Understanding the carriage and transmission of non-typhoidal Salmonella infections in Kenya.

Non-typhoidal Salmonellae (NTS) are responsible for enteric disease characterized by self- limiting enterocolitis in most parts of the world. In Africa, however, they cause severe invasive disease and a high case fatality ratio. The serotypes responsible for >80% of invasive NTS disease (iNTS) are S. Typhimurium and S. Enteritidis. Control efforts against these NTS serotypes are limited by a lack of understanding of their transmission - both mode and rates of transmission. This PhD was designed to improve the understanding of the epidemiology of carriage of NTS and its rate of transmission to enable efficient design of effective control strategies, including vaccination.

Through a cross-sectional study of faecal carriage and seroprevalence in 3 locations in Kenya with varying incidence of iNTS, I observed that Kilifi, had the highest carriage prevalence despite the lowest incidence of disease. However, the majority serotypes were neither S. Typhimurium nor S. Enteritidis. At all sites, older children and adults had the highest carriage. Analysis of anti-Salmonella antibodies showed a decay of maternal antibodies in 4-5 months and a rapid acquisition of NTS infections among infants thereafter. A catalytic model of the seroprevalence data estimated the force of infection (FOI) with Group O:4,5 serotypes and Group O:9 serotypes as ranging from 0.2 to 0.5 episodes/person/year- both were highest in Kilifi. I carried out an additional longitudinal seroprevalence study retrospectively on archived samples in Kilifi. I observed that the FOI of Group O:9 serotypes had decreased over time concurrently with a documented decrease in iNTS incidence while the FOI of Group O:4,5 increased as iNTS incidence decreased in one of the locations. These suggest that interrupting transmission can be a strategy for control of iNTS caused by these serotypes, possibly through vaccines which reduce faecal carriage. Such measures can be focused on older children and adults who are the reservoirs for infection, especially in households with infants who have the highest incidence of infection and are at highest risk of invasive disease. For control of S. Typhimurium disease, host risk factors will need to be addressed. The parameters we have estimated such as FOI could be used to inform dynamic transmission models to predict vaccine effectiveness and cost-effectiveness analyses of different potential vaccination strategies including vaccination of pregnant mothers, infants, or older children.