Author's response to reviews Title: Effects of a commercial product containing Guarana on psychological well-being, anxiety and mood: a single-blind, placebo-controlled study in heathy subjects Authors: Gianluca I Silvestrini (lab.pharm@uninsubria.it) Franca Marino (franca.marino@uninsubria.it) Marco Cosentino (marco.cosentino@uninsubria.it) Version: 2 Date: 23 April 2013 Author's response to reviews: see over
Varese, 23 April 2013 Prof Bjorn Olsen Harvard Medical School, USA Editor-in-Chief Journal of Negative Results in Biomedicine Dear Editor, Ref: MS: 1901793901915771 EFFECTS OF A COMMERCIAL PRODUCT CONTAINING GUARANÁ ON PSYCHOLOGICAL WELL-BEING, ANXIETY AND MOOD: A SINGLE-BLIND, PLACEBO- CONTROLLED STUDY IN HEALTHY SUBJECTS enclosed please find the revised version of our manuscript which has been carefully revised taking into account all the Referees' criticism and requirements. A point-by-point outline of our responses is enclosed and changes are highlighted in red throughout the manuscript. We would like to express our sincere gratefulness to the Referees for their careful evaluation and their indications which have significantly improved the quality of the manuscript, and we hope that it may be now found suitable for publication in Journal of Negative Results in Biomedicine. Sincerely, Many thanks and best regards. Marco Cosentino Marco Cosentino, MD PhD Center for Research in Medical Pharmacology University of Insubria, Via Ottorino Rossi n. 9 21100 Varese VA - Italy Phone +39 0332 217410/397410 Fax +39 0332 217409/397409 E-mail marco.cosentino@uninsubria.it
Ref: MS: 1901793901915771 EFFECTS OF A COMMERCIAL PRODUCT CONTAINING GUARANÁ ON PSYCHOLOGICAL WELL-BEING, ANXIETY AND MOOD: A SINGLE-BLIND, PLACEBO- CONTROLLED STUDY IN HEALTHY SUBJECTS Authors' responses to Referees' comments Referee #1 1. Under Results; psychological wellbeing; the final sentence "...of guarana and placebo either in either groups" - this sounds strange - please reformulate for better understanding. Corrected: " no significant difference occurred between the effects of guaranà and placebo in either groups". 2. Under Statistical power analysis. The handling of statistical power is extremely important in reports with negative findings. I find the handling in this manuscript very good. The only question I have, for clarification purposes, is if the power analysis is based on the total population or divided by genders. Maybe it could be specified which sample size has been used in the power calculations. Power calculations were performed on the whole sample, according to the main purpose of the study which did not include any specific hypothesis about possibly different effects in female vs male subjects. This has been now indicated in the text (see page 7, last paragraph). 3. In the discussion the authors discuss adverse effects. Was this measured in this study? If not, this should be stated. If it was measured, results should be reported. The first section of Results (Subjects) includes the statement that "No subjects reported any adverse event in any of the evaluation sessions." 4. In the method section it is stated that the subjects were in good health. How was this assessed? At enrollement, subjects were interviewed about their personal health and habits, including cigarette smoking, use of medicines and drugs. The sentence has been now reformulated to make clear that also information about health was self-reported (see page 13, line 6). 5. In the beginning of the method section it is stated that "...all the participants limited their coffe intake to no more than 4 cups/day...". However, under the procedure-part of the manuscript it is stated that "...subject were asked to abstain from alcoholic and caffeinecontaining beverages". Please be consistent and report what was done in the study.
We thank the Referee for the identification of an apparent contradiction in the study procedures. Actually, participants were told to abstain from caffeine-containing beverages (e.g. energy drinks, Coca-Cola, etc.) but were allowed up to a maximum of 4 cups of italian "espresso" per day, with the only exception of the morning of the weekly testing session, when they were asked to abstain also from coffee. Such conditions have been now better clarified in the text (see page 16, lines 16-18). Referee #2 The chosen study-setting is not capable of measuring any positive effect. Negative results might be independent of the study drug due to an inadequate study design: After 5 days of study drug treatment, the effect is measured more than 24 hours after the last intake. Even if the maximal effect of guaranà would be expected after more than one day after ingestion, measurement should have been performed under running treatment. To recent knowledge there is no one single oral taken substance, which would show a positive effect on measured items in this setting. It is more likely to discover withdrawal effects. Though, the study results would rather have to be quoted as no withdrawal effects after short course of guaranà therapy. The study design was chosen to measure long-term effects due to chronic intake, excluding any possible interference due to short-term effects of individual doses/servings of the product. In other words, the design actually was not chosen to detect the acute effects of caffeine contained in Guaranà, but to assess the unproven claims that support Guaranà-containing products as mood enhancers and able to increase feelings of well-being (as stated in the Introduction, from page 4, line 12, to the end of background, and in the Discussion, page 12, from line 4). While it is established that caffeine may have acute, short-term effects, Guaranà-containing products are nonetheless recommended for regular, long-term use without any clear directions for time and duration of intake (page 4, lines 5-6). Several websites also implicitly suggest that the effects may be long-lasting e.g. due to "the antioxydants [which] are what ensures a slow action, with a delayed effect, of the diffusion of the guaranine in the body, causing no nervousness or excitement" (http://www.goarana.ca/en/guarana). Some websites also deal with Guaranàcontaining products as if their regular intake would allow attainment of some steady-state condition (e.g.: "Missing a dose is probably not harmful. If you miss a dose, simply resume taking it on your previous schedule. Do not take double or extra doses." - http://www.einstein.edu/einsteinhealthtopic/). Information associated with such products never state that e.g. effects are only short term. Our purpose was therefore "to assess in healthy subjects the effects on psychological well-being, anxiety and mood of a commercially available guaranà preparation used according to the labeled dosages and instructions" (as stated at the end of the Introduction). As a consequence, the product was given according to the labelled advice, i.e. 360 mg x 3 daily just after breakfast, and tests were administered 24 hours later, implicitly assuming a steady state condition, as suggested by claims and directions of this kind of products (sse examples above). Failure to observe any effects was subsequently discussed also in view of the possibility that the main active ingredient of Guaranà products is actually caffeine (Discussion, from page 12, last 7 lines, to page 13, first 6 lines).
Finally, concerning the statement that "there is no one single oral taken substance, which would show a positive effect on measured items in this setting": indeed, as also discussed above, Guaranàcontaining products are marketed with claims of long term effects, therefore we deemed necessary to establish a study design able to prove (or disprove) such claims without any major interference due to the obvious short term effects of caffeine. Thus, once again we would like to emphasize that the purpose of the study was not to investigate the possible effects of Guaranà but to assess the validity of commercial claims. The study-setting must be clearly shown in the abstract; e.g., A daily consumption of 1080mg commercially available guaranà extract for 5 consecutive days does not show any effect after 1 day of last intake on well-being, anxiety and mood in this single-blinded study. In my opinion the measurement needs to be repeated and the study drug continued under measurements. As required by the Referee, the study design has been now explained in more detail even in the abstract. Reasons for chosing such design are discussed above. In such a small study all possible confounding effects should be eliminated. A sorrow documentation of possible other stimulants is mandatory. It is not clear what finally the caffeine intake of the participants was. On page 14 it is 4 cups/day and abstinence before testing. On page 16 were the subjects asked to abstain from alcoholic and caffeine-containing beverages. No information about the coffee consumption before and during the study is provided. We thank the Referee for the identification of an apparent contradiction in the study procedures. Actually, participants were told to abstain from alcoholic and caffeine-containing beverages (e.g. energy drinks, Coca-Cola, etc.) but were allowed up to a maximum of 4 cups of italian "espresso" per day, with the only exception of the morning of the weekly testing session, when they were asked to abstain also from coffee. Such conditions have been now better clarified in the text (see page 16, lines 16-18). Moderate coffee consumption was allowed to avoid any confounding effects due to modifications of established personal habits. The total caffeine intake was thus limited and considered not relevant for the study. Indeed, as also discussed above, the study design was chosen to exclude any interference due to the acute effects of caffeine from Guaranà (or from other sources). Thus 24 h after the last intake of Guaranà and at least an overnight without intake of coffee was deemed adequate to avoid any unintended acute effects of caffeine on the tests. Finally, no participants reported use of over-the-counter, medication and/or illicit drugs in the previous three months, thus also non-food sources of stimulants are unlikely. To evaluate well-being, anxiety and mood in females the menstrual cycle has to be taken into account. No information about it is provided. Fluctuation of psychological symptoms throughout the menstrual cycle may indeed occur in healthy women, however they are usually small and without clinical significance (see e.g. Gonda X et al., Progr Neuro-Psychopharmacol Biol Psychiatry 2008, 32: 1782 8). Concerning our subjects, there are several reasons to conclude that any bias due to differences in the individual timing of the menstrual cycle is unlikely: (i) females represented only 33% of the sample, (ii) patterns of test results in females were not significantly different from those obtained in males and (iii) in
particular standard deviations of the distribution of results were the same in females and males. As regards the scales used in the study, at least the PWB scales show no significant sex difference (see Ryff CD and Keyes CL, J Personal Social Psychol 1995, 69: 719-727). Finally, none of the female subjects were using a hormonal contraceptive method, thus also this kind of influence can be discarded. Information about hormonal contraception has now been made explicit (page 14, lines 7-8). In the study design it is questionable to perform a single-blind study without obvious reason why not double-blinding was possible. Power calculation or estimation has to be done on study planning. Double-blind was deemed unecessary since all the tests were self-administered and there was no opportunity for the occurrence of any investigator-dependent bias. As stated in the Methods, Data analysis section (from page 18, last 5 lines, to page 19, first 2 lines), sample size calculations could not be performed in advance, since there is no general consensus about the smallest average differences which would be scientifically important for each of the scales employed. We therefore decided to include a number of subjects comparable to those enrolled in studies which documented positive results concerning the acute effects of Guaranà (e.g. Haskell CF et al., J Psychopharmacol 2007, 21: 65-70), and to determine study power for each test after completion of the study to detect the least hypothetical difference between the effects of Guaranà and placebo. The section Methods stands before Results. Sections order is in agreement with the instructions for authors of the Journal of Negative Results in Biomedicine. Data on the returned capsules are missing. Subjects regularly took all the capsules in the two treatment periods, as stated on page 5, lines 13-14.