1. Schumann D, Langhorst J, Dobos G, Cramer H. Randomised clinical
trial: yoga vs a low-FODMAP diet in patients with irritable bowel
syndrome. Aliment Pharmacol Ther. 2018;47:203-211.
2. Bouman AC, ten Cate-Hoek AJ, Ramaekers BL, Joore MA. Sample
size estimation for non-inferiority trials: frequentist approach versus
decision theory approach. PLoS ONE. 2015;10:e0130531.
3. European Medicinal Agency (EMA). Guideline on the evaluation of
medicinal products for the treatment of irritable bowel syndrome.
CPMP/EWP/785/97 Rev. 1, Sept, 25, 2014
4. Dorn SD, Kaptchuk TJ, Park JB, et al. A meta-analysis of the pla-
cebo response in complementary and alternative medicine trials of
irritable bowel syndrome. Neurogastroenterol Motil. 2007;19:
5. Enck P, Aziz Q, Barbara G, et al. Irritable bowel syndrome. Nat Rev
Dis Primers. 2016;2:16014.
6. Rutherford BR, Sneed JR, Roose SP. Does study design influence
outcome? The effects of placebo control and treatment duration in
antidepressant trials. Psychother Psychosom. 2009;78:172-181.
7. Elsenbruch S, Enck P. Placebo effects and their determinants in gastroin-
testinal disorders. Nat Rev Gastroenterol Hepatol. 2015;12:472-485.
8. Ford AC, Moayyedi P. Meta-analysis: factors affecting placebo
response rate in the irritable bowel syndrome. Aliment Pharmacol
9. Locke GR 3rd, Zinsmeister AR. Perils and pitfalls of small negative
clinical trials. Am J Gastroenterol. 2006;101:2185-2186.
10. Leontiadis GI. How to interpret a negative study. Am J Gastroenterol.
Letter: all or nothing—placebo effects in a non-drug clinical
trial in IBS. Authors’ reply
With interest we have read Prof Enck’s letter
regarding our ran-
domised trial on yoga vs a low-FODMAP diet in patients with irrita-
ble bowel syndrome (IBS).
The therapeutic approach recommended by the World Gastroen-
for patients with irritable bowel syndrome
includes pharmacological, dietary and psychological approaches. Due
to the complexity of designing appropriate placebo-controlled inter-
ventions for psychological treatments, obtaining high-quality evi-
dence for these therapies is challenging. Up to date small
randomised controlled trials have suggested gastrointestinal symp-
tom relief in patients with IBS comparing yoga to a waitlist control
An alternative methodological approach is to compare the
effectiveness of such an intervention to one with a high level of evi-
dence. The low FODMAP diet has been shown to benefit patients in
a large proportion, regardless of IBS subtype and the majority of
studies support the efficacy of a low FODMAP diet when compared
to a placebo.
Furthermore, it should be considered that the dura-
bility of improvement (3 months after the intervention) seen in our
trial is usually not seen in pharmacological studies, when medication
In the letter, it is further noted that the study was “conducted
in a medical unit known for and specialised in complementary and
alternative medicine (CAM) approaches, thus likely to attract a
sub-selection of IBS patients interested in CAM treatment.” Before
randomisation, patients were asked to rate their treatment expec-
tancy for both interventions on a 10-point Likert-type scale
(higher values indicate higher expected efficacy). The yoga group
had rated their expectancy that yoga would be effective as 6.27
(SD, 2.85) while the FODMAP group had rated their expectancy
that FODMAP would be effective as 6.21 (SD, 2.62) indicating
comparable expectations for the conventional and the CAM
approach (P = 0.941). While this was not discussed in our trial,
we are not aware that potential centre effects are generally dis-
cussed in clinical trials conducted at (for example) university hospi-
tals, also likely to attract a different patient subgroup than a
The letter further questions the adequacy of our sample size cal-
culation. As noted in the publication, the study was powered to
detect a clinically meaningful group difference on the primary out-
come measure. The “noncontrolled three-arm cross-over trial of
another CAM IBS intervention” cited in the sample size calculation
section was used only to estimate the expected baseline standard
deviation in IBS symptoms. Given that this other trial was also con-
ducted in our centre, we expected a comparable standard deviation
in our new study. Of note, the prior study was not used to estimate
We totally agree that secondary analyses are prone to beta
errors and therefore mainly based our discussion on the primary out-
The authors’ declarations of personal and financial interests are
unchanged from those in the original article.
LETTERS TO THE EDITORS
© 2018 John Wiley & Sons Ltd