The age-specific concentrations and seroprevalence of antibodies against Salmonella Enteritidis (O:9) and Salmonella Typhimurium (O:4,5) across three sites in Kenya

BackgroundSeroprevalence studies can indicate the age-distribution of first infection with non-typhoidal Salmonella (NTS) and estimate the rate of infection by age. This can help target interventions, particularly vaccination. MethodWe collected 1254 paired samples of serum and stool from healthy children and adults (aged 0-82 years) in Kilifi, Nairobi, and Siaya counties in Kenya, areas of low, medium and high incidence of invasive NTS disease (iNTS), respectively. We quantified the age and site-specific geometric mean concentrations (GMCs) of IgG and IgA antibodies against S. Enteritidis O-antigen (O:9, serogroup D) and S. Typhimurium O-antigen (O:4,5, serogroup B) using an in-house standardised ELISA. A previously calibrated reference serum was used as an internal control. The threshold for seropositivity was determined using mixture modelling. Stool samples were cultured for NTS; isolates were serotyped using the Kauffman-White scheme. ResultsMaternally derived O:9 IgG and O:4,5 IgG antibodies were detectable in 100% of neonates and the GMC decreased by 40% (95%CI [25%-52%]) per month in the first six months of life. GMCs of IgA were low in neonates. After age 6 months, the O:9 and O:4,5 IgG and IgA GMCs increased sharply with age across all sites, reaching a plateau in early adulthood. The rate of increase in IgG GMCs by age was highest for O:9 in Nairobi, and for O:4,5 in Kilifi. Mixture modelling defined a threshold of 14.1 AU for O:9 IgG (87% sensitivity and 87% specificity) and 28.2 AU for O:4,5 IgG (89% sensitivity and 66% specificity). Seroprevalence also increased by age. The GMCs of O:4,5 IgG were 2 times higher for carriers of serogroup B Salmonella than non-carriers (GMC Ratio 2.4 95%CI [1.1-5.4]). DiscussionMaternal antibodies to non-typhoidal Salmonella decay rapidly from 0-5months after which incident infection with both serogroups occurs. Therefore, control efforts should be implemented in early infancy before primary infection. Author SummaryWe have used serology to describe the epidemiology of NTS infection and rates of infection across three sites in Kenya with varying incidence of iNTS disease. Using a population-based, cross-sectional study, we have demonstrated that protective maternal antibodies are present at birth but decay within first 6 months of life after which antibody concentrations rise quickly with age implying a high rate of primary infections; therefore, infection control strategies need to be implemented in early infancy, and where vaccination is considered, the optimal age window for vaccination should consider the presence of maternal antibodies. In addition, we have highlighted areas that would benefit from further scientific research, including assay standardization and the thresholds for correlates of protection against infection and against invasive disease.