Cross-sectional studycited by the Tayebat archivefull text checked A-CHK-18 in-depth review

Pneumonia near Dutch poultry farms

Increased risk of pneumonia in residents living near poultry farms: does the upper respiratory tract microbiota play a role?

Overview

Smit LAM et al.. Pneumonia (Nathan) 2017 · PMID 28702305 · DOI 10.1186/s41479-017-0027-0

People living within about 1 km of poultry farms had 11% more pneumonia.

Does the study support the specific claim? Partly · Methodological quality: moderate

Original publication

Plain-language summary

Study explained simply

What did the researchers want to find out?: Is living near poultry farms associated with pneumonia and altered upper-airway microbiota?
How was the study done?: Spatial analysis of GP records plus microbiota analysis in hospitalised CAP patients
How many people, animals or samples were studied?: 92,548; 126; 92,548 GP patients; 126 hospitalised CAP patients
What was actually found?: About 1 extra pneumonia per 1,000 people per year for each poultry farm within about 1 km (background 9 per 1,000), roughly 11% more; throat-bacteria differences were borderline and not significant after correction for multiple tests.
How large was the observed effect?: Kernel model; PERMANOVA
What do the results mean in everyday life?: Farm emissions may affect neighbours.
What are the study's limitations?: No adjustment for smoking or other confounders in the main model; distance as exposure proxy; small microbiota study without healthy controls; single year.
What does it say about Elawady's specific claim?: Not about eating chicken.
Do the results apply to Germany and Europe?: High (EU).
What can explicitly not be concluded from this study?: Cannot show dietary risk or that farm emissions cause pneumonia.

Scientific in-depth review

1. Full reference

Authors: Smit LAM, Boender GJ, de Steenhuijsen Piters WAA, Hagenaars TJ, Huijskens EGW, Rossen JWA, Koopmans M, Nodelijk G, Sanders EAM, Yzermans J, Bogaert D, Heederik D
Journal: Pneumonia (Nathan)
Year: 2017
Identifier: PMID 28702305 DOI 10.1186/s41479-017-0027-0 · PMC5471663
Verification: verified – PubMed E-utilities efetch, 2026-10-25
Retracted: no
Corrections: none
Extent checked: full text
Conflicts of interest and funding: Funded by Dutch Ministries of Economic Affairs (and former Economic Affairs, Agriculture and Innovation) and of Health, Welfare and Sport, Wilhelmina Children's Hospital Fund, ZonMW and NWO-VENI. One author received vaccine-study grants from GSK and Pfizer (declared unrelated); others none.

2. Research question

Is living near poultry farms associated with pneumonia and altered upper-airway microbiota?

3. Study design and population

Study design: Cross-sectional study – Two cross-sectional analyses: (i) re-analysis of 2009 GP electronic records with a spatial kernel model (maximum likelihood) of pneumonia risk by distance to poultry farms; (ii) exploratory 16S-rRNA oropharyngeal microbiota comparison in hospitalised CAP patients by residential proximity (<1 km vs >=1 km)
Population: (i) 92,548 GP patients (70,142 adults 18-70 y, 22,406 children) in Noord-Brabant, a region with dense intensive livestock farming; (ii) 126 adults hospitalised with CAP near Tilburg (26 living <1 km, 100 >=1 km from a poultry farm), no antibiotics in prior 2 weeks, C. burnetii-negative
Country: Netherlands (Noord-Brabant)
Period: GP data year 2009; hospitalised CAP patients November 2008 to February 2009
Sample size: 92,548 GP patients (993 GP-diagnosed CAP cases in 2009); 126 hospitalised CAP patients
Inclusion and exclusion criteria: GP analysis: all registered patients of participating practices in 2009 with geocoded address. Microbiota analysis: hospitalised CAP patients with microbiota data and address; excluded antibiotic use within 2 weeks and positive C. burnetii qPCR.

4. Methods and endpoints

Exposure / intervention: Residential proximity to poultry farms
Comparator: Living farther away
Endpoints: CAP diagnosis; oropharyngeal microbiota
Statistical methods: Spatial kernel model (3 parameters plus constant background risk) by maximum likelihood, compared with distance-independent model by likelihood-ratio test; kernel did not include individual covariates. Microbiota: Bray-Curtis nMDS, PERMANOVA, Mann-Whitney with Benjamini-Hochberg FDR, hierarchical clustering with Fisher's exact test; sensitivity analysis at 1.15 km.

5. Actual results

Results: Excess CAP risk ~0.001 per person-year per poultry farm within ~1.15 km vs background 0.009 per person-year (~11% increase); distance-dependent model improved fit (LRT p=0.0067); similar in adults and children. Microbiota: overall composition borderline different (PERMANOVA p=0.075); S. pneumoniae relative abundance 34.9% vs 22.5% (p=0.058, FDR q=0.541); Lactobacillus 1.4% vs 3.8% (p=0.049, q=0.541); S. pneumoniae cluster over-represented near farms (12/36 vs 14/90, p=0.049). At the 1.15 km cut-off differences attenuated (PERMANOVA p=0.23; S. pneumoniae p=0.14).
Absolute risks: Excess about 1 GP-diagnosed pneumonia per 1,000 person-years per poultry farm within 1.15 km, against a background of 9 per 1,000 person-years.
Authors' conclusion: Living near poultry farms associated with 11% increased CAP risk; replication needed.

6. Strengths

Large population-based GP dataset with geocoded addresses
Flexible spatial kernel replicating a prior finding
Microbiota groups similar in age, smoking, COPD and severity

7. Weaknesses

Kernel model contains no individual-level confounder adjustment (e.g. smoking, SES, age structure)
Exposure is residential distance only; no measured PM/endotoxin; misclassification from activity patterns acknowledged
GP-diagnosed pneumonia (not radiologically confirmed) as outcome
Microbiota sub-study small (26 exposed), no healthy controls, results not significant after FDR correction and attenuated in sensitivity analysis
Cross-sectional; single year; text and table inconsistently label group sizes (100/26)
Re-analysis of the same 2009 data that generated the hypothesis (not independent replication)

8. Risk of bias

Orienting tool: ROBINS-E
Type of assessment: structured assessment based on the full text by a single reviewer (no formal double review)
Full formal assessment: no
Overall risk of bias: high
Rationale: Raised from some_concerns (abstract-based) to high because the main kernel analysis lacks confounder adjustment and the microbiota findings do not survive multiple-testing correction. Single-reviewer structured judgement; a formal ROBINS-E needs two independent reviewers and a protocol.

8. Risk of bias
Domain	Judgement	Rationale
Confounding	high	Spatial kernel model has no individual covariates; smoking/SES not controlled in the main analysis. Microbiota groups balanced on measured factors but tiny.
Exposure measurement	some concerns	Residential distance as proxy for emissions; misclassification acknowledged.
Selection of participants	some concerns	GP practices in one region; microbiota study restricted to hospitalised patients without healthy controls.
Post-exposure interventions	not applicable	Not relevant to residential exposure.
Missing data	some concerns	Microbiota analysis limited to patients with available data and addresses; extent not reported.
Outcome measurement	some concerns	GP-coded pneumonia without radiological confirmation; microbiota by validated 16S sequencing.
Selection of reported result	some concerns	Multiple microbiota tests; headline S. pneumoniae difference not significant after FDR; 1 km cut-off from prior study, sensitivity analysis weaker.

Full-text review (25 October 2026): Added absolute excess (0.001 vs 0.009 per person-year), LRT p=0.0067, 993 cases, unadjusted kernel model, FDR q=0.541, attenuation at 1.15 km, funding; ROBINS-E itemised, overall some_concerns -> high.

9. Assessment of the statistical interpretation

The ~11% excess is modest and model-based without confounder control; the microbiota signal is hypothesis-generating only (q=0.541).

10. Relevance to the specific claim

C-CHK-13.

11. Independent contrary findings

Not dietary.

12. Transferability to Germany and Europe

High (EU).

13. Overall methodological judgement

Suggestive environmental-health evidence that living within ~1 km of poultry farms in an intensive livestock region is associated with modestly more GP-diagnosed pneumonia; mechanism unproven. It concerns farm emissions for neighbours, not eating chicken.

Multidimensional assessment
Multidimensional assessment
Authenticity of the source	verified
Methodological quality	moderate
Risk of bias	high
Precision of results	low
Directness	very indirect
Reproducibility	partly replicated
Consistency with the overall evidence	unknown
Regional transferability	high
Adequacy of the interpretation (by the archive)	not applicable

14. Open questions

Independent replication with confounder adjustment and measured exposures; role of PM/endotoxin vs zoonotic agents.

15. Linked fact checks

Poultry production is a source of zoonotic avian influenza and respiratory risks for people (H7N9 from wet-market poultry; pneumonia near poultry farms). – Studies supporting the claim

Use by the Tayebat archive

Note: according to the archive, the studies were compiled by the archive operators. Whether Dr. Elawady himself cited this study is not established.

What the archive says about the study: Listed in bibliography (no interpretation).
A. What was actually studied?: Pneumonia near poultry farms.
B. What did the authors find?: About 11% higher GP-diagnosed pneumonia risk within ~1.15 km of poultry farms (unadjusted spatial model); borderline, non-FDR-significant microbiota differences.
C. What conclusions do the authors themselves draw?: Replication needed.
D. What does the archive derive from it?: Implicitly a hazard of poultry industry.
E. Is this derivation scientifically justified?: Valid for farm neighbours, not for eating chicken.
Adequacy of the interpretation (by the archive): not applicable