Investigating temporal patterns in the force of infection of dengue virus transmission using longitudinal serologic data

Background/Question/Methods

Over 40% of the world’s population is currently at risk from dengue. Estimating the rate at which susceptible individuals become infected (i.e. the force of infection) is complicated by many factors. The percent of cases that are asymptomatic is difficult to estimate since these cases often occur without even the person infected being aware they’re infective. There are four distinct serotypes (DEN-1, DEN-2, DEN-3, and DEN-4) of the virus that, once a person is infected, provide lifelong immunity against that particular serotype but also influence the risk of developing severe dengue in subsequent infections by any of the other serotypes. Only through incorporating asymptomatic cases and utilizing knowledge of the serological history of each infected case can reliable estimates of the force of infection be calculated. In an attempt to estimate the force of infection of all four serotypes of dengue across time, longitudinal data was analysed from Iquitos, Peru from 1999-2010. Approximately 14000 individuals’ serotype-specific serostatus was measured several (2-14) times throughout the study, typically with an interval of 6 months between tests.

Results/Conclusions

To utilize the large but sparsely sampled data from this study, we developed a Markov Chain Monte-Carlo approach that fits a smoothed force of infection function through the use of B-splines. A preliminary analysis that treats the four serotypes as four independent viruses illuminates various patterns in the force of infection over time. DEN-1 and DEN-2 were introduced in Iquitos before the beginning of the study period, while DEN-3 and DEN-4 were introduced during the study period (~2000 and 2009 respectively). As such, the number of observed seroconversions was lower for DEN-1 and DEN-2. However, there are several times throughout the study where the force of infection for DEN-1 and DEN-2 are quite high and, at certain points, the force of infection jumps for all circulating serotypes synchronously. Two important examples of these synchronized jumps are during the introductions of both DEN-3 and DEN-4. While immunity clearly influences the number of infections, it appears that whenever the environment is ideal for dengue transmission, many or all actively circulating serotypes experience an increase in transmission. Further analysis of these data, as well as including the 2011 data which showed a deadly DEN-2 outbreak will hopefully illuminate more of the interactions between the four serotypes.