In 2015, the Brazilian Health Ministry identified a Zika virus (ZIKV) outbreak in the northeastern region of the country. The outbreak has intensified and spread, reaching into 46 previously unaffected populations, including the southern United States. The epidemic is of special concern because Zika was found to cause severe neurological disorders among infants. Government agencies across the Americas have advised populations to take the epidemic into account when making family planning decisions, including specific appeals for fertility delay. Research on Zika has focused on transmission dynamics, disease pathogenesis, and teratogenic effects. We focus on an underexplored area of impact: population fertility. We build a detailed data set characterizing demographic patterns in Brazil from 2008-2018. Sources include natality files, records of interactions with the health care system, ecological data, and two types of media resources. Because we can identify pregnancies in progress, we provide the earliest possible estimates of fertility change driven by the epidemic. Our study has three aims: 1) To estimate the effect of Zika risk on population fertility; 2) To identify the magnitude of socioeconomic heterogeneity in fertility response to the epidemic; (3) To test for evidence of behavioral mechanisms through which a fertility response occurs. We estimate municipality-level pregnancy/fertility rates, and overlay those estimates on vulnerability to epidemic spread of Zika. We model two components of vulnerability: (a) transmission risk based on the suitability of each municipality's climate and topography for reproduction of the primary Zika mosquito vector (Aedes aegypti), and (b) depiction of infection risk using temporally-and spatially-referenced data from local and social media sources. Fertility rates calculated before the epidemic, and rates calculated in areas with minimal Aedes prevalence, provide an estimate of the expected temporal patterns within exposed areas in the absence of the epidemic. The difference between the observed and the expected patterns in fertility allow us to describe the direction, magnitude, and timing of the fertility response. Health-system data on long-acting contraceptives, prenatal practices, STIs, among others, allow us to detect evidence that fertility change is accompanied by related, behavioral responses to the epidemic. Our research provides early estimates of a major public health impact of Zika. It advances our understanding of how (and which) families shift fertility timing in response to reproductive health threats, shedding light on important patterns in population dynamics during and after epidemics.
In response to mounting maternal and infant health risks, public health officials across the Americas have encouraged populations affected by the Zika virus (ZIKV) epidemic to delay childbearing. We build a novel data resource on Brazilian demographic behavior to provide the first tests of whether, and among which sub-populations, fertility delay has occurred. To better understand the public health needs of the population, our research emphasizes (a) assessment of behavioral responses to the epidemic, (b) detection of socioeconomic variability in these responses, and (c) differentiation of infection risk from depiction of infection risk across multiple sources of media communication.
