RUNPHOTO / GETTY IMAGESThe Zika virus had to circumnavigate half the globe to lose its anonymity. For almost 60 years the virus circulated in Africa and Asia almost without being noticed, but upon landing in Brazil, it found favorable conditions. Zika spread rapidly and attracted international attention by becoming the prime suspect behind an increase in cases of microcephaly, a type of congenital malformation that few in Brazil had heard of.
Microcephaly is a term of Greek origin used by doctors to describe a condition in which infants are born with heads that are too small for the length of the gestation period. Most of them, according to experts, are healthy. Only a small portion of them are born with microcephaly due to developmental problems resulting in a smaller brain. In such cases, there is no cure. A baby can be born with a brain that is too small because of a number of genetic defects—there are at least 16 known genes associated with the problem. But microcephaly can also occur for environmental reasons, such as alcohol abuse or exposure to toxic products during pregnancy, or as a result of a series of infections such as those caused by the rubella and herpes viruses, by the toxoplasmosis parasite or by syphilis bacteria.
The possibility that the Zika virus could also cause microcephaly sounded a general alarm because of the ease with which it spreads. Long considered harmless, Zika entered Brazil between 2014 and 2015 and, according to the Ministry of Health, may already have infected 1.4 million people. During the same period of time, an increase in cases of microcephaly appeared, especially in Brazil’s Northeast. Between 2000 and 2014, the Ministry recorded an annual average of 164 cases of microcephaly. But between October 2015 and February 20, 2016, the number of confirmed cases reached 583.
Amid the outbreak, politicians and health authorities concluded that Brazil was facing the most terrible epidemic of recent times, which, if not contained, could leave a generation of Brazilians with neurological damage or, as they said, “with sequelae.”
Evidence began to emerge, however, of something that many people already suspected: the number of cases of microcephaly has always been underestimated in Brazil. Not having a clear picture of the number of microcephaly cases before Zika’s entry into Brazil makes it difficult to know if the problem is actually increasing—and, if so, how much of an increase has there been and what proportion of it is due to the virus. In this scenario, collecting data to ascertain how the problem has evolved over time is as important as studying the best way to fight the virus and the mosquito.
|More about the Zika virus|
|Zika: The virus that took Brazil by surprise|
|The spread of Zika|
An important indication that the Brazilian health system has not identified most cases of microcephaly comes from a recent study conducted by researchers from Pernambuco and Paraíba, the two states that reported the most births of babies with abnormally small heads in recent months.
With the possibility of facing an outbreak of microcephaly, Dr. Sandra da Silva Mattos, a specialist in fetal cardiology in Recife, Pernambuco State, decided to challenge her team. She coordinates a cardiology network that in recent years has followed 100,000 newborns in the neighboring state of Paraíba. In late 2015, Dr. Mattos recruited 40 nurses and nursing assistants from 21 Paraíba maternity wards and asked them to search the records of delivery rooms to retrieve information about 10% of the infants.
Then they went further. In December, they reviewed the head size measurements (head circumference) of 16,208 babies born between 2012 and 2015 in Paraíba. The survey indicated that between 2% and 8% of these children could be classified as having microcephaly, depending on the criteria used to define the problem. This represents, respectively, 320 and 1,300 newborns and does not mean that all suspected cases of microcephaly are necessarily associated with the Zika virus.
The important thing is that even the smaller number, obtained by the most restrictive criteria and representing the most serious cases of microcephaly, would already amount to about half the annual average of 164 cases that the Ministry of Health recorded for all of Brazil through the Live Birth Information System (Sinasc), a national database that collects information about Brazilian newborns. This database contains a field for entry of a skull measurement, but, as many researchers suspect, it was often not filled in—perhaps because microcephaly reporting was not previously required.
In the last four months, the Ministry of Health identified a higher number of cases of microcephaly, after receiving an alert from doctors in Pernambuco who had detected an unusual increase in the number of children born with smaller heads than normal for the gestation period.
Between November 8, 2015 and February 20, 2016, at least 5,640 babies were born in Brazil with this characteristic. This figure represents an average of 46 suspected new cases of microcephaly per day, a more alarmingly high proportion than previously known. Between 2000 and 2014, the average recorded by Sinasc was approximately one every two days. The increase in possible cases and their association with a Zika virus infection during pregnancy positioned microcephaly as the main threat to national public health.
“The Paraíba study is important because it showed, by using the microcephaly criteria adopted by the Ministry, that there was a blind spot in the data and that Sinasc was not detecting most cases,” says Professor Fernando Kok, pediatric neurologist at the University of São Paulo School of Medicine (FMUSP).
In fact, the portion previously identified by Sinasc was tiny. Each year about 2.9 million children are born in Brazil, and the 164 cases of microcephaly reported annually between 2000 and 2014 represent only 0.006% of that universe. That figure is very low compared to little known data from other populations. The United States, for example, has adopted criteria similar to Brazil’s to define microcephaly, and it has a higher proportion of cases.
Close to 3.9 million babies are born in the U.S. each year and, according to a review published in 2009 in the journal Neurology, there were about 25,000 cases of microcephaly identified. This means that approximately 0.6% of American babies presented microcephaly, and that the problem in the U.S. would be 100 times more common than in Brazil.
For purposes of comparison, data were converted into a slightly more concrete number so the rate of 0.006% measured by Sinasc would indicate that only 60 Brazilian newborns in each group of 100,000 would have microcephaly and should be referred for further evaluation. Already the more conservative rate (2%) now found in Paraíba would be equal to 2,000 children in each group of 100,000—or 58,000 in the entire country.
Is that a lot? Maybe not. It depends on the criteria used to define microcephaly. In early December 2015, the Ministry started to classify infants suspected of having microcephaly using as its criterion head measurements of less than 32 centimeters (cm) in circumference at birth. Doctors, epidemiologists and statisticians often use a very simple graph to determine whether specific measurements presented by an individual deviate substantially from the standard for the population—in some cases this difference may indicate a health problem.
The graph is constructed by placing the head measurements of infants of a population on the horizontal axis and the number of infants on the vertical axis. In general, the head size of human infants is between 30 cm and 39 cm. Almost 20 years ago a survey commissioned by the World Health Organization (WHO) for an international consortium of researchers took several measurements, including the heads of 27,000 infants from different populations, including Brazil. One outcome of this work was a graph showing the distribution of skull size in the human population. It presents a bell-shaped curve that statisticians use to present well known mathematical properties.
One such property is that the mean—in this case, the total sum of head measurements divided by the total number of children—divides the graph in the middle, into two symmetrical parts (see graph on next page). Statisticians know that the total area under the curve represents the entire studied population, and they can easily calculate the proportion of people who fit into certain ranges of the curve.
Doctors and epidemiologists rely on this information to ascertain whether a particular measurement may indicate a health problem. The general idea behind this type of tool is that anything that departs substantially from the norm of most people may be a sign of trouble—these curves are used, for example, to assess whether a child is too short indicating growth problems or to ascertain whether the concentration of certain fats in the blood have reached levels harmful to health.
In the case of skull size, the 32 cm criterion adopted by the Ministry represents the cutoff point in determining whether a child is suspected of having microcephaly. This point was probably chosen because it deviates substantially from the average head size of most newborns. After 37 weeks of gestation, the head size of babies considered healthy usually measures around 34.5 cm, according to WHO data. The difference may seem small, but 2.5 cm is substantial for a baby.
Statisticians use a measurement known as standard deviation to get an idea of this degree of departure from the norm. In the bell-shaped graph, 32 cm is approximately two standard deviations below the mean. Based on the properties of normal distribution, it is known that a small part of the sample, just 2.3%, is more than two standard deviations from the mean.
This means that 2.3% of babies born in Brazil—representing 66,700 infants—could fit the definition of microcephaly established by the Ministry. A much smaller proportion of newborns (0.1% or 2,900 babies) have even smaller heads. The size of their skulls is three standard deviations below the mean, and in most cases, indicates a problem in brain development.
“The vast majority of infants classified with microcephaly in any country following the WHO recommendation [that is, those that are two standard deviations below the mean] will be normal with a small head,” says epidemiologist Cesar Victora of the Federal University of Pelotas (UFPel). He contends that the pathological cases associated with Zika and other infections or genetic problems represent a small minority of the 66,700. “The vast majority of these children are normal and have small heads for non-pathological genetic reasons. They have small heads and bodies because their parents are small, or because they were subjected to some type of intrauterine growth restriction like, for example, their mothers smoked during pregnancy,” says Victora.
“The fact that the head size falls below a certain value does not necessarily indicate illness or disease,” notes Kok, who tracks cases of microcephaly at USP’s Hospital das Clínicas. “The skull measurement has to be analyzed in conjunction with other information. Now, if the measurement deviates substantially from the mean, then there is a greater probability of a problem.”
If a measurement of two standard deviations below the mean is really a good indicator of microcephaly—some countries in Europe use three standard deviations—both the Brazilian and United States health care systems are failing to evaluate many infants who should be looked at more carefully. Some of them are known to be healthy and will not present neurological development problems in the future, but another portion may become ill and deserve to undergo a more detailed assessment.
In Brazil, São Paulo biologist Fernando Reinach was one of the first to present this data to a wider audience. In his column in the newspaper O Estado de S. Paulo, published on February 6, 2016, he called attention to the fact that the official figures for microcephaly diverged from those forecast for Brazil. In a portion of the text “Microcephaly has always existed,” he noted: “These children should have been identified and carefully examined. But they were not, because notification of this characteristic was not required. There have certainly always been children with these characteristics, but there are no statistics in the National Healthcare System (SUS). Now, with mandatory notification, and the panic caused by Zika, they are ‘showing up’. Their sudden appearance could be real, and caused by Zika, or they could be an anomaly caused by underreporting in Brazil,” wrote Reinach.
Like Reinach, some researchers interviewed by Pesquisa FAPESP have complained about the lack of reliable historical data on microcephaly for Brazil. The lack of information for prior years, they say, makes it difficult to determine if the current numbers are growing only because of Zika, or if there are other factors involved.
In late December 2015, researchers from the Latin American Collaborative Study of Congenital Malformations (ECLAMC), an international consortium that monitors malformation records in 35 hospitals in seven countries, reviewed Brazilian microcephaly data between 1967 and 2015 and cross-referenced it with the information collected by Sinasc in the last three years.
In a summary report, available on the Group’s website, the researchers say that the Sinasc numbers were underestimated. According to ECLAMC calculations, two cases of microcephaly are expected for each group of 10,000 babies born in Brazil, but this rate should be higher in the Northeast, where the problem is more common than in other regions. Using the microcephaly rate observed in Europe, they calculated that there should be 45 cases among the 147,597 babies born in Pernambuco State in 2015. But by the end of December, 2015, the state had reported 1,153 suspected cases (a rate of 26 times more). For the researchers, these figures could only be explained if all of Pernambuco’s pregnant women had been infected by the virus—the document does not explicitly say what proportion of infected women could transmit the virus to the fetus.
The researchers at ECLAMC suspect that much of the increase is due to the active identification of cases and conclude that current data do not allow an evaluation of whether there was a real increase in the prevalence of microcephaly at birth in the Northeast, the magnitude of such increase and whether it was due to Zika exposure or other causes. The ECLAMC team was approached, but declined to be interviewed.
Despite these considerations and causality that has as yet to be demonstrated, in mid-February 2016, the Minister of Health, Marcelo Castro, said that 40% of suspected cases of microcephaly reported in recent months were related to infection by the Zika virus.
Epidemiological report No. 14, issued by the Ministry in late February 2016, indicates that of the 5,640 reported cases between November 8, 2015 and February 20, 2016, 1,533 were investigated and 583 (10.3% of the 5,640) were confirmed as microcephaly. According to the document, molecular tests detected the genetic material of Zika in 67 of the 583 confirmed cases. In the remaining 516 cases, confirmation was provided by brain imaging tests that revealed lesions previously associated with Zika. Also according to the report, the Ministry suspects that most of the mothers of these children had Zika. However, it is unclear if in the 516 cases classified by imaging, the possibility of other infections that cause microcephaly (toxoplasmosis, rubella, cytomegalovirus, herpes and syphilis) was eliminated.
The Ministry did not respond to requests for clarification.
Epidemiological report No. 14 also offered no details about the 950 cases that were excluded. The document suggests that the infants did not have microcephaly of infectious origin, but they could have another form of the problem.
It is known that infections are not the only cause of microcephaly—and perhaps not even the most common. The 2009 review appearing in the journal Neurology suggested that between 15% and 50% of cases of microcephaly could be of genetic origin. There are at least 16 known genes that cause the problem when their two copies are found to be altered. In addition, environmental factors such as alcohol consumption during pregnancy or exposure to pollutants and toxic products may also cause microcephaly. How much does each contribute to the total number of cases? “I know of no studies that show this,” says Kok.
A group of doctors and epidemiologists from the states of Rio Grande do Sul, São Paulo and Ceará suspect that the strategy of considering infants born with a skull size of less than 32 cm as a potential case of microcephaly includes many babies that are healthy.
In an article published in February 2016 in the journal The Lancet, the team led by Cesar Victora of UFPel raised several technical reasons for this. The first is that adopting a single cutoff point for babies of both sexes is not appropriate, since girls on average are born with smaller skulls than boys. In addition, the researchers argue, 68% of Brazilian babies are born before completing a 40-week gestation period, in part because of the high rates of Caesarean section, and they may be smaller than normal.
To reduce the number of babies who do not have microcephaly—known as false positives—among those who will undergo further evaluation, the group suggests adopting standard growth curves better suited to the Brazilian population and more capable of detecting the truly positive cases, such as the strategy generated by the Intergrowth-21st Consortium, which the Pelotas group (UFPel) helped to develop (see Pesquisa FAPESP Issue nº 225). Currently, in addition to 32 cm for babies born at 37 weeks of gestation, the Ministry has adopted a growth curve produced for infants in developed countries, known as the Fenton curve, to screen those who are born prematurely.
For physician and epidemiologist Eduardo Massad, also a professor at FM-USP, infection by the Zika virus may explain part of the increase in the number of cases of microcephaly. “By exactly how much? We do not know,” he says. In his opinion, the important thing is that the virus was found in 67 of 583 confirmed cases, which reinforces the connection of the virus to the problem, although a relationship of causality has not yet been conclusively demonstrated.
“There is undoubtedly an association between infection by Zika during pregnancy and the birth of babies with microcephaly, and it is perfectly plausible that part of the increase in cases is due to the virus,” says Massad. “A fraction of infected fetuses develop microcephaly, but the size of that fraction is still unknown.”
In the Paraíba study, the group led by Dr. Mattos detected a notable increase in severe cases of microcephaly from the third quarter of 2015, which could be associated with the circulation of the virus. She suspects, however, that it is too early to conclude that Zika is the only cause of the problem. “We do not want to eliminate the influence of the virus, but question whether there are not more factors involved, such as other infections and malnutrition, which are prevalent in the population,” says Mattos, who is director of the Maternal-Fetal Cardiology Unit of Real Hospital Português de Beneficência, located in Recife, Pernambuco State. “We need to know what we’re dealing with.”
The epidemiological research is just beginning. In Paraíba, the group led by Dr. Mattos is participating in a study with researchers from the U.S. Centers for Disease Control and Prevention (CDC) and the Ministry of Health, the aim of which is to determine the risk of infected women having children with microcephaly. In São Paulo, researchers of the Zika Network, a consortium of 40 São Paulo universities and research institutes funded by FAPESP, will carry out a similar study.
It will take several months to learn the results. According to Massad, more and longer studies are necessary—to follow the entire population and determine what proportion of pregnant women are infected by the virus and have healthy children or children with microcephaly.
SOARES DE ARAÚJO, J.S. et al. Microcephaly in northeast Brazil: a review of 16,208 births between 2012 and 2015. Bulletin of the World Health Organization. February 4, 2016.
ASHWAL, S. et al. Practice parameter: evaluation of the child with microcephaly (an evidence-based review). Neurology, V. 73, pp. 887-97. 2009.
VICTORA, C. G. et al. Microcephaly in Brazil: how to interpret reported numbers? The Lancet. February 13, 2016.