The investigators conducted paired longitudinal household cohort studies in urban Bangladesh and rural Tanzania, enrolling 100 households per site. Each enrolled household included a child under one year of age. Households were followed for one year to capture enteric infection dynamics in young children and their household contacts.
Stool specimens were collected monthly from household participants and additionally during reported diarrheal episodes over the 12-month follow-up. From these serial samples the team estimated daily infection trajectories for each participant for two bacterial enteric pathogens: Shigella and Campylobacter. These longitudinal detection data provided the basis for modeling transmission processes.
Using the reconstructed infection trajectories, the authors fit a household transmission model to partition infections into those arising from internal household sources (within-household transmission) versus external sources introduced from the wider community (community-to-household transmission). Estimates of transmission rates were obtained by pathogen, by age group, and by study site. Forward selection was used to identify interactions in the model, including possible effect modification by age.
Across both cohorts, a minority of detected infections were symptomatic: 8.8% of Campylobacter infections and 7.6% of Shigella infections presented with symptoms. Incidence differed between sites: the incidence of Shigella was 2.3 times higher in Bangladesh compared with Tanzania, while Campylobacter incidence was 1.4 times higher in Bangladesh compared with Tanzania. Initial crude associations suggested that Campylobacter transmission had a relatively larger community-to-household component compared with Shigella.
When interactions with age were considered, the apparent pathogen-specific differences were largely explained by age-related effect modification. Children under 5 years displayed higher transmission rates for Campylobacter than for Shigella, and for both pathogens this younger age group exhibited higher community-to-household than within-household transmission rates. In contrast, participants aged 5 years and older had higher transmission estimates for Shigella compared with Campylobacter. In other words, the relative importance of community versus household transmission varied by pathogen and by age group, with younger children playing a distinct role for Campylobacter transmission.
The authors interpret their findings to mean that what might appear as intrinsic pathogen-specific transmission differences are in fact driven by age as an effect modifier. Therefore, designing interventions to reduce enteric bacterial transmission should account for age-specific patterns in addition to biological differences between pathogens. Targeting community-level exposures that affect children under five may be especially important for interrupting Campylobacter transmission, whereas strategies addressing older children and adults may be more relevant for Shigella, based on these cohort data.
Ethical approvals for the study were obtained from the Institutional Review Board of the University of Virginia, the National Institute for Medical Research in Tanzania, and the Ethical Review Committee and Research Review Committee of the International Centre for Diarrhoeal Disease Research, Bangladesh. The authors state that all participant consents and appropriate oversight were secured. Data generated in the study are available upon reasonable request to the authors. Funding declarations in the source list National Institutes of Health grants (R01 AI153254, R01 TW012183).
This report is a medRxiv preprint and has not undergone peer review; the authors note this status. The source abstract and information provided do not detail all modeling assumptions, specific numeric transmission rates, or full subgroup sample sizes beyond the cohort enrollments and summary incidence ratios. Those details are available in the full preprint and supplementary materials referenced by the authors. As with any observational cohort and modeled transmission estimates, unmeasured confounding and detection limits of stool testing can affect inferred transmission partitions.
Paired household cohorts in two epidemiologic settings showed low symptomatic proportions for both pathogens and higher pathogen incidence in Bangladesh versus Tanzania. Apparent differences in transmission routes between Shigella and Campylobacter were explained by age interactions: children under five had higher community-linked transmission for Campylobacter, whereas older individuals contributed more to Shigella transmission. Intervention planning should therefore integrate age-specific transmission patterns and pathogen biology when prioritizing control measures.