Retraction Watch

Tracking retractions as a window into the scientific process

Controversial paper on life-extending buckyballs corrected after blog readers note problems

with 16 comments

Back in April, a group of French and Tunisian researchers published a paper in Biomaterials which came to the astonishing conclusion that buckyballs (carbon tetrachloride) coated in olive oil could dramatically extend the lives of lab rodents. That news was picked up by Derek Lowe’s In the Pipeline blog, on which he expressed some bemusement about the work but ultimately praised it:

These are reasonable (but unproven) hypotheses, and I very much look forward to seeing this work followed up to see some more light shed on them. The whole life-extension result needs to be confirmed as well, and in other species. I congratulate the authors of this work, though, for giving me the most number of raised eyebrows I’ve had while reading a scientific paper in quite some time.

One of those eyebrows dropped a bit the following day, when Lowe reported that readers had pointed him to a clear case of image duplication in the article. At the time, Lowe concluded:

This is, at the very least, very sloppy work, on both the part of the authors and the editorial staff at Biomaterials. I didn’t catch this one myself, true – but I wasn’t asked to review the paper, either, and I can assure you that I spend more time critically studying the figures in a paper under review than one I’m writing a quick blog entry about. Under normal reading conditions, most of us don’t look at histology slides in a paper while constantly asking ourselves “Is this right? Or is this just a duplicate of another image that’s supposed to be something else?”

And while this image duplication does not directly bear on the most surprising and interesting results of the paper – life extension in rodents – it does not inspire confidence in those results, either. I’m emailing the editorial staff at Biomaterials and the corresponding author of the paper with this blog entry. We’ll see what happens.

Well, now we know. The journal has issued a correction for the article, which addresses not just the histology but another image, as well.

Subsequent to publication of this paper online the authors have realized that there were errors in two figures, Figures 3 and 4. The correct figures are given below. The authors apologize to readers for these errors.

Here are the images:

That’s followed by an editor’s note, which states:

It should be noted that one of these errors, referring to the inadvertent duplication of the same image within two panels of Fig 4, was pointed out to the Editor-in-Chief by several readers. The authors contacted the Editor-in-Chief with an explanation of this error and an error in Figure 3 before he requested an explanation from the authors. This paper draws conclusions that appear counter-intuitive. The Editor-in-Chief received two very detailed reports from referees who indicated that the methodology appeared sound and they both recommended acceptance after some revision. Neither referee nor the Editor-in-Chief noticed either error, and the revised paper was published. Due consideration has been given to the potential effect of these errors on the overall results and conclusions drawn, and so it has been decided the conclusions are still valid. The authors have provided explanations of how the errors were made during the preparation of graphics and images.

Hat tips: Marco de Weert, Tim Smith

Written by amarcus41

June 28th, 2012 at 11:34 am

  • jaspevacek June 28, 2012 at 11:45 am

    Buckyballs are the C60 and C70 allotropes of carbon. Carbon tetrachloride is CCl4, a common lab solvent.

    • mass_speccer June 28, 2012 at 11:52 am

      Yeah, I think it would be really really surprising if carbon tetrachloride extended life. My old department didn’t even want us to order it for health and safety reasons.

    • ivanoransky June 28, 2012 at 12:02 pm

      Fixed — thanks.

  • Neuroskeptic June 28, 2012 at 12:03 pm

    Another case of these anonymous bloggers and commenters obstructing science with their pesky ‘logic’.

    • Clare Francis June 29, 2012 at 5:51 am

      Last night Deutschlandfunk had a programme about “Schwarmintelligenz”.
      I think it is simply better communication, internet etc, and that diverse people are seeing the same things/behaviour. The explanations that stnad the test of scrutiny are the ones that fit the facts with the least amount of fudging. It has always been like that, except that now everything goes faster.
      Mundane explanations do not make for radio programmes though.

      You can make the academic out of anything. So much for the social sciences

  • Toby White June 28, 2012 at 12:18 pm

    Revised figure 3(a), with vertical lines, looks worse than the original. For example, at month 24 (or maybe 25), control rats are shown to have both 83% and 67% survival. Are these Schroedinger rats or am I missing something?

    • Marco June 28, 2012 at 12:32 pm

      It’s common to show survival curves like this.

      • Neuroskeptic June 28, 2012 at 12:34 pm

        Yeah it just means that a rat died during month 24. At month 24 survival started at 83% and at the end it was 67%.

    • Klaus Metzeler June 28, 2012 at 12:40 pm

      Are you familiar with Kaplan Meier plots? Nothing wrong with the vertical lines. They just indicate that, at that time point, one rat died and so the proportion surviving went from ~83% down to ~67%. What is a bit strange in Figure a) are the data points plotted on the line each month. Usually, such “tick marks” indicate a censoring event, but here they seem to be used differently…

      • Toby White June 28, 2012 at 1:04 pm

        I wasn’t familiar with them, but I looked it up. My thanks to you, Marco & Neuroskeptic for the clarification.

  • vhedwig June 28, 2012 at 1:28 pm

    I’ll merely re-post what I said originally at In The Pipeline (posting under the name “Virgil”)…

    I find it VERY odd that they started dosing at 10 months (just a few months before the control rats started dying), and kept the therapy up for only 7 months, but the oldest lived rats went on for over 5 years. Really? In humans this would translate to taking C60 for most of your 30s, none before, none after, and living to 120. Something’s not right.

    My other suspicion comes from the timing… the oldest lived animals were 66 months (5.5 years), and the paper was submitted in Jan 2012. Thus, the studies were begun on or before July 2006. Many of the papers they cite regarding the rationale for the dosing regimen were published AFTER the study had already begun. The paper providing the rationale for stopping at 7 months (due to accumulation of c60 in the liver) was not published until 2010, more than 3 years after they’d already stopped dosing! You can’t cite something as rationale, when it wasn’t even published back when you were designing the study (and actually, not published until well after you’d actually stopped the treatment).

    I don’t see that the correction addresses this fundamental error in the citations regarding rationale. The methods section or the citations need to be corrected also.

    • chirality June 28, 2012 at 1:56 pm

      I remember your post from In The Pipeline and I too smelled a rat (it was probably the one treated with C60).

    • Molendinarius July 8, 2012 at 5:03 pm

      The rationale for stopping at 7 months, as quoted in the paper, is because of the poor ability of rats to handle oral gavage, as they are sensitive to it. They didn’t want this to become a confounding factor, so after the first rat died, the prodecude was stopped.

      10 month old rats were used, as is standard protocol, and the reason for this is given explicitly.

      There is no problem with accumlating citations over the course of a study that runs for such a long time, as these are used in the discussion. Post hoc propter hoc does not apply here.

      The 2010 study on bio-accumulation is not the earliest, it is the latest….. but there are plenty of other generic studies that show problems with administering a high fat diet to rats long term, and this would have been a confounding factor. The 2010 study is the most recent, and reviews all the previous literature in its bibliography, so this body of literature is not directly quoted in the references. The problems of obesity, excessive steatosis, liver lipid degeneration, and insulin resistance for administering a high-oil diet to rats was well established at the time the study started. The rationale was NOT bio-accumulation of fullerene in the liver, as you incorrectly state.

  • littlegreyrabbit June 28, 2012 at 2:39 pm

    It is really life-extending olive oil, presumably bucky-balls act in some kind of slow release mechanism or sequestering action that extends the effect of the olive oil a bit further.

    Presumably the bucky-balls themselves are inert and that is why there is no bucky-ball alone control. Although others have pointed out, a rat’s life is nasty, brutish and rather short and accummulation in the liver appears to have happened after 7 months.

    I believe these results could be true – doesn’t mean they are though.

    Is the rat-life enhancing properties of olive oil well established?

    • Fernando Pessoa July 1, 2012 at 4:28 am

      In reply to littlegreyrabbit June 28, 2012 at 2:39 pm

      “I believe these results could be true”.

      We all live in hope.

  • foolishtrends July 9, 2012 at 9:51 pm

    it is VERY clear that the graph is either still quite wrong or the data is. The new graph does not show an average lifespan of 28 months for Olive Oil rats. it shows a MINIMUM lifespan in excess of 30 months with a average close to 40 months. Quite Frankly, to call such a double publication of blatantly false graphs anything less than extremely sloppy would be insulting to good researchers everywhere.

  • Post a comment

    Threaded commenting powered by interconnect/it code.