Shaken Baby Syndrome
by Rebecca Crosier

Shaken Baby Syndrome (SBS) is a serious form of abusive traumatic brain injury, inflicted on an infant through violent shaking causing rotational head acceleration (see the video below for a detailed depiction of the SBS mechanism). Death occurs in 10 to 40% of patients with SBS and survivors frequently experience chronic neurological problems including behavioural and cognitive disturbances, blindness and epilepsy[1] . The concept of SBS was extrapolated from an experiment by A.K. Ommaya in 1968, which looked at the effects of high velocity whiplash on rhesus monkeys[2] . It was hypothesized that shaking could cause similar intracranial injury in a child[3] . Since then, medical professionals have diagnosed SBS from two principal symptom findings: subdural hematoma (SDH) and retinal haemorrhage (RH). However, recent biomechanical and animal studies have cast doubt on their pathophysiologic connection [4] [5] [6] . Moreover, researchers are discovering alternative causes for both SDH and RH[5][7] . For example, SDH can occur during the birthing process in up to 46% of births[5] . This new evidence has serious medical and legal implications.

Visualizing how shaking damages the brain

1.1The Development of SBS as a Diagnosis

Rotational acceleration of the head caused by shaking

The concept of SBS was developed by C. Caffrey based on an experiment by A.K. Ommaya in 1968, which looked at the effects of high velocity whiplash on rhesus monkeys[2]. The monkeys were strapped to a chair with unrestricted head movement. The chairs were hit forcefully from behind producing significant (600 times the acceleration due to gravity) rotational acceleration to the head. Of the 19 monkeys subjected to autopsy, 79% showed evidence of subdural hematoma (SDH), and 42% had severe neck injury. Without further independent investigation, Caffrey proposed that human shaking could cause intracranial injury similar to Ommaya’s monkeys. He further theorized that shaking would damage the retinal capillaries which would account for the retinal haemorrhage (RH) often present with SDH[3]. He proposed the diagnosis “Whiplash Shaken Infant Syndrome” and later “Shaken Baby Syndrome” to describe infant patients with SDH and RH. Evidence to support the classification of SBS came from witnessing shaking events and well documented confessions and convictions[8] .

1.2 Biomechanical and Animal Studies

Biomechanical studies were the first type of study used to investigate SBS. Although their goal was to provide evidence to back up the SBS hypothesis, immediate doubt began to arise concerning the pathophysiologic connection between shaken baby syndrome and its two cardinal manifestations, SDH and RH. The first biomechanical study used dummies to mimic a 6-month old child[8]. College students shook the dummies and generated accelerations well below the threshold for intracranial damage. Above threshold levels of acceleration were achieved only through impact. Furthermore, if shaking could generate adequate acceleration, the infant's neck would be destroyed- neck damage is rarely seen in SBS cases[4]. This research stimulated further studies to investigate whether shaking alone was sufficient to cause RH and SDH. For example, Prange et al. used a 6-week old dummy model and found that shaking and a small fall onto a soft surface was sufficient to produce accelerations that theoretically could cause SDH[9] . To date, biomechanical studies often suggest that shaking alone cannot cause the classic symptoms of SBS, yet many medical professionals still believe it can[4]. Researchers are turning to finite element (FE) modelling for a more detailed interaction of the brain, skull and cerebral spinal fluid[10] . The results obtained using the new FE model would be more accurate because the paediatric features of the skull and skull-brain interface are more accurately represented, than in the classic dummy model used by previous investigators. FE modelling has recently begun being used in studies and results have found that shaking will lead to a specific pattern of brain matter motion, increased axonal injury at contact locations and deep brain structures, but have shown little support for SDH and no support for RH[11] . Taken together, these studies imply that shaking an infant will not cause SDH and RH 100% of the time. Moreover, SDH and RH alone should not be used to diagnose SBS. These results have legal implications which will be discussed in detail below.
The biomechanical studies discussed above investigated whether or not shaking or various forms of impact could produce enough acceleration to cause intracranial head injury. They were not able to determine the specific effects of shaking. The FE modelling experiment[11]was able to determine the general location of axonal damage, but neither SDH nor RH were physically demonstrated to arise from shaking or impact. An animal model was needed to develop a clearer relationship between shaking and the proposed pathology of SBS. Animal studies have only recently begun being used to investigate the pathogenesis of SBS. Finnie, Manavis and Blumbergs subjected anesthetised 7-10 day old lambs to vigorous shaking 10 times for 30 seconds over a 30 minute period[1]. The lambs were sacrificed and their brains dissected. The results found widespread neuronal perikaryal amyloid precursor protein expression likely a non-specific response to head trauma. Axonal injury,SDH and RH were, in contrast, all minimal. This means that although both SDH and RH are used to diagnose SBS, they are not specific for trauma.

2.1 Symptoms of SBS:

SBS is diagnosed principally based on the presence of subdural hematoma, which is a collection of blood between the brain and the skull (in the subdural space) and retinal haemorrhage- abnormal bleeding of the blood vessels in the retina. Encephalopothy is often present as well, but it is not required for a SBS diagnosis. Shaking is proposed to cause SDH by tearing the large bridging veins located between the dura and the brain. These veins move blood from the brain to the dural sinuses[5] . The major mechanism by which RH is produced in cases of SBS is vitreoretinal traction[12] . The hemorrhages are often numerous and multilayered and there is a correlation between the severity of the trauma and the severity of the hemorrhage[12].

Intact bridging veins

2.2 SDH mechanism in SBS:

SDH has been integral to the diagnosis SBS. In babies, SDH occurs typically as a thin bilateral film over two areas: the cerebral convexities and the posterior interhemispheric fissure[13] . Shaking is proposed to tear the large bridging veins which then cause SDH. However, there is a lack of litterature describing ruptured bridging veins in infants with SDH[13] . There are also physiological and anatomical and anatomical objections to the bridging vein hypothesis. Bridging veins carry a lot of blood; 260mL of blood is carried into the dural sinuses per minute[5]. If a bridging vein were to rupture, it would cause a massive clot, which is not observed in SBS patients. Furthermore, should the veins leak under severe tension, radiological observations indicate that the bleeding would be predominantly in the subarachnoid space as opposed to the subdrual space[13] . The lack of evidence supporting tearing of the bridging veins, as well as the physiological and anatomical shortcomings of the hypothesis, has prompted scientists to hypothesize alternate causes of SDH not caused by shaking.
Intradural bleeding is commonly found in fetal and neonatal autopsy[13] . The infantile dura is not rigid like in the adult and many of its features (ex. arachnoid granulations) are not fully formed. Careful observational analysis by Cohen et al.[14] has found that the sites most susceptible to bleeding in the fetus and the neonates by natural causes are also the sites of posterior interhemispheric haemorrhage- one of the two areas typically associated with SDH in shaken babies . Interhemispheric haemorrhage is often intradural, as opposed to subdural, and it is impossible for CT and MRI to distinguish between natural intradural bleeding experienced by some babies and the thin film subdural bleeding of shaken babies. This means that the anatomy of the infant dura questions the validity of the belief that posterior interhemispheric haemorrhage is evidence of bridging vein rupture and that it is caused by shaking. Spontaneous intradural bleeds are likely mistaken as SDH, which may lead to abuse speculations.
A case study by Guddat et al.[15] suggested a second differential diagnosis to SBS when their patient presented with SDH . The patient presented with cerebral edema, RH and SDH. The patient died and autopsy revealed that the bridging veins were intact. Histological analysis showed that severe neonatal giant cell hepatitis induced coagulopathy which caused spontaneous subdural bleeding. Neonatal giant cell hepatitis is a frequent reaction to infant injury and may be a cause of SDH and sudden death in infancy.
Ruptured intracranical vascular malformation, bleeding from a healing subdural membrane, hypoxia[14] and birth related subdural bleeds are all other possible causes of SDH[13] . However, when a baby is presented in a clinic or hospital with SDH, it is difficult to establish what its cause was unless the baby dies and an autopsy can be performed. This makes diagnosis SBS an extremely controversial and at times, inaccurate. It is for this reason that caution should be taken when using SDH in diagnosis.

A muItilayered retinal haemorrhage

2.3 Retinal Haemorrhage:

RH is found in about 85% of SBS cases[12] . Hemorrhages can be unilateral or bilateral, retinal or vitreous[13] .RH is hypothesised to arise from shaking which moves the lens back and forth, transferring tractional forces through the retina, which tears the retina from its connections, thereby disrupting the integrity of the blood vessels in the eye causing hemorrhages[16] . However, Squier[13]suggested that it is biomechanically improbable that the level of force generated by shaking would damage the eye directly. A more likely hypothesis is that increased intracranial pressure (ICP) is transferred to the optic nerve sheath causing RH[5]. It has been well documented that as intracranial pressure increases, intraocular pressure increases which causes hypoxemia in the retinal capillary thus causing RH[17] . It is possible that shaking can increase the infant’s intracranial pressure, however, increased ICP is not specific to shaking. In fact, intracranial pressure can increase due to infection, hypoxicischemic encephalopathy or post-event edema from any trauma including birth[5]. This means that if shaking can indeed cause RH, it is a nonspecific manifestation. Its frequency in SBS cases can be explained by circular reasoning: the incidence is high because it is a cardinal manifestation and its presence lead to the diagnosis[13] . Emerson[18] documented the invalidity of using any RH pattern as diagnostic of abuse. He autopsied 118 infants where abuse was suspected and found many combinations of eye pathology in significant numbers and many combinations. His findings did not fit any particular pattern and frequencies were similar across patients with blunt injury to the head and blunt injury to the body. Other studies have further solidified the speculation that RH can be caused by non-shaking trauma such as small falls[19] . Therefore, since RH is common to many forms of trauma and has other non-traumatic causes, physicians must be cautions when diagnosing SBS using RH as a symptom.

3.1 Medical and Legal Concerns:

Doctors have been diagnosing SBS for 40 years without adequate scientific basis[5]. The concept of SBS was an extrapolation from Ommaya's whiplash study on monkeys[2]and much of the 'supporting' literature exhibits circular reasoning or misinterprets the data. Also, SBS continues to be diagnosed despite the growing body of contradictory evidence. Hundreds to thousands of parents and caretakers have been imprisoned based on testimony that SDH and RH are diagnostic of abuse[5]. In the courts, little regard has been given to the alternatives, including short falls and natural causes. Recently, the Canadian government has called for a review and re-evaluation of all previous SBS convictions[5]. The careful legal and medical analysis has restored the freedom of many individuals who were falsely accused of child abuse.

3.2 Future of the SBS diagnosis:

Since the determination of child abuse is controversial, it is important for diagnosticians to be familiar with the issues and current literature. Dr. Barnes[20] suggests an evidence-based protocol for diagnosticians to follow when child abuse is suspected. Children with suspected SBS must receive protective evaluation along with complete imagery (CT and MRI). Radiologists must be familiar with the limitations in using imaging in diagnosis and should not assume that a baby has been shaken based on the presentation of RH and SDH alone. Imaging findings must be correlated with clinical findings, laboratory testing and forensic examination where possible. The child’s medical history must be thoroughly examined to determine if there are any other possible causes including hypoxia or birth trauma. Should the context of clinical, imaging, biomechanical or pathological findings support abuse, then a child protective investigation may provide the basis for inflicted injury. The physician may choose to believe in SBS or not, however when making a diagnosis, they should search for other possible causes and never assume child abuse when presented with a patient with only SDH and RH.

External Links:
National Center on SBS
Detailed video (CBC) including biomechanical studies and interviews with falsely accused parents of 'SBS' victims

  1. ^ Finnie J. Manavis J. Blumbergs P. (2010) "Diffuse neuronal perikaryal amyloid precursor protein immunoreactivity in an ovine model of non-accidental head injury (the shaken baby syndrome)" Journal of Clinical Neuroscience 17(2):237-240
  2. ^ Ommaya AK. Faas F. Yarnell P. (1968) "Whiplash injury and brain damage: an experimental study" JAMA 204(4):285-289
  3. ^ Caffey J. (1974) "The whiplash shaken infant syndrome: manual shaking by the extremities with whiplash-induced intracranial and intraocular bleedings, linked with residual permanent brain damage and mental retardation" Pediatrics 54:396-403
  4. ^ Bandak FA. (2005) "Shaken baby syndrome: A biomechanics analysis of injury mechanisms" Forensic Science International 151:71-79
  5. ^ Gabaeff SC. (2003) "Challenging the pathophysiologic connection between subdural hematoma and retinal hemorrhage and shaken baby syndrome" West JEmerg Med. 12(2):144-158
  6. ^ Cory CZ, Jones BM. (2003) "Can Shaking alone cause fatal brain injury? A biomechanical assessment of the Dunhaime shaken baby syndrome model" Med Sci Law. 43(4):317-333
  7. ^ Reddy AR, Clarke M, Long VW. (2010) "Unilateral retinal hemorrhages with subarachnoid hemorrhage in a 5-week-old infant: is this nonaccidental injury?" Eur J Ophthalmol. 20(4):799-801.
  8. ^ Squier W. (2008) "Shaken baby syndrome: The quest for evidence" Developmental Medicine and Child Neurology 50(1):10-14
  9. ^ Prange MT, Coats B, Duhaime AC, Margulies SS. (2003) "Anthropomorphic simulations of falls, shakes, and inflicted impacts in infants" J Neurosurg 99: 143-150
  10. ^ Cheng J. Howard I. C. Rennison M. (2010) "Study of an infant brain subjected to periodic motion via a custom experimental apparatus design and finite element modelling" J Biomech 43(15):2887-2896
  11. ^ Couper Z. Albermani F. (2010) "Mechanical response of infant brain to manually inflicted shaking" Proc Inst Mechanical Engeneering 224(1):1-15
  12. ^ Levin A. V. "Retinal hemorrhage in abusive head trauma" Pediatrics 126:961-970
  13. ^ Squier W. (2011) "The 'Shaken Baby' syndrome: Pathology and mechanisms" Acta Neuropathology 122:519-542
  14. ^ Cohen M. C. Whitby E. H. (2010) "Subdural haemorrhage, intradural haemorrhage and hypoxia in the pediatric and perinatal post mortem: are they related? An observations study combining the use of post mortem pathology and magnetic resonance imaging" Forensic SciInt 200:100-107
  15. ^ Guddat S. S. Ehrlich E. Marin H. Tsokos M. (2011) “Fatal spontaneous subdural bleeding due to neonatal giant cell hepatitis: a rare differential diagnosis of shaken baby syndrome” Forensic Sci Med Pathology 7(3):294-297
  16. ^ Clarke M. P. (2009) “Vitroretinal traction is a major factor in causing the haemorrhagic retinopathy of abusive head injury?- No” Eye 23:1761-1763
  17. ^ Tayal V. S. Neulander M. Norton H. J. Foster T. Saunders T. Blaivas M. (2007) “Emergency department somographic measurement of optic nerve sheath diameter to detech findings of increased intracranial pressure in adult head injury patients” AnnEmerg Med 49(4):508-514
  18. ^ Emerson M. V.Jakobs E. Green W. R. (2007) “Ocular autopsy and histopathological features of child abuse” Ophthalmology 114(7):1384-1394
  19. ^ Reddie I.C. Bhardwai G. Dauber S. L. Jacobs M. B. Moran K.T. (2010) “Bilateral retinoschisis in a 2-year-old following a threestorey fall” Eye 24:1426-1427
  20. ^ Barnes P. D. (2011) “Imaging of nonaccidental injury and the mimics: Issues and controversies in the era of evidence-based medicine” Radiological Clinics of North America 49(1):205-229