A comprehensive literature search was conducted in PubMed, Embase, Cochrane Library, Web of Science, China National Knowledge Infrastructure (CNKI), Chongqing VIP Chinese Science and Technology Journal Database, Wanfang Database, Chinese Biomedical Literature Database (CBM), and Yiigle Database from their inception until July 22, 2024, limited to studies published in Chinese or English. The search strategy utilized terms including “red meat,” “beef,” “mutton,” “colitis, ulcerative,” and “ulcerative colitis.” Detailed search strategies are provided in Supplementary Table S1. To minimize the risk of missing relevant studies, additional manual screening of reference lists from included studies and pertinent reviews was performed, complemented by expert consultation.

The inclusion criteria were established according to the Population, Exposure, Comparator, Outcomes, and Study Design (PECOS) framework. Detailed inclusion criteria are presented in Table 1. The following studies were excluded: (1) studies evaluating dietary patterns explicitly controlling for red or processed meat intake, such as the Mediterranean diet or low-fat diet; (2) studies without clear specification of nutrient origin; and (3) duplicate publications or studies where the full text was inaccessible.

Retrieved literature was imported into NoteExpress 3.4 software, and duplicate publications were removed. Two reviewers (YYZ, ZYJ) independently conducted initial screening based on titles and abstracts, followed by a comprehensive full-text review. Subsequently, reviewers (ZJA, YHD, ZYZ, ZYJ) independently extracted data using a pre-designed data extraction form. The data extraction form included study identification, study design, sample size, participant demographics (age, sex), diagnostic criteria, country, type and dose of red or processed meat, dietary assessment tools, follow-up duration, effect sizes [odds ratio (OR), hazard ratio (HR), or relative risk (RR)] with their 95% confidence intervals (CIs), and adjustment factors. When multiple estimates with different adjustments were available, the estimate adjusted for the most covariates was selected. Any discrepancies during screening or data extraction were resolved by consulting a third reviewer (TYM).

The quality of the included studies was assessed independently by two reviewers (YYZ, JHY) using the Newcastle-Ottawa Scale (NOS) for cohort and case–control studies (15). The NOS evaluates three domains: selection of study population, comparability of groups, and assessment of exposure or outcomes. A maximum of 9 points can be awarded using the NOS, with scores ≥7 considered high quality, 5–6 as moderate quality, and <5 as low quality (16). Two reviewers (YYZ, JHY) also independently assessed the certainty of evidence for the main findings using the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) approach (17). The certainty of evidence was evaluated across five domains for downgrading (risk of bias, inconsistency, indirectness, imprecision, and publication bias) and three domains for upgrading (large magnitude of effect, dose–response gradient, and effect of plausible confounding). Any disagreements were resolved through discussion with a third reviewer (TYM).

All statistical analyses were conducted using the “meta” and “metafor” packages in R version 4.4.2 within the R Studio environment. For the meta-analysis, the highest category of red or processed meat intake was compared to the lowest (reference group) using a random-effects model. Given the low incidence of UC (below 10%), the HR and OR are considered to approximate the RR. Therefore, both HR and OR were treated as equivalent to RR for the purpose of effect size pooling in this meta-analysis (18). Heterogeneity was assessed using the Q test and the I² statistic, with I² < 30% indicating low heterogeneity. A two-sided p-value < 0.05 was considered statistically significant.

Where data from studies with more than three intake categories were available, a dose-response meta-analysis was conducted. Generalized least squares regression and restricted cubic spline models with knots at the 10th, 50th, and 90th percentiles were applied. Intake levels were estimated based on reported upper and lower category boundaries. For open-ended intake ranges, it was assumed that the width of the interval equaled that of the adjacent category. When portion sizes were reported instead of weight, one serving was standardized to 100 g of red meat and 50 g of processed meat (19).

Subgroup analyses and meta-regression were performed to explore sources of heterogeneity, considering variables such as study design, dietary pattern (Western vs. Eastern), meat type, dietary assessment method, outcome definition, and adjustment for confounding variables. Where applicable, sensitivity analyses were conducted to assess the robustness of the findings. For meta-analyses including more than 10 estimates, publication bias was assessed using funnel plots and Egger’s regression test.

A total of 4,552 records were identified through database searches, and an additional 21 records were retrieved via manual searches. After automated deduplication, 2,999 unique records remained. Following title and abstract screening, 53 full-text articles were retrieved. After excluding ineligible studies (Supplementary Table S2), 18 studies were included in the final analysis, including 13 investigating UC incidence and 5 addressing UC recurrence (Figure 1).

The final analysis comprised 18 studies published between 1993 and 2024, including 9 cohort studies and 9 case–control studies, with a combined sample size of 1,384,024 participants. Seventeen studies were published in English and one in Chinese. Eleven studies focused on populations from Western countries characterized by Western dietary patterns, while seven included participants from Asian countries with Eastern dietary habits. Ten studies examined processed meat consumption, fifteen investigated red meat intake, and one specifically evaluated pork consumption during childhood (Tables 2, 3). These studies accounted for a range of confounders, including geographic region, age, sex, smoking status, energy intake, and physical activity, as detailed in Supplementary Tables S3 and S4.

The results of the quality assessment are presented in Figure 2. Among both cohort and case–control studies, three were rated as moderate quality and six as high quality. The mean quality score was 7.22 for cohort studies and 7.00 for case–control studies. Overall, the included studies were predominantly of high methodological quality.

Thirteen studies reported the association between red meat consumption and the risk of developing UC. A random-effects model revealed significant heterogeneity among the included studies (I² = 57.8%, p = 0.005). As shown in Figure 3, individuals with high red meat intake had a significantly increased risk of developing UC compared to those with low intake (RR = 1.21; 95% CI: 1.03–1.42; p = 0.020).

Seven studies assessed the association between processed meat consumption and the risk of developing UC. A random-effects model indicated substantial heterogeneity among the studies (I² = 76.5%, p < 0.001). As shown in Figure 4, individuals with high processed meat consumption exhibited a non-significant increase in UC risk compared to those with low consumption (RR = 1.54; 95% CI: 0.99–2.42; p = 0.058).

Four studies evaluated the association between red meat consumption and the risk of UC recurrence. A random-effects model indicated substantial heterogeneity among the included studies (I² = 78.6%, p = 0.003). As shown in Figure 5, the analysis revealed no significant association between red meat intake and UC recurrence risk (RR = 1.32; 95% CI: 0.54–3.21; p = 0.546).

Three cohort studies assessed the association between processed meat consumption and the risk of UC recurrence. A random-effects model revealed considerable heterogeneity among the studies (I² = 83.7%, p = 0.002). As shown in Figure 6, the meta-analysis revealed no significant difference in UC recurrence risk between individuals with high versus low processed meat intake (RR = 1.58; 95% CI: 0.56–4.47; p = 0.390).

Eight studies provided dose–response data suitable for analysis. The preceding meta-analysis suggested that red meat consumption may increase the risk of developing UC. Therefore, a dose–response analysis was conducted. A linear association was observed between red meat intake and UC risk, with a positive dose–response trend, as shown in Figure 7. Specifically, each 100 g/day increase in red meat intake was associated with an approximately 65% higher risk of UC (RR = 1.65; 95% CI: 1.30–2.09).

Considerable heterogeneity was observed among studies investigating the association between red and processed meat consumption and the risk of UC. To explore potential sources of this heterogeneity, subgroup analyses were conducted based on study design, dietary characteristics, meat type, dietary assessment tools, outcome measurements, and adjusted confounders. These variables were also included as covariates in a meta-regression analysis.

Table 4 and Supplementary Figure S1 present the results of the subgroup analysis for the association between red meat consumption and UC risk. The findings suggest that heterogeneity may be partially explained by the type of dietary assessment tool used. In studies using country-specific validated food frequency questionnaires, no significant heterogeneity was observed, and red meat intake was significantly associated with increased UC risk (RR = 1.501; 95% CI: 1.160–1.941; p = 0.002). Outcome measurement, geographical region, and family history of inflammatory bowel disease may also contribute to the observed heterogeneity. However, meta-regression analysis did not identify any variables that significantly explained the heterogeneity.

Table 5 and Supplementary Figure S2 summarize the subgroup analysis results for the association between processed meat consumption and UC risk. Heterogeneity in these studies may be influenced by study design, dietary characteristics, and the type of dietary assessment tool. In studies focusing on Western-style diets, processed meat consumption was significantly associated with increased UC risk (RR = 1.292; 95% CI: 1.012–1.649; p = 0.040). Similarly, in studies using country-specific validated food frequency questionnaires, processed meat intake significantly increased UC risk (RR = 1.352; 95% CI: 1.003–1.821; p = 0.048). However, due to the limited number of studies, meta-regression was not conducted.

Thirteen studies were included to examine the association between red meat intake and the risk of developing UC, prompting the construction of a funnel plot to assess publication bias (Supplementary Figure S3). Egger’s test revealed significant publication bias (p = 0.005). Therefore, a sensitivity analysis was conducted using the trim-and-fill method. After imputing five studies, the adjusted pooled effect size was RR = 1.026 (95% CI, 0.817–1.287), indicating no statistically significant association (p = 0.827). However, this estimate differed from the original pooled result, indicating potential instability in the findings.

According to the GRADE assessment, the overall certainty of the evidence was primarily rated as very low. This rating was mainly downgraded due to the complexity in categorizing meat types and substantial heterogeneity across studies (Supplementary Table S5).

This systematic review included 18 cohort and case–control studies of moderate to high quality, encompassing a total of 1,384,024 participants. The findings suggest that, high intake of red meat may be significantly associated with an increased risk of developing UC. Conversely, the meta-analysis did not find a statistically significant association between processed meat consumption and UC risk. Regarding UC recurrence, the available evidence remains too limited to draw definitive conclusions; however, the current findings tentatively suggest that red and processed meat consumption may not be strongly associated with an elevated risk of disease recurrence. Furthermore, the overall certainty of evidence was rated as “very low” according to the GRADE framework, primarily due to substantial heterogeneity and the potential for misclassification bias. This high degree of heterogeneity, observed across most analyses, challenges the reliability of the point estimates and suggests that the true effect may differ.

With regard to red meat and UC risk, our findings confirmed a significant linear relationship. Specifically, an increase of 100 grams of red meat per day was associated with an approximately 65% increased risk of UC. Nevertheless, the presence of significant publication bias, as indicated by Egger’s test and the trim-and-fill adjustment, introduces uncertainty. The instability of the pooled estimates following this adjustment underscores that the observed association must be interpreted with considerable caution.

Recent animal studies suggest that high red meat intake may disrupt the colonic mucosal barrier in UC mouse models, increasing the expression of pro-inflammatory M1 macrophages and decreasing anti-inflammatory M2 macrophages, thereby disturbing the M1/M2 balance (20). Several biologically plausible mechanisms have been proposed to explain this link. Red meat is rich in compounds like carnitine and choline, which gut microbiota can metabolize into trimethylamine (TMA). Hepatic oxidation of TMA produces trimethylamine-N-oxide (TMAO), a metabolite implicated in promoting inflammatory processes that could contribute to UC pathogenesis (21). However, the biological mechanisms linking red meat to UC pathogenesis remain hypothetical and are primarily derived from animal models. It is essential to distinguish these mechanistic hypotheses from our evidence-based conclusions, as direct human evidence supporting the involvement of these pathways in UC development remains limited.

Subgroup analyses were conducted to explore the sources of heterogeneity for red meat. Heterogeneity was significantly reduced in analyses limited to studies that used country-specific validated food frequency questionnaires, suggesting that such tools may better capture regional dietary patterns and minimize measurement error. Geographic location also explained some heterogeneity, which may be attributed to differences in cooking practices. For example, the predominantly lower UC risk in regions like India, where meat is often minimally processed and cooked simply, contrasts with higher-risk Western countries (22). This implies that cooking methods may modify the risk associated with red meat consumption.

Contrary to common assumptions, our results suggest that processed meat intake may not increase UC risk to the same extent as red meat. This finding aligns with existing literature (23). This finding requires careful interpretation. A significant limitation in synthesizing the evidence on processed meat is the considerable variability in its definition across different studies and regions. The lack of a standardized definition likely introduces classification bias and complicates the interpretation of null results. For instance, more heavily processed meats—such as organ meats prepared with substantial additives and seasonings—may pose greater health risks than minimally processed products like simply seasoned dried meat (24). However, such gradations in processing levels are not consistently captured in current studies on processed meat. Furthermore, UC often has an acute onset with a relatively short preclinical phase, which may reduce the observable effects of processed meat consumption (25). Additionally, the substantial heterogeneity among the included studies and the potential for residual confounding factors—such as unadjusted total energy intake—preclude definitive conclusions that processed meat is not associated with UC risk.

Regarding UC recurrence, the analysis was based on a limited number of studies. The null findings are accompanied by wide confidence intervals and considerable heterogeneity, indicating that the evidence base is currently insufficient to draw definitive conclusions. Therefore, we refrain from making dietary recommendations for disease management based on these results, and more research is urgently needed in this area.

This study investigates the association between red and processed meat consumption and the risk of UC, with a particular emphasis on the dose–response relationship. Nevertheless, several limitations should be acknowledged. First, this meta-analysis included both cohort and case–control studies; the retrospective nature of the latter may have introduced recall bias. Second, the majority of the included studies were conducted in Western populations, potentially limiting the generalizability of the findings to regions such as China and other Asian countries. Third, substantial heterogeneity among the included studies may compromise the robustness of the pooled estimates. To address this issue, subgroup analyses and meta-regression were performed to identify potential sources of heterogeneity.

Future prospective observational studies are warranted to investigate the association between UC incidence and the consumption of various types of processed meats, including smoked and cured products. Furthermore, due to individual dietary preferences and cultural habits, recommendations to reduce red and processed meat consumption in the general population have not been effectively adopted. Therefore, effective public health strategies are urgently needed to promote adherence to dietary recommendations. With regard to the risk of UC relapse, current evidence remains limited; thus, additional high-quality prospective cohort studies involving diverse populations and geographic regions are essential.