A thesis submitted to the Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Doctor of Philosophy Johannesburg 2017. / Introduction: Human rhinovirus (HRV) is the most prevalent virus detected in children with respiratory symptoms; however, its aetiological role during disease episodes remains unclear as detection of HRV is also ubiquitous among asymptomatic children. We evaluated the clinical epidemiology of HRV-associated disease among children hospitalised with severe and very severe pneumonia together with community control children living in Africa and Southeast Asia. In addition, we explored the associations between the molecular subtyping and nasopharyngeal viral loads of the HRV species and their ability to cause viraemia as potential markers for HRV disease. Methods: Using a case-control study conducted in seven countries, we compared the clinical characteristics of children (1-59 months of age) hospitalised with HRV-associated pneumonia between August 2011 - January 2014 and age-frequency matched controls. Nasopharyngeal swabs from the cases and controls were tested for HRV, together with 27 other respiratory pathogens, with quantitative real-time PCR assays. The 5’ NCR region of the HRV positive samples were sequenced to determine the species/strains of HRV and phylogenetic analysis was performed. Additionally, the blood samples from a limited number of cases (n=210) and controls (n=212) were tested for the presence of HRV viraemia and the 5’ NCR sequence of positive blood samples were further characterised. Results: Overall, HRV detection was 1.45-fold (aOR 95% CI: 1.29-1.62) higher among children hospitalised with pneumonia (24%) compared to controls (21%, P<0.005); including being 2.08-fold (28% vs 18%, aOR 95% CI: 1.75-2.47) more associated with case status among children 12-59 months of age. The HRV-associated cases were younger (13.1 months) than controls with HRV infection (15.4 months, P=0.001) and more likely to be malnourished (30% vs. 12%, P<0.001) and HIV-1 exposed (10% vs. 8%, P=0.046). HRV nasopharyngeal viral load was significantly higher among cases compared to controls (3.7 vs. 3.5 log10 copies/mL, P<0.001). Also, HRV viraemia was 7.02-fold (aOR 95% CI 1.70-28.94) more prevalent among cases (7%) compared to controls (2%, P=0.007). Moreover, HRV nasopharyngeal viral loads ≥4 log10 copies/mL differentiated between viraemia positive and negative cases. There was, however, no difference in the molecular subtyping of the HRV species prevalence among cases (HRV-A:48%; HRV-B:7%; HRV-C:45%) and controls (HRV-A:45%; HRV-B:10%; HRV-C:45%, P=0.496); as well as no evidence of seasonal or temporal clustering of the HRV species over time. Among cases, HRV detection was less likely to be associated with presence of radiographically confirmed pneumonia (40% vs 46%, P=0.001) or hospital stay >3 days (52% vs 61%, P=0.001). It was, however, positively associated with older age (13.1 months vs. 11.3 months, P<0.001) and presence of wheeze (46% vs. 31%, P<0.001) compared to the HRV uninfected cases. HRV was the sole virus detected in the 53% of cases and generally there were no differences in severity or clinical presentation among cases with HRV mono-infections compared to those with HRV-mixed infections. The HRV mono-infections, however, were associated with a 2.83-fold (aOR 95% CI: 1.44-5.53) higher case fatality ratio than cases with HRV and other viral mixed infections (10% vs. 5%, P=0.002). The HRV-associated case fatalities were more likely to have markers of bacterial co-infections compared to the HRV-associated cases that survived. Among the HRV species, HRV-C compared to HRV-A cases were older (12.1 vs. 9.4 months, P=0.033), more likely to present with wheeze (35% vs. 25%, P=0.031) and 2.59-fold (aOR 95% CI: 1.23-5.95) more likely to be associated with viraemia (12% vs. 2%, P=0.025). Conversely, the HRV-A infected cases were more likely to have radiographically confirmed pneumonia (46%) compared to HRV-C infected cases (36%, P=0.040) and HRV-A mono-infected cases were more likely to have hospital stay of >3 days (72%) than HRV-C mono-infected cases (54%, P=0.039). Conclusion: HRV detection, especially among children 1-5 years of age, was associated with severe lower respiratory tract infection; however, HRV detection was ubiquitous with a high degree of genetic diversity among both cases and controls. Thus the true etiologic role of HRV during childhood disease, especially among infants, remains uncertain. Nonetheless, HRV nasopharyngeal viral loads ≥4log10 copies/mL in conjunction with HRV viraemia are potential markers for HRV-associated severe respiratory disease. Among cases, HRV-A was associated with radiographically confirmed pneumonia and generally more severe disease than HRV-C which was more associated with viraemia and wheezing disease. / MT2017