Exploring the risk factors and genomic characteristics of Campylobacter in Australia and the region

Abstract

Campylobacter infections are a significant cause of foodborne disease globally. In high-resource settings, disease burden is primarily associated with lost productivity due to initial gastrointestinal illness and costs associated with post-infectious sequelae. In low-resource settings, endemic infections are common and are a potential contributor to malnutrition. This thesis contributes to understanding the epidemiology and diversity of campylobacteriosis in Australia and neighbouring countries by examining risk factors, disease burden, and genomic factors. I analysed case-control study data from Australia using multivariable logistic regression to identify risk factors for C. jejuni and C. coli enteritis. I found that people eating chicken, in particular undercooked chicken, and taking proton-pump inhibitors had an increased odds of illness. Other significant risk factors included contact with puppies, contact with chicken faeces and eating delicatessen products. I estimated the burden and costs attributed to the main risk factors for campylobacteriosis in Australia. These costs included healthcare, lost productivity, premature mortality, and willingness to pay to avoid the pain and suffering associated with illness. I found that the attributable risk of campylobacteriosis from eating chicken meat was approximately 30%, costing an estimated AUD 160 million per year. Excess campylobacteriosis from the use of proton-pump inhibitors cost an estimated AUD 65 million per year, with these costs predominantly from disease in adults over 65 years of age. I conducted two studies examining Campylobacter spp. genomics. First, I explored characteristics of illness in Australian cases, classifying infection severity according to a modified Vesikari scale, and conducting random forest analysis to identify genomic characteristics that predicted severe disease. I found that older cases reported milder symptoms but had a higher odds of hospitalisation. Key virulence genes were highly conserved across isolates and did not predict severe disease. However, some genes (e.g., fliK, Cj1136, and Cj1138) appeared to distinguish human C. jejuni cases from isolates from food sources. Host factors including age and comorbidities were more likely to predict severity of illness than virulence characteristics. Second, I conducted a comparative genomic analysis of Australian and New Zealand human and poultry isolates to explore diversity and characteristics within the region. Both C. coli and C. jejuni sequence types were highly diverse within each country and were generally unique to each country. This reflects poultry industries that were specific to each country and did not import or export product. Closely related isolates between countries could indicate similar food sources or travel. I analysed longitudinal birth cohort data and hospital-based surveillance data from infants and children in Timor-Leste to understand risk factors in the home environment and the prevalence of Campylobacter spp. and other enteric pathogens. Infants from a few months of age showed signs of malnutrition, particularly wasting and underweight. Infants and children had enteric pathogens including diarrhoeagenic Escherichia coli and Campylobacter spp. detected in their stool. Infants who were exclusively bottle fed had an association with enteric infection. This thesis provides evidence for the main risk factors for campylobacteriosis in Australia, how factors can differ between resource settings, and the genetic diversity of the pathogen within the region. It highlights the advantages of whole genome sequencing data for surveillance and demonstrates that multisectoral research collaboration is essential to understand Campylobacter and create effective interventions to reduce disease burden. Show more