Sage Journals: Discover world-class research

Abstract

Background:

Clonidine is considered an alternative treatment for refractory diarrhoea. The evidence in the literature is scarce and not conclusive. The present paper’s purpose is to gather available evidence and provide a systematic answer regarding the effectiveness of clonidine for diarrhoea.

Method:

We performed a systematic review of clonidine and its effect on diarrhoea. Meta-analysis was performed with a random effects model of the standardized mean difference (SMD) or the weighted mean difference and heterogeneity was quantified with I² and publication bias was assessed with Egger’s and Begg’s test. Subgroup analyses and meta-regression were performed to investigate sources of heterogeneity. Any empirical study describing use of clonidine for diarrhoea in humans independent of age was included. For the meta-analysis, papers had to provide sufficient data to produce an effect measure, while case reports were not included in the meta-analysis and are discussed narratively only.

Results:

A total of 24 trials and seven case reports were identified. Clonidine (median dose 300 μg/day) has been used for treatment of diarrhoea in irritable bowel syndrome, faecal incontinence, diabetes, withdrawal-associated diarrhoea, intestinal failure, neuroendocrine tumours and cholera; studies were also performed on healthy volunteers. Results indicate a strong effect of clonidine on diarrhoea (SMD = −1.02, 95% confidence interval [CI] −1.46 to −0.58) with a decrease of stool volume by 0.97 l/day, stool frequency by 0.4 times/day and increase in transit time by 31 minutes. In a sensitivity analysis of studies with functional diarrhoea and sample size over 10 subjects, the effect was similar −0.99 (95% CI −1.54 to −0.43). There is however significant heterogeneity and publication bias. Heterogeneity decreased in subgroup analyses by condition but not with other factors examined. A limitation of the present study includes small study effects.

Conclusion:

Clonidine is effective for treatment of diarrhoea and should be considered as an alternative when all other medications have failed.

Keywords

clonidine diarrhoea intestinal transit meta-analysis

Introduction

Diarrhoea means different things to different patients: loose stool consistency and frequent bowel movements are prime characteristics. Diarrhoea is said to be chronic when it lasts more than 4 weeks. Chronic diarrhoea is an important medical condition that troubles patients [Schiller, 2005]. Treating diarrhoea is a difficult issue for the gastroenterologist and the general physician who needs effective medications [Vohmann and Hoffmann, 2013]. The differential diagnosis is broad, and diagnostic evaluation may be complex. In spite of the large number of causes of chronic diarrhoea, and in spite of extensive investigations, no cause for the symptoms is found in a large group of patients, who are referred to as having functional diarrhoea. The precise prevalence of chronic diarrhoea is unknown [Tack, 2012]. Diarrhoea kills an estimated 2.5 million people each year, with about 60–70% of them being children under 5 years of age [Guarino et al. 2012].

The main causes of chronic diarrhoea seem to depend on the socioeconomic status of the population surveyed [Li and Vaziri, 2012; Schiller, 2012]. In developed countries, the most frequent diagnoses made in patients with chronic diarrhoea are irritable bowel syndrome, idiopathic inflammatory bowel disease, malabsorption syndrome, chronic infections and idiopathic secretory diarrhoea. In less-developed countries, chronic bacterial, mycobacterial and parasitic infections are the most common causes of chronic diarrhoea; functional disorders, inflammatory bowel disease and malabsorption (from a variety of unspecified causes) are also common in this setting. Treatment of diarrhoea includes oral rehydration therapy, dietary modifications, antibiotics, absorbents, probiotics, anticholinergics, opiates and opiate-like medications, bile acid binders, budesonide, somatostatin analogues, alpha adrenergic agonists, proton pump inhibitors and antihistamines and calcium channel blockers [Halland and Talley, 2012; Kaiser and Surawicz, 2012; Li and Vaziri, 2012; Randall et al. 2012; Tack, 2012].

An important treatment of chronic diarrhoea which has proven resistant to standard treatments is clonidine, an alpha 2-adrenergic receptor agonist [Dharmsathaphorn, 1986; Fedorak and Field, 1987; Tack, 2012]. Activation of alpha 2-adrenergic receptors on enterocytes increases fluid and electrolyte absorption and inhibits secretion while at the same time decreasing bowel transit time [Schiller et al. 1985; Gregersen et al. 1989; Rubinoff et al. 1989]. In particular, clonidine can inhibit gastrointestinal motor activity by presynaptically inhibiting acetylcholine release from nerves in the myenteric plexus and at the neuromuscular junction, while it also has antinociceptive effects mediated by alpha-2 receptors in the spinal cord, brainstem, and forebrain [Unnerstall et al. 1984; Bharucha et al. 1997]. Clonidine reduced colonic and rectal tone, increased colonic and rectal compliance, and reduced colonic and rectal perception of distention in healthy subjects [Bharucha et al. 1997; Malcolm et al. 2000]. Secondary to the potential adverse effects including hypotension, this category should be cautiously used in treating diarrhoea especially in dehydrated patients. Due to the less-central hypotensive effect, clonidine is the medication of choice in this category. Clonidine has been used in treatment of chronic secretory diarrhoea of unknown aetiology, diarrhoea associated with narcotics withdrawal and diabetic diarrhoea, diarrhoea caused by chemotherapy or graft versus host disease, among many others [Li and Vaziri, 2012].

Until now, there has been no systematic assessment of the available literature on the effectiveness of clonidine for diarrhoea. This makes it difficult for clinicians to suggest this treatment as an option, since the literature is quite sparse and fragmented. A systematic review of this topic will provide answers for the clinician regarding clinical effectiveness and potential limitations.

Methods

Study eligibility criteria

The inclusion criteria were: any empirical study (abstract and full paper) describing treatment of diarrhoea with clonidine in humans independent of age. Diarrhoea was considered in a broad manner and anything that could be considered a proxy for diarrhoea was considered (e.g. stool volume changes, intestinal transit changes, etc.). Any studies with clonidine analogues (e.g. lidamidine) [Heredia Diaz et al. 1979, 1983; Heredia Diaz and Kajeyama Escobar, 1981; Goff, 1984; Edwards et al. 1986; Edwards and Read, 1986; Gasbarrini et al. 1986; Sninsky et al. 1986; Vicencio Tovar et al. 1986; Masukawa et al. 1987; Allison et al. 1988; Prior et al. 1988; Rodriguez Magallan et al. 1997; Awad et al. 2000] or with primary focus gastric emptying [Rosa-E-Silva et al. 1995; Asai et al. 1997; Thumshirn et al. 1999; Huilgol et al. 2002; Kwiatek et al. 2009] were excluded. There was no restriction to language of papers and the types of interventions included observational studies, randomized controlled trials, case series and case reports. For the meta-analysis papers had to provide sufficient data to produce an effect measure, while case reports were not included in the meta-analysis and are going to be discussed narratively only.

Search strategy and terms

PRISMA guidelines and MOOSE checklist for systematic reviews and meta-analyses were used [Stroup et al. 2000; Moher et al. 2009]. Electronic database searches were conducted in Google Scholar, PubMed/Medline, Scopus, EMBASE and Cochrane Library with no year limits. Publisher databases were also searched (Sciencedirect.com, link.springer.com, Wiley Library Online, Highwire press and Nature.com). The keywords for search were: clonidine, bowel, intestine, diarrhoea, diabetes, jejunostomy, ileostomy, irritable bowel syndrome, transit, stool, inflammatory bowel disease, cholera, infection, withdrawal. The date of search was 1 September 2015. The bibliographies from all included manuscripts and hand searching of relevant neurogastroenterology journals were used to identify further references.

Study selection, data extraction and quality assessment

The resulting studies (in abstract form) were assessed against the inclusion criteria. When there was insufficient information available in the abstract, the full text was reviewed. The investigators then extracted data from the selected studies including: author, year of publication, aim of the study, country, continent, sample size, mean age, male percentage, study design and results. The quality of studies (risk of bias) was assessed with Cochrane Collaboration’s tool [Higgins et al. 2008]. For the meta-analysis, studies were examined for p-values, means and standard deviations, correlation coefficients or other metrics depicting the effect of clonidine in diarrhoea. Metrics were converted to the standardized mean difference (SMD) [Lipsey and Wilson, 2001; Cooper et al. 2009] or weighted mean difference (WMD) where similar outcomes were identified [Borenstein et al. 2008]. Examples of outcomes included changes in stool volume, intestinal transit, and number of spontaneous bowel movements or any other relevant outcome depicting directly or indirectly improvement in diarrhoea.

Statistical analysis

Quantitative analysis was performed with Stata 12.0 (StataCorp LP, Texas) and Review Manager 5.3 (Cochrane Collaboration, Copenhagen). SMDs/WMDs were extracted from studies when available. The strength of association was categorized as following: small, SMD = 0.2; medium, SMD = 0.5; and large, SMD = 0.8 [Cohen, 1988; Faraone, 2008]; WMD has units and reflects the units of the outcome under description (e.g. litres per day for stool volume). A random effects model was used to produce a pooled estimate of the SMDs/WMDs. Statistical heterogeneity was assessed using Cochran’s Q test, which examines the null hypothesis that all studies are evaluating the same effect [Higgins et al. 2003]. Statistical significance for heterogeneity was set as p ⩽ 0.10. Heterogeneity was quantified using the I² statistic, indicating the percentage of total variation across studies that is due to heterogeneity rather than chance [Gillum et al. 2000; Higgins et al. 2003]. An I² value of 0% was considered to indicate no observed heterogeneity whilst a value >50% was considered to indicate substantial heterogeneity [Higgins and Thompson, 2002; Huedo-Medina et al. 2006; Bowden et al. 2011; Fragkos et al. 2014]. Heterogeneity was further investigated with subgroup analysis and meta-regression. Publication bias was assessed using funnel plots and Egger’s and Begg’s test [Begg and Mazumdar, 1994; Egger et al. 1997; Fragkos et al. 2016]. A funnel plot was created for the clinical measures with more than 10 studies [Sterne et al. 2011]. This is a scatter plot of the effect estimates from individual studies against a measurement of the study’s sample size or precision. Resemblance of a symmetrical inverted funnel supports that findings are due to sampling variation alone; thus, absence of bias [Sterne et al. 2011]. Sensitivity analyses were performed to quantify the effect of clonidine on diarrhoea when restricted to trials/case series with patient groups and sample size over 10 subjects. We carried out a cumulative meta-analysis of the effect of clonidine on diarrhoea based on the date of publication and condition.

Results

Systematic review and meta-analysis of trials

Studies and quality assessment

The systematic review of the literature resulted in 24 eligible trials/case series and seven eligible case reports (Figure 1 and Table 1). The case reports [Mcarthur et al. 1982; Sacerdote, 1986; Zimmermann et al. 1986; Roof, 1987; Migliore et al. 1988; Scholz et al. 1991; Schworer et al. 1995] are discussed narratively in a separate section below because of possible high placebo rates and are not included in the meta-analysis. The remaining 24 trials/case series included 10 studies with healthy volunteers (total sample size = 150 participants) [Schiller et al. 1985; Baxter et al. 1987; Pressman et al. 1987; Baumer et al. 1989; Gregersen et al. 1989; Rubinoff et al. 1989; Thollander et al. 1997, 1999; Mann and Shinkle, 1998; Viramontes et al. 2001] and 14 studies (total sample size = 493 patients) with faecal incontinence, irritable bowel syndrome, diabetes, alcohol withdrawal, heroin withdrawal, ultra-rapid opioid detoxification, short bowel syndrome (high output ileostomy or high rectal output), intestinal transplant and cholera patients [Fedorak et al. 1985; Bretzke, 1987; Gupta and Jha, 1988; Rabbani et al. 1989; Baumgartner and Rowen, 1991; Morali et al. 1991; Rovera et al. 1997; Camilleri et al. 2003, 2009; Mcdoniel et al. 2004; Buchman et al. 2006; Farzam and Najafi, 2009; Bharucha et al. 2010, 2014]. Results regarding statistical analysis of these studies are described in the following.

Figure 1.

Flow diagram of the systematic review and meta-analysis.

Table 1.

Empirical studies for clonidine and diarrhoea.

Study	Country	Sample size (men)	Mean age	Primary condition	Type of study	Clonidine dose (μg/day) (route)	Description: main results and side effects
CLINICAL TRIALS
Fedorak et al. [1985]	USA	3 (not reported)	Not reported	Diabetes	Design: Observational uncontrolled study	700 (oral)	Results: stool volume/weight decrease, improvement in stool consistency and nocturnal incontinence. Clonidine was used for 6–24 months.
					Comparison: clonidine effect (pre- and post-treatment)		Side effects: 0% postural hypotension; 1 patient who stopped abruptly developed rebound hypertension, anxiety, headache, sleeplessness
Schiller et al. [1985]	USA	8 (5)	31.0	Healthy volunteers	Design: Observational controlled study	300 (oral)	Results: rectal effluent volume reduction by 48%; increased total gut volume significantly by altering gut motility; absorption rate stimulated by clonidine by 15%
					Comparison: Clonidine versus placebo		Side effects: p < 0.05 for drop 15mmHg in SBP and 9mmHg in DBP
Bretzke [1987]	Germany	9 (3)	64.5	Diabetes	Design: Observational uncontrolled study	225 (oral)	Results: Stool frequency improved (from eight times before to two times/day after); nocturnal incontinence also improved.
					Comparison: clonidine effect (pre- and post-treatment)		Side effects: 3/9 reduction in BP (but not orthostatic hypotension); 1/9 bradycardia; 4/9 dry mouth; 5/9 fatigue; 3/9 headache
Baxter et al. [1987]	United Kingdom	13 (13)		Healthy volunteers	Design: Observational controlled study	300 (oral)	Results: Clonidine prolonged orocaecal transit time
					Comparison: Clonidine versus placebo		Side effects: Asymptomatic drop in systolic blood pressure (SBP) by 13 mmHg and pulse rate drop by 10 units
Pressman et al. [1987]	USA	12 (not reported)	Not reported	Healthy volunteers	Design: Observational controlled study	Not reported	Results: Experimental study validating a certain lactulose breath test, using codeine to estimate the effect on bowel transit time. Clonidine did not increase mouth-to-caecal transit time but codeine did.
					Comparison: Clonidine versus codeine		Side effects: not reported
Gupta and Jha, [1988]	India	120 (120)	25.8	Heroin withdrawal	Design: Randomized single blinded controlled study	600 (oral)	Results: Diarrhoea appeared in fewer patients receiving clonidine compared with chlordiazepoxide (14/60 versus 25/60 respectively).
					Comparison: clonidine versus chlordiazepoxide/chlorpromazine		Side effects: hypotension in 27/60 compared with 22/60 with chlordiazepoxide (4/60 versus 0/60 severe hypotension)
Baumer et al. [1989]	France	10 (10)	25.0	Healthy volunteers	Design: Randomized double blinded controlled study	300 (oral)	Results: Clonidine prolonged small bowel transit significantly after treatment.
					Comparison: Clonidine versus placebo
Gregersen et al. [1989]	Denmark	16 (16)	30.0	Healthy volunteers	Design: Observational controlled study	200 (oral)	Results: Clonidine decreased antral contractile activity and decreased the number of duodenal contractions
					Comparison: Clonidine effect (pre- and post-treatment)		Side effects: 4/15 drowsiness; 2/15 headache and nausea; 1/15: autonomic symptoms
Rabbani et al. [1989]	Bangladesh	37 (29)	32.2	Cholera	Design: Randomized single blinded controlled study	900 (oral)	Results: Modest reduction of stool electrolyte loss but does not significantly reduce faecal fluid loss in patients with cholera.
					Comparison: Clonidine versus placebo (sham treatment)		Side effects: The clonidine group had consistently lower SBP and diastolic blood pressure (DBP) but these differences were not statistically significant (10–15 mmHg less in SBP/DBP). Also no differences in heart rate between the two groups
Rubinoff et al. [1989]	USA	6 (6)	24.0	Healthy volunteers	Design: Randomized single blinded controlled crossover study	300 (oral)	Results: Orocaecal transit time was prolonged by 69.5%.Side effects: 6/6 had asymptomatic fall in SBP 31 ± 6 mmHg and 6/6 had drowsiness
					Comparison: clonidine versus placebo
Baumgartner and Rowen [1991]	Canada	43 (23)	46.0	Alcohol withdrawal	Design: Randomized double blinded controlled study	600 (oral and transdermal)	Results: Patients who were on clonidine exhibited less diarrhoea compare to chlordiazepoxide
					Comparison: Clonidine versus chlordiazepoxide		Side effects: Asymptomatic heart rate drop by 10 units (p < 0.05); asymptomatic SBP/DBP drop by 10–15 mmHg (p < 0.05). No rebound hypertension, nausea (20%), vomiting (4%), dry mouth (48%), light-headedness (25%), dizziness (16%), drowsiness 51%, headache 10%, fatigue 35%
Morali et al. [1991]	Israel	13 (3)	47.4	Diabetes	Design: Observational controlled study	300 (oral)	Results: Clonidine prolonged small bowel transit significantly after treatment.Side effects: Asymptomatic drop in systolic blood pressure by 32.85 ± 8 mmHg
					Comparison: clonidine effect (pre- and post-treatment)
Rovera et al. [1997]	USA	13 (not reported)	21.5	Small bowel transplant	Design: Observational uncontrolled study	150 (oral)	Results: Clonidine was effective in reducing intestinal output in adults but not in children.
					Comparison: Clonidine effect (pre- and post-treatment)		Side effects: No changes in blood pressure in children; in adults, clonidine tended to reduce both SBP and DBP (mean drop 11 mmHg). 1/13 stopped because of hypotension.
^#Thollander et al. [1997], [1999]	Sweden	21 (21)	28.0	Healthy volunteers	Design: Observational controlled study	4 μg/kg (not reported)	Results: Clonidine affects the migrating vascular and motor complex, by decreasing blood flow microcirculation in the duodenum
							Side effects: sedation and dry mouth in all subjects; reduction in SBP/DBP, but no hypotension
Mann and Shinkle [1998]	USA	9 (9)	42.0	Healthy volunteers	Design: Observational uncontrolled study	300 (oral)	Results: Clonidine prolonged orocaecal transit after treatment
					Comparison: clonidine effect (pre- and post-treatment)
Viramontes et al. [2001]	USA	55 (20)	36.0	Healthy volunteers	Design: Randomized single blinded controlled study	110 (oral)	Results: Clonidine did not significantly alter gastrointestinal or colonic transit, but it increased colonic compliance and reduced fasting tone without altering colonic response to a meal. Clonidine significantly reduced aggregate sensation to distensions overall.
					Comparison: 1. clonidine versus placebo; 2. clonidine versus clonidine (different dose)
Camilleri et al. [2003]	USA	38 (9)	38.3	Irritable bowel syndrome	Design: Randomized double blinded controlled study	160 (oral)	Results: Satisfactory relief of irritable bowel syndrome was higher with clonidine and relief was sustained through 4 weeks of treatment; also, bowel dysfunction was reduced with clonidine. Clonidine did not significantly alter gastrointestinal transit or gastric volumes.
					Comparison: 1. clonidine versus placebo; 2. clonidine versus clonidine (different dose)		Side effects: Drowsiness 40%, Malaise/fatigue 8%, Dizziness/light-headedness 16% Tinnitus 4%, Dry mouth 24%, Excessive thirst 8%, Bronchitis 12%, Cold/cough/nasal congestion 12%, Fever 0%, Vomiting 4%, Abdominal cramps 4%, Constipation 12%, Rectal pain/harder stools 4%, Rectal bleeding (wipe) 4%, Headache 12%, Irregular menstrual cycle 4%, Nosebleed 0%, Puncture wound finger 0%, Laceration thumb 0%, Folliculitis 4%, Tingling/numbness in foot 4%
Mcdoniel et al. [2004]	USA	2 (1)	25.5	Short bowel syndrome	Design: Observational uncontrolled study	300 (oral)	Results: Case 1: Ileostomy output decrease from 4 to 1 l/day. Case 2: Ileostomy output decrease from 4 to 1.5 l/day.
					Comparison: clonidine effect (pre- and post-treatment)		Side effects: 0% hypotension
Buchman et al. [2006]	USA	8 (3)	49.9	Short bowel syndrome	Design: Observational uncontrolled study	300 (transdermal)	Results: Modest but clinically significant decrease in faecal output and faecal Na losses
					Comparison: clonidine effect (pre- and post-treatment)		Side effects: 0% hypotension
Camilleri et al. [2009]	USA	83 (3)	34.5	Irritable bowel syndrome and healthy controls	Design: Randomized controlled study	250 (oral)	Results: Reduced volume to satiation, postprandial gastric volume, and sensation threshold for pain; increased rectal compliance.
				Irritable bowel syndrome and healthy controls	Comparison: clonidine effect (irritable bowel syndrome versus healthy controls); 2. clonidine versus clonidine (different dose)		Side effects: not reported
Farzam and Najafi [2009]	Iran	88 (not reported)	Not reported	Ultra-rapid opioid detoxification	Design: Randomized double blinded controlled study	539 (oral)	Results: Use of larger dose of clonidine can effectively decrease the incidence of diarrhoea after ultra-rapid opioid detoxification.
					Comparison: Clonidine versus clonidine (different doses)		Side effects: not reported
Bharucha et al. [2010]	USA	12 (0)	58.3	Faecal incontinence	Design: Observational uncontrolled study	200 (transdermal)	Results: symptom improvement; associated with increased rectal compliance and reduced rectal sensitivity
					Comparison: clonidine effect (pre- and post-treatment)		Side effects: overall side effects 50%, dry mouth 17%, orthostatic symptoms 17%, fatigue 25%
Bharucha et al. [2014]	USA	44 (0)	58.0	Faecal incontinence	Design: Randomized double blinded controlled study	200 (oral)	Results: Unaltered affect bowel symptoms, faecal continence, or anorectal functions, compared with placebo; in patients with diarrhoea, clonidine increased stool consistency, with a borderline improvement in faecal continence.
					Comparison: Clonidine versus placebo		Side effects: overall side effects 86%, dry mouth 73%, drowsiness 23%, light-headedness 27%, fatigue 36%
CASE REPORTS
Mcarthur et al. [1982]	USA	–	–	Lung cancer with neuroendocrine features	–	500 (oral)	74-year-old man with lung cancer secreting VIP causing diarrhoea. Diarrhoea initially was between 1400 and 1861 g/day. With 200 and 800 μg over 16 days, diarrhoea decreased up to 53% compare to before treatment.
Sacerdote [1986]	USA	–	–	Diabetes	–	100 (transdermal)	50-year-old woman with refractory diarrhoea did not respond to treatment with cophenotrope and kaolin-pectin. Initially stool frequency was 15–20/day and after clonidine treatment reduced to 1–2/day. Also stool became well formed from liquid/semisolid. Well tolerated with no mouth dryness or hypotension; only mild drowsiness. Discontinuation of treatment led to recurrence of diarrhoea.
Zimmermann et al. [1986]	Germany	–	–	VIPoma	–	225 (oral)	61-year-old man with VIPoma developed related to disease and chemotherapy. Patient responded to high-dose somatostatin and prednisolone but no significant response to oral clonidine 225 μg/day and other drugs (indomethacin, lithium, trifluoperazine, and nicotinic acid)
Roof [1987]	USA	–	–	Type 1 Diabetes mellitus	–	600 (transdermal)	48-year-old woman with watery diarrhoea (intermittent, voluminous). Initially tried oral dose of 200 μg/day, but could not tolerate due to drowsiness. Then tried transdermal clonidine (2 patches Catapres TTS-3/week), which tolerated well with only local dermatitis and no hypotension.
Migliore et al. [1988]	Italy	–	–	Type 1 diabetes mellitus	–	150 (oral)	22-year-old man with watery diarrhoea (7–8 times/day), resistant to antibiotic treatment with no incontinence. Biochemical testing was normal and barium follow through showed gastroparesis with fast small bowel transit. After treatment with clonidine, diarrhoea decreased in frequency and volume, stool consistency increased and nocturnal diarrhoea disappeared. No objective decrease in blood pressure was observed.
Scholz et al. [1991]	Germany	–	–	adenocarcinoma of colon in Familial adenomatous polyposis: ended up with high-output ileostomy, after colectomy and bowel infarction	–	1200 (oral)	46-year-old patient with high=output ileostomy (up to 10 l/day) was treated with clonidine, after loperamide, tinctura opii, cholestyramine and somatostatin had failed to reduce stool volume to less than 6 l/day. Under combined treatment with clonidine (1200 µg/day) and somatostatin (6 mg/day), which was well tolerated, stool weights were normalized within 24 hours.
Schworer et al. [1995]	Germany	–	–	Carcinoid syndrome	–	375 (oral)	69-year-old man with carcinoid syndrome developed diarrhoea secondary to octreotide treatment and interferon α_2b. Stool frequency was 6–8 times/day, initially decreased by ondansetron and tropisetron. When withdrawn, patient was treated with oral clonidine 75 μg tds and 150 μg at night which reduced stool frequency to 2–3 times/day and no night defecation.

Study not included in the meta-analysis because no data could be extracted from paper.

The total number of participants was 643 (range 2–120, median 13, mean 28, males 291). The mean age was 36 years (standard deviation [SD] 6). Clonidine was used as an oral or transdermal preparation with mean daily doses ranging between 110 and 900 μg/day (mean dose 344 μg/day, median dose 300 μg/day). Regarding quality assessment of the included papers, there is a high risk of selection bias due to lack of randomization and performance and detection bias due to lack of blinding; however the presentation of results seems adequate (see Figure 2 and Supplementary Figure 1).

Figure 2.

Risk of bias graph: reviewers’ judgements about each risk of bias item presented as percentages across all included studies.

Meta-analysis of the main outcome

Main effect

Meta-analysis was initially performed on studies involving healthy volunteers and patient groups, which included 22 studies that data could be extracted from. The studies by Thollander and colleagues [Thollander et al. 1997, 1999] did not allow for numerical data extraction. Overall, clonidine had a significant effect on diarrhoea with a mean SMD of −1.02 (95% confidence interval [CI] −1.46 to −0.58; see Figure 3A), which is strong according to Cohen’s criteria [Cohen, 1988]. However, there was significant heterogeneity (I² = 82%). The funnel plot for publication bias indicates asymmetry (Figure 4) and Begg’s test was significant (p = 0.012) but not Egger’s test (p = 0.151). Hence, publication bias is possibly present due to small study effects.

Figure 3.

Forest plots of the standardized mean difference (SMD) regarding the efficacy of clonidine for diarrhoea. (A) Forest plot of all studies (n = 22). (B) Forest plot with all studies (n = 14) except healthy volunteers. (C) Forest plot with all studies (n = 10) regarding functional diarrhoea (diabetes, irritable bowel syndrome, faecal incontinence and withdrawal). (D) Forest plot with all studies (n = 8) regarding functional diarrhoea and sample size over 10 subjects.

Figure 4.

Funnel plot indicating probable publication bias in the reporting of results for clonidine’s efficacy for diarrhoea.

Sensitivity analysis and cumulative meta-analysis

The first step of sensitivity analysis was to include only the 14 studies with patient groups and not healthy volunteers [Fedorak et al. 1985; Bretzke, 1987; Gupta and Jha, 1988; Rabbani et al. 1989; Baumgartner and Rowen, 1991; Morali et al. 1991; Rovera et al. 1997; Camilleri et al. 2003, 2009; Mcdoniel et al. 2004; Buchman et al. 2006; Farzam and Najafi, 2009; Bharucha et al. 2010, 2014]. Clonidine still had a strong significant effect on diarrhoea with a mean SMD of −1.17 (95% CI −1.74 to −0.60; see Figure 3B), very similar to the previous overall SMD of −1.02. Heterogeneity was still present in a similar magnitude (I² = 86%) as well as publication bias. When we included only the 10 studies with diagnoses that essentially constitute functional diarrhoea (diabetes, irritable bowel syndrome, faecal incontinence and withdrawal) [Fedorak et al. 1985; Bretzke, 1987; Gupta and Jha, 1988; Baumgartner and Rowen, 1991; Morali et al. 1991; Camilleri et al. 2003, 2009; Farzam and Najafi, 2009; Bharucha et al. 2010, 2014], clonidine still had a significant strong effect on diarrhoea with a mean SMD of −1.04 (95% CI −1.54 to −0.53; see Figure 3C). Furthermore, when we included only the studies with sample size over 10 patients [Gupta and Jha, 1988; Baumgartner and Rowen, 1991; Morali et al. 1991; Camilleri et al. 2003, 2009; Farzam and Najafi, 2009; Bharucha et al. 2010, 2014], the mean SMD did not change much: −0.99 (95% CI −1.54 to −0.43; see Figure 3D).

Further sensitivity meta-analysis was performed to examine the influence of individual studies in the overall mean SMD. No individual study lowers the effect of clonidine significantly since the overall mean always remains less than −0.87, indicating a constant strong effect. Two studies that lower the effect of clonidine to −0.87 are those of Rovera and colleagues [Rovera et al. 1997] and Camilleri and coworkers [Camilleri et al. 2009] (Figure 5A). Cumulative meta-analysis shows that a constant strong statistically significant effect of clonidine on diarrhoea is observed after publication of the ninth trial in 1989 (SMD ⩽ −0.74), while a significant strong effect appears as early as 1985 with the second trial (SMD = −1.45); see Figure 5B. Although subsequent trials have increased the precision of the point estimate, no substantive change has occurred in the direction or magnitude of the treatment effect. Cumulative meta-analysis by condition shows that after inclusion of studies with healthy volunteers the SMD does not change substantially, ranging between −1.02 and −1.18 with more precise confidence intervals (Figure 5C).

Figure 5.

(A) Sensitivity analysis for each individual study. (B) Cumulative meta-analysis by year. (C) Cumulative meta-analysis by condition.

Heterogeneity analysis

Next, we examined the potential sources of heterogeneity. We performed subgroup analyses with respect to geographical information (country and continent of study), study criteria (randomized or not, controlled or not), the condition under which the effect of clonidine was examined and the effect type [clonidine effect (pre- and post-treatment); clonidine versus placebo; clonidine effect (condition versus healthy controls); clonidine dose effect; clonidine versus codeine; and clonidine versus chlordiazepoxide). Results are shown in Table 2 and funnel plots for subgroup analyses are shown in Supplementary Figures 2–7. With respect to study criteria and geographical information, heterogeneity remained in the subgroups (I² > 50%), indicating that they were not potential factors. Interestingly though, in certain subgroups of the effect type and condition heterogeneity was low (I² < 50%). With respect to condition, heterogeneity was not present anymore in the subgroups of diabetes, faecal incontinence and withdrawal; and regarding effect type, I² was still high when effect was compared pre- and post-clonidine (80%), against placebo (59%) and healthy controls (94%) but was 0% when clonidine was compared against chlordiazepoxide in withdrawal.

Table 2.

Subgroup analyses of the efficacy of clonidine.

	k	N	Main effect		Heterogeneity
	k	N	SMD (95% CI)	p-value	I ²	χ² (p-value)
All studies	22	622	−1.02 (−1.46 to −0.58)	<0.001	82%	117.69 (<0.001)
Condition
Healthy volunteers	8	129	−0.77 (−1.47 to −0.07)	0.031	74%	26.96 (<0.001)
Cholera	1	37	0.38 (−0.29 to1.05)	0.265	–	–
Diabetes	3	25	−0.94 (−1.60 to −0.29)	0.005	0%	1.52 (0.469)
Faecal incontinence	2	56	−0.80 (−1.31 to −0.29)	0.002	0%	0.02 (0.892)
Intestinal failure	3	23	−4.49 (−8.90 to −0.07)	0.047	92%	24.09 (<0.001)
Irritable bowel syndrome	2	121	−1.77 (−3.82 to 0.28)	0.091	95%	19.97 (<0.001)
Withdrawal	3	231	−0.64 (−0.93 to −0.34)	<0.001	0%	1.14 (0.566)
Effect Type
Clonidine effect (pre- and post-treatment)	8	72	−1.87 (−2.90 to −0.84)	<0.001	80%	34.11 (<0.001)
Clonidine versus placebo	8	211	−0.57 (−1.02 to −0.11)	0.015	59%	17.05 (0.017)
Clonidine effect (condition vs healthy controls)	2	96	−1.71 (−3.91 to 0.48)	0.126	94%	16.52 (<0.001)
Clonidine dose effect	1	88	−0.75 (−1.20 to −0.30)	0.001	–	–
Clonidine versus chlordiazepoxide	2	143	−0.55 (−0.94 to −0.16)	0.006	0.0%	0.72 (0.398)
Clonidine versus codeine	1	12	1.22 (−0.23 to 2.67)	0.100	–	–
Continent
America	14	316	−1.23 (−1.94 to −0.52)	0.001	85%	87.57 (<0.001)
Asia	4	258	−0.38 (−0.84 to 0.08)	0.106	61%	7.72 (0.052)
Europe	4	48	−1.28 (−2.02 to −0.53)	0.001	60%	7.50 (0.058)
Country
Bangladesh	1	37	0.38 (−0.29 to 1.05)	0.265	–	–
Canada	1	23	−0.90 (−1.80 to −0.004)	0.049	–	–
Denmark	1	16	−2.29 (−3.26 to −1.33)	<0.001	–	–
France	1	10	−1.04 (−2.03 to −0.05)	0.04	–	–
Germany	1	9	−1.33 (−2.35 to −0.31)	0.011	–	–
India	1	120	−0.47 (−0.91 to −0.04)	0.034	–	–
Iran	1	88	−0.75 (−1.20 to −0.30)	0.001	–	–
Israel	1	13	−0.57 (−1.46 to 0.33)	0.215	–	–
United Kingdom	1	13	−0.55 (−1.37 to 0.27)	0.187	–	–
USA	13	293	−1.27 (−2.05 to −0.50)	0.001	86%	87.38 (<0.001)
Randomization
Randomized	10	504	−0.83 (−1.39 to −0.26)	0.004	86%	65.76 (<0.001)
Nonrandomized	12	118	−1.27 (−2.01 to −0.54)	0.001	78%	49.73 (<0.001)
Controlled or not
Controlled	14	550	−0.72 (−1.19 to −0.25)	0.003	83%	74.36 (<0.001)
Uncontrolled	8	72	−1.87 (−2.90 to −0.84)	<0.001	80%	34.11 (<0.001)

k, number of studies; N, number of participants; SMD, standardized mean difference; CI, confidence interval.

We finally examined continuous variables that could be sources of heterogeneity. The results of meta-regression are shown in Table 3. Clonidine dose and mean age do not predict heterogeneity in the SMD (coefficient p-values >0.05).

Table 3.

Meta-regression results.

Variable	Coefficient	Standard error	p-value
Clonidine dose	0.0029	0.0016	0.094
Mean age	0.0444	0.0273	0.128

Other outcomes

Stool volume was described in six papers [Fedorak et al. 1985; Schiller et al. 1985; Rabbani et al. 1989; Rovera et al. 1997; Mcdoniel et al. 2004; Buchman et al. 2006]. The WMD was calculated: clonidine reduced stool volume by −0.97 l/day (95% CI −2.27 to 0.33; I² = 97%). The SMD indicated a strong significant decrease in stool volume (SMD = −2.08, 95% CI −3.88 to −0.27; I² = 89%; see the Forest plots shown in Supplementary Figures 8 and 9).

Stool frequency was described in five papers [Fedorak et al. 1985; Bretzke, 1987; Camilleri et al. 2003; Bharucha et al. 2010, 2014]. The WMD was calculated: clonidine decreased stool frequency by −0.4 times/day (95% CI −1.0 to 0.2; I² = 27%) [Bretzke, 1987; Camilleri et al. 2003; Bharucha et al. 2010, 2014]; when clonidine was compared pre- and post-treatment it decreased stool frequency by −1.5 times/day (95% CI −4.2 to 1.2; I² = 64%) [Bretzke, 1987; Bharucha et al. 2010] and when clonidine was compared with placebo it decreased stool frequency by −0.3 times/day (95% CI −0.7 to 0.3; I² = 0%) [Camilleri et al. 2003; Bharucha et al. 2014]. Finally, the overall SMD indicated decrease in stool frequency (SMD = −0.24, 95% CI −0.64 to 0.15; I² = 14%) [Fedorak et al. 1985; Bretzke, 1987; Camilleri et al. 2003; Bharucha et al. 2010, 2014] (Supplementary Figures 10 and 11). The decrease of stool frequency does not appear to be clinically significant and this result is consistent with the mode of action of clonidine.

Stool consistency was described in four papers [Fedorak et al. 1985; Camilleri et al. 2003; Bharucha et al. 2010, 2014]. The SMD indicated that there was a weak non-significant improvement in stool consistency after clonidine treatment (SMD = −0.30, 95% CI −0.77 to 0.18; I² = 29%; see Supplementary Figure 12). Nocturnal incontinence was described in two papers [Fedorak et al. 1985; Bretzke, 1987]. The SMD indicated that there was an important reduction in the presence of this symptom after clonidine treatment (SMD = −2.20, 95% CI −3.53 to -0.87; I² = 0%; see Supplementary Figure 13).

Gastrointestinal transit time was described in eight papers [Baxter et al. 1987; Pressman et al. 1987; Baumer et al. 1989; Rubinoff et al. 1989; Morali et al. 1991; Mann and Shinkle, 1998; Viramontes et al. 2001; Camilleri et al. 2003]. The SMD indicated an increase in transit time, however this was nonsignificant (SMD = 0.31, 95% CI −0.17 to 0.79; I² = 53%; see Supplementary Figure 14). The WMD was also calculated: clonidine increased significantly orocaecal transit time by 0.52 hours (31.2 minutes; 95% CI 0.06−0.98 h; I² = 61%) [Baxter et al. 1987; Pressman et al. 1987; Baumer et al. 1989; Rubinoff et al. 1989; Morali et al. 1991; Mann and Shinkle, 1998; Camilleri et al. 2003] (Supplementary Figure 15).

Furthermore, clonidine decreased rectal losses of sodium (SMD = −0.81, 95% CI −1.67 to 0.05; I² = 59%; see Supplementary Figure 16) [Schiller et al. 1985; Rabbani et al. 1989; Buchman et al. 2006] and chloride (SMD = −0.75, 95% CI −1.34 to 0.16; I² = 0%) (Supplementary Figure 17) [Schiller et al. 1985; Rabbani et al. 1989], but did not affect bicarbonate (SMD = 0.29, 95% CI −0.63 to 1.21; I² = 53%; see Supplementary Figure 18) and potassium losses (SMD = 0.11, 95% CI −0.46 to 0.67; I² = 0%; see Supplementary Figure 19) [Schiller et al. 1985; Rabbani et al. 1989] while it did not affect absorption of sodium (SMD = 0.31, 95% CI −0.44 to 1.06; I² = 0%; see Supplementary Figure 20) and potassium (SMD = 0.20, 95% CI −0.55 to 0.94; I² = 0%; see Supplementary Figure 21) [Schiller et al. 1985; Buchman et al. 2006]. Absorption of fat, nitrogen, magnesium, calcium [Buchman et al. 2006], chloride, bicarbonate and water [Schiller et al. 1985] were not affected.

Several other outcomes were described in papers which are discussed narratively. Bharucha and colleagues [Bharucha et al. 2010, 2014] performed an uncontrolled study and then a placebo-controlled study of the effect of clonidine in faecal incontinence. In the uncontrolled study there was a more prominent effect of clonidine reducing the days of faecal incontinence, faecal incontinence episodes and proportion of bowel movements being incontinent or with staining. There was greater satisfaction with therapy in the uncontrolled study than in the placebo controlled study but neither study seemed to have an effect on lifestyle, coping or depression or on loperamide use. Camilleri and colleagues [Camilleri et al. 2009] also did not find any effect of the use of clonidine for irritable syndrome on depression and anxiety as measured by the hospital anxiety depression scale.

Moreover, most studies did not find any effect of clonidine on anal pressures, rectal compliance or sensation [Camilleri et al. 2003; Bharucha et al. 2010, 2014]. However, Camilleri and colleagues [Camilleri et al. 2009] reported that clonidine reduced sensation threshold for pain and increased rectal compliance in patients with irritable bowel syndrome and Viramontes and colleagues [Viramontes et al. 2001] reported that clonidine increased colonic compliance and reduced fasting tone without altering colonic response to a meal, aggregate sensation to distensions and sensation of pain in healthy volunteers.

Finally, two studies examined the influence of certain gene variations on the effect clonidine. Camilleri and colleagues [Camilleri et al. 2009] reported that α2A, GNβ3 and SLC6A4 genotypes significantly modify responses to clonidine on sensory and motor gastrointestinal functions (rectal sensation of gas, urgency, pain) in health and irritable bowel syndrome but a later study by the group in patients with faecal incontinence [Bharucha et al. 2014] reported that the α2A-C1291G polymorphism did not predict effects of clonidine with respect to faecal incontinence severity score, Bristol stool score or quality of life.

Narrative discussion of case reports

There have been seven case reports in the literature describing the use of clonidine to treat severe refractory diarrhoea (Table 1). Their description is interesting, albeit should be treated with caution due to the lack of control and the possibility of placebo effect. The indications have been diabetes, tumours with neuroendocrine features (carcinoid syndrome, lung cancer with VIP secretion, VIPoma) and high output ileostomy. Dose range was 100–1200 μg/day and route of administration could be oral or transdermal. The highest dose of clonidine was used in a 46-year-old patient with high-output ileostomy (up to 10 l/day) who was treated with clonidine, after loperamide, tinctura opii, cholestyramine and somatostatin had failed to reduce stool volume to less than 6 l/day. Under combined treatment with clonidine 1200 μg/day and somatostatin 6 mg/day, which was well tolerated, stool weights were normalized within 24 hours. Neuroendocrine tumours seemed to require higher doses of clonidine and lower doses were used in diabetic patients.

Side effects

The side effects for each study are describe in Table 1. The most prevalent side effect was dry mouth with a mean risk ratio of 12.3 (95% CI 2.5–61.2; I² = 0%) for clonidine versus placebo [Camilleri et al. 2003; Bharucha et al. 2014], appearing up to 100% of patients who had received clonidine [Bretzke, 1987; Rubinoff et al. 1989; Baumgartner and Rowen, 1991; Thollander et al. 1997, 1999; Bharucha et al. 2010, 2014]. Drowsiness, light-headedness and fatigue had mean risk ratios over 1 for clonidine versus placebo but with nonsignificant 95% CIs (2.37, 95% CI 0.80–7.00; 3.32, 95% CI 0.89–12.42; and 1.50, 95% CI 0.63–3.61, respectively; I² = 0% for all analyses) [Camilleri et al. 2003; Bharucha et al. 2014]. Hypotension varied from between 0% [Fedorak et al. 1985; Mcdoniel et al. 2004; Buchman et al. 2006] and 45% [Gupta and Jha, 1988] in patients who received clonidine but there was a general conception that this was a drop in systolic blood pressure up to 32.9 units with no symptomatology [Rubinoff et al. 1989; Morali et al. 1991]; however, a few cases of severe orthostatic hypotension [Rovera et al. 1997] were described (Table 1). Finally, other side effects were cold/cough/fever, headache, constipation, bradycardia and tachycardia: all general autonomic symptoms.

Discussion

We performed a systematic review of studies examining the effect of clonidine against diarrhoea. This is the first such review to the best of our knowledge. As a general comment, we can observe that the literature regarding clonidine starts as far back as the 1980s, however its use and reporting comes in a more fragmented manner than one would expect: only 31 studies (mainly observational nonrandomized) spanning a period of almost 40 years.

This is not something to disregard. Although this study shows an objective decrease of diarrhoea by clonidine, the quality of studies is compromised by selection bias and the results might be limited by publication bias and heterogeneity. Interestingly, diabetes, faecal incontinence and withdrawal were three group subgroups without heterogeneity; no group however included many studies. Small studies effect was a pattern observed in the present review and we dare say it almost reflects some sort of caution regarding the use of clonidine. This may be due to its potential side effects, most serious being hypotension and rebound hypertension when stopped abruptly. This does not come as a surprise since it is something observed in our current practice at University College Hospital, London. Clonidine prescription requires pharmacy approval for unlicensed use and frequently patients need inpatient trials of this medication before commencing outpatient treatment. Moreover, its use by anaesthetists during anaesthesia could equally trouble the general practitioner or gastroenterologist who might not be willing to try this drug as an outpatient.

Nevertheless, the present study appears to dispel some of practitioners’ and specialists’ cautions. Most studies report an asymptomatic drop in systolic blood pressure, which is acceptable considering the severity of their patient’s diarrheal symptoms. Moreover, the use of clonidine appears to treat effectively their refractory diarrhoea, having failed multiple other treatments: the benefits outweighing the risk. The most reported side effect was dry mouth increasing the risk just over 12 times compared with placebo. The WMDs calculated from the study were only a subset of the whole group of studies because not all studies reported on raw values to allow for such calculations. This decreases the power of these calculations and should be kept in mind.

On another topic, this study seems to corroborate previous findings regarding clonidine’s acting mechanism. Clonidine seems to decrease on average one litre of stool per day, improving consistency and decreasing frequency by one movement every 2 days approximately while additionally increasing gastrointestinal transit up to 1 hour. It also reduces sodium and chloride rectal losses. The effect of clonidine seems to be stronger when clonidine was compared pre- and post-treatment.

Within its limitations, the present study suggests that clonidine is a considerable option for refractory diarrhoea. However, one must take into account the limitations which were the presence of heterogeneity, small sample size of individual studies which in its turn was amplified by design biases.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Conflict of interest statement

The authors declare that there is no conflict of interest.

References

Allison

Sercombe

Pounder

(1988) A double-blind crossover comparison of lidamidine, loperamide and placebo for the control of chronic diarrhoea. Aliment Pharmacol Ther 2: 347–351.

Asai

Mcbeth

Stewart

Williams

Vaughan

Power

(1997) Effect of clonidine on gastric emptying of liquids. Br J Anaesth 78: 28–33.

Awad

Llorens

Camelo

Sanchez

(2000) A randomised double-blind placebo-controlled trial of lidamidine HCL in irritable bowel syndrome. Acta Gastroenterol Latinoam 30: 169–175.

Baumer

Danays

Lion

Cosnes

Gendre

Le Quintrec

(1989) [Effect of clonidine on oro-cecal transit time in normal man]. Ann Gastroenterol Hepatol (Paris) 25: 5–9.

Baumgartner

Rowen

(1991) Transdermal clonidine versus chlordiazepoxide in alcohol withdrawal: a randomized, controlled clinical trial. South Med J 84: 312–321.

Baxter

Edwards

Holden

Cunningham

Welch

Read

(1987) The effect of two alpha 2-adrenoreceptor agonists and an antagonist on gastric emptying and mouth to caecum transit time in humans. Aliment Pharmacol Ther 1: 649–655.

Begg

Mazumdar

(1994) Operating characteristics of a rank correlation test for publication bias. Biometrics 50: 1088–1101.

Bharucha

Camilleri

Zinsmeister

Hanson

(1997) Adrenergic modulation of human colonic motor and sensory function. Am J Physiol 273: G997–G1006.

Bharucha

Fletcher

Camilleri

Edge

Carlson

Zinsmeister

(2014) Effects of clonidine in women with fecal incontinence. Clin Gastroenterol Hepatol 12: 843–851 e842; quiz e844.

10.

Bharucha

Seide

Zinsmeister

(2010) The effects of clonidine on symptoms and anorectal sensorimotor function in women with faecal incontinence. Aliment Pharmacol Ther 32: 681–688.

11.

Borenstein

Hedges

Higgins

Rothstein

(2008) Introduction to Meta-Analysis. Chichester: John Wiley & Sons.

12.

Bowden

Tierney

Copas

Burdett

(2011) Quantifying, displaying and accounting for heterogeneity in the meta-analysis of RCTs using standard and generalised Q statistics. BMC Med Res Methodol 11: 41.

13.

Bretzke

(1987) [Therapy of diarrhea in diabetic enteropathy with clonidine]. Z Gesamte Inn Med 42: 680–682.

14.

Buchman

Fryer

Wallin

Ahn

Polensky

Zaremba

(2006) Clonidine reduces diarrhea and sodium loss in patients with proximal jejunostomy: a controlled study. JPEN J Parenter Enteral Nutr 30: 487–491.

15.

Camilleri

Busciglio

Carlson

Mckinzie

Burton

Baxter

. (2009) Pharmacogenetics of low dose clonidine in irritable bowel syndrome. Neurogastroenterol Motil 21: 399–410.

16.

Camilleri

Kim

Mckinzie

Kim

Thomforde

Burton

. (2003) A randomized, controlled exploratory study of clonidine in diarrhea-predominant irritable bowel syndrome. Clin Gastroenterol Hepatol 1: 111–121.

17.

Cohen

(1988) Statistical Power Analysis for the Behavioral Sciences, 2nd edn. Hillsdale, NJ: L. Erlbaum Associates.

18.

Cooper

Hedges

Valentine

(2009) The Handbook of Research Synthesis and Meta-Analysis, 2nd edn. New York: Russell Sage Foundation.

19.

Dharmsathaphorn

(1986) Alpha 2-adrenergic agonists: a newer class of antidiarrheal drug. Gastroenterology 91: 769–770.

20.

Edwards

Cann

Read

Holdsworth

(1986) Effect of two new antisecretory drugs on fluid and electrolyte transport in a patient with secretory diarrhoea. Gut 27: 581–586.

21.

Edwards

Read

(1986) Effect of lidamidine, a proposed alpha 2-adrenoreceptor agonist, on salt and water transport in human jejunum. Dig Dis Sci 31: 817–821.

22.

Egger

Smith

Schneider

Minder

(1997) Bias in meta-analysis detected by a simple, graphical test. BMJ 315: 629–634.

23.

Faraone

(2008) Interpreting estimates of treatment effects: implications for managed care. P T 33: 700–711.

24.

Farzam

Najafi

(2009) Effect of clonidine on vomiting and diarrhea associated with ultra-rapid opioid detoxification: a randomized clinical trial. Crit Care 13: S165.

25.

Fedorak

Field

(1987) Antidiarrheal therapy. Prospects for new agents. Dig Dis Sci 32: 195–205.

26.

Fedorak

Field

Chang

(1985) Treatment of diabetic diarrhea with clonidine. Ann Intern Med 102: 197–199.

27.

Fragkos

Tsagris

Frangos

(2014) Publication bias in meta-analysis: confidence intervals for Rosenthal’s fail-safe number. Int Sch Res Notices 2014: 1–17.

28.

Fragkos

Tsagris

Frangos

(2016) Exploring the distribution for the estimator of Rosenthal’s ‘fail-safe’ number of unpublished studies in meta-analysis. Commun Stat Theory Methods: in press.

29.

Gasbarrini

Corazza

Feliciani

Albano

Stufano

Altomare

. (1986) A Multicenter double-blind controlled trial comparing lidamidine HCL and loperamide in the symptomatic treatment of acute diarrhoea. Arzneimittelforschung 36: 1843–1845.

30.

Gillum

Mamidipudi

Johnston

(2000) Ischemic stroke risk with oral contraceptives: a meta-analysis. JAMA 284: 72–78.

31.

Goff

(1984) Diabetic diarrhea and lidamidine. Ann Intern Med 101: 874–875.

32.

Gregersen

Kraglund

Rittig

Tottrup

(1989) The effect of a new selective alpha 2-adrenoreceptor antagonist, idazoxan, and the agonist, clonidine, on fasting antroduodenal motility in healthy volunteers. Aliment Pharmacol Ther 3: 435–443.

33.

Guarino

Lo Vecchio

Berni Canani

(2012) Chronic diarrhoea in children. Best Pract Res Clin Gastroenterol 26: 649–661.

34.

Gupta

Jha

(1988) Clonidine in heroin withdrawal syndrome: a controlled study in India. Br J Addict 83: 1079–1084.

35.

Halland

Talley

(2012) Fecal incontinence: mechanisms and management. Curr Opin Gastroenterol 28: 57–62.

36.

Heredia Diaz

Alcantara

Solis

(1979) [Evaluation of the safety and effectiveness of WHR-1142A in the treatment of non-specific acute diarrhea]. Rev Gastroenterol Mex 44: 167–173.

37.

Heredia Diaz

Kajeyama Escobar

(1981) [Double-blind evaluation of the effectiveness of lidamidine hydrochloride (WHR-1142A) vs loperamide vs. placebo in the treatment of acute diarrhea]. Salud Publica Mex 23: 483–491.

38.

Heredia Diaz

Romero Ramirez

Soliz Ladron De Guevara

(1983) [Lidamidine. Treatment of Staphylococcal food poisoning]. Salud Publica Mex 25: 367–371.

39.

Higgins

Green

Cochrane Collaboration (2008) Cochrane Handbook for Systematic Reviews of Interventions. Chichester: Wiley-Blackwell.

40.

Higgins

Thompson

(2002) Quantifying heterogeneity in a meta-analysis. Stat Med 21: 1539–1558.

41.

Higgins

Thompson

Deeks

Altman

(2003) Measuring inconsistency in meta-analyses. BMJ 327: 557–560.

42.

Huedo-Medina

Sanchez-Meca

Marin-Martinez

Botella

(2006) Assessing heterogeneity in meta-analysis: Q statistic or I² index? Psychol Meth 11: 193–206.

43.

Huilgol

Evans

Hellman

Soergel

(2002) Acute effect of clonidine on gastric emptying in patients with diabetic gastropathy and controls. Aliment Pharmacol Ther 16: 945–950.

44.

Kaiser

Surawicz

(2012) Infectious causes of chronic diarrhoea. Best Pract Res Clin Gastroenterol 26: 563–571.

45.

Kwiatek

Fox

Steingoetter

Menne

Pal

Fruehauf

. (2009) Effects of clonidine and sumatriptan on postprandial gastric volume response, antral contraction waves and emptying: an MRI study. Neurogastroenterol Motil 21: 928-e971.

46.

Vaziri

(2012) Treatment of chronic diarrhoea. Best Pract Res Clin Gastroenterol 26: 677–687.

47.

Lipsey

Wilson

(2001) Practical Meta-Analysis. Thousand Oaks, CA: Sage Publications, Inc.

48.

Malcolm

Camilleri

Kost

Burton

Fett

Zinsmeister

(2000) Towards identifying optimal doses for alpha-2 adrenergic modulation of colonic and rectal motor and sensory function. Aliment Pharmacol Ther 14: 783–793.

49.

Mann

Shinkle

(1998) Effect of clonidine on gastrointestinal transit time. Hepatogastroenterology 45: 1023–1025.

50.

Masukawa

Kaneko

Takahashi

Ishikawa

Suzuki

(1987) Effect of lidamidine on transepithelial potential difference evoked by prostaglandin E2 in human jejunum. Gastroenterol Jpn 22: 588–590.

51.

Mcarthur

Anderson

Durbin

Orloff

Dharmsathaphorn

(1982) Clonidine and lidamidine to inhibit watery diarrhea in a patient with lung cancer. Ann Intern Med 96: 323–325.

52.

Mcdoniel

Taylor

Huey

Eiden

Everett

Fleshman

. (2004) Use of clonidine to decrease intestinal fluid losses in patients with high-output short-bowel syndrome. JPEN J Parenter Enteral Nutr 28: 265–268.

53.

Migliore

Barone

Manna

Greco

(1988) Diabetic diarrhea and clonidine. Ann Intern Med 109: 170–171.

54.

Moher

Liberati

Tetzlaff

Altman

(2009) Preferred reporting items for systematic reviews and meta-analyses: the Prisma statement. J Clin Epidemiol 62: 1006–1012.

55.

Morali

Braverman

Lissi

Goldstein

Jacobsohn

(1991) Effect of clonidine on gallbladder contraction and small bowel transit time in insulin-treated diabetics. Am J Gastroenterol 86: 995–999.

56.

Pressman

Hofmann

Witztum

Gertler

Steinbach

Stokes

. (1987) Limitations of indirect methods of estimating small bowel transit in man. Dig Dis Sci 32: 689–699.

57.

Prior

Wilson

Whorwell

(1988) Double-blind study of an alpha 2 agonist in the treatment of irritable bowel syndrome. Aliment Pharmacol Ther 2: 535–539.

58.

Rabbani

Butler

Patte

Abud

(1989) Clinical Trial of clonidine hydrochloride as an antisecretory agent in cholera. Gastroenterology 97: 321–325.

59.

Randall

Vizuete

Wendorf

Ayyar

Constantine

(2012) Current and emerging strategies in the management of Crohn’s disease. Best Pract Res Clin Gastroenterol 26: 601–610.

60.

Rodriguez Magallan

Valadez Velazquez

Llorens Torres

Sanchez Torres

(1997) [Treatment of irritable colon with lidamidine and support psychotherapy]. Rev Gastroenterol Mex 62: 7–13.

61.

Roof

(1987) Treatment of diabetic diarrhea with clonidine. Am J Med 83: 603–604.

62.

Rosa-E-Silva

Troncon

Oliveira

Iazigi

Gallo

Jr Foss

(1995) Treatment of diabetic gastroparesis with oral clonidine. Aliment Pharmacol Ther 9: 179–183.

63.

Rovera

Furukawa

Reyes

Todo

Hutson

(1997) The use of clonidine for the treatment of high intestinal output following small bowel transplantation. Transplant Proc 29: 1853–1854.

64.

Rubinoff

Piccione

Holt

(1989) Clonidine prolongs human small intestine transit time: use of the lactulose-breath hydrogen test. Am J Gastroenterol 84: 372–374.

65.

Sacerdote

(1986) Topical clonidine for diabetic diarrhea. Ann Intern Med 105: 139.

66.

Schiller

(2005) Chronic diarrhea. Curr Treat Options Gastroenterol 8: 259–266.

67.

Schiller

(2012) Definitions, pathophysiology, and evaluation of chronic diarrhoea. Best Pract Res Clin Gastroenterol 26: 551–562.

68.

Schiller

Santa Ana

Morawski

Fordtran

(1985) Studies of the antidiarrheal action of clonidine. Effects on motility and intestinal absorption. Gastroenterology 89: 982–988.

69.

Scholz

Bause

Reymann

Durig

(1991) [Treatment with clonidine in a case of the short bowel syndrome with therapy-refractory diarrhea]. Anasthesiol Intensivmed Notfallmed Schmerzther 26: 265–269.

70.

Schworer

Munke

Stockmann

Ramadori

(1995) Treatment of diarrhea in carcinoid syndrome with ondansetron, tropisetron, and clonidine. Am J Gastroenterol 90: 645–648.

71.

Sninsky

Davis

Clench

Thomas

Mathias

(1986) Effect of lidamidine hydrochloride and loperamide on gastric emptying and transit of the small intestine. A double-blind study. Gastroenterology 90: 68–73.

72.

Sterne

Sutton

Ioannidis

Terrin

Jones

Lau

. (2011) Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. BMJ 343: d4002.

73.

Stroup

Berlin

Morton

Olkin

Williamson

Rennie

. (2000) Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis of Observational Studies in Epidemiology (MOOSE) Group. JAMA 283: 2008–2012.

74.

Tack

(2012) Functional diarrhea. Gastroenterol Clin North Am 41: 629–637.

75.

Thollander

Gazelius

Hellstrom

(1999) Adrenergic modulation of small bowel haemodynamics in interdigestive motility state of man. Eur J Gastroenterol Hepatol 11: 257–265.

76.

Thollander

Hellström

Gazelius

(1997) Semi-invasive laser-Doppler flowmetry technique. Int J Microcirc Clin Exp 17: 15–21.

77.

Thumshirn

Camilleri

Choi

Zinsmeister

(1999) Modulation of gastric sensory and motor functions by nitrergic and alpha2-adrenergic agents in humans. Gastroenterology 116: 573–585.

78.

Unnerstall

Kopajtic

Kuhar

(1984) Distribution of Α₂ agonist binding sites in the rat and human central nervous system: analysis of some functional, anatomic correlates of the pharmacologic effects of clonidine and related adrenergic agents. Brain Res 319: 69–101.

79.

Vicencio Tovar

Gonzalez Guajardo

Barrera Martinez

De Guevara

(1986) [Evaluation of the efficacy of lidamidine hydrochloride in patients with a hypersecretory ileostomy]. Rev Gastroenterol Mex 51: 199–203.

80.

Viramontes

Malcolm

Camilleri

Szarka

Mckinzie

Burton

. (2001) Effects of an alpha(2)-adrenergic agonist on gastrointestinal transit, colonic motility, and sensation in humans. Am J Physiol Gastrointest Liver Physiol 281: G1468–G1476.

81.

Vohmann

Hoffmann

(2013) Antidiarrhoika Bei Chronischer Diarrhoe. Dtsch Med Wochenschr 138: 2309–2312.

82.

Zimmermann

Rapp

Binder

Frolich

Bode

(1986) [Vipoma. 9-year observations using currently available therapy methods]. Dtsch Med Wochenschr 111: 298–301.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

2.13 MB

What about clonidine for diarrhoea? A systematic review and meta-analysis of its effect in humans

Abstract

Background:

Method:

Results:

Conclusion:

Keywords

Introduction

Methods

Study eligibility criteria

Search strategy and terms

Study selection, data extraction and quality assessment

Statistical analysis

Results

Systematic review and meta-analysis of trials

Studies and quality assessment

Meta-analysis of the main outcome

Main effect

Sensitivity analysis and cumulative meta-analysis

Heterogeneity analysis

Other outcomes

Narrative discussion of case reports

Side effects

Discussion

Footnotes

Funding

Conflict of interest statement

References

Supplementary Material