Determinants of Campylobacter species diversity in infants and association with family members, livestock, and household environments in rural Eastern Ethiopia

Campylobacter infections represent a substantial global health concern. They have been identified as a leading cause of bacterial gastroenteritis, resulting in 166 million cases of diarrhea globally in 2020, of which 96 million were foodborne [1]. While most cases are self-limiting, the potential for severe, even life-threatening outcomes cannot be overlooked, particularly in infants, the elderly, and those with weakened immune systems [2].

Campylobacter infections in high- and middle-income countries are mainly foodborne [3], but other pathways are more important in settings with poor sanitation and hygiene infrastructure [4], contributing significantly to the disease spread. Beyond acute gastroenteritis, these infections have been linked to long-term gastrointestinal health issues, particularly in pediatric populations where the early-life gut microbiome is being established. This critical period is marked by rapid changes in the intestinal microbial community driven by diet and environment and generally stabilizes by three years of age [5]. Research from the Etiology, Risk Factors, and Interactions of Enteric Infections and Malnutrition and the Consequences for Child Health and Development (MAL-ED) project has associated Campylobacter infections with adverse health outcomes such as Environmental Enteric Dysfunction (EED) leading to compromised nutrient absorption, weakened immunity, and stunting [6‐8]. These early-life growth patterns are crucial indicators of nutritional status, with significant implications for future health outcomes, including mortality, chronic disease, neurodevelopment, and economic productivity in later life [7‐10].

Outbreaks of Campylobacter are rarely reported globally [11, 12]. Nevertheless, Africa is estimated to have the highest incidence of campylobacteriosis, particularly among children residing in both rural and urban areas in East Africa [13, 14]. In 2018, a formative cross-sectional study in rural eastern Ethiopia conducted under the CAGED project’s umbrella (Campylobacter Genomics and Environmental Enteric Dysfunction) unveiled that the prevalence of Campylobacter at the genus level was 88% in stool samples from children under two years of age [15]. Meta-total RNA sequencing revealed an average of 11 distinct Campylobacter species in positive stool samples from children, heightening concerns about their potential role in chronic outcomes such as EED and stunting [16].

Although by November 2024, 49 species have been validly published within the Campylobacter genus (https://lpsn.dsmz.de/genus/campylobacter), existing research disproportionately concentrates on thermophilic Campylobacter jejuni and Campylobacter coli [17]. C. jejuni is the dominant diarrhea-associated species worldwide, with C. coli contributing 1–25% to Campylobacter-related gastroenteritis cases [18, 19]. These species, while dominant, are not exclusive agents of Campylobacter-related gastroenteritis [20, 21]. Moreover, the MAL-ED study revealed that when detecting all Campylobacter species using immunoassays, the impact on stunting was greater than when looking at just C. jejuni and C. coli alone [7‐9].

Most data on Campylobacter epidemiology are from high-income countries and it remains understudied in low- and middle-income countries, particularly the non-thermophilic species abundant in vulnerable populations like Ethiopian children enrolled in the CAGED formative study [15]. However, advancements in isolation and detection techniques have shed light on other emerging Campylobacter species, such as C. upsaliensis, C. lari, and C. hyointestinalis, due to their increased association with human illness [22]. In addition, recent studies have described a new species Candidatus C. infans (C. infans from hereon), originally detected in infant stools from the Global Enteric Multicenter Study (GEMS) study[23]. This species was detected by shotgun metagenomic sequencing in 59.1% of fecal samples from diarrheal and asymptomatic children under two years of age in Iquitos, Loreto, Peru. These samples were previously tested positive for the Campylobacter genus but negative for C. jejuni/coli by PCR. C. infans was identified as the dominant species in breastfed infants and is associated with diarrhea in humans and non-human primates [22]. Furthermore, most studies have focused exclusively on human or animal hosts; however, the role of environmental factors, including soil and water, in Campylobacter transmission remains less understood.

Building on these insights, a longitudinal study was conducted from December 2020 to June 2022 aimed at assessing the prevalence, species composition, and genomic diversity of thermotolerant and non-thermotolerant Campylobacter spp. in infants, adults, livestock, and environmental reservoirs in the Haramaya woreda in Eastern Ethiopia. The details of the study design have been previously described [24, 25]. A high prevalence of Campylobacter at the genus level was observed, with 64% of stool samples from infants testing positive, with an age-dependent prevalence, as determined through TaqMan real-time PCR analysis [25]. This study seeks to elucidate the temporal colonization patterns of Campylobacter at the species level in infants and the potential reservoirs, [humans (siblings, mothers), livestock (cattle, sheep, goats, chickens), the environment (soil samples collected at the front and inside the home)], and household determinants contributing to their infection. Using species-specific quantitative PCR, we assessed the prevalence and load of both thermotolerant and non-thermotolerant Campylobacter at the species level across these various sample types and conducted risk-factor analysis using metadata obtained through extensive face-to-face interviews [26].

Based on a preliminary screening of 15 Campylobacter species (Table 1) using a subset of samples, we identified four species most detected in infant stool: C. infans (36%, n = 171/475), C. jejuni (14%, n = 66/475), C. upsaliensis (10%, n = 48/475), and C. lari (6%, n = 30/475) (Fig. S3). Subsequently, we targeted these four species in 2,045 samples across all sources.

In low- and middle-income countries, diarrheal diseases remain a leading cause of mortality in children under 5 years of age, with Campylobacter species emerging as significant pathogens associated with both acute illness and chronic health consequences, including malnutrition, stunting, and cognitive deficits [8, 27, 31]. The CAGED longitudinal study found Campylobacter is highly prevalent in rural households in eastern Ethiopia [21]. Here, we identified a high prevalence of C. infans and C. jejuni among infants residing in Haramaya, Ethiopia, using quantitative PCR (qPCR). Over 20% of infants were colonized with C. infans within the first month of life, and the prevalence plateaued at 61% by 11–12 months. In contrast, C. jejuni colonization was observed in 6% of infants during the first month, peaking at 53% by 10 months. C. upsaliensis colonization was observed later, emerging after the first two months of life, and remained relatively low (6–12%) throughout most of the first year. However, a notable increase to 28% at 11 months suggests that C. upsaliensis colonization may occur later in infancy. These findings align with research efforts to understand Campylobacter acquisition and clearance dynamics using the MAL-ED longitudinal data, which also observed increased Campylobacter acquisition rates in infants across multiple low- and middle-income countries during the first year of life [31]. Both C. infans and C. jejuni abundances showed positive correlations with infant age, ranging between 2 to 3 log₁₀ genome copies per 50 ng DNA, which is concerning as Campylobacter is highly infectious compared to other gastrointestinal pathogens, given that C. jejuni has a minimum infectious dose of 500 to 800 organisms [32, 33]. In addition, the co-occurrence of Campylobacter species increased as infants grew older, with infants between 298 and 375 days old having 31 times higher odds of co-occurrence detection. Overall, the prevalence and loads of infants at approximately one year of age were higher than those in siblings, which in turn were higher than in mothers. These differences are likely due to the development of immune systems and differences in environmental exposure through behaviors such as crawling and play and different hygiene practices. The development of adaptive immunity may play a crucial role in age-related susceptibility to Campylobacter [31]. Acute infections trigger the production of specific immunoglobulins (IgG, IgM, and IgA) in serum and intestinal secretions, with IgA and IgM antibodies particularly elevated during acute phases [34‐37]. This acquired immunity strengthens with repeated exposures throughout life [38], likely explaining the lower infection and co-occurrence rates observed in mothers compared to their children.

In our study, many Campylobacter-positive samples belonged to asymptomatic infants; however, infants colonized with C. infans had 2.0 times higher odds of experiencing enteric symptoms, particularly diarrhea, which was linked to higher C. infans load. These findings contrast with recent observations from other LMICs, particularly from Peru, where researchers reported low C. infans prevalence in infants under two years of age and found no association between C. infans colonization and diarrheal symptoms [39, 40]. While the Peru cohort was small, this discrepancy highlights the importance of conducting region-specific epidemiological studies. It suggests that the role C. infans in infant health may be more complex than initially understood, warranting further investigation into strain-specific characteristics and local determinants of pathogenicity. Additionally, we found that infants colonized with C. jejuni had 2.3 times higher odds of experiencing diarrhea, aligning with findings from other LMIC-oriented studies, including the GEMS and MAL-ED studies, which identified C. jejuni as a leading cause of bacterial diarrhea in LMICs during early childhood [7, 41, 42]. Despite the high occurrence of Campylobacter infection and EED, with more than half of infants having EED by one year of life, no association was found between C. infans, C. jejuni, or C. upsaliensis load and EED in this population. These results align with previous work assessing the relationship between the overall Campylobacter load and gut function parameters [26] and is related to a small sample size in our study. Importantly, we observed similar loads of Campylobacter at the genus level and the three dominant species in samples from diarrheal and non-diarrheal infants, supporting the finding from the MAL-ED study that asymptomatic infants are also at risk of developing chronic gut inflammation and, consequently, EED and stunting [43]. It is important to note, however, that the associations observed between Campylobacter loads, symptoms, and environmental exposures in our study are correlational. While these findings suggest potential links, the study design does not allow us to infer causality, and further research is necessary to determine whether these environmental factors directly influence infection rates.

Several household determinants were identified to impact Campylobacter species abundance. Household dietary practices significantly influenced Campylobacter abundance, with raw milk consumption strongly associated with higher C. infans loads. This association was also observed at the genus level, suggesting that introducing complementary foods, particularly animal-source foods, may be a critical source for microbial colonization [16, 27]. Infant sex also influenced Campylobacter colonization, as females had a higher average load of C. infans, suggesting potential gender-based differences in susceptibility and/or exposure. Additionally, crawling in areas with animal droppings was linked to increased C. infans load. While C. infans was not prevalent among the four livestock species tested in this study, its detection in surface soil samples around the household underscores the potential for complex environmental exposure pathways contributing to elevated C. infans levels in infants. While sheep demonstrated the highest C. infans prevalence among tested livestock species, closer nighttime proximity to sheep was associated with lower infant C. infans loads, matching findings from genus-level testing. Similarly, keeping sheep confined inside the home was associated with a reduced abundance of C. upsaliensis. This pattern may suggest that though sheep are reservoirs of Campylobacter, the relationship between animal proximity and human colonization is complex and requires further investigation. Given that these findings originate from rural households, where there is frequent close contact with livestock and environmental exposures, it would be beneficial to conduct similar studies in urban settings. Urban environments differ significantly in terms of infrastructure, sanitation, and patterns of animal-to-human contact, which can influence the prevalence and transmission pathways of Campylobacter species. Lastly, we found that infants who put soil or animal feces in their mouths were at risk for higher C. jejuni loads, a species predominant in zoonotic reservoirs. These findings emphasize the need for targeted interventions to address the complex interplay of environmental, dietary, and behavioral factors in Campylobacter transmission. Such interventions are urgently required to reduce infection rates and mitigate the burden of diarrheal diseases in resource-limited settings.

Building on our formative research, this longitudinal study provides deeper insights into the socio-demographic and exposure dynamics within the Haramaya woreda and may also be relevant for comparable smallholder environments. Our species-specific testing across human, livestock, and environmental samples revealed that three targeted Campylobacter species accounted for 54% of genus-level detections, suggesting the presence of additional, untested Campylobacter species and undetected species using the current method in these ecological niches. Indeed, shotgun metagenomic testing identified 21 dominant Campylobacter species across 280 samples (infants, mothers, siblings, cattle, goats, sheep, and chickens), revealing a diverse Campylobacter community in which many different species were found in ruminants that do not seem to be transmitted to humans. In infant stools, however, C. infans, C. jejuni, and C. upsaliensis collectively accounted for 75% of the Campylobacter genus-level signals, establishing these species as the predominant colonizers of the infant gut. This finding aligned with metagenomic data, where C. jejuni, C. infans, and C. upsaliensis were frequently detected alongside C. concisus [50]. In addition, 83 samples tested positive for one or more species, while the test at the genus level was negative. TaqMan was used for genus testing, whereas species testing was conducted with SYBR Green, and these observations may be partly explained by the different sensitivities of the two PCR methods. Further emphasizing these limitations, Parker et al. (2022) demonstrated that even validated primers for C. jejuni can yield false negatives. They found that a significant percentage of stool samples from children in Peru tested negative for C. jejuni using qPCR. Yet, the species was detected via whole-genome shotgun metagenomic sequencing performed on the same extracts. While we utilized validated primers for species testing, we acknowledge certain limitations in our approach. We observed some cross-reactivity in our primers, particularly for C. jejuni (with C. fetus, C. lari, and C. showae), C. upsaliensis (with C. helveticus), and C. lari (with C. helveticus and C. upsaliensis). C. infans primers showed minimal cross-reactivity with C. concisus at a high CT value, indicating low-level amplification. Future studies could benefit from a combined approach using both qPCR and metagenomic sequencing to address these limitations. While qPCR offers higher sensitivity and is more suitable for the absolute quantification of target genes, metagenomic sequencing provides broader coverage and an unbiased overview of microbial species within a sample [44]. This complementary approach would leverage the strengths of both methods: the high sensitivity and quantification capabilities of qPCR and the comprehensive detection and novel species identification potential of metagenomics [45]. By integrating these techniques, researchers can achieve a more robust surveillance strategy, gaining a deeper understanding of the microbial landscape while minimizing diagnostic gaps and potential biases introduced by primer-based methods.

This study contributes to the growing body of evidence elucidating the prevalence and diversity of Campylobacter species in human and animal populations. Our investigation revealed distinct distribution patterns: C. infans showed a high occurrence in human stool (cumulatively, 43% prevalence, p < 0.001 based on the chi-square test), which increased during the first year of life, and in surface soil samples, while C. jejuni demonstrated substantial presence across environmental and zoonotic sources, particularly in livestock. Consistent with previous research, our findings reveal that human stools and livestock feces frequently harbor multiple Campylobacter species. Our species-specific quantitative PCR results, further validated through shotgun metagenomics analyses in a subset of samples (n = 280, 40 per source), not only confirmed the presence of targeted Campylobacter species but also suggested the existence of additional species, particularly in ruminants [50]. These complementary approaches strengthen our confidence in the observed prevalence patterns and underscore the value of refined detection methodologies.

In this study, we investigated the prevalence and dynamics of Campylobacter species colonization among 106 households residing in Haramaya, Ethiopia, using species-specific quantitative PCR. We identified C. infans, C. jejuni, and C. upsaliensis as the predominant colonizers of the infant gut, with notable age-dependent patterns and associations with enteric symptoms. Notably, C. infans and C. jejuni were linked to higher odds of diarrhea and fever, highlighting their clinical significance. Our findings support the recognition of Candidatus C. infans as a novel species with a distinct ecology. In this setting, C. infans was more prevalent in humans than C. jejuni, suggesting primarily human-to-human transmission. The low occurrence in livestock implies potential reverse zoonotic transmission or passage of this bacteria through the animal gut after exposure from a highly contaminated environment. The co-presence of anthroponotic C. infans and zoonotic C. jejuni and C. upsaliensis suggests that control methods in low-resource settings should address both transmission cycles. The identified risk factors and complex interplay of environmental, dietary, and behavioral factors offer potential targets for interventions to reduce Campylobacter colonization and associated health impacts.