dcyphr | Impact of preexisting dengue immunity on Zika virus emergence in a dengue endemic region


Clinical outcomes associated with Zika virus (ZIKV), a mosquito-borne virus, in the Americas are well-documented, but other aspects such as risk factors are not well-understood. In this study, the researchers prospectively followed a cohort of 1453 urban residents in Salvador, Brazil. They used an assay that measured immunoglobulin G3 (IgG3), which function as antibodies, against ZIKV NS1 antigen. The researchers estimated that 73% of individuals in the cohort were infected with ZIKV during the 2015 outbreak. Attack rates of the virus were very different across very small areas. Preexisting antibodies against dengue virus, also a mosquito-borne virus, were associated with reduced risk of ZIKV infection and symptoms. The ZIKV immunity generated by the 2015 outbreak in this population may affect the risk for future transmission.


The researchers wanted to understand how dengue and Zika viruses, which are related mosquito-borne viruses, interacted with one another in terms of immunity and susceptibility.


There is much uncertainty about the dynamics and circumstances of the Zika virus (ZIKV) outbreak in the Americas in 2015. The association of ZIKV with microcephaly (a condition where head circumference is smaller than normal, damaging the brain) makes it important to determine future dynamics of transmission.

An unanswered question is how preexisting immunity to dengue virus (DENV), which is genetically and antigenically similar to ZIKV, in suppressing ZIKV transmission. Some previous studies have shown enhanced ZIKV infection in the presence of DENV antibodies, but other studies have been less clear. Immunity to DENV has been shown to protect against ZIKV infection, and thus could reduce ZIKV infection in a DENV-immune population. However, this relationship has not been evaluated in humans.

It is difficult to characterize interactions between ZIKV and DENV due to asymptomatic infections and difficulties and diagnosis. Infections can only be understood by serological measures, which quantify and characterize antibodies in blood samples. Current serological tests are limited because of cross-reactivity between the two related viruses. These limitations have prevented full understanding of the American Zika epidemic and interactions between DENV and ZIKV.


High ZIKV attack rates in an urban population

The researchers prospectively characterized ZIKV transmission in Pau da Lima, an urban slum community in Salvador, Brazil. Northeast Brazil has been endemic for DENV for over 30 years. It was also the epicenter of the 2016 Zika epidemic. The researchers quantified ZIKV infection rate using data from 1453 individuals. In 642 individuals for whom dengue immunity data was available, the researchers investigated the impact of DENV immunity on ZIKV susceptibility.

The researchers identified ZIKV infection with an assay that measures immunoglobulin G3 (IgG3) response to ZIKV NS1 protein. They performed additional validation of the assay to ensure cross-reactivity did not affect the results.

Samples collected before the 2015 ZIKV epidemic in Salvador were mostly negative for ZIKV immunity, suggesting little transmission before the epidemic. In contrast, 63% of samples collected in October 2015 were positive. Adjusting for limitations of the assay, this level of positivity corresponds to an attack rate of 73% in the population.

Heterogeneity in attack rates at fine spatial scales

Although the attack rate was high, they were not the same across the population and differed across distances as small as 20 meters. Similar differences have been reported in DENV transmission. This variation could be due to differences in mosquito populations or in exposure of people living in different conditions.

The researchers estimated that the basic reproductive number (R0) of ZIKV was 1.8, but ranged from 1.2 and 2.1 between areas of the study site. These results are within the range of previous estimates and calculations for both ZIKV and DENV.

Influence of dengue immunity on ZIKV infection

The researchers investigated potential associations between having DENV antibodies (total IgG) and susceptibility to ZIKV infection. 642 samples collected before the ZIKV epidemic were tested for antibodies against DENV. 553 samples were positive. Among individuals who had immunity to DENV, each time the amount of antibodies present against DENV was doubled, reduction in the risk of ZIKV infection rose by 9%. To investigate this effect, the researchers estimated the risk of ZIKV infection among individuals with different levels of pre-outbreak immunity to DENV. Individuals with mid- to high-levels of antibody experienced a 38% to 44% reduction in ZIKV susceptibility respectively. Individuals with low to no levels of DENV antibodies were more susceptible.

The researchers also investigated the relationship between a specific subset of antibodies, IgG3, against DENV and the risk of ZIKV infection. The researchers found that increased levels of IgG3 actually increased susceptibility to ZIKV infection, in contrast to total IgG. This positive association may be caused by an immune profile caused by recent DENV infection that makes individuals more susceptible to ZIKV infection.

To further investigate the association between DENV exposure and susceptibility to ZIKV infection, the researchers created a mathematical model including the variables DENV total IgG, DENV IgG3, and age. Analysis showed that DENV total IgG was negatively associated with ZIKV infection. It also showed a positive association between DENV IgG3 and ZIKV infection, consistent with results above. Age was not associated with risk of ZIKV infection.

Finally, the researchers investigated associations between previous DENV immunity and Zika symptoms. Among 642 individuals, 29, 206, and 11 reported experiences with skin rash, fever, or both, respectively. Pre-outbreak DENV total IgG was associated with a 47% reduction in the odds of having fever, but no reduction for rash. DENV IgG3 did not decrease odds of fever or rash.


The researchers found that individuals in Salvador, the epicenter of the Zika epidemic in Brazil, was heavily affected with an attack rate of 73%. The pattern of cases is consistent with the emergence of a highly transmissible and asymptomatic pathogen, followed by reduced transmission after populations gain immunity. Differential infection rates within Pau da Lima, Salvador shows that susceptible individuals who were not infected in the 2015 epidemic may remain. Although this population could sustain some future ZIKV transmission, the high rates of immunity will present a barrier to future ZIKV infections.