E. coli NIZO Human Challenge With Protein Drinks and Neutralised Gastric Acid
- Study type:
- Randomised, double‑blind nutritional intervention in healthy volunteers, all of whom were challenged orally with an E. coli strain after gastric acid neutralisation.
- Topic:
- Experimental E. coli gastroenteritis model in human volunteers
- Added:
- 2026-02-13
- Status:
- published
Strength of evidence
🔴 Very low – small, short-term human challenge experiment where all volunteers received E. coli after antacid; no control group without bacteria, heavy reliance on subjective symptoms, and non-natural exposure conditions.
Source
NIZO Food Research human challenge experiment with enteric E. coli and milk-protein drinks in healthy volunteers (E. coli challenge model used to test nutritional modulation of gastro-intestinal symptoms).
In this modern NIZO Food Research study, healthy adult volunteers were pre-treated with an antacid to neutralise gastric acid and then given an oral dose of a defined, attenuated but diarrhoeagenic E. coli strain. All participants received bacteria; they were randomised only with respect to the accompanying protein drink (e.g. different milk-protein formulations), not to infection vs placebo. A substantial proportion developed mild to moderate gastrointestinal symptoms (diarrhoea, cramps, nausea), and the study analysed whether the type of protein drink modulated symptom severity. There was no group that received the drinks without E. coli, and the use of antacid created non-physiological exposure conditions. From a WorthDay perspective, this study shows that when gastric acid is suppressed and volunteers ingest an E. coli challenge, symptoms can occur, and nutritional factors may influence how strongly they are felt. However, the design does not test natural infection risk under everyday conditions and provides only very low-quality evidence for real-world causality.
Key findings
- Healthy adults were recruited and screened before being enrolled in a controlled challenge protocol at NIZO Food Research.
- Participants received an antacid or bicarbonate pre-treatment to neutralise gastric acid, followed by an oral dose of a defined E. coli strain suspended in a liquid vehicle.
- All volunteers were exposed to E. coli; randomisation concerned only the type of accompanying protein drink (e.g. different milk-protein formulations) and not the presence or absence of bacteria.
- A notable proportion of participants developed gastrointestinal symptoms such as loose stools, abdominal cramps and nausea during the observation period after challenge.
- The primary analyses evaluated whether different protein drinks altered symptom scores and stool characteristics, not whether E. coli caused illness compared to no exposure.
- No control group received the same drinks and antacid without E. coli, so baseline rates of gastrointestinal symptoms under identical dietary conditions are unknown.
- Exposure conditions (antacid pre-treatment, bolus dose of bacteria) differ substantially from typical everyday food consumption with intact gastric acid.
Background
- The NIZO E. coli human challenge model was developed to study how nutritional interventions, such as different milk-protein formulations, might influence host responses to enteric pathogens. Instead of comparing infection vs no infection, the model assumes that all participants are exposed to a standardised E. coli challenge, and then tests whether particular drinks modify symptom severity.
Methods
- Healthy adult volunteers were enrolled under controlled conditions. Before challenge, they received an antacid or bicarbonate to reduce gastric acidity. They then ingested a defined dose of an attenuated, but still diarrhoeagenic E. coli strain in a liquid vehicle. All participants were randomised in double-blind fashion to receive different protein drinks (for example, various milk-protein or whey formulations). The primary endpoints were gastrointestinal symptom scores, stool frequency/consistency and some laboratory or microbiological markers over several days. There was no group that received the drinks without E. coli.
Results
- Following the E. coli challenge, a substantial number of volunteers reported gastrointestinal symptoms ranging from mild discomfort to moderate diarrhoea. The incidence and severity of these symptoms varied somewhat between the different protein‑drink groups, which the authors interpreted as evidence that nutritional composition might modulate host response to enteric pathogens. However, because all participants were exposed to E. coli and gastric acid was intentionally suppressed, the study cannot determine how many symptoms would have occurred without bacterial exposure, nor how gastric acid alone might have protected against illness.
Interpretation
- From the conventional perspective, this NIZO model is seen as a sophisticated way to study nutritional modulation of infectious disease. From the WorthDay perspective, S0039 demonstrates several key limitations: lack of a true control group without bacteria, deliberate disruption of a major natural defence (gastric acid), and reliance on subjective symptom scores over a short time frame. The study does show that under these artificial stress-test conditions, E. coli can trigger symptoms and that diet may influence them, but it offers only very low-quality evidence about infection risk from E. coli-contaminated foods under normal, real-world circumstances.
Limitations
- • No null control arm without bacteria: all volunteers received the E. coli challenge; there is no group with antacid + drinks but no bacteria to establish baseline symptom rates.
- • Artificial suppression of gastric acid: use of antacid/bicarbonate before challenge departs from normal physiology and likely increases susceptibility to ingested bacteria.
- • Short-term, soft endpoints: the main outcomes are self-reported gastrointestinal symptoms and stool characteristics over a few days, not hard clinical endpoints.
- • Limited sample size and specific population (healthy adults) restrict generalisability to children, elderly or people with comorbidities.
- • The study answers a narrow question about nutritional modulation under forced challenge, not the broader question of whether E. coli causes disease under everyday dietary exposure.
- • Lack of blinding with respect to infection status (everyone knows they were challenged) may influence symptom reporting and expectations, even if drink allocation is blinded.
WorthDay perspective
- For WorthDay, S0039 is a useful example of how modern human challenge models can look rigorous on the surface yet still rest on weak foundations for real-world causality claims. By neutralising gastric acid and exposing all volunteers to E. coli, the study creates a stress-test of host defences rather than a natural eating scenario, and it lacks a true no-bacteria control arm. It can therefore only provide very low-quality evidence about everyday 'food-borne infection', even if it says something about symptom modulation by diet.
- Together with S0038 (Salmonella Typhi challenge), S0039 helps users see recurring methodological patterns in infection research: antacid pre-treatment, universal exposure, no null controls, short-term soft endpoints. In the Health category, these entries support WorthDay's insistence on strict control logic and careful separation between experimental stress conditions and normal life when evaluating claims about pathogens and food safety.
Permalink: https://www.worth-day.com/science-compass/s0039-e-coli-nizo-human-challenge-with-protein-drinks-and
Related studies
Science meets practice