PHILADELPHIA -- Using a specially designed, highly lifelike doll, researchers at the University of Pennsylvania have determined that rapid head rotations sustained when a baby's head contacts a hard surface during household falls may result in diffuse brain injuries.  The findings call into question earlier assessments of the seriousness of such falls by young infants, previously viewed by some as unlikely to cause widespread brain injury.

The results appear in the July issue of the Journal of Neurosurgery.

"Previously falls were considered relatively benign, because the head was assumed to move in a linear path at the terminus of a fall," said Susan S. Margulies, associate professor of bioengineering at Penn.  "Linear motions are most frequently associated with skull fractures and focal brain injuries, but it is primarily rotational movements that produce more severe diffuse brain injuries.  We found that when the head contacted a firm surface before the body, significant rotational motions were produced."

The Penn investigators found that rotational deceleration -- rapid changes in velocity as the head contacts a hard surface and then violently rebounds -- increased with higher falls and harder surfaces.  The largest rotational decelerations, however, were measured when volunteers intentionally struck the doll's head against a hard surface.  These inflicted impacts resulted in decelerations dramatically higher than those from even a five-foot fall onto concrete.

The findings by Margulies and her colleagues may help abuse investigators differentiate accidental falls from injuries caused by the striking of a child's head against a surface.  Brain injuries -- accidental and inflicted -- hospitalize or kill an estimated 150,000 children annually in the U.S.

"Traumatic brain injury is the most common cause of death in childhood, and child abuse is believed to be responsible for at least half of infant brain injuries," Margulies said.  "While accidental falls are a frequent cause of pediatric trauma, they are also a common explanation given by caretakers in suspected abuse cases."

Margulies has been using anthropomorphic dolls to study infant head injuries since 1987.  The sophisticated doll used in this experiment was designed to mimic the median body weight, weight distribution and size of a one-and-a-half-month-old infant.  The dummy's neck was hinged to replicate the compliant neck of a young infant, and its skull and scalp were made of materials closely approximating the properties of a young infant.

A sensor on the doll's head measured changes in rotational velocity and acceleration.  Such motions are known to cause a diffuse pattern of strains and injuries throughout the brain, but no previous experiment has compared the rotational motion of the head during falls and inflicted events.

An apparatus dropped the doll 134 times from heights of one, three and five feet onto common household surfaces: a concrete floor, quarter-inch-thick carpet padding and a four-inch-thick foam pad, similar to a crib mattress.  In additional tests, volunteers also shook the doll vigorously and then struck its head against one of the same three surfaces.  

"We found that vigorous shaking of this infant model had effects similar to one-foot falls and falls onto foam, but inflicted impacts of the head onto hard surfaces produced significantly greater rotational decelerations and changes in velocity than those onto foam, vigorous shakes and even a five-foot fall onto concrete," Margulies said.  "Separate studies have shown that larger rotational decelerations lead to more severe brain injuries.  Based on this evidence, our data suggest that inflicted impacts are much more likely than falls or shaking to lead to brain injury."

Comparing the results with published data from animals and children, the Margulies group concluded that it was highly unlikely that vigorous shaking or falls onto a foam mattress from distances up to five feet would result in severe or fatal brain injuries; however, five-foot falls onto concrete appeared capable of causing serious brain injury.  The researchers also concluded that inflicted impacts with a hard surface would likely produce subdural hemorrhage and possibly diffuse axonal injury.  Results were inconclusive on the effect of intermediate-height falls onto concrete or carpet padding.  

Margulies cautions that extensions of the findings to estimate likelihood of injury are tentative because little is known about how the infant or toddler brain and skull responds to rapid rotational motions and impacts.

"Children are not just miniature adults," Margulies said.  "Learning more about pediatric brain injuries will help us develop protective devices -- helmets, playground surfaces, car seats -- that better meet their specific needs."

Margulies was joined in this research by Michael T. Prange and Brittany Coats of Penn's Department of Bioengineering and Ann-Christine Duhaime of Hitchcock Medical Center in Hanover, N.H.  The research was funded by the National Institutes of Health and the Centers for Disease Control and Prevention.