Abstract
Background
The novel Dietary Index for Gut Microbiota (DI-GM) quantifies dietary patterns influencing microbiota health, yet its association with irritable bowel syndrome (IBS) risk remains unexplored. We aim to investigate prospective association of DI-GM with incident IBS in a long-term cohort.
Methods
Participants free of IBS at baseline with at least one record of 24-hour dietary recall from UK Biobank were included (N = 175,103). DI-GM was composed of 14 food/nutrient components known to influence gut microbial health. Each component was scored 0 or 1 based on sex-specific median intakes, and total scores were summed. The primary outcome was incident IBS. Multivariable Cox models were used to estimate associated risk.
Results
Over a 12.56-years of follow-up, 2,980 incident IBS cases were identified. Compared to participants with the lowest DI-GM quartile, those with the highest quartile had a 13% lower IBS risk (HR = 0.87; 95% CI: 0.79–0.97; P-value = 0.009), with significant trend (P-trend = 0.012). Moreover, per 1 score increase in DI-GM was associated with a 3% reduced IBS risk (HR = 0.97; 95% CI: 0.95–0.99; P-value = 0.009). Subgroup and sensitivity analyses consistently supported this inverse association.
Conclusions
Higher baseline DI-GM scores, reflecting dietary patterns supportive of gut microbiota, are associated with lower risk of developing IBS. These findings suggest that the DI-GM score may serve as a useful predictive tool for assessing IBS risk and facilitating the implementation of targeted dietary interventions. Future studies should examine changes in DI-GM scores over time help establish a potential causal link with IBS.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12937-025-01224-3.
Keywords: Dietary index for gut microbiota, Irritable bowel syndrome, Cohort study
Introduction
Irritable bowel syndrome (IBS), a common disorder of gut-brain interaction, is characterized by recurrent abdominal pain and altered bowel habits, including diarrhea, constipation, or a combination of both. Prior studies suggested that IBS affected 3.5%−10% of the world’s population based on Rome III diagnostic criteria [1]. A nationwide cross-sectional study in the United States showed a prevalence of IBS at 6.1% based on the Rome IV criteria [2]. Despite the absence of detectable organic abnormalities in patients with IBS, the recurrent nature of symptoms severely diminishes quality of life and often leads to frequent healthcare visits, imposing a substantial economic burden on society [3–5]. Therefore, identifying early contributing factors in IBS and implementing effective targeted interventions has become an urgent priority.
Recently, emerging evidence highlights the critical role of gut microbiota in the pathogenesis of IBS, with dysbiosis at the mucosal interface implicated through several mechanisms, such as intestinal barrier dysfunction, low-grade inflammation, immune modulation, and altered gut-brain signaling, particularly visceral hypersensitivity [6, 7]. Diet is a major determinant of gut microbiota composition and function. Accumulating evidence suggests specific dietary components, including fiber, polyphenols, and short-chain fatty acids, can modulate gut microbiota and its metabolites, potentially influencing gut homeostasis and symptom generation in IBS [8, 9]. Furthermore, dietary patterns rich in fermentable carbohydrates (i.e., Fermentable Oligosaccharides, Disaccharides, Monosaccharides and Polyols [FODMAPs]) or saturated fats have been associated with microbial shifts linked to gas production, altered motility and visceral hypersensitivity, leading to IBS pathogenesis [10].
As dietary intake involves complex interactions among multiple food components, diet quality indices have become essential tools for evaluating the cumulative impact of overall dietary patterns on health outcomes [11, 12]. The Diet Index for Gut Microbiota (DI-GM), a novel developed index based on 14 food items or nutrients, specifically captures the relationship between diet, gut microbiota diversity and host health [13]. Higher DI-GM scores have been associated with reduced risks of gastrointestinal cancer, fatty liver, hyperlipidemia and metabolic syndrome [14–17]. However, no study has yet examined the association between DI-GM and IBS risk. Given the interplay between diet, gut microbiota, metabolic factors and IBS pathophysiology [18, 19], investigating the link between DI-GM and IBS may offer valuable insights into dietary strategies for prevention and management.
Therefore, we aimed to prospectively investigate the effect of DI-GM diet pattern on long-term risk of incident IBS based on a large cohort, in order to provide novel insights into nutritional interventions for IBS.
Methods
Study population
The UK Biobank (UKB) is a large-scale prospective cohort that recruited over 500,000 participants aged 37–73 years at 22 assessment centers across England, Scotland, and Wales between 2006 and 2010. Baseline data was collected through touchscreen questionnaires, physical measures and medical condition reports. More than 210,000 participants completed a web-based 24-hour diet recall questionnaire (Oxford WebQ) over 5 cycles between April 2009 and June 2012. Details of the UKB design was reported elsewhere [20]. The study protocol was approved by the North West Multi-Centre Research Ethics Committee (21/NW/0157), and all participants provided written informed consent prior to the study.
For the current study, participants with at least one record of dietary assessment were included. Those who withdrew (N = 41), had prevalent cancer (N = 22,054), inflammatory bowel disease (N = 2,161), IBS (N = 9,582), coeliac disease diagnosis (N = 1,186) prior to 24-hour diet recall questionnaire, or extreme energy intake (< 500 kcal/d or > 4,500 kcal/day, N = 838) were excluded. All diagnoses were identified through International Classification of Disease-10 (ICD-10) codes. Consequently, a total of 175,103 participants free of IBS with available diet information were included (Fig. 1).
Fig. 1.
Flowchart of the study population. Abbreviations: UKB: UK Biobank; IBD: inflammatory bowel disease; IBS: irritable bowel syndrome
Assessment of baseline dietary index for gut microbiota
Dietary information on over 200 commonly consumed food and beverage items was collected using the Oxford WebQ across 5 cycles (cycle 1: April 2009-September 2010; cycle 2: February 2011-April 2011; cycle 3: June 2011-August 2011; cycle 4: October 2011-December 2011; cycle 5: April 2012-June 2012). This multi-cycle approach was designed to ensure comprehensive coverage of all seasons and days of the week. The food list was developed based on data from the UK National Diet and Nutrition Survey and was refined through a pilot study to enhance its relevance and accuracy. At the beginning of questionnaire, participants were asked “Would you say that what you ate and drank yesterday was fairly typical for you?”. Further instructions and help sections were provided to assist participants in reporting portions for each food item. These measures contributed to a high level of consistency between nutrient intakes estimated by the WebQ and those obtained from interviewer-administered assessments conducted on the same day [21]. Each group of food consumption (g/day) was generated by multiplying the number of portions consumed by the set quantity of each portion size as describe by Piernas C et al. [21, 22]. The nutrient intake data were derived from the UK Nutrient Databank food composition Table [23]. For participants who completed ≥ 2 dietary cycles, an average value of each group of food consumption or nutrient intake was calculated.
According to the criteria established by Kase et al. [13], the DI-GM dietary pattern incorporates a total of 14 food- and nutrient-based components, classified according to their potential impact on gut microbiota as either beneficial or unfavorable. For each gut microbiota-beneficial component (avocado, broccoli, chickpeas, coffee, cranberries, fermented dairy, fiber, green tea, soybean, and whole grains), participants who consumed at or above sex-specific median received 1 point (0 otherwise). Conversely, for each unfavorable component [red meat, processed meat, refined grains, and high-fat diet (≥ 40% energy from fat)], participants who consumed below sex-specific median received 1 point (0 otherwise). Notably, sex-specific median for DI-GM components were derived from the study population itself rather than from previously established reference cut-off values. The total DI-GM score, ranging from 0 to 14, was derived by summing the points across all 14 components, with higher scores indicating greater adherence to a gut microbiota-beneficial dietary pattern.
Outcome assessment
The primary outcome was incident irritable bowel syndrome (IBS), identified during follow-up (censoring date: 30 September 2023) using ICD-10 code (K58). IBS diagnosis was based on linked national registry data, including primary care records and hospital admissions from the Hospital Episode Statistics (England), Scottish Morbidity Record (Scotland), and Patient Episode Database (Wales). Moreover, Rome III criteria via the digestive healthcare questionnaire (DHQ) during 2017–2018 was also used to ascertain incident IBS in sensitivity analysis.
Covariates
Potential covariates at baseline were selected based on previous epidemiology evidence [24–27], including age (continuous), sex (male, female), ethnicity (white, nonwhite), Townsend deprivation index (TDI, quartiles), education level (university, non-university), smoking status (never, previous, current), drinking status (never, previous, current), body mass index (BMI, < 18.5 kg/m2, 18.5–24.9 kg/m2, 25.0–29.9 kg/m2, ≥ 30 kg/m2), physical activity, and energy intake (continuous). TDI was a measure of socioeconomic status, with lower value indicating relative affluence. Education was based on self-report of highest qualification achieved and classified as university or non-university. Physical activity was self-reported and categorized into three levels (high, moderate and low) based on the International Physical Activity Questionnaire (IPAQ). Additionally, in order to further control potential confounders, weekly alcohol intake in grams, type 2 diabetes, anxiety and depression were also considered as covariates in the sensitivity analyses. Weekly alcohol intake in grams was calculated by multiplying the average number of alcoholic drinks consumed each week by the average grams of alcohol contained in each type of drink, determined using the UK Food Standard Agency’s guidelines [28]. Baseline comorbidity was identified via ICD-10 code (E11 for type 2 diabetes, F32 for depression, F40 and F41 for anxiety) through linkage to primary care and/or hospital admission data.
Statistical analysis
Baseline characteristics were described by mean ± standard deviation (SD) for continuous variables and absolute numbers (percentages) for categorical variables. Comparison across different quartile groups were performed by ANOVA test for continuous variables and Chi-square test for categorical variables. The cumulative incidence of IBS was calculated via Kaplan–Meier method.
Cox proportional hazard model was conducted to assess the effect of DI-GM scores on risk of IBS development. For quartiles and per 1 increment of DI-GM scores, three adjustment models in addition to univariable analysis were conducted: (1) model 1, adjusted for age and sex; (2) model 2, additionally adjusted for ethnicity, TDI, education level; (3) model 3, additionally adjusted for smoking status, drinking status, BMI, IPAQ and energy intake. Considering very small percentage of missing values (0.1%−0.33%) for most variables, missing indicators were used. The follow-up period started from the first date of dietary survey until the date of the first IBS diagnosis, death, lost to follow-up or the end of study (30 September 2023). Proportional hazard assumptions were assessed via Schoenfeld residuals, indicating no significant violations. Furthermore, we have additionally conducted the statistical analysis based on the model 3 regarding the associations between individual DI-GM components and IBS to further identify which dietary factors most strongly drive the observed associations.
Subgroup analyses were performed to assess whether risk of incident IBS associated with DI-GM scores varied by age (< 60 years, ≥ 60 years), sex (male, female), BMI (<25.0 kg/m2, ≥ 25.0 kg/m2), smoking status (never, previous/current), drinking status (never/previous, current), and TDI (< −2.15, ≥ −2.15). Potential effect modifications were tested by adding interaction terms of each stratified variable with quartiles of DI-GM scores, respectively.
To test the robustness and reliability of results, additional sensitivity analyses were conducted based on model 3. Firstly, participants who had IBS diagnosis within 1 or 2 years after recruitment were excluded, in order to avoid reverse causation. Secondly, Fine and Gray’s competing risk model was conducted by considering death and lost to follow-up as competing events, as those participants might develop IBS thereafter. Thirdly, to further reduce potential misclassification bias, those who died but without developing IBS during the follow-up period were excluded from the analysis. Fourthly, considering the influence of age on IBS risk, age-scaled Cox regression with left truncation and stratified by 5-years birth cohort was also conducted, separately. Fifthly, weekly alcohol intake in grams instead of drinking status was additionally adjusted. Sixthly, anxiety, depression and type 2 diabetes were additionally adjusted to further rule out the potential confounding bias [29]. Seventhly, to accounting for the misdiagnosis of incident IBS, those who fulfilled Rome III criteria via DHQ without incident ICD-10 diagnosis (i.e. considering these IBS cases as prevalent cases at baseline) were excluded.
SAS version 9.4 software (SAS Institute, Inc., Cary, NC, USA) and R Version 4.2.2 (forestplot and ggplot2 packages) were used for all the statistical analyses and figures, respectively. All statistical tests were 2-sided, with a P-value < 0.05 indicating statistical significance.
Results
Baseline characteristics
Among 175,103 participants with a mean (SD) age of 55.75 (7.96) years, 53.26% were female (Table 1). The mean (SD) DI-GM score of all participants was 5.49 (1.56), with 3.54 (0.69), 4.93 (0.24), 5.91 (0.24), and 7.44 (0.83) from quartile 1 to quartile 4, respectively. Compared with participants in the lowest quartile, those with higher DI-GM scores were more likely to be older, female, never/previous smokers and current alcohol drinkers, had a higher level of TDI and physical activity, and had a lower BMI and a lower prevalence of type 2 diabetes (all P-values < 0.001). The median follow-up period for this cohort was 12.6 years (interquartile range: 12.1–13.4 years).
Table 1.
Baseline characteristics of participants by quartiles of DI-GM scores
| Characteristics | Overall (N = 175,103) |
DI-GM Scores | P-value | |||
|---|---|---|---|---|---|---|
| Q1(N = 43,745) | Q2(N = 42,411) | Q3(N = 41,648) | Q4(N = 47,299) | |||
| Age, yearsa | 55.75 ± 7.96 | 54.36 ± 8.06 | 55.39 ± 7.97 | 56.17 ± 7.88 | 56.99 ± 7.71 | < 0.001 |
| Sex, N (%) | < 0.001 | |||||
| Male | 81,851 (46.74) | 21,745 (49.71) | 20,420 (48.15) | 19,456 (46.72) | 20,230 (42.77) | |
| Female | 93,252 (53.26) | 22,000 (50.29) | 21,991 (51.85) | 22,192 (53.28) | 27,069 (57.23) | |
| Ethnicity, N (%) | < 0.001 | |||||
| White ethnicity | 8,094 (4.62) | 2,706 (6.19) | 1,996 (4.71) | 1,643 (3.94) | 1,749 (3.7) | |
| Non-white ethnicity | 166,435 (95.05) | 40,895 (93.48) | 40,271 (94.95) | 39,864 (95.72) | 45,405 (96.0) | |
| Unknown | 574 (0.33) | 144 (0.33) | 144 (0.34) | 141 (0.34) | 145 (0.31) | |
| Townsend deprivation index, N (%) | < 0.001 | |||||
| Mean ± SD | -1.55 ± 2.89 | -1.35 ± 2.99 | -1.54 ± 2.89 | -1.64 ± 2.82 | -1.65 ± 2.83 | < 0.001 |
| Q1 | 43,725 (24.97) | 10,296 (23.54) | 10,531 (24.83) | 10,651 (25.57) | 12,247 (25.89) | |
| Q2 | 43,716 (24.97) | 10,469 (23.93) | 10,701 (25.23) | 10,538 (25.3) | 12,008 (25.39) | |
| Q3 | 43,719 (24.97) | 10,892 (24.9) | 10,553 (24.88) | 10,471 (25.14) | 11,803 (24.95) | |
| Q4 | 43,719 (24.97) | 12,017 (27.47) | 10,559 (24.9) | 9,947 (23.88) | 11,196 (23.67) | |
| Unknown | 224 (0.13) | 71 (0.16) | 67 (0.16) | 41 (0.1) | 45 (0.1) | |
| Education level, N (%) | < 0.001 | |||||
| University | 84,480 (48.25) | 23,426 (53.55) | 20,865 (49.2) | 19,329 (46.41) | 20,860 (44.10) | |
| Non-university | 75,425 (43.07) | 15,521 (35.48) | 17,843 (42.07) | 19,007 (45.64) | 23,054 (48.74) | |
| Unknown | 15,198 (8.68) | 4,798 (10.97) | 3,703 (8.73) | 3,312 (7.95) | 3,385 (7.16) | |
| Smoking status, N (%) | < 0.001 | |||||
| Never | 99,596 (56.88) | 24,542 (56.1) | 23,852 (56.24) | 23,744 (57.01) | 27,458 (58.05) | |
| Previous | 61,131 (34.91) | 14,347 (32.8) | 14,829 (34.96) | 14,970 (35.94) | 16,985 (35.91) | |
| Current | 13,912 (7.95) | 4,719 (10.79) | 3,620 (8.54) | 2,818 (6.77) | 2,755 (5.82) | |
| Unknown | 464 (0.26) | 137 (0.31) | 110 (0.26) | 116 (0.28) | 101 (0.21) | |
| Drinking status, N (%) | < 0.001 | |||||
| Never | 5,644 (3.22) | 1,717 (3.93) | 1,367 (3.22) | 1,168 (2.8) | 1,392 (2.94) | |
| Previous | 5,107 (2.92) | 1,348 (3.08) | 1,188 (2.8) | 1,166 (2.8) | 1,405 (2.97) | |
| Current | 164,176 (93.76) | 40,625 (92.87) | 39,808 (93.86) | 39,278 (94.31) | 44,465 (94.01) | |
| Unknown | 176 (0.1) | 55 (0.13) | 48 (0.11) | 36 (0.09) | 37 (0.08) | |
| BMI, N (%) | < 0.001 | |||||
| Mean ± SD | 26.98 ± 4.65 | 27.73 ± 4.92 | 27.21 ± 4.65 | 26.78 ± 4.49 | 26.26 ± 4.38 | < 0.001 |
| < 18.5 kg/m2 | 896 (0.51) | 172 (0.39) | 185 (0.44) | 214 (0.51) | 325 (0.69) | |
| 18.5–24.9 kg/m2 | 63,725 (36.39) | 13,332 (30.48) | 14,464 (34.1) | 15,717 (37.74) | 20,212 (42.73) | |
| 25.0–29.9 kg/m2 | 72,936 (41.65) | 18,519 (42.33) | 18,014 (42.47) | 17,475 (41.96) | 18,928 (40.02) | |
| ≥ 30.0 kg/m2 | 37,053 (21.16) | 11,573 (26.46) | 9,613 (22.67) | 8,138 (19.54) | 7,729 (16.34) | |
| Unknown | 493 (0.28) | 149 (0.34) | 135 (0.32) | 104 (0.25) | 105 (0.22) | |
| IPAQ, N (%) | < 0.001 | |||||
| Low | 26,016 (14.86) | 8,077 (18.46) | 6,822 (16.09) | 5,856 (14.06) | 5,261 (11.12) | |
| Moderate | 61,123 (34.91) | 14,958 (34.19) | 14,875 (35.07) | 14,814 (35.57) | 16,476 (34.83) | |
| High | 57,819 (33.02) | 12,332 (28.19) | 13,253 (31.25) | 14,000 (33.62) | 18,234 (38.55) | |
| Unknown | 30,145 (17.22) | 8,378 (19.15) | 7,461 (17.59) | 6,978 (16.75) | 7,328 (15.49) | |
| Type2 diabetes, N (%) | 4,273 (2.44) | 1,314 (3.00) | 1,050 (2.48) | 954 (2.29) | 955 (2.02) | < 0.001 |
| Anxiety, N (%) | 6,085 (3.48) | 1,584 (3.62) | 1,450 (3.42) | 1,364 (3.28) | 1,687 (3.57) | 0.025 |
| Depression, N (%) | 12,863 (7.35) | 3,386 (7.74) | 3,110 (7.33) | 2,958 (7.1) | 3,409 (7.21) | 0.002 |
| Energy intake (kcal/day)a | 2060.32 ± 573.81 | 1936.21 ± 558.39 | 2018.29 ± 563.36 | 2098.02 ± 566.46 | 2179.6 ± 576.32 | < 0.001 |
| DI-GMa | 5.49 ± 1.56 | 3.54 ± 0.69 | 4.93 ± 0.24 | 5.91 ± 0.24 | 7.44 ± 0.83 | < 0.001 |
| DI-GM components | ||||||
| Beneficial to gut microbiota | ||||||
| Avocados (serving/day)a | 0.02 ± 0.1 | 0.01 ± 0.05 | 0.01 ± 0.07 | 0.02 ± 0.08 | 0.04 ± 0.15 | < 0.001 |
| Broccoli (serving/day)a | 0.13 ± 0.28 | 0.05 ± 0.17 | 0.09 ± 0.23 | 0.13 ± 0.28 | 0.24 ± 0.37 | < 0.001 |
| Chickpea (g/day)a | 13.71 ± 31.0 | 6.06 ± 20.36 | 10.34 ± 25.99 | 14.1 ± 30.39 | 23.48 ± 39.95 | < 0.001 |
| Coffee (g/day)a | 243.11 ± 269.26 | 165.6 ± 255.64 | 236.77 ± 271.93 | 269.69 ± 269.69 | 297.09 ± 261.69 | < 0.001 |
| Cranberries (g/day)a | 8.19 ± 17.76 | 2.83 ± 10.51 | 5.69 ± 14.26 | 8.27 ± 16.73 | 15.33 ± 23.54 | < 0.001 |
| Fermented dairy (g/day)a | 59.96 ± 63.04 | 29.55 ± 45.48 | 49.86 ± 54.81 | 66.89 ± 61.53 | 91.04 ± 69.42 | < 0.001 |
| Fiber (g/day)a | 17.78 ± 6.62 | 13.42 ± 4.82 | 16.15 ± 5.36 | 18.64 ± 5.66 | 22.51 ± 6.6 | < 0.001 |
| Green tea (serving/day)a | 0.12 ± 0.52 | 0.04 ± 0.3 | 0.07 ± 0.39 | 0.11 ± 0.49 | 0.25 ± 0.73 | < 0.001 |
| Soybean (g/day)a | 8.24 ± 41 | 1.64 ± 17.54 | 3.92 ± 26.9 | 7.08 ± 37.08 | 19.24 ± 62.45 | < 0.001 |
| Whole grains (g/day)a | 279.64 ± 138.68 | 203.56 ± 101.44 | 258.63 ± 120.6 | 300.71 ± 132.35 | 350.27 ± 148.63 | < 0.001 |
| Unfavorable to gut microbiota | ||||||
| High-fat diet (% energy)a | 0.31 ± 0.07 | 0.32 ± 0.07 | 0.31 ± 0.06 | 0.30 ± 0.06 | 0.30 ± 0.06 | < 0.001 |
| Processed meat (g/day)a | 17.91 ± 29.85 | 27.26 ± 36.78 | 19.42 ± 29.67 | 15.52 ± 26.29 | 10.03 ± 22.36 | < 0.001 |
| Red meat (g/day)a | 37.95 ± 52.06 | 56.81 ± 61.31 | 41.23 ± 51.29 | 33.7 ± 46.83 | 21.33 ± 40.44 | < 0.001 |
| Refined grains (g/day)a | 84.42 ± 85.45 | 115.9 ± 89.86 | 92.62 ± 84.86 | 77.07 ± 81.04 | 54.42 ± 73.48 | < 0.001 |
Categorical variables are presented as frequencies and percentages
Abbreviations: DI-GM dietary index for gut microbiota, Q quartile, SD standard deviation, TDI Townsend Deprivation Index, BMI Body Mass Index, IPAQ International Physical Activity Questionnaire
aData are presented as the mean ± standard deviation
DI-GM scores and the risk of incident IBS
Over 2,162,884 person-years of follow-up, 2,980 incident IBS cases were identified. The 12.6-year cumulative incidence of IBS was 1.7% (95% CI: 1.6–1.8%), with 1.9% (95% CI: 1.8–2.1%), 1.7% (95% CI: 1.6–1.8%), 1.7% (95% CI: 1.5–1.8%), and 1.7% (95% CI: 1.5–1.8%) from quartile 1 to quartile 4, respectively.
Compared to participants in the lowest quartile of DI-GM score, those in the highest quartile had a significantly 13% lower risk of IBS (HRQ4 VS Q1=0.87; 95% CI: 0.79–0.97; P = 0.009) according to the fully adjusted model. Meanwhile, compared with the lowest quartile group, both quartile 2 (HRQ2 VS Q1=0.89; 95% CI: 0.80–0.98; P = 0.023) and quartile 3 (HRQ3 VS Q1=0.88; 95% CI: 0.80–0.98; P = 0.018) groups were also associated with an evidently reduced risk of IBS. Moreover, per 1 score increment in DI-GM score was associated with a 3% reduction in IBS risk (HR = 0.97; 95% CI: 0.95–0.99, P-value = 0.009, Table 2).
Table 2.
Associations between quartiles of DI-GM scores with risk of IBS
| DI-GM scores | Cases/Total | Follow-up, person-years | Hazard ratio for incident IBS (95% CI, P-value) | |||
|---|---|---|---|---|---|---|
| Crude Model | Adjusted Model 1 | Adjusted Model 2 | Adjusted Model 3 | |||
| Quartile 1 | 830/43,745 | 539,033 | Reference | Reference | Reference | Reference |
| Quartile 2 | 704/42,411 | 524,338 | 0.87 (0.79–0.96, 0.008) | 0.86 (0.78–0.95, 0.004) | 0.88 (0.80–0.98, 0.014) | 0.89 (0.80–0.98, 0.023) |
| Quartile 3 | 680/41,648 | 515,458 | 0.86 (0.77–0.95, 0.003) | 0.84 (0.76–0.93, 0.001) | 0.87 (0.78–0.96, 0.007) | 0.88 (0.80–0.98, 0.018) |
| Quartile 4 | 766/47,299 | 584,055 | 0.85 (0.77–0.94, 0.001) | 0.81 (0.74–0.90, < 0.001) | 0.85 (0.77–0.94, 0.002) | 0.87 (0.79–0.97, 0.009) |
| P-trend = 0.002 | P-trend < 0.001 | P-trend = 0.002 | P-trend = 0.012 | |||
| Per 1 change | 2,980/175,103 | 2,162,884 | 0.96 (0.94–0.99, 0.002) | 0.95 (0.93–0.97, < 0.001) | 0.96 (0.94–0.99, 0.001) | 0.97 (0.95–0.99, 0.009) |
Adjusted Model 1: adjusted for age and sex
Adjusted Model 2: additionally adjusted for ethnicity, TDI, education level
Adjusted Model 3: additionally adjusted for smoking status, drinking status, BMI, IPAQ and energy intake (kcal/d)
Abbreviations: DI-GM dietary index for gut microbiota, IBS irritable bowel syndrome, IQR interquartile range, CI confidence interval, TDI Townsend Deprivation Index, BMI Body Mass Index, IPAQ International Physical Activity Questionnaire
Regarding individual DI-GM components, higher consumption of coffee (HR = 0.85; 95% CI: 0.79–0.92; P < 0.001), soybeans (HR = 0.78; 95% CI: 0.69–0.89; P < 0.001) and fermented dairy (HR = 0.93; 95% CI: 0.87-1.00; P = 0.063) was associated with a reduced risk of developing IBS (Supplementary Fig. 1).
Subgroup analyses
In subgroup analysis, the decreased risk of incident IBS associated with higher DI-GM scores after stratification by age, sex, BMI, smoking status, drinking status, and TDI remained statistically significant, particularly in those with age < 60 years, female, BMI ≥ 25 kg/m2, never smoking, current drinking and TDI<−2.15 (Fig. 2, Supplementary Tables 1–6). No significant interaction effects were observed across those subgroups (all P-interaction > 0.05).
Fig. 2.
Subgroup analysis for the association between DI-GM scores and incident IBS.Note: all adjusted HRs were calculated by adjusting the following covariates: age, sex, ethnicity, TDI, education level, smoking status, drinking status, BMI, IPAQ, type 2 diabetes, anxiety, depression and energy intake (kcal/d).Abbreviations: DI-GM, dietary index for gut microbiota; IBS, irritable bowel syndrome; HR, Hazard ratio; CI, confidence interval; BMI, Body Mass Index; TDI, Townsend Deprivation Index; IPAQ, International Physical Activity Questionnaire
Sensitivity analyses
Results of sensitivity analysis, either for quartiles or per 1 score increment, were all consistent with principal findings, when excluding incident IBS cases within 1 or 2 years after baseline, performing competing risk models, excluding those who died without IBS occurrence during the follow-up period, conducting age-scaled Cox regression with left truncation and stratified by 5-years birth cohort, additionally adjusting weekly alcohol intake in grams, additionally adjusting anxiety, depression and type 2 diabetes, excluding those who fulfilled Rome III criteria via DHQ without incident ICD-10 diagnosis (Table 3).
Table 3.
Sensitivity analysis regarding risk of IBS associated with DI-GM scores
| DI-GM scores | No. of participants | No. of incident IBS | Hazard ratio (95% CI, P-value) |
P-trend | |
|---|---|---|---|---|---|
| Sensitivity analysis 1: excluding IBS participants diagnosed within 1 year after baseline (N = 174,828) | |||||
| Q1 | 43,669 | 754 | Reference | 0.018 | |
| Q2 | 42,337 | 630 | 0.87 (0.79–0.97, 0.013) | ||
| Q3 | 41,591 | 623 | 0.89 (0.80–0.99, 0.028) | ||
| Q4 | 47,231 | 698 | 0.87 (0.78–0.97, 0.010) | ||
| Per 1 change | 174,828 | 2,705 | 0.96 (0.93–0.99, 0.018) | ||
| Sensitivity analysis 2: excluding IBS participants diagnosed within 2 years after baseline (N = 174,572) | |||||
| Q1 | 43,600 | 685 | Reference | 0.012 | |
| Q2 | 42,282 | 575 | 0.88 (0.79–0.98, 0.024) | ||
| Q3 | 41,536 | 568 | 0.89 (0.80-1.00, 0.047) | ||
| Q4 | 47,154 | 621 | 0.85 (0.76–0.96, 0.006) | ||
| Per 1 change | 174,572 | 2,449 | 0.95 (0.92–0.99, 0.012) | ||
| Sensitivity analysis 3: competing risk model (N = 175,103, No. of competing events = 12,591) | |||||
| Q1 | 43,745 | 830 | Reference | < 0.001 | |
| Q2 | 42,411 | 704 | 0.92 (0.87–0.96, 0.001) | ||
| Q3 | 41,648 | 680 | 0.90 (0.85–0.94, < 0.001) | ||
| Q4 | 47,299 | 766 | 0.87 (0.83–0.92, < 0.001) | ||
| Per 1 change | 175,103 | 2,980 | 0.96 (0.94–0.97, < 0.001) | ||
| Sensitivity analysis 4: excluding those who died but did not develop IBS after baseline (N = 165,996) | |||||
| Q1 | 41,345 | 830 | Reference | 0.009 | |
| Q2 | 40,203 | 704 | 0.89 (0.80–0.98, 0.019) | ||
| Q3 | 39,524 | 680 | 0.88 (0.79–0.97, 0.014) | ||
| Q4 | 44,924 | 766 | 0.87 (0.78–0.96, 0.007) | ||
| Per 1 change | 165,996 | 2,980 | 0.96 (0.93–0.99, 0.009) | ||
| Sensitivity analysis 5: age-scaled Cox regression with left truncation (N = 175,103) | |||||
| Q1 | 43,745 | 830 | Reference | 0.012 | |
| Q2 | 42,411 | 704 | 0.89 (0.80–0.98, 0.023) | ||
| Q3 | 41,648 | 680 | 0.88 (0.80–0.98, 0.018) | ||
| Q4 | 47,299 | 766 | 0.87 (0.79–0.97, 0.009) | ||
| Per 1 change | 175,103 | 2,980 | 0.96 (0.93–0.99, 0.012) | ||
| Sensitivity analysis 6: age-scaled Cox regression stratified by 5-years birth cohort (N = 175,103) | |||||
| Q1 | 43,745 | 830 | Reference | 0.011 | |
| Q2 | 42,411 | 704 | 0.89 (0.80–0.98, 0.021) | ||
| Q3 | 41,648 | 680 | 0.88 (0.79–0.98, 0.015) | ||
| Q4 | 47,299 | 766 | 0.87 (0.79–0.97, 0.008) | ||
| Per 1 change | 175,103 | 2,980 | 0.96 (0.93–0.99, 0.011) | ||
| Sensitivity analysis 7: additionally adjusted for weekly alcohol intake in grams instead of drinking status (N = 175,103) | |||||
| Q1 | 43,745 | 830 | Reference | 0.008 | |
| Q2 | 42,411 | 704 | 0.89 (0.80–0.98, 0.022) | ||
| Q3 | 41,648 | 680 | 0.88 (0.79–0.98, 0.015) | ||
| Q4 | 47,299 | 766 | 0.87 (0.78–0.96, 0.006) | ||
| Per 1 change | 175,103 | 2,980 | 0.96 (0.93–0.99, 0.008) | ||
| Sensitivity analysis 8: additionally adjusted for anxiety and depression (N = 175,103) | |||||
| Q1 | 43,745 | 830 | Reference | 0.012 | |
| Q2 | 42,411 | 704 | 0.89 (0.81–0.99, 0.028) | ||
| Q3 | 41,648 | 680 | 0.89 (0.80–0.98, 0.023) | ||
| Q4 | 47,299 | 766 | 0.87 (0.79–0.97, 0.009) | ||
| Per 1 change | 175,103 | 2,980 | 0.96 (0.93–0.99, 0.012) | ||
| Sensitivity analysis 9: additionally adjusted for type 2 diabetes (N = 175,103) | |||||
| Q1 | 43,745 | 830 | Reference | 0.013 | |
| Q2 | 42,411 | 704 | 0.89 (0.81–0.99, 0.024) | ||
| Q3 | 41,648 | 680 | 0.88 (0.80–0.98, 0.019) | ||
| Q4 | 47,299 | 766 | 0.87 (0.79–0.97, 0.010) | ||
| Per 1 change | 175,103 | 2,980 | 0.96 (0.93–0.99, 0.013) | ||
| Sensitivity analysis 10: excluding those who fulfilled Rome III criteria via DHQ without incident ICD-10 diagnosis (N = 167,109) | |||||
| Q1 | 41,870 | 830 | Reference | 0.015 | |
| Q2 | 40,485 | 704 | 0.89 (0.81–0.99, 0.026) | ||
| Q3 | 39,791 | 680 | 0.88 (0.80–0.98, 0.018) | ||
| Q4 | 44,963 | 766 | 0.88 (0.79–0.97, 0.012) | ||
| Per 1 change | 167,109 | 2,980 | 0.96 (0.93–0.99, 0.015) | ||
All adjusted HRs were calculated by adjusting the following covariates: age, sex, ethnicity, TDI, education level, smoking status, drinking status, BMI, IPAQ, and energy intake (kcal/d)
Abbreviations: IBS irritable bowel syndrome, DI-GM dietary index for gut microbiota, CI confidence interval, Q quartile, DHQ Digestive Health Questionnaire, ICD international classification disease, TDI Townsend Deprivation Index, BMI Body Mass Index, IPAQ International Physical Activity Questionnaire
Discussion
In this population-based prospective cohort study based on the UKB, we found that adherence to high DI-GM diet was significantly associated with a reduced long-term risk of IBS, with significant dose-response relationship. Of note, individuals with the highest quartile of DI-GM score had a 13% lower risk of incident IBS. Meanwhile, a 3% decreased risk of IBS was observed with per 1 score increment of DI-GM score. These findings suggest the gut microbiota-beneficial dietary patterns may represent a promising strategy for the prevention of IBS onset.
As a novel dietary index designed to reflect gut microbiota health, DI-GM emphasizes increased consumption of dietary fiber, legumes and fermented dairy, while recommending reduced intake of high-fat foods, red meat and refined grains. Prior evidence showed that high-fat diet led to the alterations of mucous layer and microbial composition, and thereby disrupted intestinal barrier function by impairing tight junction proteins [30]. In contrast, high-fiber diet could significantly increase the abundance of beneficial bacteria such as Lactobacillus spp. and Bifidobacterium spp. [31], while broccoli-rich diets in murine models demonstrated an increase in beneficial microbiota such as Lachnospiraceae NK4A136 group and Dubosiella newyorkensis [32]. Additionally, polyphenol-rich foods including coffee, green tea and cranberries, were also linked to higher abundance of Bifidobacterium, Akkermansia and Lactobacillus species [33, 34]. These favorable change of gut microbiota may further produce beneficial metabolites such as short-chain fatty acids (SCFAs), bile acids, which may enhance gut barrier integrity, reduce inflammation, regulate intestinal acid–base homoeostasis and modulate gastrointestinal motility, thereby potentially lowering IBS risk [35–37].
Furthermore, DI-GM may also modulate visceral sensitivity through the brain-gut-microbiota (BGM) axis, potentially contributing to a reduced risk of IBS [9, 38]. Current evidence indicates BGM modulation is primarily mediated through neuroimmune and neuroendocrine mechanisms, often involving the vagus nerve. It has been suggested that the communication is mediated by several microbially derived metabolites (i.e., SCFAs, secondary bile acids, tryptophan metabolites) with the interaction of enteroendocrine cells (EECs), enterochromaffin cells (ECCs) and mucosal immune system. Additionally, the varied microbiota could also independently contribute to the production of various neuroactive molecules such as γ-aminobutyric acid and 5-HT, which may also result in the variation of visceral sensitivity and further impact the pathogenesis of IBS [39].
Emerging evidence supports the relationship between higher diet quality indices and a lower risk of IBS. Shang et al. found that greater adherence to healthy dietary patterns, including Alternate Mediterranean Diet (AMED), Anti-Empirical Dietary Inflammatory Index (AEDII), Alternate Healthy Eating Index-2010 (AHEI-2010) and Healthful Plant-based Diet Index (HPDI), were associated with a lower IBS risk [40]. A case-control study from Iran further suggested a potential positive association between pro-inflammatory diet and increased IBS risk [41], whereas clinical trials indicated the Mediterranean diet may alleviate gastrointestinal symptoms among IBS patients [42]. Notably, two randomized controlled trials reported that adherence to Mediterranean diet was associated with specific alterations in gut microbiota composition, including increased abundance of beneficial bacteria such as Faecalibacterium prausnitzii and Roseburia spp., and reduced abundance of Ruminococcus gnavus, Collinsella aerofaciens and Ruminococcus torques [43, 44]. Similar to DI-GM, these dietary indices emphasize high intake of whole grains, nuts and legumes, fruits and vegetables, while discouraging intake of red and processed meat. Accumulating evidence suggested that adherence to such nutrient-dense dietary patterns may promote a balanced gut microbiota and mitigate systemic inflammation and oxidative stress [45, 46]. However, not all dietary patterns have demonstrated consistent protective effects. For instance, a cross-sectional study among 3,362 Iranian adults found no significant association between adherence to Mediterranean-DASH Diet Intervention for Neurodegenerative Delay (MIND) diet and IBS risk [47]. This null finding might be partially explained by the relatively high content of FODMAPs in MIND diet, which are known to trigger gastrointestinal symptoms in susceptible individuals [48]. These findings underscore the importance of individualized dietary management, particularly for those experiencing bloating, flatulence and abdominal discomfort, who should be encouraged to identify and avoid symptom-triggering foods and tailor dietary recommendations accordingly.
Moreover, emerging evidence supports the effectiveness of the low-FODMAP diet in alleviating IBS symptoms, primarily through the restriction of fermentable short-chain carbohydrates that induce osmotic effects and microbial fermentation in the gut [48, 49]. However, current evidence remains limited regarding whether adherence to a low-FODMAP diet reduces the risk of developing IBS. In contrast, the DI-GM was developed to reflect habitual dietary patterns associated with a gut microbiota profile linked to better metabolic and inflammatory health [13]. Although both the low-FODMAP diet and the DI-GM recognize the diet-microbiota axis and share some dietary components, their purposes differ fundamentally: the former is a therapeutic intervention for symptom management in diagnosed IBS patients, while the latter promotes long-term dietary patterns to support a beneficial gut microbiota for primary prevention at the population level [50, 51]. In this context, our study provides novel evidence that a diet reflecting more healthier gut microbiota (higher DI-GM) may lower the risk of IBS onset, a preventive effect distinct from the symptomatic benefits of low-FODMAP, underscoring the role of habitual diet in disease prevention.
Owing to data availability, we used Rome III criteria instead of Rome IV criteria in this study. Given that the Rome IV criteria are more stringent than the Rome III criteria, the impact of applying Rome IV for both prevalent and incident IBS diagnosis on the observed associations is complex. If both prevalent and incident IBS were defined using Rome IV, fewer individuals would be excluded at baseline as prevalent cases. Those meeting Rome III but not Rome IV criteria would be classified as non-cases and retained in the cohort. However, this group may have subclinical or milder symptoms and thus could be at higher risk of being classified as incident IBS during follow-up compared to the general population, possibly leading to an overestimation of the incidence rate and potentially biasing the association. Therefore, it remains unclear whether the use of Rome III versus Rome IV criteria leads to overestimation or underestimation of the true association. Future studies applying consistent Rome IV criteria from baseline onward are needed to confirm our findings.
Despite observing a 13% relative risk reduction in IBS development when comparing the highest versus the lowest quartile of the DI-GM score, the absolute risk difference was minimal (i.e., 0.2% over a period of 12.6 years). This is largely attributable to the low baseline incidence of IBS in the general population, specifically, a 12.6-year cumulative incidence of only 1.7% in our cohort. Although this suggests limited clinical significance at the population level, the population-attributable risk depends on both the relative risk and the prevalence of the exposure. Hence, the absolute benefit of a higher DI-GM diet may be more substantial in high-risk subpopulations—such as females, individuals aged less than 50 years, or those with fatty liver disease—who are likely to have a higher baseline risk of IBS [2]. In such groups, our findings could offer preliminary value for targeted preventive approaches.
Notably, an important limitation of our study is the lack of repeated dietary assessments during the extended follow-up period. As dietary habits can change substantially over time, this may have led to misclassification of exposure and potential attenuation of the observed associations. Although such non-differential misclassification would likely bias results toward the null, the possibility of meaningful shifts in dietary intake over time remains and should be considered when interpreting the results. Dietary habits represent a modifiable lifestyle factor, and promoting aligned dietary patterns may still hold promise for public health strategies aimed at reducing IBS incidence, especially when tailored to individual and cultural contexts. However, these implications should be interpreted cautiously given the modest absolute effect. Further longitudinal and interventional studies are needed to clarify the causal role of DI-GM diets in IBS prevention.
The strength of this study lies in its first demonstration of long-term risk of IBS associated with DI-GM, a novel dietary index reflecting gut microbiota profiles. Moreover, the large sample size and the prospective design enable sufficient events and enough statistical power, as well as adjustment for a wide range of covariates. In addition, we conducted a series of subgroup and sensitivity analyses to furthermore verify the robustness of our results.
Despite these strengths, potential limitations should be noted. Firstly, IBS may be underdiagnosed, even though we identified IBS cases based on ICD-10 codes from both primary care and hospital records. However, we also integrated data from DHQ with Rome III criteria, and the results were consistent. Secondly, given the underdiagnosis and misclassification of IBS in routine clinical practice, the possibility of outcome misclassification remains a potential limitation. Although several sensitivity analyses yielded similar findings, this issue may still affect the validity of the findings. Thirdly, owing the absence of Rome IV consideration during the process of data collection, we could not integrate or contrast the more recent Rome IV criteria, which have a lower estimated prevalence and different diagnostic thresholds. This inconsistency may introduce heterogeneity in outcome ascertainment and limit comparability with newer studies. Fourthly, although dietary data were collected using the Oxford WebQ, a validated dietary recall tool, recall bias may still exist. Additionally, as the dietary assessment was based on a single 24-hour recall, this one-time measurement might capture an unrepresentative or atypical day of intake rather than habitual dietary patterns. However, the use of average value from multiple dietary assessment cycles substantially mitigated such bias [52]. Fifthly, the use of a baseline dietary assessment without repeated measures over long-term follow-up may lead to exposure misclassification due to changes in dietary habits over time, potentially attenuating the observed associations. Nevertheless, despite this bias toward the null, our findings still demonstrated a significant beneficial effect of higher DI-GM scores in reducing IBS risk, which underscores the robustness of the observed association. Sixthly, owing to lack of subtype data, we could not examine the association between DI-GM and risk of specific IBS subtype (e.g., constipation-predominant vs. diarrhea-predominant IBS). Given that dietary components, particularly fiber, may differentially affect IBS pathophysiology across subtypes, this association may vary by subtype. Future studies with detailed subtype information are needed to explore this potential heterogeneity. Seventhly, as in many studies, all confounders were assessed once at baseline. However, these confounders may change during the follow-up period, which might also bias the association. Finally, considering the predominantly of White ethnicity in our cohort, our findings should be interpreted with caution when generalizing to other ethnic or geographic regions due to substantial variations in different dietary, genetic backgrounds and gut microbiota profiles. Consequently, future well-designed research is warranted to validate our findings in diverse populations.
Conclusions
In summary, higher DI-GM scores, reflecting dietary patterns beneficial to the gut microbiota, are associated with lower risk of developing IBS. These findings highlight the importance of nutrition-based strategies, particularly diet favorable to gut microbiota, for the prevention and management of IBS. However, due to the lack of repeated dietary assessments over the follow-up period, these findings should be interpreted with caution. Future research should prioritize the integration of both plasma and fecal metabolomic profiles to comprehensively elucidate the mechanistic pathways and provide novel evidence for future dietary guidelines, as well as incorporate repeated dietary measures to better capture longitudinal dietary habits.
Supplementary Information
Supplementary Material 1: Supplementary Table 1. Association between quartiles of DI-GM scores and risk of incident IBS by age. Supplementary Table 2. Association between quartiles of DI-GM scores and risk of incident IBS by sex. Supplementary Table 3. Association between quartiles of DI-GM scores and risk of incident IBS by BMI. Supplementary Table 4. Association between quartiles of DI-GM scores and risk of incident IBS by smoking status. Supplementary Table 5. Association between quartiles of DI-GM scores and risk of incident IBS by drinking status. Supplementary Table 6. Association between quartiles of DI-GM scores and risk of incident IBS by TDI. Supplementary Figure 1. Associations between individual DI-GM components with risk of IBS.
Abbreviations
- DI-GM
Dietary Index for Gut Microbiota
- IBS
Irritable Bowel Syndrome
- HR
Hazard Ratio
- CI
Confidence Interval
- FODMAPs
Fermentable Oligosaccharides, Disaccharides, Monosaccharides, and Polyols
- UKB
UK Biobank
- ICD-10
International Classification of Disease-10
- DHQ
Digestive Healthcare Questionnaire
- TDI
Townsend Deprivation Index
- BMI
Body Mass Index
- IPAQ
International Physical Activity Questionnaire
- SD
Standard Deviation
- SCFAs
Short-Chain Fatty Acids
- BGM
Brain-Gut-Microbiota
- EECs
Enteroendocrine Cells
- ECCs
Enterochromaffin Cells
- AMED
Alternate Mediterranean Diet
- AEDII
Anti-Empirical Dietary Inflammatory Index
- AHEI-2010
Alternate Healthy Eating Index-2010
- HPDI
Healthful Plant-based Diet Index
- MIND
Mediterranean-DASH Diet Intervention for Neurodegenerative Delay
- IBS-D
Diarrhea predominant-IBS
Authors’ contributions
SSW designed the study. YGL and SSW drafted the manuscript. YGL analyzed the data. SSW and QZ verified the analysis. YGL, SL and STZhang interpreted the results, incorporated comments for the co-authors and finalized the manuscript. All authors approved the final version of the paper.
Funding
This work was supported by National Key Research and Development Program of China (No. 2022YFC2504002), National Natural Science Foundation of China (No. 82570631) and Beijing Nova Program (No. 20230484349). None of the funding organizations had any role in the design and conduct of the study; in the collection, analysis and interpretation of the data; in the preparation, review and approval of the manuscript.
Data availability
All data relevant to the study were using the UK Biobank Resource under application number 74444. The data that support the findings of this study are available from the corresponding author [SSW]. The data are not publicly available due to the privacy and restrictions of UKB.
Declarations
Ethics approval and consent to participate
The UK Biobank study was approved by the North West Multicenter Research Ethical Committee (21/NW/0157), and all participants or their proxy respondents provided written informed consent.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Patient and public involvement
Patients and the public were not involved in the design, conduct, reporting or dissemination plans of this research.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Sperber AD, Bangdiwala SI, Drossman DA, Ghoshal UC, Simren M, Tack J, et al. Worldwide prevalence and burden of functional Gastrointestinal Disorders, results of Rome foundation global study. Gastroenterology. 2021;160:99–e1143. [DOI] [PubMed] [Google Scholar]
- 2.Almario CV, Sharabi E, Chey WD, Lauzon M, Higgins CS, Spiegel BMR. Prevalence and burden of illness of Rome IV irritable bowel syndrome in the united states: results from a nationwide Cross-Sectional study. Gastroenterology. 2023;165:1475–87. [DOI] [PubMed] [Google Scholar]
- 3.Shin A, Xu H. Healthcare costs of irritable bowel syndrome and irritable bowel syndrome subtypes in the united States. Official J Am Coll Gastroenterol | ACG. 2024;119:1571–9. [DOI] [PubMed] [Google Scholar]
- 4.Goodoory VC, Ng CE, Black CJ, Ford AC. Direct healthcare costs of Rome IV or Rome III-defined irritable bowel syndrome in the united Kingdom. Aliment Pharmacol Ther. 2022;56:110–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Goodoory VC, Ng CE, Black CJ, Ford AC. Impact of Rome IV irritable bowel syndrome on work and activities of daily living. Aliment Pharmacol Ther. 2022;56:844–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Staudacher HM, Scholz M, Lomer MC, Ralph FS, Irving PM, Lindsay JO, et al. Gut microbiota associations with diet in irritable bowel syndrome and the effect of low FODMAP diet and probiotics. Clin Nutr. 2021;40:1861–70. [DOI] [PubMed] [Google Scholar]
- 7.Wang Y, Ma W, Mehta R, Nguyen LH, Song M, Drew DA, et al. Diet and gut microbial associations in irritable bowel syndrome according to disease subtype. Gut Microbes. 2023;15:2262130. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Holscher HD. Dietary fiber and prebiotics and the Gastrointestinal microbiota. Gut Microbes. 2017;8:172–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Horn J, Mayer DE, Chen S, Mayer EA. Role of diet and its effects on the gut Microbiome in the pathophysiology of mental disorders. Transl Psychiatry. 2022;12:164. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Vervier K, Moss S, Kumar N, Adoum A, Barne M, Browne H, et al. Two microbiota subtypes identified in irritable bowel syndrome with distinct responses to the low FODMAP diet. Gut. 2022;71:1821–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Schulze MB, Martínez-González MA, Fung TT, Lichtenstein AH, Forouhi NG. Food based dietary patterns and chronic disease prevention. BMJ. 2018;361:k2396. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Shan Z, Wang F, Li Y, Baden MY, Bhupathiraju SN, Wang DD, et al. Healthy eating patterns and risk of total and Cause-Specific mortality. JAMA Intern Med. 2023;183:142–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kase BE, Liese AD, Zhang J, Murphy EA, Zhao L, Steck SE. The development and evaluation of a Literature-Based dietary index for gut microbiota. Nutrients. 2024;16;7:1045. [DOI] [PMC free article] [PubMed]
- 14.Niu Y, Xiao L, Feng L. Association between dietary index for gut microbiota and metabolic syndrome risk: a cross-sectional analysis of NHANES 2007–2018. Sci Rep. 2025;15:15153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Shi F, Yang D, Sun Q, Fang C. Association of the newly proposed dietary index for gut microbiota and hyperlipidemia: from the 2007–2020 NHANES study. PLoS ONE. 2025;20:e0323887. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Zheng C, Qi Z, Chen R, Liao Z, Xie L, Zhang F. The association between the dietary index for gut microbiota and non-alcoholic fatty liver disease and liver fibrosis: evidence from NHANES 2017–2020. BMC Gastroenterol. 2025;25:163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Li DR, Liu BQ, Li MH, Qin Y, Liu JC, Zheng WR, et al. Dietary index for gut microbiota and risk of Gastrointestinal cancer: a prospective gene-diet study. Nutr J. 2025;24:81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Wu S, Yuan C, Yang Z, Liu S, Zhang Q, Zhang S, et al. Non-alcoholic fatty liver is associated with increased risk of irritable bowel syndrome: a prospective cohort study. BMC Med. 2022;20:262. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Yu S, Zhou Y, Liu S, Zhang Q, Zhang S, Zhu S, et al. Both general and central obesity are associated with increased risk of irritable bowel syndrome: A large-scale prospective cohort study. Am J Clin Nutr. 2025;121:1054–62. [DOI] [PubMed] [Google Scholar]
- 20.Sudlow C, Gallacher J, Allen N, Beral V, Burton P, Danesh J, et al. UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med. 2015;12:e1001779. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Bradbury KE, Young HJ, Guo W, Key TJ. Dietary assessment in UK biobank: an evaluation of the performance of the touchscreen dietary questionnaire. J Nutr Sci. 2018;7:e6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Piernas C, Perez-Cornago A, Gao M, Young H, Pollard Z, Mulligan A, et al. Describing a new food group classification system for UK biobank: analysis of food groups and sources of macro- and micronutrients in 208,200 participants. Eur J Nutr. 2021;60:2879–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Perez-Cornago A, Pollard Z, Young H, van Uden M, Andrews C, Piernas C, et al. Description of the updated nutrition calculation of the Oxford WebQ questionnaire and comparison with the previous version among 207,144 participants in UK biobank. Eur J Nutr. 2021;60:4019–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Fujiwara Y, Kubo M, Kohata Y, Machida H, Okazaki H, Yamagami H, et al. Cigarette smoking and its association with overlapping gastroesophageal reflux Disease, functional Dyspepsia, or irritable bowel syndrome. Intern Med. 2011;50:2443–7. [DOI] [PubMed] [Google Scholar]
- 25.Johannesson E, Simrén M, Strid H, Bajor A, Sadik R. Physical activity improves symptoms in irritable bowel syndrome: A randomized controlled trial. Official J Am Coll Gastroenterol | ACG. 2011;106:915–22. [DOI] [PubMed] [Google Scholar]
- 26.Sámano R, Esparza-Juárez F, Chico-Barba G, González-Medina E, Sánchez-Jiménez B, Hernández-Trejo M. Association of Diet, body mass Index, and lifestyle on the Gastrointestinal health risk in a sample of adults. Int J Environ Res Public Health. 2022;19:10569. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Gao X, Tian S, Huang N, Sun G, Huang T. Associations of daily sedentary behavior, physical activity, and sleep with irritable bowel syndrome: A prospective analysis of 362,193 participants. J Sport Health Sci. 2024;13:72–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.UK alcohol unit guidance. CMOs’ Low Risk Drinking Guidelines. Available from: https://www.drinkaware.co.uk/
- 29.Trindade IA, Hreinsson JP, Melchior C, Algera JP, Colomier E, Törnblom H, et al. Global prevalence of psychological distress and comorbidity with disorders of Gut-Brain interactions. Official J Am Coll Gastroenterol | ACG. 2024;119:165–75. [DOI] [PubMed] [Google Scholar]
- 30.Rohr MW, Narasimhulu CA, Rudeski-Rohr TA, Parthasarathy S. Negative effects of a High-Fat diet on intestinal permeability: A review. Adv Nutr. 2020;11:77–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Oliver A, Chase AB, Weihe C, Orchanian SB, Riedel SF, Hendrickson CL et al. High-Fiber, Whole-Food dietary intervention alters the human gut Microbiome but not fecal Short-Chain fatty acids. mSystems. 2021;6:e00115–21. [DOI] [PMC free article] [PubMed]
- 32.Zhao A, Li J, Peterson M, Black M, Gaulke CA, Jeffery EH, et al. Cooked broccoli alters cecal microbiota and impacts microbial metabolism of Glucoraphanin in lean and obese mice. Mol Nutr Food Res. 2025;69:e202400813. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Rodríguez-Daza MC, Roquim M, Dudonné S, Pilon G, Levy E, Marette A, et al. Berry polyphenols and fibers modulate distinct microbial metabolic functions and gut microbiota Enterotype-Like clustering in obese mice. Front Microbiol. 2020;11:2032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Pérez-Burillo S, Navajas-Porras B, López-Maldonado A, Hinojosa-Nogueira D, Pastoriza S, Rufián-Henares JÁ. Green tea and its relation to human gut Microbiome. Molecules. 2021;26:3907. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.El-Salhy M, Valeur J, Hausken T, Gunnar Hatlebakk J. Changes in fecal short-chain fatty acids following fecal microbiota transplantation in patients with irritable bowel syndrome. Neurogastroenterol Motil. 2021;33:e13983. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Siddiqui MT, Cresci GAM. The Immunomodulatory functions of butyrate. J Inflamm Res. 2021;14:6025–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Cheng H, Zhang D, Wu J, Liu J, Zhou Y, Tan Y, et al. Interactions between gut microbiota and polyphenols: A mechanistic and metabolomic review. Phytomedicine. 2023;119:154979. [DOI] [PubMed] [Google Scholar]
- 38.Camilleri M, Boeckxstaens G. Irritable bowel syndrome: treatment based on pathophysiology and biomarkers. Gut. 2023;72:590–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Martin CR, Osadchiy V, Kalani A, Mayer EA. The Brain-Gut-Microbiome axis. Cell Mol Gastroenterol Hepatol. 2018;6:133–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Shang X, Liu J, Zhu Z, Zhang X, Huang Y, Liu S, et al. Healthy dietary patterns and the risk of individual chronic diseases in community-dwelling adults. Nat Commun. 2023;14:6704. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Eslampour E, Ghanadi K, Aghamohammadi V, Kazemi AM, Mohammadi R, Vahid F, et al. Association between dietary inflammatory index (DII) and risk of irritable bowel syndrome: a case-control study. Nutr J. 2021;20:60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Staudacher HM, Mahoney S, Canale K, Opie RS, Loughman A, So D, et al. Clinical trial: A mediterranean diet is feasible and improves Gastrointestinal and psychological symptoms in irritable bowel syndrome. Aliment Pharmacol Ther. 2024;59:492–503. [DOI] [PubMed] [Google Scholar]
- 43.Meslier V, Laiola M, Roager HM, De Filippis F, Roume H, Quinquis B, et al. Mediterranean diet intervention in overweight and obese subjects lowers plasma cholesterol and causes changes in the gut Microbiome and metabolome independently of energy intake. Gut. 2020;69:1258–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Ghosh TS, Rampelli S, Jeffery IB, Santoro A, Neto M, Capri M, et al. Mediterranean diet intervention alters the gut Microbiome in older people reducing frailty and improving health status: the NU-AGE 1-year dietary intervention across five European countries. Gut. 2020;69:1218–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Saleem MM, Masood S, Rahmatullah MM, Ayesha Imdad I, Mohammed Aslam Sange A, Nasr D. Gut microbiota dysbiosis and its role in the development of irritable bowel syndrome. Cureus. 2025;17:e83084. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Yan J, Wang L, Gu Y, Hou H, Liu T, Ding Y et al. Dietary patterns and gut microbiota changes in inflammatory bowel disease: current insights and future challenges. Nutrients. 2022;14(19):4003. [DOI] [PMC free article] [PubMed]
- 47.Nouri-Majd S, Salari-Moghaddam A, Keshteli AH, Esmaillzadeh A, Adibi P. The association between adherence to the MIND diet and irritable bowel syndrome. Dig Dis. 2022;40:198–205. [DOI] [PubMed] [Google Scholar]
- 48.Black CJ, Staudacher HM, Ford AC. Efficacy of a low FODMAP diet in irritable bowel syndrome: systematic review and network meta-analysis. Gut. 2022;71:1117–26. [DOI] [PubMed] [Google Scholar]
- 49.Tunali V, Arslan N, Ermiş BH, Derviş Hakim G, Gündoğdu A, Hora M, et al. A multicenter randomized controlled trial of Microbiome-Based artificial Intelligence-Assisted personalized diet vs Low-Fermentable Oligosaccharides, Disaccharides, Monosaccharides, and polyols diet: A novel approach for the management of irritable bowel syndrome. Am J Gastroenterol. 2024;119:1901–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Kasti AN, Katsas K, Pavlidis DE, Stylianakis E, Petsis KI, Lambrinou S et al. Clinical trial: A mediterranean Low-FODMAP diet alleviates symptoms of Non-Constipation IBS-Randomized controlled study and volatomics analysis. Nutrients. 2025;17(9):1545. [DOI] [PMC free article] [PubMed]
- 51.Nybacka S, Törnblom H, Josefsson A, Hreinsson JP, Böhn L, Frändemark Å, et al. A low FODMAP diet plus traditional dietary advice versus a low-carbohydrate diet versus Pharmacological treatment in irritable bowel syndrome (CARIBS): a single-centre, single-blind, randomised controlled trial. Lancet Gastroenterol Hepatol. 2024;9:507–20. [DOI] [PubMed] [Google Scholar]
- 52.Ambrosini GL, Hurworth M, Giglia R, Trapp G, Strauss P. Feasibility of a commercial smartphone application for dietary assessment in epidemiological research and comparison with 24-h dietary recalls. Nutr J. 2018;17:5. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Material 1: Supplementary Table 1. Association between quartiles of DI-GM scores and risk of incident IBS by age. Supplementary Table 2. Association between quartiles of DI-GM scores and risk of incident IBS by sex. Supplementary Table 3. Association between quartiles of DI-GM scores and risk of incident IBS by BMI. Supplementary Table 4. Association between quartiles of DI-GM scores and risk of incident IBS by smoking status. Supplementary Table 5. Association between quartiles of DI-GM scores and risk of incident IBS by drinking status. Supplementary Table 6. Association between quartiles of DI-GM scores and risk of incident IBS by TDI. Supplementary Figure 1. Associations between individual DI-GM components with risk of IBS.
Data Availability Statement
All data relevant to the study were using the UK Biobank Resource under application number 74444. The data that support the findings of this study are available from the corresponding author [SSW]. The data are not publicly available due to the privacy and restrictions of UKB.


