A Systematic Review of the Efficacy and Safety of Turmeric in the Treatment of Digestive Disorders

Abstract

Background:

Turmeric has been gaining popularity as a treatment option for digestive disorders, although a rigorous synthesis of efficacy has not been conducted.

Objective:

To summarize the evidence for the efficacy and safety of turmeric in the treatment of digestive disorders, including inflammatory bowel diseases (IBD), irritable bowel syndrome (IBS), dyspepsia, gastroesophageal reflux disease (GERD), and peptic ulcers.

Search methods:

Literature searches were conducted in Medline, EMBASE, AMED, the Cochrane Central Register of Control Trials, and Dissertation Abstracts from inception to 15 November 2021.

Selection criteria:

Dual independent screening of citations and full texts was conducted and studies meeting inclusion criteria were retained: randomized controlled trials (RCT) and comparative observational studies evaluating turmeric use in people of any age with one of the digestive disorders of interest.

Data collection and analysis:

Extraction of relevant data and risk of bias assessments were performed by two reviewers, independently. Meta-analysis was not conducted due to high heterogeneity.

Main results:

From 1,136 citations screened, 26 eligible studies were retained. Most studies were assessed to have high risk of bias, and many had methodological limitations. Descriptive summaries suggest that turmeric is safe, with possible efficacy in patients with IBD or IBS, but its effects were inconsistent for other conditions.

Conclusions:

The efficacy of turmeric in digestive disorders remains unclear due to high risk of bias and methodological limitations of the included studies. Future studies should be designed to include larger sample sizes, use rigorous statistical methods, employ core outcome sets, and adhere to reporting guidance for RCTs of herbal interventions to facilitate more meaningful comparisons and robust conclusions.

Funding sources:

This project is partially funded by the Li Ka Shing Post-doctoral Fellowship Program at the St. Michael’s Hospital to KT. Research reported in this publication was also partially supported by the National Center for Complementary and Integrative Health of the National Institutes of Health under award number R24AT001293. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

1. Introduction

Digestive disorders, such as inflammatory bowel disease (IBD, including ulcerative colitis (UC) and Crohn’s disease (CD)), irritable bowel syndrome (IBS), gastroesophageal reflux disease (GERD), peptic ulcers, and dyspepsia contribute significantly to the global burden of disease. Although highly prevalent in developing countries¹, the burden of gastrointestinal disease in developed countries is also high: > 300 million people were estimated to be living with a digestive disorder across 44 European countries in 2019² and > 43 million ambulatory visits with a primary gastrointestinal diagnosis occurred in the United States in 2016³, with healthcare expenditures totalling approximately 120 billion USD³. In addition to high financial costs, digestive disorders can also significantly impair the quality of life of afflicted individuals, affecting their ability to participate fully in their professional and personal lives^4,5. Effective, low-cost treatments for digestive disorders would reduce healthcare costs and improve the quality of life for millions of people globally. Complementary and alternative medicines (CAM) may be promising options in the treatment of these conditions.

Conventional treatment options for digestive disorders include pharmacological, behavioural, and psychological interventions⁶. Complementary and alternative medicines are popular among patients as adjunct or primary treatments for these disorders^7–9, with up to 50% of people living with IBD using CAMs¹⁰. Turmeric has been used for over 4,000 years in Southeast Asia as a spice and in religious events. It has traditionally been used to treat chronic illnesses such as asthma, liver problems, diabetic wounds, and abdominal pain in Ayurvedic and traditional Chinese medicine¹¹. The polyphenol curcumin, derived from turmeric, has been shown to have anti-inflammatory, antioxidant, antimicrobial, and immunomodulatory effects in some in vitro and animal studies^12,13, demonstrating potential as a treatment for digestive disorders¹⁴. The effectiveness of curcumin-containing regimens for the treatment of digestive disorders has also been assessed in various in-human controlled trials, with many reporting beneficial outcomes in terms of favourable remission rates^15,16, reduced severity of gastrointestinal symptoms¹⁷, and improved quality of life¹⁷. In contrast, other studies have reported no significant differences in several patient outcomes between curcumin and placebo or a conventional pharmacological treatment^18,19. A recent systematic review and meta-analysis found that adjuvant curcumin use produced a significant beneficial effect on clinical remission of UC compared to placebo; however, no significant effects were identified regarding clinical improvement, endoscopic remission, or endoscopic improvement²⁰. The review found a possible dose-response relationship and suggested that different therapeutic formulations, such as combinations with other elements, may increase the solubility, bioavailability, and effectiveness of curcumin²⁰. The safety profile of turmeric, particularly its active compound curcumin, has been extensively studied and is generally considered safe in animals and humans when consumed in typical dietary amounts²¹. At the time of publication, no rigorous reviews have been published regarding the efficacy and safety of curcumin in the treatment of digestive disorders, a wide range of conditions that impact the oropharynx, gastrointestinal tract, liver, biliary system, and pancreas. This systematic review aims to synthesize the available evidence regarding the efficacy and safety of curcumin and its derivatives or compounds in the treatment of common digestive disorders, including IBD, IBS, dyspepsia, peptic ulcers, and GERD.

2. Methods

A protocol for this review was prepared a priori²², and protocol amendments are described in Appendix 1. The results of this review have been reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidance²³.

2.1. Search Strategy

Search strategies were developed by an experienced information specialist with input on search terms provided by members of the research team (Appendix 2). The strategies utilized a combination of controlled vocabulary (e.g., “Inflammatory Bowel Disease,” “Irritable Bowel Syndrome,” “Dyspepsia,” “Curcuma,” “Curcumin”) and keywords (e.g., “gastric acid reflux”, “ileitis”, “GERD”). Vocabulary and syntax were adjusted across databases and were not restricted by language or date. Searches were conducted in Medline, EMBASE, AMED, the Cochrane Central Register of Control Trials, and Dissertation Abstracts on 15 November 2021. Bibliographies of included studies were manually searched to identify additional studies.

2.2. Study Eligibility Criteria

Inclusion criteria were developed based upon the Population-Intervention-Comparator-Outcome-Study design (PICOS) framework (Table 1).

Table 1.

Study eligibility criteria

PICOS Element	Description of Criteria
Population	People of any age diagnosed with IBD, IBS, dyspepsia, GERD, or peptic, gastric, or duodenal ulcers
Intervention	Curcumin or a derivative, either as a single- or multi-component product, used either as a primary or adjunct treatment of one of the conditions of interest at any dose, for any duration, and via any route of administration
Comparator	Placebo, active comparator, or no treatment
Outcomes	Measures of response to treatment (e.g., relief of gastrointestinal symptoms measured with validated scales/scores, remission), relapse, quality of life, hospitalization, physician visits, adverse events We excluded individual symptom scores (e.g., changes in bloating, nausea, abdominal pain) that were not part of a validated composite scale.
Study design	Randomized controlled trials, non-randomized or quasi-randomized controlled trials, or cohort, case-control, or longitudinal studies with a control group

2.3. Study Selection Process

Following de-duplication, citations identified by the searches were uploaded to the online systematic review software Covidence (www.covidence.org; Melbourne, Australia). Screening was conducted in two stages: (1) screening of titles and abstracts and (2) screening of the full texts of citations found to be potentially relevant at Stage 1. Two reviewers independently screened citations at both stages, with conflicts resolved through discussion or a third reviewer. The process of study selection was summarized using a flow diagram as per PRISMA recommendations.

2.4. Data Extraction and Outcome Parameters

Data extraction was conducted in duplicate independently by two reviewers, with verification by a third reviewer to ensure consistency of extraction across the review updates. The collected data included study characteristics (i.e., first author, year of publication, country, study design, sample size, recruitment period, study setting), clinical and demographic characteristics of the study sample (i.e., digestive disorder of interest, age, sex), the nature of the curcumin intervention and the comparator(s) (i.e., product name and formulation, dose, treatment duration), and outcome data (i.e., outcome definition and scale (if applicable), group sizes and numbers of events for dichotomous outcomes and mean and standard deviation (SD) for continuous outcomes at baseline and follow-up, between- and within-group differences, p-values, statistical tests used). For each condition of interest, “response to treatment” was the primary outcome of interest, as defined by the included study (e.g., disease-specific clinical scores, such as the Simple Clinical Colitis Activity Index (SCCAI) in IBD patients; presence of ulcer healing or the eradication of H. pylori infection in people with peptic ulcers).

2.5. Risk of Bias Assessment

Risk of bias of the included studies was assessed using the Cochrane Collaboration’s Risk of Bias Tool (ROB-1)²⁴. Studies were assessed to have an overall low, unclear, or high risk of bias based on the highest risk of bias determined to be present in the seven domains of the ROB-1 tool. Risk of bias assessment was first piloted using two studies, and scoring results between reviewers were compared and discussed. Discrepancies in scoring methodology were discussed to reduce subjectivity before the remaining studies were assessed independently by the two reviewers. All risk of bias assessments were verified by a third reviewer after the final review update to ensure consistency across updates. Findings from these assessments were summarized descriptively to provide an overview of strengths and weaknesses of the included evidence.

2.6. Data Synthesis

Meta-analyses were of a priori interest to combine data across studies for each condition of interest. Due to the high heterogeneity of study methodologies, patient conditions, disease severities and durations, curcumin interventions and doses, measures or definitions of treatment success, and types of analysis (e.g., between- or within-group analyses), meta-analyses were not performed, and study findings have been synthesized and reported descriptively. For synthesis and reporting, studies were grouped according to digestive disorder of interest (i.e., IBD; IBS; dyspepsia; peptic, gastric, or duodenal ulcers; other gastrointestinal conditions) and descriptive summaries of evidence regarding the outcomes of interest were reported within each digestive disorder. Where comparative analyses were not reported but raw data were available, we conducted Chi-squared or Fisher’s exact tests as necessary using Stata/SE 15.1 (StataCorp LLC, College Station, TX). For data visualization, forest plots of study findings were generated by patient condition and outcome, when possible, using Comprehensive Meta-Analaysis (CMA) software (Version 3.3.070; Biostat Inc., Englewood, NJ). To standardize the visualization of the variously reported effects, raw data were used to calculate standardized mean differences for display in forest plots by CMA. Therefore, effect estimates reported in the text (i.e., the effect estimates extracted from the studies) may differ from those reported on forest plots. As well, if findings from per-protocol dichotomous analyses were reported and baseline group sample sizes were known, we used the baseline group sample sizes to calculate the equivalent intention-to-treat analysis findings in CMA for forest plots. For dichotomous outcomes, when zero events were reported in either group, 0.5 was added to all cells (i.e., events and non-events for both treatment groups) to calculate the risk ratio. For continuous outcomes, we used a pre-post correlation of 0.75 and standardized by change score SD to calculate study-level differences of differences.

3. Results

3.1. Availability of the literature

Literature database searches returned 1,159 references after duplicate removal and an additional 14 references were identified by manual searches, resulting in 1,173 references available for screening. Of these, 1,046 references were excluded at title/abstract screening, and 101 references were excluded at full-text screening, leaving 26 studies that were included (Figure 1). A list of references excluded at full-text screening and reasons for their exclusion is available in Appendix 3.

Figure 1. Evidence Flow Diagram.

3.2. Study Characteristics

Characteristics of the 26 included studies are presented in Table 2. All studies^{15–18,25–46} were published between 1989 and 2021 and were conducted in Asia (n = 13)^{27,28,32–35,40–46}, Europe (n = 6)^{17,18,26,29–31}, the Middle East (n = 5)^{15,25,37–39}, and Australia (n = 1)³⁶, with one multi-country study (i.e., Israel, Hong Kong, and Cyprus)¹⁶. Twenty-five studies were randomized controlled trials (RCTs)^{15–17,25–46}, and one was a randomized controlled crossover trial¹⁸. The patient conditions that were studied included inflammatory bowel disease (IBD; ulcerative colitis or Crohn’s disease; n = 9 studies)^{16,28,29,32,33,37,39,41,43}; irritable bowel syndrome (IBS; n = 5)^{17,18,26,30,31}; dyspepsia (n = 5)^{38,40,42,44,46}; peptic, gastric, or duodenal ulcers (n = 4)^15,25,35,45; Helicobacter pylori infection (n = 2)^27,34; and “gastrointestinal complaints” (i.e., mild to moderately severe gastrointestinal discomfort, excluding inflammatory bowel diseases and functional gastrointestinal disease; n = 1)³⁶. Mean patient age ranged from 22.7 to 56.4 years across the studies, with studies of UC/CD ranging from 34.1 to 42.5 years^{16,28,29,32,33,37,39,41,43} and studies of IBS ranging from 40.5 to 50.3 years^17,18,26,30 (one other IBS study reported that the majority of patients were > 50 years³¹). The average proportion of female patients in the 18 studies that reported these data was 51.8% (range 22–92.6%)^{15–18,27,28,30–35,37,39,41,42,44,45}.

Table 2.

Summary of Study Characteristics by Patient Condition

Patient Condition	Study Country	Study Design Sample size Mean age Percent female	Interventions Treatment duration	Patient condition definition for inclusion in study	Overall risk of bias	Outcomes
Inflammatory Bowel Disease (IBD; Ulcerative colitis (UC) or Crohn’s Disease (CD))	Banerjee et al., 202028 India	RCT n = 69 38 years (median) 37.7%	Bio-enhanced curcumin (Valdone; Cadila Pharmaceuticals Ltd., India), 50 mg soft gel orally BID + mesalamine Placebo + mesalamine Duration NR	Mild to moderate UC disease activity with active endoscopic inflammation (partial Mayo score 2 to 6 with endoscopic Mayo score > 1)	High	RTT: partial Mayo score, Mayo endoscopic sub-score
	Bommelaer et al., 202029 France	RCT n = 62 36.3 years NR	Curcumin capsule 1,000 mg orally TID + azathioprine Placebo + azathioprine 24 weeks	CD treated with thiopurines and a recent surgical resection of all ileal, ileocolonic, or colonic macroscopic lesions, with an anastomosis that could be reached by ileocolonoscopy	Unclear	Relapse: endoscopic Rutgeerts index, CDAI QoL: IBDQ
	Sadeghi et al., 202039 Iran	RCT n = 70 40.4 years 70%	Curcumin capsule 500 mg orally TID + background therapy (salicylates +/– immunomodulators +/– corticosteroids) Placebo + background therapy (salicylates +/– immunomodulators +/– corticosteroids) 8 weeks	UC diagnosed based on clinical records, colonoscopy, and pathology; having symptoms of active mild-to-moderate UC (5–12 on the Simple Clinical Colitis Activity Index [SCCAI])	High	RTT: SCCAI QoL: IBDQ-9
	Sugimoto et al., 202043 Japan	RCT n = 30 35.1 years NR	Bio-enhanced synthetic curcumin (Theracurmin®; Theravalues Col., Ltd., Japan), capsule 180 mg orally BID + background therapy (salicylates and/or immunomodulator) Placebo + background therapy (salicylates and/or immunomodulator) 12 weeks	CD definitively diagnosed based on radiologic, endoscopic, and histologic criteria set by the Research Committee of Inflammatory Bowel Disease that is affiliated with the Japan Ministry of Health. Patients treated with medications outside of the Japanese medical insurance system were excluded.	High	RTT: CDAI; SESCD; improvement of anal lesions, based on physician’s judgement
	Masoodi et al., 201837 Iran	RCT n = 56 37.1 years 50%	Bio-enhanced curcuminoid nanomicelles 80 mg orally TID + mesalamine Placebo + mesalamine 4 weeks	Mild-to-moderate UC, according to SCCAI	Unclear	RTT: SCCAI QoL: self-reported well-being
	Kedia et al., 201733 India	RCT n = 62 34.9 years 33.9%	Curcumin capsule 150 mg orally TID + mesalamine Placebo + mesalamine 8 weeks	mild-to-moderate active UC (3–9 on the Ulcerative Colitis Disease Activity Index [UCDAI]; with a minimum sigmoidoscopic score of 2, with at least one previously documented attack of active disease	Unclear	RTT: UCDAI, healing based on endoscopy
	Lang et al., 201516 Israel, Hong Kong, and Cyprus	RCT n = 50 40.9 years 34%	Curcumin (Cur-Cure; Bara Herbs Inc., Israel), capsule 1,500 mg orally BID + mesalamine Placebo + mesalamine 4 weeks	UC with SCCAI score ≥ 5 and < 12	High	RTT: Mayo endoscopic sub-score, SCCAI
	Singla et al., 201441 India	RCT n = 45 34.1 years 48.9%	Enema of C. longa extract (NCB-02, Himalaya Drug Company, Bangalore)a, 140 mg/enema + mesalamine Placebo enema + mesalamine 8 weeks	UC with UCDAI score 3–9	Unclear	RTT: UCDAI
	Hanai et al., 200632 Japan	RCT n = 89 42.5 years 44.9%	Curcumin 1,000 mg orally BID + sulfasalazine or mesalamine Placebo + sulfasalazine or mesalamine 24 weeks	UC fulfilling radiologic, endoscopic, or histologic criteria, according to the Research Committee of Inflammatory Bowel Disease and the Japanese Ministry of Health, and in clinical remission based on a Clinical Activity Index (CAI) ≤ 4 and stable for the previous 4 weeks	Unclear	Relapse: CAI, EI
Irritable bowel syndrome (IBS)	Alt et al., 201726 Germany	RCT n = 90 45.8 years NR	Curcuminoids (IQP-CL-101; Xanthofen™)b, soft gel 660 mg curcuminoids orally BID Placebo 8 weeks	IBS diagnosed based on Rome III Criteria, with average abdominal pain/discomfort score ≥ 4 and ≤ 8 on 11-point Likert scale	High	RTT: IBS-SSS – overall score and abdominal pain domain, IBS-GIS, global assessment of benefit (4-point scale) QoL: IBS-QOL
	Lauche et al., 201618 Germany	Randomized Controlled Cross-over Trial n = 32c 50.3 years 59.4%	Herbal compound powder containing turmericd, 5 g of compound orally BID Placebo 4 weeks	IBS diagnosed based on ROME III criteria and the German IBS guidelines	Unclear	RTT: IBS-SSS QoL: IBS-QOL, EQ-5D index, EQ-5D VAS
	Portincasa et al., 201617 Italy	RCT n = 116 40.5 years 66.4%	Curcumin-fennel essential oil (CU-FEOe; Alfa Wassermann S.p.A., Italy), 42 mg curcumin capsule, orally, BID Placebo 4 weeks	IBS diagnosed based on ROME III criteria	Unclear	RTT: IBS-SSS, abdominal pain VAS QoL: IBS-QOL, IBS-SSS (interference with QoL domain)
	Brinkhaus et al., 200530 Germany	RCT n = 106 48.1 years 63.2%	C. xanthorrhiza extractf tablet 20 mg orally TID Fumaria officinalis (fumitory) 250 mg TID Placebo 18 weeks	IBS-related pain and distension between “tolerable” and “severe,” frequency of defecation “more frequent than normal” or “atypically rarely”	Unclear	RTT: global assessment scale (not defined) – IBS symptoms, Pain VAS QoL: global assessment scale (not defined) – psychosocial stress
	Bundy et al., 200431 United Kingdom	RCT n = 207 NR >70%	Turmeric extract tablet (Cynara™ Turmeric, Lichtwer Pharma (UK) Ltd., UK); 72 mg orally SID Turmeric extract tablet (Cynara™); 144 mg orally SID 8 weeks	IBS diagnosed based on Rome II criteria	Unclear	RTT: IBS prevalence based on Rome II criteria QoL: IBS-QOL
Dyspepsia	Panahi et al., 202138 Iran	RCT n = 75 38.6 years NR	Curcuminoids (C3 Complex®; Sami Labs, Ltd., Bangalore, India)g capsule 500 mg orally SID + famotidine Placebo + famotidine 4 weeks	Functional dyspepsia according to Rome III criteria	High	RTT: HKDI, H. pylori antigen stool tests
	Yongwatana et al., 202146 Thailand	RCT n = 132 56.4 years NR	C. longa Linn capsule 500 mg orally QID Omeprazole 20 mg SID Placebo 4 weeks	Functional dyspepsia according to ROME IV criteria	Unclear	RTT: SODA – pain domain and total score QoL: EQ-5D index, EQ-VAS
	Sawangroj et al., 201940 Thailand	RCT n = 78 38.6 years NR	C. longa capsule 250 mg orally TID C. longa capsule 500 mg orally TID Simethicone 80 mg TID 4 weeks	Postprandial distress syndrome Functional dyspepsia according to Rome IV criteria	High	RTT: GOS – composite of PDS symptom domains (i.e., early satiety + postprandial fullness)h and epigastric pain domain
	Sirijarugul et al., 200742 Thailand	RCT n = 120 22.7 years 92.6%	C. longai capsule 500 mg orally QID Ranitidine 4 weeks	Mild-to-moderate dyspepsia	High	RTT: SODA – pain domain
	Thamlikitkul et al., 198944 Thailand	RCT n = 106 42.0 years 49%	C. domestica Val. capsule 500 mg orally QID Flatulence™j Placebo 1 week	Acid dyspepsia, flatulent dyspepsia, or atonic dyspepsia	High	RTT: subjective improvement
Peptic, gastric, or duodenal ulcer	Abbas et al., 201725 Iraq	RCT n = 40 42.7 years NR	Curcumin capsule 500 mg orally TID + esomeprazole 20 mg BID + amoxicillin 1,000 mg BID + clarithromycin 500 mg BID Esomeprazole + amoxicillin + clarithromycin 2 weeks	Newly endoscopically diagnosed with peptic ulcer and detection of H. pylori	High	RTT: H. pylori rapid antigen and immunoglobulin M enzyme-linked immunosorbent assay (ELISA) stool tests
	Khonche et al., 201615 Iran	RCT n = 68 35.1 years 56.7%	Curcuminoids (C3 Complex®)g capsule 500 mg orally SID + pantoprazole 40 mg BID + amoxicillin 1,000 mg BID + clarithromycin 500 mg BID Placebo + pantoprazole + amoxicillin + clarithromycin 4 weeks	Gastric or duodenal ulcers identified by gastroscopy, and H. pylori infection based on urease testing of biopsy specimens of antral mucosa	High	RTT: HKDI, H. pylori urea breath test, HKDI – stomach pain and dyspepsia symptoms domains
	Van Dau et al., 199845 Vietnam	RCT n = 123 36.5 years 22.0%	C. longa tablet 2,000 mg orally TID Placebo 8 weeks	One duodenal ulcer with a minimum diameter of 5 mm, confirmed by endoscopy	Unclear	RTT: ulcer healing, using endoscopy, radiography, and laboratory testing; subjective symptom relief
	Kositchaiwat et al., 199335 Thailand	RCT n = 60 50.5 years 36.7%	C. longa capsule 250 mg orally TID Antacid (aluminium hydroxide and magnesium hydroxide) TID 6 weeks	Benign gastric ulcer with or without an associated non-gastric ulcer lesion	High	RTT: endoscopic healing
Helicobacter pylori infection	Asif et al., 201527 Pakistan	RCT n = 210 35.7 years 44.9%	C. longa in a phytomedicine-based treatment tablet (“Pylorex Plus”)k, 150 mg C. longa orally BID Omeprazole 20 mg + amoxicillin 1,000 mg + metronidazole 500 mg + bismuth 400 mg 4 weeks	Positive stool antigen test for H. pylori	High	RTT: H. pylori urea breath test and stool antigen test, overall gastrointestinal symptom score, abdominal pain score
	Koosirirat et al., 201034 Thailand	RCT n = 36 47.4 years 52.8%	Curcumin tablet 700 mg orally TID Omeprazole 20 mg BID + amoxicillin 1,000 mg BID + metronidazole 800 mg BID 4 weeks	Chronic gastritis with H. pylori infection confirmed by urease breath test (UBT), pathology, or culture	Unclear	RTT: H. pylori eradication based on rapid urease test, pathology, and culture
Gastrointestinal complaints	Lopresti et al., 202136 Australia	RCT n = 79 43 years NR	Curcuminoids extract capsule (Curcugen™)l, 500 mg orally SID Placebo 8 weeks	Self-reported symptoms of gastrointestinal complaints	High	RTT: GSRS; hydrogen and methane SIBO breath tests QoL: SF-36 (all domains)

^aNCB-02 is a standardized extract of C. longa with a composition of 72% curcumin, 18.08% demethoxy curcumin, and 9.42% bis-demethoxy curcumin

^bIQP-CL-101 contains “330 mg proprietary mixture of curcuminoids and essential oils from C. longa and C. xanthorrhiza, 70 mg fish oil, 15 mg peppermint oil, and 8 mg caraway oil as well as 263 gg thiamine, 39 gg folic acid, and 625 gg vitamin D3”

^c32 patients, however, this was a crossover trial with multiple phases: 37 intervention and 35 control phases were completed

^dThe herbal compound consisted of curry (Murraya koenigii), pomegranate (Punica granatum) and turmeric (C. longa rhizome pulvis), in a 6:3:1 ratio

^eCurcumin-Fennel Essential Oil (CU-FEO) contains 42 mg C. longa and 17.5 mg Foeniculum vulgare Miller Oil

^f“C. xanthorriza tablets contained 20 mg spray-dried extract of the herb, as well as lactose, calcium hydrogenphospate, magnesium stearate and talcum”

^gC3 Complex^® contains 500 mg curcuminoids and 5 mg piperine (Bioperine^®, Sami Labs, Ltd.)

^hThe primary outcome was a composite of the two domains of the Global Overall Symptoms score that measured Postprandial Distress Syndrome symptoms: the early satiety domain and postprandial fullness domain

ⁱEach capsule contained curcuminoid 0.024 mg and volatile oil 0.02 ml

^jFlatulence^™ contains 65 mg cascara dry extract, 16 mg nux vomica dry extract, 0.325 mg asofoetida tincture, 8 mg capsicum powder, 48 mg ginger powder, and 3 mg diastase

^kPylorex Plus contains 150 mg C. longa rhizomes, 150 mg Mallotus phillipinesis fruits, 100 mg Glycyrrhiza glabra roots, and 100 mg Zingiber officinale rhizomes

^lCurcugen^™ is 98.5% turmeric based (50% curcuminoids) and 1.5% essential oils and other native turmeric molecules

BID = two times a day; CAI = Clinical Activity Index; CD = Crohn’s disease; CDAI = Clinical Disease Activity Index; EI = Endoscopic Index; ELISA = enzyme-linked immunosorbent assay; GOS = Global Overall Symptoms score; HKDI = Hong Kong Dyspepsia Index; IBD = inflammatory bowel disease; IBDQ = Inflammatory Bowel Disease Questionnaire; IBS = irritable bowel syndrome; IBS-GIS = IBS Global Improvement Scale; IBS-QOL = IBS Quality of Life Instrument; IBS-SSS = IBS Symptom Severity Scale; NR = not reported; QoL = quality of life; RCT = randomized controlled trial; RTT = response to treatment; SCCAI = Simple Clinical Colitis Activity Index; SESCD = Simple Endoscopic Score for Crohn's Disease; SODA = Severity of Dyspepsia Assessment; SID = once daily; TID = three times a day; UBT = urease breath test; UC = ulcerative colitis; UCDAI = Ulcerative Colitis Disease Activity Index; VAS = visual analogue scale

Interventions included the following Curcuma species or derivatives: curcumin (n = 10 studies)^{16,17,25,28,29,32–34,39,43}, C. longa (n = 7)^{27,35,40–42,45,46}, curcuminoids (n = 5)^{15,26,36–38}, “turmeric” (not further described; n = 2)^18,31, C. xanthorrhiza (n = 1)³⁰, and C. domestica (n = 1)⁴⁴. Bio-enhanced oral curcumin was evaluated in three studies of IBD^28,37,43, while non-bio-enhanced products were evaluated in the remaining 23 studies. Doses ranged from 60 mg/day of C. xanthorrhiza extract tablets for IBS³⁰ to 6,000 mg/day of C. longa tablets for duodenal ulcer⁴⁷. Oral administration was used in all but one study, which administered curcumin via enema⁴¹. Adjunct medications or background treatments were included in all studies of IBD (e.g., mesalamine, azathioprine)^{16,28,29,32,33,37,39,41,43}, one of five dyspepsia studies (i.e., famotidine)³⁸, and two of four studies of ulcers (e.g., pantoprazole + amoxicillin + clarithromycin)^15,25. Oral products were formulated as capsules (n = 15 studies)^{15–17,25,29,33,35,36,38–40,42–44,46}, tablets (n = 5)^{27,30,31,34,45}, soft gels (n = 2)^26,28, and powder (n = 1)¹⁸, with two studies not reporting the drug formulation^32,37.

Twenty-four studies (92%) reported response to treatment outcomes^{15–18,25–28,30,31,33–46}, three (12%) reported relapse outcomes^29,32,40, and ten (38%) reported quality of life outcomes^{17,18,26,29–31,36,37,39,46}. Nine studies (35%) reported data on non-gastrointestinal adverse events (AEs)^{15,17,18,27,30,32,41,42,44}.

3.3. Risk of Bias Assessment

Over half of all studies were assessed to have a high risk of bias (14/26 studies; 54%)^{15,16,25–28,35,36,38–40,42–44}, with the remaining 12 studies receiving an unclear assessment. A summary of risk of bias by domain of the Cochrane Risk of Bias tool is depicted in Figure 2. High risks of bias were frequently identified for “Other risks of bias” (e.g., funding source, conflicts of interest, significant differences between groups at baseline), blinding of participants and personnel, and incomplete outcome data. Summaries of risk of bias by study are available in Appendix 4 and Dataset 1.

Figure 2. Risk of bias summary by domain.

Inflammatory Bowel Disease (Ulcerative Colitis and Crohn’s Disease)

Nine RCTs focused on IBD (UC or CD)^{16,28,29,32,33,37,39,41,43} and included a total of 533 patients (mean age: 34.1–42.5 years; proportion female: 33.9–70%). All seven studies of UC^{16,28,32,33,37,39,41} included patients with mild-to-moderate active disease, defined with a variety of scales (e.g., partial Mayo score of 2–6²⁸, Simple Clinical Colitis Activity Index (SCCAI) of 5–12^16,37,39, Ulcerative Colitis Disease Activity Index (UCDAI) of 3–9^33,41, Clinical Activity Index (CAI (not further defined) ≤ 4)³². Patients included in the two studies of CD had either (1) been treated with thiopurines and had a recent surgical resection and anastomosis²⁹, or (2) been definitively diagnosed with CD based on radiologic, endoscopic, and histologic criteria⁴³.

All studies included a concomitant therapy in both the turmeric and placebo arms, either mesalamine as an adjunct (n = 5)^{16,28,33,37,41} or the established background therapy of the patients: salicylates or an immunomodulator (with or without corticosteroids; n = 2)^39,43, azathioprine (n = 1)²⁹, or sulfasalazine or mesalamine (n = 1)³². Across the nine UC/CD studies, turmeric interventions included C. longa (n = 7)^{16,29,32,33,39,41,43}, bio-enhanced curcumin (n = 1)²⁸, and bio-enhanced curcuminoid nanomicelles (n = 1)³⁷. One study administered C. longa and placebo by enema to UC patients⁴¹. Duration of therapy ranged from 4 to 24 weeks.

3.3.1. Response to Treatment

Seven of the nine studies that focused on UC/CD reported response to treatment outcomes^{16,28,33,37,39,41,43}, either as dichotomous or continuous data, evaluating a variety of turmeric interventions (e.g., curcumin either natural extract or synthetic, bio-enhanced or not) administered from 100–3,000 mg/day for 4–12 weeks, orally or by enema, either with mesalamine as adjunct therapy or with the established background therapy of each patient (i.e., 5-aminosalicylates and/or immunomodulators and/or corticosteroids) (Table 2). Seven different treatment response measures were used (Table 2), with six studies reporting fourteen different dichotomous outcome definitions (Figure 3, top)^{16,28,33,39,41,43}, and four studies reporting four different continuous outcome definitions (Figure 3, bottom)^33,37,39,43. There was minimal overlap of outcome definitions across studies.

Figure 3. Forest plot of the response to treatment with curcumin vs placebo in people with inflammatory bowel disease (ulcerative colitis or Crohn’s disease).

Effect estimates in forest plots (risk ratios, differences of differences, and standardized differences of differences) may not match those reported in the text due to use of raw data to generate plots and standardization of some estimates.

Abbreviations. ASA = acetylsalicylic acid; Bio-en = bio-enhanced; CDAI = Crohn’s disease Activity Index; CS = corticosteroid; Diff = difference; IM = immunomodulator; MES = mesalamine; SCCAI = Simple Clinical Colitis Activity Index; SE = standard error; SESCD = Simple Endoscopic Score for Crohn’s Disease; Std diff = standardized difference; tx = treatment; UCDAI = Ulcerative Colitis Activity Index

All studies reported beneficial responses to treatment (Figure 3), although significant findings were only reported in studies that evaluated either curcumin extract administered orally at doses ≥ 1,000 mg/day^16,39 or bio-enhanced curcumin at doses < 1,000 mg/day^28,37,43. Single studies of curcumin extract at doses < 1,000 mg/day³³ and a curcumin enema⁴¹ reported non-significant findings. Studies with small sample sizes and zero events demonstrated reduced precision of estimates of treatment effect^16,43. One study did not report baseline data from which a difference of differences could be calculated³⁷; however, no significant difference between groups was calculated at final follow-up (mean difference_SCCAI = −0.97 (SE = 0.53)).

3.3.2. Relapse

Two trials focusing on the maintenance phase of IBD reported disease relapse findings, one of CD²⁹ and one of UC³². To be eligible, participants in the CD study were required to have had recent surgery to remove all CD lesions²⁹, while participants in the UC study were in corticosteroid-free clinical remission from UC, according to a stable Clinical Activity Index (an undefined scale; CAI ≤ 4) for at least four weeks prior to entry and successful cessation of corticosteroid or alternative medication that had been used to achieve remission³². Both studies compared oral curcumin extract > 1,000 mg/day to placebo.

There were no significant differences between curcumin extract (1,500 mg/day) and placebo in the occurrence of either endoscopic or clinical relapse of CD at six months (Figure 4, top)²⁹. However, a significantly greater proportion of people in the curcumin group experienced severe endoscopic recurrence at six months compared to the placebo group (p = 0.034)²⁹. For UC patients in remission, at six months there was a trend toward fewer cases of clinical relapse in the curcumin group compared to placebo; however, low event rates impacted precision of the estimate (Figure 4, top)³². At 12 months (i.e., six months after the treatment period ended), no significant differences in relapse rates were found³². In the same patients, at six months, the mean CAI and EI scores in the curcumin group slightly decreased from baseline, whereas those of the placebo group increased, resulting in a significant difference of differences in means, favouring the curcumin group (Figure 4, bottom)³².

Figure 4. Forest plot of the effects of curcumin vs placebo on relapse of ulcerative colitis or Crohn’s disease.

Effect estimates in forest plots (risk ratios, differences of differences, and standardized differences of differences) may not match those reported in the text due to use of raw data to generate plots and standardization of some estimates.

Abbreviations. AZA = azathioprine; CAI = Clinical Activity Index; CDAI = Crohn’s disease Activity Index; Diff = difference; EI = Endoscopic Index; MES = mesalamine; SE = standard error; Std diff = standardized difference; SZ = sulfasalazine; tx = treatment

3.3.3. Quality of Life

Quality of life outcomes were reported in three studies either using the “general well-being” domain of the SCCAI, reported as a five-level ordinal outcome³⁷, or the IBDQ²⁹ or IBDQ-9 scales³⁹ as continuous outcomes. Two studies evaluated curcumin extract administered at 1,500 mg/day^29,39 and one study evaluated a bio-enhanced curcumin intervention³⁷.

For curcumin extract at 1,500 mg/day, the change from baseline to eight weeks in mean IBDQ-9 scores for the curcumin group did not differ significantly from that of the placebo group, after adjusting for differences in baseline values (p = 0.08)³⁹. In another study, mean IBDQ scores at 24 weeks of treatment were not significantly different between curcumin and placebo groups (curcumin: 178.5 ± 25.8 vs placebo: 181.5 ± 24.7; p = 0.80)³⁹. For bio-enhanced curcumin, the distribution of well-being responses for the curcumin group was not significantly different from that of the placebo group after two weeks of treatment (p = 0.455); however, by the end of the four-week treatment period, well-being responses approached a statistically significant difference favouring the curcumin group (p = 0.050)³⁷.

3.4. Irritable Bowel Syndrome

Five RCTs focused on IBS^{17,18,26,30,31} and included a total of 551 patients (mean age: 40.5–50.3 years; proportion female: 59.4–70%). A variety of definitions of IBS were used as inclusion criteria across the studies, including ROME III criteria alone^17,26, ROME III criteria and the German IBS guidelines¹⁸, ROME II criteria³¹, and clinical diagnosis following an intensive clinical exam to rule out an organic cause³⁰. One study included only patients with diarrhea-predominant IBS¹⁸, one included 62% diarrhea-predominant and 38% constipation-predominant IBS patients¹⁷, and three studies did not report IBS type^26,30,31. Three studies reported baseline severity of IBS as moderate, measured by abdominal pain/discomfort Likert scales^17,18,26, with¹⁷ or without^18,26 the IBS severity scoring system (IBS-SSS). Baseline severity was not reported by the other two studies^30,31.

Three studies compared mixed-formula curcumin products (i.e., herbal products with several components of which one was curcumin) to placebo^17,18,26. The remaining two studies compared two doses of turmeric extract³¹ and C. xanthorrhiza extract to Fumaria officinalis (fumitory) and placebo³⁰, respectively. All curcumin doses in all trials were less than 1,000 mg/day and no products were reported to be bio-enhanced. No specific adjunct therapies were administered alongside the trial interventions; however, one study reported that concomitant treatments were allowed at stable and minimal doses throughout the study and that these treatments were comparable between groups¹⁷. One study required participants to maintain their diet for the entire study period¹⁷. Dietary intake of curcumin was not monitored.

3.4.1. Response to Treatment

All five studies that focused on IBS reported response to treatment outcomes^{17,18,26,30,31}. Treatment response definitions varied and included dichotomous outcomes (n = two studies^17,26, reporting six definitions, with no overlap), continuous outcomes (n = four studies^17,18,26,31, reporting two definitions, with some overlap), and an ordinal outcome (n = one study³⁰, reporting one definition). Treatment duration and longest follow-up time ranged from 4^17,18 to 18³⁰ weeks.

Individuals receiving mixed-formula curcumin products (i.e., Xanthofen^™ and curcumin-fennel essential oil) in two studies were significantly more likely to have beneficial responses to treatment compared to placebo for all response measures, after four¹⁷ and eight weeks of treatment²⁶, respectively (Figure 5). These findings were also confirmed in analyses of continuous data (IBS-SSS scores) in the same two studies^17,26 (significant difference of differences after eight weeks following adjustment for baseline values, p < 0.001²⁶; significant relative mean decreases and adjusted mean difference at four weeks, p < 0.001¹⁷). However, a third study of a mixed-formula curcumin product found no significant difference in mean IBS-SSS scores at four weeks of treatment compared to placebo (MD = 24.10 (−17.12 to 65.32), p = 0.26)¹⁸.

Figure 5. Forest plot of response to treatment for patients with IBS, curcumin vs placebo.

Effect estimates in forest plots (risk ratios, differences of differences, and standardized differences of differences) may not match those reported in the text due to use of raw data to generate plots and standardization of some estimates.

Abbreviations. Diff = difference; IBS = irritable bowel syndrome; IBS-GIS = IBS Global Improvement Scale; IBS-SSS = IBS Symptom Severity Scale; SE = standard error; Std diff = standardized difference; tx = treatment; VAS = Visual analogue scale

Studies of curcumin extracts^30,31 found no significant differences in response to treatment compared to placebo, Fumaria officinalis (fumitory), or between doses of curcumin. A comparison of the effects of C. xanthorrhiza extract, fumitory, and placebo, using an ordinal global assessment scale (i.e., improved, unchanged, worsened) demonstrated no significant differences in response distributions between the groups after 18 weeks of treatment (overall comparison of interventions: p = 0.96; curcumin vs placebo: p = 0.74; fumitory vs placebo: p = 0.85)³⁰. As well, there was no significant difference in the proportion of individuals whose IBS had subsided (according to Rome II criteria) after eight weeks of treatment with either lower-dose or higher-dose curcumin extract, although within both groups, significant proportions of individuals had their IBS subside from baseline to eight weeks (p < 0.001 in both groups)³¹.

3.4.2. Quality of Life

All five IBS studies reported outcomes related to quality of life (QoL), using either IBS-specific scales (e.g., IBS-QoL total score, IBS-SSS: interference with QoL domain)^17,18,26,31, a general QoL scale (i.e., EQ-5D)¹⁸, or improvement in psychosocial stress caused by IBS measured on a global assessment scale³⁰.

Curcumin extracts demonstrated no significant difference in measures of QoL compared to placebo or fumitory at 18 weeks (no differences in distributions of responses regarding improvement of psychosocial stress caused by IBS)³⁰, or between lower and higher doses after 8 weeks (67% vs 70% reporting some or definite improvement of IBS-QoL)³¹.

Compared to placebo, mixed-formula curcumin products had varied effects on the IBS-QoL scale in three studies (Figure 6)^17,18,26. The IBS-QoL total score improved from baseline to the end of treatment in the mixed-formula curcumin groups across all three studies, and these changes were significantly different in the mixed-formula curcumin groups compared to placebo in two studies^17,26 (Alt et al., 2017²⁶: MD_curcumin = 3.79 (2.93) vs MD_placebo = 2.19 (3.25), p = 0.008; Portincasa et al., 2016¹⁷: adjusted difference of differences = 9.19 [95% CI = 3.13 to 15.25]). There was no significant difference between groups in IBS-QoL total score after four weeks of treatment in the third study (Lauche et al. 2016¹⁸: MD_{curcumin-placebo} at four weeks = 0.25 [95% CI = −3.70 to 4.19], p = 0.90). As well, in this third study¹⁸, there was no significant difference between groups in the EQ-5D Index (MD_{curcumin-placebo} at four weeks = 0.02 [95% CI = −0.08 to 0.04], p = 0.48) or the EQ-5D VAS (MD_{curcumin-placebo} at four weeks = 2.5 [95% CI = −10.85 to 5.86], p = 0.57). One study found a significant effect of mixed-formula curcumin product compared to placebo on the Interference with QoL domain of the IBS-SSS after four weeks of treatment (adjusted difference of differences = −14.0 [95% CI = −21.2 to −6.8], p < 0.001).

Figure 6. Forest plot of the effects of curcumin vs placebo on the quality of life of people with IBS.

Figure footnote: Effect estimates in forest plots (differences of differences and standardized differences of differences) may not match those reported in the text due to use of raw data to generate plots and standardization of some estimates.

Abbreviations. Diff = difference; EQ-5D = a standardized measure of health-related quality of life; IBS = irritable bowel syndrome; IBS-QoL = Irritable Bowel Syndrome Quality of Life; IBS-SSS = Irritable Bowel Syndrome Symptom Severity Scale; QoL = quality of life; SE = standard error; Std diff = standardized difference; tx = treatment; VAS = Visual analogue scale

3.5. Dyspepsia

Five trials^{38,40,42,44,46} focused on dyspepsia and included a total of 511 patients (mean age: 22.7–56.4 years; proportion female: 48–92.6%). Three studies included patients with functional dyspepsia, defined using Rome criteria (III³⁸ or IV^40,46), and two studies did not use formal disease definitions but included patients with “acid, flatulent, or atonic dyspepsia⁴⁴”, or mild-to-moderate “uninvestigated” dyspepsia⁴².

Four studies evaluated curcumin extracts^40,42,44,46 at doses from 750 to 2,000 mg/day and one study evaluated a mixed-formula curcumin/piperine product³⁸. Curcumin interventions were compared to placebo (n = 3 studies)^38,44,46 and/or an active comparator (i.e., ranitidine⁴²; simethicone⁴⁰; omeprazole⁴⁶; or “Flatulence,” a combination herbal product⁴⁴). One study administered famotidine as an adjuvant treatment in both curcumin and placebo arms³⁸. Duration of therapy ranged from one to four weeks.

3.5.1. Response to Treatment

All five studies that focused on dyspepsia reported response to treatment outcomes^{38,40,42,44,46}. Treatment response definitions varied, with one study⁴⁴ reporting a dichotomous outcome (i.e., subjective improvement in symptoms) and four others reporting continuous outcomes^38,40,42,46. Two studies reported Severity of Dyspepsia Assessment (SODA) domains^42,46 but used different comparators.

Both a mixed-formula curcumin/piperine product and placebo demonstrated significant reductions in Hong Kong Dyspepsia Index (HKDI) in repeated measures testing from baseline to immediately after four weeks of treatment and to four weeks after the end of treatment (curcumin: p < 0.001; placebo: p < 0.033)³⁸. No comparison was made between groups. However, only the mixed-formula curcumin/piperine group experienced significant reductions in the prevalence of H. pylori over the study period (curcumin: p = 0.004; placebo: p = 0.126)³⁸. There were no significant differences between groups in the prevalence of H. pylori at either four (curcumin vs placebo: 20% vs 24%; p = 0.132) or eight weeks (curcumin vs placebo: 18% vs 23%; p = 0.095)³⁸.

Overall, curcumin extract was found to have significantly beneficial responses to treatment compared to placebo but not to active comparators. Compared to placebo, 2,000 mg/day curcumin extract was demonstrated to significantly reduce all SODA domains after four weeks of treatment (pain: p < 0.001; non-pain: p = 0.008; satisfaction: p < 0.001)⁴⁶ and to result in a significantly greater proportion of participants reporting subjective assessments of “improved” or “cured⁴⁴.” No significant differences were found in most SODA domains when compared to active comparators after four weeks (curcumin 2,000 mg/day vs omeprazole⁴⁶: p_pain = 0.302, p_non-pain = 0.486, p_satisfaction = 1.000; curcumin 2,000 mg/day vs ranitidine, per-protocol analyses⁴²: p_pain = 0.766, p_non-pain = 0.280); however, the SODA satisfaction domain of the curcumin extract group improved significantly more after four weeks than that of the ranitidine group (per protocol analysis: p_satisfaction = 0.037)⁴². Similarly, in a comparison of two dosages of curcumin extract and simethicone, there were significant changes in composite scores of a Global Overall Symptom scale from baseline to four weeks of treatment for all groups (within-group differences: p < 0.001 for all groups)⁴⁰; however, the differences of differences between the curcumin extract and simethicone groups were not significant (curcumin 750 mg/day vs simethicone: p = 0.123; curcumin 1,500 mg/day vs simethicone: p = 0.828)⁴⁰. Finally, the proportion of participants with subjective improvement after one week of treatment was not significantly different between 2,000 mg/day curcumin extract and a mixed-formula herbal product that did not contain curcumin (87% vs 83%; p = 0.685)⁴⁴.

3.5.2. Quality of Life

One dyspepsia study reported QoL outcomes⁴⁶. Compared to placebo, 2,000 mg/day curcumin extract significantly improved the EQ-VAS (p = 0.003) and trended to improve the EQ-5D index (p = 0.053) after four weeks of treatment⁴⁶. For the same QoL outcomes, no significant differences were found between curcumin extract and omeprazole (p_EQ-VAS = 0.602; p_{EQ-5D index} = 0.388) or placebo and omeprazole (p_EQ-VAS = 0.116; p_{EQ-5D index} = 1.000)⁴⁶.

3.6. Peptic, gastric, and duodenal ulcers

Four trials focused on gastrointestinal ulcers, either peptic^15,25, gastric³⁵, or duodenal⁴⁵ by definition and included a total of 291 patients (mean age: 35.1–50.5 years; proportion female: 22–65%). Study inclusion criteria for three studies required confirmation of the presence of ulcers by endoscopy^15,25,45. Both studies of peptic ulcers were recent publications (published in 2016 and 2017) and required detection of H. pylori for inclusion^15,25, whereas neither of the studies of duodenal (published in 1998)³⁵ or gastric (published in 1993)³⁵ ulcers did.

The two peptic ulcer studies compared H. pylori triple therapy (i.e., either esomeprazole²⁵ or pantoprazole¹⁵ + amoxicillin + clarithromycin) in combination with 1,500 mg/day curcumin extract²⁵ or a mixed-formula curcumin/piperine product¹⁵ to the same triple therapy without curcumin, while the gastric and duodenal ulcer studies compared curcumin extract (750³⁵ or 6,000⁴⁵ mg/day) to either an aluminium hydroxide/magnesium hydroxide antacid³⁵ or placebo⁴⁵. Duration of treatment ranged from two to eight weeks.

3.6.1. Response to Treatment

All four studies that focused on gastrointestinal ulcers reported response to treatment outcomes^15,25,35,45. Treatment response definitions varied, including the HKDI¹⁵; eradication of H. pylori, as indicated by the urea breath test¹⁵ or rapid antigen/immunoglobulin M antibody enzyme-linked immunosorbent assay (ELISA) tests²⁵; ulcer healing (complete or incomplete), as demonstrated endoscopically^35,45 or radiographically⁴⁵; and subjective symptom reduction from baseline⁴⁵.

In people with peptic ulcers, H. pylori triple therapy plus curcuminoids/piperine significantly reduced the HKDI score compared to triple therapy plus placebo (curcumin vs placebo: −12.90 ± 2.81 vs −9.60 ± 3.39; p < 0.001)¹⁵; however, there was no significant difference between groups in the proportion of participants who experienced H. pylori eradication after four weeks of treatment¹⁵. In another study, people with duodenal ulcers were significantly more likely to experience H. pylori eradication after two weeks of treatment with triple therapy plus 1,500 mg/day curcumin compared to triple therapy alone²⁵; however, there was no significant difference between the treatments for people with gastric ulcers²⁵.

In people with gastric ulcers, antacid significantly increased the probability of complete ulcer healing compared to 750 mg/day curcumin extract after twelve weeks of treatment (curcumin vs antacid: 70.6% vs 94.1%; p < 0.05)³⁵.

In people with duodenal ulcers, there was no significant difference between 6,000 mg/day curcumin extract and placebo groups in the proportion of people with endoscopic or radiologic evidence of ulcer healing after eight weeks of treatment (curcumin vs placebo: 25.4% vs 28.1%; p = 0.736)⁴⁵. Symptom scores reduced to approximately 24% and 32% of baseline for the curcumin extract and placebo groups, respectively (no statistical comparison made)⁴⁵.

3.7. Other gastrointestinal conditions

Three trials focused on either H. pylori infection²⁷, chronic gastritis with confirmed H. pylori infection³⁴, or self-reported gastrointestinal complaints³⁶ and included a total of 325 patients (mean age: 35.7–47.4; proportion female: 44.9–87.3%). H. pylori was confirmed by stool antigen⁴⁸ or by urease test, pathology, or culture³⁴.

The two studies focused on H. pylori infection compared a curcumin intervention—either a mixed-formula curcumin phytomedicine treatment⁴⁸ or 2,100 mg/day curcumin extract³⁴—to either H. pylori quadruple⁴⁸ or triple therapy³⁴ (i.e., omeprazole + amoxicillin + metronidazole ± bismuth). The study of gastrointestinal complaints compared 500 mg/day curcumin extract plus essential oils to placebo³⁶. Duration of treatment ranged from four to eight weeks.

3.7.1. Response to Treatment

All three studies that focused on other gastrointestinal complaints reported response to treatment outcomes^27,34,36. Treatment response definitions included eradication of H. pylori as indicated by either C-urea breath test⁴⁸ or by negative findings for H. pylori on all three of a urease test, pathology, and culture³⁴; two different gastrointestinal symptom scores^27,36; and the proportion of people with negative breath tests for small intestinal bacterial overgrowth (SIBO; hydrogen, methane, and both hydrogen and methane breath tests)³⁶.

In people with H. pylori, use of 2,100 mg/day curcumin extract was associated with a significantly lower likelihood of eradication of H. pylori compared to triple therapy after six weeks of treatment (curcumin vs triple therapy: 5.9% vs 78.9%; p < 0.001)³⁴. However, use of a mixed-formula curcumin phytomedicine treatment was demonstrated to have similar effectiveness to eradicate H. pylori compared to quadruple therapy after four weeks (curcumin vs quadruple therapy: 55.8% vs 62.2%; p > 0.05)⁴⁸. In the latter study, patients experienced significantly reduced overall gastrointestinal symptom scores from baseline to four weeks within the curcumin group, regardless of their H. pylori eradication status (curcumin—H. pylori eradicated: p < 0.0001; H. pylori not eradicated: p < 0.001)⁴⁸; however, in the quadruple therapy group, only those for whom H. pylori was eradicated experienced a significant reduction in overall gastrointestinal symptom score (quadruple therapy—H. pylori eradicated: p < 0.001; H. pylori not eradicated: p > 0.05)⁴⁸.

In people with self-reported gastrointestinal complaints, the mean Gastrointestinal Symptoms Rating Scale (GSRS) total score decreased significantly from baseline with repeated measures at four and eight weeks of treatment within both the 500 mg/day curcumin extract plus essential oils group and the placebo group (curcumin: p < 0.001; placebo: p < 0.001), and there was a significant difference between groups over time (curcumin vs placebo: p = 0.021)³⁶. However, there were no significant differences either within or between groups in the change in the proportion of patients with positive SIBO breath tests per protocol³⁶.

3.7.2. Quality of Life

Quality of life outcomes were reported in one study of self-reported gastrointestinal complaints, using domains of the SF-36 tool³⁶. No significant differences were found from baseline with repeated measures at four and eight weeks of treatment between the 500 mg/day curcumin extract plus essential oils group and the placebo group for any domain. Some significant changes over time within groups were found (see Dataset 1).

3.8. Safety

Nine studies reported data for at least one non-gastrointestinal AE^{15,17,18,27,30,32,41,42,44}. A summary of the reported AEs is provided in Appendix 5. No studies suggested significant differences in safety between curcumin, placebo, or active comparator groups, but two studies reported study withdrawal due to AEs in the turmeric group^16,31. Three other studies reported AEs or “side-effects” that were common symptoms of the underlying condition (e.g., burning sensation in stomach, diarrhea) or were indicators of non-response to treatment (e.g., exacerbation of UC symptoms necessitating corticosteroids)^16,31,45.

4. Discussion

This systematic review of the literature identified 26 unique studies that evaluated the efficacy and safety of curcumin compounds in the treatment of digestive disorders. Despite a moderate volume of evidence, this review was limited by certain aspects of the included primary studies, including variation in turmeric preparations, clinical and methodological heterogeneity, high risks of bias, small sample sizes, and limitations reporting. The limitations identified in this review reflect the less stringent regulations in the nutraceutical sector compared to pharmaceuticals⁴⁹. This lack of rigorous oversight may lead to variations in product quality and effectiveness, methodological inconsistencies in research, and a higher likelihood of bias and inadequate reporting in studies. This situation underscores the need for manufacturers of nutraceuticals, like turmeric, to enforce standards that parallel those in the pharmaceutical industry, ensuring rigorous testing for efficacy, safety, and quality of their products.

The wide range of available turmeric preparations and dosage forms makes it challenging to identify those that are best for gastrointestinal disorders. Curcuminoid extracts are not well absorbed when consumed and are insoluble in water. Numerous strategies have been developed to increase the bioavailability of turmeric and curcuminoids and prevent their metabolism so that they are available at higher levels and remain active longer in the body⁵⁰. These methods include using natural ingredients like black extract as well as coating curcumin in lipids, nanomicelles, or colloidal preparations⁵⁰. Due to the scarce details of the potency of these additives in commercially available turmeric formulations, we are unable to make conclusions about how they impact the therapeutic effect of turmeric. Additionally, studies evaluating turmeric’s efficacy in treating IBS or IBD patients commonly administered it alongside standard therapy. This concurrent use makes it difficult to isolate and understand the specific effects of turmeric alone on IBD and IBS symptoms.

Within each patient condition of interest, a high degree of clinical and methodological heterogeneity across the studies precluded meta-analysis. Descriptive summaries of the evidence demonstrated a mix of statistically significant and non-significant responses to treatment that generally favoured curcumin interventions over placebo in studies of IBD and IBS. For IBD, the patient condition with the greatest volume of evidence, magnitudes of effect for response to treatment with bio-enhanced curcumin or high-dose curcumin extract (e.g., > 1,000 mg/day) appeared to be greater than for lower doses of curcumin extract; however, strong inferences could not be drawn due to inconsistent outcome definitions and clinical and methodological heterogeneity across studies. Generally, for other gastrointestinal conditions included in our review, response to curcumin was inconsistent across patient populations, with occasional beneficial effects compared to placebo and no benefits compared to active comparators. Risk of bias was assessed to be high across the majority of the included studies and unclear across the remaining, further limiting the inferences that should be drawn and the reducing strength of the available evidence. The globally high or unclear potential for bias to have impacted the findings of the included studies was a substantial limitation of this review.

Earlier systematic reviews have also demonstrated limitations within the available evidence^20,51,52. A recent systematic review by Goulart et al. investigating the effects of curcumin on IBD identified challenges due to clinical and methodological heterogeneity that prevented quantitative synthesis of evidence from primary studies⁵¹. The same eight studies in the Goulart et al. review were included in our review^{16,28,32,33,37,39,41,43}, plus an additional study²⁹. A recent systematic review by Yin et al. ignored clinical and methodological heterogeneity of seven studies investigating curcumin use in UC and included meta-analyses²⁰; however, statistical heterogeneity was high (I² = 80–85%) across almost all outcomes, indicating that most of the observed variability in effects—which were highly dispersed—was real and there was likely no true common effect size across studies⁵³. Results of planned explorations of the heterogeneity were not reported, nor was high heterogeneity identified as a limitation of the meta-analyses²⁰. An earlier systematic review by Grammatikopoulou et al. that also investigated the effects of curcumin use in UC excluded the RCT by Banerjee et al.²⁸ from analyses due to several inconsistencies between the trial registry and final publication that caused the reviewers to question the rigour of the research⁵². In our review, methodological heterogeneity across the included studies, in part, was due to use of a diversity of outcome measures and follow-up times. Core outcome sets (COSs) have recently been developed for IBD in adults⁵⁴ and for functional abdominal pain disorders in children⁵⁵, with the aim to harmonize measurement and reporting of outcomes in clinical trials of these conditions. Of the nine included studies that focused on IBD, only a single study by Sugimoto et al.⁴³ reported at least one outcome measure from the COS at a follow-up time recommended for induction trials⁵⁴. In future studies, measuring and reporting core outcomes and using recommended cut-offs at recommended follow-up times would reduce methodological heterogeneity, thereby increasing the feasibility of meaningful meta-analyses.

Geographically, the sites of the included studies were distributed across 11 countries, which potentially added to clinical heterogeneity. Different countries and regions have different clinical practice guidelines for digestive disorders, as well as different criteria for insurance coverage, which may affect the adjuvant treatments available and inclusion criteria for patients. For example, Sugimoto et al. conducted a curcumin trial with CD patients in Japan⁴³. Patients were excluded if they used treatments not allowed or covered under the Japanese medical insurance system, such as > 3 g/day of mesalamine or some monoclonal antibodies. Other IBD studies may not have had such restrictions (e.g., Lang et al., 2015, conducted in Israel, Hong Kong, and Cyprus, in which patients must have received at least 4 mg/day mesalamine for inclusion¹⁶), adding to the clinical heterogeneity of the studies.

4.1. Limitations

Systematic reviews and meta-analyses are at the top of the hierarchy of clinical evidence⁵⁶; however, their ability to generate robust findings can be hindered by limitations with the available evidence. In addition to the limitations discussed above regarding high risk of bias and clinical and methodological heterogeneity, our review syntheses were limited by reporting and statistical issues within the included studies. The quality of reporting of RCTs of herbal interventions, generally, has been demonstrated to be low^57,58, despite the availability of reporting guidance for almost two decades⁵⁹. Sub-optimal reporting of methods and findings within the included studies prevented robust inferences from being made and adequate assessments of risk of bias from being conducted. Clinically, severity of disease for patient inclusion was poorly reported across studies, and details of the curcumin interventions deemed to be important by reporting guidance appeared to be reported rarely⁵⁹. Methodologically, despite all included studies being RCTs with a control group, not all studies reported between-group comparisons for key outcomes, making estimates of the relative efficacy of curcumin impossible. Additionally, rather than intention-to-treat analyses, several studies reported findings from per-protocol analyses that did not account for losses to follow-up, further calling into question the robustness of the evidence. Reporting of adverse events and safety data were inadequate across most studies; however, where reported, no significant differences in adverse events were found between curcumin and placebo or active comparators. We also found that some studies did not use appropriate statistical methods to analyse their data (e.g., dichotomous data reported as proportions were analysed using continuous data methods). Finally, small sample sizes and zero cells limited the ability of many studies to identify significant differences between treatment groups.

5. Conclusions

Although a moderate volume of evidence is available investigating the efficacy and safety of curcumin use for digestive disorders, these findings should be interpreted with caution, given the low quality of the evidence base due to clinical and methodological heterogeneity, high risks of bias, poor reporting, and statistical issues. However, turmeric appears to be safe; it therefore could be used as a nutraceutical option or add-on therapy in the treatment of digestive disorders, particularly in patients with IBD or IBS. Future trials should include larger samples, use study designs and statistical methods that minimize the risks of bias, and use outcome sets (for IBD) to improve the available evidence, facilitating more meaningful comparisons and robust conclusions.

Availability of data and materials

Extracted data is available as part of supplementary materials.