Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Royaume-Uni → 2014 → Exploring vulnerability to infectious disease in a smallholder farming community in rural western Kenya

University of Edinburgh (2014)

Exploring vulnerability to infectious disease in a smallholder farming community in rural western Kenya

Glanville, W.A. de

Titre : Exploring vulnerability to infectious disease in a smallholder farming community in rural western Kenya

Auteur : Glanville, W.A. de

Université de soutenance : University of Edinburgh.

Grade : Doctor of Philosophy (PhD) 2014

Résumé partiel
More than 2 billion people live on less than 2 US dollars per day. People in these conditions often have inadequate access to basic sanitation, safe water, and medical services. These individuals, households and communities may be at high risk for a wide range of preventable and treatable infectious diseases. The aims of this study were to : 1) describe the individual- and household-level prevalence of a number of endemic helminth, protozoal, bacterial and viral infections of people in a small-holder farming community in western Kenya ; 2) explore the spatial distribution of risk of infection for this wide range of pathogens ; 3) quantify the association between social and environmental conditions and individualand household-level risk of infection ; 4) identify shared risk factors for household-level infection. All data were collected between July 2010 and July 2012 as part of a cross-sectional survey of 416 households containing 2113 people. This sample was considered to be representative of a population of 1.4 million people living in a rural, mixed farming area of western Kenya that is characterised by high levels of poverty. Sampled individuals were tested for exposure to, or infection with, at least 21 infectious agents using a range of faecal, blood and serological tests. Extensive questionnaire-based data were also collected at both the individual and household level. The spatial distribution of infectious disease risk was assessed using a kernel smoothing approach, with spatial congruence in areas of elevated risk for multiple pathogens examined as a potential indicator of shared contextual or compositional effects. Individual- and household-level risk factors for infection with a range of prevalent pathogens were explored using multilevel logistic regression, with a particular focus on examining the impact of socioeconomic position (SEP). A hierarchical multispecies zero-inflated binomial (ZIB) regression was used to derive an estimate of household ‘species richness’ for six neglected helminth infections (viz. Ascaris lumbricoides, hookworm, Trichuris trichiura, Schistosoma mansoni, Taenia solium, and Strongyloides stercoralis) with correction for detection error. This modelling framework also allowed assessment of the relationship between household-level infection with each parasite and a range of social and environmental conditions, and, uniquely for a single study setting, the average response of the ‘group’ of parasites to these conditions. This study found very high levels of parasitism in the community, particularly with hookworm (individual survey-adjusted prevalence : 36.3% (95% CI 32.8 – 39.9)), Entamoeba histolytica/dispar (30.1% (27.5 – 32.8)), Plasmodium falciparum (29.4% (95% CI 26.8 – 32.0)), and Taenia spp. (19.7% (16.7 – 22.7)). Some degree of within-household clustering of infection was found for all pathogens under study, and this was particularly large for the helminth species (intra-cluster correlation coefficient (ICC) for A. lumbricoides = 61.8% ; T. trichiura = 54.1% ; Taenia spp. = 48.0% ; Hookworm = 35.3%) and HIV (27.2%). Virtually all of the (map-able) pathogens showed spatial heterogeneity in disease risk, with evidence of spatial clustering in household-level infection for most infectious agents, notably HIV, S. mansoni, A. lumbricoides, T. trichiura, P. falciparum and hookworm. There was substantial overlap in spatial clusters for several infections, with some evidence of a geographic gradient in risk for multiple pathogens in the study area.



Version intégrale (48,5 Mb)

Page publiée le 9 janvier 2018, mise à jour le 14 novembre 2018