Microcephaly in north-east Brazil: a retrospective study on neonates born between 2012 and 2015

Abstract Objective To assess the number of children born with microcephaly in the State of Paraíba, north-east Brazil. Methods We contacted 21 maternity centres belonging to a paediatric cardiology network, with access to information regarding more than 100 000 neonates born between 1 January 2012 and 31 December 2015. For 10% of these neonates, nurses were requested to retrieve head circumference measurements data from delivery-room books. We used three separate criteria to classify whether a neonate had microcephaly: (i) the Brazilian Ministry of Health proposed criterion: term neonates (gestational age ≥ 37 weeks) with a head circumference of less than 32 cm; (ii) Fenton curves: neonates with a head circumference of less than −3 standard deviation for age and gender; or (iii) the proportionality criterion: neonates with a head circumference of less than ((height/2))+10) ± 2. Findings Between 1 and 31 December 2015, nurses obtained data for 16 208 neonates. Depending on which criterion we used, the number of neonates with microcephaly varied from 678 to 1272 (4.2–8.2%). Two per cent (316) of the neonates fulfilled all three criteria. We observed temporal fluctuations of microcephaly prevalence from late 2012. Conclusion The numbers of microcephaly reported here are much higher than the 6.4 per 10 000 live births reported by the Brazilian live birth information system. The results raise questions about the notification system, the appropriateness of the diagnostic criteria and future implications for the affected children and their families. More studies are needed to understand the epidemiology and the implications for the Brazilian health system.


Introduction
Congenital microcephaly is a neurological condition defined by an occipital-frontal head circumference that is smaller than expected for the gestational age and gender. Head circumference is a validated measurement for intracranial brain volume, since the growth of the cranium depends on the forces of the expanding brain. Microcephaly could, therefore, be used as an indicator of an undersized brain. However, controversy exists concerning the suitable lower limit for this measurement, as well as over the need for ethnically controlled data. 1 Furthermore, establishing the clinical implications of an undersized brain is difficult. Common causes of microcephaly are genetic disorders, severe malnutrition during pregnancy and intrauterine infections -such as syphilis and toxoplasmosis.
Microcephaly is a rare condition. In the United States of America, the prevalence has been estimated to range from 2.0 to 12.0 newborns with microcephaly per 10 000 live births 2 and the European Surveillance of Congenital Anomalies centre reports 2.9 newborns with microcephaly per 10 000 live births. 3 In Brazil, the live birth information system, SINASC, reported a prevalence of 0.6 newborns with microcephaly per 10 000 live births in 2010. 4 However, the reporting of microcephaly was neither compulsory nor had clearly defined criteria. Between November 2015 and February 2016, however, the reported number of newborns with microcephaly reached a total of 5280 in 25 of the 27 Brazilian states. More than 80% of them were from north-east Brazil and the State of Paraíba reported 776 newborns with microcephaly. 5 Since 2012, the government in Paraíba has run a paediatric cardiology network, 6 in collaboration with Círculo do Coração, a nongovernmental organization. The network has screened and stored cardiovascular data from more than 100 000 neonates from this state. However, the original data set does not include head circumference data. To increase the knowledge about microcephaly prevalence in Paraíba, the network carried out a four-week exercise between 1 and 31 December 2015, to obtain head circumference data from 10% of the neonates in the data set. This article summarizes the results from this research.

Methods
The network collected data on all neonates born in the 21 participating maternity centres. To retrieve data on head circumference for a subset of these neonates, we exported data -including an identification number, the date of birth and the mother's name, for all neonates in the network -from the database to an electronic spreadsheet. We added an extra column to the spreadsheet with the heading head circumference. We sent the spreadsheet to each participating centre by email, and in the email asked the nurses to retrieve the head circumference measurements from the delivery-room books. In Brazil, a newborn's occipital-frontal head circumference is measured in the delivery room by a nurse or a paediatrician, who uses a common measuring tape. The measurement is noted in the delivery room book as part of the routine perinatal practice in Paraíba state.
The goal was to retrieve data on approximately 10% of the neonates born between 1 January 2012 and 31 December 2015. To distribute the collection of data evenly over time, we suggested that the nurses should randomly select 10% of Objective To assess the number of children born with microcephaly in the State of Paraíba, north-east Brazil. Methods We contacted 21 maternity centres belonging to a paediatric cardiology network, with access to information regarding more than 100 000 neonates born between 1 January 2012 and 31 December 2015. For 10% of these neonates, nurses were requested to retrieve head circumference measurements data from delivery-room books. We used three separate criteria to classify whether a neonate had microcephaly: (i) the Brazilian Ministry of Health proposed criterion: term neonates (gestational age ≥ 37 weeks) with a head circumference of less than 32 cm; (ii) Fenton curves: neonates with a head circumference of less than −3 standard deviation for age and gender; or (iii) the proportionality criterion: neonates with a head circumference of less than ((height/2))+10) ± 2. Findings Between 1 and 31 December 2015, nurses obtained data for 16 208 neonates. Depending on which criterion we used, the number of neonates with microcephaly varied from 678 to 1272 (4.2-8.2%). Two per cent (316) of the neonates fulfilled all three criteria. We observed temporal fluctuations of microcephaly prevalence from late 2012. Conclusion The numbers of microcephaly reported here are much higher than the 6.4 per 10 000 live births reported by the Brazilian live birth information system. The results raise questions about the notification system, the appropriateness of the diagnostic criteria and future implications for the affected children and their families. More studies are needed to understand the epidemiology and the implications for the Brazilian health system.

Microcephaly in north-east Brazil
Juliana Sousa Soares de Araújo et al.
each month's deliveries. The nurses first printed the spreadsheet and wrote down the head circumference of the selected neonates on the list. Then, they logged on to the network's database, typed in the identification number for each selected neonate and manually entered the measurement in the newly added field for the head circumference data.
We obtained additional information about the mother's name and address, the gestational age, the length and weight, and the gender of the neonate at birth, from the network's database. Data on ethnicity were not collected. We only included neonates with complete data.
We classified neonates as having microcephaly by using one of the following criteria: (i) the criterion proposed by the Brazilian Ministry of Health, for which a term neonate (gestational age ≥ 37 weeks) is diagnosed with microcephaly if the head circumference is less than 32 cm; 7 (ii) Fenton curves, for which neonates are classified as having microcephaly if the head circumference is less than −3 standard deviation for age and gender; 4 or (iii) the proportionality criterion, for which a neonate is classified as having microcephaly if the head circumference is less than ((height/2)+10) ± 2. 8 We made four different analyses: one for each criterion and one where the neonates had to fulfil all three criteria. In addition, within each of the criterion, a neonate was classified as having severe microcephaly when the measurement was −3 standard deviation or more for that criterion.
We used the software R, version 3.3.0 (R Foundation, Vienna, Austria) for all data analysis. For statistical analysis we used the Friedman test.
Depending on which criterion we used, 4.2% (678) to 8.2% (1272) of neonates were classified as having microcephaly and 316 (2.0%) neonates fulfilled all three criteria. Of the neonates that fulfilled all three criteria, three were classified as having severe microcephaly ( Table 2).
Between 2012 and 2015, the number of neonates with microcephaly fluctuated, all of the criteria showed a similar pattern over this time. From the end of 2012, the numbers were higher than expected, with the highest peak in mid-2014 ( Fig. 1). When we only considered neonates with severe microcephaly, we observed a significant increase in numbers (P = 0.001) from the third quarter of 2015 (Fig. 2).

Discussion
Between 2012 and 2015, the Brazilian live birth information system reported a microcephaly prevalence of 6.4 per 10 000 live births in Paraíba. 9 Our study indicates that the prevalence could have been even higher. When projecting our findings to the total number of live births in Paraíba in 2014 (n = 58 147), if we use the proposed criterion from the health ministry, the estimated number of neonates born with microcephaly in that year is 4652; if we use the Fenton curves, the estimated number is 2442; and if we use the proportionality criterion, the estimated number is 2907. The estimated number of neonates fulfilling all three criteria is 1105.
These observations raise several questions. First, what is the true prevalence of microcephaly in north-east Brazil? The discrepancy in numbers between the Brazilian Ministry of Health and this study may reflect underreporting in recent years associated with an even greater incidence of microcephaly than presumed. It is possible that a high prevalence of the nonsevere forms of microcephaly had been occurring before the current outbreak, but health workers had only notified the live birth information system about the cases of neonates with typical severe phenotypes. As a result of the number of neonates with severe microcephaly increasing in the last quarter of 2015, the health workers' awareness might have increased, therefore causing them to notify the live birth information system about the milder forms as well. Second, what is the most appropriate diagnostic criterion for microcephaly in this setting? If we only consider the number of neonates with severe microcephaly in our study, this number is in the expected range of 2-12 neonates with microcephaly per 10 000 live births, reported by the United States Centers for Disease Control and Prevention. 2 However, most of the neonates in our study had milder forms of microcephaly. The clinical significance of these milder forms is not well established. For example, we do not know if a head circumference of 31 m or 32 cm in a term neonate could be within normal limits for this particular population. We also do not know if ethnical or nutritional factors could explain these findings. We need to determine if the population is facing large numbers of children with a neurological disease or if the observation is an anthropometric variation of normality.
Third, what is the cause of the increase in microcephaly prevalence? This question has attracted a lot of attention and the most likely explanation is the Zika virus outbreak that started in mid-2014 in Brazil. It is believed that the Zika virus was introduced to Brazil during the 2014 FIFA World Cup. 10 Researchers have postulated a possible association between microcephaly and the Zika virus intrauterine infection. 11 Evidence favouring this hypothesis are: perinatal transmission of the Zika virus; 12 the virus's strong neurotropism; 13 and the detection of the virus in amniotic fluid of fetuses with microcephaly. 11 The Zika virus was identified in Africa more than 50 years ago 14 and despite the numerous outbreaks of the virus, both inside and outside Africa, an increase in the number of neonates with microcephaly has not been reported. However, during a Zika virus outbreak in French Polynesia in 2013, the virus was associated with several conditions, including Guillain Barré syndrome and microcephaly. 12,15 The Zika virus is transmitted by Aedes aegypti mosquitoes and infections transmitted by this vector demonstrate temporal fluctuations similar to the fluctuations of microcephaly that we present here. 16 Researchers are considering whether other concurrent infections transmitted by Ae. aegypti, such as dengue and chikungunya, might explain the increase in microcephaly seen in Brazil. The hypothesis is that concurrent infections have an additive effect that promotes microcephaly. A study from the island of La Réunion-an overseas department of France -showed an association between chikungunya infections and microcephaly. 17 Teratogens' exposure -such as vaccines or drugs used in early pregnancy -is another potential factor to consider as a potential cause of microcephaly. 1,18 Furthermore, malnutrition, which has previously been  Year and quarter  6 In Brazil, controlling the Ae. aegypti vector has been a major public health strategy to combat the arboviruses. This strategy is justified both by its potential to reduce the number of babies born with microcephaly -if the association with Zika virus infections proves trueand for the reduction of other Ae. aegypti transmitted diseases, such as dengue and chikungunya infections.
At this stage, we can only conclude that Brazil is facing a new and challenging public health problem. The current epidemiological and clinical data are insufficient to make conclusions concerning the risk factors of microcephaly and the pathogenic mechanisms of the Zika virus. Further retrospective studies and follow-up investigations on children with well-defined or borderline microcephaly will be important to clarify the etiology as well as the neurological consequences of these diagnoses. Children born with microcephaly can also have other birth defects, which could further aggravate the neurological manifestations. For the affected children and the families it is paramount that the government provides management strategies, such as social inclusion programmes and access to specialized health care. ■ Year and quarter  Цель Оценить количество детей, рожденных с микроцефалией, в штате Парайба на северо-востоке Бразилии.

Microcephaly in north-east Brazil
Juliana Sousa Soares de Araújo et al.