Antioxidant Supplements and Gastrointestinal Diseases: A Critical Appraisal

The problems in the gastrointestinal (GI) tract can lead to many diseases, but in this review only irritable bowel syndrome (IBS) and inflammatory bowel disease (IBD), which includes Crohn's disease (CD) and ulcerative colitis (UC), are considered. IBS affects 10-25% of the population globally with a slightly higher prevalence in females compared to males [1]. CD inflicts individuals in the prime of life between 10 and 30 years of age [2]. In developed countries, the prevalence of CD exceeds 4 million [3] and, surprisingly, the prevalence of CD has increased following the improvement in human hygiene and industrialization [3]. The worldwide prevalence of UC varies between 0.5 and 24.5 per 100,000 [3]. While CD is more common among smokers, UC is more common in ex-smokers and nonsmokers [4]. For these diseases, a variety of therapies are used to attain remission of the symptoms and/or avoid relapse once the patient is in remission. The current understanding of the pathology suggests that oxidative stress may contribute to these diseases [5]. This review contains a critical appraisal of the efficacies of these supplements based on the available literature on human trials. No studies were found on antioxidant supplements and celiac disease and hence it was not discussed here. This review covers the structure and function of the GI tract, diagnosis, pathophysiology, and current treatment of the diseases followed by a rationale for the use of antioxidant supplements. The efficacies of different supplements are discussed, leading to a synopsis, recommendations, and a critical appraisal. This not a meta-analysis type of study, but instead is based on opinions of the authors because the vast variety of antioxidant supplements were used for different GI diseases at different stages and examined by different methods.

The GI tract is a tube with mucosa lining its lumen, followed by a set of layers of submucosa, and smooth muscle with a rich supply of nerves and blood vessels. The mucosa enables the GI tract to perform its main functions: digestion and absorption of nutrients from the diet, protection of the body against physical and chemical damage from luminal contents, and provision of immunity against them [5,6]. The lining is made up of a single layer of epithelial cells. The small intestinal epithelium is extensively folded to provide a very large absorptive surface area, resulting in distinct villus and crypt regions. The mucosa contains several types of cells: enterocytes, Paneth, goblet, and enteroendocrine [5,6,7,8]. A large proportion (80%) of the mucosal cells are enterocytes whose main role is the absorption of nutrients and unconjugated bile salts. They may also be involved in chemical processing of food and cooperate with other cells in the induction of immunological tolerance to the ingested proteins [5,6]. Paneth cells synthesize and secrete substantial quantities of antimicrobial peptides which are key mediators of host-microbe interactions, including homeostatic balance with colonizing microbiota and innate immune protection from enteric pathogens [5,8]. They also secrete factors that help sustain and modulate the epithelial stem and progenitor cells. Goblet cells secrete mucin 2 polymers which form mucus that plays a key role in mucosal immunology [7]. Based on their location, there are several types of goblet cells. In the colon, the surface goblet cells secrete continuously to maintain the inner mucus layer, whereas goblet cells of the colonic and small intestinal crypts secrete upon stimulation such as after endocytosis or in response to acetylcholine [7]. Thus, the mucus system itself and the regulation of its secretion differ between the small and the large intestine [5,7]. Enteroendocrine cells, distributed throughout the GI tract, form a large endocrine organ to aid in the control of GI secretion and motility, the regulation of food intake, postprandial glucose levels, and metabolism. They also communicate directly with neurons innervating the GI tract and thus form part of the gut “brain axis.”Beneath the endothelium is the lamina propria, which supports the mucosal epithelium. It consists ofthe extracellular matrix, blood and lymphatic vessels, neuronal and smooth muscle cells, and various immune cells (lymphocytes, macrophages) [5,7,8].

The mucosa is constantly being renewed as the stem cells in the crypt divide into progenitor cells, which can differentiate into secretory (Paneth, mucin-secreted goblet, and enteroendocrine cells) or absorptive (enterocyte) lineages [9]. The microenvironment surrounding the progenitor cells may determine the resulting cell types and their fate. The differentiated cells undergo apoptosis and are shed into the lumen in about 3 days. Damage to these structures, cross-talk between them, or imbalance in individual cell types diminishes barrier functions and sensitizes the GI tract with dietary antigens and GI normal microbiota, leading to various GI diseases [9].

IBD is a multifactorial condition resulting from an interaction between genes and environmental factors [10]. Based on genome-wide association studies, there are 163 loci for IBD of which 110 are shared between CD and UC [11]. Single-nucleotide polymorphisms in the nucleotide-binding oligomerization domain gene (NOD2) may play an important role in CD. The NOD2 gene regulates barrier function, autophagy, and hyposensitization of the gut to normal microflora and health of Paneth cells. Three alleles of NOD2 (Arg702Trp, Gly908Arg, and 980fs981) have been studied extensively [10]. A large segment of CD patients in Western countries carry at least one mutated NOD2 allele. Persons with 2 of the mutated alleles have a 20- to 40-fold higher risk of developing CD. Thus, mutations in NOD2 gene are strong genetic risk factors for CD. The NOD2 binds muramyl dipeptide, which is a component of bacterial cell wall peptidoglycan [10], thereby leading to an upregulation of NF-κB expression. NOD2 may directly influence the composition of the gut microbiota by regulating the production of a subgroup of intestinal antimicrobial peptides known as cryptdins, which are produced by Paneth cells in the intestinal crypts. The patients with CD may also show decreased populations of Faecalibacterium prausnitzii and Roseburia species and increase in bacteria such as Escherichia coli [11].

Unlike CD, UC is confined to the large bowel with diffuse inflammation of the colonic and rectal mucosa. It is associated with colorectal cancer risk, which increases with the duration of symptoms [12]. In a study on Indian patients with UC, the risk of colorectal cancer was 900 times higher than the general population [12]. The basic mechanisms of UC may be similar to those of CD: increased membrane permeability, immune exaggeration, and decreased autophagy [13]. The hepatocyte nuclear factor 4-alpha (HNF4A) gene has been associated with UC [13]. HNF4A may regulate epithelial permeability since it controls adherens junctions, tight junctions, and desmosomes, which play an important role in cell-cell junctions. In UC, the basement membrane may also be compromised by mutations in the E-cadherin encoding gene CDH1 and the laminin β1 subunit encoding gene LAMB1 [13]. Another defense is through the production of anti-inflammatory cytokines such as IL-10 and TG-Fβ, which activate TH17 (Treg cells) which in turn suppresses antigen-presenting cells and dendritic cells to secrete IL-17 [13]. The solute carrier family 9 member (SLC9A) 3 gene, encoding an epithelial sodium/hydrogen exchanger, has also been implicated in UC, but not in CD [13]. It was hypothesized that there is a considerable overlap between IBS and UC [5]. Compared to healthy controls, the mucosal biopsy samples obtained from the colon, rectum, and terminal ileum of IBS patients show more inflammation [5]. There may also be similarities between IBS and UC in the alterations in the serotonergic and cytokine-related pathways. Yet, IBS may lack the expression of several proteins which are markers of UC [5].

There is a symbiotic relationship between the microbiota and the gut: the human host contributes the nutrients needed for the survival of the microbes, which in turn aid the host in nutrition (e.g., converting indigestible dietary fibers into short-chain fatty acids butyrate, propionate, and acetate), protection against pathogens, and regulation of the immune responses [10,14]. A disruption in the ecological balance of the GI microbiota can lead to GI diseases or be caused by them. In IBD there is a decrease in the population of several protective bacteria and an increase in the population of harmful bacteria [14]. Food poisoning may be viral, but is mostly due to bacterial contamination in food and water. A large portion of the population also has Helicobacter pylori infection, which can lead to secondary diseases [10,14,15]. The infecting bacteria, or the antibiotics used to eliminate them, may also alter the balance of the indigenous microbiota. Often these problems are overcome on their own with time, but sometimes probiotic supplementation may be needed [10,14,15,16,17,18].

IBS is a chronic condition of the large bowel that is characterized by abdominal pain, cramping, bloating, gas, diarrhea, and/or constipation [1]. Patients may be categorized as diarrheal-predominant, constipation-predominant, and as having both. Unlike IBD, there are no macroscopic or microscopic changes in the bowel tissue structure, and therefore IBS is also called a “functional disorder of gastrointestinal tract” [1,19]. It is diagnosed by exclusion criteria for cancer, IBD, and infections.

IBD may be divided into CD and UC, which share several common symptoms except that the affected area in UC is the colon but in CD it may be the colon or ileum [20]. CD symptoms may include diarrhea, fever, fatigue, abdominal pain and cramping, blood in the stool (bright red blood in toilet bowl or darker blood mixed with stools), mouth sores, reduced appetite, weight loss, and anal fistulas [21]. For UC the part of the colon being inflamed is considered and the general symptoms are: diarrhea (often with blood or pus), abdominal pain and cramping, rectal pain, blood in the stool (bright red blood in toilet bowl or darker blood mixed with stools), urgency and inability to defecate, weight loss, fatigue, and fever [22]. There is no single test to diagnose CD or UC. The patient may undergo all or some of the following tests: blood tests for anemia or infection, fecal occult blood test, colonoscopy (presence of clusters of inflammatory cells called granulomas point to CD), flexible sigmoidoscopy, computerized tomography, magnetic resonance imaging, capsule endoscopy, double-balloon endoscopy, and small bowel imaging. Diagnosis of either disease is made by ruling out other factors.Several indices and scoring systems of GI diseases have been used by investigators to determine the efficacies of supplements as indicated in Table 1.

CD and UC are chronic GI inflammatory diseases with periods of relapse and remission, and the treatments aim to achieve and maintain remission. The current treatments include a combination of anti-inflammatory (steroids), immunosuppressive agents (e.g., azathioprine, sulfasalazine, mesalamine, 6-mercaptopurine, and methotrexate) and anti-tissue necrosis factor-α (anti-TNF-α) even though some treatments may have serious adverse effects [23,24]. For CD, a phase II clinical trial reported that 21 oligomeric antisense specific for SMAD7, which suppresses TGF-β1, achieved 65% remission as compared to 15% with placebo [25,26]. The compound is ready for the phase III clinical trial. Probiotics and prebiotics may also be included in the treatment so as to normalize the GI microfloral ecosystem [14,15,16,17,18].

IBS treatment depends on the most bothersome symptoms of the patient. The treatment may include the use of antispasmodics, laxatives, dopamine antagonists, 5-hydroxy tryptamine-3 antagonists and/or 5-hydroxy tryptamine-4 agonists, sedatives, antibiotics, probiotics, modifications in diet and lifestyle, and complementary and alternative therapies [19]. For C-IBS, linaclotide and lubiprostone may be used and for D-IBS, cilansetron or ramosetron may be given. Rifaximin has emerged as a strong option for the treatment of IBS because of its broad-spectrum bactericidal activity in vitro, and its favorable tolerability profile [19]. Nitric oxide (NO) donors, opioid receptor agonists, and ketotifen have also been proposed for IBS treatment [19].

Normal physiological processes result in the generation of reactive oxygen species (ROS) such as peroxide, superoxide, hydroxyl radical, and peroxynitrite. The ROS occur normally in the body at very low concentrations (nanomolar to micromolar) [27]. The body needs them for survival but excess ROS might have deleterious effects [28,29] because the GI tract maintains an appropriate balance of the levels of ROS. The enzyme superoxide dismutase decreases the level of superoxide by catalyzing its conversion into molecular oxygen or hydrogen peroxide [27,30]. Similarly, catalase also catalyzes the degradation of hydrogen peroxide into oxygen and water. Glutathione peroxidases decrease the levels of peroxides using reduced glutathione (GSH) [31]. This requires maintenance of an appropriate ratio of GSH and its oxidized form (GSSG). Whereas γ-glutamylcysteine ligase and GSH synthase are pivotal for the GSH synthesis, nicotinamide adenine dinucleotide phosphate (NADPH)-dependent GSSG reductase is key to the conversion of GSSG to GSH. Ascorbate may also quench ROS and is converted into a transient unstable species and then into dehydroascorbate. Dehydroascorbate reductase uses GSH to recycle it back into ascorbate [27]. The reduced peroxiredoxin can also quench hydrogen peroxide and gets oxidized in the process. Thioredoxin can convert the peroxiredoxin back by oxidation of thioredoxin which in turn is reduced by thioredoxin reductase and NADPH⁺[27]. The cysteine-cystine cycle also plays a key role in maintaining the appropriate balance of ROS [32].

High levels of ROS may damage cells by oxidizing lipids and DNA and protein in various tissues [27]. In the GI tract, the cellular redox potential may even determine the fate of the differentiating cells [9]. Wnt/β-catenin and Notch signaling pathways are major determinants of commitment of the cells for specialization. There may be a direct relationship among NADPH oxidase 1 (NOX-1)-dependent ROS generation, differential activation of Wnt/β-catenin or Notch signaling pathways, and intestinal proliferation and lineage commitment. Changes in the redox potentials of GSH/GSSG and cysteine/cystine (Cys/CySS) have been correlated with intestinal phenotypic transitions. Reducing potentials may favor cell proliferation, but a switch to oxidizing potentials may lead to cell growth arrest [9,33].

Studies with animal models have shown that IBD leads to decreased glutathione concentrations, high DNA oxidation, lipid peroxidation, myeloperoxidase, and elevated TNF-α and malondialdehyde levels [34]. Catalase and superoxide dismutase levels are also altered in some models [34]. The gene ablation of inducible NO synthase (iNOS) in mice showed significant attenuation of UC indicating that iNOS and NO play a major role in UC [35]. In UC, the inflamed human lamina propria showed an increased infiltration by neutrophils and myeloperoxidase activity [24]. At early stages of CD, there is a patchy necrosis of leukocytes near the crypts and an increased number of macrophages and granulocytes [35]. These phagocytotic cells produce large amounts of NO and superoxide, resulting in the formation of peroxynitrite which can cause tyrosine nitration and also cause DNA damage. Thus, in CD and UC, there is an increased production of ROS and peroxynitrite by phagocytotic cells (neutrophils, monocytes, and macrophages) that infiltrate the mucosa. The inflammation also results in increased TNF-α which in turn can activate iNOS, thereby producing even larger amounts of NO. TNF-α also activates NF-κB, which translocates to the nucleus and activates mucosal inflammation genes such as the cytokines IL-6, IL-8 IL-1β, IL-10, TNF-α, ICAM, and P-selectin [34,35]. This inflammatory pathway further triggers the generation of ROS. The importance of oxidative stress in this pathway is further confirmed by the use of sulfasalazine and mesalamine in the therapy for IBD [23,24,34,35].

The conventional hypothesis is that antioxidant supplements simply scavenge ROS. However, based on the kinetic constraints, it is not possible for any concentrations of antioxidant supplements to scavenge ROS sufficiently to prevent biological damage [32]. An alternative hypothesis is that antioxidant supplements modulate endogenous mechanisms that either decrease the production of ROS or increase the enzymes that decompose ROS. One key pathway by which several antioxidants work involves the nuclear factor erythroid-related factor 2 (Nrf2) and the electrophile response element (EpRE), which is also called the antioxidant response element (ARE) [32,33]. Nrf2 is normally bound to Keap1 in the cytoplasm, but polyphenols dissociate this complex to release Nrf2 which is then phosphorylated and transported to the nucleus where it binds EpRE and results in the transcription of genes for several enzymes which can either degrade ROS or prevent their production [32,33]. Several antioxidants including curcumin, boswellic acids, planar aromatics, phenols, and rosmarinic acid can activate the Nrf2 pathway in animals [32,36,37,38,39]. Antioxidant supplements may also decrease ROS levels through action on other enzymes [40]. For example, one of the enzymes that generates superoxide anions is xanthine oxidase, and allopurinol inhibits it [85]. Curcumin inhibits the inflammatory pathway by decreasing the activities of the aryl hydrocarbon receptor, IL-1β and COX, the AKT/mTOR pathway that controls the energy, and also the insulin growth factor and insulin pathways [40]. Resveratrol influences the aryl hydrocarbon receptor, COX, protein kinase C, adenosine monophosphate (AMP) kinase, and sirtuin 1 [40]. Since individual tissues differ in the activities of the enzymes involved, the action of antioxidant supplements may also be tissue and disease dependent. Hence, it is important to determine which antioxidant supplements are efficacious specifically against GI diseases.

Vitamins, minerals, and related compounds which act as cofactors for key enzymes are important to human health and these, as well as the effects of exercise, have been examined extensively [41,42,43]. Here the focus is human trials that examined the efficacies of various antioxidant supplements against GI diseases.

The initial general Medline search using the following keywords resulted in 372 references: “celiac or Crohn's or inflammatory bowel or irritable bowel or colitis” and “antioxidants” and “clinical trials”. A modified search using the following keywords and the name of each antioxidant led to a total of 340 publications: “celiac” or “Crohn's” or “inflammatory bowel” or “irritable bowel” or “colitis”, and “trial”. However, very few of these searches were human clinical trials. Most of them were reviews or in vitro or animal studies. The original studies on human clinical trials are listed in Tables 1, 2, 3. These studies differed not only in the antioxidants tested, but also in the study methods, criteria, sample sizes, study duration, use of placebos or controls, rigor, and subjectivity (random, blind, open and questionnaires, types of clinical symptoms measured) [113,114,115,116,117,118]. Benefit ratings were provided for some guidance. A rating of “++” was given when there were either statistically significant benefits to the treated groups over the controls or there was a significant increase in the percent of patients with improvements and there were little or no adverse effects. The “+” rating was given for the cases where there were only marginal differences in some of the parameters, A rating of “±” was given when there were no statistically significant benefits. The studies where the adverse effects outweighed benefits received the rating “-”.

Allopurinol had been tested as a combined therapy with routinely used drugs such as sulfasalazines or steroids [44,45,46,47,48,85]. In 3 open studies on IBD, allopurinol combination therapy increased thioguanine nucleotide (6-TGN) levels to greater than with monotherapy levels with sulfasalazine, mesalamine or steroids (Table 1) [44,45,46]. In UC patients, combination therapy with allopurinol gave better relapse rates than sulfasalazine and prednisone enema alone, but the results on combination therapy with mesalamine were not as promising (Table 1) [47,48]. Thus, allopurinol may be useful as an adjunct therapy to enhance the levels of the immunosuppressive metabolites like 6-TGN.

Boswellia serrata resins contain boswellic acids such as acetyl-keto-β-boswellic acid which is a lipoxygenase inhibitor and may have anti-inflammatory properties [49]. Boswellia resins have been used in Ayurvedic medicine (shallaki or salai) [50] and in Europe (olibanum or frankincense) [51]. Boswellia extracts proved useful against collagenous colitis in one study but not in another (Table 1) [52,94]. A random double-blind control study showed Boswellia extracts to be useful in CD but less efficacious than mesalamine [49]. Of greater interest is a small study on juvenile CD where capsules containing Boswellia plus several other ingredients were efficacious in maintaining remission over several years [53]. A small open study in patients with UC for ≥6 weeks indicated Boswellia extracts to be superior over sulfasalazine in maintaining remission [54]. Thus, Boswellia has been reported to be useful as a replacement for mesalamine and sulfasalazine in GI diseases, but its usefulness along with other antioxidants may be of greater value.

Curcumin is anantioxidant inturmeric (the rhizome of Curcuma longa or Curcuma domestica) which is used extensively as a spice. It inhibits inflammatory pathways by decreasing activities of the aryl hydrocarbon receptor, IL-1β, and COX; the AKT/mTOR pathway that controls the energy; and insulin pathways [40]. It is well tolerated. A random blind study with 207 patients over 8 weeks showed curcumin to be beneficial for IBS patients, but the study did not contain a placebo control [55]. In 2 independent random double-blind placebo controlled studies, combined therapy with curcumin and mesalamine was superior over mesalamine alone [56,57]. In 2 small open studies curcumin showed marginal efficacy for IBD in adult and pediatric patients [58,59]. It is noted that these 2 studies contained both UC and CD patients. Thus, the evidence for curcumin alone as a therapy for UC may be marginal, but the evidence is better for its usefulness in its combined therapy with mesalamine.

The term chamomile is used for members of Asteraceae, particularly Matricaria chamomilla. Their flowers contain antioxidant and anti-inflammatory compounds such as bisabolol, chamazulene, apigenin, and luteolin [60]. Two random double-blind controlled studies on UC patients found that a chamomile preparation which also contained myrrh and coffee charcoal was as effective as mesalamine [61,62]. However, since the preparation contained many components, one cannot assign the beneficial effects to chamomile alone.

Wormwood or Artemesia species are members of Asteraceae and contain antioxidants such as artemesinin, myrcene, and camphor [63]. A small random double-blind placebo controlled study reported that therapy with wormwood extracts decreased or eliminated the need for steroids in CD patients [64]. Another open study also found wormwood to be beneficial for CD patients [65]. Both studies, however, had small sample sizes and short duration.

Peppermint (Mentha piperita) contains several antioxidants such as rosmarinic acid, eriocitrin, luteolin, and terpenes [66]. Table 1 lists 7 studies that examined the benefits of peppermint in IBS [67,68,69,70,71,72]. Based on these studies, at best peppermint has a marginal efficacy. However, it is important to note that there are no adverse reactions although there was a high dropout rate in one trial [72].

The following antioxidants have also been examined, but there is insufficient evidence concerning their benefits: Aloe vera [86,87,88,89]Andrographis paniculata [90,91], bilberry [92,93], Capsicum [95,96,97], carnosine [98], ferrous fumarate [99], green tea [100], kiwifruit [101], mastic gum [102], N-acetyl cysteine [103], oxpentifylline [104], pomegranate [106], psyllium [107,108], Pycnogenol [109], resveratrol (red wine) [110], superoxide dismutase [11,111], and tormentil [112] (Table 1).

Omega-3 and omega-6 polyunsaturated fatty acids (PUFA) are present in fish oil and also to some degree in plant flax seeds (Linum usitatissimum). PUFA and fish oil have been examined for their therapeutic effects on CD, UC, and IBS (Table 2) [73,74,75,76,77,78,119,120,121,122,123,124,125,126,129,130]. Some of the studies examined PUFA along with several other ingredients and hence it is not possible to ascertain whether the observed effects were due to PUFA. Several other studies had small sample sizes and/or were not properly controlled. Only 3 random double-blind placebo controlled trials were found [73,74,75]. One study found mesalamine to control remission in pediatric CD patients but 95% of patients relapsed within 1 year. Adjunct therapy with PUFA decreased this relapse rate to 61% [73]. The multicentered Epanova Program in Crohn's (EPIC) Study included 2 trials named EPIC1 and EPIC2 in which patients were in remission. In 1 year, relapse occurred in 32% with PUFA and 36% with placebo in EPIC1, and 48% with PUFA and 49% with placebo in EPIC2 [75]. Thus, PUFA were ineffective in maintaining the remission. Another study used enteric coated PUFA capsules for remission maintenance in patients who had a high risk of relapse and found PUFA to be more effective than placebo [74]. Thus, there is only a marginal evidence for a low-grade efficacy of PUFA against CD. The evidence seems to be similar or slightly better for PUFA efficacy against UC (Table 2). Of interest is a small random double-blind placebo crossover study where patients on prednisone or sulfasalazine reported significant improvement with PUFA [76]. Another random double-blind placebo crossover study showed that PUFA improved the oxidative stress status in active UC patients receiving sulfasalazine [77]. Another small study showed that PUFA have some efficacy against UC, but it is less than that of sulfasalazine [78]. Thus, the efficacy of PUFA or fish oil against CD or UC is, at best, marginal; however, none of the studies reported any adverse effects of PUFA supplementation.

Léi gōng téng (Tripterygium wilfordii), which is rich in antioxidant polyglycosides, has been tested for its efficacy against CD (Table 3). The trials were random (some of them blind) with very small to medium sample sizes over a period of 1 year. Léi gōng téng performed better than azathioprine, mesalamine, and sulfasalazine [79,80,81,82]. It is cautioned that Léi gōng téng has male antifertility properties and this side effect is not monitored in these studies.

The remaining Chinese and Ayurvedic medicines reported in Table 3 contain multiple herb blends. There were 2 studies [83,84] on UC in which the herbal medicines were significantly efficacious. The first was a study [83] on ulcerative proctitis patients using an enema of the multiherb drug xilei san. In this random double-blind placebo controlled trial, xilei san (Chinese medicine) showed improvement within 14 days, and after 6 months the remission rate was 81.8% with xilei san versus only 16.7% with the placebo. There was an overall improvement based on several parameters measured. The second was an open study [84] that used an Ayurvedic medicine containing the herbs Aegle marmelos Correa plus Bacopa monnieri plus a Ficus glomerata enema. In this study the patients were allowed to continue with steroids or sulfasalazine as needed. In 4 weeks, the patients showed significant improvement in their symptoms and stopped taking the other drugs. However, the sample sizes in both the studies were small and longer-term assessment is also needed. There are several other studies shown in Table 3 in which marginal or no benefits of the multiherbal preparations were observed on IBS or UC [131,132,133,134,135,136,137,138]. Thus, the multiple herbs were a mixture and even when they worked the active principles in them were not identified.

The results reported on the effects of antioxidant supplements in GI diseases in Tables 1, 2, 3 are very heterogeneous. There may be many reasons for this. First, the definitions of different GI diseases overlap and possibly so do the diagnosis, etiology, and pathophysiology. The consequence is that different researchers have used a variety of indices to monitor the efficacies of the antioxidants. Some of the methods were quite subjective. Also, a lack of uniformity in the studies in following the Cochrane criteria for clinical trials added to the confusion. This led to assigning the ++, +, ±, and - ratings for the efficacies and separate star ratings for the rigor and quality of each study so that the reader may choose which substances they should explore more thoroughly. The reader is further cautioned that most studies did not comment on the quality of the antioxidant substances being used. These might vary with the subspecies used, suppliers, age of the products, and how the extracts were prepared.

Combined therapy of current treatments with allopurinol (that may decrease the generation of superoxide) has proved useful for CD and UC patients. Curcumin may also have provided a benefit for UC patients in combined therapy with mesalamine. Other substances for which benefits against CD might be Boswellia (frankincense or shallaki), Artemesia sp. (wormwood), T. wilfordii (Léi gōng téng), and fish oil, and for UC such antioxidants might be allopurinol, M. chamomilla(chamomile), and fish oil. Some studies showed benefits of multiherbal Chinese and Ayurvedic medicines in UC but the studies were not replicated in large clinical trials. For IBS patients, a number of substances that have small benefits and no adverse effects are listed in Tables 1 and 3.

Based on these findings, 3 levels of recommendations have been made. The first recommendation is that allopurinol and possibly curcumin should be used for combined therapy in IBD patients with the current treatments using mesalamine, azathioprine, sulfasalazine, and steroids. Secondly the combined therapy trials should be considered using the current treatments and the following antioxidants: Boswellia, Artemesia, T. wilfordii, fish oil, and chamomile. Studies on more promising antioxidants should be replicated that include Chinese and Ayurvedic medicine-based antioxidants. The third recommendation is that there was insufficient evidence for the benefits of several antioxidants against IBS, but these did not have any adverse effects.