Thu Oct 9 2003
Experiments Cast Doubt on Shaken Baby Syndrome
Experimental results published in July suggest that shaking a baby cannot cause serious brain injury unless it includes head impact on a hard surface. The series of laboratory experiments was part of scientific efforts to distinguish accidental from assaultive cases of serious infant brain injury-- injury characterized by subdural hematoma, traumatic axonal injury, and retinal hemorrhage.
The published report, "Anthropomorphic Simulations of Falls, Shakes, and Inflicted Impacts in Infants,” was written by researchers Michael T. Prange, Brittany Coats, Ann-Christine Duhaime, and Susan S. Margulies. It appears in the Journal of Neurosurgery, Volume 99, pages 143-150, July 2003.
In some cases of serious infant brain injury, the skin of the baby’s head bears no signs of impact. Many physicians believe “shaken baby syndrome” explains these cases. The brain injury, they conclude, has come entirely from an adult’s manually shaking the baby. Impact is not necessary, they have argued, to produce serious, even fatal, infant brain injury.
In a 1987 experiment, though, adults could not shake a doll hard enough to generate the forces necessary to cause serious brain injury. To this some physicians have responded that a shake which included striking an infant’s head against a soft object, such as a mattress, could generate the necessary forces without leaving evidence of impact.
The recent series of experiments elaborated on the 1987 experiment. It used a doll more meticulously crafted to approximate a six-week old infant. It also discarded the 1987 experiment’s assumption that a shaken baby’s head would have a fixed center of rotation.
In the recent experiments, six adults (four men and two women, weighing between 110 and 220 pounds) shook the doll repeatedly. Almost all the shaking episodes included striking the doll’s head against a stone bench top, a carpet, or unencased foam from a crib mattress.
Results of the recent experiments, like those of the 1987 experiment, suggest that a human cannot shake an infant forcefully enough to cause serious brain injury. They also cast doubt on the theory that a human can generate the necessary forces by striking an infant’s head against a surface too soft to leave evidence of impact. The study report concluded, “There are no data showing that the [maximum changes in angular velocity and peak angular accelerations] of the head experienced during shaking and inflicted impact against unencased foam [are] sufficient to cause subdural hematomas or primary traumatic axonal injuries in an infant.”
The researchers do not regard their results as definitive because of limitations in experimental design. They could not produce a perfectly “biofidelic” model of an infant’s head. And their foam was unencased, unlike foam in mattresses and padded furniture.
The recent experiments also expanded on the 1987 experiment by dropping the doll from various heights onto surfaces of differing hardness. The purpose of this was to contribute to the investigation of cases in which impact is clear, but the suspected abuser reports that impact occurred in an accidental fall of a few feet, such as from a bed.
Copyright © 2003 David S. Marshall
