A Sensitivity Problem in Pain Imaging: #cingulategate

Screen Shot 2015-12-31 at 1.03.04 AM

I know that the #cingulategate horse has been beaten to death, but I thought it would be worth putting it here, just in case someone had managed to miss it. It all started with a recent paper in PNAS by Lieberman and Eisenberger1 (L&E) that (in my opinion) makes some egregious statements about the specificity of the anterior midcingulate cortex based on Neurosynth – a nifty and useful meta-analysis software, ideal for hypothesis forming. Tal Yarkoni (TY) – the creator of Neurosynth – published a blogpost discussing the various problems with the paper.

L&E responded. And TY responded, and [correctly] suggested a retraction may be in order. Others have also weighed in, including a commentary by Alex Shackman, who also wrote an excellent review of cingulate function2 [although you should really read Vogt’s work on the cingulate3], and a commentary by Tor Wager on the Pain Research Forum (an excellent online resource for pain research). I recommend that every neuroimager, budding or expert, read this series of articles as they provide key information on inference and #hownottodoimaging. I say this because of a pervasive problem in pain imaging, and I suspect imaging in general. Many of these points are also discussed in a forthcoming book chapter4, and in a perspective piece in TiCS5. The latter focuses on L&E’s notion that activation of a loose set of brain regions activated by nociceptive stimuli – the so-called ‘pain matrix’ – is specific to pain. They then use this activation to conclude that “social rejection is painful.” The paper that led to #cingulategate (you can follow it on twitter; hashtag kudos to @dmellingson) basically uses a similarly flawed approach to make a conclusion about the function of the ‘anterior’ cingulate cortex [although it’s the anterior mid-cingulate cortex]. In a future post, I’ll discuss why this line of reasoning is incorrect, and will draw on examples from more recent literature.

References

1          Lieberman, M. D. & Eisenberger, N. I. The dorsal anterior cingulate cortex is selective for pain: Results from large-scale reverse inference. Proc Natl Acad Sci U S A 112, 15250-15255, doi:10.1073/pnas.1515083112 (2015).

2          Shackman, A. J. et al. The integration of negative affect, pain and cognitive control in the cingulate cortex. Nat Rev Neurosci 12, 154-167 (2011).

3          Vogt, B. A. in Cingulate neurobiology and disease   (ed B. A. Vogt) Ch. 1, 5-30 (Oxford University Press, 2009).

4          Moayedi, M. & Salomons, T. V. in An Introduction to Pain and Nervous System Disorders   (ed A. Battaglia)  (John Wiley & Sons Ltd, 2016).

5          Iannetti, G. D., Salomons, T. V., Moayedi, M., Mouraux, A. & Davis, K. D. Beyond metaphor: contrasting mechanisms of social and physical pain. Trends Cogn Sci 17, 371-378, doi:10.1016/j.tics.2013.06.002 (2013).

Leave a Reply

Your email address will not be published. Required fields are marked *