Oct. 7, 2004 -- Seven years ago, the trial of Louise Woodward brought to national attention a terrible kind of child abuse called "shaken baby syndrome."
Woodward was the teenage au pair convicted of second-degree murder in 1997 after authorities in Massachusetts accused her of shaking to death 8-month-old Matthew Eappen.
She was sentenced to mandatory life in prison, but a judge reduced the conviction to manslaughter and sentenced her to the nine months she had already served in prison awaiting trial.
Woodward returned to her native Britain, still protesting her innocence. But where does science stand on SBS?
Hemorrhages and Swelling
It seems like common sense: the widely held notion that a baby's neck is so weak — its brain so soft — that vigorous shaking could kill the child.
Public service announcements warn: "Shaking a baby can result in brain damage, blindness and even death. Never ever shake a baby."
The American Academy of Pediatrics cites estimates that as many as 350 infants die every year from shaken baby syndrome.
Dr. Randell Alexander, a professor of pediatrics at the University of Florida and a highly regarded expert for the prosecution in shaken baby cases, says shaking for just a few seconds is enough to cause bilateral retinal hemorrhages and swelling of the brain.
"And it can even kill," he said.
Little wonder that just two weeks ago New York Gov. George Pataki issued an initiative to educate parents about the dangers of shaking a baby.
A Study Debated
The science appears to be pretty well settled — but it is not. Dr. Jan Leestma — a leading neuropathologist who has testified for defendants in shaken baby cases, including Woodward — says there's "not much" science supporting SBS.
"The theory that is popular is that in the course of this flexing back and forth of the head, that forces are operating in the back of the eye [that] could tear blood vessels and so forth," he said.
"The problem with that is, the amount of force that can be generated in shaking isn't enough to do that."
"[There have been] experiments done with models like this with sensors in there and they had some football players shaking the baby as hard as they could," he said.
The experiment Leestma refers to was done at the University of Pennsylvania. Many consider it the best published on the subject to date.
According to Leestma, it shows that the forces generated from shaking are actually far less than what is thought to be needed to cause fatal injury.
However, Alexander called it a "limited study," looking at the equivalent of a single shake.
"It's the repetitive shakes that do this," Alexander said. "In the same way that buildings in earthquakes get knocked down, not with just the first shake, but with multiple ones all coming on top of each other."
SBS is usually diagnosed by a number of tell-tale signs: most importantly bleeding in the brain and behind the eyes — but there is controversy about that, too.
Alexander says there is no doubt in his mind that shaking a baby hard enough can cause retinal hemorrhages.
However, he added, "it's not always going to do that. I mean, we have some children that die from shaken baby syndrome and don't have retinal hemorrhages. But the vast majority do."
He said, "what's so unusual about that is that there's almost nothing that gives retinal hemorrhages in childhood in those early days."
Leestma disagreed. There could be medical conditions. "The child could have an infection in the brain," he said. "The child may have fallen from the changing table."
No one suggests that shaking a baby is good, but the science is extremely complicated.
Leestma says when individual cases come before doctors, police and prosecutors, too often there's a rush to judgment.
He said he has "no doubt" that there have been defendants prosecuted for shaking their babies to death who in fact had not harmed their children.
This summer, Ken Marsh was freed after serving 21 years in a California prison. He was convicted of killing his girlfriend's baby. But new medical evidence showed an accidental fall from a couch could have caused the child's death.
In 1999, Brandy Briggs says she was pressured to plead guilty to shaking her baby to death and was sentenced to 17 years in prison. But after reviewing the medical evidence in her case, a new medical examiner in Houston changed the manner of death from homicide to "undetermined," and Briggs is seeking to have her conviction overturned. Prosecutors are confident in their evidence against Briggs and just this week a judge agreed with them. Her case is now before the court of appeals.
And just before Christmas in 2002, Christina Diaz was accused of shaking to death a baby at her day-care center. After two trials, Diaz was finally acquitted, but not before she lost her business, her home and reputation. Now she's back in court to get the charges wiped off her record. Many friends and even a juror showed up to support her.
"I love doing day care. I love working with children," she told ABC News. "It was my life. I would never think of hurting any child."