On 2014 Jul 29, Amanda Capes-Davis commented:

Please be aware that "Chang liver" cells are actually not from liver. The Chang liver cell line is known to be cross-contaminated with HeLa and comes from cervical adenocarcinoma. For a list of known cross-contaminated cell lines, see http://iclac.org/databases/cross-contaminations/.

On 2015 Sep 16, M Mangan commented:

An MD has taken a look at the claims in this paper. You can see that discussion here: https://www.sciencebasedmedicine.org/seneff-claims-gmos-cause-concussions/

On 2014 Oct 29, Roman Stilling commented:

Thank you for the comment. We realized that the paper is not associated with the search term "Gcn5" in Pubmed, hence I posted this comment. I had hoped that if the term appears in the comments it would appear in the search results, but it doesn't, as you pointed out correctly. While personally I'm all for "rationalizing" gene names, since today most genes are associated with a lot more functions than the name would suggest, I think databases should be able to deal with synonyms, i.e. Pubmed should show results for Gcn5 when Kat2a was searched and vice versa. It works well in the NCBI Gene database, so why not in Pubmed?

On 2014 Oct 22, L Charles Murtaugh commented:

A good example of the unintended consequences of "rationalizing" gene names - rendering opaque almost 30 years of literature! Note that if one searches Pubmed for "Gcn5," this paper does not appear among the hits.

On 2014 Oct 18, Roman Stilling commented:

For clarification of this abstract, it may be worth noting that K-Lysine acetyltransferase 2a (Kat2a) is more widely known by its former name General control of amino-acid synthesis 5 (Gcn5) or General control of amino-acid synthesis 5 like 2 (Gcn5l2). http://www.ncbi.nlm.nih.gov/gene/14534

On 2014 Aug 13, David Keller commented:

Smoking reduces risk of PD, but increases risk of COPD and of vascular Parkinsonism

It is well established that smoking is a major risk factor for COPD. Smoking has also been shown to reduce the risk of Parkinson disease (1), thought to be due to neuroprotection by nicotine (2). Therefore, based on the prior observed associations with smoking, one would expect the risk of PD to be lower in COPD patients than in patients with no smoking history. I propose that the unexpected association between COPD and PD may be due to misdiagnosis of vascular Parkinsonism as PD.

Smoking is a major risk factor for COPD and for atherosclerotic vascular disease (ASVD). Parkinsonism can be caused by ASVD, in which case it is categorized as vascular Parkinsonism (VP), a disease distinct from classic neurodegenerative PD. VP would be expected to correlate directly with COPD, while PD should be inversely correlated with COPD, based on the known association of smoking with ASVD, and its inverse association with PD.

COPD is expected to be associated with vascular Parkinsonism, but not with Parkinson's disease, based on the known relationships of these diseases with smoking. The results of this study would be consistent with these known associations if enough VP patients were misdiagnosed as having PD. I suggest that the data be re-examined to consider this possibility.

Lastly, the authors refer to dementia as a neurodegenerative disease. Dementia is actually a symptom which can be caused by neurodegenerative diseases (such as Alzheimer's disease or dementia with Lewy bodies), atherosclerotic disease (as in multi-infarct vascular dementia), other disorders, or a combination thereof. It is incorrect to categorize dementia due to atherosclerotic infarctions as a "neurodegenerative disorder".

References

1: van der Mark M, Nijssen PC, Vlaanderen J, Huss A, Mulleners WM, Sas AM, van Laar T, Kromhout H, Vermeulen R. A case-control study of the protective effect of alcohol, coffee, and cigarette consumption on Parkinson disease risk: time-since-cessation modifies the effect of tobacco smoking. PLoS One. 2014 Apr 30;9(4):e95297. doi: 10.1371/journal.pone.0095297. eCollection 2014. PubMed PMID: 24788751; PubMed Central PMCID: PMC4005732.

2: Ross GW, Petrovitch H. Current evidence for neuroprotective effects of nicotine and caffeine against Parkinson's disease. Drugs Aging. 2001;18(11):797-806. Review. PubMed PMID: 11772120.

On 2016 Dec 05, Anne Carpenter commented:

Please note that there is a typo for the microscope's TexasRed filter, which is mentioned in the Supplemental Materials. It is listed as excitation/emission (562/642 nm) when it is actually (562/624 nm). This was confirmed with the screening center that performed the microscopy described in the paper.

On 2016 Dec 05, Anne Carpenter commented:

As one of the senior authors on this paper, I report an error on page 14 of the Supporting Materials and Methods: the words "after cell fixation with 16% paraformaldehyde" should be replaced by "after cell fixation with 3.2% formaldehyde".

On 2016 Jan 06, Lydia Maniatis commented:

This article finally clarified what the authors of this and similar papers mean by “biologically-generated” or “evolved” stimuli. Here are the relevant statements:

The visual system achieves its goals on the basis of “the creation of light energy patterns whose frequency of occurrence tracks reproductive success...these patterns are “stimuli” ...Photon energy in the environment causes receptor cells in the retina to respond...It is only after the evolved apparatus of the eye collects, focuses, and selects a particular fraction of this energy into a biologically determined pattern that a stimulus exists...light stimuli are determined by a collaboration between physics and evolved biology.”

Thus, Purves et al “stimulus” is simply the physiological state of the retina, which is contingent on incoming light that is focussed. I don't think anyone would argue with that. It acknowledges that the “stimulus” is connected with the physics of the world around us, specifically, the light impinging on the retina.

However, this description of the visual stimulus does not in any way justify or explain the notion that the “frequency of occurrence” of these stimuli "tracks reproductive success.” The pattern of retinal stimulation changes with each movement of the eye, and the patterns, depending as they do on the chance arrangements of the environment and on the line of sight, on the time of day, etc, could fairly be described as “random.” Even if we could somehow define, categorise and track particular “patterns,” why should their frequency of occurrence track reproductive success? Also, since the slate, insofar as the frequency of occurrence of these light patterns is concerned, is wiped clean with every generation, how are these frequency distributions preserved?

The repeated claims that the visual system can't use the stimulus to infer features of the physical world is contrdicted by the authors, who say that “the visual system does not take the measure of physical reality...that we readily assess with the instruments of physics.” In fact, we readily assess using our perception of the physical properties of the instruments of physics. In addition, the fact that focussing of light achieves a point to point correspondence between points on the retina and points in the world that reflected the light is suggestive of an attempt to infer the shapes and locations of actual physical objects.

The claims about the inability of the v.s. to access properties of the physical world is made based on our perception of pictorial mimics of projections from real-world scenes. If I compare my perceptions of the color of objects in a naturally-lit real-life scene under inhomogeneous illumination, then my perception of the intrinsic surface properties of objects (e.g. reflectance, chromaticity) will appear quite reliable. If I then take a photo of that scene, and compare my perceptions of its surface properties to the actual properties of the photo surface, then, of course, errors all around. This doesn't justify denying the veridical outcome in the natural scene, or discounting it in accounts of how perception works.

The methods are oddly weak. "Three hundred and eight high-resolution natural red–green–blue (RGB) images from The University of Texas at Austin “set 1” database (7) served as a proxy for human spectral experience.” [The sampling problem should speak for itself. The sample was collected by a different team with different priorities. We are not given a description of the images, or of the rationale with which they were collected. The lighting conditions are not controlled for before or after the fact. Perceptual effects are not tested. The claim that it is a proxy for human evolutionary experience is to be taken on faith. The argument of this paper hinges on this database.]

On 2014 Nov 24, Guillaume Filion commented:

This article is one of the "CISCOM meta-analyses", which are very similar papers written by different authors. For more information about the CISCOM meta-analyses, check the blog post "A flurry of copycast on PubMed" at the following link http://blog.thegrandlocus.com/2014/10/a-flurry-of-copycats-on-pubmed

On 2014 Aug 04, MARK VRABEL commented:

Good to see this article. My (librarian) contributions have been acknowledged on www.guidelines.gov; for example, "With the assistance of the Oncology Nursing Society (ONS) librarian" at http://www.guideline.gov/content.aspx?id=15702

Mark Vrabel, MLS, AHIP, ELS Oncology Nursing Society twitter.com/ONSmark

On 2015 Mar 28, Edward Archer commented:

Hébert et al.’s defense of the status quo is an impediment to both scientific progress and empirically supported public nutrition policy.

For details please see the following:

"Implausible data, false memories, and the status quo in dietary assessment." Archer, E., Blair, S.N. Adv Nutr. 2015 Mar 13;6(2):229-30. doi: 10.3945/an.114.007799.

and

"Validity of U.S. nutritional surveillance: National Health and Nutrition Examination Survey caloric energy intake data, 1971-2010." Archer et al., PLoS One. 2013 Oct 9;8(10):e76632. doi: 10.1371/journal.pone.0076632.

Link: http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0076632

On 2014 Aug 30, Michelle Lin commented:

Excellent article on this relatively unrecognized epidemic. Week-long discussion on ALiEM blog with webcast discussion with authors.

http://www.aliem.com/opioid-prescription-epidemic-annals-em-resident-perspectives-article/

On 2014 Jul 28, Changting Liu commented:

Hi Willem, thanks for your attention on our manuscript. Some of your questions are really important and here is my answering. Firstly, the large differences on genome size. We noticed that the assembly results of these two strains were not good and some genome sequences were not obtained. But it was shown that the sequences of repeats and plasmids were much longer in strain LCT-KP289 and the some of the results were exhibited in Table1 and Table2. And the assembly results of repeats and plasmids can be affected by the sequencing data size and software parameters distinctively. On the other hand, the strains with special phenotypes were sequenced after space-travelling and laboratory-culturing, so there could be some mutations laid in the genome of LCT-KP289. Secondly, the acquisition of several antibiotic resistance genes and plasmids (some of enterococcal origin) in strain LCT-KP289. The manuscript has illustrated that the gene sul1 was multi-copy gene in Klebsiella pneumonia and there was one more copy exhibited in LCT-KP289. And other drug resistance genes which contained in integron were found both in LCT-KP214 and LCT-KP289, so it only explained the multiple drug resistances of Klebsiella pneumonia. As for the enterococcal plasmids, we should notice that it exist both in LCT-KP214 and LCT-KP289, so the plasmid should be captured by Klebsiella pneumonia before space traveling and it was not acquired from the medium or sequencing contaminant. Thirdly, the manuscript contains several errors (including a reported %GC of 84.42% for one of the strains). I apologize for this writing error, and the GC% of LCT-KP289 should be 56.78% and we are contacting the editor of BMC for the typo correction. Thanks again for your question. And if you have any more problems, please don't be hesitated to contact us. Thank you.

On 2014 Jul 15, Willem Schaik commented:

This paper attempts to show genomic differences between a Klebsiella pneumoniae strain that was incubated in the Shenzhou VIII unmanned spacecraft for 17 days and a control that remained on Earth. The reported differences between the two strains are large and include a 200 kbp difference in genome size and the acquisition of several antibiotic resistance genes and plasmids (some of enterococcal origin). These data are highly unlikely because when strains are grown in isolation, they should not be able to acquire foreign DNA from the medium. The authors do not explain these observations and simply claim that this is a result of spaceflight. In my opinion it is more likely that strains have been mixed-up or a contaminant has been sequenced. I also note that the manuscript contains several errors (including a reported %GC of 84.42% for one of the strains) and unsubstantiated claims.

On 2016 Aug 07, Nidal Jaradat commented:

Hard work

On 2017 Oct 05, ROBERT HURST commented:

Unfortunately the significance of these findings for bladder cancer is unclear because KU7 cells have been shown to be HeLa cells. PMC3805942

On 2014 Jul 15, T Eugene Day commented:

This is a very high quality paper. The validation metrics are appropriate and statistically validated. The data collection is high quality, and the model supports the conclusions the authors assert. The incorporation of patient punctuality, and separating it by new and follow-up distributions, is a particularly elegant piece.

And then, finally, they showed what happened when a simulated strategy was implemented in the real world, which is missing in a lot of the literature. One quibble there is that the table showing the results isn't completely clear. I'd like to see something along the lines of {Pre-implementation Real World, Simulated Performance Prediction, Post-Implementation Real World, Statistical test of Sim to Post-I Real, Statistical test of Pre-I real to Post-I real}. This would make it very clear how well the prediction fared in a single table.

All around extremely important work.

On 2014 Jul 18, William Cawthorn commented:

Several thoughts about this very interesting study:

1. Larger adipocytes are often taken as a surrogate for impaired adipogenesis, whereas smaller adipocytes are seen as indicative of enhanced adipogenesis. But lipogenesis, leading to adipocyte hypertrophy, also contributes to adipose tissue expandability (perhaps more so than adipogenesis). Therefore, I wonder if the smaller adipocytes at baseline actually represent impaired lipogenesis, and therefore impaired adipose tissue expandability?

2. Was there any association between adipocyte size at baseline, and changes in circulating lipids? I.e. the subjects with larger adipocytes had smaller change in SAT, VAT, and total fat mass, without accumulating more lipid in liver or SkM; therefore, is the excess lipid circulating instead?

3. The adipocyte size measurements are for subcutaneous WAT. It would be interesting to see if there were similar associations for the sizes of adipocytes in visceral WAT, which tends to be more inflamed.

4. The authors here find no association between SAT adipocyte size and expression of markers of inflammation. Another study in humans found that smaller SAT adipocytes were associated with increased inflammation (McLaughlin et al, 2010; PMID 19816674). Each of these studies are in contrast to the concept that adipocyte hypertrophy is associated with increased inflammation (e.g. Gustafson et al, 2009; PMID 19622783). Clearly, further studies in humans, both in subcutaneous and visceral WAT, are needed to determine if adipocyte hypertrophy is associated with increased WAT inflammation.

On 2014 Dec 03, MARY PATTI commented:

Thank you very much for your interest in our study. We agree that genetic variation is an important consideration when conducting comparative epigenetic studies. However, we would like to clarify that we used samples of sperm from a total of 8 control (C) and 8 undernutrition (UN) mice for MeDIP-seq analysis. We prepared 2 pools for each condition, each comprising 4 individuals from 4 independent litters. These pools were used for 2 independent analyses (C pool 1 vs. UN pool 1, and C pool 2 vs. UN pool 2); differentially methylated loci common to both analyses were identified comprising clusters of CGs behaving similarly, not single site differences. Subsequent validation of these loci was very important. Pyrosequencing validation of differential methylation was performed on unpooled sperm using 12 individual control males from 5 litters and 11 individual UN males from 4 litters. This approach minimized outcomes potentially linked to inter-individual genetic differences.

On 2014 Oct 17, Andrew Sharp commented:

Readers should note that the whole genome-profiling in this study used outbred mice, with a sample size of n=4 per group. In my opinion, this is a very serious limitation, as it will result in the identification of many differences between groups that are simply due to the underlying genetic variation of the mice being tested, and are unrelated to nutrition status.

On 2014 Aug 24, David Keller commented:

Please see the response posted on the BMJ website by Dr. Yedy Israel, who explains why caution must be used in projecting the results of this study to persons with wild-type genes for alcohol dehydrogenase. The effects of ethanol consumption on normals can only be determined by studying normals. The consistently-observed minimum risk for cardiovascular disease in normals occurs with a chronic daily ingestion of about 1 to 2 ounces of ethanol per day (1), the so-called "J-shaped curve".

Reference

1: O'Keefe JH, Bhatti SK, Bajwa A, DiNicolantonio JJ, Lavie CJ. Alcohol and cardiovascular health: the dose makes the poison…or the remedy. Mayo Clin Proc. 2014 Mar;89(3):382-93. doi: 10.1016/j.mayocp.2013.11.005. Review. PubMed PMID: 24582196.

On 2014 Jul 14, Dale D O Martin commented:

Once again, the myristoylation site 'predicted' in this paper is unlikely. N-myristoylation occurs on N-terminal Glycines hence the name N-myristoylation. It can only happen on internal glycine residues following proteolysis of the protein before the Glycine. As this protein has no cleavage site, it is unlikely to be myristoylated. I have pointed this out to one of the authors who agreed with this assessment and yet continues to publish this data.

Also, as far as I can tell, they do not reference what prediction tool they used for myristoylation.

On 2014 Jul 31, Amanda Capes-Davis commented:

The authors use HeLa, KB and MCF-7 as examples of human cancer cell lines. However, KB is known to be cross-contaminated with HeLa and so is cervical, not oral. For a list of cross-contaminated or otherwise misidentified cell lines, see http://iclac.org/databases/cross-contaminations/.

On 2014 Nov 24, Guillaume Filion commented:

This article is one of the "CISCOM meta-analyses", which are very similar papers written by different authors. For more information about the CISCOM meta-analyses, check the blog post "A flurry of copycast on PubMed" at the following link http://blog.thegrandlocus.com/2014/10/a-flurry-of-copycats-on-pubmed

On 2015 Feb 07, Airton Stein commented:

very interesting article.

On 2016 Dec 06, Pan Pantziarka commented:

Concerns about the validity of this review have been raised in the South African Medical Journal by Z. Harcombe and T. Noakes: "The universities of Stellenbosch/Cape Town low carbohydrate diet review: Mistake or mischief?", (S Afr Med J 2016;106(12):1179-1182. DOI:10.7196/SAMJ.2016.v106i12.12072. http://www.samj.org.za/index.php/samj/article/view/11605/7753.

On 2016 Jun 17, David C. Norris commented:

Four deficiencies in Dr Kessler's reply may be noted:

• Kessler describes as a "strong statement" my assertion that "a unitary PTSD construct cannot be presumed to support meaningful comparisons of prevalence across environments differing in ambient violence." Aside from his merely technical error (properly speaking, it is such a presumption and not its negation that is logically the 'strong' statement^1§15 ), by labeling my assertion as 'strong' Kessler overlooks its sharp qualification by "unitary" and "meaningful" -- words that point the way to an opportunity for productive critical discourse.

• Kessler asserts that my negating this presumption "dismisses psychiatric epidemiological research on PTSD, which is based [on it]". In so doing, he seems to employ no fewer than 3 logical fallacies² : He advances an argument from authority [of said research programme], and employs the fallacy of the excluded middle to produce a straw man caricature of my point. Far from 'dismissing' such epidemiologic research, my sharply qualified caution indicates how such research may be improved. From a point of view in which the meaning in "meaningful" is to be sought not in the tabulation of p-values but in the critical examination of open concepts³ , a PTSD epidemiologist may readily acknowledge a "unitary" PTSD construct -- ungraded as to severity and undifferentiated as to causation -- as worthy of critical appraisal.

• To my question, unmistakably cast in terms of construct validity, Kessler then responds in terms of criterion validity, ignoring a distinction that was articulated even at the very birth⁴ of construct validity 60 years ago. The phrase "construct validity" appears twice in my Letter; each time, I speak of the "construct validity of the instruments used" relative to "mental health" as (unmistakably) the target construct. Dr Kessler's discussion of "concordance" in para. 3 of his reply is completely irrelevant to this issue.

• When at last, in para. 4 of his reply, Kessler does begin to address a point of construct validity, he does so only from a perspective that is "primarily concerned with the adequacy of tests as measures of a construct rather than with the adequacy of a construct"^5p100 . Indeed, even then, only a convergent validity argument is pursued, with no effort at an exploration of discriminant validity (such as I solicited in the 2nd paragraph of my letter), which might have gone some way toward addressing my criticism.

[1] Popper, Karl. The Logic of Scientific Discovery. 2 edition. New Delhi: Routledge, 2002.

[2] Sagan, Carl. “Chapter 13: The Fine Art of Baloney Detection.” In The Demon-Haunted World: Science as a Candle in the Dark, 1996. See pp. 212-216.

[3] Meehl, Paul E. “Theoretical Risks and Tabular Asterisks: Sir Karl, Sir Ronald, and the Slow Progress of Soft Psychology.” Journal of Consulting and Clinical Psychiatry 46 (1978): 806–34. Full text

[4] CRONBACH LJ, 1955 Full text

[5] CAMPBELL DT, 1959 Full text

On 2014 Jul 31, Paul Glasziou commented:

We agree with Bastian's comments, but getting authors to do this will not occur soon! However, journals could potentially use automated tools to assist with extraction of trial details and provision of UTNs. Manual extraction, whether by author, journal, or reviewer, should become a thing of the past.

On 2014 Jul 26, Hilda Bastian commented:

A thorough and valuable breakdown of what could and should be automated in systematic reviewing. One additional important strategy lies in the hands of everyone doing (and publishing) clinical trials and systematic reviews: following the IJCME recommendation to include the clinical trial registry identification number of every trial at the end of abstracts. This needs to be done using the specific, unaltered formats for each registry in which a study is included, so that IDs are easily retrievable - and the IDs should be with every cited study inside the systematic review, too. Using the WHO's Universal Trial Number (UTN) would also help with the critical, and time-consuming, task of study de-duplication.

The issue raised in this article of some databases of manually extracted trial data not being publicly available is an important one. It's worth noting, though, that this is not because it's not possible: systematic reviewers have the option of using the open and collaborative public infrastructure of the SRDR (Systematic Review Data Repository) (Ip S, 2012).

Another option to add to the list of ways of improving the snowballing technique for identifying studies: using the related articles function in PubMed. That's been found to be useful in empirical studies of techniques for updating systematic reviews (Shojania KG, 2007).

(Disclosure: I work on projects related to systematic reviews at the NCBI (National Center for Biotechnology Information, U.S. National Library of Medicine), which is also responsible for ClinicalTrials.gov.)

On 2015 May 30, Torsten Seemann commented:

The software is available at: https://sourceforge.net/projects/simulatepcr/

On 2014 Jul 15, Elisabeth Bik commented:

There seems to be a typo in the title: "Escherchia".

On 2016 Aug 23, Ben Goldacre commented:

This trial has the wrong trial registry ID associated with it on PubMed: both in the XML on PubMed, and in the originating journal article. The ID given is NCT00602603. We believe the correct ID, which we have found by hand searching, is NCT00603603.

This comment is being posted as part of the OpenTrials.net project^[1] , an open database threading together all publicly accessible documents and data on each trial, globally. In the course of creating the database, and matching documents and data sources about trials from different locations, we have identified various anomalies in datasets such as PubMed, and in published papers. Alongside documenting the prevalence of problems, we are also attempting to correct these errors and anomalies wherever possible, by feeding back to the originators. We have corrected this data in the OpenTrials.net database; we hope that this trial’s text and metadata can also be corrected at source, in PubMed and in the accompanying paper.

Many thanks,

Jessica Fleminger, Ben Goldacre*

[1] Goldacre, B., Gray, J., 2016. OpenTrials: towards a collaborative open database of all available information on all clinical trials. Trials 17. doi:10.1186/s13063-016-1290-8 PMID: 27056367

* Dr Ben Goldacre BA MA MSc MBBS MRCPsych<br> Senior Clinical Research Fellow<br> ben.goldacre@phc.ox.ac.uk<br> www.ebmDataLab.net<br> Centre for Evidence Based Medicine<br> Department of Primary Care Health Sciences<br> University of Oxford<br> Radcliffe Observatory Quarter<br> Woodstock Road<br> Oxford OX2 6GG

On 2014 Nov 20, Neven Patrick commented:

Although results from ongoing prospective trials looking into the predictive value of multigene signatures (MGS) for chemotherapy benefit are pending, many use them for prognostic purposes. In general, this leads to less adjuvant chemotherapy. This paper reported on how a multidisciplinary team (MDT) changed treatment decisions studying 75 patients with a grade 1-2 cT1-2N0 ER-positive, HER2-negative breast cancer. Clinical risk by MDT based on (St Gallen criteria) pathological tumor size, grade, nodal stage, steroid receptors and Ki-67 was compared with MammaPrint® (MP) risk. There was discordance between clinical and MP-risk in 29 out of 75 patients; 21 had a clinical high, MP low and 8 a clinical low, MP high risk. In this group of 29, 4 out of 8 with a clinical low, MP high risk did and 10 out of 21 in the clinical high, MP low risk did not get chemotherapy. All together, based on clinical risk alone, 31 of 75 women would have received adjuvant chemotherapy while this changed to 25 of 75 when incorporating the MP-risk. The use of MGS in this cohort was cost effective. The paper however, lacks some important details that are relevant and will help understand the use of MGS in this MDT. First, the discordance rates between clinical and MGS-risk in this paper are much higher for the clinical high risk patients (67%)compared with other experiences with MGS, like MP in RASTER or MINDACT (Bueno-de-Mesquita et al., 2007; Rutgers et al., 2011). This might be related to an overestimation of the clinical or underestimation of the MGS-risk. The definition they used for clinical risk is lacking and suggests an overestimation of clinical risk as 41% of women with a low to intermediate risk were stratified by the MDT as high risk. Otherwise, MGS might underestimate relapse risk. It would be interesting to know from the patients’ demographics whether some of the patients stopped exogenous hormones prior to breast cancer surgery. Short term changes in the hormonal environment following an ER-positive breast cancer diagnosis on core needle biopsy might confound MGS results. This has been shown for proliferation markers other than MGS for example with hormone replacement therapy withdrawal. Second, the authors state why they decided to give adjuvant chemotherapy in 11 out of 21 women with a clinical high, MP low risk, but they did not indicate why adjuvant chemotherapy was or was not given to half of the patients with a clinical low, MP high risk.

On 2014 Jul 12, David Keller commented:

The USPSTF statement confuses sensitivity and specificity

The following error appears in the USPSTF Recommendation statement posted on their own website (1) as well as in the identical document published online by Annals of Internal Medicine (2):

"Although screening with ultrasonography has few direct harms, all screening strategies, including those with or without confirmatory tests (that is, digital subtraction or magnetic resonance angiography), have imperfect sensitivity and could lead to unnecessary surgery and result in serious harms, including death, stroke, and myocardial infarction."

Correction: The author meant to write that these screening tests "have imperfect specificity, and could lead to unnecessary surgery..."

Rationale: Imperfect specificity is defined as a rate of false-positive test results greater than zero, which can lead to a diagnosis of disease which does not exist, and hence to unnecessary surgery. In contrast, imperfect sensitivity is defined as a rate of true-positive test results less than 100%, which does not lead to unnecessary surgery - but it can lead to the opposite kind of harm, which is a missed diagnosis, which can lead to the failure to perform necessary surgery.

The correct use of these terms, and the distinction between their definitions, is important in any discussion of public health measures.

The erroneous word was bolded by this commentator.

9/22/2014: As of today, the above error remains uncorrected on both the USPSTF and Annals websites. I have notified the editor of Annals by email. I encourage readers to submit their comments regarding this error to USPSTF, Annals, and PubMed Commons.

11/9/2014: As of today, the above error remains uncorrected on the Annals of Internal Medicine website. However, it has been corrected on the USPSTF website, at the following new web address:

http://www.uspreventiveservicestaskforce.org/Page/Document/RecommendationStatementFinal/carotid-artery-stenosis-screening

I will again attempt to have it corrected by the editors of Annals.

12/26/2014: I exchange emails with the editor-in-chief of Annals, who assures me that the matter will be looked into.

1/21/2015: A recheck of the Annals website discloses that the error has been corrected, and now reads "Although screening with ultrasonography has few direct harms, all screening strategies, including those with or without confirmatory tests (that is, digital subtraction or magnetic resonance angiography), have imperfect sensitivity and specificity and could lead to unnecessary surgery and result in serious harms, including death, stroke and MI".

To date, I have received no acknowledgement or thanks for pointing out the above error, from either the USPSTF or from Annals. This is the typical response I have received when correcting errors, including other, more serious errors in JAMA and the NEJM, which were eventually corrected after I expended almost preposterous efforts. Errors in the scientific literature can propagate unless corrected, leading to further errors. The NIH should establish an impartial board of experts, to whom outstanding errors in PubMed-indexed publications can be appealed, in cases where readers have spotted significant errors and the responsible editors refuse to correct them.

References

1: U.S. Preventative Services Task Force website, accessed on 7/12/2014, search for the second instance of "imperfect sensitivity" http://www.uspreventiveservicestaskforce.org/uspstf13/cas/casfinalrs.htm

2: Annals of Internal Medicine website, accessed on 7/12/2014, search for the second instance of "imperfect sensitivity" http://annals.org/article.aspx?articleid=1886690

On 2014 Jul 12, David Keller commented:

This USPSTF statement contains two different recommendations, which must be discussed separately

1) The USPSTF recommends against screening for asymptomatic carotid artery stenosis in the general adult population. Grade D

2) The USPSTF recommends against invasive mechanical treatment (ie: surgery or angioplasty) for asymptomatic carotid artery stenosis in the general adult population. Grade D.

The USPSTF statement convinced me that their grade of "D" is valid for recommendation #2 above; invasive treatments of asymptomatic carotid stenosis probably do more harm than good.

However, the USPSTF is clearly wrong on recommendation #1 above. Their statement admits that there is insufficient data to determine whether high-dose statin therapy will prove to be as beneficial for carotid atherosclerotic disease as it has proven for coronary atherosclerotic disease. Therefore, by their own grading standards, recommendation #1 should be classified as Grade "I" for indeterminate.

Further, many experts maintain that atherosclerotic stenoses, whether discovered in a carotid artery or a coronary artery, are manifestations of the same underlying disease entity, systemic atherosclerosis, which if present in a significant degree, should be treated with high-dose statin therapy. For those who hold this view, screening asymptomatic patients for carotid stenosis should be given a grade of "C", meaning each individual should discuss with their physician whether they should get screened, based on their risk profile.

I screen asymptomatic patients who have a reasonable risk of atherosclerosis, to identify systemic atherosclerosis in a convenient artery which is accessible to ultrasonic evaluation. If carotid stenosis is present, I treat is as a coronary artery disease risk equivalent, and initiate treatment with high-dose statin therapy. This strategy is reasonable, based on extrapolation of the findings of benefit in every cardiac statin trial.

The arguments supporting carotid screening for the purpose of risk-stratification are biologically plausible and lead to active changes in therapy for patients without known coronary artery disease, who are found to have asymptomatic carotid stenosis.

The USPSTF should rewrite this statement to narrowly focus on recommending against invasive treatment of carotid stenosis, and encourage physicians to continue screening appropriate patients for asymptomatic carotid atherosclerosis as a way to identify additional patients who can benefit from the use of high dose statin therapy.

On 2016 Mar 18, ZHONGMING ZHAO commented:

My lab recently moved to the University of Texas Health Science Center at Houston. The dataset now available at https://bioinfo.uth.edu/kinomenetworkX/.

On 2014 Sep 19, Ian Dworkin commented:

In addition to all of the data and scripts being on Dryad, we also have them in a repository on github.

https://github.com/DworkinLab/Pitchers_PTRS2014

On 2014 Nov 22, Ruchira Engel commented:

We recently discussed this article in our journal club. Our comments can be found on our journal-club blog Sanquin Digests.

On 2014 Jul 24, Ryan Radecki commented:

Post-publication commentary:

"ED Hocus POCUS ... or Just a Hoax?"

A landmark paper recently published in Lancet Respiratory Medicine is certainly destined to send the ED Ultrasound world into a tizzy. This is the first RCT examining the utility of Emergency Department based Point of Care ultrasound (POCUS) for patients presenting with undifferentiated respiratory complaints. Authors randomized patients presenting to the ED with signs or symptoms concerning for a respiratory etiology to either a standard work up as determined by the treating physician or the addition of POCUS performed by a single experienced operator. The US protocol consisted of sonographic examination of the heart, lungs and lower extremity deep veins to identify possible causes of patients' symptoms. The authors' primary outcome was the percentage of patients with a correct presumptive diagnosis 4 hours after presentation to the Emergency Department as determined by two physicians blinded to ED POCUS findings, but with access to the records of the entire hospital stay....

http://www.emlitofnote.com/2014/07/ed-hocus-pocus-or-just-hoax.html

On 2014 Jul 28, Amanda Capes-Davis commented:

Please be aware that HEp-2 is not a laryngeal squamous carcinoma cell line. HEp-2 is known to be cross-contaminated with HeLa and is actually cervical adenocarcinoma. For a database of known cross-contaminated cell lines, please see http://iclac.org/databases/cross-contaminations/.

On 2014 Aug 07, Madhusudana Girija Sanal commented:

This is an interesting observation. There is another aspect. It is known that platelets carry serotonin and they play an important role in distributing extra-neural serotonin.Platelet depletion impairs liver regeneration. Lesurtel M et al reported in 2006 that platelet-derived serotonin mediates liver regeneration (Science. 2006 Apr 7;312(5770):104-7.). Another paper by Paulose and others, back in 1998 reported increased mitotic activity in liver cells in-vitro in response to serotonin. (Hepatology. 1998 Jan;27(1):62-6.). It is therefore possible that platelets may play a role in hepatocellular carcinoma-proliferation and metastasis. Thus, serotonin may have an additional role different from the growth factors such as PDGF in liver cancer secreted by platelets. However, I doubt if this has some direct clinical implication in the management of this cancer considering its extremely dynamic nature.

On 2014 Jul 09, Ryan Radecki commented:

Post-publication commentary:

"Build a New EDIS, Advertise it in Annals for Free"

As everyone who has switched from paper to electronic charting and ordering has witnessed, despite some improvements, many processes became greatly more inefficient. And – it doesn’t matter which Emergency Department information system you use. Each vendor has its own special liabilities. Standalone vendors have interoperability issues. Integrated systems appear to have been designed as an afterthought to the inpatient system. We have, begrudgingly, learned to tolerate our new electronic masters....

http://www.emlitofnote.com/2014/07/build-new-edis-advertise-it-in-annals.html

On 2014 Dec 30, William Grant commented:

Geographical variations in solar ultraviolet-B doses may explain some of the spatial-temporal excess mortality patterns of the 1918-1919 influenza pandemic in Spain

Disclosure I receive funding from Bio-Tech Pharmacal (Fayetteville, AR), the Sunlight Research Forum (Veldhoven), and Medi-Sun Engineering, LLC (Highland Park, IL).

The paper by Chowell and colleagues presents maps of excess respiratory death rates during the 1918-1919 influenza pandemic in Spain 1. The summer and winter peak rates are in the center and south of Spain, while the fall peak rates are predominantly in the north. The authors suggested that colder temperatures and lower humidity in the north and in fall-winter might explain the observed patterns. I agree that this suggestion is generally correct. However, there is another factor that should also be considered - the role of solar UVB irradiance and vitamin D production. In an ecological study of the 1918-1919 influenza pandemic in the United States, it was reported that both summertime and wintertime solar UVB doses were inversely correlated with case-fatality rates for a dozen communities 2. Most of the deaths associated with the influenza were due to pneumonia, and occurred about ten days after the start of influenza. The mechanisms suggested were induction of cathelicidin, which has antibiotic properties, and reduction of the cytokine storm due to the body fighting influenza. The UVB-vitamin D-influenza hypothesis was proposed in 2006 by John Cannell and colleagues 3. This hypothesis has gained further support from vitamin D randomized controlled trials involving people with low 25-hydroxyvitamin D-25OHD concentrations at the time of enrollment in the trials 4-6.

Broadband solar UVB doses have been measured for 14 locations in Spain 7. In summer, values at solar noon ranged from 1234 to 1726 units but do not follow a strict latitude gradient. In winter, the values ranged from 160 to 348 and showed a strong latitudinal gradient. Based on these values, the higher mortality rates in winter and in the north may be due in part to lower UVB doses in winter. The higher rates in south and central Spain in summer may be related to where influenza was first introduced into Spain. A related paper found that cancer mortality rates were often inversely correlated with latitude and directly correlated with nonmelanoma skin cancer 8.

This analysis suggests that 25OHD concentrations be raised by vitamin D supplementation in winter in Europe to reduce the risk of respiratory infections. Optimal 25OHD concentrations are between 75 and 125 nmol per L 9, which can be achieved with 800-2000 IU per day vitamin D3 for most people 10. There are also many other beneficial effects of higher 25OHD concentrations 9, 11.

References 1. Chowell G, Erkoreka A, Viboud C, Echeverri-Dávila B. Spatial-temporal excess mortality patterns of the 1918-1919 influenza pandemic in Spain. BMC Infect Dis. 2014;14:371. 2. Grant WB, Giovannucci E. The possible roles of solar ultraviolet-B radiation and vitamin D in reducing case-fatality rates from the 1918–1919 influenza pandemic in the United States. Dermatoendocrinol. 2009;1:215-9. 3. Cannell JJ, Vieth R, Umhau JC, Holick MF, Grant WB, Madronich S, Garland CF, Giovannucci E. Epidemic influenza and vitamin D. Epidemiol Infect. 2006;134:1129-40.<br> 4. Aloia JF, Li-Ng M. Re: epidemic influenza and vitamin D. Epidemiol Infect. 2007 Oct;135:1095-6; author reply 1097-8. 5. Urashima M, Segawa T, Okazaki M, Kurihara M, Wada Y, Ida H. Randomized trial of vitamin D supplementation to prevent seasonal influenza A in schoolchildren. Am J Clin Nutr. 2010;91:1255-60. 6. Camargo CA Jr, Ganmaa D, Frazier AL, Kirchberg FF, Stuart JJ, Kleinman K, Sumberzul N, Rich-Edwards JW. Randomized trial of vitamin D supplementation and risk of acute respiratory infection in Mongolia. Pediatrics. 2012;130:e561-7. 7. Martínez-Lozano JA, Utrillas, MP, Nunez JA, Esteve AR, Gomea-Amo JL, Estelles V, Pedros R. Measurement and Analysis of Broadband UVB Solar Radiation in Spain. Photochemistry and Photobiology, 2012;88:1489–96 8. Grant WB. An ecologic study of cancer mortality rates in Spain with respect to indices of solar UV irradiance and smoking. Int J Cancer. 2007;120:1123-7. 9. Pludowski P, Holick MF, Pilz S, Wagner CL, Hollis BW, Grant WB, Shoenfeld Y, Lerchbaum E, Llewellyn DJ, Kienreich K, Soni M. Vitamin D effects on musculoskeletal health, immunity, autoimmunity, cardiovascular disease, cancer, fertility, pregnancy, dementia and mortality- a review of recent evidence. Autoimmun Rev. 2013;12:976-89. 10. Pludowski P, Karczmarewicz E, Bayer M, Carter G, Chlebna-Sokół D, Czech-Kowalska J, Dębski R, Decsi T, Dobrzańska A, Franek E, Głuszko P, Grant WB, Holick MF, Yankovskaya L, Konstantynowicz J, Książyk JB, Księżopolska-Orłowska K, Lewiński A, Litwin M, Lohner S, Lorenc RS, Lukaszkiewicz J, Marcinowska-Suchowierska E, Milewicz A, Misiorowski W, Nowicki M, Povoroznyuk V, Rozentryt P, Rudenka E, Shoenfeld Y, Socha P, Solnica B, Szalecki M, Tałałaj M, Varbiro S, Zmijewski MA. Practical guidelines for the supplementation of vitamin D and the treatment of deficits in Central Europe - recommended vitamin D intakes in the general population and groups at risk of vitamin D deficiency. Endokrynol Pol. 2013;64:319-27. 11. Hossein-Nezhad A, Holick MF. Vitamin D for Health: A Global Perspective. Mayo Clin Proc. 2013;88:720-55.

On 2016 May 09, Attila Becskei commented:

Keung et al write that “An interesting result that emerged from our CR [chromatin regulator] library screens is that transcription can be repressed but not activated from downstream of a gene. Furthermore, activators do not appear to display the long-range properties that some repressors do”. This result confirms the conclusion of an earlier work (1). At the same time, Keung et al implicitly re-classified some repressors as activators (2).

(1) Basic mechanisms of transcription are conserved among eukaryotes; however, there are interesting quantitative differences. While genes can be activated from distant sites in higher eukaryotes, yeast genes can be controlled only from nearby sites in wild-type cells (Dobi KC, 2007). Only the Sir (silent information regulator) proteins have been known to display long-range (> 1kb) repression in yeast, which have a strong effect on chromatin structure(Nasmyth KA, 1982). Later, our study revealed that even classical repressors, such as Ssn6, can repress genes from distant sites provided they are positioned downstream of a gene (Ratna P, 2009,Kelemen JZ, 2010). An activator recruited to these downstream sites did not affect gene expression. In a set of synthetic gene reporters, three regulators (Ssn6, Sum1 and Sir3) had similar ranges of repression (Figure 3A in Becskei A, 2011). The range of repression from upstream (U) of a gene was considerably shorter than that from downstream (D) of the gene.

Using similar set of synthetic gene reporters, Keung et al recruited all known regulators to downstream sites. Interestingly, the our above conclusion was found to be true for all regulators: genes can be repressed but not activated from downstream sites.

(2) This uniform behavior of the repressors at the downstream sites is also interesting because a variable, non-uniform, behavior was observed by the authors at the upstream sites, i.e. at the promoter. For example, the classical yeast repressors, Ssn6, Tup1 and Ume6 did not repress expression when recruited to the promoter. Instead, they turned out to be the most potent activators of all regulators: they activated gene expression around 100 times (Figure 2B) (< PMID: 24995982>). Previous studies indicate that they are potent repressors both in wild-type and synthetic contexts (Keleher CA, 1992,Smith RL, 2000,Kadosh D, 1997).

On 2015 Feb 27, Geriatric Medicine Journal Club commented:

This cross sectional and longitudinal analysis of a prospective cohort of osteoporosis in men was critically appraised at the November 2014 Geriatric Medicine Journal Club (follow #GeriMedJC on Twitter). Full details of the discussion can be found at: http://gerimedjc.blogspot.com/2014/11/gerimedjc-november-28-2014.html?spref=tw One of the highlights was the concern that the Multicenter Osteoporotic Fractures in Men (MrOS) cohort did not include many people on steroids.

On 2014 Jul 17, Alexander Weiss commented:

Wilson et al. tested how participants reacted to doing no more than thinking. They couched this research question in the human ability to engage in “directed conscious thought” and asked whether this activity was pleasant. It was not. In fact, many participants self-administered electric shocks when given the opportunity to do so during the thinking task. Discussing their findings, the authors stress the “thinking” aspect of their tasks and concluded that “people prefer doing to thinking”.

I question this interpretation. Removal of stimuli individuals find pleasing (“doing”), is a means of shaping behavior and is as aversive to chimpanzees and pigeons (Ferster, 1958) as it was, presumably, to the participants of this study. The importance of “doing” for animals by providing environmental enrichment in the form of puzzles, games, and social contact is enshrined in animal welfare regulation (see the 1985 amendment to the 1966 United States Animal Welfare Act) based on decades of research showing, for example, that a lack of enrichment is associated with self-injurious behavior (Lutz & Novak 2005).

Engaging in “directed conscious thought” is thus either common across a range of different species or it is not the aversive stimuli in question. The over-interpretation of well-known phenomena, however, is probably unique to humans and most definitely aversive.

References

C. B. Ferster, Control of behavior in chimpanzees and pigeons by time out from positive reinforcement. Psychological Monographs: General and Applied 72, 1-38 (1958).

C. K. Lutz, M. A. Novak, Environmental enrichment for nonhuman primates: Theory and application. ILAR J. 46, 178-191 (2005).

On 2015 Aug 06, David Keller commented:

Possible bias in the reporting of test results for this weight-loss product

The main author of this study reports conflicts of interest with regard to the green-plant membrane weight-loss product being tested, which is sold under the brand name "Appethyl" [1]. Her positions as scientific adviser, part-owner and member of the board of directors of companies that produce Appethyl could bias her evaluation of its efficacy. Results from her research are presented in promotional materials for Appethyl [2] and quoted in the popular media [3].

An example of possible bias involves two subjects who dropped out of this study after they were randomized to the control group. Their weight data was simply deleted, yielding an average weight loss in the control group of 3.5 kg. Intention-to-treat analysis, performed by carrying forward the starting weight in Table 1 for each of the control group drop-outs, yields an average weight loss in the control group of 3.9 kg. A smaller weight loss in the control group makes the Appethyl-treated group's average weight loss appear larger and more significant, by comparison to placebo. The U.S. Food and Drug Administration requires intention-to-treat analysis in studies of products submitted for their approval [4], which would be required if Appethyl is to be offered in the U.S.

The NIH suggests the following measures be taken to ameliorate financial conflicts of interest [5]:

"Change of personnel or personnel responsibilities, or disqualification of personnel from participation in all or a portion of the research;

"Reduction or elimination of the financial interest (e.g., sale of an equity interest); or

"Severance of relationships that create financial conflicts."

A safe and effective weight loss product derived entirely from natural green plant sources is a welcome possibility. The existence of significant conflicts of interest requires great care to prevent bias in the reporting of experimental results for Appethyl.

References:

1: Science Direct website, accessed on 8/6/2015, at the following URL: http://www.sciencedirect.com/science/article/pii/S0195666314003493

2: Online press release published 3/10/2014, "New Clinical [sic] Shows That Greenleaf Medical's Appethyl(TM) Decreases Urge to Eat Sweet Foods", SOURCE: Greenleaf Medical AB, accessed on 8/20/2015 at http://www.marketwatch.com/story/new-clinical-shows-that-greenleaf-medicals-appethyltm-decreases-urge-to-eat-sweet-foods-2014-03-10

3: Commercial sale of Appethyl on Amazon.com, accessed on 9/12/2015, at: http://www.amazon.com/Appethyl-30-ct/dp/B00RM5KY48/ref=cm_cr_pr_product_top?ie=UTF8

4: Guideline for the Format and Content of the Clinical and Statistical Sections of Applications, Center for Drug Evaluation and Research, Food and Drug Administration, Department of Health and Human Services, 5600 Fishers Lane Rockville, Maryland 20857 (301) 443-4330. 1988 Jul 24

5: ELECTRONIC CODE OF FEDERAL REGULATIONS, Subpart F—Promoting Objectivity in Research, accessed on 8/20/2015 at http://www.ecfr.gov/cgi-bin/text-idx?c=ecfr&SID=992817854207767214895b1fa023755d&rgn=div5&view=text&node=42:1.0.1.4.23&idno=42#sp42.1.50.f

On 2015 Sep 14, David Keller commented:

Blinding of study subjects was not assessed, and likely failed, causing Pygmalion effect

In this study, the control group subjects were given a beverage containing 2.8 grams of rapeseed oil and 50 grams of blueberry "soup". The active intervention group subjects were given the same beverage, with the addition of 5 grams of a commercial green spinach extract powder called "Appethyl". No assessment was made of whether the subjects remained blinded after seeing and tasting these beverages, which must have differed markedly in appearance and taste. It is difficult to imagine that the study subjects would not quickly figure out who was getting the placebo beverage, and who was getting the active intervention beverage. Once this was known, uncontrolled Pygmalion effect [1] is likely in a weight loss clinical trial. The Pygmalion effect is defined as enhanced benefit that occurs when a subject becomes aware that she is receiving the active experimental intervention. Pygmalion effect is closely related to placebo effect; in the limit, if the product being studied is actually ineffective, then the Pygmalion effect becomes equal to the placebo effect.

Clinical studies should evaluate the quality of the blinding achieved and maintained throughout the study, in order to exclude or quantify the biases and and confounding factors introduced when subjects know whether they received placebo or the experimental therapy. Assessment of blinding quality could have been obtained by asking the subjects to provide their best guess as to whether they received the placebo beverage with or without added Appethyl, and tracking those results over time. With perfect blinding, all subjects should guess with the random accuracy of a fairly flipped coin, on the average (50%). With complete unblinding, all subjects guess correctly whether they received placebo or the therapy being studied.

Perhaps Higgs and Geary can reveal further heretofore unpublished data to settle this question.

References:

1: Draper,S.W. The Hawthorne, Pygmalion, Placebo and other effects of expectation, published on 5/14/2002, and visited on 5/14/2015, at:<br> http://www.psy.gla.ac.uk/~steve/hawth.html#Hawthorne overall

2: Bang H, Ni L, Davis CE. Assessment of blinding in clinical trials. Control Clin Trials. 2004 Apr;25(2):143-56. PubMed PMID: 15020033.

On 2015 Aug 04, David Keller commented:

The data in Table 1 contain important discrepancies, which require discussion & correction

Table 1 states that the average starting weight for the 19 control subjects who began the trial was 80.2 kg, and the average ending weight was 76.3 kg for the 17 control subjects who finished the trial. Given this data, one can calculate the average weight loss of 3.9 kg by intention-to-treat analysis using the "last data point carried forward" technique. The authors of this paper chose to report the per-protocol average weight loss, which is impossible to calculate from Table 1, because it is missing the average starting weight of the 17 control patients who completed the study. None of this was disclosed anywhere in the paper, nor was it disclosed that the weight loss results were not analyzed by intention-to-treat analysis, which preserves the statistical qualities of the original randomization. Deleting control patients from a study degrades its statistical randomness and increases the effect of confounding factors. That is why intention-to-treat analysis is the scientific standard for analyzing data from clinical trials.

Further, there is no way for readers to check these calculations or the integrity of the paper's conclusions, due to the data missing from Table 1. This is an important point, because dropping the data from the 2 control patients decreased the average weight loss by control group subjects from 3.9 kg to 3.5 kg. This, in turn, increased the apparent average net weight loss among subjects treated with Appethyl.

The editors of Appetite should encourage readers to report published errors, because such reports are the last line of defense of the truth in scientific publishing. False counter-charges of error made against such readers, such as in their reply to me, will only discourage readers from pointing out published errors in the future.

On 2015 Aug 04, David Keller commented:

The authors DID specify the data variability as SEM, not SD, as falsely claimed by Higgs and Geary

The following quote is taken from the last paragraph of the Materials and Methods section of this paper: "In figures and text, data are expressed as the mean,  plus or minus the standard error". There is no other statement in this paper defining the data format. Interpreted using the data format specified by the authors, the study's reported weight loss with Appethyl is not significantly better than achieved with placebo, which I reported in an earlier comment. In their reply above, Higgs & Geary, the co-editors of Appetite, state that the study's final results are actually reported in the format of the mean plus or minus the standard deviation, a format not specified anywhere in the paper. It is only by substituting the use of the undocumented SD variability format that the weight loss with Appethyl is made to appear significantly better than with placebo. If, indeed, the authors decided to change data formats, that should have been documented in the paper. Rather than thanking me for identifying this important error in the paper, the editors of Appetite attack my conclusion that Appethyl is statistically ineffective, which is the only possible interpretation of the data using the format specified by the authors. My observations exposed an important error in this paper, which they should be grateful to have corrected by a reader.

On 2015 Aug 04, Suzanne Higgs commented:

Dr. David Keller commented that, contrary to the authors' conclusion, green-plant membrane consumption did not significantly reduce body weight in this twelve-week trial, that the control weight-loss data in Table 1 and in the abstract are discrepant, that an intention-to-treat analysis was not done, and that work may have been biased by the authors' financial interests. The authors did not consistently describe the numbers of subjects in the control group and did not explicitly identify whether the variabilities reported in Table 1 were SD or SEM. Dr. Keller’s suspicions regarding the significance of the weight-change data arise from this lack of clarity. In fact, the variabilities reported in the abstract and Table 1 are mean ± SEM and the mean (95% CI) three-month weight losses are 5.0 (3.97 – 6.03) kg for the 19 subjects receiving green-plant-membrane and 3.5 (2.41 – 4.59) kg for the 17 of the 19 initial control subjects who finished the study, and these two means are significantly different by t-test. A correction to the original article has now been published that clarifies these issues. The analytic approach, which was not an intention-to-treat approach, and how the missing data were handed are described in the methods and results section. Finally, the senior author fully reported her conflicts of interest to the satisfaction of the journal. Appetite appreciates Dr Keller's close study of this report.

Suzanne Higgs, PhD

Nori Geary, PhD

Co-Executive Editors, Appetite

On 2015 Jun 29, David Keller commented:

Results of this study are reported inconsistently and without intention-to-treat analysis

The weight-loss results of this study are inconsistently reported as the mean value, plus or minus the standard error of the mean, or as the mean value plus or minus the standard deviation. This uncertainty makes it impossible for the reader to evaluate the results of the experiment, and the senior investigator (C. Erlanson-Albertsson) has acknowledged this error and described plans to correct it [1].

In addition, there is an apparent discrepancy between the average weight loss in the control group, calculated from Table 1 of the paper as 3.9 kg, but reported in the abstract as 3.5 kg. The editor of Appetite has acknowledged that this difference is due to the fact that the results are reported by per-protocol analysis, rather than by intention-to-treat analysis, and that the paper does not disclose that fact. Addressing the lack of intention-to-treat analysis, she pointed out that "the study is not registered as a clinical trial in the USA, and it was not funded by the US government, so does not fall under its guidelines." [2] However, according to the Cochrane Reviews, "An intention-to-treat (ITT) analysis is often recommended as the least biased way to estimate intervention effects in randomized trials" [3]

Reference

1: Erlanson-Albertsson C, private correspondence, August 7, 2015

2: Higgs, S, private correspondence, August 10, 2015

3: Cochrane Handbook, Section 8.13.1, Rationale for concern about bias, accessed online on 9/10/2015 at the following website: http://handbook.cochrane.org/chapter_8/8_13_1_rationale_for_concern_about_bias.htm

On 2014 Dec 17, Sean Ekins commented:

link to post mentioning it http://www.collabchem.com/2014/09/15/minding-the-gaps-in-tuberculosis-research-what-science-translational-medicine/

On 2015 May 04, Andrew Brown commented:

See our letter about our concerns about this article here Li P, 2015, the authors' replies here Sichieri R, 2015 and the Editor's comments here Hauner H, 2015.

On 2017 Sep 13, Michael Cole commented:

As first author of this paper, I am grateful to have the chance to evaluate the possibility of a logical error in one of the reported analyses, in order to correct the scientific record and possibly strengthen the conclusions drawn in that paper. Indeed, it appears that the reported negative correlation between resting-state functional connectivity (FC) strength and task-to-rest FC differences could (in theory) be driven by regression to the mean. Critically, having this as a logical possibility does not mean that regression to the mean actually drove the reported result. I therefore conducted a new analyses of publicly-available data to determine whether these results were due to regression to the mean.

I used a slightly different (but overlapping) set of subjects from what was used in the 2014 paper, due to easy availability of these data in my lab and also given that this set of subjects is more standard in the field than the one used in the 2014 paper. Specifically, I used the “100 unrelated” healthy young adult subjects from the WashU-Minn Human Connectome Project dataset (https://www.humanconnectome.org/) . I also switched to using a different set of functionally-defined brain regions – the Glasser et al. 2016 (Nature) region set (360 cortical regions) – rather than the original Power et al. 2011 (Neuron) region set. Additionally, unlike the original study, global signal regression was not used, given concerns about its effect on fMRI signals when comparing resting-state FC to task-state FC (based on simulations conducted in my lab). Finally, a finite impulse response (FIR) general linear model was used to fit the task-evoked activations instead of a canonical hemodynamic response general linear model, given that FIR modeling removes more cross-trial mean variance than canonical hemodynamic response modeling. (As before, the residuals from such a general linear model were used for computing task-state FC estimates). Other than these changes the data are, to my knowledge, identical to those reported in the 2014 paper. As expected, the identified resting-state FC to task-rest FC difference correlation was similar to the original r-value (which was r=-0.49, averaged across the 7 tasks) when using this slightly different dataset: r=-0.63 (averaged across the 7 tasks).

My approach to testing whether regression to the mean could explain the negative correlation between resting-state FC and task-rest FC was to use a split-half analysis, as suggested by Joern in his comment. Specifically, I split the 100 subjects into two sets of 50 each. Then the resting-state FC data from the first half were compared to the task-state FC data in the second half. Thus, correlations were computed for X1 with Y2-X2, such that the two to-be-compared FC matrices were completely independent (since each subject’s data were collected and processed independently).

The resulting split-half cross-validated r-value (averaged across the 7 tasks) was r=-0.60. This is highly similar to the original non-split-half r-value (r=-0.63), despite using only half of the data. Similar r-values were found for each task independently. Specifically (by task): -0.50, -0.69, -0.60, -0.56, -0.62, -0.62, -0.62. This demonstrates that the originally reported effect is likely not driven by regression to the mean.

As an aside, it might be surprising that such a similar result would arise despite using half of the data used in the original analysis. A quick test of repeat reliability, however, indicates that the two halves are highly similar to each other, explaining the strong replication. Specifically, the task-rest FC differences across all 360 regions correlated across the split halves at r=0.86 (mean across 7 tasks). The task-state FC values correlated across the split halves at r=0.97 (mean across 7 tasks). Similarly, the resting-state FC values correlated across the split halves at r=0.98.

Finally, I thought it would be interesting to test whether there was anything special about the resting-state FC values with respect to how anti-correlated they are with task-rest FC differences. I therefore repeated the identical split-half cross-validation analysis, but using each task (one at a time) in place of the resting-state FC data. This indicated that there was a much weaker anti-correlation when using task-state FC in place of resting-state FC: r=-0.10 (mean across the 7 tasks, each separately “standing in” for resting-state FC in the analysis). This was consistent across all 7 tasks, as they each had correlations much less negative (and even positive in the case of the Gambling task) than the resting-state FC data.

Even though these analyses cast strong doubt on the possibility that regression to the mean drove the previously-reported negative correlation, it is still important to take this potential confound into account going forward when conducting analyses similar to this (and more generally).

On 2017 Sep 11, Jörn Diedrichsen commented:

In this otherwise very interesting paper, there is a potential problem with the analysis presented on page 244 and Figure 8b. Here, the authors correlate a connectivity weight estimated during rest (X) with the difference between connectivity weights estimated from task-based data and rest (Y-X). They found a negative correlation. This result could of course be trivial, as both X and Y are measured with noise, and a negative correlation can emerge simply by regression to the mean (https://en.wikipedia.org/wiki/Regression_toward_the_mean). Note that this phenomenon can be substantial – even with a halfway decent re-test reliability of r=0.5 for both X and Y and a large positive correlation between the true X and Y of r=0.9, the expected correlation between X and Y-X will be r=-0.52. One way to test the idea of a relationship would be to split the resting-state data in half and correlate X1 with Y – X2. Assuming that X1 and X2 are now independent (which will depend on the preprocessing and details of the analysis), the resulting correlation coefficient should be unbiased.

On 2015 Oct 06, Siddharudha Shivalli commented:

Stringent and detailed methodology to avoid ambiguity

I read this article with curiosity. Authors’ efforts are commendable. It reiterates malaria as a major public health problem in the study area. However, following issues need to be addressed. Authors have used 7 open ended questions to assess the malaria knowledge among the study participants. The participants’ knowledge level was categorized as good (responded well to all questions), adequate (responded to at least five questions) and poor (could not answer >3 questions). However, ambiguity persists when categorizing the responses to questions 4, 5 and 6. Authors should have mentioned the minimum expected response to each question to consider the participant as knowledgeable. For instance, mentioning at least 3 measures to avoid malaria for question 4 was regarded as knowledgeable. Furthermore, such categorization should be done by 2 people independently and should sought 3rd opinion in case of disagreement in order to avoid inter observer bias. However, authors have not mentioned about who did the categorization and how the bias was tackled. Data regarding correct responses to all the 7 questions would have helped to fix the priority while designing health education material. I would have also considered the ‘knowledge of availability of free diagnosis and treatment in government healthcare setup’ while assessing the knowledge component. In addition, treatment seeking pattern (preferred source and type of treatment) would have been more interesting. Inclusion of 'body complexion, odour, and clothing [Table 1] as study variables need to be justified. Is it related to personal hygiene and/or vector bionomics? If so, their association with occurrence of malaria should have been mentioned. Definition for house type (bamboo, kacha and pucca) should have been explicitly mentioned in the method to avoid ambiguity. Use of existing guidelines to define the economic status would have been more apt. Use of ‘socio-economic determinants’ in the title is debatable as the study was cross sectional and multivariate analysis was not done.

None the less, I must congratulate the authors for investigating an important public health problem in the study area.

Conflict of Interests: The author declares that there is no conflict of interest about this publication.

On 2014 Sep 17, Klaas Vandepoele commented:

An interactive and pre-configured genome browser as well as a bulk download of all CNSs described in this paper are available via http://bioinformatics.psb.ugent.be/cig_data/Ath_CNS/Ath_CNS.php

On 2015 Aug 10, Simon Young commented:

In the article by Tiger et al the levels reported for both serotonin and 5-hydroxyindoleacetic acid (5-HIAA) in human cerebrospinal fluid (CSF) are not consistent with reports in the literature. Concerning serotonin they cite two references, to Anderson et al Anderson GM, 2002 and to Hubbard et al Hubbard KE, 2010. Anderson et al Anderson GM, 2002 report a mean level of serotonin in cisternal CSF of rhesus monkey of 87 pg/ml, which is well below the mean level reported by Tiger et al for human lumbar cerebrospinal fluid. However, the relevance of rhesus monkey cisternal CSF serotonin for human lumbar CSF levels is not clear. The paper by Hubbard et al Hubbard KE, 2010 describes a method for measuring human CSF serotonin. Fig. 2D in that paper shows a chromatogram obtained with a human CSF sample that has a very small peak at the same retention time as the serotonin standard. However, no information is give on the concentration of serotonin detected in CSF. The article which Tiger et al cite by Anderson et al Anderson GM, 2002 mentions that levels of serotonin in human lumbar CSF reported in the literature varied over several orders of magnitude. The paper reports an attempt to determine the true level. Human lumbar CSF from 25 participants was collected on ice and aliquots were frozen at -70C. Measurements were performed in two different laboratories using different HPLC columns and eluting buffers. One lab used an electrochemical detector (detection limit 7 – 8 pg/ml for serotonin) and the other a fluorometric detector (detection limit 7 – 15 pg/ml). In both labs N-methylserotonin was used as an internal standard and a sample was injected directly into the HPLC after removal of proteins. Neither system was able to detect serotonin in any CSF sample. The conclusion was that the real value was less than 10 pg/ml. Anderson et al Anderson GM, 2002 suggest that the higher values reported in the litterature can be attributed to some extent to a failure to carry out rigorous validation steps needed to insure that a peak in HPLC is in fact the analyte of interest and not another compound with a similar retention time and fluorescent or electrochemical properties. Given the rigor of the study of Anderson et al Anderson GM, 2002 the values reported by Tiger et al, which are in the range 400 – 800 pg/ml, must be considered with skepticism. The same is true for the CSF 5-HIAA values reported by Tiger et al which are in a range from less than 1 ng/ml to about 4 ng/ml. Other studies report lumbar CSF 5-HIAA levels in the range 10 – 30 ng/ml, e.g. Kishida I, 2007 Chotai J, 1999 Carpenter LL, 2003.

On 2016 Oct 16, Jaime A. Teixeira da Silva commented:

This paper has been retracted: "BioMed Research International has retracted this article. The article was found to contain images with signs of duplication and manipulation in Figures 1(a), 1(b), 1(d), 2(b), 3(a), 3(b), 3(c), 3(d), 3(k), 4(d), 4(g), 4(m), 4(p), 8, 10(c), 10(d), 10(e), 10(f), 10(g), 10(h), 10(i), 10(j), 10(k), 10(l), and 10(o) and duplication from Talukdar D. An induced glutathione-deficient mutant in grass pea (Lathyrus sativus L.): Modifications in plant morphology, alteration in antioxidant activities and increased sensitivity to cadmium. Biorem. Biodiv Bioavail. 2012; 6: 75–86 in Figure 2B and from Dibyendu Talukdar and Tulika Talukdar, “Superoxide-Dismutase Deficient Mutants in Common Beans (Phaseolus vulgaris L.): Genetic Control, Differential Expressions of Isozymes, and Sensitivity to Arsenic,” BioMed Research International, vol. 2013, Article ID 782450, 11 pages, 2013, doi: 10.1155/2013/782450 in Figure 10."

On 2014 Jul 29, Amanda Capes-Davis commented:

The title refers to endothelial cells. However, ECV-304 is not from normal endothelium. This cell line is known to be cross-contaminated and is actually bladder carcinoma cell line T-24. For a list of known cross-contaminated cell lines, see http://iclac.org/databases/cross-contaminations/.

On 2014 Jul 31, Amanda Capes-Davis commented:

This study focuses on HNSCC. Please be aware that KB cells are not an appropriate model for this disease. The KB cell line is cross-contaminated with HeLa and is actually cervical adenocarcinoma. For a list of known cross-contaminated cell lines, see http://iclac.org/databases/cross-contaminations/.

On 2014 Aug 07, Gabe (RG) Boldt commented:

The validity of this literature search is questionable at best since 'SBRT' and 'liver metastases' are not actual Medical Subject Headings (MeSH). If you're not searching properly you're not going to get accurate results, and that will skew the outcome of your literature search, meaning you may be missing important research in your analysis.

On 2014 Nov 16, Ivan Oransky commented:

This paper has been subjected to an Expression of Concern: http://retractionwatch.com/2014/11/13/lancet-journal-puts-icu-paper-on-watch-after-authors-acknowledge-potentially-fatal-flaw/

On 2015 Feb 24, Janet Kern commented:

In Table 4 of this study, the geometric mean maternal serum mercury concentrations between the general population (0.32) and the ASD group (0.48) had P value of 0.05. The maternal serum blood mercury levels were statistically significantly higher in the ASD group than in the general population. Thus, the higher the maternal serum mercury levels, the more likely the child would be diagnosed with an ASD. Based upon this significant difference, the conclusion of this study should have included this finding.

On 2015 Jan 09, T Eugene Day commented:

This is a very interesting paper, applying DES to a major public health issue on a phenomenally large scale. Modeling individuals for a nation-wide analysis of risk is ambitious and relevant. The authors do an excellent job of explaining the value of simulation over, for example, Markov modeling in this setting, and provide useful and significant implications for policy makers.

I might quibble with their definition of "internal validation", but the model is overall very well validated, and they describe the process usefully and thoroughly. This is an excellent example of leveraging the strengths of simulation to provide both economic and health policy insights in a very complex environment. Well done.

On 2015 Feb 01, Ivan Oransky commented:

The Office of Research Integrity found that many of the figures in this paper had been falsified: http://retractionwatch.com/2015/01/23/former-pitt-cancer-researcher-admits-faking-findings/

On 2014 Nov 24, Guillaume Filion commented:

This article is one of the "CISCOM meta-analyses", which are very similar papers written by different authors. For more information about the CISCOM meta-analyses, check the blog post "A flurry of copycast on PubMed" at the following link http://blog.thegrandlocus.com/2014/10/a-flurry-of-copycats-on-pubmed

On 2014 Jul 25, R. Shane Tubbs commented:

Our review of the available literature identified both anatomic depictions and a smattering of clinical evidence that many cutaneous nerves of the human body do not respect the midline and will in fact, cross it. The implications of this concept and its absence in the anatomic literature are discussed. Such data may be of interest to clinicians and anatomists alike.

On 2014 Aug 16, A Martinez-Arias commented:

There is much that is interesting and significant in this work but the section on ES cells needs to be looked at closely, particularly as the conclusions drawn here (that YAP/TAZ loss of function favours pluripotency) are at odds with others published (that YAP/TAZ are necessary for pluripotency [1]).

Here Azzolin et al. show that loss of function for YAP/TAZ can substitute for Chiron in 2i conditions. As Chiron leads to the upregulation of ß-catenin, the authors conclude that in keeping with their other findings, the effects of YAP/TAZ in pluripotency, are due to ß-catenin. Something supported by the dependence of this effect on the presence of ß-catenin. The experiments are clear but the interpretation should be given some thought.

There are three facts that need to be considered in the interpretation of these results. (a) 2i conditions are special non-physiological conditions which set cells in a peculiar, though experimentally useful, state; in fact, 2i really means “2 out of 3 (PD03, CHIR and LIF)” as LIF+PD03 and LIF+CHIR will work as well as PD03+CHIR [2] in maintaining pluripotency. In fact, ß-catenin mutant cells can be maintained in PD03+LIF, an observation which has been used to state that ß-catenin is not required for pluripotency [2]. (b) There is evidence that the levels of Oct4 are crucial for the maintenance of pluripotency [3-5] and that there is a very close relationship between ß-catenin and Oct4 [6-8]. This might be important in the interpretation of the results here. (c) The function of ß-catenin is pluripotency is not mediated by its transcriptional activity. It is curious that the authors quote Wray et al [2] in support of ß-catenin transcriptional activity in pluripotency when the manuscript shows and states that the transcriptional activity of ß-catenin is dispensable for pluripotency. Something that is further supported by studies in GSK3 mutants that lead to high levels of ß-catenin transcriptional activity which upon introduction of a dominant negative Tcf4, can be shown to be irrelevant for pluripotency [8 see also 2, 6 and 7].

The results on transcriptional activities of YAP/TAZ in ES cells and their overlap with ß-catenin/Tcf are therefore not particularly relevant to the interpretation of the results but the other two might be, as there is a need to understand the discrepancy with previous experiments. It would have been good if the authors had repeated some of their experiments in Serum and LIF or LIF and BMP to have a proper comparison. As it is, we can only consider possibilities. Lin et al. [1] show that YAP/TAZ leads to upregulation of Oct4 and it is known that lowering the levels of ß-catenin leads to increases in Oct4. As low Oct4 favours pluripotency [3, 4], could it be that lowering YAP/TAZ in 2i lowers the levels of Oct4 which then balance the ratios of the different factors and maintains pluripotency? This would provide a partial explanation for the discrepancy. There are two other possibilities worth considering. Could it be that in the same manner that 2i is ‘two out of three’, could be that it is “two out of four”, the fouth being YAP/TAZ? Finally, a key event in pluripotency is the downregulation of Tcf3 whose mechanism remains open [9]. Could it be that YAP/TAZ helps ß-catenin to downregulate Tcf3?

Removal of YAP/TAZ in 2i conditions has no effect on pluripotency, which is not surprising as in 2i condition cells are, for the most part, at the maximum capacity for pluripotency. However, to date the factors involved in 2i have been shown to also maintain pluripotency in Serum or Serum free conditions (at this it is surprising to observe the inability of PD03 to maintain pluripotency, which contradicts multiple previous results). For this reason it would be good to know if the authors have performed knock downs of YAP/TAZ in conditions other than 2i as only then one can consider the portrayed discrepancy.

References

[1] Lian I, Kim J, Okazawa H, Zhao J, Zhao B, Yu J, Chinnaiyan A, Israel MA, Goldstein LS, Abujarour R, Ding S, Guan KL. (2010) The role of YAP transcription coactivator in regulating stem cell self-renewal and differentiation. Genes Dev. 24,1106-18.

[2] Wray, J., Kalkan, T., Gomez-Lopez, S., Eckardt, D., Cook, A., Kemler, R. and Smith, A. (2011) 'Inhibition of glycogen synthase kinase-3 alleviates Tcf3 repression of the pluripotency network and increases embryonic stem cell resistance to differentiation', Nat Cell Biol. 13, 838-45.

[3] Radzisheuskaya A, Le Bin Chia G, dos Santos R, Theunissen TW, Castrro LF, Nichols J, Silva J (2013) Defined Oct4 level governs cell state transitions of pluripotency entry and differentiation into all embryonic lineages. Nature Cell Biology 15, 579-90. [4] Karwacki-Neisius V, Goke J, Osorno R, Halbritter F, Ng JH, Weisse AY, Wong FC, Gagliardi A, Mullin NP, Festuccia N, Colby D, Tomlinson SR, Ng HH, Chambers I (2013) Reduced oct4 expression directs a robust pluripotent state with distinct signaling activity and increased enhancer occupancy by oct4 and nanog. Cell Stem Cell 12: 531-45.

[5] Muñoz Descalzo S, Rué P, Faunes F, Hayward P, Jakt LM, Balayo T, Garcia-Ojalvo J, Martinez Arias A. (2013) A competitive protein interaction network buffers Oct4-mediated differentiation to promote pluripotency in embryonic stem cells. Mol Syst Biol. 9:694.

[6] Faunes F, Hayward P, Muñoz-Descalzo S, Chatterjee SS, Balayo T, Trott J, Christophorou A, Ferrer-Vaquer A, Hadjantonakis AK, DasGupta R, Martinez Arias A (2013) A membrane-associated β-catenin/Oct4 complex correlates with ground-state pluripotency in mouse embryonic stem cells. Development 140: 1171-83.

[7] Livigni A, Peradziryi H, Sharov AA, Chia G, Hammachi F, Migueles RP, Sukparangsi W, Pernagallo S, Bradley M, Nichols J, Ko MS, Brickman JM. (2013) A conserved Oct4/POUV-dependent network links adhesion and migration to progenitor maintenance. Curr Biol. 123:2233-44.

[8] Kelly, K. F., Ng, D. Y., Jayakumaran, G., Wood, G. A., Koide, H. and Doble, B. W. (2011) 'beta-catenin enhances Oct-4 activity and reinforces pluripotency through a TCF-independent mechanism', Cell Stem Cell 8, 214-27.

[9] Yi F, Pereira L, Hoffman JA, Shy BR, Yuen CM, Liu DR, Merrill BJ. (2011) Opposing effects of Tcf3 and Tcf1 control Wnt stimulation of embryonic stem cell self-renewal. Nat Cell Biol. 13, 762-70.

On 2016 Nov 21, Alessandro Rasman commented:

Franz Schelling, M.D. Open letter to Dennis Bourdette & Jeffrey Cohen re the editorial Venous angioplasty for “CCSVI“ in multiple sclerosis. Ending a therapeutic misadventure. Neurology 2014 (July 29); 83:1–2 https://medium.com/@franzschelling/open-letter-to-dennis-bourdette-jeffrey-cohen-re-the-editorial-ae6043a19dac#.elgrw1op3

On 2014 Sep 13, Vahid Rakhshan commented:

Thanks for the authors’ clarifying response. I would like to answer to their response. My answers follow each paragraph of the response of the authors. The authors’ response is in bold font. My response to their response is in normal font.

1. The estimation of power for this study was not simply based on a pilot study, but also based on the only other study carried out using MRI to assess condylar position/displacement, and where displacements of 1 mm were detected. More importantly, is Dr Rakhshan actually advising the readership that mean (absolute) differences of 0.141 mm (ranging from 0.01 to 0.36 mm) in the glenoid fossae are something of a clinical health concern?

Dr Rakhshan is not advising something regarding the clinical concerns. An error magnitude should simply be validated. Considering the results showing that the mean difference is as small as 0.1 mm, detecting an error of 1 mm can lead to serious problems.

The calculated power was reportedly 80% but none of the comparisons was significant. This is what I was talking about. There is a huge difference between the clinical significance with the notion of statistical power calculation which is relevant to "statistical significance", and not to the clinical significance.

Should we have actually adjusted the detection of differences to 0.1 mm? Would this have been clinically relevant?

Of course, you needed to adjust it to about 0.1 mm and then increase the sample size to be able to obtain real powers above 80% and not an inflated power of 80%. Again, this has nothing to do with the notion of clinical significance [or “clinical health concern” as put by the author]. Mentioning “clinical health concern” when talking about test powers is misleading.

2. Repeated-measures ANOVA can be used to assess the variability of techniques. We also agree that with smaller samples a nonparametric test like the Friedman could have also been used, but we used parametric methods.

I think studying the assumptions of the repeated-measures ANOVA would help us in understanding my humble words. How did normality hold in your case of 2 × 3 sample?

In addition, I think we should re-read the other part of my paragraph 2 as well which was about the lack of power of that ANOVA and the unreliability of its non-significant result.

3. In contradiction to the claims made by Dr Rakhshan, in our “Results” section, there is reference to the P values. All P values were lower than 0.05 and, as such, were not statistically significant. The Tukey was the post-hoc test to use if any significant differences were detected; however, there were none.

First, P values smaller than 0.05 mean significant, in contrast to what was stated by the authors.

Secondly, I could not find any contradiction in my critique. The authors had used two different ANOVAs but had not mentioned what ANOVA was used when and with which results.

4. The underlying distribution was deemed normal, and as such, ANOVA was used. Variable results do not necessarily indicate nonnormal underlying distributions.

Thanks for the clarification that distribution was normal.

5. Condylar positions were discussed in relation to CO. Because the Roth power and CR bite registrations were supposedly capable of positioning condyles in the glenoid fossae in relation to CO, we discussed the differences in relation to CO.

I was fine with the information presented. My concern was about the lack of comparisons between CR and Roth power.

Moreover, from the response of the authors, it appears to me that they “supposed” that the other two methods were the gold standards and therefore they only assessed the CO in relation to them. This needed to be clarified in the original article [that these two methods were assumed as gold standards], and be substantiated by proper citations. Nevertheless, these methods are not real gold standards and cannot be supposedly capable of positioning condyles in the desired position. This is implied even from the introduction of the original article, in which it was stated that these methods are not verified by evidence.

6. We do not agree that lines 3 to 8 are unclear; they are in fact quite the opposite. If there was any significant positioning difference that was detectable, we would have found it, but we did not. This was more than safe to infer.

No; the important but forgotten formula is that the absence of evidence does not mean the evidence of absence. This becomes more serious when the power is low. When the sample size is small and the real power is not high enough, there is a great chance that a false negative error happens. This error means that there is some difference, but the data is inadequate to show it.

7. It appears that Dr Rakhshan has sadly and simply fixated his whole critique on his own interpretation of the statistics, claiming that our findings were unreliable or unsubstantiated. We are acutely aware that limitations do exist with any research. Our study simply and openly shows that the differences detected between the registrations were so small and highly variable that using certain bite registrations to accurately and predictably position condyles into specific locations in the glenoid fossae is not evidence based. Our findings are reliable and substantiated.

Your study showed that there was a lack of evidence. Unlike what the authors think, their results did not indicate an evidence for the absence.

Although this result is valuable as a pilot finding, the authors needed to approach their conclusions with a much greater deal of caution. The limitations that the authors are “acutely aware of” disallow them to conclude so strongly. As I had already stated in my original letter to the editor, “The authors needed to state their limitations and warn the reader that their results were inconclusive.” Being aware of the limitations is one thing and clearly disclosing them is another thing.

Any statistically non-significant result in the absence of a high power is considered inconclusive.

I think if we study the notion of statistical significance more deeply, we would agree that my points were not my own interpretation of statistics, but were simply the very basics of statistics.

On 2014 Jul 09, Carl L von Baeyer commented:

For a more extensive review of this topic see Tomlinson D, 2010. A systematic review of faces scales for the self-report of pain intensity in children. Pediatrics. 2010 Nov;126(5):e1168-98. Open access.

On 2014 Sep 07, Jonathan Eisen commented:

Embriette Hyde wrote a blog post about this paper: Diversify Your Microbiome by Rock Climbing Indoors at the microbiology of the built environment network blog that I co-run.

On 2016 Dec 15, M Mangan commented:

This paper is being used by nonsense peddlers to suggest that GMO insulin is harmful to people, and to scare people away from proper treatment for diabetes. Please see this piece for some guidance on this issue: http://kfolta.blogspot.com/2014/06/when-liars-cross-line-gmo-insulin.html

Another science blogger has taken this on as well. http://theness.com/neurologicablog/index.php/abusing-science-you-dont-understand/

On 2014 Sep 26, Swapnil Hiremath commented:

This study was discussed on Sep 23rd 2014 in the open, online, nephrology journal club, #NephJC, on twitter. Introductory comments are available at the PBFluids blog and at the NephJC website. It was a scintillating discussion, with more than 20 participants and over 300 tweets. A transcript and a curated (i.e. Storified) version of the tweetchat are available from the NephJC website. It was also the subject of a write-up on MedPage today. The highlights of the tweetchat were:

· Most clinicians already use renin-angiotensin system blockade for treatment of diabetic nephropathy and this is necessary, but far from sufficient to stop the rising tide of diabetic patients starting dialysis.

· Pentoxifylline does represent an extremely promising and inexpensive agent for these patients, and the investigators should be commended for choosing the appropriate population to study, and following through with a well-designed and conducted trial.

· The next step that most participants would like to see is a large, multi-centre trial which will study clinical endpoints such as need for renal replacement therapy.

Lastly, there was universal, unmistakable and unanimous unhappiness with the current journal style of relegating the Methods after the discussion at the end of the article.

Interested individuals can track and join in the conversation by following @NephJC and #NephJC on twitter, or by visiting the webpage at NephJC.com.

On 2014 Aug 27, Jim Woodgett commented:

Regarding the inhibitor used, SB415286, a quote from reference 29 (Coghlan et al. PMID: 11033082), "SB-216763 and SB-415286 are structurally distinct maleimides that inhibit GSK-3alpha in vitro, with Ki's of 9 nM and 31 nM respectively, in an ATP competitive manner. These compounds inhibited GSK-3beta with similar potency."

As such, the >25 references in this study to specific inhibition of GSK-3beta should be replaced by GSK-3 as both isoforms (alpha and beta, which are ubiquitously expressed) will be similarly inhibited by this compound.

On 2015 Mar 20, DAVID ALLISON commented:

Apparent Statistical and Reporting Errors in “Effect of transcutaneous auricular vagus nerve stimulation on impaired glucose tolerance: a pilot randomized study”

1. Huang et al. compared the efficacy of transcutaneous auricular vagus nerve stimulation (taVNS) and sham taVNS on patients with impaired glucose tolerance (IGT) in a pilot randomized clinical trial. The authors state in the Abstract, specifically in the Results section, that “100 participants completed the study and were included in data analysis." Thereafter, Huang et al. state in the Methods section of the article that “After screening, 72 IGT participants were randomly assigned to either the transcutaneous auricular vagus nerve stimulation group (taVNS) or sham taVNS group (randomization sequence created using SAS System 7.0)”. In the Results section, Huang et al. state that “A total of 102 participants completed the study (72 females, 54.4 ± 7.4 years (mean ± SD), range (33–68).” (4) The number of participants reported by the authors is inconsistent. It is not clear as to whether the number of participants is 100, 72, or 102.

2. Though described as a randomized controlled trial, this is misleading. Patients were not randomized to control. There was non-uniformity of recruitment for study participants in different groups (vagus nerve stimulation subjects were recruited from an acupuncture hospital, while control subjects were recruited from free clinics). Although the authors acknowledge the control group is inappropriate to compare to the other groups as it was recruited from another population, they cite the comparison to the control at the beginning of the discussion as one of the core findings of the study.

3. The authors report on many test statistics with 2 numerator degrees of freedom in their analysis, which appears to be accurate for this data set for a comparison between three groups (df=3-1) or of a profile analysis with two groups and three time points (df=[2-1][3-1]). However, the authors also report tests with 1 numerator degree of freedom for certain outcomes in the exact same context without explanation, which would be incorrect and makes it unclear what tests were performed in these cases. For example, the rmANOVA for a treatment-by-time interaction on 2hPG used two numerator degrees of freedom but the test for the same effect on HbAlc used only one (p.4-5).

4. Tables 2, 4, and 5 report on pairwise differences over time between the three groups. They also contain confidence intervals for the difference between groups that do not correspond to any such difference over time as given in Table 3. For instance, comparing taVNS and sham taVNS for 2hPG the CI is centered around 8.271, but the baseline-to-12-week between-group difference is only 1.1. It is therefore unclear where these values came from. It also is inappropriate to report a single confidence interval for a test with multiple numerator degrees of freedom

Note: This comment has been posted on both PubMed Commons and the Journal’s website.

Lisa H. Antoine, University of Alabama at Birmingham, School of Engineering, Birmingham, Alabama 35294, antoinel@uab.edu

Brandon J. George, Office of Energetics, University of Alabama at Birmingham, Birmingham, Alabama 35294, brgeorge@uab.edu

David B. Allison, Nutrition Obesity Research Center and Department of Biostatistics, University of Alabama at Birmingham, Ryals Public Health Building, Room 140J, Birmingham, Alabama 35294, dallison@uab.edu

On 2017 Feb 04, James Hemp commented:

Ducluzeau et al. erroneously cited our paper 'Adaptation of aerobic respiration to low O2 environments.; PMID: 21844375' multiple times as making claims that aerobic respiration existed in LUCA.

"Aerobic respiration is thus proposed to have been operating in the last universal common ancestor (LUCA) of cellular life in one scenario [1,12,13], whereas others favour the later emergence of O2 respiration from within an anaerobic type of respiration such as the denitrifying pathway [7,11]." - Our paper never mentioned LUCA, nor the presence of aerobic respiration or heme-copper oxygen reductases in LUCA.

"These lines of reasoning lead Brochier-Armanet et al. to propose SoxM (A), the enzyme most widely distributed in extant organisms, as the most ancestral O2R present already in the LUCA, whereas all other groups of O2Rs and NORs would be later evolutionary emanations from this ancestral SoxM (A)-type enzyme. This scenario was subsequently adopted by the Gennis group and is hence used by these colleagues as a basis for interpreting structural and functional results [1]." - We definitely did not adopt the proposal by Brochier-Armanet et al. to interpret our results, nor did we cite their work in the paper. The phylogenetic framework used to interpret our results was from our work "Diversity of the heme-copper superfamily in archaea: insights from genomics and structural modeling.; PMID: 18183358". In it we state "Phylogenetic analyses strongly suggest that the heme–copper oxygen reductases originated within the Bacteria and members were later transferred to the Archaea.", and show a phylogenetic tree of the B-family and relatives evolving from the A-family.

"Consequently, Han et al. [1] consider that LUCA invented the low-affinity SoxM (A) enzyme in a high O2 environment, and the high-affinity SoxB (B)- and cbb3 (C)-type O2Rs would have appeared only when life started to colonize microaerobic habitats. This scenario therefore predicts a LUCA thriving at O2 levels of higher than 1 μM. To the best of our knowledge, no geochemical evidence for such high concentrations of O2 prior to the GOE has been reported so far." - We never said that any oxygen reductases were present in LUCA or before the GOE.

On 2016 Apr 01, Zhiyong Mi commented:

Just downloaded ImageJ from https://imagej.nih.gov/ij/ Image J (bundled with 32-bit Java 1.6.0-24; 44MB )​. Running imageJ on Lenovo 7483 with Windows 7 OS.<br> The glitch is not fixed yet; the black box still shows up. Users should be aware.

On 2016 Mar 29, PAUL KUO commented:

I am the corresponding author and PI for this paper; I am submitting this for the record.

At the time of the retraction, the original data were no longer available and we were unclear as to the reason for the "black box". Subsequently, the following has occurred: 1. The study has been repeated using the original samples and the original findings were replicated, 2. Dr. Guo discovered an ImageJ software artifact or glitch that produced the "black box". The glitch has been submitted to both ImageJ for correction and to an imaging journal for publication. It is displayed in this video: https://www.dropbox.com/s/7sl0qh7woqxljvt/Backspace.wmv?dl=0 3. The original paper and our subsequent discoveries have been reviewed by the Duke Institutional Research Review Board; their findings were that the research record was not misrepresented and there was no evidence for research misconduct. 4. When the journal was informed of these, they declined to retract or correct the original retraction. Hence, the reason for this post.

On 2014 Aug 26, Joyce E B Backus commented:

Readers of this analysis should be aware of the background of retractions in PubMed and, in particular, the peculiarities of the examples cited. Publishers often submit a new record for a retraction so that the original citation and retraction notice citation can be linked together but remain separate. In this case, the publisher chose to make a change to the record for the original paper. The retraction notice Anonymous, 2011 (featured in Table 2), is the record upon which the authors base their finding that the “number of retraction notices became stable 35 months after the retraction” but this record has some unusual circumstances PubMed.

The original paper was published electronically ahead-of-print on 2008 Jul 23; then the paper was retracted (for the first time in 2009), and was never published in print. The publishers substituted the retraction notice online in place of the original paper, but the electronic date of publication remained that of the original paper even though most of the record is related to the retraction notice. The situation is further complicated because the additional retraction notice for the original paper, Lindstrom A, 2009, was published on 2009 Mar 15, and correctly lacks the electronic date of publication of the article. Neither of these retraction notices was published in 2008; Lindstrom A, 2009 was not part of the study and Anonymous, 2011 should also not have been part of the study. Another retraction notice example listed in Table 1 that has a similar publishing practice involves Wolfort RM, 2011 (which incorrectly has the October 24, 2008 electronic publication date of the original article) and Anonymous, 2011. In these cases, the records carry publication dates that do not pertain to the retraction notice itself because the publisher chose to send a replacement record which only partially replaced the record for the original paper.

Also, these few examples can’t be used to draw broad conclusions about the timing of retraction notices, which can be issued at any time, sometimes many years after publication. For example, a 2014 PubMed record Anonymous, 2014 retracts a 2005 article, and a 2013 PubMed record Van Le TS, 2013 retracts a 2004 article.

Retractions are explained in detail in the NLM Fact Sheet, Errata, Retractions, Partial Retractions, Corrected and Republished Articles, Duplicate Publications, Comments, Updates, Patient Summaries, and Republished (Reprinted) Articles Policy for MEDLINE.”

Joyce Backus, Associate Director for Library Operations, National Library of Medicine

On 2014 Jul 07, Ryan Radecki commented:

Post-publication commentary:

"Lives Saved … or Profiteering by Overdiagnosis?"

Following an initial acute ischemic stroke, a search for the cause must be undertaken – for small vessel vasculitis, atherosclerotic emboli, thrombi from the systemic circulation, and so forth, beyond the domain of the Emergency Physician. However, what the Emergency Physician does encounter is the sequelae of this search, in the form of oral anticoagulants....

http://www.emlitofnote.com/2014/07/lives-saved-or-profiteering-by.html

On 2014 Jul 07, Ryan Radecki commented:

Post-publication commentary:

"Lives Saved … or Profiteering by Overdiagnosis?"

Following an initial acute ischemic stroke, a search for the cause must be undertaken – for small vessel vasculitis, atherosclerotic emboli, thrombi from the systemic circulation, and so forth, beyond the domain of the Emergency Physician. However, what the Emergency Physician does encounter is the sequelae of this search, in the form of oral anticoagulants....

http://www.emlitofnote.com/2014/07/lives-saved-or-profiteering-by.html

On 2017 Feb 13, Massimiliano Tognolini commented:

None

On 2014 Dec 13, Massimiliano Tognolini commented:

• CPD1 is a not reliable pharmacological tool as it spontaneously forms a mixture of polymers, with a molecular weight of 40 kDa, which are likely responsible for its biological activity (Baell and Holloway, 2010; Zhu et al., 2013). Conversely pure, freshly synthesized CPD1 is inactive (Noberini et al., 2011) on EphA4 receptor, consequently data reported in Figure 2F, showing that the inhibition of EphA4 by CPD1 prevents the Eph-mediated neurotransmission impairment, should be carefully considered.

• A not-reported concentration of biotinylated rhynchophylline (and not rhynchophylline alone), co-precipitates with the chimeric protein composed by the Fc region of immunoglobulin and the extracellular domain of EphA4. As a signal can be detected also for the control (Figure 4A, lane 1 column 1) the interpretation of this data is unclear.

• Rhynchophylline inhibition of EphA4 phosphorylation in rat neurons was claimed. However, the poor reduction of EphA4 phosphorylation reported in Figures 4B and 4C (20-25% of inhibition), with an unknown concentration of rhynchophylline, is not supported by a cell-viability study nor by an inhibition assay on the isolated kinase domain of EphA4. Without these data, the mechanism of action of rhynchophylline is not proved.

• The indication of concentration is essential (fig 4A, 4B, 4C). The evaluation of Rhynchophylline Ki, Kd or IC50 is crucial.

• Considering the large number of pharmacological activities attributed to Rhynchophylline, i.e. interference with : i) platelet aggregation (Jin et al., 1991), ii) 5HT and dopamine release (Shi et al., 1993), iii) calcium channels (Wang et al., 1994; Shimada et al., 1999), apoptosis (Shi and Kenneth, 2002), iv) NMDA receptors (Kang et al., 2002), v) voltage-gated channel (Chou et al., 2009), vi) NF-kappaB and AP-1(Hsieh et al., 2009), NR2B expression (Zhou et al., 2010), is rather complex demonstrating that this compound can act on a single target.

• In our hands, when tested in our binding study (Giorgio, Plos one, 2011 and Tognolini, ACS chem neurosci, 2014) rhynchophylline does not interfere with Eph-ephrin interaction.

Thank you for your attention.

On 2016 Nov 13, Robin P Clarke commented:

My original comment here turns out to be in error, at least partially. The authors did not make the mistake I suggested there. However, the wording of the text causes confusion. The relevant sentence in the Results is confusing because it begins with "Crude rates of being diagnosed with AD....." but then goes on to include the "fully adjusted" rates. Meanwhile the Results statement in the Abstract refers to the raised risk but without mentioning that is only the "fully adjusted" result.

A problem is that these adjustments are not facts but only theoretical inferences, unlike the crude ratios which are arguably more "factual" facts. Though as it happens I incline to reckon the adjusted ratios are the more correct; my own evidence strongly indicates that the main causal factor in modern autism is dental amalgams, and these would be more prevalent in poorer parents, hence black parents, hence the higher risk for them. (Note that older mothers would tend to have more amalgams and more mercury leaked therefrom, also adding to risk as per the adjustment rationales.)

So not an actual error here, but could have been presented a bit more clearly. Generally it's best to break up sentences into separate units rather than conglomerate them into blockbusters.

Original comment: ~~~~~~ There is clearly an error in either Table 1 or the sentences that refers to it. The results (and abstract) state:

"and 13% to 14% higher in US-born Hispanics and blacks compared with US-born white mothers (Table 1)."

And yet Table 1 shows the opposite, with the per 10,000 rate being lower in US-born Hispanics and blacks. I can only guess that someone accidentally put "higher" when they should have put "lower". This needs a correction mark in the text as otherwise people will assume that Table 1 says what that sentence does.

On 2014 Sep 06, Ryan Radecki commented:

Post-publication commentary:

"Emergency Department or Return-to-Daycare Department?"

Have you ever seen a child brought to the Emergency Department for pinkeye? Not only that, but specifically to fulfill the requirement they receive a prescription for antibiotic ointment in order to return to daycare? You would not be alone, my friends.

In this convenience sample of 303 surveyed parents, 26% reported taking their child to an Urgent Care, and 25% reported taking their child to an Emergency Department, rather than primary care, for minor ailments – including pinkeye.

http://www.emlitofnote.com/2014/08/emergency-department-or-return-to.html

On 2014 Dec 05, Gilles van Wezel commented:

We'd be interested in reports by others on Streptomyces species that are in fact Kitasatosporae.

Gilles

On 2014 Aug 30, Michelle Lin commented:

Congratulations on writing this great review of the benefits, perils, and challenges in integrating social media into a formalized curriculum in graduate medical education. The week-long ALiEM-Annals of EM discussion can be found here, including a videocast (Google Hangout on Air) with the authors and experts.

http://www.aliem.com/social-media-in-the-em-curriculum-annals-em-resident-perspective-article/

On 2014 Jun 29, David Keller commented:

2014 United States Preventative Services Task Force report omitted mandatory analysis by sex

The following is quoted from the NIH Guideline on The Inclusion of Women and Minorities as Subjects in Clinical Research (1):

"In the case of any clinical trial in which women or members of minority groups will be included as subjects, the Director of NIH shall ensure that the trial is designed and carried out in a manner sufficient to provide for valid analysis of whether the variables being studied in the trial affect women or members of minority groups, as the case may be, differently than other subjects in the trial."

While this regulation may appear to apply to women subjects in clinical trials but not to men, implicit protection for men can be derived from the fact that all trial subjects are either men or women, so variables which affect men differently than women must necessarily also affect women differently than men. Thus, any tested variable which affects either sex differently actually affects both sexes differently, and must therefore be examined in a sex-specific analysis.

In their recently-updated report on multivitamin supplements, the United States Preventative Services Task Force (USPSTF) again found "insufficient evidence" to recommend their general use. The report dismissed the significant reduction in cancers seen in men taking multivitamins in both the SU.VI.MAX and the Physicians' Health Study, citing "the lack of effect in women" (2). This failure to analyze and apply the data separately for women and for men violates the NIH mandate to quantify gender differences in clinical trials.

The USPSTF should re-analyze the data separately for women and for men, which would almost certainly yield a "C" recommendation for men to discuss their diet and nutritional status with their doctors and consider taking a multivitamin supplement daily to reduce their risk of cancer, based on the significant findings of benefit in two large well-designed studies.

References

1) NIH website, accessed on 6/29/2014 http://grants.nih.gov/grants/funding/women_min/guidelines_update.htm

2) Moyer VA. Vitamin, Mineral, and Multivitamin Supplements for the Primary Prevention of Cardiovascular Disease and Cancer: U.S. Preventive Services Task Force Recommendation Statement. Ann Intern Med. 2014 Feb 25. doi: 10.7326/M14-0198. [Epub ahead of print] PubMed PMID: 24566474.

On 2015 May 14, Suzanne Gage commented:

Power and colleagues¹ report evidence that individuals with genetic risk of schizophrenia smoke cannabis more, and in greater quantities, than the general population, arguing that these findings are evidence of gene-environment correlation, and suggest that part of the association between schizophrenia and cannabis use is due to shared genetic aetiology. However, an alternate explanation, which is hinted at in the discussion, is that genetic predisposition to schizophrenia may increase the risk of using cannabis. We discuss how causal associations from observational data can be assessed using Mendelian randomisation (MR) designs.

The study by Power and colleagues¹ is in some ways analogous to a MR design, whereby genetic information is used as a proxy for an exposure, in order to ascertain causal relationships from observational data.² In principle, this approach can overcome problems of confounding and reverse causality. While MR has more traditionally been utilised to study the consequence of modifiable risk factors, rather than disease processes, it can be utilised to study the consequences of disease-like states. For example, genetic variants related to hypercholesterolaemia are associated with higher prevalence of the use of cholesterol-lowering drugs (primarily statins).³ The findings of Power and colleagues are consistent with the interpretation that schizophrenia (or its prodrome) might increase the likelihood of cannabis use. The concept of a schizophrenia prodrome, whereby changes in brain chemistry may occur long before symptoms reach a clinical threshold, is increasingly accepted.⁴ Observational studies of cannabis use indicating an association with later psychosis, psychotic symptoms, or schizophrenia are not inconsistent with the self-medication hypothesis. This could be directly tested, using indicators that predict later transition to schizophrenia. For example, psychotic-like experiences might be indicative of prodromal status, and would therefore be expected to be associated with a genetic risk score for schizophrenia, if this predicts later transition to clinical psychosis. In this case, a MR design using a schizophrenia polygenic risk score as the exposure would provide some insight into whether the association between schizophrenia and cannabis use is due to self-medication.

The recent genome-wide association study (GWAS) by the Psychiatric Genomics Consortium (PGC) includes a polygenic risk score for schizophrenia risk which captures 18% of phenotypic variance,⁵ compared with 7% for the score used by Power and colleagues derived from an earlier study by the PGC.⁶ This is likely to be a suitable instrument for MR-style analyses, allowing further insight in to the relationship between cannabis use and schizophrenia through the use of bidirectional MR.⁷ However, an MR design using a schizophrenia risk score would also predict cannabis use if the association were due to genetic covariance (as Power and colleagues argue), rather than a causal effect on cannabis use. Fractionating the polygenic score could help here, as the hypothesis that a schizophrenia prodrome increases cannabis use would suggest that any combination of variants related to schizophrenia should relate to cannabis use.⁷ Genetic covariance would not be expected to exist for the whole gamut of the schizophrenia-related genetic variants. Furthermore, a more complex design combining risk scores and longitudinal data would be required to disentangle these possibilities and provide better evidence of the direction of causation.

Finally, it is well established that cannabis use is also heritable,⁸ and although no specific loci have yet been robustly identified, as larger GWAS of cannabis phenotypes are undertaken it is likely that they will be. MR studies of cannabis use as an exposure will then become possible, allowing the investigation of the causal association between cannabis use and schizophrenia (i.e., that schizophrenia is a consequence of cannabis use). Such studies have already been used to investigate causal associations of tobacco on various outcomes.^9,10 MR methods come with caveats, such as the need for large sample sizes, and may be invalid if certain assumptions are violated (e.g., pleiotropy, canalization, linkage disequilibrium and population stratification).^2,7 Nevertheless, they offer an opportunity to better understand the complex question of the nature of the relationship between cannabis use and schizophrenia.

Suzanne H. Gage, George Davey Smith and Marcus R. Munafò

References

1.Power RA, Verweij KJ, Zuhair M, Montgomery GW, Henders AK, Heath AC et al. Genetic predisposition to schizophrenia associated with increased use of cannabis. Molecular psychiatry 2014.

2.Davey Smith G, Ebrahim S. Mendelian randomization: prospects, potentials, and limitations. International journal of epidemiology 2004; 33(1): 30-42.

3.Benn M, Tybjaerg-Hansen A, Stender S, Frikke-Schmidt R, Nordestgaard BG. Low-density lipoprotein cholesterol and the risk of cancer: a mendelian randomization study. Journal of the National Cancer Institute 2011; 103(6): 508-519.

4.Dominguez MD, Wichers M, Lieb R, Wittchen HU, van Os J. Evidence that onset of clinical psychosis is an outcome of progressively more persistent subclinical psychotic experiences: an 8-year cohort study. Schizophrenia bulletin 2011; 37(1): 84-93.

5.Schizophrenia-Working-Group-of-the-Psychiatric-Genomics-Consortium. Biological insights from 108 schizophrenia-associated genetic loci. Nature 2014; 511(7510): 421-427.

6.Ripke S, O'Dushlaine C, Chambert K, Moran JL, Kahler AK, Akterin S et al. Genome-wide association analysis identifies 13 new risk loci for schizophrenia. Nature genetics 2013; 45(10): 1150-1159.

7.Davey Smith G, Hemani G. Mendelian randomization: genetic anchors for causal inference in epidemiological studies. Hum Mol Genet in press.

8.Agrawal A, Lynskey MT. The genetic epidemiology of cannabis use, abuse and dependence. Addiction 2006; 101(6): 801-812.

9.Asvold BO, Bjorngaard JH, Carslake D, Gabrielsen ME, Skorpen F, Davey Smith G et al. Causal associations of tobacco smoking with cardiovascular risk factors: a Mendelian randomization analysis of the HUNT Study in Norway. International journal of epidemiology 2014.

10.Freathy RM, Kazeem GR, Morris RW, Johnson PC, Paternoster L, Ebrahim S et al. Genetic variation at CHRNA5-CHRNA3-CHRNB4 interacts with smoking status to influence body mass index. International journal of epidemiology 2011; 40(6): 1617-1628.

On 2014 Aug 16, Miguel Lopez-Lazaro commented:

Improving Selectivity versus Improving Cytotoxicity

This article reports that a 2'-modified-4'-selenoarabinofuranosyl-pyrimidine improves the cytotoxic potency of the standard anticancer drug cytarabine (ara-C) in several human cancer cell lines. Like most researchers involved in anticancer drug discovery, the authors evaluate the in vitro anticancer potential of their compounds addressing the question: Can my compounds improve the cytotoxicity of the standard anticancer drugs?

We typically assess the in vitro therapeutic potential of our compounds following a cytotoxic potency-based approach: the lower the concentration required to kill cancer cells, the higher the potential for cancer therapy (I have followed this approach many times in the past; e.g., http://www.ncbi.nlm.nih.gov/pubmed/23174318). Despite its widespread use, this approach is illogical and unreliable.

Cancer patients do not need drugs that kill their cancer cells at low concentrations. They need drugs that kill their cancer cells at concentrations that do not significantly affect their normal cells; otherwise, the maximum doses tolerated by the patients will be insufficient to reach the drug concentrations required to eliminate the cancer cells of their bodies. If a drug kills cancer cells without affecting nonmalignant cells, it does not really matter at what concentrations it kills the cancer cells.

It is also important to note that the ability of a compound to kill cancer cells at low concentrations does not reliably predict its ability to kill cancer cells selectively. Therefore, when we show that our compounds kill cancer cells at low concentrations, we are not revealing whether or not they have potential for cancer therapy.

To properly reveal the in vitro therapeutic potential of our compounds, we only need cancer cells, nonmalignant cells, standard anticancer drugs, a simple cytotoxicity test, and an answer to the question: Can my compounds improve the ability of the standard drugs to kill cancer cells without significantly affecting nonmalignant cells from appropriate tissues? (1).

A basic protocol for addressing this question can be found in reference (1). I can also e-mail a copy of the screening protocol included in this reference to anyone who requests it at mlopezlazaro@us.es

The authors should consider using this selectivity-based approach to assess if their compounds improve the selectivity of cytarabine; this would reveal their cancer therapeutic potential. Meanwhile, readers should interpret the findings of this article with caution.

(1) Lopez-Lazaro, M. Experimental Cancer Pharmacology for Researchers: At What Concentration Should my Drug Kill Cancer Cells so that it has Potential for Cancer Therapy? 2014, ASIN: B00MMO25NM http://www.amazon.com/dp/B00MMO25NM/

On 2016 Dec 19, induprabha yadev commented:

this article helps in surgeons on the learning curve to follow a standardised approach in lap hernia repair.

On 2014 Jul 08, Siyu Sun commented:

It is really interesting to perform EUS for 3-year-old child.

On 2014 Jun 27, Lauren F Laker commented:

Thanks for feedback and interesting questions. In response to your comments regarding validation, Figure 3 illustrates the arrivals for the real-world system which was approximated in our simulation model. It was replicated by the simulation within 0.5% of the empirical data. Unfortunately, the decision to release actual operational data would require the approval of senior hospital administration, so we can't do that at this time. While we could have discussed validation more extensively, the manuscript length was a concern and the primary goal of the paper was to compare flexibility policies to each other rather than to a real-world benchmark. As for implementation, these results influenced part of a redesign of the UCMC ED's patient intake area, which has been modified to explicitly include some flexible beds that can be shared between the fast-track area and the regular ED. That project is ongoing and construction is expected to begin soon. Thanks again for your interest in the paper.

On 2014 Jun 24, T Eugene Day commented:

This is a really cool model, and the concept of a flex-bed in the fast track area is intriguing. Using simulation to determine the preferred number of bed under varying assumptions is an ideal use of the technology.

Triage time was estimated from conversations with providers. This is sub-ideal (but common practice - preferable would be real-world observations), but they found that the model was robust to changes in triage time, justifying the method. My main concern is that though the authors assert that the model was validated against real world data, that data isn't presented or described. No model perfectly tracks the real world system, and having that data available would provide boundaries on how well we'd expect the model results to translate back to the real world. Similarly, it is not always clear (for example for figure 3) when they are discussing simulated or real-world data.

With the caveat that I'd like to see the baseline real-world to simulated results presented clearly, this seems like a strong paper with a useful result. Up next: are there plans to test this solution in the real world? And how well do such tests compare with predicted results? How much has the demography of the patient population changed from the AY2011 data set to the current population? etc.

On 2016 Aug 24, Ben Goldacre commented:

This trial has the wrong trial registry ID associated with it on PubMed: both in the XML on PubMed, and in the originating journal article. The ID given is NCT0168577. We believe the correct ID, which we have found by hand searching, is NCT01685775.

This comment is being posted as part of the OpenTrials.net project^[1] , an open database threading together all publicly accessible documents and data on each trial, globally. In the course of creating the database, and matching documents and data sources about trials from different locations, we have identified various anomalies in datasets such as PubMed, and in published papers. Alongside documenting the prevalence of problems, we are also attempting to correct these errors and anomalies wherever possible, by feeding back to the originators. We have corrected this data in the OpenTrials.net database; we hope that this trial’s text and metadata can also be corrected at source, in PubMed and in the accompanying paper.

Many thanks,

Jessica Fleminger, Ben Goldacre*

[1] Goldacre, B., Gray, J., 2016. OpenTrials: towards a collaborative open database of all available information on all clinical trials. Trials 17. doi:10.1186/s13063-016-1290-8 PMID: 27056367

* Dr Ben Goldacre BA MA MSc MBBS MRCPsych<br> Senior Clinical Research Fellow<br> ben.goldacre@phc.ox.ac.uk<br> www.ebmDataLab.net<br> Centre for Evidence Based Medicine<br> Department of Primary Care Health Sciences<br> University of Oxford<br> Radcliffe Observatory Quarter<br> Woodstock Road<br> Oxford OX2 6GG

On 2014 Dec 31, William Grant commented:

The paper by Huss and colleagues studied the relation between 25-hydroxyvirtamin D [25(OH)D] concentrations and breast cancer mortality rates for women enrolled in the Malmö Diet and Cancer Study observed between 1991 and 2006 [1]. 17,035 women completed the baseline examination, and from these, 672 women were diagnosed with invasive unilateral breast cancer after baseline, of whom 101 died from breast cancer and 54 died from other causes. The mean time between enrollment and breast cancer was 6.0 to 8.3 years with standard deviations of 7.4 to 8.9 years. They found a U-shaped relation between 25(OH)D concentration at baseline and mortality rate for three tertiles with ranges <75 nmol/L 76-99 nmol/L, and >100 nmol/L. Each tertile had 216 to 221 invasive breast cancer cases and 21 to 30 breast cancer specific deaths. While no explanation was given why high 25(OH)D concentrations should be associated with increased risk. In this comment, I present three reasons why the increased mortality rate at the highest 25(OH)D tertile is probably not representative of natural 25(OH)D concentrations, i.e., derived from solar UVB or diet.

First, breast cancer develops very rapidly, so baseline 25(OH)D concentrations are not useful for long-term studies. For example, there is a global seasonality of breast cancer incidence rates, with peaks in spring and fall [2]. The authors of this paper suggested that UVB and vitamin D reduce incidence rates in summer while dark days and melatonin reduce incidence rates in winter. In addition, breast cancer screening using mammography is recommended annually, unlike colorectal cancer, for which decadal screening is generally used. There are large seasonal variations in 25(OH)D concentrations in Europe [3]. In addition, there are significant drifts in 25(OH)D concentrations over periods of years such that for periods longer than about three years, observational studies generally do not find significant inverse correlations between baseline 25(OH)D concentration and incidence of breast cancer, although they do for colorectal cancer [4]. Long follow-up times also adversely affect observational studies of all-cause mortality rate with respect to baseline 25(OH)D concentration [5].

Second, 25(OH)D concentrations above 100 nmol/L are most likely due to vitamin D supplementation. In addition, the supplementation may have begun late in life, perhaps after a physician noticed a vitamin D deficiency disease such as osteoporosis, or in order to prevent osteoporosis. Support for this hypothesis is found in a pair of observational studies on frailty as a function of 25(OH)D concentrations. In a cross-sectional study of elderly men in the United States, there was a monotonic decrease in frailty status with increasing 25(OH)D concentration [6]. However, in a similar study of elderly women, there was a U-shaped relation, with those having 25(OH)D concentration >75 nmol/L having significantly increased prevalence of frailty than those with 50 nmol/L <25(OH)D <75 nmol/L [7]. In the United States, postmenopausal women are often advised to take vitamin D supplements while men are not.

Third, there is no support for a U-shaped breast cancer mortality rate relation from geographical ecological studies with respect to solar UVB doses. In Ref. 8, there is a graph of breast cancer mortality rate with respect to July solar UVB doses in over 400 state economic areas of the United States. The regression line shows a monotonically decreasing mortality rate with respect so solar UVB doses from 3.5 kJ/m2 to 10 kJ/m2 with no evidence in the data of an upturn at higher UVB doses. Other ecological studies from mid- to high-latitude countries also do not show evidence of a U-shaped relation [9].

Thus, for these three reasons the finding of increased breast cancer specific mortality rate for 25(OH)D concentrations >100 nmol/L is most likely not due to long term natural 25(OH)D concentrations but, rather, vitamin D supplementation late in life. This does not imply that taking vitamin D supplements to raise 25(OH)D concentrations above 100 nmol/L is not recommended. However, it suggests that taking supplements should begin earlier in life since many chronic diseases including cancer can develop over periods of decades. That high concentrations are not adverse per se is demonstrated in a recent meta-analysis of 32 prospective observational studies of all-cause mortality rate with respect to baseline 25(OH)D concentration. There was a nearly linear decrease in the hazard ratio from 0-25 nmol/L to 50-73 nmol/L with a minimum reached at 90 nmol/L, after which there was no change [10].

References 1. Huss L, Butt S, Borgquist S, Almquist M, Malm J, Manjer J. Serum levels of vitamin D, parathyroid hormone and calcium in relation to survival following breast cancer. Cancer Causes Control. 2014;25:1131-1140. 2. Oh EY, Ansell C, Nawaz H, Yang CH, Wood PA, Hrushesky WJ. Global breast cancer seasonality. Breast Cancer Res Treat. 2010;123:233-243. 3. Hyppönen E, Power C. Hypovitaminosis D in British adults at age 45 y: nationwide cohort study of dietary and lifestyle predictors. Am J Clin Nutr. 2007;85:860-868. 4. Grant WB. Effect of interval between serum draw and follow-up period on relative risk of cancer incidence with respect to 25-hydroxyvitamin D level; implications for meta-analyses and setting vitamin D guidelines. Dermatoendocrinol. 2011;3:199-204. 5. Grant WB. Effect of follow-up time on the relation between prediagnostic serum 25-hydroxyitamin D and all-cause mortality rate. Dermatoendocrinol. 2012;4:198-202. 6. Ensrud KE, Blackwell TL, Cauley JA, Cummings SR, Barrett-Connor E, Dam TT, Hoffman AR, Shikany JM, Lane NE, Stefanick ML, Orwoll ES, Cawthon PM; Osteoporotic Fractures in Men Study Group. Circulating 25-hydroxyvitamin D levels and frailty in older men: the osteoporotic fractures in men study. J Am Geriatr Soc. 2011;59:101-106. 7. Ensrud KE, Ewing SK, Fredman L, Hochberg MC, Cauley JA, Hillier TA, Cummings SR, Yaffe K, Cawthon PM; Study of Osteoporotic Fractures Research Group. Circulating 25-hydroxyvitamin D levels and frailty status in older women. J Clin Endocrinol Metab. 2010;95:5266-5273. 8. Grant WB. An estimate of premature cancer mortality in the U.S. due to inadequate doses of solar ultraviolet-B radiation. Cancer. 2002;94:1867-75. 9. Moukayed M, Grant WB. Molecular link between vitamin D and cancer prevention. Nutrients. 2013;5:3993-4023. 10. Garland CF, Kim JJ, Mohr SB, Gorham ED, Grant WB, Giovannucci EL, Baggerly L, Hofflich H, Ramsdell J, Zeng K, Heaney RP. Meta-analysis of all-cause mortality according to serum 25-hydroxyvitamin D. Am J Pub Health. 2014;104:e43-50.

Disclosure I receive funding from Bio-Tech Pharmacal (Fayetteville, AR) and Medi-Sun Engineering, LLC (Highland Park, IL).

On 2014 Nov 24, Guillaume Filion commented:

This article is one of the "CISCOM meta-analyses", which are very similar papers written by different authors. For more information about the CISCOM meta-analyses, check the blog post "A flurry of copycast on PubMed" at the following link http://blog.thegrandlocus.com/2014/10/a-flurry-of-copycats-on-pubmed

On 2014 Aug 13, Janet Kern commented:

The following studies that showed an association between Thimerosal and the risk of autism were not included in the Yoshimasu et al. (2014) meta-analysis even though they were published within the same time frame as the studies that were included.

Gallagher CM, Goodman MS. Hepatitis B vaccination of male neonates and autism diagnosis, NHIS 1997-2002. J Toxicol Environ Health A 2010;73(24):1665-1677.

Geier DA, Hooker BS, Kern JK, King PG, Sykes LK, Geier MR. A two-phase study evaluating the relationship between Thimerosal-containing vaccine administration and the risk for an autism spectrum disorder in the United States. Transl Neurodegener 2013;2(1):25.

Young HA, Geier DA, Geier MR. Thimerosal exposure in infants and neurodevelopmental disorders: an assessment of computerized medical records in the Vaccine Safety Datalink. J Neurol Sci 2008;271(1-2):110-118.

On 2014 Jul 24, Ryan Radecki commented:

Post-publication commentary:

"How Electronic Health Records Sabotage Care"

Our new information overlords bring many benefits to patient care. No, really, they do. I’m sure you can come up with one or two aspects of patient safety improved by modern health information technology. However, it’s been difficult to demonstrate benefits associated with electronic health records in terms of patient-oriented outcomes because, as we are all well aware, many EHRs inadvertently detract from efficient processes of care....

http://www.emlitofnote.com/2014/07/how-electronic-health-records-sabotage.html

On 2015 Jul 01, Bill Cayley commented:

A good example of when Less is More: https://lessismoreebm.wordpress.com/2015/07/01/a-comparison-of-various-contemporary-methods-to-prevent-a-wet-cast/

On 2014 Jun 29, Ivan Oransky commented:

This paper has been retracted, one of three by the same group: http://retractionwatch.com/2014/06/23/cancer-genetics-group-retracts-three-papers-for-inappropriate-presentation-of-data/

On 2014 Aug 30, Michelle Lin commented:

This summary publication of the discussion (held at http://www.aliem.com/journal-club-clinical-decision-rule-subarachnoid-hemorrhage/), was novel because of the objective analytics that could be tracked over time.

On 2015 Mar 12, James Tsung commented:

Supplemental video curvilinear vs. linear probe: https://youtu.be/8xZE8_QeKwc