Can anything be worse for a parent than going to wake a sleeping baby and finding the child dead? The number of deaths from SIDS, or sudden infant death syndrome, has dropped by more than 50 percent since 1992, when the American Academy of Pediatrics first said that all babies should be put to sleep on their back because sleeping on the stomach appeared to increase the risk of SIDS.

Despite that progress, SIDS remains the leading cause of death in infants between 1 month and 1 year old, annually killing more than 2,000 babies in this country. In an effort to reduce the toll, the academy this week released new and far stronger recommendations. For the first time, the AAP takes on advocates of "co-sleeping," stating unequivocally that babies shouldn't sleep in adults' beds.

Advocates of co-sleeping say it encourages breastfeeding and infant-parent bonding.

"There are over 10 case-controlled studies that show an increased risk," says John Kattwinkel, a professor of neonatology at the University of Virginia and head of the academy's SIDS task force. "We've tried very carefully not to make statements and speculations that aren't borne out by data."

The pediatricians, well aware that the co-sleeping contingent approaches the subject with near-religious fervor, point to studies showing that "room sharing"—having the baby sleep in a cradle or crib in the parents' room—reduces SIDS risk and propose that as a safe alternative to co-sleeping.

"We hope we've put that in a positive light," Kattwinkel says.

What makes the issue hard to resolve is that the cause of SIDS is still unknown. In some cases, autopsy reveals that the child suffocated in soft bedding or wedged between a bed and a wall. But in other cases, there's no evidence of suffocation. The leading theory is that in some babies, the arousal centers in the brain stem aren't fully developed, so these infants don't awaken and move when they're in a position that forces them to rebreathe their own oxygen-deficient exhaled air. Left unable to say why SIDS happens, the pediatricians have to rely on population studies that show the relative risk of different situations.

The academy's key recommendations:

• Babies should sleep on their back only. New data show that side sleeping, once also thought to be acceptable, is not as safe as back sleeping.
• Babies should sleep close to their parents but not in the same bed. New data show that SIDS risk is reduced when the child sleeps in the same room as the mother. Thus, the academy recommends that babies sleep in a crib or bassinet in the parents' room for the first six months.
• Using a pacifier at naptime and bedtime reduces SIDS risk, according to six studies. Researchers speculate that the pacifier may help open the infant's airway or possibly encourages an arousal reflex. Although both doctors and parents worry that pacifiers can damage a child's teeth, dentists say that using a pacifier in the first year of life doesn't cause problems. Breastfeeding advocates also worry that pacifiers may make it more difficult to establish breastfeeding. In order to minimize that, the pediatricians suggest not introducing a pacifier until the baby is 1 month old.
• The baby should sleep on a firm mattress. Pillows, quilts, stuffed toys, sheepskins, and loose bedding such as blankets increase SIDS risk and should be avoided. The pediatricians recommend using sleep sacks to keep babies warm without the possible hazard of covering the head.
• Don't use sleep positioners designed to prop babies up on their side. Although these and other devices (such as mattresses with tiny fans inside) are heavily marketed to new parents as anti-SIDS weapons, there is no evidence that they are effective or safe. Electronic breathing monitors also don't decrease SIDS risk and should be avoided.
• Babies who sleep on a sofa or chair with an adult are at an especially high risk of SIDS.
• Secondary caregivers, including grandparents and day-care providers, need to be educated on the importance of putting babies to sleep on their back.

That last point is particularly important—many mothers go back to work when their babies are 2 or 3 months old, which is just when the SIDS risk peaks. A study published along with the academy's recommendations in the October 10 issue of Pediatrics points out that 20 percent of SIDS deaths in the 1990s happened in child-care settings, far higher than the 8 percent that would be expected statistically. SIDS risk goes up as much as 18 times when a baby not used to stomach sleeping is placed that way. The researchers speculate that a caregiver might discount a new parent's cautions and lay a baby down on her stomach because "it's always been done that way."

"That way" is clearly wrong.