Funded from 2009 – 2014, the Environmental Effects project studied how seasonality in incidence is a pervasive phenomenon among infectious diseases. The mechanisms responsible for this variation are known for very few systems. Possible mechanisms include variation in climate and its impact on pathogen persistence in the environment, variation in human contact rates, variation in vector or alternative host populations, and oscillations in host biological processes. In the project, we took several approaches to determine what mechanisms drive seasonality of transmission, and account for seasonality in models of spatial dynamics and control measures.
We used data from influenza and vector-borne diseases to estimate the seasonality of transmission using methods such as those previously described for measles and chickenpox. We compared the seasonality of influenza to these other respiratory diseases and to school terms and climatic variation. The project used meta-population models developed using data from the United States to investigate the optimal timing of both existing influenza vaccines and improved vaccines that might become available in future years.
The project also modeled mosquito-borne diseases, such as malaria, including demographic dynamics and epidemic dynamics. This spatial-temporal component was useful in collaborating with the behavioral modeling project. Dr. Cummings is now on the Data & Parameters team.
H1N1pdm in the Americas
This study characterizes the early spread of the H1N1 pandemic in the Americas. We found that the reproductive number (R) estimates of H1N1pdm were most associated with latitude; to the extent that latitude is a proxy for seasonal changes in climate and behavior, this association suggests a strong seasonal component to H1N1pdm transmission.
Serotype-Specific Differences in the Risk of Dengue Hemorrhagic Fever
It is not entirely understood why some individuals develop more severe disease than others when exposed to the dengue virus. This study found significant and non-significant correlations between dengue serotype 2 infection and more severe dengue disease. Additionally, individual serotypes varied in disease severity between study years, supporting the hypothesis that particular sequences of dengue virus infections may influence disease severity.
From Re-Emergence to Hyperendemicity: The Natural History of the Dengue Epidemic in Brazil
The re-introduction of Dengue virus into Brazil occurred in 1986, and by 2007, a shift in the age distribution of disease was reported; where previously, the majority of dengue hemorrhagic fever cases occurred among adults, in 2007, 53% of cases occurred in children under 15 years old. We used discrete-time simulation to estimate the accumulation of monotypic and multitypic immunity over time in a population previously completed susceptible to dengue. Simulations show that as time since re-emergence of dengue goes by, multitypic immunity accumulates in adults while an increasing proportion of susceptible individuals and those with monotypic immunity are among younger age groups. Assuming persons who are monotypically exposed are at highest risk for severe dengue disease, the shift observed in Brazil can be partially explained by the accumulation of multitypic immunity in older age groups, 22 years after the reintroduction of dengue.
Prediction of Dengue Incidence Using Search Query Surveillance
Internet search terms predict incidence and periods of large incidence of dengue with high accuracy and may prove useful in areas with underdeveloped surveillance systems.
Derek Cummings, PhD