For weeks, Zika virus has been all over the news. Stark photos of babies with microcephaly, a birth defect in which the infant’s head is abnormally small and its brain often not fully developed, have accompanied stories about the ongoing spread of Zika across more than 20 countries. Several Latin American countries have advised women (many of whom lack access to reliable birth control) to avoid pregnancy for the time being, and the Centers for Disease Control and Prevention has recommended that pregnant women avoid traveling to countries with the Zika virus if they can. These scary bulletins can overshadow an important fact: The Zika virus has not yet been proven to cause microcephaly.
The proposed link between Zika and microcephaly began taking shape in September after some pediatricians in northeast Brazil noticed an unusual number of babies born with microcephaly. This spike in microcephaly cases coincided with an outbreak of Zika, a virus related to the dengue virus and spread by mosquitoes, and when a task force set up a registry to track the problem, it discovered that in 26 of the first 35 cases entered, the mother had developed an illness with a rash potentially indicative of Zika during her pregnancy.
Given the apparent overlap between microcephaly and Zika virus, “The first impression is that there’s a relationship one to the other,” said Denise Cavalcanti, a geneticist at the State University of Campinas in Campinas, Brazil. First impressions can sometimes deceive, though, and she said the evidence linking the two is circumstantial and not yet confirmed. “We should prove this scientifically.”
But confirming a link between Zika and microcephaly will take more research and time — months, if not years — and the confirmation may end up as a preponderance of evidence rather than a definitive link. What’s easy to forget about the oft-cited scientific mantra, correlation is not causation, is that causation is incredibly difficult to prove. When you’re a doctor who’s seeing an alarming number of infants born with birth defects, correlation may be all you have to go on. If you’re a public health official with the job of protecting your community from a scary new threat, you can’t always wait for certainty before you act.1 And that means you must give advice based on the imperfect evidence you have, rather than the definitive data you wish you had.
The evidence available at the moment led experts at the World Health Organization to conclude that “a causal relationship between Zika infection during pregnancy and microcephaly is strongly suspected, though not yet scientifically proven.” Which is why, as my colleague Anna Maria Barry-Jester explained on FiveThirtyEight recently, the WHO’s declaration of a “public health emergency of international concern” referred to the cluster of birth defects and a potentially paralyzing condition called Guillain-Barré syndrome,2 not the Zika virus itself. Sometimes lost amid the worry is the fact that Zika produces relatively minor symptoms, and only in about 20 percent of people infected. And at the moment, there’s no quick and easy test for Zika that can be used outside a research setting.
The evidence connecting Zika and microcephaly remains circumstantial, but it’s growing. “There’s not going to be any single perfect piece of data to link Zika to microcephaly,” said Melinda Moore, a pediatrician at the RAND Corp. who previously studied infectious diseases at the CDC. Instead, researchers must piece together multiple lines of imperfect evidence to come up with a clear picture.
But efforts to construct that clear picture face some major challenges. As Brazilian pediatrician Sandra da Silva Mattos and her colleagues recently wrote in the Bulletin of the World Health Organization, “The first question to be addressed is the real incidence of microcephaly in Northeast Brazil.” Before scientists can confirm that Zika is causing microcephaly, they need to know how common microcephaly was before Zika and how widespread it is now. Right now, the answer to both of those questions remains muddy.
In search of numbers, da Silva Mattos’s group put together a local task force to go back and look for microcephaly in a database of newborns in northeast Brazil born mostly before the current outbreak. They found that the incidence of microcephaly in the area was much higher than expected — from 2 percent to 8 percent, depending on which criteria were used for diagnosis (more on that in a minute). However, when they reviewed only extreme cases, the incidence fell in the range of 0.02 percent to 0.19 percent, which is in line with the incidences reported for microcephaly in other parts of the world. They also found that microcephaly cases varied seasonally, in a pattern that could reflect the peak activity of mosquitoes that carry Zika.
But here’s the puzzle: This fluctuation was seen as far back as late 2012, several years before Zika is known to have appeared in Brazil in mid-2014. In fact, cases of microcephaly seem to have peaked early in 2014, before Zika’s appearance in Brazil, most likely during a sporting event, later that year. This finding could suggest another cause, or it could mean that Zika entered this area earlier than was previously known. It’s important to note that da Silva Mattos’s review is a preliminary report.
What first looked like a startling epidemic of microcephaly cases could turn out to be, at least in part, an example of the “awareness effect” — an uptick in cases that happens when increased awareness of a problem leads to the identification (or, in some cases, misidentification) of cases that otherwise would have gone undetected. Microcephaly was probably underreported before Zika, and Jorge Lopez-Camelo and Ieda Maria Orioli from the Latin American Collaborative Study of Congenital Malformations note in a report they had translated into English for Nature that Brazil’s Live Birth Information System recorded only about a third as many birth defect cases as would be expected based on incidence in other parts of the world. Unless the area had an unusually low baseline rate of birth defects, any new effort to look for a specific defect could easily result in a two-thirds increase in numbers, simply by finding cases that would otherwise have gone uncounted.
Some of the confusion may come down to how microcephaly is classified and counted. A commentary published in the Lancet this month describes how different standards for diagnosing microcephaly can drastically alter how many cases are found. Under criteria in use by Brazil’s Ministry of Health up to Dec. 8, northeast Brazil (where the problem is thought to be most severe) would have had an estimated 158,000 suspected cases. But the standard adopted by the ministry on Dec. 8 reduced that estimate to 46,000 — less than a third of what it was previously — simply by setting the cutoff for microcephaly at a head circumference further from the norm.
In other words, the newer standard allowed for more natural variation. It’s not the head size but the brain and its development that matter here, and some babies can have smallish heads and still have normally developing brains. (The earlier standard found more cases but would have had more than three times the false-positive rate, meaning that it would label lots of healthy babies as having microcephaly.) The Pan American Health Organization uses an even more stringent cutoff, which would estimate 29,000 cases in this part of Brazil.
None of this means cases of microcephaly aren’t increasing, but the uptick doesn’t appear to be as drastic as first reported. According to an update published on Wednesday by the Brazilian Health Ministry, there are currently 5,280 suspected cases of microcephaly or other central nervous system problems. Of these, 3,935 are still being investigated, 837 have been discarded as something else and 508 have been confirmed.
While the case numbers are sorted out, researchers are gathering and assessing evidence regarding Zika and microcephaly. The Pan American Health Organization reported in December that Zika virus had been identified in the amniotic fluid of two pregnant women whose fetuses had been diagnosed with microcephaly, and scientists also found RNA from the Zika virus in the brain of an infant with microcephaly who had died shortly after birth. A recent case report published in the New England Journal of Medicine described a European woman who developed symptoms of Zika while living in Brazil during the first trimester of a pregnancy. She went to Europe for the latter part of her pregnancy, and after an ultrasound revealed severe microcephaly, the pregnancy was terminated. Tests on the fetus showed Zika virus in the brain tissue. The finding of Zika in babies is compelling, but it’s still possible that it’s an incidental finding.
Zika would not be the first virus to lead to microcephaly — rubella can also cause the condition — but it would be the first virus of its kind known to trigger it. Other flaviviruses, such as dengue and West Nile, are not known to cause microcephaly, and that’s perplexing, said Moore, the RAND pediatrician. “There are plenty of other viruses in this family, and none of them cause this.”
If the link between Zika and microcephaly is real, and the experts I talked with suspect that it is, then one outstanding question is why microcephaly hasn’t been seen in other areas where Zika outbreaks have occurred. The virus was discovered in 1947 in the Zika forest of Uganda, and it has also appeared in Southeast Asia and French Polynesia.
Eskild Petersen, an expert in vaccines and infectious diseases at Aarhus University in Denmark, has posed this question to colleagues in Southeast Asia, and “their answer was that the surveillance system was really poor.” It’s a problem that’s common across areas that have experienced Zika.
In French Polynesia, an outbreak of Zika in 2013 and 2014 was not associated with microcephaly at the time, but when researchers went back later to look, they found 17 potential cases. Microcephaly thought to be linked to Zika still hasn’t been seen in Brazil’s neighbor Colombia, which also is experiencing an outbreak, although it’s possible that most babies in Colombia whose mothers have been exposed to Zika haven’t been born yet.
What’s needed now are rigorous case-control studies (some of which are underway) of otherwise similar women whose babies had or didn’t have microcephaly. That could help explain why some babies developed the condition and others didn’t, Moore said. “The good studies now underway will hopefully help nail down whether [the link between Zika and microcephaly] is strong, weak or non-existent,” she said. There are numerous other possible explanations for the cases seen in northeast Brazil, including malnutrition, environmental factors or other viruses, and these have not yet been fully explored.
Even if Zika isn’t causing new cases of microcephaly, we’re going to need science to help us understand what is. If there really is an uptick in microcephaly cases and they are not the result of Zika, then prematurely pinning them on the virus could obscure the true cause and stand in the way of doing something about it.
Anna Maria Barry-Jester contributed reporting.