Abstract
Epidemiological data regarding coffee and tea consumption and risk of esophageal inflammation, Barrett’s esophagus (BE), and adenocarcinoma are sparse and inconclusive. This study examined the association between consumption of tea or coffee with risk of BE. We conducted a cross-sectional study among United States veterans, comparing 310 patients with histologically confirmed BE with 1728 individuals with no endoscopic or histopathologic features of BE (controls). Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated using logistic regression models. In univariate models, we found a statistically significant association between risk of BE and consumption of coffee (OR, 1.41; 95% CI, 1.06–1.87) or tea (OR, 1.34; 95% CI, 1.05–1.71). However, in multivariate analysis, in which models were adjusted for confounders including sex and race, we found no association between risk of BE and consumption of coffee (adjusted OR, 1.04; 95% CI, 0.76–1.42) or tea (adjusted OR, 1.11; 95% CI, 0.85–1.44). These data do not support an association between consumption of coffee or tea and risk of BE. It is unlikely that avoidance of coffee or tea will protect against BE.
Keywords: risk factors, epidemiology, EAC, caffeine, decaffeinated
Barrett’s esophagus (BE) is the precursor lesion for esophageal adenocarcinoma (EAC), a rapidly increasing cancer.1 BE affects up to 15% of persons with frequent symptoms of gastroesophageal reflux disease (GERD) and 1%-2% of the general adult population.2 In patients with nondysplastic BE, the annual risk of EAC is 0.12 to 0.40%.3
Caffeine from coffee and non-coffee beverages induces gastric acid secretion, decreases lower esophageal sphincter (LES) pressure and promotes reflux.4,5 Data regarding associations between coffee, tea and EAC risk are mixed. Recently, a large, prospective study of >400,000 participants (142 EAC during follow up) reported no association between coffee and tea consumption and EAC.6 A previous prospective study found an inverse association with coffee consumption.7 No studies have examined the association with BE. Hence, we examined the associations between caffeinated and decaffeinated coffee and tea consumption and risk of BE. We hypothesize that consumption increases BE risk.
Methods
We performed a cross-sectional study to examine risk factors for BE. Patients scheduled for an elective esophagogastroduodenoscopy for upper gastrointestinal symptoms (40-80 years) or screening colonoscopy in seven primary care clinics (50-80 years) were recruited between February 15, 2008 and August 20, 2013. Primary care patients underwent the study endoscopy as an additional procedure at the same time as their scheduled colonoscopy. We excluded patients with prior gastroesophageal surgery or cancer; history of lung, colon or breast cancer; contraindication for biopsy; major liver disease (ascites or varices); or cognitive impairment. At least one biopsy was taken and BE diagnosed by intestinal columnar epithelium with goblet cells. Patients with endoscopically suspected
BE but without histologic evidence on biopsy were not included in this analysis
Prior to the study endoscopy, participants, with help from a trained research assistant, completed a computerized survey detailing their coffee and tea consumption (Supplementary Appendix). We calculated odds ratios (OR) and 95% confidence intervals (CIs) using unconditional logistic regression models. The adjusted model included terms for age, sex, race, waist-to-hip ratio, duration of GERD symptoms, smoking status, alcohol use, Helicobacter pylori infection, use of aspirin/NSAIDs, use of acid suppressants, and recruitment source. Statistical analyses were completed using SAS 9.4 (SAS Institute, Cary, NC). Tests were 2-sided, with statistical significance determined at α=.05.
Results
This study involved 310 cases and 1728 controls (466 primary care and 1262 endoscopy) (Table 1). Compared to controls, cases were older and more likely to be male, white, have abdominal obesity and GERD symptoms.
Table 1.
Characteristics of controls and BE cases
| Controls (n=1728) |
BE cases (n=310) |
|||
|---|---|---|---|---|
| Variable | n (%) | n (%) | P valuea | |
| Age at time of study endoscopy, y |
||||
| Mean (SD) | 60.0 (9.2) | 61.6 (7.6) | .002 | |
| Sex | <.001 | |||
| Male | 1567 (90.7) | 302 (97.4) | ||
| Female | 161 (6.3) | 8 (2.6) | ||
| Race | <.001 | |||
| White | 1058 (61.2) | 273 (88.1) | ||
| Black | 617 (35.7) | 32 (10.3) | ||
| Other | 53 (3.1) | 5 (1.6) | ||
| Waist-to-hip ratio | .001 | |||
| Low (<0.85 or 0.9) | 271 (16.1) | 26 (8.4) | ||
| High (>0.85 or 0.9) | 1412 (83.9) | 282 (91.6) | ||
| Missinga | 45 | 2 | ||
| GERD symptoms | <.001 | |||
| No symptoms | 914 (53.0) | 116 (37.4) | ||
| 1-4 y | 27 (1.6) | 6 (1.9) | ||
| 5-9 y | 55 (3.2) | 7 (2.3) | ||
| ≥10 y | 729 (42.3) | 181 (58.4) | ||
| Missinga | 3 | 0 | ||
| PPI/H2RA use | <.001 | |||
| No | 751 (43.6) | 79 (25.6) | ||
| Yes | 970 (56.4) | 229 (74.4) | ||
| Missinga | 7 | 2 | ||
| Aspirin/NSAID use | .38 | |||
| None | 1570 (91.3) | 282 (92.8) | ||
| Less than daily | 9 (0.5) | 0 (0.0) | ||
| At least daily | 140 (8.1) | 22 (7.2) | ||
| Missinga | 9 | 6 | ||
| H. pylori positive | 560 (32.4) | 62 (20.0) | <.001 | |
| Smoking status | .27 | |||
| Never | 479 (27.8) | 73 (23.5) | ||
| Former | 738 (42.8) | 137 (44.2) | ||
| Current | 506 (29.4) | 100 (32.3) | ||
| Missinga | 5 | 0 | ||
| Alcohol drinking status | .69 | |||
| Never drank | 150 (8.7) | 23 (7.4) | ||
| Former drinker | 665 (38.6) | 117 (37.9) | ||
| Current drinker | 907 (52.7) | 169 (54.7) | ||
| Missinga | 6 | 1 | ||
Missing values were not included in the analyses.
Cases were more likely than controls to drink coffee (77% vs 70%) and tea (58% vs 51%) (Table 2). In the unadjusted analysis, coffee (OR, 1.41; 95% CI, 1.06–1.87) and tea (OR, 1.34; 95% CI, 1.05–1.71) were statistically significantly associated with BE. BE risk increased linearly with increased frequency of caffeinated coffee (P-trend = .002) and risk was highest for those drinking strong (OR, 1.69; 95% CI, 1.19–2.40) and hot or extremely hot coffee (OR, 1.47; 95% CI, 1.10–1.96). Risk of BE was associated with cold tea (OR, 1.45; 95% CI, 1.12–1.86), but not warm or hot or extremely hot tea.
Table 2.
ORs and 95% CIs for associations between coffee and tea consumption and BE
| Controls (n=1728) N |
BE (n=310) N |
Crude OR (95% CI) |
Adjusted ORa (95% CI) |
||
|---|---|---|---|---|---|
| Non-coffee drinker | 510 | 71 | Referent | Referent | |
| Ever coffee drinkers | 1218 | 239 | 1.41 (1.06-1.87) | 1.04 (0.76-1.42) | |
| Caffeinated (drinks/day) |
|||||
| 0-<1 | 88 | 14 | 1.14 (0.62-2.12) | 1.19 (0.62-2.28) | |
| 1-<2 | 280 | 33 | 0.85 (0.55-1.31) | 0.73 (0.46-1.17) | |
| 2-<3 | 327 | 66 | 1.45 (1.01-2.08) | 1.07 (0.72-1.59) | |
| 3-<4 | 219 | 54 | 1.77 (1.20-2.61) | 1.11 (0.72-1.70) | |
| ≥4 | 130 | 38 | 2.10 (1.35-3.26) | 1.22 (0.76-1.98) | |
| P-trend in coffee drinkers | 0.002 | 0.30 | |||
| Decaffeinated (drinks/day) |
|||||
| 0-<1 | 27 | 3 | 0.80 (0.24-2.70) | 0.74 (0.21-2.65) | |
| ≥1 | 85 | 16 | 1.35 (0.75-2.44) | 1.20 (0.63-2.30) | |
| Coffee strength | |||||
| Weak | 53 | 8 | 1.08 (0.50-2.37) | 0.77 (0.34-1.75) | |
| Medium | 771 | 146 | 1.36 (1.00-1.85) | 1.01 (0.72-1.41) | |
| Strong | 324 | 76 | 1.69 (1.19-2.40) | 1.18 (0.80-1.73) | |
| Coffee temperature | |||||
| Cold | 6 | 2 | 2.39 (0.47-12.1) | 1.63 (0.30-9.01) | |
| Warm | 156 | 26 | 1.20 (0.74-1.94) | 0.90 (0.54-1.50) | |
| Hot or extremely hot | 1028 | 210 | 1.47 (1.10-1.96) | 1.08 (0.79-1.49) | |
| Non-tea drinker | 850 | 130 | Referent | Referent | |
| Ever tea drinkers | 878 | 180 | 1.34 (1.05-1.71) | 1.11 (0.85-1.44) | |
| Caffeinated tea (drinks/day) |
|||||
| 0-<1 | 356 | 80 | 1.47 (1.08-1.99) | 1.25 (0.90-1.73) | |
| 1-<2 | 189 | 44 | 1.52 (1.05-2.22) | 1.18 (0.79-1.76) | |
| ≥2 | 149 | 28 | 1.23 (0.79-1.92) | 0.88 (0.55-1.40) | |
| P-trend in tea drinkers | 0.45 | 0.14 | |||
| Decaffeinated tea | |||||
| Ever | 30 | 5 | 1.09 (0.42-2.86) | 0.66 (0.23-1.87) | |
| Green tea | |||||
| Ever | 85 | 13 | 1.00 (0.54-1.84) | 1.10 (0.57-2.11) | |
| Tea temperature | |||||
| Cold | 724 | 160 | 1.45 (1.12-1.86) | 1.13 (0.86-1.48) | |
| Warm | 24 | 1 | 0.27 (0.04-2.03) | 0.45 (0.06-3.53) | |
| Hot or extremely hot | 116 | 18 | 1.02 (0.60-1.72) | 1.09 (0.62-1.91) |
Adjusted for age, sex, race, WHR, GERD symptoms, smoking, alcohol use, H pylori infection, use of aspirin/NSAIDs, PPIs/H2RAs, and recruitment source.
However, in multivariate analysis, we found no significant association between coffee (adjusted OR, 1.04; 95% CI, 0.76–1.42) and tea (adjusted OR, 1.11; 95% CI, 0.85–1.44) consumption and BE. The associations were similar when we compared cases separately with endoscopy and colonoscopy controls.
Discussion
In this large and well-characterized study we found no association between coffee and tea consumption and BE. Coffee is hypothesized to increase the risk of BE and EAC by promoting reflux through decreased LES pressure and increased acidity. Although patients with GERD are often advised to avoid caffeine,8 the current literature does not support an association between coffee and GERD.9 Likewise, based on the findings of this study and in light of prior contradictory findings for EAC,6,7 there does not appear to be an association between coffee and tea consumption and BE/EAC.
Although cases consumed more coffee and tea than controls, the associations with coffee and tea were almost entirely confounded by sex and race after additional adjustment. Men had higher consumption than women and Caucasians more than non-Caucasians, and these known BE risk factors explained the association.
Strengths of this study include the large sample size, use of a comprehensive questionnaire to collect detailed information on coffee and tea consumption as well as potential confounders, and the strict inclusion criteria. We limited the potential for recall and interviewer bias by conducting interviews prior to the study endoscopy. Assuming a real association does exist, it is possible that the null association here may be due to cases refraining from coffee/tea consumption after enduring prolonged reflux discomfort (or by medical advice). Endoscopy controls may also reduce their coffee/tea consumption due to underlying reflux, thus making them similar in that regard to BE cases. However, for comparisons with endoscopy controls, we observed a similar null finding, suggesting that coffee/tea avoidance by BE cases is unlikely to explain the lack of association. Because our study consisted mostly of older male veterans, our findings may not be generalizable to a wider population.
In conclusion, our findings do not support an association between coffee and tea consumption and BE. Currently, there are not enough data to support avoidance of caffeine to reduce the risk of GERD or BE. Additional studies are needed to confirm these findings, and whether coffee and tea impact risk of progression from BE to EAC needs to be investigated in cohorts of BE patients.
Supplementary Material
Acknowledgments
Funding: Supported in part by a National Institutes of Health grant (NCI R01 116845), the Houston VA HSR&D Center for Innovations in Quality, Effectiveness and Safety (CIN#13-413), the Texas Digestive Disease Center National Institute of Health (DK58338), and by the National Institute of Diabetes and Digestive and Kidney Diseases (K24-04-107 to H.B.E.-S.). The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the United States government.
Abbreviations used in this paper
- BE
Barrett’s esophagus
- CI
confidence interval
- EAC
esophageal adenocarcinoma
- GERD
gastroesophageal reflux disease
- LES
lower esophageal sphincter
- OR
odds ratio
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflicts of interest: The authors disclose no conflicts.
Specific author contributions: Conception and design: APT and HES. Patient recruitment and acquisition of the data: HES. Data preparation and analysis: KCS and APT. Interpretation of the data: KCS, HES and APT. Manuscript preparation and review: KCS, HES and APT.
REFERENCES
- 1.Thrift AP, Whiteman DC. The incidence of esophageal adenocarcinoma continues to rise: analysis of period and birth cohort effects on recent trends. Ann Oncol. 2012;23:3155–3162. doi: 10.1093/annonc/mds181. [DOI] [PubMed] [Google Scholar]
- 2.Ronkainen J, Aro P, et al. Prevalence of Barrett's esophagus in the general population: an endoscopic study. Gastroenterology. 2005;129:1825–1831. doi: 10.1053/j.gastro.2005.08.053. [DOI] [PubMed] [Google Scholar]
- 3.Rustgi A, El-Serag HB. Esophageal carcinoma. N Engl J Med. 2015;372:1472–1473. doi: 10.1056/NEJMc1500692. [DOI] [PubMed] [Google Scholar]
- 4.Thomas FB, Steinbaugh JT, et al. Inhibitory effect of coffee on lower esophageal sphincter pressure. Gastroenterology. 1980;79:1262–1266. [PubMed] [Google Scholar]
- 5.Cohen S, Booth GH., Jr. Gastric acid secretion and lower-esophageal-sphincter pressure in response to coffee and caffeine. N Engl J Med. 1975;293:897–899. doi: 10.1056/NEJM197510302931803. [DOI] [PubMed] [Google Scholar]
- 6.Zamora-Ros R, Lujan-Barroso L, et al. Tea and coffee consumption and risk of esophageal cancer: the European prospective investigation into cancer and nutrition study. Int J Cancer. 2014;135:1470–1479. doi: 10.1002/ijc.28789. [DOI] [PubMed] [Google Scholar]
- 7.Ren JS, Freedman ND, et al. Tea, coffee, carbonated soft drinks and upper gastrointestinal tract cancer risk in a large United States prospective cohort study. Eur J Cancer. 2010;46:1873–1881. doi: 10.1016/j.ejca.2010.03.025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ness-Jensen E, Hveem K, et al. Lifestyle Intervention in Gastroesophageal Reflux Disease. Clin Gastroenterol Hepatol. 2015 doi: 10.1016/j.cgh.2015.04.176. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Kim J, Oh SW, et al. Association between coffee intake and gastroesophageal reflux disease: a meta-analysis. Dis Esophagus. 2014;27:311–317. doi: 10.1111/dote.12099. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.