Interventions for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Summary of findings 1. Low FODMAP diet compared to sham diet for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Low FODMAP diet compared to sham diet for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with inflammatory bowel disease Setting: multicentre, 2 gastroenterology clinics in the UK and an unstated setting in Italy Intervention: low FODMAP diet Comparison: sham diet
Outcomes	Risk with sham diet	Risk with low FODMAP diet	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Note measured
Pain frequency IBD (measured in days of pain on the IBS‐SSS questionnaire)	‐	MD 2 lower (15.86 lower to 11.86 higher)	‐	52 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Pain frequency IBD (measured in days with moderate or severe pain on the GSRS questionnaire)	‐	MD 0.4 higher (0.44 lower to 1.24 higher)	‐	52 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Pain frequency CD (measured in days of pain on the IBS‐SSS questionnaire)	‐	MD 12 lower (114.55 lower to 90.55 higher)	‐	26 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Pain intensity IBD (0‐10cm visual analogue scale)	‐	MD 8.46 lower (15.76 lower to 1.16 lower)	‐	82 (2 studies)	⊕⊝⊝⊝ very low^{a b}
Pain intensity IBD (0‐3 point GSRS scale)	‐	MD 8 lower (66.27 lower to 50.27 higher)	‐	26 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Pain intensity CD (0‐10cm visual analogue scale)	‐	MD 0.2 higher (8.67 lower to 9.07 higher)	‐	52 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Withdrawal due to adverse events	4 per 1000	7.4 per 1000 (1 to 77)	RR 1.85 (0.18 to 19.19)	52 (1 study)	⊕⊝⊝⊝ very low^{a b}
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CD: Crohn's disease; CI: confidence interval; FODMAP: fermentable oligo‐, di‐, monosaccharides and polyols; GSRS: Gastrointestinal Symptom Rating Scale; IBD: inflammatory bowel disease; IBS‐SSS: Irritable Bowel Syndrome Severity Scoring System; MD: mean difference; RR: risk ratio
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded one level due to risk of bias. ^bDowngraded two levels due to imprecision from very sparse data.

Summary of findings 2. Medicine‐separated moxibustion combined with acupuncture compared with wheat bran‐separated moxibustion combined with shallow acupuncture for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Medicine‐separated moxibustion combined with acupuncture compared with wheat bran‐separated moxibustion combined with shallow acupuncture for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with Crohn's disease Settings: Shanghai Acupuncture Meridian Institute Medical Clinic Acupuncture Inflammatory Bowel Disease Specialist Clinic, Zhongshan Hospital Affiliated Endoscopic Center, and Yueyang Hospital, Shanghai University of Traditional Chinese Medicine Intervention: medicine‐separated moxibustion combined with acupuncture Comparison: wheat bran‐separated moxibustion combined with shallow acupuncture
Outcomes	Risk with wheat bran‐separated moxibustion combined with shallow acupuncture	Risk with medicine‐separated moxibustion combined with acupuncture	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Abdominal pain frequency or change in frequency	‐	‐	‐	‐	Not measured
Abdominal pain intensity or change in intensity	‐	‐	‐	‐	Not measured
Withdrawals due to adverse events	0 per 1000	0 per 1000 (0 to 0)	Not estimable	102 (1 study)	⊕⊝⊝⊝ very low ^{a b}
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded two levels due to imprecision from very sparse data. ^bDowngraded one level due to risk of bias.

Summary of findings 3. Mindfulness with cognitive behavioural therapy (CBT) compared with no treatment (both groups received standard medical therapy) for the management of abdominal pain in Crohn's Disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Mindfulness with cognitive behavioural therapy (CBT) versus no treatment (both groups received standard medical therapy) for the management of abdominal pain in Crohn's Disease and inflammatory bowel disease
Patient or population: people with inflammatory bowel disease Setting: multicentre, hospitals in the UK Intervention: mindfulness with CBT Comparison: no treatment
Outcomes	Risk with standard medical therapy	Risk with mindfulness with CBT + standard medical therapy	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Pain frequency and intensity (measured as part of the 0‐500 point IBS‐SSS questionnaire)	‐	MD 37 lower (87.29 lower to 13.29 higher)	‐	66 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Withdrawal due to adverse events	0 per 1000	0 per 1000 (0 to 0)	Not estimable	66 (1 study)	⊕⊝⊝⊝ very low ^{a b}
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CBT: cognitive behavioural therapy; CI: confidence interval; MD: mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded two levels due to imprecision from very sparse data. ^bDowngraded one level due to risk of bias.

Summary of findings 4. soft non‐manipulative osteopathic treatment compared to no intervention for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Soft non‐manipulative osteopathic treatment compared to no intervention for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with Crohn's disease Setting: single centre, hospital in Spain Intervention: soft non‐manipulative osteopathic Comparison: no intervention
Outcomes	Risk with no intervention	Risk with soft non‐manipulative osteopathic	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Abdominal pain frequency or change in frequency	‐	‐	‐	‐	Not measured
Pain intensity (0‐10cm visual analogue scale)	‐	MD 0.01 higher (1.81 lower to 1.83 higher)	‐	30 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Withdrawals due to adverse events	0 per 1000	0 per 1000 (0 to 0)	Not estimable	30 (1 study)	⊕⊝⊝⊝ very low ^{a b}
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded two levels due to imprecision from very sparse data. ^bDowngraded one level due to risk of bias.

Summary of findings 5. Directed stress management (it is unclear whether these interventions replaced standard treatment or were added to standard treatment) versus standard treatment for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Stress management (it is unclear whether these interventions replaced standard treatment or were added to standard treatment) versus standard treatment for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with Crohn's disease Setting: single centre, Inflammatory Intestinal Disease Unit of Asturias Central Hospital, Spain Intervention: stress management Comparison: no stress management
Outcomes	Risk with no stress management	Risk with stress management	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Pain frequency or change in pain frequency	‐	‐	‐	‐	Not measured
Pain intensity (author‐derived formula based on a 1‐3 point scale)	‐	MD 34.3 lower (61.99 lower to 6.61 lower)	‐	30 (1 study)	⊕⊝⊝⊝ very low^{a b}
Withdrawals due to adverse events	‐	‐	‐	‐	Not measured
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded two levels due to imprecision from very sparse data. ^bDowngraded one level due to risk of bias.

Summary of findings 6. Self‐directed stress management (it is unclear whether these interventions replaced standard treatment or were added to standard treatment) versus standard treatment for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Self‐directed stress management (it is unclear whether these interventions replaced standard treatment or were added to standard treatment) versus standard treatment for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with Crohn's disease Setting: single centre, Inflammatory Intestinal Disease Unit of Asturias Central Hospital, Spain Intervention: stress management Comparison: no stress management
Outcomes	Risk with no stress management	Risk with stress management	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Pain frequency or change in pain frequency	‐	‐	‐	‐	Not measured
Pain intensity (author‐derived formula based on a 1‐3 point scale)	‐	MD 30.5 lower (58.45 lower to 2.55 lower)	‐	30 (1 study)	⊕⊝⊝⊝ very low^{a b}
Withdrawals due to adverse events	‐	‐	‐	‐	Not measured
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded two levels due to imprecision from very sparse data. ^bDowngraded one level due to risk of bias.

Summary of findings 7. Enteric‐release glyceryl trinitrate compared to placebo for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Enteric‐release glyceryl trinitrate compared to placebo for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with Crohn's disease Setting: unstated (centres in the UK) Intervention: enteric‐release glyceryl trinitrate Comparison: placebo
Outcomes	Risk with placebo	Risk with enteric‐release glyceryl trinitrate	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Treatment success as defined by the authors	‐	‐	‐	‐	‐	Not measured
Abdominal pain frequency or change in frequency	‐	‐	‐	‐	‐	Not measured
Abdominal pain intensity or change in intensity	‐	‐	‐	‐	‐	Not measured
Withdrawal due to adverse events	Study population		RR 3.18 (0.94 to 10.76)	70 (1 study)	⊕⊝⊝⊝ very low ^{a b}	IG: Headache = 1, worsening clinical condition = 4, generalised rash = 1, mood change/irritability = 1, loss of consciousness/memory = 1 CG: Headache = 2, worsening clinical condition = 1
Withdrawal due to adverse events	83 per 1000	265 per 1000 (78 to 897)	RR 3.18 (0.94 to 10.76)	70 (1 study)	⊕⊝⊝⊝ very low ^{a b}
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded one level due to risk of bias. ^bDowngraded two levels due to imprecision from very sparse data. IG: Intervention group CG: Control group

Summary of findings 8. 100 mg olorinab 3 times/day compared to 25 mg olorinab 3 times/day for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
100 mg olorinab 3 times/day compared to 25 mg olorinab 3 times/day for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with Crohn's disease Setting: unstated (multicentre, USA) Intervention: 100 mg olorinab 3 times/day Comparison: 25 mg olorinab 3 times/day
Outcomes	Risk with 25 mg olorinab 3 times/day	Risk with 100 mg olorinab 3 times/day	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Pain frequency or change in pain frequency	‐	‐	‐	‐	Not measured
Pain intensity (30% reduction in weekly AAPS)	Study population		RR 0.66 (0.38 to 1.15)	14 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Pain intensity (30% reduction in weekly AAPS)	1000 per 1000	660 per 1000 (380 to 1000)	RR 0.66 (0.38 to 1.15)	14 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Withdrawal due to adverse events	0 per 1000	0 per 1000 (0 to 0)	Not estimable	14 (1 study)	⊕⊝⊝⊝ very low ^{a b}
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded one level due to risk of bias. ^bDowngraded two levels due to imprecision from very sparse data.

Summary of findings 9. Relaxation training compared to waitlist for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Relaxation training compared to waitlist for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with inflammatory bowel disease Setting: Hadassah Medical Center in Jerusalem Intervention: relaxation training Comparison: waitlist
Outcomes	Risk with waitlist	Risk with relaxation training	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Pain frequency or change in pain frequency	‐	‐	‐	‐	Not measured
Pain intensity (0‐10cm visual analogue scale)	‐	MD 0.72 lower (1.85 lower to 0.41 higher)	‐	56 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Withdrawals due to adverse effects	‐	‐	‐	‐	Not measured
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded one level due to risk of bias. ^bDowngraded two levels due to imprecision from very sparse data.

Summary of findings 10. Web‐based education compared to standard book‐based education for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Web‐based education compared to standard book‐based education for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with inflammatory bowel disease Setting: single centre, gastroenterology unit in Turkey Intervention: web‐based education Comparison: standard book‐based education
Outcomes	Risk with standard book‐based education	Risk with web‐based education	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Abdominal pain frequency or change in frequency	‐	‐	‐	‐	Not measured
Pain intensity (0‐10cm visual analogue scale)	‐	MD 0.13 lower (1.25 lower to 0.99 higher)	‐	60 (1 study)	⊕⊕⊝⊝ very low ^{a b}
Withdrawals due to adverse events	0 per 1000	0 per 1000 (0 to 0)	Not estimable	60 (1 study)	⊕⊝⊝⊝ very low^{a b}
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded one level due to risk of bias. ^bDowngraded two levels due to imprecision from very sparse data.

Summary of findings 11. Yoga intervention compared to no treatment (both groups received standard medical therapy) for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Yoga intervention compared to no treatment (both groups received standard medical therapy) for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with inflammatory bowel disease Settings: single centre, All India Institute of Medical Science (AIIMS), New Delhi, India Intervention: yoga Comparison: no yoga
Outcomes	Risk with standard medical therapy	Risk with yoga plus standard medical therapy	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Abdominal pain frequency or change in frequency	‐	‐	‐	‐	Not measured
Abdominal pain intensity or change in intensity	‐	‐	‐	‐	Not measured
Withdrawal due to adverse events	0 per 1000	0 per 1000 (0 to 0)	Not estimable	100 (1 study)	⊕⊝⊝⊝ very low ^{a b}
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded one level due to risk of bias. ^bDowngraded two levels due to imprecision from very sparse data.

Summary of findings 12. Transcranial direct current stimulation compared to sham stimulation for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Transcranial direct current stimulation compared to sham stimulation for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with inflammatory bowel disease Setting: single centre, Medical Department I (Gastroenterology, Infectious Diseases, Rheumatology) of the Charite‐Campus Benjamin Franklin, Germany Intervention: transcranial direct current stimulation Comparison: sham stimulation
Outcomes	Risk with sham stimulation	Risk with transcranial direct current stimulation	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Abdominal pain frequency or change in frequency	‐	‐	‐	‐	Not measured
Pain intensity (0‐10cm visual analogue scale)	‐	MD 1.65 lower (3.29 lower to 0.01 lower)	‐	20 (1 study)	⊕⊕⊝⊝ low ^a
Withdrawal due to adverse events	0 per 1000	0 per 1000 (0 to 0)	Not estimable	20 (1 study)	⊕⊕⊝⊝ low ^a
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference; PPT: Pressure Pain Threshold
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded two levels due to imprecision from very sparse data.

Summary of findings 13. Kefir ( Lactobacillus bacteria) compared to no intervention for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
*Kefir (Lactobacillus* bacteria) compared to no intervention for the management of abdominal pain in Crohn's disease and inflammatory bowel disease**
Patient or population: people with inflammatory bowel disease Setting: unstated (single centre, Turkey) Intervention: kefir Comparison: no intervention
Outcomes	Risk with no intervention	Risk with kefir	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Abdominal pain frequency or change in frequency	‐	‐	‐	‐	Not measured
Pain intensity IBD (4‐point rating scale from 0 to 3)	‐	MD 0.62 higher (0.17 higher to 1.07 higher)	‐	48 (1 study)	⊕⊕⊝⊝ very low ^{a b}
Pain intensity CD (4‐point rating scale from 0 to 3)	‐	MD 1.10 lower (1.67 lower to 0.53 lower)	‐	20 (1 study)	⊕⊕⊝⊝ very low ^{a b}
Withdrawals due to adverse events	0 per 1000	0 per 1000 (0 to 0)	Not estimable	20 (1 study)	⊕⊝⊝⊝ very low ^{a b}
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CD: Crohn's disease; CI: confidence interval; IBD: inflammatory bowel disease; MD: mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded one level due to risk of bias. ^bDowngraded two levels due to imprecision from very sparse data.

Background

Description of the condition

Inflammatory Bowel Disease (IBD) is a term for a group of disorders characterised by inflammation of the gut. The two main types of IBD are Crohn's Disease (CD) and Ulcerative Colitis (UC), which have several subtypes based on onset, disease location and clinical behaviour for CD, and extent for UC. There also exists a categorisation for the much rarer condition indeterminate colitis (Satsangi 2006).

CD is a remitting and relapsing disease of the gastrointestinal tract that affects over 3.5 million young people and adults in the USA and Europe (Kaplan 2015), and more than 6.8 million globally (Alatab 2020). Active CD symptoms include abdominal pain, fatigue, weight loss, and diarrhoea. There is no known cure; however, the disease can be managed via lifelong treatment and care, and therefore places a huge financial burden on healthcare systems. The annual care costs of inflammatory bowel disease (IBD) to the National Health Service (NHS) was estimated at over GBP 1000 million in 2010 (RCP 2012). For CD alone, this amounts to over GBP 6000 per patient annually (Ghosh 2015). Treatment of the disease may involve surgical intervention or immunosuppression using thiopurines and anti‐tumour necrosis factor (anti‐TNF) medications (Gjuladin‐Hellon 2019). These interventions aim to induce remission, maintain remission, and manage symptoms (Greenley 2013).

Abdominal pain is a common and debilitating symptom of CD and other IBDs, which is multifaceted, with multiple causes and contributors, such as a symptom of disease relapse, an adverse effect of medication, surgical complication, or due to problems related to CD itself, such as strictures or adhesions secondary to IBD (Srinath 2012). The pain may also vary between adults and children and may also be influenced by disease activity. Even in the absence of the aforementioned factors, around 20% to 50% of people with CD in remission still experience pain (Bielefeldt 2009). This has been attributed to functional abdominal pain disorders (FAPD) such as irritable bowel syndrome (IBS), abdominal migraine, and functional dyspepsia (Odes 2017), although the definition of such disorders involves explicit exclusion of pathology such as IBD. Evidence is lacking to indicate whether there is a specific variant of functional pain coexisting within people with IBD, or a separate pain disorder that can be attributed to IBD pathologic mechanisms. The aetiology and management of abdominal pain in CD may therefore vary in ways that cannot be fully explained. A common suggestion is that the inflammation in the intestinal wall can lead to the perception of abdominal pain (Docherty 2011).

Description of the intervention

Pharmacological interventions

Medication for CD can reduce inflammation and associated pain by inducing remission. Where pain persists in the absence of inflammation, it has been managed with a variety of agents, including pain‐relieving medication such as antispasmodics, non‐steroidal anti‐inflammatory drugs (NSAIDs), laxatives, antidepressants, antiemetic agents, cyclo‐oxygenase‐2 (COX‐2) inhibitors, and psychoactive drugs such as cannabis and opioids (Srinath 2012). Due to the potential adverse effects of some of these drugs, short‐term use is advised.

Non‐pharmacological interventions

Non‐pharmacological interventions used in managing pain may include dietary, psychological, lifestyle advice, and alternative medicine. These interventions are considered by some as less invasive and may be used as adjuvant treatment. Cognitive behavioural therapy (CBT), stress management, and coping skills training are the most commonly used psychological interventions. These therapies can be very heterogeneous, therefore it is key to consider the specific evidence and conceptual alignment of the approach delivered to understand 'what' the therapy was, as well as 'whether' it was effective. Alternative treatments such as acupuncture and transcutaneous electrical nerve stimulation, which have been used in other conditions such as IBS, are increasingly used in people with IBD, albeit based on limited evidence (Srinath 2012). Dietary interventions studied include the avoidance of FODMAP (fermentable oligo‐, di‐, monosaccharides and polyols) and the use of supplements with prebiotic properties; however, the evidence on their effectiveness appears to be weak and conflicting (Norton 2017).

Whilst some interventions such as neurochemicals and acupuncture have mostly been used in conditions such as IBS, others such as COX‐2 inhibitors have been associated with little to no effect in IBD (Paiotti 2012).

How the intervention might work

The mechanism of action of different interventions depends on the nature or cause of the abdominal pain.

Antispasmodics suppress intestinal spasms which cause pain from inflammation or obstruction (Srinath 2012). Pain related to strictures can be improved by following a low‐residue diet, which can pass through with ease, thereby preventing intestinal pain (Srinath 2012). A low FODMAP diet also aims to limit the intake of non‐absorbable nutrients, and has recently gained considerable attention in the management of IBS (Prince 2016).

Psychological techniques such as CBT, mindfulness, and stress management tend to help people with CD change negative behaviours that might be worsening their pain and provide coping mechanisms (Norton 2017). Yoga‐based programmes are thought to work by improving depression and anxiety (Ewais 2019). Other complementary or alternative therapies, such as the use of herbal and dietary supplements, traditional Chinese practices, and mind‐body techniques, have been proposed to have anti‐inflammatory, stress‐reducing, or other therapeutic modes of action (Lin 2018).

Why it is important to do this review

Abdominal pain in people with CD can lead to depressive symptoms, reduced quality of life, and an increase in the use of healthcare facilities (Srinath 2012). Effective pain management is therefore vital.

There are concerns regarding the safety of pharmacological interventions, such as the relative inefficacy of currently available analgesics and their potential toxicity. Opioids can offer short‐term relief; however, they are associated with such problems as narcotic bowel syndrome and other symptoms like constipation (Thapa 2019). Furthermore, there may be concern amongst people with IBD about the stigma of addiction associated with the use of opioids. The use of opioids in chronic pain can lead to people exhibiting withdrawal symptoms that are similar to CD symptoms (Pauly 2017), which can further complicate treatment. There are also concerns about the use of NSAIDs, which can have effects that mimic or potentially exacerbate CD activity (Long 2016).

Pain management has been highlighted as a priority topic for research by IBD patient groups and charities, but is currently not covered in the National Institute for Health and Care Excellence guidelines (NICE 2019), European Crohn's and Colitis Organisation guidelines (ECCO 2010), or Crohn's and Colitis Foundation guidance. Whilst several non‐Cochrane systematic reviews have assessed interventions for pain management in IBD, none has currently assessed the efficacy and safety of these interventions in Crohn's disease. Although this review covers interventions that have been previously assessed in published Cochrane Reviews in the group portfolio (Iheozor‐Ejiofor 2019; Kafil 2018; Limketkai 2019; Timmer 2011), the focus of this review was only on studies that have been conducted for the purpose of providing relief for abdominal pain.

Objectives

To assess the efficacy and safety of interventions for managing abdominal pain in people with Crohn's disease and IBD (where data on ulcerative colitis and Crohn's disease could not be separated).

Methods

Criteria for considering studies for this review

Types of studies

All published, unpublished, and ongoing randomised controlled trials (RCTs) that compared interventions for the management of abdominal pain in the setting of CD and IBD, with other active interventions or standard therapy, placebo, or no therapy. We excluded studies that did not report on any abdominal pain outcomes.

Types of participants

Adults and children with Crohn's disease or IBD who are experiencing abdominal pain. If studies included participants with ulcerative colitis as well as those with CD, these studies were included, and separate data sought for analysis. Where separate data could not be obtained, IBD patients were included as a whole.

Types of interventions

Pain‐relieving drugs such as antispasmodics, antidepressants, laxatives, antidiarrhoeal agents, antibiotics, analgesics, antireflux agents, antiemetic agents, antimigraine agents, antihistaminic agents, serotonergic agents, and psychoactive drugs.
Behaviour therapy, e.g. cognitive behavioural therapy (CBT), hypnotherapy.
Lifestyle advice, e.g. advice on physical activity including exercise.
Dietary interventions such as reduced intake of FODMAP; additional fibre intake; decrease in gas‐producing foods; extra fluid intake; lactulose‐, gluten‐, and histamine‐free diet.
Pre‐ and probiotics.
Other alternative therapies, e.g. acupuncture, homeopathy, body‐oriented therapy, musculoskeletal therapy (osteopathy/chiropractic), yoga.

Types of outcome measures

Both dichotomous and continuous outcomes were valid for inclusion.

Primary outcomes

Treatment success as defined by the authors.
Abdominal pain frequency or change in frequency of pain using any validated scale.
Abdominal pain intensity or change in pain intensity using any validated scale.
Withdrawal due to adverse events.

Secondary outcomes

Anxiety/depression using any validated scale.
Adverse events (total number of participants with any event).
Serious adverse events (as defined by the authors within the primary study)

Search methods for identification of studies

Electronic searches

We searched the following sources from the inception of each database to the date of search on 29 April 2021:

the Cochrane Central Register of Controlled Trials (CENTRAL) (via Ovid EBMR) (inception to Issue 03, 2021);
MEDLINE (via Ovid) (1946 to 29 April 2021);
PsycINFO (via Ovid) (1987 to 29 April 2021);
AMED (via Ovid) (Allied and Complementary Medicine) (1985 to 29 April 2021);
CINAHL (via EBSCO) (Cumulative Index to Nursing and Allied Health Literature) (1984 to 29 April 2021).

We also searched the following trial registers on 29 April 2021:

US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (www.clinicaltrials.gov);
World Health Organization International Clinical Trials Registry Platform (WHO ICTRP) (www.who.int/trialsearch/).

We placed no restrictions on language of publication. For detailed search strategies, see Appendix 1.

Searching other resources

As complementary search methods, we carefully checked relevant systematic reviews for potentially eligible studies. We also scrutinised the references of included studies. We sought unpublished trials by contacting experts in the field, and scanned the Internet and abstracts submitted to major international congresses from the three years prior to the search to capture any studies presented but not yet published in full.

In the case of foreign language papers, we planned to obtain translations of papers if necessary.

Data collection and analysis

We carried out data collection and analysis according to the methods recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2021).

Selection of studies

Five review authors independently screened the titles and abstracts identified by the literature search, excluding studies that based on title and abstract did not meet our inclusion criteria. We obtained the full reports of studies deemed potentially eligible. Five review authors independently assessed the full texts for inclusion in the review. Any disagreements were resolved by discussion or by consulting another review author if necessary. We recorded the studies excluded at this or subsequent stages, and the main reason for their exclusion, in the 'Characteristics of excluded studies' tables.

Where there were multiple publications for a given study, we collated the reports of the same study so that each study, rather than each report, was the unit of interest in the review; such studies have a single identifier with multiple references.

Studies that had the primary goal of inducing or maintaining remission in IBD were excluded, regardless as to whether they reported pain outcomes. This reflected the fact that any such pain imporvement was as a result of disease state and not an indepedent pain intervention and all such interventions are considered in seperate reviews. Such studies were excluded at the title and abstract screening stage. Studies that did not report any abdominal pain related outcomes, were excluded at the full text screening stage, and only after contact with the authors whenever the reporting of pain related outcomes was unclear.

Data extraction and management

Five review authors independently performed data extraction using piloted data extraction forms. We extracted the following data from the included studies:

trial setting: country and number of trial centres;
methods: study design, total study duration and date;
participant characteristics: age, sociodemographics, ethnicity, diagnostic criteria, pain location, and total number of participants;
eligibility criteria: inclusion and exclusion criteria;
intervention and comparator;
outcomes: outcome definition, unit of measurement, and time of collection;
results: number of participants allocated to each group, missing participants, and sample size;
funding source.

All treatment arms are described in the 'Characteristics of included studies' tables.

Assessment of risk of bias in included studies

Five review authors independently assessed risk of bias in the included studies based on the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2021). We assessed the following 'Risk of bias' domains:

sequence generation (selection bias);
allocation concealment (selection bias);
blinding of participants and personnel (performance bias);
blinding of outcome assessment (detection bias);
incomplete outcome data (attrition bias);
selective reporting (reporting bias);
other bias such as imbalance in participants' baseline characteristics.

We judged the studies to be at low, high, or unclear risk of bias for each domain assessed, based on the guidance in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2021).

After data extraction, the five review authors compared the extracted data, discussing and resolving any discrepancies before transfer of data into the 'Characteristics of included studies' tables.

Measures of treatment effect

We expressed treatment effect as risk ratios (RR) with corresponding 95% confidence intervals (CI) for dichotomous outcomes, and mean difference (MD) with 95% CI for continuous outcomes. Where endpoint and change score were both reported, we used endpoint scores for data analysis. However, if the studies assessed the same continuous outcome in different ways, we estimated the treatment effect using the standardised mean difference (SMD) (Cohen 1988).

Unit of analysis issues

The unit of analysis was the participant. For studies comparing more than two intervention groups, we planned to make multiple pair‐wise comparisons between all possible pairs of intervention groups. To avoid double counting, we would divide shared intervention groups evenly amongst the comparisons. For dichotomous outcomes, we planned to divide both the number of events and the total number of participants. For continuous outcomes, we would only divide the total number of participants, and leave the means and standard deviations (SDs) unchanged. We planned to include cross‐over studies for quantitative analysis only if data were separately reported before and after cross‐over, and use only pre‐cross‐over data. We did not anticipate finding any cluster‐RCTs; we would only use study data from such trials if the authors employed appropriate statistical methods in taking the clustering effect into account. We would also exclude cluster‐RCTs in a sensitivity analysis to assess their impact on the results.

Dealing with missing data

We contacted study authors in the case of missing data or studies that did not report data in sufficient detail. We attempted to estimate missing SDs using relevant statistical tools and calculators available in Review Manager 5 if studies reported standard errors (Review Manager 2020). Studies that failed to report measures of variance were judged as at high risk of reporting bias.

Assessment of heterogeneity

We assessed the included studies to determine their homogeneity in terms of participants, intervention, comparator, and outcome. To test for statistical heterogeneity, we employed a Chi² test using a P value of less than 0.1 to give an indication of the presence of heterogeneity. Inconsistency was quantified and represented by the I² statistic. We interpreted the thresholds as follows (Higgins 2021):

0% to 40%: might not be important;
30% to 60%: may represent moderate heterogeneity;
50% to 90%; may represent substantial heterogeneity;
75% to 100%: considerable heterogeneity.

Assessment of reporting biases

Most reporting biases were minimised by using an inclusive search strategy. We planned to investigate publication bias using a funnel plot if there were 10 or more studies. The magnitude of publication bias would be determined by visual inspection of the asymmetry of the funnel plot. In addition, we would test funnel plot asymmetry by performing a linear regression of intervention effect estimate against its standard error, weighted by the inverse of the variance of the intervention effect estimate (Egger 1997).

Data synthesis

To summarise the study characteristics, we conducted a narrative synthesis of all the included studies. We then carried out a meta‐analysis if two or more studies assessed similar populations, interventions, and outcomes. We planned to analyse studies of children, adults, and different sub‐intervention types separately. We used Review Manager 5 (Review Manager 2020). We synthesised study data using the random‐effects model if there was statistical heterogeneity (I² > 0%); otherwise, we used the fixed‐effect model. We combined effect estimates of studies that reported data in a similar way in the meta‐analysis. We pooled RRs for dichotomous outcomes, and MDs or SMDs for continuous outcomes, alongside 95% CIs. Where we were unable to carry out a meta‐analysis (e.g. due to lack of uniformity in data reporting), we presented a narrative summary of the included studies.

Subgroup analysis and investigation of heterogeneity

If we identified heterogeneity, we investigated possible causes and addressed them using the methods described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2021). We planned to undertake subgroup analyses of potential effect modifiers if sufficient data were available. We identified several potential modifiers of effect:

disease activity (active versus inactive disease);
pain location;
disease location.

Sensitivity analysis

We planned to undertake a sensitivity analysis on the primary outcome of treatment success in order to assess whether the findings of the review were robust to decisions made during the review process. In particular, we planned to exclude studies at high or unclear risk of bias from analyses. Where data analyses included studies with reported and estimated SDs, we excluded studies with estimated SDs to assess whether this affected the findings of the review. We investigated whether the choice of model (fixed‐effect versus random‐effects) affected the results.

Summary of findings and assessment of the certainty of the evidence

We have presented our primary outcomes results in 'Summary of findings' tables. Each comparison and primary outcome was exported to GRADEpro GDT software for quality assessment (GRADEpro GDT). Based on risk of bias, inconsistency, imprecision, indirectness, and publication bias, we graded the quality of the evidence for each outcome as high, moderate, low, or very low. These ratings have been defined as follows:

high: further research is very unlikely to change our confidence in the estimate of effect;
moderate: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate;
low: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate;
very low: any estimate of effect is very uncertain.

We justified all decisions to downgrade the quality of studies using footnotes, and made comments to aid the reader's understanding of the review where necessary.

Results

Description of studies

Information on the results of the search, included and excluded studies, and 'Risk of bias' assessment is provided below.

Results of the search

We completed our literature search on 29 April 2021 (Appendix 1), identifying a total of 3654 records through database searching and 13 additional records from alternative sources. After removal of duplicates, 3282 unique records remained. Title and abstract screening revealed 63 records for full‐text review. After assessing all 63 records, we identified 22 records of 14 studies that met the inclusion criteria and were included in the review. We also identified 9 records of 9 ongoing studies and 24 records of 19 studies awaiting classification. We excluded 8 records of 8 studies for various reasons (see Characteristics of excluded studies). The results of the search are presented in a PRISMA flow diagram (Figure 1). There are two fully published RCTs (Bao 2021; Lee 2021) which are under awaiting classification and not included in our results and meta‐analyses, as they were identified during our updated search. They will be included in future updated of this review.

Figure 1

Study flow diagram.

Included studies

Setting

Fourteen RCTs involving a total of 743 participants met our inclusion criteria. One study was conducted in China (Bao 2016), three in the UK (Berrill 2014; Cox 2020; Hawkes 2001), two in Spain (Espi Lopez 2018; Garcia‐Vega 2004), one in the USA (Higgins 2019), two in Turkey (Ozgursoy Uran 2019; Yilmaz 2019), one in India (Sharma 2015), one in Israel (Mizrahi 2012), two in Italy (Tapete 2018; Tapete 2019), and one in Germany (Volz 2016). All of the included studies were conducted in hospitals, medical centres, and gastroenterology units, except Sharma 2015, which was conducted in an institute of medical science, and three studies for which no information was provided about setting (Tapete 2018; Tapete 2019; Yilmaz 2019). Seven studies were single‐centre (Espi Lopez 2018; Garcia‐Vega 2004; Mizrahi 2012; Ozgursoy Uran 2019; Sharma 2015; Volz 2016; Yilmaz 2019); five were multicentre (Bao 2016; Berrill 2014; Cox 2020; Hawkes 2001; Higgins 2019); and two studies did not provide this information (Tapete 2018; Tapete 2019).

Participants

All studies reported age in mean (SD), except for two studies that did not report it at all (Tapete 2018; Tapete 2019); one study that reported mean and range of ages (Yilmaz 2019); and one study that only mentioned their accepted age range for participants (Sharma 2015). The average age of participants ranged from 31.7, in Garcia‐Vega 2004, to 44.9, in Berrill 2014. The lowest accepted age was 16 years (Hawkes 2001; Sharma 2015), and the highest 80 years (Higgins 2019; Volz 2016); 4 studies did not have an upper age limit (Cox 2020; Hawkes 2001; Mizrahi 2012; Ozgursoy Uran 2019). Four studies did not mention age in their inclusion/exclusion criteria (Bao 2016; Garcia‐Vega 2004; Tapete 2018; Tapete 2019).

Five studies examined exclusively CD populations (Bao 2016; Espi Lopez 2018; Garcia‐Vega 2004; Hawkes 2001; Higgins 2019), whilst the remaining studies examined a mix of IBD patients (Berrill 2014; Cox 2020; Mizrahi 2012; Ozgursoy Uran 2019; Sharma 2015; Tapete 2018; Tapete 2019; Volz 2016; Yilmaz 2019). Cox 2020 reported separate CD results. We contacted the authors of the other studies providing mixed IBD results to ask for separate outcome results for their CD participants, and one was able to provide this information (Yilmaz 2019).

Four studies examined participants in an active stage of the disease (Bao 2016; Hawkes 2001; Mizrahi 2012; Volz 2016); six studies participants in an inactive stage of the disease (Berrill 2014; Cox 2020; Garcia‐Vega 2004; Sharma 2015; Tapete 2018; Tapete 2019); one study participants in an inactive or mild stage of the disease (Higgins 2019); one study participants in inactive to moderate stages of the disease (Yilmaz 2019); and one study participants with a mix of inactive and active disease (Ozgursoy Uran 2019). One study did not report on activity of the disease (Espi Lopez 2018).

Nine studies reported disease duration (Bao 2016; Cox 2020; Garcia‐Vega 2004; Hawkes 2001; Higgins 2019; Ozgursoy Uran 2019; Volz 2016; Yilmaz 2019). All of these studies presented disease duration in mean (SD) except Cox 2020, which only provided the mean; Ozgursoy Uran 2019, which reported disease duration in incremental ranges in months; and Yilmaz 2019, which provided the mean and range. Average disease duration ranged from 3 years, in Yilmaz 2019, to 12 years, in Higgins 2019.

Interventions

The following interventions were assessed in the included trials.

Low FODMAP (fermentable oligo‐, di‐, monosaccharides and polyols) diet versus sham diet (Cox 2020; Tapete 2018).
Low FODMAP diet versus high FODMAP/normal diet (Tapete 2019).
Medicine‐separated moxibustion combined with acupuncture versus wheat bran‐separated moxibustion combined with shallow acupuncture (Bao 2016).
Mindfulness with cognitive behavioural therapy (CBT) versus no treatment (both groups received standard medical therapy) (Berrill 2014).
Soft non‐manipulative osteopathic versus no treatment besides doctor advice (Espi Lopez 2018).
Stress management versus self‐directed stress management (it is unclear whether these interventions replaced standard treatment or were added to standard treatment) versus standard treatment (Garcia‐Vega 2004).
Enteric‐release glyceryl trinitrate versus placebo (Hawkes 2001).
100 mg olorinab three times per day versus 25 mg olorinab three times per day (Higgins 2019).
Relaxation training versus waitlist (Mizrahi 2012).
Web‐based education versus standard book‐based education (Ozgursoy Uran 2019).
Yoga intervention versus no treatment (both groups received standard medical therapy) (Sharma 2015).
Transcranial direct current stimulation versus sham stimulation (Volz 2016).
Kefir diet (Lactobacillus bacteria) versus no intervention (Yilmaz 2019).

Outcomes

The length of the interventions ranged from 5 days, in Volz 2016, to 12 months, in Berrill 2014.

Primary outcomes

Treatment success as defined by the authors

Only two studies clearly defined their criteria for treatment success: Sharma 2015, which measured pain as a dichotomous outcome (presence or absence of pain), and Higgins 2019, which defined success as achieving ≥ 30% reduction in weekly average abdominal pain from baseline to week 8. Treatment success was not explicitly mentioned in the remaining studies. Study authors reported pain as a continuous outcome and did not report numbers of responders for their interventions per any definition.

Abdominal pain frequency or change in frequency of pain

Three studies measured pain frequency. Bao 2016 evaluated pain frequency using Traditional Chinese Medicine (TCM) symptom scores on an adapted pain intensity scale of 0 to 3: 0 (none), 1 (light), 2 points (moderate), 3 points (severe). The wording of the adapted scale for pain frequency was not mentioned. Cox 2020 measured pain frequency in days using the Irritable Bowel Syndrome Severity Scoring System (IBS‐SSS) 0‐to‐100 scoring scale, and in days where pain was reported as moderate or severe using the Gastrointestinal Symptom Rating Scale (GSRS). Berrill 2014 used the total IBS‐SSS scores, which measure frequency and severity of abdominal discomfort, severity of abdominal bloating, satisfaction with bowel habit, and impact of symptoms on life in general. Each domain is scored 0 to 100, and an overall score of 0 to 500 is obtained. A higher score is indicative of more severe symptoms.

Abdominal pain intensity or change in pain intensity using any validated scale

All studies except Sharma 2015 measured pain intensity. Five studies used a 0‐to‐100‐millimetre visual analogue scale (VAS), where 0 indicates no pain and 100 (or 10 if measured in centimetres) the worst pain possible (Espi Lopez 2018; Mizrahi 2012; Tapete 2018; Tapete 2019; Volz 2016). One study, Ozgursoy Uran 2019, used the 0‐to‐100‐millimetre VAS to separate participants into groups that experienced no pain, pain from 1 to 5, and pain from 6 to 10. Four studies used scoring scales between 0 and 3. Bao 2016 evaluated pain intensity using TCM symptom scores on a scale of 0 to 3, where 0 = none, 1 = light, 2 = moderate, 3 = severe; Hawkes 2001 used scores that represented the sum of a 7‐day diary card score on a scale of 0 = no pain to 3 = severe pain; Yilmaz 2019 used a symptoms diary where participants rated their pain on a scale of 0 to 3, where 0 = none, 1 = mild, 2 = moderate, and 3 = severe; and Cox 2020 used the IBS‐SSS 0‐to‐100 scale and the GSRS, which measures severity of pain on a 0‐to‐3 scale. Garcia‐Vega 2004 used a symptom diary where participants rated the severity of their daily pain on a scale of 1 to 3 (1 = mild, 2 = moderate, 3 = severe) and devised their own formulas based on diary card scores to calculate their results. Berrill 2014 used the total IBS‐SSS scores, which measure frequency and severity of abdominal discomfort, severity of abdominal bloating, satisfaction with bowel habit, and impact of symptoms on life in general. Each domain is scored 0 to 100, and an overall score of 0 to 500 is obtained. A higher score is indicative of more severe symptoms. Higgins 2019 used a weekly Average Abdominal Pain Score (AAPS), which was the daily pain scores averaged over one week that were larger than 4 on a scale of 0 (no pain) to 10 (worst ever).

Withdrawal due to adverse events

This was reported or could be extracted based on the text in 10 studies (Bao 2016; Berrill 2014; Cox 2020; Espi Lopez 2018; Hawkes 2001; Higgins 2019; Ozgursoy Uran 2019; Sharma 2015; Volz 2016; Yilmaz 2019).

Secondary outcomes

Anxiety/depression

Five studies mentioned having measured anxiety or depression, or both (Berrill 2014; Espi Lopez 2018; Mizrahi 2012; Sharma 2015; Volz 2016). Two of these studies used the Hospital Anxiety and Depression Scale (HADS) (Berrill 2014; Espi Lopez 2018); however, Berrill 2014 only reported baseline scores, and Espi Lopez 2018 did not report any scores. Mizrahi 2012 used Spielberger’s State‐Trait Anxiety Inventory to measure anxiety and the 0‐to‐10‐centimetre VAS to measure depression; Sharma 2015 also used Spielberger’s State‐Trait Anxiety Inventory to measure anxiety. Volz 2016 used the Beck Depression Inventory to measure depression, but only reported baseline scores.

Adverse events (total number of participants with any event)

Six studies reported the total number of participants with adverse events (Bao 2016; Cox 2020; Espi Lopez 2018; Hawkes 2001; Higgins 2019; Sharma 2015). Volz 2016 reported total occurrences of adverse events, but not the number of participants with adverse events.

Serious adverse events (as defined by the authors within the primary study )

The same six studies that reported numbers of participants with adverse events also reported numbers of participants with serious adverse events (Bao 2016; Cox 2020; Espi Lopez 2018; Hawkes 2001; Higgins 2019; Sharma 2015).

Funding sources and conflicts of interest

Eight studies reported their sources of funding (Berrill 2014; Cox 2020; Hawkes 2001; Higgins 2019; Ozgursoy Uran 2019; Sharma 2015; Volz 2016; Yilmaz 2019). Three studies were funded via government grants (Berrill 2014; Sharma 2015; Volz 2016), two by private foundations (Cox 2020; Hawkes 2001), one by an industrial partner (Higgins 2019), and two studies reported having received no funding (Ozgursoy Uran 2019; Yilmaz 2019).

Eight studies made declarations on conflicts of interest (Berrill 2014; Cox 2020; Espi Lopez 2018; Higgins 2019; Ozgursoy Uran 2019; Sharma 2015; Volz 2016; Yilmaz 2019). Six studies declared no conflicts of interest (Berrill 2014; Espi Lopez 2018; Ozgursoy Uran 2019; Sharma 2015; Volz 2016; Yilmaz 2019); one study declared industry connections and ownership of an invention connected to their intervention (Cox 2020); and one study declared that one of the authors is an employee of the industrial partner that provided funding (Higgins 2019).

Excluded studies

We excluded eight studies for various reasons. The reasons for exclusion of each study are presented in the Characteristics of excluded studies table and are summarised below.

Wrong outcomes (3 studies) (ACTRN12617000876392; Engel 2016; Tripp 2017).
Wrong interventions (2 studies) (Forbes 2019; Gearry 2009).
Not RCTs (2 studies) (McCormick 2010; Spagnuolo 2017).
Wrong indication (1 study) (ISRCTN98226923).

Risk of bias in included studies

The results of our ‘Risk of bias’ assessment are presented below (Figure 2; Figure 3). Further details can be found in the ‘Risk of bias’ tables in the Characteristics of included studies tables.

Figure 2

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Figure 3

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Allocation

Randomisation was described clearly in nine studies, which we rated as at low risk of bias (Bao 2016; Berrill 2014; Cox 2020; Espi Lopez 2018; Hawkes 2001; Ozgursoy Uran 2019; Sharma 2015; Volz 2016; Yilmaz 2019), and insufficiently described in five studies, which were assessed as at unclear risk of bias (Garcia‐Vega 2004; Higgins 2019; Mizrahi 2012; Tapete 2018; Tapete 2019).

Four studies were assessed as at low risk of allocation concealment bias (Berrill 2014; Cox 2020; Espi Lopez 2018; Sharma 2015), as allocation concealment was judged to be adequately described. We rated the other 10 studies as at unclear risk of bias for allocation concealment (Bao 2016; Garcia‐Vega 2004; Hawkes 2001; Higgins 2019; Mizrahi 2012; Ozgursoy Uran 2019; Tapete 2018; Tapete 2019; Volz 2016; Yilmaz 2019), as they provided insufficient or no information.

Blinding

We rated two studies as having a low risk of performance bias since they used as control a placebo that was identical to the intervention product, Hawkes 2001, or a sham procedure that could not be differentiated from the intervention, Volz 2016. We rated Bao 2016 as at unclear risk of performance bias as there was no information about blinding, and we received no response to our enquiries to the author. We assessed the other 11 studies as at high risk of performance bias, as neither participants nor study personnel were blinded to the interventions, or the studies were open‐label (Berrill 2014; Cox 2020; Espi Lopez 2018; Garcia‐Vega 2004; Higgins 2019; Mizrahi 2012; Ozgursoy Uran 2019; Sharma 2015; Tapete 2018; Tapete 2019; Yilmaz 2019); however, Cox 2020 used a sham diet to blind their participants to the intervention, which is not typical in diet RCTs due to the difficulty it entails.

Five studies provided sufficient information about blinding of outcome assessment and were assessed as at low risk of detection bias (Espi Lopez 2018; Garcia‐Vega 2004; Ozgursoy Uran 2019; Volz 2016; Yilmaz 2019). The remaining nine studies provided insufficient information for judgement and were judged to be at unclear risk of detection bias (Bao 2016; Berrill 2014; Cox 2020; Hawkes 2001; Higgins 2019; Mizrahi 2012; Sharma 2015; Tapete 2018; Tapete 2019).

Incomplete outcome data

Ten studies provided sufficient information for judgement and were assessed as at low risk of attrition bias (Bao 2016; Cox 2020; Espi Lopez 2018; Hawkes 2001; Higgins 2019; Mizrahi 2012; Ozgursoy Uran 2019; Sharma 2015; Tapete 2018; Tapete 2019). We assessed three studies as at unclear risk of bias for this domain (Garcia‐Vega 2004; Tapete 2018; Tapete 2019), and one study as at high risk of attrition bias (Berrill 2014).

Selective reporting

Nine studies reported all outcomes they set out to assess and were judged to be at low risk of reporting bias (Bao 2016; Berrill 2014; Cox 2020; Hawkes 2001; Mizrahi 2012; Ozgursoy Uran 2019; Tapete 2018; Volz 2016; Yilmaz 2019). Four studies provided insufficient information in the reports or protocols to permit a judgement as to whether all outcomes had been reported; these studies were assessed as at unclear risk of reporting bias (Garcia‐Vega 2004; Higgins 2019; Sharma 2015; Tapete 2019). We rated the remaining study as at high risk of bias for this domain (Espi Lopez 2018).

Other potential sources of bias

We rated eight studies as at low risk of other potential sources of bias (Berrill 2014; Cox 2020; Espi Lopez 2018; Ozgursoy Uran 2019; Sharma 2015; Tapete 2018; Volz 2016; Yilmaz 2019). We assessed two studies as at unclear risk of bias for this domain, as there were imbalances in the baseline characteristics of participants, and it is unclear how this may have influenced the results (Hawkes 2001; Tapete 2019). We rated four studies as having a high risk of other bias for the following reasons: significant differences in the baseline characteristics of participants that were highly likely to have affected the results (Garcia‐Vega 2004; Mizrahi 2012); the study declared that one or more authors were directly employed by the pharmaceutical companies funding the study (Higgins 2019); bias was detected in the way the efficacy of the studied intervention was presented in their introduction (Bao 2016).

Effects of interventions

A summary of the interventions and key outcome definitions and data are presented in Table 1 and Table 2 and explained below.

Table 1. Primary outcome details

Comparison	Study ID	Disease type	Disease activity	Length of intervention	Measurement of pain	Number of randomised participants
Low FODMAP diet vs sham diet	Cox 2020	CD/IBD	Inactive	4 weeks	Pain frequency and intensity: IBS‐SSS for pain rating scale 0 to 100, GSRS rating scale 0 to 3	52 (IG: 27; CG: 25)
Low FODMAP diet vs sham diet	Tapete 2018	IBD	Inactive	6 to 8 weeks	Pain intensity: VAS rating scale 0 to 10 cm	30 (IG: 10; CG: 20)
Low FODMAP diet vs high FODMAP/normal diet	Tapete 2019	IBD	Inactive	8 weeks cross‐over (4 weeks x 2)	Pain intensity: VAS rating scale 0 to 10 cm	50 (IG: 25; CG: 25)
Medicine‐separated moxibustion combined with acupuncture vs wheat bran‐separated moxibustion combined with shallow acupuncture	Bao 2016	CD	Mild to moderate	12 weeks	Pain frequency and intensity: Traditional Chinese Medicine rating scale 0 to 3	102 (IG: 51; CG: 51)
Mindfulness with CBT + standard medical therapy vs standard medical therapy	Berrill 2014	IBD	Inactive	12 months	Pain frequency and severity of abdominal discomfort, severity of abdominal bloating, satisfaction with bowel habit, and impact of symptoms on life in general. Each domain is scored 0 to 100, and an overall score of 0 to 500 is obtained.	66 (IG: 33; CG: 33)
Soft non‐manipulative osteopathic vs no intervention	Espi Lopez 2018	CD	Unclear	30 days	Pain intensity: VAS rating scale 0 to 10 cm	30 (IG: 16; CG: 14)
Stress management vs self‐directed stress management vs standard treatment	Garcia‐Vega 2004	CD	Inactive	8 weeks	Pain intensity: rating scale 1 to 3 and author formulas	45 (IG1: 15; IG2: 15; CG: 15)
Enteric‐release glyceryl trinitrate vs placebo	Hawkes 2001	CD	Moderate to severe	12 weeks	Pain intensity: rating scale 0 to 3	70 (IG: 34; CG: 36)
100 mg olorinab 3 times/day vs 25 mg olorinab 3 times/day	Higgins 2019	CD	Inactive to mild	8 weeks	Pain intensity: AAPS of 0‐to‐10 Likert scale	14 (IG: 8; CG: 6)
Relaxation training vs waitlist	Mizrahi 2012	IBD	Active	5 weeks	Pain intensity: VAS rating scale 0 to 10 cm	56 (IG: 28; CG: 28)
Web‐based education vs standard book‐based education	Ozgursoy Uran 2019	IBD	Mix of active and inactive	8 weeks	Pain intensity: VAS rating scale 0 to 10 cm	60 (IG: 10; CG: 10)
Yoga intervention + standard medical therapy vs standard medical therapy	Sharma 2015	IBD	Inactive	8 weeks	Presence or absence of pain	100 (IG: 50; CG: 50)
Transcranial direct current stimulation vs sham stimulation	Volz 2016	IBD	Active	5 days	Pain intensity: Pressure Pain Threshold taken with algometer, VAS rating scale 0 to 10 cm	20 (IG: 10; CG: 10)
Kefir diet (Lactobacillus bacteria) vs no intervention	Yilmaz 2019	CD/IBD	Inactive to moderate	4 weeks	Pain intensity: rating scale 0 to 3	48 (IG: 28; CG: 20)

AAPS: average abdominal pain score

CBT: cognitive behavioural therapy

CD: Crohn's disease

CG: control group

FODMAP: fermentable oligo‐, di‐, monosaccharides and polyols

IG: intervention group

GSRS: Gastrointestinal Symptom Rating Scale

IBD: inflammatory bowel disease

IBS‐SSS: Irritable Bowel Syndrome Severity Scoring System

VAS: visual analogue scale

Table 2. Primary outcome data

Comparison	Study ID	Treatment success end of study data IG/CG	Pain frequency end of study data IG/CG	Pain intenstiy end of study data IG/CG	Withdrawals due to adverse events IG/CG
Low FODMAP diet vs sham diet	Cox 2020	NR	IBS‐SSS all participants mean(SD) IG= 36(26); CG= 38(25) IBS‐SSS CD mean (SD) IG = 36(138.4); CG=48(128.2) GSRS all participants mean(SD) IG= 1.5(1.6); CG= 1.1(1.5)	IBS‐SSS all participants mean(SD) IG=22(15.6) ; CG=30(15) IBS‐SSS CD mean (SD) IG = 24(82.3); CG= 32(69.3) GSRS all participants mean(SD) IG=0.9(2.6) ; CG= 0.7(22.5)	IG: 2 CG: 1
Low FODMAP diet vs sham diet	Tapete 2018	NR	NR	mean(SD): IG=3.3 (1.9) CG= 4.3(2.2)	NR
Low FODMAP diet vs high FODMAP/normal diet	Tapete 2019	NR	NR	mean(SD): IG=1.1(1.6) CG=3.1(2.3) (unclear if average of both cross‐over arms or from the first or second phase of the cross‐over)	NR
Medicine‐separated moxibustion combined with acupuncture vs wheat bran‐separated moxibustion combined with shallow acupuncture	Bao 2016	NR	mean (no SD provided) IG=0 CG=0	mean (no SD provided) IG=2 CG=1	IG=0 CG=0
Mindfulness with CBT + standard medical therapy vs standard medical therapy	Berrill 2014	NR	Frequency and intensity measured together as part of the IBS‐SSS score mean(SD) IG= 187 (97) CG= 224 (111)		IG=0 CG=0
Soft non‐manipulative osteopathic vs no intervention	Espi Lopez 2018	NR	NR	mean(SD): IG=2.72(2.66) CG=2.71(2.43)	IG=0 CG=0
Stress management vs self‐directed stress management vs standard treatment	Garcia‐Vega 2004	NR	NR	mean(SD): IG=13.3(28) CG=47.6(47)	NR
Enteric‐release glyceryl trinitrate vs placebo	Hawkes 2001	NR	NR	mean(no SD presented) IG=8.1 CG=8.6	IG=9 CG=3
100 mg olorinab 3 times/day vs 25 mg olorinab 3 times/day	Higgins 2019	≥ 30% reduction in weekly AAPS IG: 5 CG: 6	NR	1.5 hours postdose mean(SD could not be calculated from figure) IG=1.9 CG= 1.9 Mean change in AAPS (no SD) IG= ‐4.6 CG= ‐4.6	IG=0 CG=0
Relaxation training vs waitlist	Mizrahi 2012	NR	NR	mean(SD): IG= 2.23(1.83) CG= 2.95(2.44)	NR
Web‐based education vs standard book‐based education	Ozgursoy Uran 2019	NR	NR	mean(SD): IG= 1.8(2.04) CG= 1.93(2.39)	IG=0 CG=0
Yoga intervention + standard medical therapy vs standard medical therapy	Sharma 2015	NR	NR	NR	IG=0 CG=0
Transcranial direct current stimulation vs sham stimulation	Volz 2016	NR	NR	mean(SD): IG= 2.8(2.3) CG= 4.45(1.3)	IG=0 CG=0
Kefir diet (Lactobacillus bacteria) vs no intervention	Yilmaz 2019	NR	NR	CD pain intensity mean(SD) IG= 0.2(0.63) CG= 1.3(0.67) IBD pain intensity mean(SD) IG= 0.9(0.97) CG= 0.28(0.61)	IG=0 CG=0

AAPS: average abdominal pain score

CBT: cognitive behavioural therapy

CD: Crohn's disease

CG: control group

FODMAP: fermentable oligo‐, di‐, monosaccharides and polyols

IG: intervention group

GSRS: Gastrointestinal Symptom Rating Scale

IBD: inflammatory bowel disease

IBS‐SSS: Irritable Bowel Syndrome Severity Scoring System

VAS: visual analogue scale

Low FODMAP diet versus sham diet

Two studies compared a diet low in FODMAP to a sham diet (Cox 2020; Tapete 2018). Both studies included participants with either CD or UC and who were at an inactive stage of their disease. Tapete 2018 (n = 30) included participants with a Crohn's Disease Activity Index (CDAI) score < 150 for CD and a full Mayo score < 3 for UC, whilst in Cox 2020 (n = 52) quiescent IBD was defined by all of the following: physician global assessment, stable medications, no IBD flare in the previous six months, faecal calprotectin < 250 mg/g, and serum C‐reactive protein (CRP) < 10 mg/L. The length of the intervention was four weeks in Cox 2020 and six to eight weeks in Tapete 2018.

Primary outcomes

Treatment success was not reported. Pain in Cox 2020 was measured using the pain subscale of the IBS‐SSS that rates pain on a scale of 0 to 100 and the GSRS pain‐rating scale of 0 to 3, whilst Tapete 2018 used the 0‐to‐100‐millimetre VAS.

Pain frequency

Only Cox 2020 measured pain frequency.

At the end of study, the mean (SD) IBS‐SSS pain frequency in days was 36 (26) days for the 27 CD and UC participants in the low FODMAP group, and 38 (25) days for the 25 participants in the sham diet group. There was no clear difference in days of pain for CD and UC participants when a low FODMAP diet was compared to a sham diet in Cox 2020 (mean difference (MD) −2.00, 95% confidence interval (CI) −15.86 to 11.86). The certainty of evidence was very low due to unclear risk of bias and imprecision (Analysis 1.1; summary of findings Table 1).

The mean (SD) days in which pain was rated as moderate or severe on the GSRS during the final week of the diet was 1.5 (1.6) for the low FODMAP group and 1.1 (1.5) for the sham diet group. There was no clear difference in days of pain during the last week of the diet for CD and UC participants when a low FODMAP diet was compared to a sham diet in Cox 2020 (MD 0.40, 95% CI −0.44 to 1.24). The certainty of evidence was very low due to unclear risk of bias and imprecision (Analysis 1.2; summary of findings Table 1).

Cox 2020 also provided separate data for their CD participants. At end of study, the mean (SD) IBS‐SSS pain frequency in days was 36 (138.4) days for the 14 CD participants in the low FODMAP group, and 48 (128.2) days for the 12 participants in the sham diet group. There was no clear difference in days of pain for CD participants when a low FODMAP diet was compared to a sham diet in Cox 2020 (MD −12.00, 95% CI −114.55 to 90.55). The certainty of evidence was very low due to unclear risk of bias and imprecision (Analysis 1.3; summary of findings Table 1).

We requested separate CD data for the GSRS but were not able to obtain this information from the authors.

Pain intensity

Both studies measured pain intensity, in Cox 2020 on a 0‐to‐100mm VAS and in Tapete 2018 on a 0‐to‐10cm VAS. To combine them for analysis we multiplied the Tapete 2018 results by 10.

When the effects of both studies were analysed together, there was a small difference in intensity of pain for CD and UC participants (MD −8.46, 95% CI −15.76 to −1.16). The certainty of evidence was very low due to high risk of bias and imprecision, so no conclusions could be drawn (Analysis 1.4; summary of findings Table 1).

In Cox 2020, the mean (SD) pain intensity on the GSRS was 0.9 (2.6) for the low FODMAP group and 0.7 (22.5) for the sham diet group. There was no clear difference in pain intensity for CD and UC participants when a low FODMAP diet was compared to a sham diet on a 4‐point scale in Cox 2020 (MD 0.20, 95% CI −8.67 to 9.07). The certainty of evidence was very low due to unclear risk of bias and imprecision (Analysis 1.5; summary of findings Table 1).

Cox 2020 also provided separate data for their CD participants. At the end of study, the mean (SD) IBS‐SSS pain intensity was 24 (82.3) for the 14 CD participants in the low FODMAP group and 32 (69.3) for the 12 CD participants in the sham diet group. There was no clear difference in intensity of pain for CD participants when a low FODMAP diet was compared to a sham diet in Cox 2020 (MD −8.00, 95% CI −66.27 to 50.27). The certainty of evidence was very low due to unclear risk of bias and imprecision (Analysis 1.6; summary of findings Table 1).

We requested separate CD data for the GSRS but were not able to obtain this information from the authors.

Withdrawals due to adverse events

Only Cox 2020 reported withdrawals due to adverse events. There were two withdrawals due to adverse events in the low FODMAP group (one IBD relapse, one commencement of antibiotics) and one in the sham diet group (one IBD relapse). There was no clear difference in the effect of the intervention on withdrawals due to adverse events (risk ratio (RR) 1.85, 95% CI 0.18 to 19.19) (Analysis 1.7; summary of findings Table 1). The certainty of the evidence was very low due to risk of bias and imprecision. We requested data from Tapete 2018 but were unable to obtain this information.

Secondary outcomes

Neither Cox 2020 nor Tapete 2018 reported on anxiety/depression. Only Cox 2020 reported adverse events and serious adverse events. There were no serious adverse events, and adverse events were reported in six participants in total. Three of these were the ones mentioned above in withdrawals due to adverse events. The other three were one case of worsening of abdominal pain (sham diet group), and two cases of flu‐like symptoms and sinusitis (one in each group).

One participant each in the intervention and control groups reported flu‐like symptoms and sinusitis, and one participant in the intervention group reported worsening of abdominal pain. One participant each in the intervention and control groups reported IBD relapse.

Low FODMAP diet versus high FODMAP/normal diet

One study (n = 50) compared a diet low in FODMAP to a high FODMAP/normal diet (Tapete 2019). The study included participants with either CD or UC who were at an inactive stage of their disease. Study participants had normal inflammatory parameters (CRP, white blood cells) and were deemed to be in clinical remission (CDAI < 150 for CD and a full Mayo < 2 for UC) but had residual abdominal symptoms (abdominal pain, diarrhoea, bloating). The length of the intervention was eight weeks; the study had a cross‐over design whereby groups alternated interventions after four weeks. We wrote to the authors for pre‐cross‐over data but received no response.

Primary outcomes

Treatment success was not reported. Tapete 2019 measured pain using the 0‐to‐10‐centimetre VAS. Pain frequency and withdrawals due to adverse events were not reported.

Pain intensity

Tapete 2019 reported that at the end of study the mean (SD) pain score was 1.1 (1.6) in the low FODMAP diet group and 3.1 (2.3) in the high FODMAP/normal diet group. However, it is unclear whether these scores were averages of both cross‐over arms, or if they were from the first or second phase of the cross‐over. The cross‐over design of the study precluded any analyses of the results.

Secondary outcomes

Tapete 2019 did not report on anxiety/depression, serious adverse events, or total number of participants with any event.

Medicine‐separated moxibustion combined with acupuncture versus wheat bran‐separated moxibustion combined with shallow acupuncture

One study (n = 102) compared medicine‐separated moxibustion combined with acupuncture to wheat bran‐separated moxibustion combined with shallow acupuncture (Bao 2016). The study included only CD participants who were at a mild to moderate stage of the disease (defined as a CDAI between 151 and 350). The length of the intervention was 12 weeks.

Primary outcomes

Bao 2016 did not report on treatment success. Pain was measured on the TCM 4‐point rating scale from 0 to 3.

Pain frequency

Pain frequency was reduced from an average score of 3 at baseline to 0 at end of study in the medicine‐separated moxibustion group, and from an average score of 1 to 0 in the wheat bran‐separated moxibustion group. The authors did not provide SD values, so we could not conduct any analyses of the results; we requested this information but received no response. It is uncertain whether medicine‐separated moxibustion combined with acupuncture leads to a difference in average pain frequency when compared with wheat bran‐separated moxibustion combined with shallow acupuncture.

Pain intensity

Pain intensity was reduced from an average score of 3 at baseline to 2 at end of study in the medicine‐separated moxibustion group, and remained at an average score of 1 for the wheat bran‐separated moxibustion group. The authors did not provide SD values, so we could not conduct any analyses of the results; we requested this information but received no response. It is uncertain whether medicine‐separated moxibustion combined with acupuncture leads to a difference in average pain intensity when compared with wheat bran‐separated moxibustion combined with shallow acupuncture.

Withdrawals due to adverse events

There were no withdrawals due to adverse events in either the medicine‐separated moxibustion or wheat bran‐separated moxibustion group, so there was no estimable relative effect of the intervention on withdrawals due to adverse events (Analysis 2.1; summary of findings Table 2).

Secondary outcomes

The authors reported that there were no serious adverse events. Two participants experienced any adverse event, one in each treatment group. Anxiety or depression was not reported.

Mindfulness with cognitive behavioural therapy (CBT) versus no treatment (both groups received standard medical therapy)

One study (n = 66) compared mindfulness with CBT to no treatment (Berrill 2014). The study included participants with CD or UC who were at an inactive stage of their disease. Study participants had a diagnosis of UC or CD that was in remission based on a clinical activity index score and a CRP level < 10 mg/L. The length of the intervention was 12 months.

Primary outcomes

Berrill 2014 did not report on treatment success. Pain was measured on the 0‐to‐500 IBS‐SSS, which is the sum of five separate 0‐to‐100 scales that measure five different pain‐related symptoms: pain frequency and severity of abdominal discomfort, severity of abdominal bloating, satisfaction with bowel habit, and impact of symptoms on life in general.

Pain frequency and intensity

At the end of the study, mean (SD) IBS‐SSS score (which includes pain frequency and severity) was 187 (97) for the mindfulness with CBT group and 224 (111) for the no‐treatment group. There was no clear difference in IBS‐SSS symptoms between groups (MD −37.00, 95% CI −87.29 to 13.29). The certainty of evidence was very low due to unclear risk of bias and imprecision (Analysis 3.1; summary of findings Table 3).

Withdrawals due to adverse events

There were no withdrawals due to adverse events in either group, so there was no estimable relative effect of the intervention on withdrawals due to adverse events (Analysis 3.2; summary of findings Table 3).

Secondary outcomes

Anxiety and depression at the end of study were not presented. Serious adverse events or total number of participants with any adverse event was not reported.

Soft non‐manipulative osteopathic versus no intervention

One study (n = 30) compared soft non‐manipulative osteopathic treatment to no intervention (Espi Lopez 2018). The study included only CD participants, and it was unclear whether participants had active or inactive stage of the disease. The length of the intervention was 30 days.

Primary outcomes

Espi Lopez 2018 did not report on treatment success. Pain was measured on the 0‐to‐10‐centimetre VAS. Pain frequency was not measured.

Pain intensity

The mean (SD) score at end of the study was 2.72 (2.66) for the osteopathic group and 2.71 (2.43) for the no‐intervention group. There was no clear difference in pain intensity in the osteopathic group when compared to the no‐intervention group (MD 0.01, 95% CI −1.81 to 1.83). The certainty of evidence was very low due to high risk of bias and imprecision (Analysis 4.1; summary of findings Table 4).

Withdrawals due to adverse events

There were no withdrawals due to adverse events in either group, so there was no estimable relative effect of the intervention on withdrawals due to adverse events (Analysis 4.2; summary of findings Table 4).

Secondary outcomes

The authors reported that no adverse events occurred in this study. Anxiety/depression results were not reported.

Directed stress management versus self‐directed stress management (it is unclear whether these interventions replaced standard treatment or were added to standard treatment) versus standard treatment

One study (n = 45) compared directed stress management to standard treatment and self‐directed management to standard treatment (Garcia‐Vega 2004). It was unclear whether the interventional treatments included standard treatment or not. We attempted to contact the author but received no response.

The study included only CD participants who were at an inactive stage of the disease based on the Harvey‐Bradshaw Index. The length of the intervention was eight weeks.

Primary outcomes

Garcia‐Vega 2004 did not report on treatment success. Pain scores were calculated with author‐derived formulas that were based on a 3‐point rating scale from 1 to 3. Pain frequency and withdrawals due to adverse events were not reported in this study.

Pain intensity

At end of the study the mean (SD) pain score was 13.3 (28.8) for the directed stress management group; 17.1 (29) for the self‐directed stress management group; and 47.6 (47) for the standard treatment group. There was a difference in the directed stress management group when compared to the standard treatment group ( MD ‐34.30, 95% CI ‐61.99 to ‐6.61 ) and in the self‐directed stress management group when compared to the standard treatment group ( MD ‐30.50, 95% CI ‐58.45 to ‐2.55 ). The certainty of evidence was very low due to high risk of bias and imprecision, therefore no conclusions could be drawn (Analysis 5.1; Analysis 6.1 summary of findings Table 5; summary of findings Table 6).

Secondary outcomes

Garcia‐Vega 2004 did not report anxiety/depression, serious adverse events, or total number of participants with any adverse event.

Enteric‐release glyceryl trinitrate versus placebo

One study (n = 70) compared enteric‐release glyceryl trinitrate to placebo (Hawkes 2001). The study included only CD participants at a moderate to severe stage of the disease with a CDAI of between 150 and 450. The length of the intervention was 12 weeks.

Primary outcomes

Hawkes 2001 did not report on treatment success. Pain was measured on a 4‐point scale from 0 to 3. Pain frequency was not reported in this study.

Pain intensity

At end of study the mean score was 8.1 for the glyceryl group and 8.6 for the placebo group. No analyses could be conducted with these results as SD values were unobtainable from the authors. It is uncertain whether enteric‐release glyceryl trinitrate leads to any differences in pain intensity compared to placebo.

Withdrawal due to adverse events

Nine participants in the glyceryl group (headache = 1, worsening clinical condition = 4, generalised rash = 1, mood change/irritability = 1, loss of consciousness/memory = 1) and three participants in the placebo group (headache = 2, worsening clinical condition = 1) experienced adverse events that led to withdrawal from the study. There was no clear difference in withdrawals due to adverse events when enteric‐release glyceryl trinitrate was compared to placebo (RR 3.18, 95% CI 0.94 to 10.76). The certainty of the evidence was very low due to unclear risk of bias and high imprecision (Analysis 7.1; summary of findings Table 7).

Secondary outcomes

The authors of Hawkes 2001 reported that there were no serious adverse events in either group. Nineteen participants in the enteric‐release glyceryl trinitrate group and 13 participants in the placebo group experienced any adverse event. Anxiety/depression was not measured in this study.

100 mg olorinab three times/day versus 25 mg olorinab three times/day

One study (n = 14) compared 100 mg olorinab taken three times per day to 25 mg olorinab taken three times per day (Higgins 2019). The study only included CD participants who were at an inactive or mild stage of the disease with a simple endoscopic score CD < 10 or faecal calprotectin < 500 μg/g. The length of the intervention was eight weeks.

Primary outcomes

Treatment success in Higgins 2019 was defined as a ≥ 30% reduction in weekly Average Abdominal Pain Score (AAPS) from baseline to week 8, as measured on a 0‐to‐10 Likert scale. Pain frequency was not measured in this study.

Pain intensity

At end of study, five of eight participants had achieved a 30% reduction in weekly AAPS in the 100 mg olorinab group, whilst six of six participants had achieved a 30% reduction in the 25 mg olorinab group. There was no clear difference in pain intensity between groups (RR 0.66, 95% CI 0.38 to 1.15). The certainty of the evidence was very low due to high risk of bias and imprecision (Analysis 8.1; summary of findings Table 8).

Withdrawals due to adverse events

There were no withdrawals due to adverse events in either group, so there was no estimable relative effect of the intervention on withdrawals due to adverse events (Analysis 8.2; summary of findings Table 8).

Secondary outcomes

One participant in the 100 mg olorinab group experienced a serious adverse event, which the authors described as unrelated to the intervention. No serious adverse events were reported in the 25 mg olorinab group.

Six of eight participants in the 100 mg olorinab group and four of six participants in the 25 mg olorinab group experienced any adverse event.

Anxiety/depression was not reported in this study.

Relaxation training versus waitlist

One study (n = 56) compared relaxation training to a waitlist (the people in the waitlist would receive the training after the end of the study) (Mizrahi 2012). The study included CD and UC participants who were at an active stage of the disease according to the "Disease Activity Questionnaire", by meeting one of the following criteria: more than five bowel movements a day, more than one hospitalisation a year over the previous two years, and had either suffered a fistula during the previous year or was using corticosteroids. The length of the intervention was five weeks.

Primary outcomes

Mizrahi 2012 did not report on treatment success. Pain was measured on the 0‐to‐10‐centimetre VAS. Pain frequency and withdrawals due to adverse events were not reported.

Pain intensity

At end of study the mean (SD) score was 2.23 (1.83) for the relaxation training group and 2.95 (2.44) for the waitlist group. There was no clear difference in pain intensity when relaxation training was compared to a waitlist (MD −0.72, 95% CI −1.85 to 0.41). The certainty of the evidence was very low due to high risk of bias and imprecision (Analysis 9.1; summary of findings Table 9).

Secondary outcomes

Mean (SD) anxiety as measured on the Spielberger’s State‐Trait Anxiety Inventory at end of study was 36.67 (10.65) for the relaxation training group and 40.51 (12.57) for the waitlist group.

Mean (SD) depression measured using the 0‐to‐10‐centimetre VAS at end of study was 1.39 (2.23) for the relaxation training group and 1.9 (1.99) for the waitlist group.

Serious adverse events or total number of participants with any adverse event was not reported in this study.

Web‐based education versus standard book‐based education

One study (n = 60) compared web‐based education to standard book‐based education (Ozgursoy Uran 2019). The study included participants with CD or UC who were in either an active or inactive stage of the disease. The length of the intervention was eight weeks.

Primary outcomes

Ozgursoy Uran 2019 did not report on treatment success. Pain intensity was measured on the 0‐to‐10‐centimetre VAS. Pain frequency was not reported in this study.

Pain intensity

At end of study, the mean (SD) score was 1.8 (2.04) for the web‐based education group and 1.93 (2.39) for the standard book‐based education group. There was no clear difference in pain intensity when web‐based education was compared with standard book‐based education (MD −0.13, 95% CI −1.25 to 0.99). The certainty of the evidence was very low due to unclear risk of bias and imprecision (Analysis 10.1; summary of findings Table 10).

Withdrawals due to adverse events

There were no withdrawals due to adverse events in either group, so there was no estimable relative effect of the intervention on withdrawals due to adverse events (Analysis 10.2; summary of findings Table 10).

Secondary outcomes

Ozgursoy Uran 2019 did not report on anxiety/depression, serious adverse events, or total number of participants with any adverse event.

Yoga intervention versus no treatment (both groups received standard medical therapy)

One study (n = 100) compared a yoga intervention to no yoga (Sharma 2015). The study only included participants with CD or UC who were at an inactive stage of the disease with a CDAI score < 150; UC activity was measured on the Truelove and Witts Severity Index. The length of the intervention was eight weeks.

Primary outcomes

Treatment success in this study was measured as a dichotomous outcome of presence or absence of pain; however, results were reported only for the UC participants and not for CD participants. We contacted the authors for the missing data but received no response. Pain frequency or intensity was not reported.

Withdrawals due to adverse events

There were no withdrawals due to adverse events in either group, so there was no estimable relative effect of the intervention on withdrawals due to adverse events (Analysis 11.1; summary of findings Table 11).

Secondary outcomes

The authors reported that no serious adverse events occurred in either group.

There was one reported total adverse event in the yoga group (bone fracture).

Anxiety level results were presented only for UC and not for CD participants. Depression was not reported.

Transcranial direct current stimulation versus sham stimulation

One study (n = 20) compared transcranial direct current stimulation to sham stimulation (Volz 2016). The study included CD and UC participants who were at an active stage of the disease as measured with either the Simple Clinical Colitis Activity Index for UC participants or Harvey–Bradshaw Index for CD participants. The length of the intervention was five days.

Primary outcomes

Treatment success was not reported. Pain was measured on the 0‐to‐10‐centimetre VAS and mechanically on the Pressure Pain Threshold (PPT) via an algometer on the right and left abdomen. Pain frequency was not reported.

Pain intensity

At the end of study, the mean (SD) VAS score for pain intensity was 2.8 (2.3) for the transcranial stimulation group and 4.45 (1.3) for the sham stimulation group. Transcranial direct current stimulation may improve pain intensity when compared to sham stimulation (MD −1.65, 95% CI −3.29 to −0.01). The certainty of the evidence was low due to high imprecision (Analysis 12.1; summary of findings Table 12).

Withdrawals due to adverse events

There were no withdrawals due to adverse events in either group, so there was no estimable relative effect of the intervention on withdrawals due to adverse events (Analysis 12.2; summary of findings Table 12).

Secondary outcomes

The authors reported that no serious adverse events occurred in this study. They reported on total adverse events, but as number of incidents rather than number of participants with adverse events.

Depression was measured at baseline, but no end‐of‐study results were reported. Anxiety was not reported.

Kefir diet (Lactobacillus bacteria) versus no intervention

One study (n = 48) compared a kefir diet with Lactobacillus bacteria to no intervention (Yilmaz 2019). The study included participants with CD or UC whose disease activity ranged from inactive to moderate. The length of the intervention was four weeks.

Primary outcomes

Yilmaz 2019 did not report on treatment success. Pain was measured on a 4‐point rating scale from 0 to 3. Pain frequency was not reported.

Pain intensity

At the end of study, mean (SD) pain score for pain intensity was 0.2 (0.63) for the CD kefir group and 1.3 (0.67) for the no‐intervention group. No clear difference was detected in pain intensity scores when kefir diet was compared to no intervention (MD −1.10, 95% CI −1.67 to −0.53). The certainty of the evidence was very low due to unclear risk of bias and imprecision (Analysis 13.1; summary of findings Table 13).

At the end of study, mean (SD) pain score for pain intensity was 0.9 (0.97) for the IBD kefir group and 0.28 (0.61) for the no‐intervention group. No clear difference was detected in pain intensity scores when kefir diet was compared to no intervention (MD 0.62, 95% CI 0.17 to 1.07). The certainty of the evidence was very low due to unclear risk of bias and imprecision (Analysis 13.2; summary of findings Table 13).

Withdrawals due to adverse events

There were no withdrawals due to adverse events in either group, so there was no estimable relative effect of the intervention on withdrawals due to adverse events (Analysis 13.3; summary of findings Table 13).

Secondary outcomes

Yilmaz 2019 did not report on anxiety/depression, serious adverse events, or total number of participants with any adverse event.

Discussion

Summary of main results

This review included a wide range of interventions. Three were forms of diet (Cox 2020; Tapete 2018; Tapete 2019; Yilmaz 2019); three were forms of psychological management (Berrill 2014; Garcia‐Vega 2004; Mizrahi 2012); two were forms of alternative therapies (Bao 2016; Espi Lopez 2018); two were drug interventions (Hawkes 2001; Higgins 2019); one was an educational intervention (Ozgursoy Uran 2019); one was a form of exercise (Sharma 2015); and one was a form of brain stimulation (Volz 2016). Five studies looked exclusively at CD (Bao 2016; Espi Lopez 2018; Garcia‐Vega 2004; Hawkes 2001; Higgins 2019), whilst the remaining studies evaluated participants with both CD and UC; two of these studies provided separate data for their CD population (Cox 2020; Yilmaz 2019). The studies included a range of disease states.

Only two studies defined or reported our primary outcome of treatment success (Higgins 2019; Sharma 2015). In the remaining studies pain was only measured as a continuous outcome by improvement on a rating scale: 0‐to‐100 scales (Berrill 2014; Cox 2020); 0‐to‐10‐centimetre VAS or 0‐to‐10 Likert scale (Espi Lopez 2018; Mizrahi 2012; Ozgursoy Uran 2019; Tapete 2018; Tapete 2019; Volz 2016); 4‐point 0‐to‐3 scale (Bao 2016; Cox 2020; Hawkes 2001; Yilmaz 2019); or a combination of a 3‐point scale and author‐derived formulas (Garcia‐Vega 2004). In all studies a lower rating indicated less pain, whilst a higher rating indicated more pain. One study measured only the absence or presence of pain as a dichotomous outcome (Sharma 2015). Except for this study, all studies measured pain intensity, whilst only three studies measured pain frequency (Bao 2016; Berrill 2014; Cox 2020). Withdrawals due to adverse events were directly or indirectly reported in nine studies (Bao 2016; Berrill 2014; Cox 2020; Espi Lopez 2018; Hawkes 2001; Higgins 2019; Ozgursoy Uran 2019; Sharma 2015; Volz 2016).

This heterogeneity in outcome measures reported and interventions used severely limited our scope for meta‐analysis.

The only low‐certainty evidence found was for the comparison transcranial direct current stimulation versus sham stimulation (Volz 2016); this evidence showed that there may be an improvement in pain intensity for CD and UC participants (no separate CD data were provided), with direct current stimulation when that was measured using visual analogue scales.

In two pooled studies comparing a low FODMAP diet and a sham diet, we found very low‐certainty evidence for a difference between a low FODMAP diet and a sham diet for the outcome pain intensity for CD and UC participants (no separate CD data provided by Tapete 2018) (Cox 2020; Tapete 2018), therefore no conclusions could be drawn about the efficacy of this intervention in improving pain intensity.

In one study comparing stress management to standard treatment and self‐directed stress management to standard treatment (it is unclear whether the stress management interventions replaced standard treatment or were added to standard treatment), we found very low‐certainty evidence for a difference between both stress management and self‐directed stress management in comparison to standard treatment for the outcome pain intensity in CD participants (Garcia‐Vega 2004), therefore no conclusions could be drawn about the efficacy of these interventions in improving pain intensity.

There was no clear difference in pain intensity or frequency between interventions for any other direct comparisons, although the certainty of the evidence was very low due to imprecision from sparse data and risk of bias ratings of unclear and high risk.

Of the studies that reported withdrawals due to adverse events, all reported no withdrawals except for Cox 2020 and Hawkes 2001. Cox 2020 evaluated a low FODMAP diet versus a sham diet, and Hawkes 2001 evaluated a drug intervention comparing enteric‐release glyceryl trinitrate to placebo. The two studies showed no clear effect on withdrawals due to adverse events, although the certainty of the evidence was very low due to unclear risk of bias and high imprecision. Lack of evidence prevented us from drawing any conclusions about the effects of the other interventions on withdrawals due to adverse events.

The reporting of serious adverse events and total number of participants with any adverse event as secondary outcomes was inconsistent. Seven studies reported on serious adverse events and total number of participants with any adverse event (Bao 2016; Cox 2020; Espi Lopez 2018; Hawkes 2001; Higgins 2019; Sharma 2015; Volz 2016). The most adverse events were reported in the two drug interventions (Hawkes 2001; Higgins 2019), whilst adverse events tended to be very low or zero in the non‐drug interventions. However, no conclusions could be drawn regarding adverse events for any of the interventions studied due to the low number of events.

Mizrahi 2012 measured and reported anxiety and depression at end of study. Three studies mentioned measuring anxiety or depression at baseline, but no anxiety or depression end‐of‐study results were presented in Berrill 2014; no anxiety end‐of‐study results for CD participants were reported in Sharma 2015; and no depression end‐of‐study results were reported in Volz 2016. Consequently, no meaningful conclusions could be drawn regarding this outcome.

Overall completeness and applicability of evidence

The included studies not only considered a wide range of interventions, but most considered a mix of disease activity and form of IBD. This is rather distinct from the majority of reviewed studies in the wider Cochrane IBD portfolio, which almost always consider UC and CD separately, or at the very least report results for both groups in a single study. Similarly, the reviews reflect a mixture of stage of CD and IBD. As such, the evidence in this review is very different, with a mixture of both conditions and disease states in all papers, which raises significant issues regarding clinical heterogeneity.

It is also very clear that, despite there being a reasonable number of included studies, the great range of interventions and the small numbers of included participants leave the individual evidence at significant risk of imprecision. This is pervasive across the evidence presented in this review, with no single comparison not at risk of imprecision.

One of the issues experienced when considering the evidence was that, despite our inclusion criteria requiring interventions to be focussed on pain, not all papers provided explicit descriptions of baseline pain. Inclusion criteria could be complex, and therefore pain change scores, as well as clear baseline pain data, are areas which limit the completeness of the evidence for clinical interpretation.

The incidence of IBS amongst IBD patients is higher than in people without IBD (Halpin 2012), and it is very difficult to separate whether IBD patients' continuing pain symptoms are related to their previous disease state. Determining what gastrointestinal symptoms are truly functional symptoms is not possible, which may have affected the overall completeness and applicability of our evidence.

Another issue was the number of participants with pain at baseline, in order to determine what denominator to use for our analyses. All papers explicitly or close to explicitly included participants with gut symptoms or pain, although in same cases there was a lack of clarity. However, considering this issue pragmatically, all pain outcomes studied were continuous, and therefore used the denominator of all participants randomised, as it would not be statistically sound to do otherwise. One exception was the Higgins 2019 trial, which reported a dichotomous pain outcome, but all participants at baseline were experiencing abdominal pain, so again we used the number of all participants randomised as the denominator for our analysis.

We also noted the use of multiple scales to measure pain within the same study (Cox 2020), or proxy or measures, such as the use of a mechanical algometer to measure pain thresholds (Volz 2016). It’s important that future studies consider using the same internationally recognised measures for pain, such as visual analogue scales, in order to limit the heterogeneity that might be caused by using a variety of different methodologies.

Quality of the evidence

There were significant issues related to risk of bias in the studies included in this review. Despite our requests to authors of all included studies, we received no data to change our judgements in these key areas.

No single comparison surpassed 100 participants as a total, which meant that all outcomes were judged to be highly imprecise, and the certainty of the evidence was downgraded twice for this reason.

A number of studies did not clearly describe randomisation, Garcia‐Vega 2004; Higgins 2019; Mizrahi 2012; Tapete 2018; Tapete 2019, or allocation concealment, Bao 2016; Garcia‐Vega 2004; Hawkes 2001; Higgins 2019; Mizrahi 2012; Ozgursoy Uran 2019; Tapete 2018; Tapete 2019; Volz 2016; Yilmaz 2019.

Blinding of participants and personnel was understandably not possible in most of the included studies. However, most studies did not discuss whether their outcome assessors were blinded either.

A number of studies had issues with selective reporting (Espi Lopez 2018; Garcia‐Vega 2004; Higgins 2019; Sharma 2015; Tapete 2019), leading to further issues related to risk of bias and downgrading of the certainty of the evidence. Attrition bias was also an issue, with three studies judged as unclear risk of bias (Garcia‐Vega 2004; Tapete 2018; Tapete 2019), and one as high risk of bias for this domain (Berrill 2014).

Finally, other key sources of bias were present, mainly a potential imbalance in baseline characteristics, which was observed or was unclear in four studies (Garcia‐Vega 2004; Hawkes 2001; Mizrahi 2012; Tapete 2019), and that further impacted the quality of the evidence.

Potential biases in the review process

Clinical heterogeneity is a key area of concern in this review. As previously discussed, the inclusion of studies focusing exclusively on CD patients together with a majority of studies that included CD and UC patients as an indecipherable IBD cohort, as well as the variety of disease states, reflects the evidence as a whole, but ignores the issue of clinical heterogeneity to some extent. It would have been possible to exclude studies that did not differentiate between CD and UC, and this was a discussion point for the team when completing the review. If we decided to exclude these studies from this review, and similarly exclude them from our concomitant UC review, a potentially important body of evidence would have been unaccounted for. We felt that representing this evidence base was vital, together with accepting and transparently presenting the difficulties this led to, related to clinical heterogeneity. On balance, given the extremely limited implications of the current evidence base, we decided to include studies that included CD and UC patients as one IBD cohort and use them for our analysis (we also decided that in our UC review, data exclusively on UC would be included). However, as further data become available, this would not be the approach the team or indeed readers would endorse and is a potential source of bias. As this has not clearly been defined within the original protocol, it must be recognised as a potential source of bias in the review.

Agreements and disagreements with other studies or reviews

This is the first Cochrane Review on this topic.

Considering the international guidelines for IBD, few of the major societies mention the treating of pain in IBD.

The recent British Society of Gastroenterology guidelines do offer recommendations (BSG 2019), citing several of the studies included in this review. They state that psychological interventions may be useful as adjunct therapy, describing this as a weak recommendation with low‐quality evidence, which would be supported by the evidence in this review. They do not define such psychological interventions, and do not comment on any of the other intervention types included in this review.

The current UK National Institute for Health and Care Excellence (NICE) guidelines do not discuss pain relief as a standalone treatment goal (NICE 2019). The American Gastroenterological Association guidelines make no mention of guidelines in this area (AGA 2020).

The European Crohn's and Colitis Organisation guidelines also do not mention therapies in this area (ECCO 2020).

Figure 1

Study flow diagram.

Figure 2

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Figure 3

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Analysis 1.1

Comparison 1: Low FODMAP diet versus sham diet, Outcome 1: Pain frequency defined by days of pain by IBS‐SSS scores (0 to 100) for IBD

Analysis 1.2

Comparison 1: Low FODMAP diet versus sham diet, Outcome 2: Days with moderate or severe pain by GSRS scores for IBD (0 to 3) for IBD

Analysis 1.3

Comparison 1: Low FODMAP diet versus sham diet, Outcome 3: Pain frequency defined by days of pain by IBS‐SSS scores (0 to 100) for CD

Analysis 1.4

Comparison 1: Low FODMAP diet versus sham diet, Outcome 4: Pain intensity IBD (0 to 100)

Analysis 1.5

Comparison 1: Low FODMAP diet versus sham diet, Outcome 5: Pain intensity IBD (0 to 3)

Analysis 1.6

Comparison 1: Low FODMAP diet versus sham diet, Outcome 6: Pain intensity CD (0 to 100)

Analysis 1.7

Comparison 1: Low FODMAP diet versus sham diet, Outcome 7: Withdrawals due to adverse events

Analysis 2.1

Comparison 2: Medicine‐separated moxibustion combined with acupuncture versus wheat bran‐separated moxibustion combined with shallow acupuncture, Outcome 1: Withdrawals due to adverse events

Analysis 3.1

Comparison 3: Mindfulness with cognitive behavioural therapy (CBT) versus no treatment (both groups received standard medical therapy), Outcome 1: Pain frequency and severity of abdominal discomfort, severity of abdominal bloating, satisfaction with bowel habit, and impact of symptoms on life in general

Analysis 3.2

Comparison 3: Mindfulness with cognitive behavioural therapy (CBT) versus no treatment (both groups received standard medical therapy), Outcome 2: Withdrawals due to adverse events

Analysis 4.1

Comparison 4: soft non‐manipulative osteopathic treatment versus no intervention, Outcome 1: Pain intensity

Analysis 4.2

Comparison 4: soft non‐manipulative osteopathic treatment versus no intervention, Outcome 2: Withdrawals due to adverse events

Analysis 5.1

Comparison 5: Directed stress management versus standard treatment, Outcome 1: Pain intensity

Analysis 6.1

Comparison 6: Self‐directed stress management vs conventional therapy, Outcome 1: Pain intensity

Analysis 7.1

Comparison 7: Enteric‐release glyceryl trinitrate versus placebo, Outcome 1: Withdrawals due to adverse events

Analysis 8.1

Comparison 8: 100 mg olorinab 3 times/day versus 25 mg olorinab 3 times/day, Outcome 1: Pain intensity

Analysis 8.2

Comparison 8: 100 mg olorinab 3 times/day versus 25 mg olorinab 3 times/day, Outcome 2: Withdrawals due to adverse events

Analysis 9.1

Comparison 9: Relaxation training versus waitlist, Outcome 1: Pain intensity

Analysis 10.1

Comparison 10: Web‐based education versus standard book‐based education, Outcome 1: Pain intensity

Analysis 10.2

Comparison 10: Web‐based education versus standard book‐based education, Outcome 2: Withdrawals due to adverse events

Analysis 11.1

Comparison 11: Yoga intervention versus no treatment (both groups received standard medical therapy), Outcome 1: Withdrawals due to adverse events

Analysis 12.1

Comparison 12: Transcranial direct current stimulation versus sham stimulation, Outcome 1: Pain intensity (0 to 100)

Analysis 12.2

Comparison 12: Transcranial direct current stimulation versus sham stimulation, Outcome 2: Withdrawals due to adverse events

Analysis 13.1

Comparison 13: Kefir versus no intervention, Outcome 1: CD pain intensity (0 to 3)

Analysis 13.2

Comparison 13: Kefir versus no intervention, Outcome 2: IBD pain intensity (0 to 3)

Analysis 13.3

Comparison 13: Kefir versus no intervention, Outcome 3: Withdrawals due to adverse events

Summary of findings 1. Low FODMAP diet compared to sham diet for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Low FODMAP diet compared to sham diet for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with inflammatory bowel disease Setting: multicentre, 2 gastroenterology clinics in the UK and an unstated setting in Italy Intervention: low FODMAP diet Comparison: sham diet
Outcomes	Risk with sham diet	Risk with low FODMAP diet	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Note measured
Pain frequency IBD (measured in days of pain on the IBS‐SSS questionnaire)	‐	MD 2 lower (15.86 lower to 11.86 higher)	‐	52 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Pain frequency IBD (measured in days with moderate or severe pain on the GSRS questionnaire)	‐	MD 0.4 higher (0.44 lower to 1.24 higher)	‐	52 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Pain frequency CD (measured in days of pain on the IBS‐SSS questionnaire)	‐	MD 12 lower (114.55 lower to 90.55 higher)	‐	26 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Pain intensity IBD (0‐10cm visual analogue scale)	‐	MD 8.46 lower (15.76 lower to 1.16 lower)	‐	82 (2 studies)	⊕⊝⊝⊝ very low^{a b}
Pain intensity IBD (0‐3 point GSRS scale)	‐	MD 8 lower (66.27 lower to 50.27 higher)	‐	26 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Pain intensity CD (0‐10cm visual analogue scale)	‐	MD 0.2 higher (8.67 lower to 9.07 higher)	‐	52 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Withdrawal due to adverse events	4 per 1000	7.4 per 1000 (1 to 77)	RR 1.85 (0.18 to 19.19)	52 (1 study)	⊕⊝⊝⊝ very low^{a b}
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CD: Crohn's disease; CI: confidence interval; FODMAP: fermentable oligo‐, di‐, monosaccharides and polyols; GSRS: Gastrointestinal Symptom Rating Scale; IBD: inflammatory bowel disease; IBS‐SSS: Irritable Bowel Syndrome Severity Scoring System; MD: mean difference; RR: risk ratio
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded one level due to risk of bias. ^bDowngraded two levels due to imprecision from very sparse data.

Summary of findings 1. Low FODMAP diet compared to sham diet for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Medicine‐separated moxibustion combined with acupuncture compared with wheat bran‐separated moxibustion combined with shallow acupuncture for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with Crohn's disease Settings: Shanghai Acupuncture Meridian Institute Medical Clinic Acupuncture Inflammatory Bowel Disease Specialist Clinic, Zhongshan Hospital Affiliated Endoscopic Center, and Yueyang Hospital, Shanghai University of Traditional Chinese Medicine Intervention: medicine‐separated moxibustion combined with acupuncture Comparison: wheat bran‐separated moxibustion combined with shallow acupuncture
Outcomes	Risk with wheat bran‐separated moxibustion combined with shallow acupuncture	Risk with medicine‐separated moxibustion combined with acupuncture	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Abdominal pain frequency or change in frequency	‐	‐	‐	‐	Not measured
Abdominal pain intensity or change in intensity	‐	‐	‐	‐	Not measured
Withdrawals due to adverse events	0 per 1000	0 per 1000 (0 to 0)	Not estimable	102 (1 study)	⊕⊝⊝⊝ very low ^{a b}
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded two levels due to imprecision from very sparse data. ^bDowngraded one level due to risk of bias.

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Mindfulness with cognitive behavioural therapy (CBT) versus no treatment (both groups received standard medical therapy) for the management of abdominal pain in Crohn's Disease and inflammatory bowel disease
Patient or population: people with inflammatory bowel disease Setting: multicentre, hospitals in the UK Intervention: mindfulness with CBT Comparison: no treatment
Outcomes	Risk with standard medical therapy	Risk with mindfulness with CBT + standard medical therapy	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Pain frequency and intensity (measured as part of the 0‐500 point IBS‐SSS questionnaire)	‐	MD 37 lower (87.29 lower to 13.29 higher)	‐	66 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Withdrawal due to adverse events	0 per 1000	0 per 1000 (0 to 0)	Not estimable	66 (1 study)	⊕⊝⊝⊝ very low ^{a b}
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CBT: cognitive behavioural therapy; CI: confidence interval; MD: mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded two levels due to imprecision from very sparse data. ^bDowngraded one level due to risk of bias.

Summary of findings 4. soft non‐manipulative osteopathic treatment compared to no intervention for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Soft non‐manipulative osteopathic treatment compared to no intervention for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with Crohn's disease Setting: single centre, hospital in Spain Intervention: soft non‐manipulative osteopathic Comparison: no intervention
Outcomes	Risk with no intervention	Risk with soft non‐manipulative osteopathic	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Abdominal pain frequency or change in frequency	‐	‐	‐	‐	Not measured
Pain intensity (0‐10cm visual analogue scale)	‐	MD 0.01 higher (1.81 lower to 1.83 higher)	‐	30 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Withdrawals due to adverse events	0 per 1000	0 per 1000 (0 to 0)	Not estimable	30 (1 study)	⊕⊝⊝⊝ very low ^{a b}
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded two levels due to imprecision from very sparse data. ^bDowngraded one level due to risk of bias.

Summary of findings 4. soft non‐manipulative osteopathic treatment compared to no intervention for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Stress management (it is unclear whether these interventions replaced standard treatment or were added to standard treatment) versus standard treatment for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with Crohn's disease Setting: single centre, Inflammatory Intestinal Disease Unit of Asturias Central Hospital, Spain Intervention: stress management Comparison: no stress management
Outcomes	Risk with no stress management	Risk with stress management	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Pain frequency or change in pain frequency	‐	‐	‐	‐	Not measured
Pain intensity (author‐derived formula based on a 1‐3 point scale)	‐	MD 34.3 lower (61.99 lower to 6.61 lower)	‐	30 (1 study)	⊕⊝⊝⊝ very low^{a b}
Withdrawals due to adverse events	‐	‐	‐	‐	Not measured
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded two levels due to imprecision from very sparse data. ^bDowngraded one level due to risk of bias.

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Self‐directed stress management (it is unclear whether these interventions replaced standard treatment or were added to standard treatment) versus standard treatment for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with Crohn's disease Setting: single centre, Inflammatory Intestinal Disease Unit of Asturias Central Hospital, Spain Intervention: stress management Comparison: no stress management
Outcomes	Risk with no stress management	Risk with stress management	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Pain frequency or change in pain frequency	‐	‐	‐	‐	Not measured
Pain intensity (author‐derived formula based on a 1‐3 point scale)	‐	MD 30.5 lower (58.45 lower to 2.55 lower)	‐	30 (1 study)	⊕⊝⊝⊝ very low^{a b}
Withdrawals due to adverse events	‐	‐	‐	‐	Not measured
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded two levels due to imprecision from very sparse data. ^bDowngraded one level due to risk of bias.

Summary of findings 7. Enteric‐release glyceryl trinitrate compared to placebo for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Enteric‐release glyceryl trinitrate compared to placebo for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with Crohn's disease Setting: unstated (centres in the UK) Intervention: enteric‐release glyceryl trinitrate Comparison: placebo
Outcomes	Risk with placebo	Risk with enteric‐release glyceryl trinitrate	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Treatment success as defined by the authors	‐	‐	‐	‐	‐	Not measured
Abdominal pain frequency or change in frequency	‐	‐	‐	‐	‐	Not measured
Abdominal pain intensity or change in intensity	‐	‐	‐	‐	‐	Not measured
Withdrawal due to adverse events	Study population		RR 3.18 (0.94 to 10.76)	70 (1 study)	⊕⊝⊝⊝ very low ^{a b}	IG: Headache = 1, worsening clinical condition = 4, generalised rash = 1, mood change/irritability = 1, loss of consciousness/memory = 1 CG: Headache = 2, worsening clinical condition = 1
Withdrawal due to adverse events	83 per 1000	265 per 1000 (78 to 897)	RR 3.18 (0.94 to 10.76)	70 (1 study)	⊕⊝⊝⊝ very low ^{a b}
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded one level due to risk of bias. ^bDowngraded two levels due to imprecision from very sparse data. IG: Intervention group CG: Control group

Summary of findings 7. Enteric‐release glyceryl trinitrate compared to placebo for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Summary of findings 8. 100 mg olorinab 3 times/day compared to 25 mg olorinab 3 times/day for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
100 mg olorinab 3 times/day compared to 25 mg olorinab 3 times/day for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with Crohn's disease Setting: unstated (multicentre, USA) Intervention: 100 mg olorinab 3 times/day Comparison: 25 mg olorinab 3 times/day
Outcomes	Risk with 25 mg olorinab 3 times/day	Risk with 100 mg olorinab 3 times/day	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Pain frequency or change in pain frequency	‐	‐	‐	‐	Not measured
Pain intensity (30% reduction in weekly AAPS)	Study population		RR 0.66 (0.38 to 1.15)	14 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Pain intensity (30% reduction in weekly AAPS)	1000 per 1000	660 per 1000 (380 to 1000)	RR 0.66 (0.38 to 1.15)	14 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Withdrawal due to adverse events	0 per 1000	0 per 1000 (0 to 0)	Not estimable	14 (1 study)	⊕⊝⊝⊝ very low ^{a b}
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded one level due to risk of bias. ^bDowngraded two levels due to imprecision from very sparse data.

Summary of findings 8. 100 mg olorinab 3 times/day compared to 25 mg olorinab 3 times/day for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Summary of findings 9. Relaxation training compared to waitlist for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Relaxation training compared to waitlist for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with inflammatory bowel disease Setting: Hadassah Medical Center in Jerusalem Intervention: relaxation training Comparison: waitlist
Outcomes	Risk with waitlist	Risk with relaxation training	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Pain frequency or change in pain frequency	‐	‐	‐	‐	Not measured
Pain intensity (0‐10cm visual analogue scale)	‐	MD 0.72 lower (1.85 lower to 0.41 higher)	‐	56 (1 study)	⊕⊝⊝⊝ very low ^{a b}
Withdrawals due to adverse effects	‐	‐	‐	‐	Not measured
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded one level due to risk of bias. ^bDowngraded two levels due to imprecision from very sparse data.

Summary of findings 9. Relaxation training compared to waitlist for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Summary of findings 10. Web‐based education compared to standard book‐based education for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Web‐based education compared to standard book‐based education for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with inflammatory bowel disease Setting: single centre, gastroenterology unit in Turkey Intervention: web‐based education Comparison: standard book‐based education
Outcomes	Risk with standard book‐based education	Risk with web‐based education	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Abdominal pain frequency or change in frequency	‐	‐	‐	‐	Not measured
Pain intensity (0‐10cm visual analogue scale)	‐	MD 0.13 lower (1.25 lower to 0.99 higher)	‐	60 (1 study)	⊕⊕⊝⊝ very low ^{a b}
Withdrawals due to adverse events	0 per 1000	0 per 1000 (0 to 0)	Not estimable	60 (1 study)	⊕⊝⊝⊝ very low^{a b}
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded one level due to risk of bias. ^bDowngraded two levels due to imprecision from very sparse data.

Summary of findings 10. Web‐based education compared to standard book‐based education for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Yoga intervention compared to no treatment (both groups received standard medical therapy) for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with inflammatory bowel disease Settings: single centre, All India Institute of Medical Science (AIIMS), New Delhi, India Intervention: yoga Comparison: no yoga
Outcomes	Risk with standard medical therapy	Risk with yoga plus standard medical therapy	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Abdominal pain frequency or change in frequency	‐	‐	‐	‐	Not measured
Abdominal pain intensity or change in intensity	‐	‐	‐	‐	Not measured
Withdrawal due to adverse events	0 per 1000	0 per 1000 (0 to 0)	Not estimable	100 (1 study)	⊕⊝⊝⊝ very low ^{a b}
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded one level due to risk of bias. ^bDowngraded two levels due to imprecision from very sparse data.

Summary of findings 12. Transcranial direct current stimulation compared to sham stimulation for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Transcranial direct current stimulation compared to sham stimulation for the management of abdominal pain in Crohn's disease and inflammatory bowel disease
Patient or population: people with inflammatory bowel disease Setting: single centre, Medical Department I (Gastroenterology, Infectious Diseases, Rheumatology) of the Charite‐Campus Benjamin Franklin, Germany Intervention: transcranial direct current stimulation Comparison: sham stimulation
Outcomes	Risk with sham stimulation	Risk with transcranial direct current stimulation	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Abdominal pain frequency or change in frequency	‐	‐	‐	‐	Not measured
Pain intensity (0‐10cm visual analogue scale)	‐	MD 1.65 lower (3.29 lower to 0.01 lower)	‐	20 (1 study)	⊕⊕⊝⊝ low ^a
Withdrawal due to adverse events	0 per 1000	0 per 1000 (0 to 0)	Not estimable	20 (1 study)	⊕⊕⊝⊝ low ^a
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference; PPT: Pressure Pain Threshold
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded two levels due to imprecision from very sparse data.

Summary of findings 12. Transcranial direct current stimulation compared to sham stimulation for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Summary of findings 13. Kefir ( Lactobacillus bacteria) compared to no intervention for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
*Kefir (Lactobacillus* bacteria) compared to no intervention for the management of abdominal pain in Crohn's disease and inflammatory bowel disease**
Patient or population: people with inflammatory bowel disease Setting: unstated (single centre, Turkey) Intervention: kefir Comparison: no intervention
Outcomes	Risk with no intervention	Risk with kefir	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)
Treatment success as defined by the authors	‐	‐	‐	‐	Not measured
Abdominal pain frequency or change in frequency	‐	‐	‐	‐	Not measured
Pain intensity IBD (4‐point rating scale from 0 to 3)	‐	MD 0.62 higher (0.17 higher to 1.07 higher)	‐	48 (1 study)	⊕⊕⊝⊝ very low ^{a b}
Pain intensity CD (4‐point rating scale from 0 to 3)	‐	MD 1.10 lower (1.67 lower to 0.53 lower)	‐	20 (1 study)	⊕⊕⊝⊝ very low ^{a b}
Withdrawals due to adverse events	0 per 1000	0 per 1000 (0 to 0)	Not estimable	20 (1 study)	⊕⊝⊝⊝ very low ^{a b}
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CD: Crohn's disease; CI: confidence interval; IBD: inflammatory bowel disease; MD: mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aDowngraded one level due to risk of bias. ^bDowngraded two levels due to imprecision from very sparse data.

Summary of findings 13. Kefir ( Lactobacillus bacteria) compared to no intervention for the management of abdominal pain in Crohn's disease and inflammatory bowel disease

Table 1. Primary outcome details

Comparison	Study ID	Disease type	Disease activity	Length of intervention	Measurement of pain	Number of randomised participants
Low FODMAP diet vs sham diet	Cox 2020	CD/IBD	Inactive	4 weeks	Pain frequency and intensity: IBS‐SSS for pain rating scale 0 to 100, GSRS rating scale 0 to 3	52 (IG: 27; CG: 25)
Low FODMAP diet vs sham diet	Tapete 2018	IBD	Inactive	6 to 8 weeks	Pain intensity: VAS rating scale 0 to 10 cm	30 (IG: 10; CG: 20)
Low FODMAP diet vs high FODMAP/normal diet	Tapete 2019	IBD	Inactive	8 weeks cross‐over (4 weeks x 2)	Pain intensity: VAS rating scale 0 to 10 cm	50 (IG: 25; CG: 25)
Medicine‐separated moxibustion combined with acupuncture vs wheat bran‐separated moxibustion combined with shallow acupuncture	Bao 2016	CD	Mild to moderate	12 weeks	Pain frequency and intensity: Traditional Chinese Medicine rating scale 0 to 3	102 (IG: 51; CG: 51)
Mindfulness with CBT + standard medical therapy vs standard medical therapy	Berrill 2014	IBD	Inactive	12 months	Pain frequency and severity of abdominal discomfort, severity of abdominal bloating, satisfaction with bowel habit, and impact of symptoms on life in general. Each domain is scored 0 to 100, and an overall score of 0 to 500 is obtained.	66 (IG: 33; CG: 33)
Soft non‐manipulative osteopathic vs no intervention	Espi Lopez 2018	CD	Unclear	30 days	Pain intensity: VAS rating scale 0 to 10 cm	30 (IG: 16; CG: 14)
Stress management vs self‐directed stress management vs standard treatment	Garcia‐Vega 2004	CD	Inactive	8 weeks	Pain intensity: rating scale 1 to 3 and author formulas	45 (IG1: 15; IG2: 15; CG: 15)
Enteric‐release glyceryl trinitrate vs placebo	Hawkes 2001	CD	Moderate to severe	12 weeks	Pain intensity: rating scale 0 to 3	70 (IG: 34; CG: 36)
100 mg olorinab 3 times/day vs 25 mg olorinab 3 times/day	Higgins 2019	CD	Inactive to mild	8 weeks	Pain intensity: AAPS of 0‐to‐10 Likert scale	14 (IG: 8; CG: 6)
Relaxation training vs waitlist	Mizrahi 2012	IBD	Active	5 weeks	Pain intensity: VAS rating scale 0 to 10 cm	56 (IG: 28; CG: 28)
Web‐based education vs standard book‐based education	Ozgursoy Uran 2019	IBD	Mix of active and inactive	8 weeks	Pain intensity: VAS rating scale 0 to 10 cm	60 (IG: 10; CG: 10)
Yoga intervention + standard medical therapy vs standard medical therapy	Sharma 2015	IBD	Inactive	8 weeks	Presence or absence of pain	100 (IG: 50; CG: 50)
Transcranial direct current stimulation vs sham stimulation	Volz 2016	IBD	Active	5 days	Pain intensity: Pressure Pain Threshold taken with algometer, VAS rating scale 0 to 10 cm	20 (IG: 10; CG: 10)
Kefir diet (Lactobacillus bacteria) vs no intervention	Yilmaz 2019	CD/IBD	Inactive to moderate	4 weeks	Pain intensity: rating scale 0 to 3	48 (IG: 28; CG: 20)
AAPS: average abdominal pain score CBT: cognitive behavioural therapy CD: Crohn's disease CG: control group FODMAP: fermentable oligo‐, di‐, monosaccharides and polyols IG: intervention group GSRS: Gastrointestinal Symptom Rating Scale IBD: inflammatory bowel disease IBS‐SSS: Irritable Bowel Syndrome Severity Scoring System VAS: visual analogue scale

Table 1. Primary outcome details

Table 2. Primary outcome data

Comparison	Study ID	Treatment success end of study data IG/CG	Pain frequency end of study data IG/CG	Pain intenstiy end of study data IG/CG	Withdrawals due to adverse events IG/CG
Low FODMAP diet vs sham diet	Cox 2020	NR	IBS‐SSS all participants mean(SD) IG= 36(26); CG= 38(25) IBS‐SSS CD mean (SD) IG = 36(138.4); CG=48(128.2) GSRS all participants mean(SD) IG= 1.5(1.6); CG= 1.1(1.5)	IBS‐SSS all participants mean(SD) IG=22(15.6) ; CG=30(15) IBS‐SSS CD mean (SD) IG = 24(82.3); CG= 32(69.3) GSRS all participants mean(SD) IG=0.9(2.6) ; CG= 0.7(22.5)	IG: 2 CG: 1
Low FODMAP diet vs sham diet	Tapete 2018	NR	NR	mean(SD): IG=3.3 (1.9) CG= 4.3(2.2)	NR
Low FODMAP diet vs high FODMAP/normal diet	Tapete 2019	NR	NR	mean(SD): IG=1.1(1.6) CG=3.1(2.3) (unclear if average of both cross‐over arms or from the first or second phase of the cross‐over)	NR
Medicine‐separated moxibustion combined with acupuncture vs wheat bran‐separated moxibustion combined with shallow acupuncture	Bao 2016	NR	mean (no SD provided) IG=0 CG=0	mean (no SD provided) IG=2 CG=1	IG=0 CG=0
Mindfulness with CBT + standard medical therapy vs standard medical therapy	Berrill 2014	NR	Frequency and intensity measured together as part of the IBS‐SSS score mean(SD) IG= 187 (97) CG= 224 (111)		IG=0 CG=0
Soft non‐manipulative osteopathic vs no intervention	Espi Lopez 2018	NR	NR	mean(SD): IG=2.72(2.66) CG=2.71(2.43)	IG=0 CG=0
Stress management vs self‐directed stress management vs standard treatment	Garcia‐Vega 2004	NR	NR	mean(SD): IG=13.3(28) CG=47.6(47)	NR
Enteric‐release glyceryl trinitrate vs placebo	Hawkes 2001	NR	NR	mean(no SD presented) IG=8.1 CG=8.6	IG=9 CG=3
100 mg olorinab 3 times/day vs 25 mg olorinab 3 times/day	Higgins 2019	≥ 30% reduction in weekly AAPS IG: 5 CG: 6	NR	1.5 hours postdose mean(SD could not be calculated from figure) IG=1.9 CG= 1.9 Mean change in AAPS (no SD) IG= ‐4.6 CG= ‐4.6	IG=0 CG=0
Relaxation training vs waitlist	Mizrahi 2012	NR	NR	mean(SD): IG= 2.23(1.83) CG= 2.95(2.44)	NR
Web‐based education vs standard book‐based education	Ozgursoy Uran 2019	NR	NR	mean(SD): IG= 1.8(2.04) CG= 1.93(2.39)	IG=0 CG=0
Yoga intervention + standard medical therapy vs standard medical therapy	Sharma 2015	NR	NR	NR	IG=0 CG=0
Transcranial direct current stimulation vs sham stimulation	Volz 2016	NR	NR	mean(SD): IG= 2.8(2.3) CG= 4.45(1.3)	IG=0 CG=0
Kefir diet (Lactobacillus bacteria) vs no intervention	Yilmaz 2019	NR	NR	CD pain intensity mean(SD) IG= 0.2(0.63) CG= 1.3(0.67) IBD pain intensity mean(SD) IG= 0.9(0.97) CG= 0.28(0.61)	IG=0 CG=0
AAPS: average abdominal pain score CBT: cognitive behavioural therapy CD: Crohn's disease CG: control group FODMAP: fermentable oligo‐, di‐, monosaccharides and polyols IG: intervention group GSRS: Gastrointestinal Symptom Rating Scale IBD: inflammatory bowel disease IBS‐SSS: Irritable Bowel Syndrome Severity Scoring System VAS: visual analogue scale

Table 2. Primary outcome data

Comparison 1. Low FODMAP diet versus sham diet

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Pain frequency defined by days of pain by IBS‐SSS scores (0 to 100) for IBD Show forest plot	1	52	Mean Difference (IV, Random, 95% CI)	‐2.00 [‐15.86, 11.86]

1.2 Days with moderate or severe pain by GSRS scores for IBD (0 to 3) for IBD Show forest plot	1	52	Mean Difference (IV, Random, 95% CI)	0.40 [‐0.44, 1.24]

1.3 Pain frequency defined by days of pain by IBS‐SSS scores (0 to 100) for CD Show forest plot	1	26	Mean Difference (IV, Random, 95% CI)	‐12.00 [‐114.55, 90.55]

1.4 Pain intensity IBD (0 to 100) Show forest plot	2	82	Mean Difference (IV, Random, 95% CI)	‐8.46 [‐15.76, ‐1.16]

1.5 Pain intensity IBD (0 to 3) Show forest plot	1	52	Mean Difference (IV, Random, 95% CI)	0.20 [‐8.67, 9.07]

1.6 Pain intensity CD (0 to 100) Show forest plot	1	26	Mean Difference (IV, Random, 95% CI)	‐8.00 [‐66.27, 50.27]

1.7 Withdrawals due to adverse events Show forest plot	1	52	Risk Ratio (M‐H, Random, 95% CI)	1.85 [0.18, 19.19]

Comparison 1. Low FODMAP diet versus sham diet

Comparison 2. Medicine‐separated moxibustion combined with acupuncture versus wheat bran‐separated moxibustion combined with shallow acupuncture

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 Withdrawals due to adverse events Show forest plot	1	102	Risk Ratio (M‐H, Random, 95% CI)	Not estimable

Comparison 2. Medicine‐separated moxibustion combined with acupuncture versus wheat bran‐separated moxibustion combined with shallow acupuncture

Comparison 3. Mindfulness with cognitive behavioural therapy (CBT) versus no treatment (both groups received standard medical therapy)

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
3.1 Pain frequency and severity of abdominal discomfort, severity of abdominal bloating, satisfaction with bowel habit, and impact of symptoms on life in general Show forest plot	1	66	Mean Difference (IV, Random, 95% CI)	‐37.00 [‐87.29, 13.29]

3.2 Withdrawals due to adverse events Show forest plot	1	66	Risk Ratio (M‐H, Random, 95% CI)	Not estimable

Comparison 3. Mindfulness with cognitive behavioural therapy (CBT) versus no treatment (both groups received standard medical therapy)

Comparison 4. soft non‐manipulative osteopathic treatment versus no intervention

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
4.1 Pain intensity Show forest plot	1	30	Mean Difference (IV, Random, 95% CI)	0.01 [‐1.81, 1.83]

4.2 Withdrawals due to adverse events Show forest plot	1	30	Odds Ratio (M‐H, Random, 95% CI)	Not estimable

Comparison 4. soft non‐manipulative osteopathic treatment versus no intervention

Comparison 5. Directed stress management versus standard treatment

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
5.1 Pain intensity Show forest plot	1	30	Mean Difference (IV, Random, 95% CI)	‐34.30 [‐61.99, ‐6.61]

Comparison 5. Directed stress management versus standard treatment

Comparison 6. Self‐directed stress management vs conventional therapy

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
6.1 Pain intensity Show forest plot	1	30	Mean Difference (IV, Random, 95% CI)	‐30.50 [‐58.45, ‐2.55]

Comparison 6. Self‐directed stress management vs conventional therapy

Comparison 7. Enteric‐release glyceryl trinitrate versus placebo

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
7.1 Withdrawals due to adverse events Show forest plot	1	70	Risk Ratio (M‐H, Random, 95% CI)	3.18 [0.94, 10.76]

Comparison 7. Enteric‐release glyceryl trinitrate versus placebo

Comparison 8. 100 mg olorinab 3 times/day versus 25 mg olorinab 3 times/day

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
8.1 Pain intensity Show forest plot	1	14	Risk Ratio (M‐H, Random, 95% CI)	0.66 [0.38, 1.15]

8.2 Withdrawals due to adverse events Show forest plot	1	14	Odds Ratio (M‐H, Random, 95% CI)	Not estimable

Comparison 8. 100 mg olorinab 3 times/day versus 25 mg olorinab 3 times/day

Comparison 9. Relaxation training versus waitlist

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
9.1 Pain intensity Show forest plot	1	56	Mean Difference (IV, Random, 95% CI)	‐0.72 [‐1.85, 0.41]

Comparison 9. Relaxation training versus waitlist

Comparison 10. Web‐based education versus standard book‐based education

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
10.1 Pain intensity Show forest plot	1	60	Mean Difference (IV, Random, 95% CI)	‐0.13 [‐1.25, 0.99]

10.2 Withdrawals due to adverse events Show forest plot	1	60	Risk Ratio (M‐H, Random, 95% CI)	Not estimable

Comparison 10. Web‐based education versus standard book‐based education

Comparison 11. Yoga intervention versus no treatment (both groups received standard medical therapy)

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
11.1 Withdrawals due to adverse events Show forest plot	1	100	Risk Ratio (M‐H, Random, 95% CI)	Not estimable

Comparison 11. Yoga intervention versus no treatment (both groups received standard medical therapy)

Comparison 12. Transcranial direct current stimulation versus sham stimulation

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
12.1 Pain intensity (0 to 100) Show forest plot	1	20	Mean Difference (IV, Random, 95% CI)	‐1.65 [‐3.29, ‐0.01]

12.2 Withdrawals due to adverse events Show forest plot	1	20	Odds Ratio (M‐H, Random, 95% CI)	Not estimable

Comparison 12. Transcranial direct current stimulation versus sham stimulation

Comparison 13. Kefir versus no intervention

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
13.1 CD pain intensity (0 to 3) Show forest plot	1	20	Mean Difference (IV, Random, 95% CI)	‐1.10 [‐1.67, ‐0.53]

13.2 IBD pain intensity (0 to 3) Show forest plot	1	48	Mean Difference (IV, Random, 95% CI)	0.62 [0.17, 1.07]

13.3 Withdrawals due to adverse events Show forest plot	1	20	Risk Ratio (M‐H, Random, 95% CI)	Not estimable

Comparison 13. Kefir versus no intervention