After the Zika virus turned up in Brazil two years ago, hundreds of babies were born with severe brain damage and underdeveloped skulls — a birth defect known as microcephaly.

The reports of microcephaly terrified pregnant women and prompted Brazil to declare a national health emergency.

But researchers in the central Brazilian state of Sao Paulo now say that Zika may be more likely to produce a miscarriage than a baby with a smaller than normal head.

Dr. Benedito Fonseca, a professor of internal medicine at the School of Medicine of Ribeirão Preto in Sao Paulo, says that back in 2015, people worried in part because Zika was a mysterious disease that had never been detected before in Brazil. "In that time every pregnant woman wanted to know what was going to happen with their pregnancy," he says.

Little was known about the virus and the effects it could have on a developing fetus — or even what the relationship was between Zika and microcephaly.

Women who came down with Zika also appeared to be having miscarriages at a higher rate than normal but even that wasn't clear.

So last year Dr. Fonseca set out to track the pregnancies of nearly 1,200 women in his home city of Ribeirão Preto in the state of Sao Paulo.

Over the course of their pregnancies, 178 of the women tested positive for Zika exposure. Nine of the Zika-positive women lost their babies due to miscarriages while only four gave birth to babies with microcephaly.

Fonseca says the severity of the viral infections caused the miscarriages: "The infections were so bad there was no way for the pregnancy to go on. Their own body just terminated the pregnancy due to the Zika infection."

Rubella is another examples of a viral infection during pregnancy that crosses the placenta, infects the fetus and can lead to higher than normal rates of miscarriage.

Because Fonseca had 1,125 pregnant women enrolled in the study, he could see that the rate of miscarriage — nine out of 178 Zika-positive mothers — was far higher than among the women who never tested positive for the virus. There were only two spontaneous abortions among the 947 Zika-negative women. So roughly five percent of the Zika-positive mothers had miscarriages versus fewer than 0.2 percent of the other women. Most of these women entered the study after their first trimester, when most spontaneous abortions occur, which could be part of why the overall miscarriage rate was so low.

The microcephaly findings also surprised the researchers. Among the Zika-positive mothers, there were four babies born with microcephaly. Among the much larger group of women who showed no signs of Zika there were two cases of microcephaly.

"I thought we'd find a lot of microcephaly but I wasn't expecting such a high rate of miscarriages," Dr. Fonseca said just before presenting his findings this week at the American Society of Tropical Medicine and Hygiene's annual convention in Baltimore.

"It's true that babies with microcephaly will have a big impact in terms of long-term care. But we have to also really deal with the short-term care of these women who are having miscarriages," he says. Some may miscarry at home. He warns that these mothers are at increased risk for infections and that Zika could potentially lead to a rise in maternal mortality.

Albert Ko, a researcher from Yale who's been studying the Zika outbreak in Northeast Brazil since it first emerged, says there may have been a rash of Zika-related miscarriages in the spring and early summer of 2015 that went unnoticed.

When Zika first hit in early 2015, the symptoms of the disease were fairly mild. It wasn't until nine months later when babies started being born with abnormally small heads that health officials and the public became concerned. If during that first wave of the outbreak, Zika was causing a spike in miscarriages, that spike would have occurred before the public alarm about the disease had been raised.

"That adverse outcome [miscarriage] of Zika infection, we probably missed it," Ko says. "We probably missed it because we don't have very good surveillance systems especially for women in the public sector health-care system in Brazil." Some women may have miscarried at home and not gone to a doctor. A miscarriage can also look like an induced abortion, which is illegal in Brazil, so that too could have led to an under-reporting of the phenomenon.

"It's a very sensitive issue in Brazil," Ko says.

So in early 2015, health officials simply weren't looking for miscarriages linked to Zika.

As quickly as the Zika outbreak emerged, it now is subsiding.

In May of this year the Brazilian government declared the Zika health emergency to be over.

Fonseca says in Ribeirão Preto, a city of 650,000 people, Zika disappeared as abruptly as it arrived.

"Last year in our city we had over 7,000 cases of Zika," he says. "This year we didn't have any. None."

But he warns that this does not mean that the threat of Zika has evaporated.

Fonseca says it appears that in places where Zika was rampant in 2015 and 2016, people developed immunity to the disease and infections plummeted. But Zika wasn't rampant everywhere across Brazil.

"Let's just take Sao Paulo State," he says. "If you plot the data there are a few cities where Zika hit hard but there are a lot of places where there was almost no Zika whatsoever. So in those places, they're still at risk of having Zika."

Copyright 2017 NPR. To see more, visit http://www.npr.org/.