Articles
Association between microcephaly, Zika virus infection, and other risk factors in Brazil: final report of a case-control study

https://doi.org/10.1016/S1473-3099(17)30727-2Get rights and content

Summary

Background

A Zika virus epidemic emerged in northeast Brazil in 2015 and was followed by a striking increase in congenital microcephaly cases, triggering a declaration of an international public health emergency. This is the final report of the first case-control study evaluating the potential causes of microcephaly: congenital Zika virus infection, vaccines, and larvicides. The published preliminary report suggested a strong association between microcephaly and congenital Zika virus infection.

Methods

We did a case-control study in eight public maternity hospitals in Recife, Brazil. Cases were neonates born with microcephaly, defined as a head circumference of 2 SD below the mean. Two controls without microcephaly were matched to each case by expected date of delivery and area of residence. We tested the serum of cases and controls and the CSF of cases for detection of Zika virus genomes with quantitative RT-PCR and for detection of IgM antibodies with capture-IgM ELISA. We also tested maternal serum with plaque reduction neutralisation assays for Zika and dengue viruses. We estimated matched crude and adjusted odds ratios with exact conditional logistic regression to determine the association between microcephaly and Zika virus infection.

Findings

We screened neonates born between Jan 15 and Nov 30, 2016, and prospectively recruited 91 cases and 173 controls. In 32 (35%) cases, congenital Zika virus infection was confirmed by laboratory tests and no controls had confirmed Zika virus infections. 69 (83%) of 83 cases with known birthweight were small for gestational age, compared with eight (5%) of 173 controls. The overall matched odds ratio was 73·1 (95% CI 13·0–∞) for microcephaly and Zika virus infection after adjustments. Neither vaccination during pregnancy or use of the larvicide pyriproxyfen was associated with microcephaly. Results of laboratory tests for Zika virus and brain imaging results were available for 79 (87%) cases; within these cases, ten were positive for Zika virus and had cerebral abnormalities, 13 were positive for Zika infection but had no cerebral abnormalities, and 11 were negative for Zika virus but had cerebral abnormalities.

Interpretation

The association between microcephaly and congenital Zika virus infection was confirmed. We provide evidence of the absence of an effect of other potential factors, such as exposure to pyriproxyfen or vaccines (tetanus, diphtheria, and acellular pertussis, measles and rubella, or measles, mumps, and rubella) during pregnancy, confirming the findings of an ecological study of pyriproxyfen in Pernambuco and previous studies on the safety of Tdap vaccine administration during pregnancy.

Funding

Brazilian Ministry of Health, Pan American Health Organization, and Enhancing Research Activity in Epidemic Situations.

Introduction

In August, 2015, physicians reported a cluster of cases of microcephaly in the state of Pernambuco, northeast Brazil. Microcephaly is an abnormality in birth that was rarely reported before the Zika virus epidemic.1 Microcephaly is a clinical sign that can reflect abnormal brain development, but it can be also be found in healthy neonates. By definition, microcephaly is any insult that disturbs early brain growth, and it can be caused by genetic variations, teratogenic compounds, or other congenital infections (such as cytomegalovirus, rubella, herpes, or toxoplasmosis).2

At the start of this microcephaly epidemic, the main causal hypothesis was Zika virus infection during pregnancy,3 but other possible causes were proposed; two of these causes were of particular interest because of their potential implications. The first of these possible causes was larvicide use in reservoirs of drinking water to control Aedes aegypti, namely pyriproxyfen, which was introduced in 2014 by the Brazilian Ministry of Health).4 The other possible cause of interest was vaccine administration during pregnancy.5, 6

Research in context

Evidence before this study

We searched PubMed and LILACS using the search terms “Zika” and “case-control study”. We searched for articles published up to Sept 30, 2017, including publications in English, Portuguese, and Spanish. The causal link between Zika virus infection and microcephaly, as part of the congenital Zika virus syndrome, is now well established; however, we did not identify any case-control studies of Zika virus infection and microcephaly. The final piece of the puzzle, providing epidemiological evidence, was the preliminary finding of a strong association in a case-control study of Zika virus infection and microcephaly in Recife, Pernambuco (Brazil), the hotspot of the microcephaly epidemic. Other risk factors have been suggested but never investigated at individual level, the more crucial being vaccines during pregnancy and use of the larvicide pyroxifen in containers of drinking water for mosquito control.

Added value of this study

This is the final report of a case-control study, with a much larger sample size than a preliminary analysis of a subset of these data. This analysis supports the strength of association with Zika virus and, for the first time, investigates other potential risk factors including use of larvicides and vaccination during pregnancy. We confirm the strong association between Zika virus infection and microcephaly at birth and provide evidence that use of larvicides and vaccines during pregnancy did not increase the risk of microcephaly. We also provide information regarding all cases of microcephaly born during the study period: about half had either laboratory confirmation of Zika virus or typical brain image abnormalities. No controls had laboratory-confirmed Zika virus infection. There was some association between laboratory-confirmed Zika virus infection and cerebral abnormalities; 60% of those with brain abnormalities were negative for Zika virus when tested with specific IgM and PCR, and about half of those who were Zika virus-positive had no cerebral abnormalities. A high proportion of cases of microcephaly were small for gestational age. The high prevalence of serological markers of Zika virus infection in the mothers of controls indicate a high transmission of infection in the study area.

Implications of all the available evidence

This study supports the magnitude of risk of microcephaly associated with congenital Zika virus infection; provides evidence that neither larvicide or vaccinations during pregnancy caused the epidemic; highlights that neither a negative laboratory result for Zika virus nor an absence of cerebral abnormalities alone are sufficient to discard Zika virus as a cause of individual cases of microcephaly.

Microcephaly was the first postnatal clinical finding to be reported at the beginning of the epidemic.7, 8, 9 However, rapidly accumulating evidence showed that congenital Zika syndrome could cause more than isolated microcephaly.10, 11, 12 In the early months of the marked increase in the prevalence of microcephaly, we designed a case-control study13 to investigate an association between microcephaly and congenital Zika virus infection and other potential causes. The previously published preliminary report13 documented a strong association with Zika virus; we now report the final results, with the aim of assessing the association between microcephaly and congenital Zika virus infection, along with a comprehensive investigation of other potential risk factors in an epidemic context in Pernambuco, Brazil.

Section snippets

Study design and participants

We present the final analysis of our case-control study of neonates who were consecutively recruited at birth. The protocol can be accessed online.

The study population consisted of neonates born from women residing in Pernambuco, Brazil, and delivered in eight public maternity hospitals in Recife. Cases—neonates with microcephaly (livebirths or stillbirths)—had head circumferences at least 2 SD smaller than the mean for their sex and gestational age on the Fenton growth chart.14 Microcephaly

Results

The preliminary analysis included participants recruited from Jan 15 to May 2, 2016;13 this Article includes participants recruited up to Nov 30, 2016. We did this analysis before reaching 200 cases for two reasons: first, we reached the necessary power for statistical analysis because the proportion of controls who were exposed to Zika virus was lower than expected (as evidenced by the absence of Zika virus infection in all controls); second, the epidemic slowed down in Recife and cases became

Discussion

The association between microcephaly and Zika virus infection, confirmed by qRT-PCR, capture-IgM ELISA, or both, was strong after controlling for confounders. The association was strong with severe and non-severe microcephaly. None of the other risk factors investigated was associated with microcephaly in multivariable analyses; these factors include the use of the larvicide, pyriproxyfen, and vaccine administration during pregnancy. These data support our preliminary findings that the increase

References (34)

  • MF Albuquerque et al.

    Pyriproxyfen and the microcephaly epidemic in Brazil—an ecological approach to explore the hypothesis of their association

    Mem Inst Oswaldo Cruz

    (2016)
  • FM Munoz et al.

    Safety and immunogenicity of tetanus diphtheria and acellular pertussis (Tdap) immunization during pregnancy in mothers and infants: a randomized clinical trial

    JAMA

    (2014)
  • T Walls et al.

    Infant outcomes after exposure to Tdap vaccine in pregnancy: an observational study

    BMJ Open

    (2016)
  • A Panchaud et al.

    Emerging role of Zika virus in adverse fetal and neonatal outcomes

    Clin Microbiol Rev

    (2016)
  • Microcephaly in infants, Pernambuco State, Brazil, 2015

    Emerg Infect Dis

    (2016)
  • DB Miranda-Filho et al.

    Initial description of the presumed congenital Zika syndrome

    Am J Public Health

    (2016)
  • P Brasil et al.

    Zika virus infection in pregnant women in Rio de Janeiro

    N Engl J Med

    (2016)
  • Cited by (0)

    *

    Contributed equally

    Investigators contributing on behalf of these organisations are listed in the appendix

    View full text