View PDF VersionPrevious ArticleNext Article
 
DOI: 10.1039/D5FO03778F (Paper) Food Funct., 2026, 17, 166-175

Association between hot beverage intake and gastric cancer risk: a prospective cohort study from the UK Biobank

Fengyi Huang ab, Xite Zheng ab, Yanling Qi c, Xiaorui Zhang de, Changwei Li *f, Deqiang Zheng *ab and Fen Liu *ab
aDepartment of Epidemiology and Health Statistics, School of Public Health, Capital Medical University, No. 10 Xitoutiao, Youanmenwai Street, Fengtai District, Beijing, 100069, China. E-mail: dqzheng@ccmu.edu.cn; lfmail0225@sina.com
bBeijing Key Laboratory of Environment and Aging, Capital Medical University, Beijing, 100069, China
cZhou Enlai School of Government, Nankai University, Tianjin, 300072, China
dBeijing Key Laboratory of Mental Disorders, National Clinical Research Center for Mental Disorders & National Center for Mental Disorders, Beijing Anding Hospital, Capital Medical University, Beijing, 100088, China
eAdvanced Innovation Center for Human Brain Protection, Capital Medical University, Beijing, 100088, China
fDepartment of Epidemiology, O'Donnell School of Public Health, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd Dallas, TX 75390, United States. E-mail: Changwei.li@utsouthwestern.edu

Received 4th September 2025 , Accepted 1st December 2025

First published on 16th December 2025

Abstract

Hot beverage consumption has been hypothesized to increase the risk of esophageal cancer; however, its impact on gastric cancer (GC) is still inconclusive. This study aimed to investigate the prospective associations between hot beverage intake and the risk of GC. We examined the association between daily consumption and preferred temperature of hot beverages (tea and coffee) and the incidence of GC in 328[thin space (1/6-em)]752 UK Biobank participants. The consumption of beverages and preferred temperature were collected using a food frequency questionnaire. We applied Cox proportional hazard regression to estimate the multivariable hazard ratios (HRs) and 95% confidence intervals (95% CIs) for the association of hot beverage intake and GC risk and further stratified the analysis by anatomical subsites. During a follow-up period of 11.6 years, 523 incident GC cases were identified. Compared with non-drinkers and warm temperature drinkers, drinking over 8 cups per day (HR, 1.54; 95% CI, 1.08–2.21) of hot beverages was associated with a higher risk of GC. The risk was further elevated among participants who preferred very hot beverages, the HRs (95% CI) comparing reference were 1.69 (1.06–2.68) for 6–8 cups per day and 2.03 (1.26–3.27) for >8 cups per day. These results remained consistent in subgroups and sensitivity analyses. Our findings provide new evidence that drinking hot or very hot beverages is a risk factor for GC in the UK where drinking hot tea and coffee is common.

Introduction

Gastric cancer (GC) remains a major global health burden, ranking fifth in incidence and mortality worldwide, with an estimated 968[thin space (1/6-em)]350 new cases and 659[thin space (1/6-em)]853 deaths annually.1 The incidence of GC varies markedly across regions, with the highest rates observed in Eastern Asia and Eastern Europe, and the lowest in most parts of Africa. GC comprises two major anatomical subtypes – gastric cardia adenocarcinoma (GCA) and gastric non-cardia adenocarcinoma (GNCA) – which demonstrate distinct epidemiological trends and risk factors.2 Notably, while the overall incidence of GC has declined in many countries, the incidence of GCA has increased since the 1960s in Western populations, particularly in the United Kingdom (UK).3

Diet is believed to be an important modifiable factor associated with GC risk. Low consumption of vegetables and fruits, high sodium intake, greater intake of processed meat, alcoholic drinks, and tobacco smoking have been implicated in increasing GC risk.4,5 Tea and coffee, widely consumed beverages globally, are often ingested at elevated temperatures. Current studies regarding the association between tea and coffee consumption and GC risk remain inconclusive. Experimental studies and epidemiological studies have provided evidence for the inhibitory effects of tea and coffee in the development of GC.6,7 Conversely, chronic thermal injury to the upper gastrointestinal tract resulting from the consumption of hot beverages may contribute to the development of cancer.8 In 2016, the International Agency for Research on Cancer (IARC) classified drinking hot beverages (>65 °C) as “probably carcinogenic to humans”.8

Multiple studies have demonstrated a positive association between consumption of hot or very hot tea or coffee and the risk of esophageal cancer (EC).9,10 However, despite the anatomical proximity of the gastric to the esophagus, evidence concerning the potential link between hot beverage consumption and GC risk remains limited and inconsistent. A prospective cohort study within U.S. cohorts has shown that no significant associations were found between hot tea consumption and the GC risk.11 However, a systematic review and dose–response meta-analysis of thirteen observational studies have found that drinking very hot tea was significantly associated with a high risk of GC.12 Moreover, case-control studies in Mongolia, Iran, and China have reported a higher risk of GC among individuals who consume hot beverages compared to non-drinkers.13–15 Furthermore, the majority of existing studies exploring this association have relied on case-control design, which is inherently prone to biases such as recall bias and reverse causality, while prospective evidence considering beverage temperature remains limited.

In the UK, hot beverages (tea and coffee) are frequently consumed and often at elevated temperatures. Therefore, the present study aimed to prospectively examine the associations between the hot beverage temperature and intake level and the incidence of GC, using data from a large, population-based cohort in the UK.

Methods

Study population

The UK Biobank is a large, population-based, multicenter prospective cohort study. Detailed descriptions of the study have been presented previously.16 Briefly, over 500[thin space (1/6-em)]000 residents aged 37–73 years were recruited across 22 assessment centers in the United Kingdom from March 2006 to December 2010. Participants provided information on sociodemographic factors, habitual diet, lifestyle factors, and medical history through touchscreen questionnaires and face-to-face interviews. The baseline summary data of the cohort can be viewed in the data showcase on the UK Biobank website (https://www.ukbiobank.ac.uk). The UK Biobank study received approval from the National Health Service (NHS) North West Multicenter Research Ethics Committee. All participants gave their written informed consent at the time of recruitment.

For the current analysis, we excluded participants who were confirmed prevalent cancer (except for non-melanoma skin cancer) or self-reported prevalent cancer on the baseline questionnaire (n = 47[thin space (1/6-em)]305), who were pregnant or unsure (n = 352), who with missing or incomplete data on tea or coffee intake (n = 2779) or preferred hot beverage temperature (n = 154), who did not drink tea or coffee but reported their preferred hot beverage temperature (n = 5918), and who with missing or incomplete data for covariables (n = 117[thin space (1/6-em)]097). A total of 328[thin space (1/6-em)]752 participants is included in the final analysis (SI Fig. S1).

Exposure assessment

Dietary data, including hot beverage intake, were collected using a food frequency questionnaire (FFQ). Tea consumption information was collected based on the question “How many cups of tea do you drink each day (including black and green tea)?” with the following response alternatives: the exact number of cups of tea drinking per day, less than one, do not know, or prefer not to answer. Similarly to tea consumption, the intake of coffee was collected based on the question “How many cups of coffee do you drink each day? (Include decaffeinated coffee)”. If over 20 cups of tea or 10 cups of coffee per day were reported, the participant was requested to confirm the number. Responses with the answer “do not know” or “prefer not to answer” were classified as missing. The total of hot beverage intake was calculated by the sum of tea intake and coffee intake. For the participants who indicated that they drank at least 1 cup of coffee each day or less than one cup each day, they were recorded coffee type, including decaffeinated coffee, instant coffee, ground coffee, and any other type of coffee. Additionally, participants were also asked their preferred temperature of hot beverages based on the question “How do you like your hot drinks? (Such as coffee or tea)”, by selecting from “very hot”, “hot”, “warm”, “do not drink hot drinks”, or “prefer not to answer”. Responses with the answer “prefer not to answer” were classified as missing. For this analysis, participants were classified based on their preference for hot beverage temperature (hot or very hot) and their daily total intake of hot beverages (≤4, >4–6, >6–8, and >8 cups per day). Those who consumed neither tea nor coffee, as well as individuals who reported drinking these beverages at a warm temperature, were grouped together as the reference category.

Assessment of outcomes

Incident cancer cases were identified through data linkage to national cancer and mortality registries. Cancer status was classified using the International Classification of Diseases, 10th edition (ICD-10). In the present study, the outcome was incident overall GC (C16) and GC by anatomical subsites (GCA: C16.0; GNCA: C16.1–C16.9).

Follow-up time was calculated from the date of recruitment to the date of the GC diagnosis, death, loss of follow-up date, or end of the follow-up (December 31st, 2020 in England, December 31st, 2016 in Wales, and November 30th, 2021 in Scotland), whichever came first.

Covariates

Covariates involved in this study were selected based on both previous studies,17 which were obtained via touchscreen questionnaires at recruitment. The covariates included age (years), sex (male, female), race (White, others), body mass index (BMI, kg m−2) Townsend Deprivation Index (TDI) (an area–based deprivation measure for socioeconomic status), education level (college or university degree, others), smoking status (never, previous, current), alcohol intake frequency (never, moderate, heavy), total metabolic equivalent task (MET, minutes per week), family history of gastrointestinal cancer (yes, no), history of esophageal disorder (yes, no), history of gastric disorder (yes, no), intake of vegetables (tablespoons per day), intake of fruits (tablespoons per day), intake of red meat (servings per week), and intake of processed meat (times per week). Dietary variables were derived from the baseline touchscreen FFQ. Daily vegetable intake was derived from separate questions for average daily intake of cooked vegetables and salad/raw vegetables. Daily fruit intake was derived from separate questions for average daily intake of fresh fruit and dried fruit. Red meat intake frequency was calculated from separate questions on the average weekly intake of beef, lamb, and pork. Both red and processed meat intakes were derived from weekly frequency questions using standardized response categories (never, <1 time per week, 1 time per week, 2–4 times per week, 5–6 times per month, or ≥1 time per day). BMI was computed using height and weight measured at the initial physical assessment central visit. MET minutes per week were calculated for all activity, which included walking, moderate and vigorous activity. Alcohol intake frequency was categorized into three groups: never, moderate (3–4 times per week, 1–2 times per week, 1–3 times per month, or special occasions only), and heavy (daily or almost daily). History of esophageal disorder mainly included gastro-esophageal reflux (GERD)/gastric reflux and esophagitis/Barrett's esophagus, history of gastric disorder mainly included gastric/stomach ulcers and gastritis/gastric erosions.

Statistical analyses

Baseline characteristics were stratified by hot beverage temperature and intake categories and participants developed GC or not. Continuous variables were presented as mean (SD), and categorical variables were presented as n (%). Student's t-test for continuous variables and Chi-square test for categorical variables.

Cox proportional hazard regression was used to estimate the hazard ratios (HRs) and 95% confidence intervals (CIs) for GC and its anatomical subtypes across nine combined categories of beverage temperature and intake levels. The proportional hazards assumption for the Cox models was checked using Schoenfeld residuals, and no violation was found. The adjustment variables included in the Cox model described below comprise the minimum sufficient set selected through directed acyclic graph (DAG) analysis (SI Fig. S2) and important factors associated with GC identified in previous studies.10,18,19 Models were adjusted for potential confounders, including age, sex, race, TDI, BMI, education level, smoking status, alcohol intake frequency, MET, family history of gastrointestinal cancer, history of esophageal disorder, history of gastric disorder, and dietary intake, which included vegetables, fruits, red meat, and processed meat. We further analyzed the associations of hot tea and hot coffee consumption individually with the incidence of GC, adjusting for the same covariates as in the main cox proportional hazards models. Additionally, we performed subgroup analysis. We tested the linear trend of GC risk by entering the consecutive categories of each beverage intake group as continuous variables in the regression models.

We performed several sensitivity analyses to test the robustness of the results. Firstly, to minimize the potential reverse causality, we excluded the individuals who developed GC within two years of follow-up. Secondly, to account for the potential influence of coffee type, we repeated the analysis of hot beverage consumption in relation to GC risk after excluding participants with missing data on coffee type and additionally adjusted for it in the model. Thirdly, we explored the potential unmeasured confounding factors between hot beverage consumption and GC risk by calculating E-values.20

All the statistical analyses were performed using R software (Version 4.3.2; R Foundation for Statistical Computing, Vienna, Austria). P < 0.05 was considered to indicate statistical significance in two-sided tests.

Results

Baseline characteristics of participants

Overall, we enrolled 328[thin space (1/6-em)]752 participants in the present study, of whom 51.0% were female. The baseline characteristics of the study population are presented in Table 1. The mean age of the participants was 56.0 years (±8.1). Among the cohort participants, 67.1% drank both tea and coffee, 19.4% drank only tea, 12.4% drank only coffee, and 1.1% reported drinking neither tea nor coffee. Regarding beverage temperature preference, 16.3% of participants reported that they did not drink tea or coffee or preferred drinking warm beverages, 66.5% of participants reported preferred drinking hot beverages, and 17.3% of them preferred drinking very hot beverages. Overall, 24.3% of participants consumed ≤4 cups per day of hot beverages, and 6.8% consumed >8 cups per day. Similarly, 5.4% of participants consumed ≤4 cups per day of very hot beverages and 2.5% consumed >8 cups per day.
Table 1 Baseline demographic and lifestyle characteristics by preferred temperature and daily intake of hot beverages (tea or coffee)
Characteristic All participants Tea or coffee drinking
Non-drinkers or warm temperature Hot temperature (cups per day) Very hot temperature (cups per day)
≤4 >4–6 >6–8 >8 ≤4 >4–6 >6–8 >8
BMI, body mass index; TDI, townsend deprivation index; MET, total metabolic equivalent task; GIC, gastrointestinal cancer.
Numbers (%) 328[thin space (1/6-em)]752 (100.00) 53[thin space (1/6-em)]429 (16.25) 80[thin space (1/6-em)]328 (24.34) 76[thin space (1/6-em)]611 (23.30) 39[thin space (1/6-em)]386 (11.98) 22[thin space (1/6-em)]196 (6.75) 17[thin space (1/6-em)]598 (5.35) 19[thin space (1/6-em)]297 (5.87) 11[thin space (1/6-em)]615 (3.53) 8292 (2.52)
Age (years), mean (SD) 55.93 (8.10) 55.42 (8.24) 55.64 (8.33) 56.67 (7.95) 56.69 (7.78) 56.09 (7.88) 54.65 (8.21) 55.57 (7.99) 55.60 (7.85) 55.10 (7.83)
Sex (%)
Male 161[thin space (1/6-em)]113 (49.01) 26[thin space (1/6-em)]639 (49.86) 38[thin space (1/6-em)]688 (48.16) 38[thin space (1/6-em)]183 (49.84) 20[thin space (1/6-em)]448 (51.92) 12[thin space (1/6-em)]850 (57.89) 6964 (39.57) 8127 (42.12) 5015 (43.18) 4199 (50.64)
Female 167[thin space (1/6-em)]639 (50.99) 26[thin space (1/6-em)]790 (50.14) 41[thin space (1/6-em)]640 (51.84) 38[thin space (1/6-em)]428 (50.16) 18[thin space (1/6-em)]938 (48.08) 9346 (42.11) 10[thin space (1/6-em)]634 (60.43) 11[thin space (1/6-em)]170 (57.88) 6600 (56.82) 4093 (49.36)
Race (%)
White 313[thin space (1/6-em)]331 (95.31) 50[thin space (1/6-em)]252 (94.05) 74[thin space (1/6-em)]300 (92.50) 74[thin space (1/6-em)]514 (97.26) 38[thin space (1/6-em)]667 (98.17) 21[thin space (1/6-em)]720 (97.86) 15[thin space (1/6-em)]831 (89.96) 18[thin space (1/6-em)]593 (96.35) 11[thin space (1/6-em)]350 (97.72) 8104 (97.73)
Others 15[thin space (1/6-em)]421 (4.69) 3177 (5.95) 6028 (7.50) 2097 (2.74) 719 (1.83) 476 (2.14) 1767 (10.04) 704 (3.65) 265 (2.28) 188 (2.27)
BMI (kg m −2 ), (%)
<18.5 1601 (0.49) 213 (0.40) 414 (0.52) 332 (0.43) 142 (0.36) 92 (0.41) 154 (0.88) 124 (0.64) 68 (0.59) 62 (0.75)
18.5 to <25 109[thin space (1/6-em)]849 (33.41) 15[thin space (1/6-em)]174 (28.40) 28[thin space (1/6-em)]697 (35.72) 25[thin space (1/6-em)]849 (33.74) 12[thin space (1/6-em)]507 (31.75) 6486 (29.22) 6957 (39.53) 7245 (37.54) 4157 (35.79) 2777 (33.49)
25 to <30 141[thin space (1/6-em)]709 (43.11) 22[thin space (1/6-em)]714 (42.51) 33[thin space (1/6-em)]765 (42.03) 34[thin space (1/6-em)]063 (44.46) 17[thin space (1/6-em)]753 (45.07) 9964 (44.89) 6793 (38.60) 8127 (42.12) 4981 (42.88) 3549 (42.80)
≥30 75[thin space (1/6-em)]593 (22.99) 15[thin space (1/6-em)]328 (28.69) 17[thin space (1/6-em)]452 (21.73) 16[thin space (1/6-em)]367 (21.36) 8984 (22.81) 5654 (25.47) 3694 (20.99) 3801 (19.70) 2409 (20.74) 1904 (22.96)
Townsend deprivation index (%)
Tertile 1 (least deprived) 109[thin space (1/6-em)]603 (33.34) 16[thin space (1/6-em)]713 (31.28) 25[thin space (1/6-em)]665 (31.95) 27[thin space (1/6-em)]359 (35.71) 14[thin space (1/6-em)]075 (35.74) 6998 (31.53) 5461 (31.03) 6734 (34.90) 4053 (34.89) 2545 (30.69)
Tertile2 109[thin space (1/6-em)]572 (33.33) 17[thin space (1/6-em)]288 (32.36) 26[thin space (1/6-em)]273 (32.71) 26[thin space (1/6-em)]333 (34.37) 13[thin space (1/6-em)]529 (34.35) 7269 (32.75) 5627 (31.98) 6598 (34.19) 3914 (33.70) 2741 (33.06)
Tertile3 (most deprived) 109[thin space (1/6-em)]577 (33.33) 19[thin space (1/6-em)]428 (36.36) 28[thin space (1/6-em)]390 (35.34) 22[thin space (1/6-em)]919 (29.92) 11[thin space (1/6-em)]782 (29.91) 7929 (35.72) 6510 (36.99) 5965 (30.91) 3648 (31.41) 3006 (36.25)
Education level (%)
College or University Degree 119[thin space (1/6-em)]918 (36.48) 20[thin space (1/6-em)]573 (38.51) 30[thin space (1/6-em)]757 (38.29) 28[thin space (1/6-em)]012 (36.56) 13[thin space (1/6-em)]991 (35.52) 7284 (32.82) 6186 (35.15) 6592 (34.16) 3879 (33.40) 2644 (31.89)
Others 208[thin space (1/6-em)]834 (63.52) 32[thin space (1/6-em)]856 (61.49) 49[thin space (1/6-em)]571 (61.71) 48[thin space (1/6-em)]599 (63.44) 25[thin space (1/6-em)]395 (64.48) 14[thin space (1/6-em)]912 (67.18) 11[thin space (1/6-em)]412 (64.85) 12[thin space (1/6-em)]705 (65.84) 7736 (66.60) 5648 (68.11)
Smoking status (%)
Never 182[thin space (1/6-em)]408 (55.48) 29[thin space (1/6-em)]324 (54.88) 46[thin space (1/6-em)]964 (58.47) 42[thin space (1/6-em)]673 (55.70) 20[thin space (1/6-em)]806 (52.83) 10[thin space (1/6-em)]269 (46.27) 10[thin space (1/6-em)]678 (60.68) 11[thin space (1/6-em)]240 (58.25) 6393 (55.04) 4061 (48.97)
Previous 113[thin space (1/6-em)]743 (34.60) 18[thin space (1/6-em)]207 (34.08) 27[thin space (1/6-em)]558 (34.31) 27[thin space (1/6-em)]339 (35.69) 14[thin space (1/6-em)]103 (35.81) 7811 (35.19) 5652 (32.12) 6444 (33.39) 3939 (33.91) 2690 (32.44)
Current 32[thin space (1/6-em)]601 (9.92) 5898 (11.04) 5806 (7.23) 6599 (8.61) 4477 (11.37) 4116 (18.54) 1268 (7.21) 1613 (8.36) 1283 (11.05) 1541 (18.58)
Alcohol intake frequency (%)
Never 22[thin space (1/6-em)]066 (6.71) 4900 (9.17) 5350 (6.66) 3957 (5.17) 2138 (5.43) 1749 (7.88) 1413 (8.03) 1143 (5.92) 700 (6.03) 716 (8.63)
Moderate 235[thin space (1/6-em)]337 (71.58) 37[thin space (1/6-em)]710 (70.58) 55[thin space (1/6-em)]701 (69.34) 55[thin space (1/6-em)]577 (72.54) 29[thin space (1/6-em)]098 (73.88) 16[thin space (1/6-em)]319 (73.52) 12[thin space (1/6-em)]289 (69.83) 13[thin space (1/6-em)]989 (72.49) 8574 (73.82) 6080 (73.32)
Heavy 71[thin space (1/6-em)]349 (21.70) 10[thin space (1/6-em)]819 (20.25) 19[thin space (1/6-em)]277 (24.00) 17[thin space (1/6-em)]077 (22.29) 8150 (20.69) 4128 (18.60) 3896 (22.14) 4165 (21.58) 2341 (20.15) 1496 (18.04)
MET (minutes per week), (%)
<300 28[thin space (1/6-em)]518 (8.67) 5351 (10.02) 6682 (8.32) 6252 (8.16) 3277 (8.32) 2044 (9.21) 1521 (8.64) 1592 (8.25) 1004 (8.64) 795 (9.59)
300 to <600 31[thin space (1/6-em)]194 (9.49) 5341 (10.00) 7618 (9.48) 7178 (9.37) 3726 (9.46) 2086 (9.40) 1624 (9.23) 1802 (9.34) 1044 (8.99) 775 (9.35)
≥600 269[thin space (1/6-em)]040 (81.84) 42[thin space (1/6-em)]737 (79.99) 66[thin space (1/6-em)]028 (82.20) 63[thin space (1/6-em)]181 (82.47) 32[thin space (1/6-em)]383 (82.22) 18[thin space (1/6-em)]066 (81.39) 14[thin space (1/6-em)]453 (82.13) 15[thin space (1/6-em)]903 (82.41) 9567 (82.37) 6722 (81.07)
Family history of GIC (%)
No 292[thin space (1/6-em)]231 (88.89) 47[thin space (1/6-em)]547 (88.99) 71[thin space (1/6-em)]849 (89.44) 68[thin space (1/6-em)]060 (88.84) 34[thin space (1/6-em)]868 (88.53) 19[thin space (1/6-em)]549 (88.07) 15[thin space (1/6-em)]645 (88.90) 17[thin space (1/6-em)]067 (88.44) 10[thin space (1/6-em)]323 (88.88) 7323 (88.31)
Yes 36[thin space (1/6-em)]521 (11.11) 5882 (11.01) 8479 (10.56) 8551 (11.16) 4518 (11.47) 2647 (11.93) 1953 (11.10) 2230 (11.56) 1292 (11.12) 969 (11.69)
History of esophageal disorder (%)
No 314[thin space (1/6-em)]854 (95.77) 50[thin space (1/6-em)]969 (95.40) 77[thin space (1/6-em)]172 (96.07) 73[thin space (1/6-em)]380 (95.78) 37[thin space (1/6-em)]622 (95.52) 21[thin space (1/6-em)]146 (95.27) 16[thin space (1/6-em)]932 (96.22) 18[thin space (1/6-em)]551 (96.13) 11[thin space (1/6-em)]142 (95.93) 7940 (95.75)
Yes 13[thin space (1/6-em)]898 (4.23) 2460 (4.60) 3156 (3.93) 3231 (4.22) 1764 (4.48) 1050 (4.73) 666 (3.78) 746 (3.87) 473 (4.07) 352 (4.25)
History of gastric disorder (%)
No 325[thin space (1/6-em)]389 (98.98) 52[thin space (1/6-em)]753 (98.73) 79[thin space (1/6-em)]616 (99.11) 75[thin space (1/6-em)]885 (99.05) 38[thin space (1/6-em)]980 (98.97) 21[thin space (1/6-em)]936 (98.83) 17[thin space (1/6-em)]428 (99.03) 19[thin space (1/6-em)]098 (98.97) 11[thin space (1/6-em)]495 (98.97) 8198 (98.87)
Yes 3363 (1.02) 676 (1.27) 712 (0.89) 726 (0.95) 406 (1.03) 260 (1.17) 170 (0.97) 199 (1.03) 120 (1.03) 94 (1.13)
Hot beverages type (%)
Neither tea nor coffee 3449 (1.05) 3449 (6.46) 0 (0.00) 0 (0.00) 0 (0.00) 0 (0.00) 0 (0.00) 0 (0.00) 0 (0.00) 0 (0.00)
Tea and coffee 220[thin space (1/6-em)]658 (67.12) 30[thin space (1/6-em)]909 (57.85) 44[thin space (1/6-em)]040 (54.83) 56[thin space (1/6-em)]457 (73.69) 32[thin space (1/6-em)]853 (83.41) 18[thin space (1/6-em)]055 (81.34) 9153 (52.01) 13[thin space (1/6-em)]410 (69.49) 9262 (79.74) 6519 (78.62)
Tea only 63[thin space (1/6-em)]848 (19.42) 10[thin space (1/6-em)]733 (20.09) 19[thin space (1/6-em)]675 (24.49) 13[thin space (1/6-em)]387 (17.47) 4777 (12.13) 3137 (14.13) 4907 (27.88) 4059 (21.03) 1773 (15.26) 1400 (16.88)
Coffee only 40[thin space (1/6-em)]797 (12.41) 8338 (15.61) 16[thin space (1/6-em)]613 (20.68) 6767 (8.83) 1756 (4.46) 1004 (4.52) 3538 (20.10) 1828 (9.47) 580 (4.99) 373 (4.50)
Tea (cups per day), mean (SD) 3.43 (2.82) 2.90 (2.82) 1.72 (1.27) 3.36 (1.73) 4.75 (2.06) 7.52 (4.26) 1.81 (1.30) 3.47 (1.79) 4.92 (2.10) 7.92 (4.71)
Coffee (cups per day), mean (SD) 2.05 (2.06) 1.92 (2.09) 1.37 (1.17) 2.08 (1.71) 2.62 (2.07) 3.63 (3.51) 1.31 (1.20) 1.99 (1.76) 2.46 (2.11) 3.49 (3.67)
Vegetables (tablespoons per day), mean (SD) 4.91 (3.33) 5.03 (3.70) 4.89 (3.37) 4.80 (3.00) 4.87 (3.10) 5.01 (3.47) 5.01 (3.53) 4.86 (3.09) 4.94 (3.24) 5.06 (3.97)
Fruits (pieces per day), mean (SD) 3.09 (2.54) 3.13 (2.65) 3.11 (2.59) 3.05 (2.36) 3.08 (2.45) 3.05 (2.72) 3.11 (2.62) 3.04 (2.50) 3.08 (2.56) 3.03 (2.78)
Red meat (servings per week), mean (SD) 3.67 (1.75) 3.62 (1.80) 3.57 (1.75) 3.74 (1.68) 3.78 (1.71) 3.82 (1.80) 3.50 (1.82) 3.68 (1.76) 3.74 (1.77) 3.77 (1.86)
Processed meat (times per week) (%)
0 30[thin space (1/6-em)]195 (9.18) 5545 (10.38) 8393 (10.45) 5745 (7.50) 2955 (7.50) 1726 (7.78) 2226 (12.65) 1783 (9.24) 1035 (8.91) 787 (9.49)
<1 99[thin space (1/6-em)]497 (30.27) 16[thin space (1/6-em)]045 (30.03) 25[thin space (1/6-em)]972 (32.33) 23[thin space (1/6-em)]045 (30.08) 10[thin space (1/6-em)]978 (27.87) 5847 (26.34) 5904 (33.55) 6115 (31.69) 3373 (29.04) 2218 (26.75)
1 96[thin space (1/6-em)]077 (29.22) 14[thin space (1/6-em)]834 (27.76) 23[thin space (1/6-em)]182 (28.86) 23[thin space (1/6-em)]487 (30.66) 11[thin space (1/6-em)]829 (30.03) 6426 (28.95) 4838 (27.49) 5744 (29.77) 3419 (29.44) 2318 (27.95)
2 to 4 89[thin space (1/6-em)]967 (27.37) 14[thin space (1/6-em)]650 (27.42) 20[thin space (1/6-em)]068 (24.98) 21[thin space (1/6-em)]502 (28.07) 12[thin space (1/6-em)]008 (30.49) 6953 (31.33) 4050 (23.01) 4941 (25.61) 3305 (28.45) 2490 (30.03)
5 to 6 10[thin space (1/6-em)]470 (3.18) 1841 (3.45) 2192 (2.73) 2345 (3.06) 1305 (3.31) 985 (4.44) 448 (2.55) 598 (3.10) 375 (3.23) 381 (4.59)
≥7 2546 (0.77) 514 (0.96) 521 (0.65) 487 (0.64) 311 (0.79) 259 (1.17) 132 (0.75) 116 (0.60) 108 (0.93) 98 (1.18)

Hot beverage intake and risk of GC

During a mean follow-up period of 11.6 years (±1.8), 523 incident GC events were identified. Participants who developed GC were more likely to be older, male, have a higher BMI, lack a college or university education, be former or current smokers, have heavy alcohol intake frequency, and have a history of esophageal disorder compared to those who did not develop GC (SI Table S1).

Table 2 presents the results of the Cox regression examining the associations between hot beverage intake and GC risk. According to the multivariable adjusted model, higher intake of hot beverages was significantly associated with increased GC risk (Ptrend = 0.002). Relative to participants reported not drinking tea or coffee or preferred drinking warm beverages, those who preferred hot beverages had a higher risk of GC; the HR (95% CI) comparing reference was 1.54 (1.08–2.21) for those consuming >8 cups per day. The risk was further elevated among participants who preferred very hot beverages, the HRs (95% CI) comparing reference were 1.69 (1.06–2.68) for 6–8 cups per day and 2.03 (1.26–3.27) for >8 cups per day.

Table 2 Risk of gastric cancer by preferred temperature and daily intake level of hot beverages (tea and coffee)
Variables Participants (cases) Model 1 Model 2 Model 3
HR (95% CI) P trend1 † P trend2 ‡ HR (95% CI) P trend1 † P trend2 ‡ HR (95% CI) P trend1 † P trend2 ‡
HR, hazard ratio; 95% CI, 95% confidence interval. Model 1, not adjusted model; Model 2, adjusted for age and sex; Model 3, additionally adjusted for race, Townsend Deprivation Index, body mass index, education level, smoking status, alcohol intake frequency, total metabolic equivalent task, family history of gastrointestinal cancer, history of esophageal disorder, history of gastric disorder, and dietary intake including vegetables, fruits, red meat, and processed meat. †P values for trend1 were calculated by assigning consecutive integers to the five beverage consumption categories within each temperature group (hot or very hot), separately. ‡P values for trend2 were calculated by assigning consecutive integers to all nine combined categories of beverage consumption and temperature.
Non-drinkers or warm temperature 53[thin space (1/6-em)]429 (73) 1.00 (reference) 0.009 1.00 (reference) 0.002 1.00 (reference) 0.002
Hot temperature
≤4 cups per day 80[thin space (1/6-em)]328 (113) 1.02 (0.76–1.37) 1.01 (0.75–1.35) 1.09 (0.81–1.46)
>4–6 cups per day 76[thin space (1/6-em)]611 (126) 1.20 (0.90–1.60) 1.10 (0.83–1.47) 1.02 (0.90–1.61)
>6–8 cups per day 39[thin space (1/6-em)]386 (63) 1.17 (0.84–1.64) 1.07 (0.77–1.51) 1.13 (0.81–1.59)
>8 cups per day 22[thin space (1/6-em)]196 (52) 1.73 (1.21–2.47) 0.003 1.59 (1.11–2.27) 0.027 1.54 (1.08–2.21) 0.028
Very hot temperature
≤4 cups per day 17[thin space (1/6-em)]598 (22) 0.90 (0.56–1.46) 1.03 (0.64–1.65) 1.10 (0.68–1.77)
>4–6 cups per day 19[thin space (1/6-em)]297 (28) 1.06 (0.68–1.63) 1.12 (0.72–1.73) 1.22 (0.79–1.88)
>6–8 cups per day 11[thin space (1/6-em)]615 (24) 1.50 (0.95–2.39) 1.59 (1.00–2.52) 1.69 (1.06–2.68)
>8 cups per day 8292 (22) 1.96 (1.22–3.16) 0.006 2.06 (1.28–3.32) 0.002 2.03 (1.26–3.27) 0.002

When examining GC anatomical subsites, very hot beverages consumption of >8 cups per day was associated with a significantly increased risk of GCA (HR comparing reference: 2.93, 95% CI: 1.52–5.64) (Fig. 1). A similar positive association was observed for GNCA, participants who preferred very hot beverages and consumed 6–8 cups per day had an increased risk of GNCA (HR comparing reference: 1.98, 95% CI: 1.13–3.48).


image file: d5fo03778f-f1.tif
Fig. 1 Risk of gastric cancer by preferred temperature and daily intake level of hot beverages (tea and coffee) by anatomical subtypes. GCA, gastric cardia adenocarcinoma; GNCA, gastric noncardia adenocarcinoma; HR, hazard ratio; 95% CI, 95% confidence interval. Models adjusted for age, sex, race, Townsend Deprivation Index, body mass index, education level, smoking status, alcohol intake frequency, total metabolic equivalent task, family history of gastrointestinal cancer, history of esophageal disorder, history of gastric disorder, and dietary intake including vegetables, fruits, red meat, and processed meat. P values for trend1 were calculated by assigning consecutive integers to the five beverage consumption categories within each temperature group (hot or very hot), separately. P values for trend2 were calculated by assigning consecutive integers to all nine combined categories of beverage consumption and temperature.

We further examined the associations of hot tea and hot coffee consumption with GC risk separately. In the fully adjusted models, very hot tea consumption of >3 cups per day was associated with a significantly increased risk of GC (HR comparing reference: 1.51, 95% CI: 1.12–2.03) (SI Table S2). No significant differences were observed in the association between hot coffee consumption and GC risk (SI Table S3).

Subgroup analyses

Associations of hot beverage intake with the risk of GC remained consistent across the subgroups (SI Fig. S3–S7). The same relationships were observed between subgroups stratified by age, sex, obese or not, smoking status, and alcohol consumption frequency (all Pinteraction > 0.05).

Sensitivity analyses

In the sensitivity analyses, after excluding the GC events that occurred in the first 2 years of follow-up, the significant positive association between hot beverage consumption and GC risk remained stable (SI Table S4). To account for the effect of coffee type, participants with missing data on coffee type were excluded, and the results were robust in the models additionally adjusted for coffee type (SI Table S5). To further assess the potential impact of unmeasured confounding, we calculated E-values for the associations between hot beverage consumption and risk of GC and its anatomical subtypes (GCA and GNCA), and the findings remained generally robust (SI Table S6).

Discussion

In this large prospective cohort study involving 328[thin space (1/6-em)]752 participants based on the UK Biobank, we found that hot beverage intake was positively associated with the risk of GC, particularly GCA. The results of the subgroup and sensitivity analyses were also found to be consistent. Our findings indicated a positive correlation between the consumption of hot beverages and the development of GC. Notably, this is the first prospective cohort study to comprehensively investigate the association between hot beverages (tea and coffee) and GC among UK adults.

The present study showed that higher consumption of hot beverages was positively associated with the risk of GC. Consistent with our results, an umbrella review including 96 meta-analyses found a harmful association between hot tea and GC risk.21 Another meta-analysis of five cohort studies and eight case-control studies found similar results.12 Moreover, an early case-control study conducted in China found that an inverse relationship between green tea drinking and GC risk was limited to the intake of lukewarm tea.22 The biological mechanism for hot beverage consumption could increase the risk of GC may be thermal injury. Thermal damage from hot beverage consumption may compromise the mucosal barrier of the gastric cardia, similar to its effects in the esophagus, thereby increasing susceptibility to luminal carcinogens.23 In addition, chronic thermal irritation can trigger localized inflammation, promoting the production of reactive nitrogen species (RNS) and reactive oxygen species (ROS), which are known to induce DNA damage in gastric epithelial cells. Indeed, studies have demonstrated that persistent inflammation in the gastric cardia is associated with increased DNA damage response, a key driver of carcinogenesis.24 Hence, these biological mechanisms provide biological plausibility for the observed association between hot beverage intake and risk of GC.

An early US population-based prospective cohort study in 2010 showed that non-significant associations between hot tea intake and both GCA and GNCA risk, while an inverse association was observed between coffee intake and GCA risk (regardless of temperature).11 In contrast, our stratified analysis by GC anatomical subtypes found that a significant positive association between hot beverages (tea and coffee) and both GCA and GNCA risk, though the relationship was lower in GNCA. These findings suggested potential carcinogenic effects related to beverage temperature. Differences in the distribution of GC subtypes across study populations may be partly explained the inconsistent findings. The previous U.S. cohort included a relatively balanced number of GCA (n = 231) and GNCA (n = 224) cases, whereas in our analysis, GNCA accounted for a larger proportion, with 309 GNCA and 214 GCA cases. Another possible explanation for the discrepancies lies in the selection of the reference group. In our analysis, individuals who either did not consume hot beverages or preferred warm beverages were grouped as the reference category. In contrast, previous studies classified intake categories for hot tea, iced tea and coffee in cups and used non-drinkers as the sole reference group. This difference in reference definitions may have introduced exposure misclassification or diluted effect estimates. The stronger association with GCA may be due to its anatomical proximity to the esophagus, making it more directly exposed to thermal injury from hot beverages than distal gastric regions.25

The observed associations of hot beverage drinking with the risk of GC may be attributed to the individual beverage types. According to our stratified analyses for beverage type, very hot tea remained significantly associated with an increased risk, whereas associations with hot tea and hot coffee were attenuated and became non-significant. Several explanations may account for this discrepancy. One potential explanation lies in population-level consumption patterns. Tea has historically been more widely consumed than coffee in the UK, potentially resulting in greater cumulative exposure to hot beverages among tea drinkers.26 Another contributing factor may be differences in preparation and drinking habits. In the UK, coffee is often consumed with milk or sugar; indeed, data from the UK Biobank indicate that only 14.4% of coffee drinkers take it without any additives.27 The additives may lower the beverage temperature, thus reducing thermal exposure to the gastrointestinal mucosa. In addition, experimental evidence suggests that high temperature brewing can increase or release potentially carcinogenic compounds from tea.28

This study has several strengths, specifically the prospective design, large sample size, long term follow-up, and the availability of rigorous standardized data collection and organization. More importantly, approximately 99% of participants in this study reported drinking tea or coffee, allowing for a comprehensive evaluation across a wide spectrum of intake levels and preferred beverage temperatures. Several limitations should be clarified. First, we used only baseline data on tea and coffee consumption and preferred beverage temperature in the present analysis. Therefore, we cannot exclude the possibility of exposure misclassification due to changes in consumption habits during the follow-up period. Nonetheless, in a subsample of 20[thin space (1/6-em)]348 participants, tea and coffee consumption showed good reproducibility (κ = 0.83 and 0.78, respectively), and preferred beverage temperature also demonstrated acceptable agreement (κ = 0.68), supporting the reliability of baseline assessments.29 Second, the consumption of tea or coffee was obtained from baseline FFQ. Although more detailed 24-hour dietary recalls were available in a subset, using them would have reduced the sample size and statistical power. Previous studies showed strong correlations between FFQ and 24-hour recall data from a subset of 122[thin space (1/6-em)]283 participants who both completed baseline FFQ and ≥2 of five 24 h diet recall questionnaires (r = 0.82 for coffee, r = 0.81 for tea).29 Third, the preferred temperature of hot beverages relied on qualitative self-report data and was not validated with actual temperature measurement. To obtain a valid and reliable estimate of the temperature at which participants typically drink hot beverages is a challenge for most of the large-scale population-based epidemiological studies. Besides, previous study has provided evidence supporting the moderate agreement between questionnaire responses and measured temperatures (κ = 0.46).30 Fourth, we were also unable to estimate or adjust for H. pylori infection, which was only available for a small subset of participants (approximately 2%). To assess the potential impact of unmeasured confounding factors, we conducted E-value analyses, which suggested that only a confounder strongly associated with both exposure and outcome could fully explain the observed associations, supporting the robustness of our findings. Finally, the UK Biobank is not a nationally representative cohort, and healthy volunteer selection bias has been discussed previously, which may underestimate risks in socioeconomically disadvantaged populations.16 External validation with more representative cohorts and targeted recruitment for underrepresented groups would be valuable to confirm our findings across different socioeconomic strata.

Conclusions

The results of this study demonstrate a significant association between hot beverage consumption and increased risk of GC. These findings address the lack of evidence on the association between hot beverage consumption and GC risk in Western populations, and emphasize the potential preventive value of reducing intake of very hot beverages. Further studies in which the actual beverage temperature is measured and more detailed exposure information is collected are particularly encouraged.

Author contributions

Fengyi Huang, Changwei Li, Deqiang Zheng, and Fen Liu conceived and designed the study; Fengyi Huang, Xite Zheng, Xiaorui Zhang conducted the study; Fengyi Huang and Xite Zheng contributed to statistical analysis; Fengyi Huang drafted the manuscript; Fengyi Huang, Yanling Qi, Changwei Li, Deqiang Zheng, and Fen Liu critically reviewed and edited the manuscript. All authors approved the final manuscript, and accept responsibility for the decision to submit for publication.

Conflicts of interest

There is no conflict of interest.

Data availability

The data that support the findings of this study are available from the UK Biobank (https://www.ukbiobank.ac.uk/) and can be accessed with reasonable request.

Supplementary information (SI) is available. See DOI: https://doi.org/10.1039/d5fo03778f.

Acknowledgements

This research utilized the UK Biobank Resource (Application Number 95259). We extend our gratitude to the participants and the team involved in creating and maintaining this valuable resource.

References

  1. F. Bray, M. Laversanne, H. Sung, J. Ferlay, R. L. Siegel, I. Soerjomataram and A. Jemal, Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries, CA-Cancer J. Clin., 2024, 74, 229–263 CrossRef PubMed.
  2. P. Karimi, F. Islami, S. Anandasabapathy, N. D. Freedman and F. Kamangar, Gastric cancer: descriptive epidemiology, risk factors, screening, and prevention, Cancer Epidemiol. Biomarkers Prev., 2014, 23, 700–713 CrossRef PubMed.
  3. J. Powell and C. C. McConkey, Increasing incidence of adenocarcinoma of the gastric cardia and adjacent sites, Br. J. Cancer, 1990, 62, 440–443 CrossRef PubMed.
  4. P. Rawla and A. Barsouk, Epidemiology of gastric cancer: global trends, risk factors and prevention, Przegl. Gastroenterol., 2019, 14, 26–38 Search PubMed.
  5. M. Laszkowska, S. Rodriguez, J. Kim and C. Hur, Heavy alcohol use is associated with gastric cancer: analysis of the national health and nutrition examination survey from 1999 to 2010, Am. J. Gastroenterol., 2021, 116, 1083–1086 CrossRef PubMed.
  6. C. S. Yang, H. Wang, G. X. Li, Z. Yang, F. Guan and H. Jin, Cancer prevention by tea: Evidence from laboratory studies, Pharmacol. Res., 2011, 64, 113–122 CrossRef PubMed.
  7. A. D. Nigra, A. J. Teodoro and G. A. Gil, A decade of research on coffee as an anticarcinogenic beverage, Oxid. Med. Cell. Longevity, 2021, 2021, 4420479 CrossRef PubMed.
  8. D. Loomis, K. Z. Guyton, Y. Grosse, B. Lauby-Secretan, F. El Ghissassi, V. Bouvard, L. Benbrahim-Tallaa, N. Guha, H. Mattock and K. Straif, Carcinogenicity of drinking coffee, mate, and very hot beverages, Lancet Oncol., 2016, 17, 877–878 CrossRef PubMed.
  9. C. Yu, H. Tang, Y. Guo, Z. Bian, L. Yang, Y. Chen, A. Tang, X. Zhou, X. Yang, J. Chen, Z. Chen, J. Lv and L. Li, Hot tea consumption and its interactions with alcohol and tobacco use on the risk for esophageal cancer: a population-based cohort study, Ann. Intern. Med., 2018, 168, 489–497 CrossRef PubMed.
  10. M. Inoue-Choi, Y. Ramirez, C. O'Connell, A. Berrington de Gonzalez, S. M. Dawsey, C. C. Abnet, N. D. Freedman and E. Loftfield, Hot beverage intake and oesophageal cancer in the UK Biobank: prospective cohort study, Br. J. Cancer, 2025, 132, 652–659 CrossRef PubMed.
  11. J. S. Ren, N. D. Freedman, F. Kamangar, S. M. Dawsey, A. R. Hollenbeck, A. Schatzkin and C. C. Abnet, 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 CrossRef PubMed.
  12. Y. Huang, H. Chen, L. Zhou, G. Li, D. Yi, Y. Zhang, Y. Wu, X. Liu, X. Wu, Q. Song, L. Liu and D. Yi, Association between green tea intake and risk of gastric cancer: a systematic review and dose-response meta-analysis of observational studies, Public Health Nutr., 2017, 20, 3183–3192 CrossRef PubMed.
  13. B. Gantuya, D. Bolor, K. Oyuntsetseg, Y. Erdene-Ochir, R. Sanduijav, D. Davaadorj, T. Tserentogtokh, D. Azzaya, T. Uchida, T. Matsuhisa and Y. Yamaoka, New observations regarding Helicobacter pylori and gastric cancer in Mongolia, Helicobacter, 2018, 23, e12491 CrossRef PubMed.
  14. F. Pourfarzi, A. Whelan, J. Kaldor and R. Malekzadeh, The role of diet and other environmental factors in the causation of gastric cancer in Iran–a population based study, Int. J. Cancer, 2009, 125, 1953–1960 CrossRef PubMed.
  15. Y. Wang, H. Duan and H. Yang, A case-control study of stomach cancer in relation to Camellia sinensis in China, Surg. Oncol., 2015, 24, 67–70 CrossRef PubMed.
  16. A. Fry, T. J. Littlejohns, C. Sudlow, N. Doherty, L. Adamska, T. Sprosen, R. Collins and N. E. Allen, Comparison of sociodemographic and health-related characteristics of UK Biobank participants with those of the general population, Am. J. Epidemiol., 2017, 186, 1026–1034 CrossRef PubMed.
  17. Y. Cai, C. Hong, J. Han, L. Fan, X. Xiao, J. Xiao, Y. Wei, Y. Zhu, J. Tian, X. Zhu, M. Jin and X. Miao, Healthy dietary patterns, genetic risk, and gastrointestinal cancer incident risk: a large-scale prospective cohort study, Am. J. Clin. Nutr., 2024, 119, 406–416 CrossRef PubMed.
  18. S. W. Lee, Risk factors and prevention of stomach cancer, excluding helicobacter pylori, Korean J. Helicobacter Upper Gastrointest. Res., 2024, 24, 243–251 CrossRef PubMed.
  19. M. Banks, D. Graham, M. Jansen, T. Gotoda, S. Coda, M. di Pietro, N. Uedo, P. Bhandari, D. M. Pritchard, E. J. Kuipers, M. Rodriguez-Justo, M. R. Novelli, K. Ragunath, N. Shepherd and M. Dinis-Ribeiro, British Society of Gastroenterology guidelines on the diagnosis and management of patients at risk of gastric adenocarcinoma, Gut, 2019, 68, 1545–1575 CrossRef PubMed.
  20. T. J. VanderWeele and P. Ding, Sensitivity analysis in observational research: introducing the E-value, Ann. Intern. Med., 2017, 167, 268–274 CrossRef PubMed.
  21. M. Yi, X. Wu, W. Zhuang, L. Xia, Y. Chen, R. Zhao, Q. Wan, L. Du and Y. Zhou, Tea consumption and health outcomes: umbrella review of meta-analyses of observational studies in humans, Mol. Nutr. Food Res., 2019, 63, e1900389 CrossRef.
  22. S. Deandrea, R. Foschi, C. Galeone, C. La Vecchia, E. Negri and J. Hu, Is temperature an effect modifier of the association between green tea intake and gastric cancer risk?, Eur. J. Cancer Prev., 2010, 19, 18–22 CrossRef PubMed.
  23. N. Yioris, S. Ivankovic and T. Lehnert, Effect of thermal injury and oral administration of N-methyl-N’-Nitro-N-nitrosoguanidine on the development of esophageal tumors in Wistar rats, Oncology, 1984, 41, 36–38 CrossRef PubMed.
  24. R. Lin, D. Xiao, Y. Guo, D. Tian, H. Yun, D. Chen and M. Su, Chronic inflammation-related DNA damage response: a driving force of gastric cardia carcinogenesis, Oncotarget, 2015, 6, 2856–2864 CrossRef.
  25. R. D. Odze, Unraveling the mystery of the gastroesophageal junction: a pathologist's perspective, Am. J. Gastroenterol., 2005, 100, 1853–1867 CrossRef.
  26. E. Landais, A. Moskal, A. Mullee, G. Nicolas, M. J. Gunter, I. Huybrechts, K. Overvad, N. Roswall, A. Affret, G. Fagherazzi, Y. Mahamat-Saleh, V. Katzke, T. Kühn, C. La Vecchia, A. Trichopoulou, E. Valanou, C. Saieva, M. Santucci de Magistris, S. Sieri, T. Braaten, G. Skeie, E. Weiderpass, E. Ardanaz, M. D. Chirlaque, J. R. Garcia, P. Jakszyn, M. Rodríguez-Barranco, L. Brunkwall, E. Huseinovic, L. Nilsson, P. Wallström, B. Bueno-de-Mesquita, P. H. Peeters, D. Aune, T. Key, M. Lentjes, E. Riboli, N. Slimani and H. Freisling, Coffee and tea consumption and the contribution of their added ingredients to total energy and nutrient intakes in 10 european countries: benchmark data from the late 1990s, Nutrients, 2018, 10, 725 CrossRef PubMed.
  27. Z. Ye, M. Liu, S. Yang, Y. Zhang, Y. Zhang, P. He, C. Zhou, X. Gan, H. Xiang, Y. Huang, F. F. Hou and X. Qin, Coffee consumption with different additives and types, genetic variation in caffeine metabolism and new-onset acute kidney injury, Br. J. Nutr., 2024, 132, 1645–1653 CrossRef PubMed.
  28. S. M. Chiang, K. C. Ueng, H. S. Chen, C. J. Wu, Y. S. Yang and D. J. Yang, Effects of manufacturing procedures and preparation conditions on European Union priority polycyclic aromatic hydrocarbons in Oolong tea samples, Food Chem., 2021, 358, 129885 CrossRef PubMed.
  29. K. E. Bradbury, H. J. Young, W. Guo and T. J. Key, Dietary assessment in UK Biobank: an evaluation of the performance of the touchscreen dietary questionnaire, J. Nutr. Sci., 2018, 7, e6 CrossRef PubMed.
  30. M. M. Mwachiro, R. K. Parker, N. R. Pritchett, J. O. Lando, S. Ranketi, G. Murphy, R. Chepkwony, S. L. Burgert, C. C. Abnet, M. D. Topazian, S. M. Dawsey and R. E. White, Investigating tea temperature and content as risk factors for esophageal cancer in an endemic region of Western Kenya: Validation of a questionnaire and analysis of polycyclic aromatic hydrocarbon content, Cancer Epidemiol., 2019, 60, 60–66 CrossRef PubMed.
This journal is © The Royal Society of Chemistry 2026
Click here to see how this site uses Cookies. View our privacy policy here.