Adverse fetal and perinatal outcomes associated with Zika virus infection during pregnancy: an individual participant data meta-analysis

Abstract

Background: Zika virus (ZIKV) infection during pregnancy is associated with an increased risk of congenital malformations. The prevalence of short and long-term consequences, however, remains uncertain due to heterogeneity across studies. Individual Participant Data Meta-Analysis (IPD-MA) offers an alternative approach to provide more precise and generalisable estimates through data harmonisation across studies, allowing for standardised definitions and exploration of heterogeneity. This project was undertaken to estimate absolute and relative risks of adverse outcomes for individuals with ZIKV infection during pregnancy.

Methods: IPD-MA studies and their datasets were identified through a systematic search conducted in 2018 with the following criteria: observational longitudinal or surveillance-based studies investigating ZIKV during pregnancy or at birth, measured fetal, infant, or child outcomes, and included at least 10 participants. Here we used IPD data shared by March 2022 from 18 studies from international health organisations and research networks, comprising 24 unique datasets, in 11 countries. Datasets were harmonised with standardised definitions, using variables related to pregnant individuals, methods used for ZIKV diagnoses, fetal characteristics and outcomes, and pooled for analysis. Frequentist and Bayesian regression methods were applied to estimate outcome prevalence and evaluate the association between maternal ZIKV infection and fetal loss, microcephaly and congenital zika syndrome as primary outcomes.

Findings: Data including 9568 pregnant individuals and 9608 newborns, were harmonised. The risk of severe primary microcephaly was significantly higher in ZIKV-positive pregnancies (1.5%, CI 0.8%-2.7%) compared to ZIKV-negative ones (0.3%, CI 0.1%-1.0%), with a relative risk of 4.5 (CI 1.5-13.3) in the one-stage meta-analysis. While some risk estimates were consistent between Bayesian and Frequentist methods, estimates for other outcomes varied, underscoring the influence of both the analytical approach and the definition of ZIKV on the associations.

Interpretation: Our findings align with previously published meta-analyses and indicate an added burden to adverse pregnancy outcomes with higher prevalence compared to pre-epidemic population-based average values. Future research should focus on additional outcomes with clear definitions of maternal infection. Women of reproductive age should be informed about the risks of Zika infection during pregnancy to support reproductive planning.

Funding: This project was supported by the Wellcome Trust grant number 206532/Z/17/Z, the WHOHealth Emergencies Programme Global Arbovirus Initiative, and the WHO Department of Sexual and Reproductive Health and Research, including the Human Reproduction Special Programme (HRP).

Keywords: Congenital Zika Syndrome; Individual participant data meta-analysis; Microcephaly; Perinatal outcomes; Pregnancy; Zika virus.

Fig. 1

Study selection.

Fig. 2

(a) Absolute Risk of Microcephaly (<2SD) in sZIKV-positive, (b) Absolute Risk of Microcephaly (<2SD) in sZIKV-negative, (c) Relative Risk of Microcephaly (<2SD), (d) Absolute Risk of Microcephaly (>3SD) in sZIKV-positive, (e) Absolute Risk of Microcephaly (>3SD) in sZIKV-negative, (f) Relative Risk of Microcephaly (>3SD), (g) Absolute Risk of Fetal Loss in sZIKV-positive, (h) Absolute Risk of Fetal Loss in sZIKV-negative, (i) Relative Risk of Fetal Loss, (j) Absolute Risk of CZS (WHO definition) in sZIKV-positive, (k) Absolute Risk of severe CZS (WHO definition) in sZIKV-positive, (l) Absolute Risk of Congenital Zika in sZIKV-positive. Absolute and relative risks of microcephaly, fetal loss, and Congenital Zika Syndrome (CZS), stratified by maternal ZIKV infection status. Error bars indicate 95% confidence intervals (CI). Squares represent study-specific estimates and diamonds the overall pooled estimates (red: one-stage meta-analysis; blue: two-stage meta-analysis). Dashed vertical lines indicate no-effect values. Abbreviations: BRA, Brazil; COL, Colombia; ESP, Spain; GUF, French Guiana; GRD, Grenada; GUA, Guadeloupe; MTQ, Martinique; REU, Réunion; VEN, Venezuela; ZIKV, Zika virus; CZS, Congenital Zika Syndrome; WHO, World Health Organization; SD, Standard deviations.