About the SA Blog Network

The Scicurious Brain

The Scicurious Brain

The Good, Bad, and Weird in Physiology and Neuroscience
The Scicurious Brain Home

The IgNobels, 2013! Playing opera to your mice.

The views expressed are those of the author and are not necessarily those of Scientific American.

Email   PrintPrint

I have to say that the winners of this IgNobel won in my heart PURELY because of the costumes they wore to the ceremony. Yes. Costumes.

(The best image I could find, from the South China Post coverage of the story, which you can read here, photo originally via Reuters)

Because if you’re going to win an IgNobel, you might as well win it dressed as a giant mouse and holding a plush heart. With CLASS. Because that’s what the IgNobels are all about.

So they definitely won points for style. But while the science is indeed funny…I don’t know that it’s going to cause people to put opera in operating rooms soon. Let’s go through it, and I’ll tell you why.

Uchiyama et al. “Auditory stimulation of opera music induced prolongation of murine cardiac allograft survival and maintained generation of regulatory CD4+CD25+ cells” Journal of Cardiothoracic Surgery, 2012.*

The authors of this study were investigating the healing powers of music. Music does lots of good things for humans: relaxation, comfort, etc. There are many studies that show positive effects of music in humans on a psychological level. But can it affect physical healing?

And can it do it in mice? We all know, after all, that mice are the most severe of art critics.

So the scientists wanted to look at music in mice receiving…heart transplants. Some mice got no sound. Some got opera before the transplant.

Some got Mozart for 6 days after the transplant.

Some got Enya for 6 days after the transplant.

Other mice just got tones at different frequencies for 6 days.

(Figure 1)

Above you can see the results. The music or sound choice is in columns on the far left, how each heart did (the survival rate) of each individual mouse is second from right, and the average is on the right.

And it looks like Mozart wins hands down, with a huge increase in survival time. As for Enya…ouch. I mean, it’s not bad, but reduced heart transplant survival time compared to Mozart has to be a rough artistic criticism.

They did the experiment again, this time just with opera music. The opera-listening mice showed increased survival time. When they burst the ear drums of the mice (so they could not hear the music), the effect apparently went away. They also saw changes in the immune response, with mice receiving Opera having a changed immune response, showing evidence of anti-inflammatory effects.

(Figure 3)

Above you can see measures of cell proliferation (on the left), higher in mice getting transplants. On the right are measures of cytokines. If you look at IL-10, on the top right, that’s a cytokine thought to have anti-inflammatory effects. It’s much higher in the opera listening mice.

The authors conclude that classical music in mice (and particularly opera) as an auditory stimulus may affect the immune system.

Perhaps it may. I have to say, though, that I’m not sure that they’ll be able to replicate this effect. The groups of mice they used were VERY small (n=4-5 per group), and though the Mozart group had longer survival time, two of the mice had VERY high survival times indeed (80-ish days as opposed to an average of 7-11 in controls, and as opposed to 20 in the rest of the opera group), which might skew the data and be outliers in a larger sample. There’s also an additional issue with the statistics to analyze the data. They used T-tests to analyze the cytokine and flow cytometry data…even though there were three groups in each set. I personally think they should have used the one-way ANOVA for that test, which takes into account the variance within the samples. If they had used the one-way ANOVA, I’m not sure they would have gotten such significant results (though I am not a stats specialist, and I would welcome other opinions on that). Finally, they use very different groups for the cytokine analysis than they do for other types. Where’s Enya in the cytokine data?! There it’s just opera, no music, or no operation. I think you need a different kind of music or sound there (probably both) as another control group.

So when it comes down to it, I’m not particularly convinced that opera can save mouse hearts. They might want to do a replication, both for science and because, if they don’t, and I were Enya, I’d sue to defamation of character. There’s no art critic quite like a mouse rejecting a heart transplant.

*It should be considered that, while very impressively named, the Journal of Cardiothoracic Surgery has an impact factor of 0.90.

Scicurious About the Author: Scicurious is a PhD in Physiology, and is currently a postdoc in biomedical research. She loves the brain. And so should you. Follow on Twitter @Scicurious.

The views expressed are those of the author and are not necessarily those of Scientific American.

Rights & Permissions

Comments 5 Comments

Add Comment
  1. 1. dpryan 4:57 pm 09/19/2013

    Your critique of their statistics is…weird. A T-test can easily handle differences in within-group variances (this is common enough that even the T-test function in Excel can do it). Remember that a T-test is just a special case of an ANOVA. BTW, the post-hoc tests that you would do after a one-way ANOVA to get pair-wise p-values are often modified T-tests!

    A better criticism would be based on the fact that their survival data is count-based, so a T-test (or ANOVA) is simply wrong to use. They should have done standard survival analysis to get their results. Given the miniscule group sizes, I wouldn’t be surprised if nothing is significant. I’ll let someone else actually do this analysis (it’s easy enough in R).

    BTW, it looks like the survival times themselves cluster into 3 groups. I wonder if a couple different people were doing the surgeries and the mice weren’t randomized evenly.

    As you aptly pointed out, it’s no surprise that the journal has a crap impact factor.

    Link to this
  2. 2. scicurious 5:02 pm 09/19/2013

    dpryan: Thank you for this comment about the stats! They aren’t my specialty, but for this I would have done a one-way ANOVA (had this been mine). Would that have been inappropriate? I also appreciated the additional critique.

    Link to this
  3. 3. dpryan 1:52 pm 09/20/2013

    Yeah, a one-way ANOVA would have been as inappropriate as a T-test (a T-test is just a one-way ANOVA with 2 groups). I see a lot of people doing one-way ANOVAs followed by post-hoc tests to get all of the pair-wise p-values because they somehow think that’s better than just doing all of the t-tests straight-off and then adjusting the p-values (or not, depending on the context).

    They should have just treated survival data as is instead of breaking the assumptions of a T-test (or ANOVA, which have basically the same assumptions).

    BTW, keep up the good work, I quite enjoy the blog!

    Link to this
  4. 4. scicurious 1:54 pm 09/20/2013

    Ah, I see! But would you still use survival data for things like the cytokine levels? That’s where I would have used the one-way ANOVA…would that have been wrong?

    Link to this
  5. 5. dpryan 7:45 am 09/21/2013

    Ah, for cytokine levels or something like that either a one-way ANOVA or just directly doing the T-tests would be fine. Which one is preferred is largely dependent on the question you’re asking. If you want to know all of the pair-wise comparisons, then just do the T-tests and be done with it. If you just want to know if there’s a general treatment effect, then the one-way ANOVA is the tool of choice. With just 3 groups, you’ll inevitably just be doing a post-hoc test to get the pair-wise p-values anyway, so there’s really no benefit to an ANOVA in that case.

    Link to this

Add a Comment
You must sign in or register as a member to submit a comment.

More from Scientific American

Email this Article