The Impact of Dietary Interventions on the Microbiota in Inflammatory Bowel Disease: A Systematic Review

Abstract Background and Aims Diet plays an integral role in the modulation of the intestinal environment, with the potential to be modified for management of individuals with inflammatory bowel disease [IBD]. It has been hypothesised that poor ‘Western-style’ dietary patterns select for a microbiota that drives IBD inflammation and, that through dietary intervention, a healthy microbiota may be restored. This study aimed to systematically review the literature and assess current available evidence regarding the influence of diet on the intestinal microbiota composition in IBD patients, and how this may affect disease activity. Methods MEDLINE, EMBASE, Scopus, Web of Science, and Cochrane Library were searched from January 2013 to June 2023, to identify studies investigating diet and microbiota in IBD. Results Thirteen primary studies met the inclusion criteria and were selected for narrative synthesis. Reported associations between diet and microbiota in IBD were conflicting due to the considerable degree of heterogeneity between studies. Nine intervention studies trialled specific diets and did not demonstrate significant shifts in the diversity and abundance of intestinal microbial communities or improvement in disease outcomes. The remaining four cross-sectional studies did not find a specific microbial signature associated with habitual dietary patterns in IBD patients. Conclusions Diet modulates the gut microbiota, and this may have implications for IBD; however, the body of evidence does not currently support clear dietary patterns or food constituents that are associated with a specific microbiota profile or disease marker in IBD patients. Further research is required with a focus on robust and consistent methodology to achieve improved identification of associations.


Introduction
Inflammatory bowel diseases [IBD] encompass a collection of chronic, relapsing-remitting disorders of the gastrointestinal [GI] tract, characterised by an inflammatory process that requires extensive medical and surgical management.Crohn's disease [CD] and ulcerative colitis [UC] are the two most prevalent manifestations of IBD and affect up to 0.5% of the global population. 1,2Although the aetiology of IBD is yet to be fully elucidated, the current pathogenesis paradigm involves a complex interplay between multiple factors, including genetic susceptibility and environmental triggers. 3mong these environmental exposures, diet may be a key driver of intestinal inflammation, as it plays a fundamental role in modulating the intestinal milieu and physiology, with modifications to diet strongly influencing the intestinal microbiota composition and function. 4,5Excessive changes to the intestinal microbiota may result in a disrupted equilibrium between commensal and harmful microbes, often referred to as 'dysbiosis' and characterised by a reduction in microbial diversity. 6Microbial dysbiosis has been shown to perturb intestinal barrier integrity, which facilitates the translocation of potentially antigenic microbiota and dietderived components into the underlying mucosa, thus driving an aberrant immune response and an ongoing inflammatory cycle. 76][17] Furthermore, diet may also exert functional changes to the microbiota by facilitating the production of metabolites such as short chain fatty acids [SCFAs] which are known to have an immune regulatory role and serve as a primary energy source for intestinal epithelial cells. 18This underscores diet as a key modifiable factor that can directly influence the GI environment.
The efficacy of exclusive enteral nutrition [EEN] [which replaces solid foods with a nutritionally complete liquid formula for up to 8 weeks] in paediatric patients with active CD 19 provides the strongest evidence for the link between dietary exposure and disease. 20,21However, EEN is underused in adult populations 22 due to unpalatability, low tolerability and poor compliance with such a restrictive diet. 22,23herefore, dietary approaches that allow access to whole foods, to maximise food variety, while optimally mitigating symptoms need to be explored.To date, evidence regarding the implementation of predefined diets [ie, dietary regimens that involve a significant restriction or complete exclusion of one or more suspect food groups] for ameliorating the course of IBD is conflicting and inconsistent. 24,25Many patients perceive diet to be an initiating factor in their disease or in evoking relapse, and have long sought dietary advice as part of IBD management. 26However, there is still no consensus nor clear indications toward an optimal dietary therapy beyond healthy eating guidelines, due to limited available research data. 27he Western-style dietary pattern is considered a predominant trigger implicated in the development of IBD, however, the specific dietary components responsible are yet to be identified. 28,29This dietary pattern, which is typically devoid in fibre from fruits and vegetables and rich in saturated fat and digestible carbohydrates, tends to evoke dysbiosis [reduced microbial diversity], resulting in dysregulated production of SCFAs by bacteria in the gut, which has profound effects on maintaining intestinal homeostasis. 30,31dditionally, this diet facilitates the expansion and activity of colonic mucus-degrading bacteria, which drives intestinal inflammation and gut barrier impairment. 30,32Dietary therapy has the potential to manipulate the gut microbiota to restore a healthy profile and alleviate intestinal inflammation in IBD patients.Therefore, the impact of diet on the intestinal microbiota and its potential to shift the future course of the disease should be considered when developing nutritional guidelines for IBD patients.An updated 2022 systematic review and meta-analysis analysed 27 controlled trials of solid food diets for the induction or maintenance of remission in IBD, and concluded that currently available evidence remains low. 33Moreover, this study did not consider the potential impact of modulation of the intestinal microbiota, through diet, on disease outcomes.Given this knowledge gap, the aim of this review is to describe relationships between whole-food dietary exposure and the intestinal microbiota in adult patients with IBD, and whether this influences disease activity.

Methods
A systematic literature review was conducted in agreement with the Preferred Reporting Items for Systematic reviews and Meta-Analyses [PRISMA] guidelines 34 and by using the PRISMA 2020 Checklist [Supplementary Table S1].Details of the methods, including the search strategy, eligibility criteria, extraction process, and analysis, were registered with the International Prospective Register of Systematic Reviews [registration no.CRD42022292986; http://www.crd.york.ac.uk/PROSPERO]. 35The PICOS [Participants, Interventions, Comparisons, Outcomes and Study design] model was used to define the research question [Supplementary Table S2].

Search strategy
A systematic literature search was performed in the electronic databases MEDLINE, EMBASE, Cochrane Library, Scopus, and Web of Science, using a combination of freetext words, medical subject headings [MeSH] terms, and synonyms relevant to this review, with the assistance of two experienced research librarians [AS and JB].Our search strategy included terms related to [i] inflammatory bowel diseases, [ii] microbiome/microbiota, and [iii] diet/ food/nutrition [Supplementary Table S3].To ensure the studies reflected contemporary microbiome sequencing techniques, the search strategy was restricted to those published after the completion of the first phase of the Human Microbiome Project in June 2012, which characterised the microbial communities from five major sites of the human body including the GI tract. 36,37The databases were searched from January 1, 201 to June 6, 2023, and limited to the English language, as translation services were not available.Abstracts, review articles, editorials, case reports, case series, theses, and study protocols were excluded.

Risk of bias assessment
Quality appraisal of all included studies was undertaken to reduce the risk of bias [Supplementary Table S4].Risk of bias was independently assessed during the selection of studies by two reviewers [CN and JP] using the Joanna Brigg's Institute [JBI] critical appraisal tools for randomised controlled trials [RCTs] [13 criteria], and quasi-experimental [nine criteria] and cross-sectional studies [eight criteria]. 38JBI includes questions about suitability of the study sample and selection, description of subjects and setting, completeness of presented data and analysis, appropriateness of measuring the condition, and identification of confounding factors.

Data availability statement
The data underlying this article are available in the article and in its online Supplementary material.

Risk of bias
Overall, 100% [n = 9] of the intervention studies [n = 7 RCTs and n = 2 quasi-experimental] 39-44,49-51 met at least 60% [8/13 and 7/9, respectively] of the JBI study quality criteria [Supplementary Table S4].RCTs scored poorly for criteria relating to blinding of participants and evaluators to allocation and/or treatment, due to the nature of food-based dietary interventions.Of cross-sectional studies, 45,47,48 75% [n = 3 of 4] satisfied at least five of eight items, with one study scoring as low as two of eight.For 50% [n = 2 of 4] 46,48 of these, potential biases stemmed from the lack of validated or reliable measures and neglecting to identify and adjust for confounding factors.According to these assessments, the overall methodological quality of included studies was deemed to be moderate.
Five of the nine experimental studies investigated UC, two studies referred to CD, and two examined both disease types [Table 1].Three of the four observational studies reported on both disease types and healthy controls [HC], and one was performed in a UC cohort only.The population analysed in the research studies included a total of 3427 individuals with IBD: 1878 diagnosed with CD and 1549 with UC, with sample sizes ranging between 15 and 1202 participants.All studies included both male and female participants, with the mean ages ranging from 29 to 53 years.Healthy mixed-sex cohorts were included in three of the four cross-sectional studies [total n = 228], with participant mean ages ranging from 48 to 55 years.

Cross-sectional studies' characteristics and outcomes
The four cross-sectional studies investigated habitual dietary habits of IBD patients and healthy individuals [Table 2].Three of the four studies evaluated the normal or habitual diet of IBD patients compared with healthy participants.The dietary assessment methods used were unique between studies and the food data analysis tools used were not reported in any of the studies except for Weng et al. 48This study demonstrated significantly lower intake of vegetables, fish, and shellfish, and higher intake of eggs, milk, and dairy products, in a cohort of IBD patients compared with HC, especially in CD patients.Comparing with HC, Teofani et al. 47 also reported significantly lower intake of vegetables, legumes, fruits, and cereals, and high intakes of dairy products in IBD patients.Conversely, Berbisá et al. 45 did not find any significant differences in the patterns of eating habits between UC patients and HC, with both groups consuming a Western-style dietary pattern.Schreiner et al. 46 investigated restrictive dietary habits of IBD patients, and defined the participants as either being vegetarian, non-vegetarian [meat-eating], gluten-free diet, or non-gluten-free, based on the frequency of their meat or gluten consumption, respectively.However, this study did not further analyse differences in intakes of specific dietary components between groups.

Dietary interventions' characteristics and outcomes
A total of nine unique dietary patterns was identified across the nine diet intervention studies [Table 3].These included a  and standard-of-care [SOC] diet.The majority of intervention studies had two trial arms, [39][40][41][42]44 except for one single-arm study 51 and one with three trial arms 43 which included patients who underwent faecal transplant [FT] using microbiota obtained from healthy donors. Theduration of the dietary interventions ranged from 7 days to 12 weeks. Diseae states varied considerably within and between studies, from quiescent to active or mild to moderate or severe.Dietary data were captured at different time points and frequencies throughout the trials from baseline to end of trial.The dietary assessment and food analysis methods used to obtain information on dietary intake were highly variable.Three studies applied food diaries, 39,40,44 four studies used diet adherence questionnaires, 41,43,49,50 and five studies administered 24-h food recall questionnaires.40,41,[49][50][51] The diet protocol inclusion and restrictions highlight that some dietary modifications centred around the restriction of specific dietary components, and others were focused on the complete exclusion of single or multiple food groups.Most studies [n = 7 of 9] implemented increased intakes of fruits, vegetables, and fibre and restricted consumption of certain meats such as red and processed meats.Five studies reduced the intake of specific dietary fats 40,43,44,49,51 [eg, animal fat, saturated and unsaturated fat, and omega 6:3 ratio], and four of these studies restricted consumption of food additives such as emulsifiers.43,44,49,51 The impact of a 12-week Mediterraneaninspired diet, which is chiefly characterised by increased intake of plant-based foods and limited consumption of red and processed meat, was investigated by Zhang et al., 44 Lewis et al., 41 and Haskey et al. 49 All three studies reported significant increase in the consumption of fruits and vegetables from baseline to end of trial in IBD participants, with Lewis et al. and Haskey et al. also noting significant improvement in diet quality and adherence in the MD groups compared with the habitual diet groups.Similar results were observed in the SCD arm as the MD arm by Lewis et al.;, however, the adherence rates for both diets were insufficient [64% and 68%, respectively].
By contrast, Strauss et al. 50trialled an 8-week MD diet with a special focus on reducing sulphur intake in UC patients, but found no significant differences in sulphur consumption and diet quality between the MD group and those consuming a habitual diet.A 4-week low-fat diet by Fritsch et al. 40 led to a significant reduction in total fat [saturated and unsaturated] and omega 6:3 ratio in UC patients, and had a similar high adherence rate [87%] as the comparator iSAD group [85%].Similarly, Cox et al. 39 reported high compliance with a 4-week low FODMAP diet [88%] in UC patients, as well as significant reductions in FODMAP, protein, fat, and sugar intake compared with a sham diet.The UCED diet used by Shabat et al., 43 which restricted total protein and fat [especially saturated fats] from animal sources as well as food additives, demonstrated the highest adherence rate [100%] after 8 weeks.Olendzki et al. 51 trialled an 8-week IBD-AID diet in UC patients which also involved limiting consumption of 'adverse foods' such as animal protein and fat, and food additives from processed foods, while increasing intake of 'beneficial foods' eg, fruits, vegetables, and omega-3.This study did not measure diet adherence or quality, but a significant increase in consumption of prebiotics, probiotics, and 'beneficial foods', and decrease in consumption of 'adverse foods' according to the diet protocol [Table 3], was reported at the end of the trial compared with baseline [habitual diet].
Although studies investigating enteral nutrition was an exclusion criterion in the literature search process, Sahu et al. 42 was included in this review due to the comparator group receiving standard-of-care IBD diet.The SOC diet involved a normal whole-food diet coupled with dietary counselling about calorie and protein intake, which was found to be comparable to exclusive enteral nutrition [EEN] in UC patients; however, corticosteroid failure occurred in more patients in the EEN group compared with SOC.

Changes in GI symptoms and quality of life
Only two of the nine intervention studies evaluated GI symptoms using a range of validated tools [Table 4].Lewis et al. 41 reported significant symptom improvement in CD patients on both the MD and SCD diets from screening to Week 6, but this change did not significantly differ between the treatment arms.In contrast, Cox et al. 39 showed that a low FODMAP diet was significantly more effective in facilitating symptom relief, with reductions in bloating and flatulence severity and daily stool frequency, compared with the sham [low FODMAP] diet.This improvement in symptoms was defined by the reduction in IBS severity scoring system, and when patients were stratified by IBD type, this was more prominent in UC .
These same two studies 39,41 and Fritsch et al. 40 also reported on the impact of diet on the quality of life [QoL] of patients using a variety of validated methods [Table 4].Lewis et al. 41 showed that both the MD and SCD diets resulted in a significant QoL improvement in a cohort of CD patients at the end of the trial compared with baseline.However, no difference was seen when the two groups were compared.The iSAD diet investigated by Fritsch et al. also improved the overall QoL in IBD patients, but to a lesser extent than its low-fat diet comparator.When stratified by IBD subtypes, this effect was observed in UC but not in the CD group.The MD diet 49 also led to an improvement in QoL, represented by a significant relief in some bowel symptoms of UC patients when compared with the habitual diet group.The low FODMAP diet used by Cox et al., 39 but not the sham diet, also led to better healthrelated quality of life [HR-QoL] when CD and UC were analysed together, but not separately.5].Notably, eight out of nine studies used FC testing with a significant reduction in FC concentration only being observed in UC patients who were on EEN 42 and in CD patients who were on an SCD diet. 41Similarly, most studies also measured CRP and found no difference in levels following the diets except for the UC patients in the EEN group, where a reduction in CRP concentration was observed after just 5 days of the 7-day trial.

Changes in clinical activity and inflammatory markers
Cox et al. 39 did not display any changes in disease activity in any of the parameters measured other than the UC subgroup demonstrating a greater patient-perceived control of IBD following a low FODMAP diet.Levels of pro-inflammatory cytokines were quantified by Fritsch et al. 40 and Olendzki et al. 51 ; however, neither study detected significant differences

Microbiota analysis methods
Overall, 12 out of 13 studies evaluated the intestinal microbiota from stool, two studies examined the intestinal tissue [inflamed and non-inflamed from the ileo-rectal region], and one study 48 investigated both specimen types [Table 6].Most of the studies implemented commercial extraction kits according to the manufacturer's instructions for DNA extraction, but two studies 39,42 followed protocols established elsewhere.The sample collection kits used varied between studies, and samples were stored mostly at -80°C or at -20°C.Eight studies 40,42,44,45,[47][48][49][50] assessed the intestinal microbiota by 16S rRNA amplicon gene sequencing which targeted the V3-V4 variable regions, except for one study 46 that used the V5-V6 regions.The remaining studies 39,41,43,51 employed metagenomic shotgun sequencing [MGS].Analysis pipelines and reference databases used to assess taxonomy were also highly variable between studies.

Changes in diversity indices and relative abundances
All nine trials characterised the intestinal microbiota using stool samples and reported on alpha diversity [within sample variation] either by richness, observed species, Shannon, Simpson, Faith's phylogenetic diversity, or Pielou's evenness [Supplementary Table S7].However, no significant shifts were detected within or between any of the groups following diet intervention.Of note, Shabat et al. 43 did not perform microbiota sequencing of stool from UC patients who received either the UCED diet alone or the UCED or free diet in combination with FT.Instead, stool from healthy donors was analysed before and after conditioning with a specific diet for 2 weeks, but no changes in alpha diversity were observed.All four cross-sectional studies measured alpha diversity either by richness, Shannon, Simpson, Chao1, or observed operational taxonomic units [OTUs], and only two studies yielded significant differences.Weng et al. 48demonstrated a significantly lower Shannon index score in CD patients consuming their habitual diet compared with the HC group.Teofani et al. 47 also found significantly reduced alpha diversity, as measured by Shannon, Simpson, and Chao1, in IBD patients compared with HC consuming a habitual diet.
Beta diversity [across sample variation] was evaluated in all studies either by Bray-Curtis, Jaccard, Aitchison, Manhattan and Gower, or weighted or unweighted Unifrac.Statistical differences in beta diversity matrices were evaluated using either PERMANOVA or Mann-Whitney testing.Zhang et al. 44 reported a significant difference in beta diversity between a habitual diversified diet [similar to an MD] and a non-diversified diet at baseline.However, after CD patients with a non-diversified diet had followed a 12-week MD diet, this difference was no longer evident.Haskey et al. 49 also reported a significant shift in microbial composition between MD and CHD over time.Schreiner et al. 46 noted significant differences in beta diversity between IBD patients who ate a vegetarian, gluten-free, or regular diet.Finally, Teofani et al. 47 demonstrated reduced beta diversity in the IBD population consuming a normal diet compared with HC.

Data collection time points Changes in clinical disease activity and inflammatory markers
Baseline versus end of trial in relative microbial abundance were measured in 11 of the 13 included studies [Supplementary Tables S5 and S6].Cox et al. 39 analysed stool samples at the end of a low FODMAP diet and only recorded a significant decrease in Faecalibacterium prausnitzii and changes in three Bifidobacterium species, namely a decrease in Bifidobacterium adolescentis and dentium and an increase in Bifidobacterium longum; however, these significant differences were not evident at the genus level.Similarly, Lewis et al. 41 also observed a reduction in Faecalibacterium prausnitzii following an SCD diet intervention compared with an MD diet, and noted an increase in the family Enterobacteriaceae and lower Eubacterium eligens and rectale.Haskey et al. 49 found that species primarily belonging to the phylum Firmicutes, including Ruminococcus, Flavonifractor, Clostridium, Lactococcus, and Blautia A spp., were most positively associated with an MD diet.The most negatively associated species belonged to a mixture of phyla including Firmicutes, Actinobacteria, and Bacteriodota, with Bifidobacterium, Blautia, Veillonella, Streptococcus, and Massilioclostridium spp.being the most abundant.It is worth noting that Strauss et al. 50also examined changes in microbial taxa; however, the profiles were associated with changes in FC levels rather than relating to the MD diet.
The low-fat diet by Fritsch et al. 40 identified significant alterations in the faecal microbiota of UC participants at the phylum level, with an enrichment of Bacteroidetes and depletion of Actinobacteria reported after a LFD, whereas no changes in these phyla were observed with iSAD.Higher Prevotella and Faecalibacterium prausnitzii were also reported after an LFD compared with iSAD.The anti-inflammatory diet by Olendzki et al. 51 led to increased abundance of SCFA-producing bacteria belonging to Clostridia class at the species level, with Roseburia hominis being the most dominant in the stool samples of all IBD subjects, as well as in CD participants only.UC patients exhibited high abundances of Faecalibacterium prausnitzii, Eubacterium eligens, Coprococcus come, and multiple Bacteroides spp., whereas overall, Parabacteroides distasonis was consistently depleted in all IBD participants during the IBD-AID intervention.Conversely, Sahu et al. 42 only yielded significant increases in the genera Bifidobacterium and Anaerosinus, and a decline in Catenibacterium in the faecal samples of UC patients who received SOC relative to those who were on EEN.In the study by Zhang et al., 44 no significant microbial differences were found in the stool of CD patients eating a habitual diversified diet or a non-diversified diet when compared with their respective baseline samples.At baseline, significantly higher Proteobacteria [Escherichia Shigella] and Paraprevotella, and lower Faecalibacterium were reported in the diversified diet group compared with its counterpart.However, the opposite trend was observed in the non-diversified diet groups after being prescribed a Mediterranean-inspired diet at Weeks 4 and 12.
Of the cross-sectional studies, Schreiner et al. 46 identified significantly lower mucosal abundance of Barnesiellaceae family, Clostridiales order, and genera within the phylum Firmicutes such as Ruminococcus and Faecalibacterium prausnitzii, in IBD participants who were gluten-free compared with those consuming a regular diet.Representative genera from Bacteroidetes were also reduced in the glutenfree CD participants.By contrast, vegetarian UC patients had a higher abundance of Firmicutes including Blautia, Coprococcus, Dorea, and Ruminococcus, whereas Faecalibacterium prausnitzii was found to be lower when compared with meat-eating [regular diet] participants.When comparing CD and UC patients on a normal diet with HC, Teofani et al. 47 only reported significantly higher levels of faecal Atopobiaceae and Enterobacteriaceae in IBD, whereas Berbisá et al. 45 found significantly increased abundance of families and genera belonging to the phylum Verrucomicrobia, such as Akkermansiaceae and Veillonellaceae, in the stool of UC participants.Of the included studies, Weng et al. 48dentified the highest number of differentially abundant taxa [primarily at the class, order, and family levels] in the microbiota of IBD patients and HC, using a combination of mucosal and faecal samples.This study found significant enrichment of numerous bacteria belonging mainly to the phyla Proteobacteria, Actinobacteria, Acidobacteria, Deinococcota, Deferribacterota, and Fusobacteria, in both the CD and the UC patient samples, whereas HC only displayed a high abundance of Firmicutes.When sample source was considered, Weng et al. 48found that there was a significantly more abundant mucosa-associated microbiota compared with the faecal communities in both UC and CD patient samples.

Discussion
The current systematic review of 13 studies indicated inconsistent associations between diet and microbiota in IBD.
Overall, this study collated information from a total of 1775 IBD participants who were on a habitual diet and 469 IBD participants who received a predefined dietary intervention.The primary feature of the types of diets trialled focused on the reduction of foods suspected of triggering inflammation, such as animal fat, red and processed meat, alcohol, and food additives [eg, emulsifiers and artificial sweeteners], with an emphasis on increased intake of fresh fruits and vegetables, and fibre. 40,41,43,44Others excluded carbohydrates responsible for fuelling microbial overgrowth and inducing functional symptoms. 39,41Of the nine intervention studies, only four studies demonstrated positive changes in either GI symptoms or QoL.Cox et al. 39 assessed the impact of a low FODMAP diet on IBD patients with persistent GI symptoms, which met the Rome III criteria for IBS and found improvement in symptom scores and QoL, which is consistent with previous studies. 55,56However, a potential limitation of this study is the inability to differentiate between IBD and IBSassociated mediators of non-specific symptoms.As there is a high co-prevalence of IBD and IBS, 57 dietary changes may be alleviating IBS rather than IBD symptoms; therefore measures of disease severity are needed.Lewis et al. 41 demonstrated improvement in symptoms with both the SCD and MD diets, but neither diet was superior over the other in inducing symptomatic remission in CD patients.This may be due to similarities between the interventions, namely the emphasis on fresh fruits and vegetables, and the ratio of monounsaturated to polyunsaturated fatty acid consumption.The beneficial effect of an MD diet is supported by Haskey et al. 49 through the achievement of reduced bowel symptoms in UC patients following the intervention.Fritsch et al. 40 also reported improvement in clinical symptoms and QoL with a low-fat diet, and to a lesser extent awith n improved Western diet [iSAD diet], in UC patients.However, a higher fibre intake was also reinforced in both diets, and therefore, it is unclear whether these responses were a product of a low-fat or a high-fibre content.Dietary components that have been reported to prevent or induce IBD get confused with factors that may assist in management.Findings from this review indicate the potential for diet modulation to substantially improve symptoms and QoL.
Remarkably, a UC-specific exclusion diet by Shabat et al. 43 induced endoscopic remission and mucosal healing [regarded as the optimal outcome in IBD research] 58 in patients, which was highest among participants who were on the diet alone compared with those who either followed an exclusion or habitual diet but also underwent FT.The impact of this diet on the microbiota composition of these patients would have been enlightening, but this study only characterised the microbiota profile of the healthy faecal donors at pre-baseline and not the UC patients, which is a major limitation of this work.Moreover, the pre-conditioning diet provided to the donors was different from the exclusion diet prescribed to the UC patients, so it is difficult to draw conclusions regarding its efficacy.In concert, findings by Cox et al. 39 and Shabat et al. 43 suggest that complete dietary exclusion of specific food groups may have a more potent effect than a mere reduction of intake.However, improvements in clinical symptoms and QoL following a low-fat diet by Fritsch et al. 40 indicate that this may be dependent on the type of dietary component restricted.
With regards to the intestinal microbiota, none of the intervention or observational studies exhibited changes in alpha diversity metrics post-diet, and a significantly lower beta diversity was only observed in two of the four cross-sectional studies investigating habitual diets in IBD patients compared with non-IBD HCs.In agreement with findings from another trial in IBS patients, 59 Cox et al. 39 also reported increased abundance of faecal Bifidobacteria longum and adolescentis and lower total Faecalibacterium prausnitzii following a low FODMAP diet compared with sham diet.Lewis et al. 41 also reported decreased Faecalibacterium prausnitzii abundance in response to SCD.
Although SCD and low FODMAP dietary approaches have distinct principles, both diets target dietary carbohydrates.A low FODMAP diet excludes specific fermentable carbohydrates, whereas the more restrictive SCD prohibits consumption of disaccharides and complex carbohydrates known to support the growth of beneficial bacteria like Faecalibacterium prausnitzii. 60The reduction in dietary substrate with either low FODMAP or SCD may account for the decrease in their population compared with the other diets that showed an increase in Faecalibacterium prausnitzii abundance.Halmos et al. 61 have shown a dysbiotic pattern in quiescent CD patients, whereby the relative abundances of butyrate-producing Clostridium cluster XIVa [phylum Firmicutes] and mucus-associated Akkermansia muciniphila [phylum Verrucomicrobia] were lower and the mucus-degrading Ruminococcus torques higher, in the low FODMAP diet compared with a typical Australian diet. 61,62his is further corroborated by Png et al., 63 and together suggests that a restricted FODMAP intake may promote an unfavourable GI tract environment and predispose the intestinal mucosa to inflammation.Whereas a low FODMAP diet is recommended for managing individuals with IBS, the efficacy of this treatment in patients with quiescent IBD with concurrent IBS-like symptoms is controversial and is yet to established.Overall, four studies 41,44,49,50 featured the MD, which broadly appears to have positive associations bacteria within the dominant Firmicutes phylum, such as Faecalibacterium, Clostridium, and Blautia, consistently found to be significantly depleted in the gut of IBD patients, 8,64 indicating that diet is imperative in shifting the intestinal microbial dysbiosis towards a balanced community.
In a similar vein, Olendzki et al. 51 also showed that the IBD-AID can also favour these commonly diminished Firmicutes, including potent butyrate-producing species Roseburia hominis, Faecalibacterium prausnitzii, and Eubacterium eligens, which have been reported to have an anti-inflammatory capacity. 65,66In line with previous studies, Berbisá et al. 45 reported reduced abundance of mucus-fortifying bacteria Akkermansia in UC patients compared with healthy subjects; however, it is uncertain whether this is a cause or consequence of disease. 67,68The absence of Akkermansia may be due to decreased mucins in UC patients. 69Recent research demonstrated that oral supplementation of Akkermansia muciniphila in mice counteracted the low-grade intestinal inflammation and alterations in species composition that are otherwise induced by consumption of the emulsifiers carboxylmethylcellulose [CMC] and polysorbate 80 [P80]. 70Degradation of colonic mucin as a primary energy source by this bacterium results in the synthesis of SCFAs such as propionate and acetate, 67 which have anti-inflammatory effects and are vital substrates for maintaining intestinal barrier integrity. 71Propionate and acetate, along with butyrate, are also produced during anaerobic bacterial fermentation of non-digestible carbohydrates such as dietary fibre and resistant starch. 31In IBD patients, a significant depletion in dominant butyrate-producing bacteria such as Faecalibacterium prausnitzii and Roseburia intestinalis has been noted. 64,72Furthermore, SCFA concentrations may be associated with disease activity, as adult patients with active IBD have been found to have significantly reduced levels of butyric, acetic, and propionic acids in the faeces compared with healthy subjects. 65,73Given the ability of these SCFAs to influence the intestinal microbiota composition and function, investigation into metabolomics in future studies is needed.A schematic representation of significant changes relating to symptoms or QoL, and microbiota diversity and relative abundance found in IBD participants, based on dietary interventions is provided [Figure 2].
This systematic review has notable limitations that warrant discussion.First, the retrieved studies were substantially heterogeneous in terms of design, participants, dietary intervention type, duration, and outcomes, which prohibited a meta-analysis.The relatively small sample size of the included studies makes it challenging to draw generalisable conclusions from results.Poor dietary data reporting was highly prevalent, culminating from the lack of unified standards to assess dietary intake.Inadequate recording of background and diet information throughout the study periods further affected our ability to describe changes attributed to diet.Given that the study location of the included studies was divided between North America, Northern Europe, and Asia, regional diet variations are another potential source of heterogeneity in findings, adding to the layers of complexity and challenge in IBD research.Variations in culture, dietary beliefs and practices, culinary preferences, and food availability and accessibility, have an impact on dietary patterns based on geographical location.Although elements of the Western diet are most prevalent in North America [USA and Canada] and Northern Europe [UK and Denmark], there can be significant variations within countries.For example, an American diet consists of larger portions and excess sugar compared with the UK, 74 and other parts of Northern Europe, such as the Faroe Islands [Denmark], include a mix of traditional foods [eg, dried fish, whale, and blubber] within their diet. 45In contrast to Western diets, which often No change QoL IBS symptoms faecal slgA 49 Speci c carbohydrate diet 41 Mediterranean diet 40,41,44,49,50 IBD-antiin ammatory diet 51 Low FODMAP diet 39   75,76 Food preparation techniques also differ between regions, wherein roasting, frying, and baking are more prevalent in Western countries whereas Asian populations tend to prefer cooking methods like steaming and stir-frying, which can greatly influence food composition and content. 76,77Regional diet variations are therefore important to consider in IBD research, as they not only impede the identification of consistent dietary triggers, but can also influence the degree to which individuals adhere to a specific dietary pattern or intervention, and lead to variabilities in baseline composition of the gut microbiota. 78he area of microbiome research is rapidly evolving.As a result, there are limitations to the methodological approaches undertaken in this field.To date, 16S rRNA-based methods have been more commonly employed to analyse the microbiota, and this was used in eight of the 13 studies included in this review.However, due to the lack of species-level sensitivity of 16S amplicon sequencing, future studies should focus on using whole metagenome sequencing techniques to evaluate the effect of diet on individual species.The origin of microbial sampling is also important to consider for intestinal microbiota profiling, as faecal samples cannot accurately reveal the alterations of intestinal microbiota, whereas biopsy samples provide a more precise reflection of mucosaassociated microbiota. 79As such, this review is limited by the fact that only two of the 13 studies examined the microbiota using tissue samples.Seven 40,44,45,[47][48][49]51 of the 12 studies that examined faecal microbiota used samples obtained using at-home self-collection kits provided to participants, and four studies 39,[41][42][43] did not specify how stool samples were collected. On study 50 required participants to undergo baseline and end-of-trial endoscopic assessment, and although stool specimens were collected at these time points, it was unclear whether these were spontaneously passed or stool aspirate samples.Microbial profiles from aspirate samples can vary significantly from faecal profiles from the sample individual, 79,80 and therefore is important to consider the type of stool specimen analysed when interpreting and comparing study results.Furthermore, the great variability in the sample collection and extraction methods used, as well as the analysis pipelines and taxonomic databases referred to between the studies, may also impact on the outcomes.The duration of a dietary intervention is another limiting factor that can possibly explain some of the difficulty in detecting compositional changes in the microbiota, with the longest interval being described as 12 weeks, 41,43,44,49 and as short as 1 week in one study.42 Whereas direct effects of diet on the microbiota are likely observable within days, 4 short-term dietary changes lead to immediate but transient shifts which may not necessarily reflect sustained improvements in IBD symptoms.The type and intensity of a dietary intervention can also determine the extent to which the microbiota is altered, as studies have demonstrated that some diets, such as those with substantial modification in fibre content, macronutrient ratio, or 'beneficial foods' [eg, prebiotics, probiotics, or fermented foods], may have more profound effects on certain microbial populations.4,81 Longer interventions, spanning weeks to months, are likely required for observing significant and stable responses of the microbiota to a dietary intervention, and in turn, microbiome-mediated effects on host disease, phenotype, or biomarkers. 82 Conseuently, this study highlights the considerable potential for future research in diet-microbiota interactions in IBD.

Conclusion
This systematic review suggests that current evidence regarding the modulatory effect of a specific dietary pattern, intervention, or constituents, on the intestinal microbiota and disease course of IBD patients is rather limited.The included studies were highly variable and changes in the diversity and abundance of intestinal microbial communities in IBD patients post-diet intervention, were minimal.Dietary patterns including the low FODMAP and low fat diets were associated with improvements in disease severity and QoL, but these were not necessarily associated with microbiota profiles, and therefore conclusions regarding diet-microbiota interactions in IBD are limited.Greater consistency in dietary and microbiota assessment methods and future studies that focus on similar interventions are necessary to identify signatures and predict diet-induced microbiota responses.Understanding precisely how to modulate the microbiota composition, metabolism, and function through a dietary approach could allow the development of personalised IBD dietary recommendations.

Figure 1
Figure1Flow diagram depicting records identified, included, and excluded for review and the reasons for exclusion.
To measure the effect of diet on disease activity, researchers mainly used scores such as the Harvey-Bradshaw [HBI], Partial Mayo Score [PMS], IBD Control Questionnaire, and Simple Clinical Colitis Activity Index [SCCAI], and objective markers such as C-reactive protein [CRP] and faecal calprotectin [FC] [Table alpha or beta diversity in faecal samples bowel symptoms

Figure 2
Figure 2 Schematic representation, summarising significant changes in symptoms or quality of life [blue] and microbiota diversity and relative abundances [purple] found in IBD patients post-dietary intervention [green] trial.IBD, inflammatory bowel disease; FODMAP, fermentable oligosaccharides, disaccharides, monosaccharides, and polyols; QoL, quality of life; sIgA, serum immunoglobulin A; IBS, irritable bowel syndrome.Created with BioRender.com.
Search results were collated and merged into the reference management software EndNote 20 [Thompson Reuters], and de-duplicated before uploading to Covidence systematic review software [Veritas Health Innovation, Melbourne, Australia; https://www.covidence.org].Title and abstract screening were conducted by two reviewers [CN and JP or KD], with conflicts resolved by third reviewer [JP or KD].Full-text articles were retrieved and uploaded into Covidence for full-text review by two reviewers [CN and KD], with conflicts resolved by third reviewer [JP or ECH].

Table 1
The Impact of Dietary Interventions on the Microbiota in Inflammatory Bowel Disease: A Systematic Review 923 Characteristics of randomised trials and cross-sectional studies for inflammatory bowel diseases

First author Country Year Study design Study intervention group Sample size Mean age, y Sex, male, n [%] Disease state Study comparator group Sample size Mean age, y Sex, male, n [%] Primary outcome/ aim Duration/ recruitment period
For this study by Schreiner et al., the groups were taken from the same total cohort of 1313 individuals with IBD. a

Table 2
Summary of findings: dietary characteristics, and outcomes of cross-sectional studies.

author, year Diet type/ ample size Dietary assessment methosd Dietary data time point Food data analysis Dietary outcomes [CD, UC, or HC]
* ↓ fish and shellfish * ↑ intake of eggs * ↑ milk and dairy products [yoghurt] * compared with HC, especially in CD * ↓ dairy products * ↓ cereals * compared with HC, especially in CD ↑ dairy starch, meat, vegetables, fruits, coffee, or tea, sweets, processed meats [cold meat cuts] ↓ Traditional Faroese foods [eg, fermented lamb, dried fish, pilot whale meat, and blubber] CD, Crohn's disease; UC, ulcerative colitis; HC, healthy controls; CNHS2010-F, 2010 Chinese Residents of Nutrition and Health Status Monitoring Semi-Quantitative Food Questionnaire; FFQ, Food Frequency Questionnaire. a For this study, Schreiner et al., the questionnaire assessed dietary habits, not intake.

Table 3
Summary of findings: dietary characteristics and outcomes of intervention studies.

protocol inclusion Diet protocol restrictions Diet rationale Dietary assessment method Dietary data collection time point Food data analysis Dietary outcomes Baseline versus end-of- trial [within group] compared with other diet [between groups]
The Impact of Dietary Interventions on the Microbiota in Inflammatory Bowel Disease: A Systematic Review

Table 3 .
ContinuedThe Impact of Dietary Interventions on the Microbiota in Inflammatory Bowel Disease: A Systematic Review * Table 3. Continued 930 C. Nieva et al. author,

year Diet type/ sample size Diet protocol inclusion Diet protocol restrictions Diet rationale Dietary assessment method Dietary data collection time point Food data analysis Dietary outcomes Baseline versus end-of- trial [within group] compared with other diet [between groups]
Western diet; AMeD, Alternate Mediterranean Diet Score; MUFA, monounsaturated fatty acid; PUFA, polyunsaturated fatty acid; HEI-2015, Healthy Eating Index 2015; MDS, Mediterranean Diet Score; MDSS, Mediterranean Diet Serving Score; MARS, Medication Adherence Rating Scale; IBD-AID, Inflammatory Bowel Disease-Anti-Inflammatory Diet; EEN, exclusive enteral nutrition.

Table 3 .
ContinuedThe of Dietary Interventions on the Microbiota in Inflammatory Bowel Disease: A Systematic Review 931

4 .
Summary of findings: changes in GI symptoms and quality of life following dietary intervention IBS-SSS, Irritable Bowel Syndrome Severity System Scores; GSRS, GI Symptom Rating scale; N/A, not applicable; SIBDQ, Short Inflammatory Bowel Disease Questionnaire; SF-36, Short Form-36 Health Survey; QoL, quality of life; sCDAI, short Crohn's Disease Activity Index; CDAI, Crohn's Disease Activity Index; PROMIS, Patient-Reported Outcomes Management Information System; RAPID-3, Routine Assessment of Patient Index Data 3; BASFI, Bath Ankylosing Spondylitis Functional Index.a Results taken from previously published work using the same cohort of participants.* Indicates statistically significant p-values [p <0.05].The of Dietary Interventions on the Microbiota in Inflammatory Bowel Disease: A Systematic Review 933

Table 5
Summary of findings: changes in clinical disease activity and inflammatory markers following dietary intervention.any of the cytokines measured following the LFD and the IBD-AID, respectively, in UC patients.Fritsch et al.only reported a significant reduction of serum amyloid-A [SAA] concentration in UC patients on an LFD at baseline versus.end of trial, as well as compared with iSAD.

Table 5 .
ContinuedThe of Dietary Interventions on the Microbiota in Inflammatory Bowel Disease: A Systematic Review 935

Table 6
Characteristics of included studies investigating intestinal microbiota profiles of IBD patients and healthy controls.

First author, year Specimen type Tissue site Collection time points Sample collection and storage Extraction method Profiling approach
International Human Microbiome Standards Standard Operating Procedure; NCBI, National Centre for Biotechnology Information; THSTI, Translational Health Sciences and Technology Institute; QIIME, Quantitative Insights Into Microbial Ecology; SGB, Species-level Genome Bins; PSP, Pre-analytical Sample Processing.

Table 6 .
Continued The of Dietary Interventions on the Microbiota in Inflammatory Bowel Disease: A Systematic Review 939 emphasise wheat-based products [eg, bread and pasta], red meat, and dairy, Asian diets typically incorporate rice as dietary staple as well as a greater range of plant-based foods [eg, vegetables, tofu, soy products, and spices].