Adding nitric oxide to the air administered to premature infants may not boost survival odds or reduce lung injuries for all of them but appears to help some babies in the short run and enhances the mental development of even more over time, new studies show.

However, one of the studies indicates nitric oxide may increase the risk of bleeding in the brain for the smallest preemies.

Nitric oxide works by dilating blood vessels in lungs to allow for better delivery of oxygen and removal of carbon dioxide.

Two groups of researchers — one from Stanford University, the other from the University of Chicago — focused on problems faced by preemies born with lungs too small to function properly on their own. About 60,000 babies are born in the United States each year with underdeveloped lungs that require mechanical help.

Their studies were published Thursday in The New England Journal of Medicine.

The Stanford team led a national study that looked at whether the compound could help the very sickest and smallest premature infants suffering from respiratory failure.

All 420 infants in the study were born before full term and weighed between 14 ounces and 3.3 pounds. Half of them were randomly assigned to receive either nitric oxide or a placebo, with neither physicians nor the family knowing which option the infant was getting.

The researchers found that nitric oxide was no better at improving survival rates or preventing long-term lung damage than the placebo. But when they broke the group down further by weight, they saw a sharp difference.

Infants who weighed 2.2 pounds or less who got nitric oxide had a significantly higher rate of bleeding in the brain and were more likely to die than those who got the placebo — 43 percent vs. 33 percent for hemorrhage and 62 percent vs. 48 percent for mortality.

Although the compound's effects aid blood vessels in the lungs and help send blood to the most oxygenated areas of the lung, its relaxing effect on such vessels also elevates an already significant risk for brain hemorrhage. That is why physicians have been cautious about using it, despite favorable reports for preemies over the past two years.

But for those weighing more than 2.2 pounds, the treatment decreased the likelihood of death or long-term lung damage to 50 percent, compared to 69 percent for the placebo group.

"This study suggests that the key for effective use of inhaled nitric oxide may lie in choosing the right patients," said Dr. Krisa Van Meurs, lead author of the Stanford study.