Modification role of tea in the association of dietary meat intake and obesity risk Authors: Sijia Cai, Chuhan Wu, Junhan Zhang, Xiaona Na, Yuefeng Tan, Ai Zhao Published: 2025 ABSTRACT AbstractBackgroundThere is no prospective epidemiologic evidence to support whether tea consumption has protective effect on the relationship between meat intake and obesity.MethodsThis study utilized data from the UK Biobank, including 206,142 participants for primary analysis. Dietary tea and meat intake was calculated based on food consumption from 24-hour dietary recall, and subsequently categorized into three groups (low, medium, and high) using tertiles. Time-dependent Cox models were used to estimate the hazard ratios (HRs) and 95% confidence intervals (CIs) for the separate and joint associations of dietary tea and meat intake with risk of incident obesity. To assess the potential modified effect of tea in the association of meat intake with the risk of incident obesity, we examined the both multiplicative and additive interactions. Stratified analysis was performed by levels of tea intake. Subgroup analyses were performed by sex, polygenic risk score (PRS), the Townsend Deprivation Index (TDI), age and cholesterol.ResultsDuring a median follow-up of 11.51 years, 11,627 incident obesity cases were identified. A near-linear dose-response relationship was observed between higher consumption of total, processed, and white meat—though not red meat—and increased obesity risk, independent of tea intake. For instance, even when accompanied by high tea consumption, high intake of each meat type remained significantly associated with elevated obesity risk (total meat: HR = 1.425, 95% CI: 1.286 to 1.579; red meat: HR = 1.217, 95% CI: 1.086 to 1.365; processed meat: HR = 1.521, 95% CI: 1.326 to 1.745; white meat: HR = 1.508, 95% CI: 1.314 to 1.732).No significant statistical or additive interactions were detected between tea and meat consumption. Consistent with this, among high meat consumers, even the highest level of tea intake did not attenuate the risk (total meat: HR = 1.483, 95% CI: 1.347 to 1.632; red meat: HR = 1.210, 95% CI: 1.086 to 1.348; processed meat: HR = 1.403, 95% CI: 1.243 to 1.584; white meat: HR = 1.536, 95% CI: 1.350 to 1.747). Subgroup analyses yielded generally consistent results but suggested potential sex-specific differences.ConclusionsThere is no interactive effect between meat and tea consumption, implying that tea intake could not serve as a remedy for high meat consumption. Independent of the amount of tea consumed, there is still an elevated risk when consuming high level of meat.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12937-025-01258-7.Keywords: Meat, Tea, Incident obesity risk, Modification effect Background The prevalence of obesity has nearly tripled since 1975 and is projected to increase to 1.02 billion (18% of adults) by 2030 [1]. Obesity is not a single disorder but a heterogeneous group of conditions with multiple causes. Body weight is determined by an interaction between genetic, environmental and psychosocial factors acting through the physiological mediators of energy intake and expenditure [2]. Among the above factors, dietary factors played a pivotal role in the onset and development of obesity, making them a key focus in the prevention and treatment of obesity. Meat is one of the staples of the human diet, which provides high-quality nutrients, but that also constitutes a relevant source of cholesterol and saturated fatty acids. The relationship between meat consumption and obesity has been widely discussed, with numerous studies highlighting a positive association between the intake of red and processed meats and the risk of obesity [3–5]. Meanwhile, certain foods are considered to have a preventive effect against obesity, among which, tea is one of the most acknowledged ones [6–9]. Tea, one of the world’s three major beverages, has been proven to be associated with lower risk of plenty health problems, which might attribute to its abundant bioactive substances, such as tea polyphenols, amino acids and polysaccharides [10–14]. Some mechanistic studies have suggested the functional components of tea may modulate obesity risk through adjustment on metabolic burden caused by overdose meat consumption. Specifically, tea may contribute to the antioxidant [15], anti-inflammatory [16], immune regulation [17, 18], and enzyme functional homeostasis [19]. For example, supplementation with green tea was shown to improve the malondialdehyde (MDA) and total antioxidant capacity (TAC) [20]. While high consumption of meat and meat products is associated with an increase in oxidative stress, particularly in the case of red meat and processed meat products [21]. What’s more, drinking tea is widely regarded as a digestive aid for oily food and is believed to promote fat loss—a practice embraced across various cultures and followed by diverse groups of people [22]. However, to the best of our knowledge, there is no prospective epidemiologic evidence to support whether tea consumption has protective effect on the relationship between meat intake and obesity. We, therefore, conducted the present study to investigate the modified role of tea in the association of dietary meat intake and obesity risk, through a longitudinal cohort that involves over half million adults in UK. Methods Study population This cohort study uses data from the UK Biobank. The UK Biobank is a large, population-based cohort study, which recruited approximately half a million participants aged 40–69 years in 2006–2010 across the England, Scotland, and Wales. Each participant completed touchscreen questionnaires, underwent physical examinations, and provided biological samples [23]. In current study design, the participants with no diagnosed obesity at baseline were recruited. We excluded participants who failed to complete dietary survey, or had a zero calorie intake. In total, our primary analysis involved 205,087 participants, among whom participants have simultaneous intake of tea and total meat (n = 129,415), tea and red meat (n = 76,511), tea and processed meat (n = 74,822), tea and white meat (n = 69,187) were included in the secondary analysis for exploring the joint associations of tea and meat intake. The flow path of inclusion and exclusion was presented in Fig. 1. Flowchart of participant enrollment Assessment of meat and tea intake Diet information was collected by a web-based 24-h recall questionnaire on five occasions between 2009 and 2012. We calculated cumulative mean values from the available data to represent long-term dietary intake and reduce within-person variation. In present study, the exposures were defined following the food group classification described by UK Biobank [24]. To be specific, pork, beef and lamb were defined as red meat; poultry and breaded/battered chicken were defined as white meat. Processed meat included sausage, bacon, and ham. Total meat included items above and additionally other meat (offal). Tea and meat consumption was examined by tertiles in all models. The three tertiles are defined as low, medium, and high, respectively. Assessment of obesity Incident obesity was determined through the analysis of the UK Biobank health data that included primary care data, hospital inpatient records, death register entries, and self-reported medical conditions. Additionally, hospital-admission data were available until 30 November 2022, and death register data were available up to 30 November 2022 for England, 31 July 2021 for Scotland, and 28 February 2018 for Wales. Therefore, the follow-up period was calculated from the baseline to the first record of incident obesity, death, loss to follow-up, or the last follow-up date, whichever occurred first. Assessment of covariates The baseline questionnaire was used to assess the following potential confounders: (1) sociodemographic factors: age, sex, ethnicity, education qualifications, Townsend Deprivation Index (TDI); (2) lifestyles: physical activity, sleep duration, smoking status, alcohol status, dietary intake (vegetables, fruits and cereal); (3) genetic factor: polygenic risk score (PRS) for BMI, which provided directly from database. (4) biochemistry markers: cholesterol, which was measured in the blood sample collected at recruitment. (5) health-related outcomes: hypertension is defined as essential (primary) hypertension, coded as I10; type 2 diabetes is defined as non-insulin-dependent diabetes mellitus, coded as E11. Ethnicity was categorized as White, Asian or Asian British, Black or Black British, and others. The TDI is an indicator of socioeconomic status, and positive TDI values indicate relative privation. Physical activity was categorized as low, moderate, and high level, according to the guidelines of the International Physical Activity Questionnaire (IPAQ) [25]. Smoking or alcohol status were categorized as current, former, or never user. Information on total energy intake, cereal, fruit and vegetable consumption was obtained from the 24-hour dietary recall. The cumulative averages of food intake from baseline to the censoring events were used [26]. The standard PRS for BMI was calculated by meta-analyzing two external genome wide association study sources, which could be used for assessing the obesity genetic susceptibility, and a higher PRS indicated more risk for obesity [27]. Statistical analysis In descriptive analyses, continuous variables were expressed as means (standard deviations) for normally distributed variables or medians (interquartile ranges) for skewed variables, and categorical variables were represented by frequency and percentage. The difference between groups were compared by a Student t test, Wilcoxon test, or χ2 test. It is assumed that data were missing at random, and we used multiple imputation chain-equation (MICE) to impute missing covariate values [28]. We generated 5 datasets with 100th iteration by random forest method. We evaluated the separate and joint associations of tertiles of tea and meat intake with the risk of incident obesity. We used Time-dependent Cox proportional hazards regression to estimate cause-specific hazard ratios (HRs) and 95% confidence intervals (CIs) for associations of obesity risk with tea and meat intake. To test the linear trend, we used a term for tertile number and included it as a continuous variable in the models. We examined the assumption of proportionality of the hazards using Schoenfeld residuals. In model1, we adjusted for baseline age, sex (female or male), ethnicity (White, Asian or Asian British, Black or Black British, and other), PRS, education qualifications (degree or no degree), TDI. Model2 was further adjusted for physical activity (low, moderate and high level), sleep(integer), smoking status (current, former, or never), alcohol consumption (continuous) based on model1. Model3 (fully adjusted model) was further adjusted for energy intake(continuous), cereal(continuous), fruit(continuous) and vegetable(continuous) consumption based on model2. The lowest tertile of tea and the lowest tertile of meat intake was the reference group in the separate analysis. The lowest tertile of tea combined with the lowest tertile of meat was the reference group in the joint analysis. We assessed the dose-response relationships of meat and tea intake with the outcomes using restricted cubic spline regression with four knots. To assess the potential modified effect of tea in the association of meat intake with the risk of incident obesity, we examined the both multiplicative and additive interactions. Multiplicative interaction between tea and meat intake was assessed by the likelihood ratio test (LRT), where a product term was included. The significance threshold was adjusted through the Bonferroni method. The additive interaction term was assessed with two indices: the relative excess risk due to the interaction (RERI) and the attributable proportion (AP) due to the interaction [29]. When 0 was within the CIs of the RERI and AP, it meant that there was no additive interaction. Stratified analysis was performed by levels of tea intake. Subgroup analyses were performed by sex (male, female), PRS level (low, high), TDI level (low, high), age (< 60, ≥ 60 years) and cholesterol (5.2 mmol/L as the cut-off point). Furthermore, several sensitivity analyses were also conducted to test the robustness of our findings. First, we excluded participants who had an obesity event during the first year of follow-up to minimize the influence of reverse causation. Second, we repeated main analyses in participants who did not have type 2 diabetes or hypertension at baseline. Third, we added the number of questionnaires as a covariate. All statistical analyses were done in R (version 4.3.1). p values were two-sided, with statistical significance set at p < 0.05. Results Baseline characteristics Baseline characteristics of participants across quartile of tea and meat consumption were showed in Table 1. Compared to the low tea intake group, the high intake group was more likely to be female, less educated, had a higher physical activity level and lower alcohol intake. In terms of total meat, compared to the low intake group, the high intake group was more likely to be older and White race, had a higher PRS level, longer sleep duration, lower TDI and lower physical activity level. Table 1Baseline characteristics of participants across the different tea and meat intake levelTea intakeMeat intakeCharacteristicsLow(n = 74426)Medium(n = 70366)High(n = 60295) P Low(n = 71947)Medium(n = 69714)High(n = 63426) P Sex, n (%)< 0.001< 0.001 Female41,530 (55.8)38,182 (54.3)33,343 (55.3)43,864 (61.0)38,758 (55.6)30,433 (48.0) Male32,896 (44.2)32,184 (45.7)26,952 (44.7)28,083 (39.0)30,956 (44.4)32,993 (52.0)Age, mean (SD), years55.3 (8.16)56.3 (7.91)56.6 (7.69)< 0.00156.0 (7.96)56.6 (7.81)55.5 (8.08)< 0.001Ethnicity, n (%)< 0.001< 0.001 White70,261 (94.4)67,148 (95.4)58,922 (97.7)68,176 (94.8)67,550 (96.9)60,605 (95.6) Asian or Asian British1035 (1.4)1361 (1.9)495 (0.8)1609 (2.2)565 (0.8)717 (1.1) Black or Black British1649 (2.2)706 (1.0)197 (0.3)905 (1.3)631 (0.9)1016 (1.6) Other1481 (2.0)1151 (1.6)681 (1.1)1257 (1.7)968 (1.4)1088 (1.7)Education, n (%)< 0.0010.594 No degree6005 (8.1)5249 (7.5)6013 (10.0)6010 (8.4)5927 (8.5)5330 (8.4) Degree68,421 (91.9)65,117 (92.5)54,282 (90.0)65,937 (91.6)63,787 (91.5)58,096 (91.6)Polygenic Risk Score, mean (SD)−0.205(0.977)−0.244(0.975)−0.248(0.975)< 0.001−0.236(0.974)−0.243(0.973)−0.213(0.981)< 0.001Townsend Deprivation Index, mean (SD)−1.37 (2.99)−1.68 (2.81)−1.72 (2.78)< 0.001−1.39 (2.96)−1.73 (2.79)−1.64 (2.85)< 0.001Physical activity level, n (%)< 0.001< 0.001 Low13,734 (18.5)12,035 (17.1)10,487 (17.4)11,782 (16.4)12,545 (18.0)11,929 (18.8) Moderate31,237 (42.0)30,797 (43.8)25,730 (42.7)30,645 (42.6)30,361 (43.6)26,758 (42.2) High29,455 (39.6)27,534(39.1)24,078 (39.9)29,520 (41.0)26,808 (38.5)24,739 (39.0)Sleep duration, mean (SD), hours/day7.13 (1.04)7.18 (0.974)7.18 (1.00)< 0.0017.14 (1.02)7.18 (0.995)7.15 (1.01)< 0.001Smoking status, n (%)< 0.001< 0.001 Never42,029 (56.5)40,264(57.2)34,273 (56.8)41,682 (57.9)39,744 (57.0)35,140 (55.4) Former25,848 (34.7)25,348 (36.0)21,241 (35.2)25,016 (34.8)24751 (35.522,670 (35.7) Current6549 (8.8)4754 (6.8)4781 (7.9)5249 (7.3)5219 (7.5)5616 (8.9)Alcohol, mean (SD), g/day17.5 (23.1)17.9 (21.8)15.3 (20.5)< 0.00113.9 (19.6)17.1 (21.0)20.4 (24.7)< 0.001Energy, mean (SD), kJ/day8430 (2570)8680 (2450)8860 (2550)< 0.0018080 (2440)8530 (2210)9400 (2760)< 0.001Tea intake, mean (SD), g/day: Low, 52.84(78.36); Medium, 438.88(104.82); High, 867.17(169.11)Meat intake, mean (SD), g/day: Low, 19.56(23.07); Medium, 95.35(19.79); High, 170.28(60.36) Baseline characteristics of participants across the different tea and meat intake level Tea intake, mean (SD), g/day: Low, 52.84(78.36); Medium, 438.88(104.82); High, 867.17(169.11) Meat intake, mean (SD), g/day: Low, 19.56(23.07); Medium, 95.35(19.79); High, 170.28(60.36) Associations of tea and meat intake with incident of obesity During a median follow-up of 11.51 years, 11,627 obesity cases were developed. We used time-varying Cox proportional hazard models to examine the associations of tea and different meat groups with the obesity risk (Table 2). In all models, compared to low intake level, higher tea intake was significantly associated with lower incidence of obesity, while higher total meat intake was associated with higher risk obesity. After multivariable adjustments (model3) including socioeconomic, lifestyle and dietary factors, the HR for obesity in participants with the highest intake of tea compared with lowest group was 0.870 (95% CI: 0.831 to 0.910). In contrast, highest intakes of total, red, processed and white meat groups were associated with highest risk of obesity, which were 1.458 (95% CI: 1.391 to 1.527), 1.164 (95% CI: 1.118 to 1.211), 1.161 (95% CI: 1.116 to 1.209) and 1.118 (95% CI: 1.074 to 1.165), respectively. In our analysis, medium intake levels of processed meat and white meat tended to be associated with a lower incidence of obesity. Conversely, for medium intake levels of red meat, we did not detect a statistically significant association with obesity incidence. Results from the multivariable adjusted restricted cubic spline regression showed that the risk of incident obesity changed non-linearly with increasing intake of tea, whereas the association between meat and incident obesity was more linear (Figure S1). Table 2Association of obesity with tea and meat intakeModel1Model2Model3Tea Low(n = 74426)1(ref)1(ref)1(ref) Medium(n = 70366)0.841(0.806,0.879)0.848(0.812,0.886)0.838(0.803,0.876) High(n = 60295)0.882(0.844,0.923)0.885(0.846,0.925)0.870(0.831,0.910)Total meat Low(n = 71947)1(ref)1(ref)1(ref) Medium(n = 69714)1.197(1.143,1.254)1.196(1.142,1.253)1.174(1.121,1.230) High(n = 63426)1.481(1.415,1.549)1.481(1.415,1.549)1.458(1.391,1.527)Red meat Low(n = 107316)1(ref)1(ref)1(ref) Medium(n = 29949)0.976(0.923,1.033)0.977(0.924,1.035)0.965(0.912,1.021) High(67822)1.200(1.154,1.249)1.201(1.154,1.250)1.164(1.118,1.211)Processed meat Low(n = 110068)1(ref)1(ref)1(ref) Medium(n = 26704)0.922(0.868,0.979)0.926(0.872,0.984)0.920(0.866,0.978) High(n = 68315)1.196(1.150,1.244)1.188(1.142,1.236)1.161(1.116,1.209)White meat Low(116069)1(ref)1(ref)1(ref) Medium(n = 29246)0.910(0.860, 0.963)0.911(0.860,0.964)0.914(0.864,0.968) High(n = 59772)1.103(1.059,1.148)1.103(1.059,1.149)1.118(1.074,1.165)Model 1 was adjusted for baseline age, sex (female or male), ethnicity (White, Asian or Asian British, Black or Black British, and other), Polygenic Risk Score (PRS), education qualifications (degree or no degree), Townsend Deprivation Index (TDI) Model 2 was further adjusted for physical activity (low, moderate and high level), sleep (integer), smoking status (current, former, or never), alcohol consumption (continuous) based on model1Model 3 was further adjusted for energy intake (continuous), cereal (continuous), fruit (continuous), vegetable (continuous) consumption based on model2Red meat intake, mean (SD), g/day: Low, 0.00(0.00); Medium, 33.56(8.50); High, 99.66(45.88)Processed meat intake, mean (SD), g/day: Low, 0.00(0.00); Medium, 10.08(3.72); High, 49.58(33.59)White meat intake, mean (SD), g/day: Low, 0.00(0.00); Medium, 33.53(9.00); High, 101.70(50.67) Association of obesity with tea and meat intake Model 1 was adjusted for baseline age, sex (female or male), ethnicity (White, Asian or Asian British, Black or Black British, and other), Polygenic Risk Score (PRS), education qualifications (degree or no degree), Townsend Deprivation Index (TDI) Model 2 was further adjusted for physical activity (low, moderate and high level), sleep (integer), smoking status (current, former, or never), alcohol consumption (continuous) based on model1 Model 3 was further adjusted for energy intake (continuous), cereal (continuous), fruit (continuous), vegetable (continuous) consumption based on model2 Red meat intake, mean (SD), g/day: Low, 0.00(0.00); Medium, 33.56(8.50); High, 99.66(45.88) Processed meat intake, mean (SD), g/day: Low, 0.00(0.00); Medium, 10.08(3.72); High, 49.58(33.59) White meat intake, mean (SD), g/day: Low, 0.00(0.00); Medium, 33.53(9.00); High, 101.70(50.67) Stratified analyses Stratified by tea intake levels (Fig. 2, and Table S1), the positive dose-response associations between higher consumption of total, processed, and white meat—though not red meat—and increased obesity risk remained consistent, regardless of tea consumption. Notably, among participants with high tea intake, medium red meat consumption was not significantly associated with obesity (HR = 0.933, 95% CI: 0.775 to 1.123). Stratified hazard ratios for incident obesity by meat intake levels and tea intake levels. Adjusted for baseline age, sex(female or male), ethnicity (White, Asian or Asian British, Black or Black British, and other), Polygenic Risk Score(PRS), education qualifications (degree or no degree), Townsend Deprivation Index (TDI), physical activity (low, moderate and high level), sleep(integer), smoking status (current, former, or never), alcohol consumption (continuous), energy intake (continuous), cereal (continuous), fruit (continuous), vegetable (continuous). P refers to the p-trend of tea intake. Figure a represents total meat intake, Figure b represents red meat intake, Figure c represents processed meat intake and Figure d represents white meat intake Joint analyses No evidence of statistical interaction was found between tea and meat intake (Table 3). No evidence of additive interaction was found between tea and meat intake (Table 4). Table 3Main and interaction effects of meat and tea intake to incident obesity riskaMain effectInteraction effectTotal meatHazard Ratio(95%CI)Total meat×TeaHazard Ratio(95%CI)P-interactionb, c Q11(ref)Q1Low1(ref)0.692 Q21.174(1.121,1.230)Q2Medium1.047(0.901,1.217)High0.993(0.861,1.144) Q31.458(1.391,1.527)Q3Medium1.070(0.925,1.237)High0.957(0.836,1.097)Red meatHazard Ratio(95%CI)Red meat×Tea Q11(ref)Q1Low1(ref)0.059 Q20.965(0.912,1.021)Q2Medium1.118(0.895,1.398)High0.773(0.610,0.980) Q31.164(1.118,1.211)Q3Medium1.047(0.888,1.233)High0.970(0.832,1.131)Processed meatHazard Ratio(95%CI)Processed meat×Tea Q11(ref)Q1Low1(ref)0.765 Q20.920(0.866,0.978)Q2Medium0.982(0.810,1.190)High1.068(0.884,1.289) Q31.161(1.116,1.209)Q3Medium1.051(0.872,1.266)High1.118(0.934,1.338)White meatHazard Ratio(95%CI)White meat×Tea Q11(ref)Q1Low1(ref)0.823 Q20.914(0.864,0.968)Q2Medium0.910(0.748,1.108)High0.955(0.783,1.164) Q31.118(1.074,1.165)Q3Medium0.964(0.793,1.172)High1.037(0.860,1.250)aAdjusted for baseline age, sex(female or male), ethnicity (White, Asian or Asian British, Black or Black British, and other), Polygenic Risk Score(PRS), education qualifications (degree or no degree), Townsend Deprivation Index (TDI), physical activity (low, moderate and high level), sleep(integer), smoking status (current, former, or never), alcohol consumption (continuous), energy intake (continuous), cereal(continuous), fruit(continuous), vegetable(continuous)bLikelihood ratio test was used to examine the statistical interaction between meat and tea intakecBonferroni adjusted significance threshold = 0.05/3 (3 levels of intake) = 0.017 Main and interaction effects of meat and tea intake to incident obesity riska aAdjusted for baseline age, sex(female or male), ethnicity (White, Asian or Asian British, Black or Black British, and other), Polygenic Risk Score(PRS), education qualifications (degree or no degree), Townsend Deprivation Index (TDI), physical activity (low, moderate and high level), sleep(integer), smoking status (current, former, or never), alcohol consumption (continuous), energy intake (continuous), cereal(continuous), fruit(continuous), vegetable(continuous) bLikelihood ratio test was used to examine the statistical interaction between meat and tea intake cBonferroni adjusted significance threshold = 0.05/3 (3 levels of intake) = 0.017 Table 4Additive effects of meat and tea intake to incident obesity riskMeat intakeTea intakeMediumHighRERI (95% CI)AP (95% CI)RERI (95% CI)AP (95% CI)Total meat Medium0.043 (−0.118, 0.203)0.036 (−0.184, 0.256)0.003 (−0.157, 0.163)0.002 (−0.211, 0.215) High0.071 (−0.101, 0.242)0.048 (−0.141, 0.237)−0.037 (−0.205, 0.132)−0.025 (−0.220, 0.171) Red meat Medium0.134 (−0.132, 0.400)0.103 (−0.129, 0.335)−0.273 (−0.525, −0.020)−0.289 (−0.678, 0.101) High0.051 (−0.135, 0.238)0.041 (−0.174, 0.256)−0.032 (−0.207, 0.142)−0.027 (−0.251, 0.198)Processed meat Medium−0.029 (−0.226, 0.169)−0.028 (−0.340, 0.285)0.063 (−0.132, 0.258)0.055 (−0.222, 0.333) High0.040 (−0.171, 0.251)0.031 (0.227, 0.289)0.126 (−0.075, 0.327)0.090 (−0.146, 0.326) White meat Medium−0.101 (−0.316, 0.114)−0.092 (−0.425, 0.241)−0.048 (−0.264, 0.167)−0.043 (−0.360, 0.274) High−0.033 (−0.277, 0.211)−0.023 (−0.291, 0.246)0.070 (−0.158, 0.298)0.046 (−0.199, 0.290)Adjusted for baseline age, sex(female or male), ethnicity (White, Asian or Asian British, Black or Black British, and other), Polygenic Risk Score(PRS), education qualifications (degree or no degree), Townsend Deprivation Index (TDI), physical activity (low, moderate and high level), sleep(integer), smoking status (current, former, or never), alcohol consumption (continuous), energy intake (continuous), cereal(continuous), fruit(continuous), vegetable(continuous)RERI: the relative excess risk due to the interactionAP: the attributable proportion Additive effects of meat and tea intake to incident obesity risk Adjusted for baseline age, sex(female or male), ethnicity (White, Asian or Asian British, Black or Black British, and other), Polygenic Risk Score(PRS), education qualifications (degree or no degree), Townsend Deprivation Index (TDI), physical activity (low, moderate and high level), sleep(integer), smoking status (current, former, or never), alcohol consumption (continuous), energy intake (continuous), cereal(continuous), fruit(continuous), vegetable(continuous) RERI: the relative excess risk due to the interaction AP: the attributable proportion In the joint analyses by tea and meat intake level (Fig. 3), the increased risks were shown within the groups of low tea intake combined with medium and high meat intake, medium tea intake combined with high meat intake, high tea intake combined with high meat intake (total, red, processed and white meat). Among participants with high meat intake, there was an increased risk even with highest doses of tea. Association of meat intake and tea consumption with incident obesity. Adjusted for baseline age, sex (female or male), ethnicity (White, Asian or Asian British, Black or Black British, and other), Polygenic Risk Score (PRS), education qualifications (degree or no degree), Townsend Deprivation Index (TDI), physical activity (low, moderate and high level), sleep (integer), smoking status (current, former, or never), alcohol consumption (continuous), energy intake (continuous), cereal (continuous), fruit (continuous), vegetable (continuous) Sensitivity and subgroup analyses In the sensitivity analyses (Figure S2, Figure S3 and Figure S4), results remained consistent when we excluded participants who developed obesity during the first year of follow-up. Among participants who did not have type 2 diabetes or hypertension at baseline, the increased risk disappeared when low tea intake combined with medium level of red meat (HR = 1.210, 95% CI: 0.987 to 1.483), processed meat (HR = 1.151, 95% CI: 0.976 to 1.358) and white meat (HR = 1.163, 95% CI: 0.977 to 1.385). After additionally controlling for the number of recall times, we found that while the elevated risk trends persisted, they were no longer statistically significant when low tea intake combined with medium level of processed (HR = 1.127, 95% CI: 0.995 to 1.277) and white meat (HR = 1.116, 95% CI: 0.978 to 1.273), and high tea intake combined with medium level of processed meat (HR = 1.125, 95% CI: 0.987 to 1.282). Subgroup analyses across sex, genetic risk, socioeconomic status, age, and cholesterol levels (5.2 mmol/L as the cut-off point) generally reaffirmed the main findings (Figure S3). However, several associations became nonsignificant in specific subgroups. For instance, among females, multiple combinations of medium or high meat and tea intake lost statistical significance. A protective effect was observed in females (HR = 0.855, 95% CI: 0.737 to 0.992) and participants under 60 years (HR = 0.838, 95% CI: 0.716 to 0.982) with both medium tea and low total meat consumption. Additionally, participants with elevated cholesterol levels showed no significant risk increase with medium or high red meat intake when tea consumption was low. Discussion In this population-based study, we investigated the separate and joint associations of dietary tea and meat intake with incident obesity risk, and further examined the interaction between meat and tea intake using both multiplicative and additive interaction models. We did not find a statistical interaction between meat and tea intake in obesity development. Although the highest dose of tea was still associated with an increased obesity risk in participants with high meat intake, the subgroup analysis revealed a sex-specific association. The association between high meat consumption and increased obesity risk, as well as the potential protective role of tea intake, has been widely reported in epidemiological studies. However, in the present study, we did not observe significant statistical or additive interaction between tea and meat consumption on obesity risk. This suggests that, within our population, the combined effect of tea and meat intake does not deviate from their independent effects. While several overlapping biological mechanisms—such as oxidative stress response, lipid metabolism, and gut microbiota—have been proposed to underlie diet–obesity relationships, our null interaction findings indicate that such pathways may not play a substantial role in modifying the joint effect of tea and meat consumption in this cohort. For instance, although experimental studies suggest that meat-derived advanced glycation end products (AGEs) might suppress Nrf2-mediated antioxidant responses [30–32], and tea polyphenols such as EGCG may activate Nrf2 [33], our results do not provide epidemiological support for a meaningful interplay between these mechanisms in affecting obesity risk. Further research is needed to clarify the conditions under which these mechanistic pathways may become clinically relevant in human populations. Although, the above mechanism support there is an interaction between tea and meat to the obesity, our results showed that, in general, tea intake could not offset the obesity risks caused by meat consumption. The current results demonstrated that among all the combinations, the increased risks disappeared only among medium level of red meat combined with high level of tea intake. At the same time, based on the results of the joint analyses, we can see that meat appears to have a greater impact on obesity risk than tea. The variations of tea consumption brought no substance change of HRs, when it sticked to same meat consumption. Conversely, incremental meat consumption contributed to more evident augment of HRs, when tea consumption stayed similar. A dose-response relationship was observed in some cases—for example, at medium tea intake levels, HRs rose progressively with low (HR = 0.945, 95%CI: 0.845 to 1.056), medium (HR = 1.179, 95%CI: 1.061 to 1.311), and high (HR = 1.470, 95%CI:1.327 to 1.628) meat consumption. Theoretically, tea consumption has the ability to counteract the risks of obesity caused by meat consumption through anti-inflammatory, antioxidant, adipose tissue reduction, etc. However, when there is a high meat intake, the protective effects of tea are not shown. On the one hand, this could be due to the fact that there were not enough antioxidants and anti-inflammatory substances from tea to quench the free radicals and inflammation derived from meat. On the other hand, a systematic review and meta-analysis found that a kind of antioxidant supplements demonstrated significant antioxidant effects in the short term. However, the long-term intervention effects were not significant [34], suggesting that the effects of antioxidants may require accumulation over a period before they become apparent, while there might be a time lag between tea and meat consumption in our cohort. The lack of tea types examined in this study may have contributed to the null results. However, due to the absence of comparative data on different teas in this dataset, future studies are needed to conduct detailed comparisons based on the classification of tea types. Additionally, it may also possibly due to the low prevalence of obesity in our cohort. This study additionally accounts for differences across populations. Overall, the moderating effect of tea intake on the relationship between meat intake and the risk of incident obesity remained similar. Interestingly, our further analyses stratified by sex suggested that women had lower obesity risk than men when consuming all combinations of red meat and tea. Similarly, women had lower obesity risk than men when consuming all combinations of total meat and tea except for medium level of total meat intake combined with low level of tea. It was even showing a protective effect (HR = 0.855, 95%CI: 0.737 to 0.992) within low level of total meat intake combined with medium level of tea intake among women. It is plausible that estrogen could account for such results, given its known antioxidant properties [35] and potential to act synergistically with antioxidant substances in tea. However, we did not measure estrogen levels in the present study, which limits our ability to explore potential hormonal pathways; therefore, these associations may be chance findings and further mechanistic studies are needed to confirm this hypothesis. Contrary to expectations, our subgroup analysis revealed that the increased risks within groups of low tea intake combined with medium and high level of red meat disappeared among participants with abnormal cholesterol. This may be because individuals with high cholesterol are more likely to engage in proactive health management—such as taking medications or dietary supplements—which reduces their obesity risk and thereby obscures the effects of tea and red meat consumption on obesity. Overall, the net effect of tea and meat is not simply offset, and it may depend on factors such as dosage, metabolic environment, and temporal dynamics. For instance, the composition of the baseline gut bacteria is closely related to the individual’s metabolic response to green tea and the presence of certain specific gut bacteria may affect the body’s absorption and metabolism of polyphenolic substances. Some strains of Clostridium can decompose tea polyphenol components such as catechins, producing biologically active metabolites, such as 3,4-dihydroxybenzoic acid [36]. Future research is crucial for clarifying the underlying mechanisms between tea and meat intake in obesity prevention to support individualized dietary interventions. The present study indicates the potential benefits of a healthy diet for obesity prevention stem from the combined interactions and synergistic effects of various dietary components, enhancing overall health advantages that surpass the benefits of relying solely on the protective properties of individual foods. Our results highlight that when meat is habitually consumed in excessive quantities, the mitigating effects of other foods are relatively constrained. However, it should be noticed that some counterintuitive associations such as medium intake levels of processed meat and white meat tended to be associated with a lower incidence of obesity, it may partly be due to chance, particularly given the number of comparisons made. Residual confounding is also possible; for instance, individuals consuming moderate amounts of these meats may overall have healthier lifestyle patterns that we were unable to fully account for. Furthermore, dietary substitution likely plays an important role. It is plausible that moderate consumption of processed or white meat may often replace the intake of other, more obesogenic food sources—such as refined carbohydrates or ultra-processed foods—rather than simply being added to the diet. This potential substitution effect could result in a seemingly protective association, even if the meat itself is not inherently beneficial. Therefore, these observations should not be interpreted as evidence of a beneficial effect of meat consumption. Rather, they highlight the complexity of dietary patterns and energy balance. Further studies, particularly with detailed dietary substitution models and larger sample sizes, are needed to clarify these associations. Strengths and limitations To the best of our knowledge, it is the first study to explore the joint associations between tea and meat consumption and to investigate the interaction effects. This study has several strengths. Firstly, we involved almost half a million participants in UK, which provided adequate statistical power and reliability. Secondly, stratified analyses were conducted across several socio-demographic factors, to explore the population heterogeneity. Thirdly, we conducted several sensitivity analyses whose results showed high reliability and robustness of our study. Nevertheless, we also acknowledge several limitations. First, diet intake was measured using 24 h dietary recall methods, although the reliability and validity of this method have been thoroughly verified previously, potential misclassification may still exist. For example, this method is still more susceptible to capturing day-to-day variation for meal-specific foods like meat than for consistently consumed items like tea. Second, we could not confirm the accurate time point of tea and meat intake, therefore impossible to know whether the consumption of tea and meat at the same time has an extra effect. Third, we considered different types of meat, but the types of tea were not subdivided due to database limitations. Fourth, the number of participants and events might be insufficient among the non-white subgroup in the UK Biobank, and the results should be cautiously generalized; future studies are needed to be conducted in other populations to confirm the relevant conclusions. Fifth, in this study, we used ICD-10 as major outcome. In further experimental studies, more comprehensive and precise measures of obesity, such as BMI and body composition should be considered to thoroughly investigate the relationship between tea and meat consumption and obesity. Finally, although we controlled for key personal characteristics, residual confounding was still possible and causal inference cannot be made because of the nature of observational studies. Conclusions In conclusion, our study found that higher meat consumption was associated with elevated incident obesity risk, although higher tea consumption contributed to lower obesity risk, but there is no interactive effect between them, implying that tea intake could not serve as a remedy for high meat consumption. Future research should explore the roles and interactions of tea and meat intake using different types of tea extended to wider populations to provide insights into obesity prevention mechanisms and correct incorrect dietary rumors. Supplementary Information Supplementary Material 1: Tables S1 and Figures S1-S5. Acknowledgements This research has been conducted using the UK Biobank Resource (https://www.ukbiobank.ac.uk) under Application ID: 91486. The authors would like to express gratitude to the UKB investigators, staff, and participants, and acknowledge the valuable contributions of the staff and analysts involved in the UKB access management system. Clinical trial number Not applicable. Abbreviations Hazard ratios Confidence intervals Polygenic risk score UK Biobank International Classification of Diseases 10th revision Townsend Deprivation Index International Physical Activity Questionnaire Multiple imputation chain-equation Authors’ contributions All authors conceived and supervised the study. SC completed data analyses and drafted the manuscript with significant contributions from CW, JZ, XN, YT, and AZ. SC and AZ have full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. All authors mentioned above made substantial contributions to the content of the paper. All authors read and approved the final manuscript. Funding No funding. Data availability No datasets were generated or analysed during the current study. Declarations Ethics approval and consent to participate Ethical approval for the original UK Biobank study was granted by Northwest Multi-Centre Research Ethics Committee (11/NW/0382). Informed consent was obtained from each participant. Our study was conducted in accordance to the Declarations of Helsinki. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1: Tables S1 and Figures S1-S5. Data Availability Statement No datasets were generated or analysed during the current study.