At Roberto Santos General Hospital in Salvador, Brazil, Dr. Antonio Almeida and a team of specialists are closely following two groups of women: Those who deliver babies with abnormally small heads and those who deliver apparently normal babies.

The hospital is one of three in this city on Brazil’s eastern coast where investigators are studying the most urgent question of the Zika outbreak: Is the virus causing a spike in birth defects, and, if so, how great is the risk?

The answer will help shape the response to the rapid spread of Zika throughout the Americas. Concerns over the potential link to microcephaly have prompted a U.S. alert advising pregnant women against travel to 31 countries and territories with outbreaks.

Officials in El Salvador, Colombia, Ecuador and Jamaica have urged women to delay pregnancies. The many unknowns about Zika's effects and transmission have cast a shadow over plans for the Olympic Games in Rio, set to begin in early August.

The evidence so far is compelling enough that the World Health Organization declared an international health emergency Feb. 1. But it is mostly circumstantial: Suspected cases of microcephaly - a rare birth defect characterized by small head size and an underdeveloped brain - spiked 30-fold in late 2015, months after Zika's arrival in Brazil.

There is harder evidence as well. Brazilian researchers have isolated the virus in brain tissue and amniotic fluid of 17 babies and stillborn fetuses with microcephaly whose mothers had symptoms of Zika infection during pregnancy.

The finding provides scientific plausibility that Zika could cause microcephaly, a condition that can result in developmental delays as well as seizures, hearing loss, vision problems and trouble swallowing. They can range from mild to severe.

"We know the virus can cross the placenta," said Dr. Albert Ko, a tropical disease specialist at Yale University and the Oswaldo Cruz Foundation, a government research institute in Salvador. "It can infect the baby, and it's likely the reason why these fetuses were stillborn or had congenital malformations. That is what we know up to this point."

The study in the Salvador hospitals is one of several aimed at proving or debunking the suspicion that Zika causes microcephaly.

In the northeastern city of Recife, Dr. Ernesto Marques, a researcher from the University of Pittsburgh and Fiocruz Pernambuco, the state arm of the Oswaldo Cruz Foundation, is finalizing plans for a similar study in 12 hospitals.

Separately, the U.S. Centers for Disease Control and Prevention is planning a retrospective study looking at microcephaly cases already reported and comparing them to healthy births.

ACHIEVING CONSENSUS

The virus was discovered in 1947 in Uganda's Zika Forest. For decades, Zika had been associated only with mild symptoms, making it less of a concern than similar mosquito-transmitted viruses, such as dengue and yellow fever.

Brazil is investigating more than 4,000 suspected cases of microcephaly reported since October and, so far, has confirmed more than 400. That is a significant increase from about 150 cases in a typical year.

Experts believe it will take six to nine months to begin to understand whether Zika can cause microcephaly. But it may take longer to reach scientific consensus.

To establish proof that a microbe causes a disease, researchers often follow Koch's postulates, a method developed in the 19th century and used to show the HIV virus was responsible for AIDS in 1984, three years after the first cases emerged in the U.S.

The method requires researchers to establish that the suspected pathogen is strongly associated with disease and that it can be passed from one infected host to an uninfected host and cause disease. The pathogen also must be isolated and grown outside the host, a step that Marques said had been accomplished with Zika.

The epidemiological studies are case-controlled - following and comparing pregnancies that result in healthy babies to those that result in cases of microcephaly. They are looking for evidence that the Zika virus is passed from infected mothers onto fetuses who develop microcephaly.

THE STUDIES

The study in Salvador hospitals is funded by Brazil's Ministry of Health and the U.S. National Institutes of Health. It aims to rule out Zika as a suspect in cases where other causes of microcephaly - including genetic disorders, cytomegalovirus and rubella infections, or exposure to drugs or alcohol during pregnancy - are found.

In apparently normal babies, Ko and his colleagues will look for problems other than microcephaly that could be related to fetal Zika infection, such as vision and hearing impairment, intellectual disability, bone and growth problems, and liver and spleen damage. Such problems can also occur with exposure to other viruses, such as rubella, that cause microcephaly.

"One possibility is that those microcephaly babies are just the severest cases and that they are only a small proportion of all the babies who were infected," Ko said.

Along with full clinical and imaging tests, Ko's team is testing babies and mothers for active Zika infections or antibodies.

Pregnant women are asked about any fever, rash, itching or red eyes to determine possible exposure, said Almeida, hospital director at Roberto Santos General. They receive fetal ultrasounds to look for signs of microcephaly, such as brain calcifications.

The study now includes more than 80 babies with microcephaly, and hundreds of normal births. Within the microcephaly group, some of the babies have unusual lesions in the eyes and ears. Preliminary results could be shared with the scientific community within the next month, Almeida said.

In a recent case, Zika virus was found in the brain, bone marrow and spinal fluid of a stillborn baby who had severe brain lesions, he said. Head measurements suggest the baby also had microcephaly. The mother had no Zika symptoms during pregnancy, suggesting that the virus may cause microcephaly even in asymptomatic women.

"We see the patients, and we are very, very scared," Almeida said. "You can't imagine the suffering of the pregnant women in Brazil."

In Recife, Marques is gearing up a similar study that aims to enroll 200 babies with microcephaly and 400 healthy babies by June or July in 12 hospitals. Marques hopes to have some answers by August.

No one wants answers more than Dr. Manoel Sarno, a fetal medicine specialist on Ko's team. In July, Sarno began to notice a spike in microcephaly cases among patients who recalled having Zika symptoms. Since then, he has identified 80 cases of microcephaly and has wept with many of the mothers.

"I know all the world has a lot of questions," he said. "We do also."

(Reporting by Julie Steenhuysen; Editing by Michele Gershberg and Lisa Girion)