Enteral Nutrition in Patients with Inflammatory Bowel Disease. Systematic Review, Meta-Analysis, and Meta-Regression

Inflammatory bowel disease (IBD) is a chronic disease mediated by the immune system and is characterized by inflammation of the gastrointestinal tract. One of the possible treatments for this pathology is a change in the type of diet, of which enteral nutrition (EN) is one. This study is to understand how the use of EN can affect the adult population diagnosed with IBD. We conducted a systematic review, meta-analysis, and a meta-regression. On the different databases (MEDLINE, Scopus, Cochrane, LILACS, CINAHL, WOS), we found 363 registers with an accuracy of 12% (44 registers). After a full-text review, only 30 research studies were selected for qualitative synthesis and 11 for meta-analysis and meta-regression. The variables used were Crohn’s disease activity index (CDAI), C-reactive protein (CRP), and erythrocyte sedimentation rate (ESR). EN has been shown to have efficacy for the treatment of Crohn’s disease and is compatible with other medicines. As for the CDAI or rates of remission, there were no differences between enteral and parenteral nutrition. Polymeric formulas have shown better results with respect to the CRP. The long-term treatment could dilute the good CDAI results that are obtained at the start of the EN treatment.


Introduction
Inflammatory bowel disease (IBD) is a chronic disease mediated by the immune system and characterized by the inflammation of the gastrointestinal tract. IBD includes Crohn's disease (CD) as well as ulcerative colitis (UC) [1]. UC affects the large intestine and is generally observed as a superficial ulcer due to an inflammatory reaction localized to the mucosa and the submucosa. However, CD occurs all along the intestinal tract (from mouth to anus) and involves the entire intestinal layer [2].
The prevalence and incidence of IBD has increased worldwide and is increasingly diagnosed in young individuals [3]. As it is a chronic, incurable, and low-mortality disease, it is expected that the decrease of the global burden of the disease in the next decade will require a two-pronged solution that implies research on prevention interventions as well as innovations in the care of these patients [3,4].
The etiology of IBD is still greatly unknown, and recent evidence indicates that the genetic susceptibility of the individual, the environment, the intestinal microbial flora, and the immune responses are all factors that are involved and functionally integrated in the pathogenesis of IBD [5].

Materials and Methods
To achieve this objective, a systematic review was conducted in agreement with the procedures and verification list described by PRISMA [14]. Afterward, a meta-analysis on the more common results and a meta-regression with the co-variables, type of enteral nutrition, and period of treatment, were conducted.

Systematic Review
A search of scientific works was conducted in the MEDLINE database through the system of open retrieval system on the Internet such as PubMed, Cochrane, Scopus, Web of Science, CINAHL, and LILACS. Studies conducted over time up to 5 January 2019 were compiled.

Inclusion and Exclusion Criteria
The studies selected had to comply with the following inclusion criteria: refer to an adult population (older than 18) diagnosed with some type of IBD; study the effect of enteral nutrition within IBD; were clinical trials; and in the English, Spanish, Portuguese, French, or German languages.
The following articles were excluded: those that referred to the infant population; to animals, to the use of EN in a healthy adult population; those that sought the effect of oral exclusion diets on IBD; observational studies; and those based on secondary sources.

Search Equation
To include content linked to the intervention EN, a specific descriptor was used (MeSH) such as "Enteral Nutrition", and the term "Enteral Nutrition" in the title or abstract.
For the content linked to the population, we utilized the descriptor that referred to the disease "Inflammatory bowel diseases", and its equivalent term in the title or abstract.
Additionally, the filters "Humans", "Adult", and "Clinical Trial" were utilized to achieve our objective. Therefore, the main search equation designed for this study was: (("Inflammatory Bowel Diseases" [Mesh]

OR "Inflammatory Bowel Diseases"[Title/Abstract]) AND ("Enteral Nutrition"[Mesh] OR "Enteral Nutrition"[Title/Abstract])) AND (Clinical Trial[ptyp] AND Humans[Mesh] AND adult[MeSH])
The search equation was adapted to each and all of the databases described previously. The process was conducted between the months of May and June, 2019.

Selection Process
After eliminating the duplicate records, the process of selection was conducted in two phases. The first consisted of reviewing the titles and abstracts of all the article records that resulted from the adapted search equations and were shown by the databases by using the inclusion and exclusion criteria and the objective of the study as the screening measure. The screening and selection of the records/articles were conducted independently by the two researchers, both experts in the fields of nutrition. These researchers agreed on the discrepancies found in order to define the final suitability of the records/articles found in the databases. The precision of the search was calculated, based on the ratio of the full-text articles selected for the review, divided by the number of records found by the search equation, and multiplied by one hundred.
The second phase was conducted by applying the inclusion/exclusion criteria to the complete texts of all of the scientific studies selected in the first phase, thus ensuring the relevance of each one. In order to obtain studies that were not accessible via the Internet, we used three methods: Researchgate, the correspondence author, and interlibrary loan. Only three were recovered through interlibrary loan.

Evaluation of the Quality of the Studies
The evaluation of the methodological quality of the included studies was performed by two independent researchers by using the Cochrane risk of bias tool (RoB 2) for clinical trials [15]. This tool is structured into five domains through which bias might be introduced into the result: the randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported results. For each study, the response options for an overall risk-of-bias judgement were "low risk of bias", "some concerns", and "high risk of bias".

Meta-Analysis and Meta-Regression
To calculate the effect size of the enteral nutrition on the variables: the Crohn's disease activity index (CDAI), C-reactive protein (CRP), and erythrocyte sedimentation rate (ESR), a meta-analysis, were performed. For this, the model of fixed effects and the model of random effects were utilized. The results were presented as a forest-plot, along with the percent heterogeneity and its confidence interval at 95%, the T value, and the heterogeneity test.
To explore the influence of each study over the effect size, we used a leave-one-out method; pooled estimates were calculated by omitting one study at a time. In addition, we plotted a scatter plot introduced by Baujat et al. [16] On the x-axis, the contribution of each study to the overall heterogeneity statistic was plotted. On the y-axis, the standardized difference of the overall treatment effect with and without each study was plotted; this quantity describes the influence of each study on the overall treatment effect. Therefore, studies that fell in the top right quadrant of the Baujat plot had the most influence.
Publication bias occurs when only favorable results are published, and this could have consequences on the results of the meta-analyses if these were included. To analyze the publication bias, a non-parametric analysis was conducted as proposed by Duval and Tweedie [17], based on the funnel-plot, estimating and adjusting for the number and outcomes of missing studies in the meta-analysis. Another less-conservative proposal to estimate the number and outcomes of missing studies was proposed by Copas et al. [18].
The meta-regression was utilized to understand if the type of enteral nutrition (polymeric or elemental), age (years), or the duration of the intervention (days) modified the effect size of the resulting variables CDAI, CRP, and ESR as a function of the type of nutrition. All the calculations were conducted within an R programming environment by utilizing the packages meta version 4.9-6 [19] and metasens version 0.4-0 [20].

Systematic Review
As a result of the specific search equations used on the different databases, a total of 438 records of scientific articles were found. A total of 75 records were duplicated, leaving a total of 363 records without duplication. In the first phase of the study, exactly 319 study records were discarded, leaving 44 full-text studies to review, so the accuracy was 12%. The reasons for not including them were that 131 records showed that the study utilized a design that was not adequate, 100 did not use an adult population, 50 did not study the effect of EN, three were written in another language other than the ones cited above, (two in Japanese and one in Chinese), 12 did not refer to humans, six did not refer to the IBD, and 17 were conducted without showing results ( Figure 1). As for the designs of the studies included, 16 controlled and randomized clinical studies (53.3%), nine non-randomized, controlled clinical trials (30%), and five non-randomized, non-controlled clinical trials (16.7%) were found. In addition, 28 of the studies found showed results that specifically referred to CD, and two studies had results on UC and CD, under the category of IBD. Additionally, 23 studies mentioned results of the disease in its active form, four studies in the shape of remission, and the rest did not indicate any. Figure 2 shows this information in a chronological manner. In the second phase, seven studies were not utilized as they could not be obtained in electronic format, not even after using the methods explained in the methodology. In addition, seven trials were removed, one for being written in Turkish, another due to defects in its design, another for not studying the effects of EN, and four because the population studied was not diagnosed with IBD. Therefore, only 30 research studies [10, were selected, as shown in Figure 1.
As for the designs of the studies included, 16 controlled and randomized clinical studies (53.3%), nine non-randomized, controlled clinical trials (30%), and five non-randomized, non-controlled clinical trials (16.7%) were found. In addition, 28 of the studies found showed results that specifically referred to CD, and two studies had results on UC and CD, under the category of IBD. Additionally, 23 studies mentioned results of the disease in its active form, four studies in the shape of remission, and the rest did not indicate any. Figure 2 shows this information in a chronological manner. As for the variety of the types of formulas employed, 19 studies utilized an elemental formula, 11 studies utilized a polymeric formula, two studies a semi-elemental formula, and three studies a type of parenteral nutrition (PN). Likewise, it should be mentioned that various types of formulas were often used in the same study. Thus, the following commercial formulas were employed: "E028", "Novasource", "Peptisorb", "Elental", "E028 Extra", "Vivonex-TEN", "Peptison", "Peptamen", "Vivonex HN", "Realmentyl", "Triosrbon", "Vital", "Vivonex", "Fortison", "Precision-Isotonic", "Uniasa", "Guarantee Plus", and "liquid Pepti-2000 LF".
In addition, a total of six types of objectives were found: 10 studies sought to compare two different types of EN, among which five of the works compared an elemental formula with a polymeric one, two compared an elemental formula with another elemental one that contained a greater concentration of fats, one work compared two types of polymeric formulas, one work compared two types of elemental formulas, and one work compared an elemental formula with a semi-elemental one.
Moreover, seven studies compared a type of EN with an oral diet, five studies sought to experiment with a type of EN, three studies compared a type of EN with a type of PN, three studies sought to compare a type of EN with another type of medication plus an oral diet, and finally, two studies sought to compare a partial EN plus a diet with an oral diet.
As for the manner of administration of the EN, 15 research studies employed a nasogastric tube, three studies utilized a nasoduodenal catheter, two studies used a nasointestinal catheter, seven studies administered the formula orally, and three did not specify the manner of administration.
The total population analyzed in the research studies found included a total of 1070 individuals with IBD, with 1016 diagnosed with CD and 25 with UC.
In addition, a total of six types of objectives were found: 10 studies sought to compare two different types of EN, among which five of the works compared an elemental formula with a polymeric one, two compared an elemental formula with another elemental one that contained a greater concentration of fats, one work compared two types of polymeric formulas, one work compared two types of elemental formulas, and one work compared an elemental formula with a semi-elemental one.
Moreover, seven studies compared a type of EN with an oral diet, five studies sought to experiment with a type of EN, three studies compared a type of EN with a type of PN, three studies sought to compare a type of EN with another type of medication plus an oral diet, and finally, two studies sought to compare a partial EN plus a diet with an oral diet.
As for the manner of administration of the EN, 15 research studies employed a nasogastric tube, three studies utilized a nasoduodenal catheter, two studies used a nasointestinal catheter, seven studies administered the formula orally, and three did not specify the manner of administration.
The total population analyzed in the research studies found included a total of 1070 individuals with IBD, with 1016 diagnosed with CD and 25 with UC.
The main tools utilized by the researchers to obtain results were scores, biomarkers, and tests to measure the activity of the disease: "Harvey-Bradshaw Index" (HBI), the CDAI, the Van Hees activity index (VHAI), the qualification in the classification of the International Organization of Inflammatory Bowel Disease (IOIBD), the Subjective Global Assessment (SGA), the Truelove and Witts index, the simple clinical index, the Crohn's disease activity score (CDAS); biomarkers such as CRP, ESR, the white blood cell count (WBC), levels of albumin, pre-albumin, transferrin, hemoglobin, platelet count, total bilirubin, alkaline phosphatase, etc.; and medical tests such as ileocolonoscopy. Specific quality of life questionnaires such as the "Inflammatory Bowel Disease Questionnaire" (IBDQ) were also used as were complementary tests such as urine and feces samples and tests to measure the body's composition such as anthropometries and bioimpedence.   Although at the beginning of the study the proportion of patients with a severe UC in the EG was higher than in the CG (8/9, and 7/16 respectively), there were no significant differences at the end of the study (p > 0.05). In both groups the improvements in disease activity of patients with UC were significant, but non-significant positive changes were observed in the clinical findings during the hospitalization period. Significant improvements of the SGA in both groups.  The intention-to-treat analysis showed that the remission rates were 20%, 52% and 79% for PEN1, PEN2 and ESTG (p = 0.001). Withdrawal from treatment, remission rates were 27%, 63% and 79%, respectively (p = 0.008).

IOIBD, ESR and CRP
The cumulative rates of continuous remission after 1, 2 and 4 years were in the ENG 94%, 63% and 63%; in the ENG + D 75%, 66% and 66% in the DG 63%; 42% and 0%, and in the CG 50%, 33%. and 0%. The ENG had a higher rate than DG (p < 0.05) and CG (p < 0.01). The ENG + D had a higher rate than the CG (p < 0.05). Patients who received more than 30 kcal of EN showed a higher continuous remission rate (p < 0.001).  ACT CD G1/G2 14/14 120 GB G1: 2 months ordinary diet followed by 2 months supplementation with the non-elementary low-waste formula "Guarantee Plus". G2: same intervention than G1 with invested order.
Nutritional measurements (height, weight, MAMC and thickness of the skin fold), biochemical measurements (serum prealbumin, serum, red cell folate, creatinine height index, platelets, T lymphocytes, etc.) and urine tests parameters.
The general effect of EN during the 2 months was to increase serum albumin, serum protein and prealbumin levels, creatinine height index and T-lymphocyte count. With EN decreased levels of orosomucoids and serum alkaline phosphatase and its activity (p < 0.05) Patients felt better when they received EN, although their monthly symptom scores showed no significant benefit.

Meta-Analysis and Meta-Regression
Only 11 clinical trials had common quality and variables needed to be used in the meta-analysis. These 11 trials worked with a total of 15 groups. The final size of the sample was comprised of 272 individuals, all with CD, to which an EN treatment had been given. The common variables were the CDAI, the CRP, and the ESR, and the co-variables type of nutrition, age, and duration of the intervention. Figure 5 shows the effect size of the use of EN. For the three indicators of disease, the effects were positive when comparing the situation at the start and finish of the treatment with EN independently, if the situation with fixed effects (less probable) or random effects (more acceptable) is considered.

Meta-Analysis and Meta-Regression
Only 11 clinical trials had common quality and variables needed to be used in the meta-analysis. These 11 trials worked with a total of 15 groups. The final size of the sample was comprised of 272 individuals, all with CD, to which an EN treatment had been given. The common variables were the CDAI, the CRP, and the ESR, and the co-variables type of nutrition, age, and duration of the intervention. Figure 5 shows the effect size of the use of EN. For the three indicators of disease, the effects were positive when comparing the situation at the start and finish of the treatment with EN independently, if the situation with fixed effects (less probable) or random effects (more acceptable) is considered.

Meta-Analysis and Meta-Regression
Only 11 clinical trials had common quality and variables needed to be used in the meta-analysis. These 11 trials worked with a total of 15 groups. The final size of the sample was comprised of 272 individuals, all with CD, to which an EN treatment had been given. The common variables were the CDAI, the CRP, and the ESR, and the co-variables type of nutrition, age, and duration of the intervention. Figure 5 shows the effect size of the use of EN. For the three indicators of disease, the effects were positive when comparing the situation at the start and finish of the treatment with EN independently, if the situation with fixed effects (less probable) or random effects (more acceptable) is considered. The influence of each study on the results of the meta-analysis are shown on Table 2, considering a model of random effects. Figure 6 shows this influence through the Baujat plot. The numbers shown in the figure correspond to the articles shown in the table in the ID column. Table 2. Influence analysis in meta-analysis using leave-one-out method (random effect).  The influence of each study on the results of the meta-analysis are shown on Table 2, considering a model of random effects. Figure 6 shows this influence through the Baujat plot. The numbers shown in the figure correspond to the articles shown in the table in the ID column. Table 2. Influence analysis in meta-analysis using leave-one-out method (random effect). The results show that the articles did not influence the results in the case of the CDAI and the ESR, however, study 6 (D. Royall et al. 1994 with Elemental Nutrition) and to a lesser degree, study 12 (Yun Feng et al. 2013 with Polymeric Nutrition) may compromise the results of the meta-analysis for the CRP. However, the heterogeneity, omitting these works, was 95.1% and 96.6% when compared to the overall 97.2%, therefore, a great influence of the CRP on the meta-analysis could not be determined.

Meta-Analysis for
A funnel plot represents the effects observed in the different studies (x-axis), and the standard error (y-axis). In the absence of heterogeneity and publication bias, the dots shown in the funnel plot should jointly adopt the aspect of a funnel, with the wider part corresponding to the smaller and more precise studies. A lack of symmetry could be due to this publication bias. The funnel plot is shown in Figure 7, where a lack of symmetry can be observed. Therefore, the non-parametric analysis proposed by Duval and Tweedie to analyze this asymmetry should show a lack of articles, and therefore a publication bias.
The results of this non-parametric analysis for the fixed-effects model and the random-effects model are shown in Table 3. These results show a possible publication bias in the three variables studied, if a fixed-effects model is assumed; however, the random-effects models did not show this bias. The results show that the articles did not influence the results in the case of the CDAI and the ESR, however, study 6 (D. Royall et al. 1994 with Elemental Nutrition) and to a lesser degree, study 12 (Yun Feng et al. 2013 with Polymeric Nutrition) may compromise the results of the meta-analysis for the CRP. However, the heterogeneity, omitting these works, was 95.1% and 96.6% when compared to the overall 97.2%, therefore, a great influence of the CRP on the meta-analysis could not be determined.
A funnel plot represents the effects observed in the different studies (x-axis), and the standard error (y-axis). In the absence of heterogeneity and publication bias, the dots shown in the funnel plot should jointly adopt the aspect of a funnel, with the wider part corresponding to the smaller and more precise studies. A lack of symmetry could be due to this publication bias. The funnel plot is shown in Figure 7, where a lack of symmetry can be observed. Therefore, the non-parametric analysis proposed by Duval and Tweedie to analyze this asymmetry should show a lack of articles, and therefore a publication bias. The results of this non-parametric analysis for the fixed-effects model and the random-effects model are shown in Table 3. These results show a possible publication bias in the three variables studied, if a fixed-effects model is assumed; however, the random-effects models did not show this bias.   The results show that the articles did not influence the results in the case of the CDAI and the ESR, however, study 6 (D. Royall et al. 1994 with Elemental Nutrition) and to a lesser degree, study 12 (Yun Feng et al. 2013 with Polymeric Nutrition) may compromise the results of the meta-analysis for the CRP. However, the heterogeneity, omitting these works, was 95.1% and 96.6% when compared to the overall 97.2%, therefore, a great influence of the CRP on the meta-analysis could not be determined.
A funnel plot represents the effects observed in the different studies (x-axis), and the standard error (y-axis). In the absence of heterogeneity and publication bias, the dots shown in the funnel plot should jointly adopt the aspect of a funnel, with the wider part corresponding to the smaller and more precise studies. A lack of symmetry could be due to this publication bias. The funnel plot is shown in Figure 7, where a lack of symmetry can be observed. Therefore, the non-parametric analysis proposed by Duval and Tweedie to analyze this asymmetry should show a lack of articles, and therefore a publication bias. The results of this non-parametric analysis for the fixed-effects model and the random-effects model are shown in Table 3. These results show a possible publication bias in the three variables studied, if a fixed-effects model is assumed; however, the random-effects models did not show this bias.   With respect to the meta-regression, the results are shown in Table 4. There was a dependence of the CDAI score with the period, losing efficacy in prolonged interventions (p < 0.05). The CRP showed better results in the EN when using polymeric formulas that were elemental (p < 0.001).

Discussion
Our results included 30 studies (1070 participants). All trials included had a broad scope and had a very varied methodological and clinical heterogeneity. The variables collected were very diverse, with CDAI, CRP, and ESR being the most common. The sample sizes of the studies included were generally small (n < 30), thus, a meta-analysis was needed in order to arrive at better conclusions.
A cure for IBD is not known, however, there is evidence of remission and improvement of the symptoms with EEN, which implies the exclusive consumption of an elemental or polymeric substance for many weeks [50], as shown by many of our results. Despite the lack of correlation between IBDQ and the CDAI, correlations were observed between both indexes starting at week 4 of the treatment. A study that focused on the gall bladder was even found, which showed its improvement after day 36 of treatment administration; therefore, aside from reducing the activity or inducing the remission of the disease, this diet could have beneficial effects on organs related with the digestive system [23,25].
The EN formulas tended to contain macronutrients such as amino acids or simple carbohydrates, along with micronutrients such as vitamins. The proteins, carbohydrates, and fats do not reach the ilium or the colon as they are absorbed in the duodenum and jejunum. As for the amino acids they contain, they were named as elemental formulas if they contained free amino acids, semi-elemental if they contained peptides, and polymeric if they contained whole proteins [51]. Different formulations exist, but the ones that do seem to have a positive effect on the maintenance and remission of the disease are elemental and polymeric diets [10,22]. The efficacy of an EN diet does not depend much on it being elemental or polymeric, as shown by some of our results [42,44], since, a priori, both have the same potential for inducing a remission [32,39,46]. However, the meta-regression conducted indicated that a polymeric diet could decrease the CRP better than an elemental one. Additionally, a distinction could be made between them when looking at the economic burden entailed by the use of one or the other and the acceptability by the patients, meaning that, in the adherence to the dietary treatment, the polymeric ones tend to be more accepted by the patients, as they are better tasting [52,53]. The elemental foods are less tolerated with mouth feeding, and generally require a nasogastric tube, which entails complications and patient discomfort. In contrast, the polymeric EN is more tolerable through the mouth by patients, making it the first option for the ill [33,54].
As for the formulation of the EN, studies have also been conducted on the benefits or not of an EN diet rich in fats as opposed to an elemental EN. Just as in other studies, the results of the clinical trials are controversial. Some studies have demonstrated the beneficial effect of the enteral formula rich in fats [55], while others did not show any effect [56] or less beneficial effects [29]. Despite what has been said, some studies have suggested that an EN high in fats could improve gastrointestinal motility and improve the ilium after an operation [57], reducing damage to the intestinal mucosa barrier and the underlying mechanism that could be associated with its antioxidant action after surgical intervention [58].
EEN, combined with some types of medication such as antibiotics [59], seem to improve the disease's symptoms. Just as shown by our results, EN combined with other types of pharmaceuticals such as prednisone, corticosteroids, and sulfasalazines show a significantly continuous high rate of remission [41,60]. On the other hand, the combination of EN with steroids does not seem to have significant differences in the probability of a relapse [40], perhaps because the steroids do not address the damage produced in the intestinal mucosa, which is the greatest predictor of complications over time [61,62].
Although the mechanism that nurtures the healing of the mucosa by the EN has not been completely determined as of yet, it has been shown that a polymeric formula was as effective as the Infliximab inhibitor of tumor necrosis factor (TNF)-α, and is higher than the hydrocortisone in the maintenance of the function of the intestinal barrier [63]. This is perhaps the reason why significant differences were not found in a study conducted by Gasull et al. in 2001 that utilized two polymeric EN formulas, with the response being similar in both [30]. Additionally, in another study conducted with two polymeric formulas for five weeks, a significant relationship was not found between the treatment with different EN and the changes produced at the level of the disease's activity [37].
As for the use of elemental and semi-elemental EN, the results were very similar. For example, Mansuf et al. achieved the clinical remission of 16 patients in four weeks with both formulas, and the reduction of the CRP was significant in both groups [36]. The mechanism of action of the semi-elemental diet could be multifunctional, just like as the elemental one, decreasing the intestinal permeability and thus decreasing the loss of fluid. The semi-elemental diet could also reduce the commensal intestinal bacteria that play a role in intestinal inflammation [64,65]. Thus, the use of these types of diets is advisable, either with the use of an elemental or semi-elemental formula for the management of different gastro-intestinal disorders [66].
In 2013, Yun Feng et al. [24] found significant differences between groups subjected to EEN and EN plus an oral diet, although it is interesting to note how the patients refer to a greater subjective well-being when they take EN together with the oral diet when compared to those who are not treated with supplemented EN, despite the biochemical parameters being very similar [21,49]. Some studies suggest that a partial enteral nutrition supplemented with different diets such as elimination, anti-inflammatory, auto-immune diets, or diets low in FODMAP (Fermentable Oligo-, Di-, Mono-saccharides and Polyols) could be beneficial for UC and CD [25,32], although larger controlled assays are needed to back their use [67]. Even patients who were subjected to EN before their operation experienced benefits, not only in their nutritional state, but also with a reduction of inflammation in their disease [68], with patients also experiencing improvements after said intervention [27].
Historically, EN was used and is used as a complementary nutritional treatment for patients with complicated IBD that leads to worrying malnutrition, thus improving their nutritional state [53]. However, the meta-regression from our study showed an inverse relationship between the period of treatment with EN and the improvement shown through the CDAI, that foreseeably, the patients who are subjected to prolonged EN could stop noticing its benefits.
The EN was utilized as an induction therapy for active IBD [53], but it is important to know which EN formulas can be used to boost their anti-inflammatory effects, as there is evidence that supplementary EN is not sufficient for inducing remission, so that it would have to be used exclusively to be able to obtain its anti-inflammatory effect [69,70]. At present, it is well-established that EEN has a strong anti-inflammatory effect with a reduction in the systemic and mucosa inflammatory parameters in a few days, however, the EEN as a long-term therapy is still a challenge, given its lack of palatability and the lack of data to analyze the efficiency of EEN as a maintenance diet [71].
Diverse studies have shown that clinical remission and healing of the mucosa is possible through different nutritional regimes [72]. As for the debate about which is healthier, EN or PN, our results showed that there are studies in which the effectiveness of both seems to be significantly the same for the improvement of the CDAI [45,48]. Bearing in mind that the dietary antigens could be important stimulants for the immune system of the mucosa, intestinal rest with total parenteral nutrition (TPN) is considered as the main option for achieving this rest and for correcting possible nutritional deficits [73]; however when compared to the EN, it does not seem to provide greater benefits. In fact, in one of the studies included in our review [47], EN was the one that seemed to provide the greatest benefits to the patients and to reduce the costs, personal as well as economic, of the different dietary treatments [74].
According to European guidelines, the acceptability and the obligatory compliance of the EN are the greatest obstacles found by different researchers when dealing with EN studies. There are clear differences between the studies shown in terms of healing of the mucosa, and therefore the remission of the activity of the disease, which makes them difficult to compare. What is known, however, is that the EEN is a real alternative to immunosuppressive therapy, which exerts its main therapeutic effect on the microbiota, thus reducing intestinal permeability, enhancing barrier defense, and promoting a reduction of pro-inflammatory cytokines [75,76].
This study is not exempt of limitations. With respect to the systematic review, seven articles were not recovered, so our results could be altered. However, the small variability observed through the meta-analysis implies that these articles could substantially vary the results obtained, and in fact, the sensitivity study that analyzed the publication bias showed little alterations on the effect size as well as its confidence interval. Although the quality of these studies has not been introduced in the meta-regression, all the articles that were utilized with this technique were considered to have sufficient quality, so we do not believe that the quality could introduce bias in the findings.

Conclusions
EN has been shown to have efficacy for the treatment of CD and is compatible with other medicines. As for the CDAI or the rates of remission, there were no differences between EN and PN. Polymeric formulas, when compared to elemental ones, have shown better results with respect to the CRP. The long-term treatment could dilute the good CDAI results that are obtained at the start of EN treatment.