But the study also found that erythropoietin is just part of the molecular conspiracy that destroys vision in the diabetic eye. The researchers detected extremely elevated levels of vascular endothelial growth factor (VEGF), another molecule that has long been identified as a contributor to diabetic retinopathy by stimulating blood vessel growth.

VEGF will occur in 60 percent of persons with diabetes unless something is done to prevent it, said an accompanying editorial in the journal by Dr. Lloyd Paul Aiello, who heads the section on eye research at the Joslin Diabetes Center in Boston.

Some 700,000 Americans have diabetic retinopathy, with 63,000 new cases and 5,000 cases of diabetes-caused blindness every year, he wrote.

"This is the first time folks have shown with extensive human data that erythropoietin is involved in diabetic retinopathy," Aiello said. "Its action is independent of that of VEGF."

Attention has been focused on VEGF. Several drugs designed to block its action are in advanced testing, and one has been approved by the U.S. Food and Drug Administration for use against a different cause of blindness, age-related macular degeneration, he said.

"But preclinical data suggests that VEGF might not be totally responsible, and this work now points out that another molecule is independently involved," Aiello said.

The finding opens a new area of research, he said, adding that "a good deal of work needs to be done to see if inhibiting erythropoietin can have a beneficial effect."

One issue is the safety of blocking erythropoietin in persons with diabetes, since there are studies indicating that it also acts to protect the retina from damage during times of stress, he said.

"This is groundbreaking work," said Dr. John Loewenstein, associate chief of ophthalmology at the Massachusetts Eye and Ear Infirmary. "And like all groundbreaking work, it has to be confirmed by others. First it has to be confirmed by more experiments, and then we can look for inhibitors of erythropoietin."