appearance be evaluated in a well-baby exam.7,8,18
Smaller Mandibular Size In Infants With A History Of An Apparent Life-Threatening Event
503
We thank David Wypij, PhD, for statistical consultation, and Susan Longobucco-Hynes, Linda Dempsey, Julia Mancuso Perkins, Rosella Mancuso, Kathy Nygren Austin, and Herminia Shermont for their collegial support.
REFERENCES 1. National Institutes of Health Consensus Development Conference on Infantile Apnea and Home Monitoring, Sept 29 to Oct 1, 1986. Pediatrics 1987;79:292-9. 2. Beal S. Apparent life threatening events with serious sequelae in infants and young children. J Paediatr Child Health 1992;28:151-5. 3. Hunt CE. Sudden infant death syndrome and other causes of infant mortality: diagnosis, mechanisms, and risk for recurrence in siblings. Am J Respir Crit Care Med 2001;164:346-57. 4. Farkas LG. Athropometry of the Head and Face. 2nd ed. New York: Raven Press; 1994. 5. Hirji F, Mehta C, Patel N. Exact inference for matched case-control studies. Biometrics 1988;44:803-14. 6. SAS Institute. SAS 9.1.3 Help and Documentation. Cary, NC: SAS Institute Inc.; 2002-2004. 7. Brooks JG. Apparent life-threatening events and apnea of infancy. Clin Perinatol 1992;19:809-38. 8. Davies F, Gupta R. Apparent life threatening events in infants presenting to an emergency department. Emerg Med J 2002;19:11-6. 9. King B. Pediatric apnea. eMed J 2001;2:1-9. 10. Brand DA, Altman RL, Purtill K, Edwards KS. Yield of diagnostic testing in infants who have had an apparent life-threatening event. Pediatrics 2005;115:885-93. 11. Tonkin SL, Gunn TR, Bennet L, Vogel SA, Gunn AJ. A review of the
504
Horn et al
anatomy of the upper airway in early infancy and its possible relevance to SIDS. Early Hum Dev 2002;66:107-21. 12. See CC, Newman LJ, Berezin S, Glassman MS, Medow MS, Dozor AJ, et al. Gastroesophageal reflux-induced hypoxemia in infants with apparent life-threatening event(s). Am J Dis Child 1989;143:951-4. 13. Arens R, McDonough JM, Costarino AT, Mahboubi S, Tayag-Kier CE, Maislin G, et al. Magnetic resonance imaging of the upper airway structure of children with obstructive sleep apnea syndrome. Am J Respir Crit Care Med 2001;164:698-703. 14. Siebert JR, Haas JE. Enlargement of the tongue in sudden infant death syndrome. Pediatr Pathol 1991;11:813-26. 15. Stark AR, Thach BT. Mechanisms of airway obstruction leading to apnea in newborn infants. J Pediatr 1976;89:982-5. 16. Ferraro N. Craniofacial development and the airway during sleep. In: Loughlim G, Marcus C, eds. Sleep and Breathing in Children a Developmental Approach. New York: Mercel Dekker Inc; 2000:293-309. 17. Finkelstein Y, Wexler D, Horowitz E, Berger G, Nachmani A, Shapiro-Feinberg M, et al. Frontal and lateral cephalometry in patients with sleep-disordered breathing. Laryngoscope 2001;111(4 Pt 1):634-41. 18. Tonkin S. Sudden infant death syndrome: hypothesis of causation. Pediatrics 1975;55:650-61. 19. Arens R, Sin S, McDonough JM, Palmer JM, Dominguez T, Meyer H, et al. Changes in upper airway size during tidal breathing in children with obstructive sleep apnea syndrome. Am J Respir Crit Care Med 2005;171:1298-304. 20. Tsuiki S, Almeida FR, Bhalla PS, Lowe AA, Fleetham JA. Supinedependent changes in upper airway size in awake obstructive sleep apnea patients. Sleep Breath 2003;7:43-50. 21. Byard RW, Kennedy JD. Diagnostic difficulties in cases of sudden death in infants with mandibular hypoplasia. Am J Forensic Med Pathol 1996;17:255-9.
The Journal of Pediatrics • October 2006