Meta-analysisindependent cross-checkfull text checked★ key study I-INTA-01 in-depth review
Does tea cause heartburn?
Association between tea consumption and gastroesophageal reflux disease: A meta-analysis.
Overview
Cao H et al.. Medicine 2019 · PMID 30681584 · DOI 10.1097/MD.0000000000014173
Across 30 studies, tea drinkers did not have significantly more reflux disease overall (OR 1.12, 0.98-1.27).
Does the study support the specific claim? Partly · Methodological quality: low
Plain-language summary
Study explained simply
- What did the researchers want to find out?
Link between tea and reflux disease
- How was the study done?
Meta-analysis of observational studies
- How many people, animals or samples were studied?
30 studies from 23 articles
- What was actually found?
No significant overall association; higher odds in East Asia, lower odds in the Middle East; cohort studies null
- How large was the observed effect?
OR 1.12 (0.98-1.27)
- What do the results mean in everyday life?
Tea is not a proven general cause of reflux
- What are the study's limitations?
Mostly cross-sectional, mostly unadjusted studies; high heterogeneity; no protocol; reporting inconsistencies
- What does it say about Elawady's specific claim?
Contradicts 'tea equals reflux' as a general rule
- Do the results apply to Germany and Europe?
Moderate
- What can explicitly not be concluded from this study?
Whether tea triggers symptoms in a particular person
Scientific in-depth review
1. Full reference
- Authors
- Cao H et al.
- Journal
- Medicine
- Year
- 2019
- Identifier
- PMID 30681584 DOI 10.1097/MD.0000000000014173
- Verification
- verified – PubMed efetch 2026-10-25
- Retracted
- no
- Corrections
Erratum published (PMID 30855514); erratum content not read.
- Extent checked
- full text
- Conflicts of interest and funding
Authors declare no competing interests; no funding statement in the full text.
2. Research question
Is tea consumption associated with gastro-oesophageal reflux disease (GERD)?
3. Study design and population
- Study design
- Meta-analysis – Systematic review and meta-analysis of observational studies
- Population
Participants of 23 observational studies (30 estimates) published 2004-2017: 17 estimates from East Asia, 5 Europe, 5 Middle East, 2 North America, 1 South Asia
- Country
Multinational (East Asia, Europe, Middle East, North America, South Asia)
- Period
PubMed and Embase searched to 20 December 2017 per Methods (abstract states March 2018)
- Sample size
23 articles, 30 estimates (18 cross-sectional, 9 case-control, 3 cohort)
- Inclusion and exclusion criteria
Included case-control, cohort or cross-sectional studies reporting OR/RR with 95% CI (highest vs lowest tea intake) for GERD, NERD, reflux/erosive oesophagitis or Barrett's oesophagus, with Newcastle-Ottawa score >=6. Excluded reviews, case reports, duplicates, insufficient data and non-pure tea (tea with salt or milk, tea combined with coffee). 7365 records, 98 full texts, 23 included.
4. Methods and endpoints
- Exposure / intervention
Tea drinking (highest vs lowest category; amounts, tea type, strength and temperature heterogeneous)
- Comparator
No or low tea drinking
- Endpoints
GERD and related diagnoses (symptom-based in 13 estimates, endoscopy in 17)
- Statistical methods
Pooled ORs; random effects if I2 >50%, otherwise fixed effects; Q and I2; meta-regression and subgroups (region, design, exposure assessment, diagnosis, symptoms); Egger and Begg tests; leave-one-out sensitivity.
5. Actual results
- Results
Overall OR 1.12 (0.98-1.27), I2 79.3%. East Asia OR 1.27 (1.07-1.51); Middle East OR 0.78 (0.63-0.93) in results (abstract: 'Middle Asia' 0.77, 0.63-0.95). Cross-sectional OR 1.18 (1.02-1.36); case-control 1.04 (0.74-1.46); cohort 1.02 (0.82-1.27). Interview-assessed 1.29 (1.02-1.62); questionnaire 1.07 (0.91-1.24). Symptom-only diagnosis 1.02 (0.87-1.20); endoscopy 1.19 (0.99-1.43). Asymptomatic subgroup OR 1.47 (1.11-1.93) (abstract). Region and symptoms explained about 25% and 15% of heterogeneity. Egger P=0.674; leave-one-out stable.
- Absolute risks
not reported in full text
- Authors' conclusion
No significant overall association between tea consumption and GERD; tea may increase risk in East Asia and decrease it in the Middle East ('Middle Asia'); limiting intake may be considered in some GERD patients; better-designed (prospective) studies needed.
6. Strengths
- First meta-analysis on tea and GERD; two databases
- Duplicate data extraction; quality threshold (NOS >=6)
- Subgroup and meta-regression analyses; publication bias and sensitivity analyses
7. Weaknesses
- Mostly cross-sectional and case-control data; only 3 cohort estimates (all null)
- Majority of included studies did not adjust for confounders (authors' statement)
- High heterogeneity (I2 79.3%); model choice driven by I2
- Heterogeneous exposure definitions (amount, tea type, strength, temperature)
- Multiple estimates from the same articles treated as separate studies
- Internal inconsistencies (search date; 'Middle Asia' vs Middle East figures); no protocol, no list of excluded studies, no funding statement
- Multiple subgroup tests without correction
8. Risk of bias
- Orienting tool
- AMSTAR 2
- 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
Strict AMSTAR 2: critically low confidence (critical items 2 and 7 not met, others partial) and primary studies largely cross-sectional and unadjusted. The overall null result is nonetheless consistent across designs (cohorts null). Simplified single-reviewer judgement from the full text; a formal AMSTAR 2 assessment needs two independent reviewers.
| Domain | Judgement | Rationale |
|---|---|---|
| Item 1 PICO | low | Exposure and outcomes defined. |
| Item 2 (critical) protocol | high | No protocol or registration reported. |
| Item 4 (critical) comprehensive search | some concerns | PubMed and Embase only; no grey literature or reference checking described; inconsistent search dates. |
| Items 5-6 duplicate selection and extraction | low | Two independent authors. |
| Item 7 (critical) excluded studies | high | Only a flow diagram; no list of excluded full texts. |
| Item 9 (critical) RoB of individual studies | some concerns | Newcastle-Ottawa scale used as inclusion threshold; no domain-level RoB reported. |
| Item 10 funding of included studies | high | Not reported. |
| Item 11 (critical) meta-analytic methods | some concerns | Pooling of mainly unadjusted, heterogeneous estimates; fixed/random choice based on I2; multiple estimates per article. |
| Item 13 (critical) RoB in interpretation | some concerns | Limitations discussed, but subgroup findings presented as possible effects. |
| Item 14 heterogeneity | low | Explored by meta-regression and subgroups. |
| Item 15 (critical) publication bias | low | Egger and Begg tests reported. |
| Item 16 review COI | some concerns | No competing interests declared; funding not reported. |
Full-text review (25 October 2026): Added study counts by design/region, inclusion criteria (NOS >=6; non-pure tea excluded), I2 79.3%, design/exposure/diagnosis subgroups (cohort OR 1.02), Egger P=0.674, reporting inconsistencies (search date; 'Middle Asia' vs Middle East), note that most primary studies were unadjusted. AMSTAR 2 itemised: bias overall raised from some_concerns to high (critically low confidence); methodological quality lowered to low.
9. Assessment of the statistical interpretation
Overall CI includes no effect; regional subgroups point in opposite directions (East Asia higher odds, Middle East lower odds); the only prospective (cohort) estimates are null (1.02, 0.82-1.27). Subgroup results are exploratory and subject to multiple testing.
10. Relevance to the specific claim
Tests the claim that tea causes acid reflux.
11. Independent contrary findings
Regional subgroup associations in East Asia.
12. Transferability to Germany and Europe
Moderate; 5 European estimates and 5 Middle Eastern estimates included (Middle East subgroup showed lower odds); no Egyptian data reported.
13. Overall methodological judgement
Low-certainty evidence; no consistent population-level association between tea and GERD. Does not support 'tea causes reflux' as a general rule; individual symptom triggers remain possible.
| Multidimensional assessment | |
|---|---|
| Authenticity of the source | verified |
| Methodological quality | low |
| Risk of bias | high |
| Precision of results | moderate |
| Directness | direct |
| Reproducibility | partly replicated |
| Consistency with the overall evidence | inconsistent |
| Regional transferability | moderate |
| Adequacy of the interpretation (by the archive) | not applicable |
14. Open questions
Effects of strength, temperature and amount of tea; Egyptian drinking habits.