Back Pains, Rubbery Brains, Doubts Remain

A while back Ben Wand blogged here about grey matter density changes in the brain and chronic pain – and particularly about a study demonstrating grey matter density reductions in patients with arthritic hip pain that reversed to a degree after successful hip replacement surgery.

A new study, published in the Journal of Neuroscience by David Seminowicz and colleagues tells what appears to be a similar story. They looked at a few things but I want to focus on their measures of structural brain abnormalities. Comparing 18 chronic back pain patients with 18 pain-free folk they collected clinical data and measured the thickness of the cortex across the brain with MRI scans. Then they measured them again 6 months later (10 of the healthy volunteers and 14 of the back pain patients). Over those 6 months they collected information regarding the treatments the patients had received for their back pains. MRI studies are a statistical nightmare with multiple comparisons being juggled about, but to my non-expert reading the researchers are clear, transparent and reasonable in how they have gone about things.

They found reductions in the thickness of the cortex in back pain patients, particularly in the dorsolateral prefrontal cortex (DLPFC). At follow up they found that where patients had experienced “successful treatment” (anyone whose pain had reduced at all), cortical thickness had increased again in this region and this increase correlated with reductions in pain and disability. They suggest that this indicates that treating back pain leads to increased cortical thickness. I would take issue on this point of interpretation. Those labelled with “successful treatment” all had treatments that aren’t well known for their efficacy, like spinal surgery or facet joint injections and no control group was in place for these treatments. So surely it is fairer to say that reductions in pain and disability, for whatever reason, are associated with a return of cortical thickness?

These results are fascinating. When Vania Apkarian and his lab first demonstrated grey matter density reductions in the right DLPFC in chronic back pain there was much worrying about it possibly representing neural degeneration. Perhaps this stark change was driving the chronic pain state, and perhaps it is irreversible. It sounded pretty catastrophic.  Cortical thickness is not the same thing as grey matter density but these results happily suggest that structural brain changes detectable in chronic pain are not a cause of irreversible brain rot, and never were. The funny thing about all of these studies is that no one really knows what actual mechanism underpins the structural changes that we think we are measuring – and that makes interpretation very difficult. At best we might say that the experience of pain is reflected in what appear to be structural changes in the brain. That still seems surprising to me, that the brain would change in that way in association with perceptual phenomena. Rubber Soul indeed.

Lots of studies have demonstrated brain density changes in a variety of chronic pain conditions. “Great!” you might think – “consistent replicable findings”. But there are important inconsistencies. Apkarian’s group found grey matter changes in the right DLPFC whereas Seminowicz and friends find it in the left. While some studies show a negative correlation between grey matter density and pain duration, others show a positive correlation (see this review by Patrick Wood) and some show no correlation at all. The lack of correlation might simply be a function of the small samples used but this still throws up some unwelcome doubt.

Not perhaps as much doubt as a new review in separate but similar field of research from the epidemiologist John Ioannidis. He reviewed all of the data from studies of brain volume abnormalities in various mental health conditions. Neuroskeptic’s excellent blog covered the study. He specifically went hunting for what he calls “excess significance bias”. In any field of research you expect only the studies that have adequate numbers of participants (statistical power) to actually find a positive result. Working backwards to calculate the power of original studies he found way too many positive findings given the numbers of people included. Even if the abnormalities were there, many of these studies were too small to detect it and yet way too many did.

How can that be? Publication bias, selective reporting of only positive findings and some cheekily lenient analyses have probably all played a part. Lots of small studies and lots of positive results? We would be muppets if we ignored the parallels in our own field.

So where does that leave us? The brain seems to demonstrate structural changes in people with chronic pain, we don’t know what they mean, they don’t appear to be catastrophic and we should be cautious about the body of evidence. It makes my head hurt (and brain shrivel?). I reckon Gustave Flaubert had it right:

“The deplorable mania of doubt exhausts me. I doubt about everything, even my doubts.”

Neil O’Connell

Neil OConnellAs well as writing for Body in Mind, Neil O’Connell is a researcher in the Centre for Research in Rehabilitation, Brunel University, West London, UK. He divides his time between research and training new physiotherapists and previously worked extensively as a musculoskeletal physiotherapist. He also tweets! @NeilOConnell

He is currently fighting his way through a PhD investigating chronic low back pain and cortically directed treatment approaches. He is particularly interested in low back pain, pain generally and the rigorous testing of treatments.


Seminowicz DA, Wideman TH, Naso L, Hatami-Khoroushahi Z, Fallatah S, Ware MA, Jarzem P, Bushnell MC, Shir Y, Ouellet JA, & Stone LS (2011). Effective treatment of chronic low back pain in humans reverses abnormal brain anatomy and function. The Journal of neuroscience, 31 (20), 7540-50 PMID: 21593339

Apkarian, A. (2004). Chronic Back Pain Is Associated with Decreased Prefrontal and Thalamic Gray Matter Density Journal of Neuroscience, 24 (46), 10410-10415 DOI: 10.1523/JNEUROSCI.2541-04.2004

Wood PB (2010). Variations in brain gray matter associated with chronic pain. Current rheumatology reports, 12 (6), 462-9 PMID: 20857244

Ioannidis JP (2011). Excess Significance Bias in the Literature on Brain Volume Abnormalities. Archives of general psychiatry PMID: 21464342


  1. I have been wary of reporting on brain changes correlated with pain, despite the juicy headlines, because I wasn’t sure what to make of them, and I rather suspected no one else did either. You’ve certainly settled that for me, and it’s off my maybe-to-do list. Thanks for lightening my load!

    I’m a little gobsmacked that Seminowicz et al. would conclude “that treating back pain leads to increased cortical thickness.” Really? Over reach the evidence much, guys? Strange. And good of you to challenge that interpretation.

    And that Ioannidis paper is great. Yay Ioannidis.

    Neil O'Connell Reply:

    Thanks a lot Paul. Agreed: ioannidis is an important dude!

  2. The difference in change in the side (2 research studies) as measured is to do with male /female and the context of the emotion associated ith chronic pain .Also it is better to look at CT scans as they resonant at same frequency as emotions in body as we are dealing with chronic pain . There are measureable changes seen with ‘rings ‘disapperaing on healing of the problem. Exciting stuff.

    Neil O'Connell Reply:

    Hi Alison,

    Thanks for your comments. I don’t share your interpretation though, although i may not have understood you correctly. The concept of gender specifity here seems tenuous to me. Similarly the idea of emotions resonating at given frequency is a little “out there” from a scientific perspective. the “rings” reference leaves me baffled.

    alison lingwood Reply:

    Thanks for getting back.I will try to find some references for you.

    As ct scans measure water content to some extent and the water stores
    vibratory information then our negative emotions associated with an event can be seen
    in the area of the brain that is representative of where we have a physical symptom. When the event is resolved the
    ring/oedema disappears. The brain layers,medulla for m/s pain, rather than
    grey matter or cortical thickness is the place to look. Worked out
    New ideas and stuff but works clinically!

    Neil O'Connell Reply:

    Thanks Alison,

    I would have to contest that ideas such as water storing vibrations associated with emotion is not neuroscience (or indeed science) as we know it. It reminds me of some of the “memory of water” bumpf discussed around homeopathy which, in my view, is totally false.

    Often alt med articles misrepresent scientific ideas or misappropriate them into often quite outlandish theories and I wonder if the papers you might have read on this have done the same.

    Also why would you look specifically to the medulla?

    How would one apply the idea that chronic pain is represented in the medulla by water holding onto special vibrations into clinical practice in order for it to “work clinically”?