Wednesday 19 June 2013

Research Briefing: Is there a neural link between ‘neglect’ and ‘pseudoneglect’?

Source Article: Varnava, A., Dervinis, M. & Chambers, C.D. (2013). The predictive nature of pseudoneglect for visual neglect: Evidence from parietal theta burst stimulation. PLOS ONE 8(6): e65851. doi:10.1371/journal.pone.0065851. [pdf] [data and analyses] 

I’m excited about this latest research briefing for several reasons.

First, as I’ll explain below, I think the study tells us something new about how the human brain represents space, with potential clinical applications in neuropsychology. Second, the study represents my group’s first excursion into the world of open access publishing and open science (including open data sharing) – something I feel strongly about and have committed to pursuing in our recently awarded BBSRC project. And finally, the manuscript itself has a rocky history that left me disillusioned with the journal Neuropsychologia and, soon after, motivated me to join others in calling for publishing reform. 

The Research 

Lets start by talking about the science. Our aim in this study was to test for a link between two types of visual spatial bias called ‘neglect’ and ‘pseudoneglect’.

Neglect (also known as ‘unilateral neglect’) is a neurological syndrome that arises after brain injury – most often due to a stroke that permanently damages the right hemisphere. Patients with neglect present with a striking lack of attention and awareness to objects presented on the left side of their midline. Such behaviours may include ignoring food on the left side of a dinner plate or failing to draw the left side of objects. Importantly, the patients aren’t simply blind on their left side. The visual parts of the brain are generally intact while the damage is limited to parietal, temporal, or frontal cortex.

Neglect has been studied for many years and we know a lot about how and why it arises. But one unanswered question is how the spatial bias of neglect relates to other spatial biases that are completely normal. We felt this was an important question because we don’t know enough at a basic level about how the brain represents space, so testing for neurocognitive links between spatial phenomena helps us build better theories. Furthermore, if there happens to be a predictive relationship between neglect and other forms of bias, we may be able to estimate the likely severity of neglect before a person has a stroke. This could have a range of useful applications in clinical therapy and management.

Enter ‘pseudoneglect’. Pseudoneglect is a normal bias in which people ignore a small part of their left or right side of space. One simple way to measure this is to ask someone to cross the centre of a straight horizontal line. Most people will misbisect the line to the left or right of its true centre. This effect is tiny (in the order of millimetres) but reliable.

In this study we wanted to know whether patterns of pre-existing bias, as reflected by pseudoneglect, predict the patterns of actual neglect following neurological interference. Of course, we couldn’t give our participants permanent brain injury, so we decided to use transcranial magnetic stimulation (TMS) to simulate some of the effects of a brain lesion. Using a particular kind of repetitive TMS called ‘theta burst stimulation’, we temporarily suppressed activity in parts of the brain while people did tasks that measured their spatial bias. To see if there was a link between systems, we then related these effects of TMS on spatial bias to people’s intrinsic pseudoneglect.

As expected by previous studies, we found that TMS of the right parietal cortex induced neglect-like behaviour – compared to a sham TMS condition (placebo), people bisected lines more to the right of centre, indicating that TMS caused a subtle neglect of the left side of space. This effect lasted for an hour (upper figure on the left). But what was particularly striking was that the effect only happened in the participants who already showed an intrinsic pattern of left pseudoneglect. In contrast, those with right pseudoneglect at baseline were immune to the effects of TMS (lower figure on the left).

There were a number of other aspects to the study too. We compared the effect of TMS using two different methods of estimating bias, and we also asked whether the TMS influenced people’s eye movements (it didn't). I won’t go into these details here but the paper covers them in depth.

What do these results mean? I think they have two implications. First, they provide evidence that neglect and pseudoneglect arise from linked or common brain systems for representing space – and they provide a biological substrate for this association in the right parietal cortex. Second, the results provide a proof of principle that initial spatial biases can predict subsequent effects of neurological interference. In theory, this could one day lead to pre-diagnostic screening to determine whether a person is at risk of more severe neglect symptoms in the event of suffering a stroke.

All that said, we need to be cautious. There is a world of difference between the subtle and reversible effects of TMS and the dramatic effects of brain injury. We simply don't know whether the predictive relationship found here would translate to patients – that remains to be established. Also, our study had a small sample size, has yet to be replicated, and provides no indication of diagnostic or prognostic utility. But I think these preliminary results provide enough evidence that this avenue is worth pursuing. 

Open Access, Open Science, and Publishing Reform 

Apart from the science, our paper represents a milestone for my group in terms of our publishing practices. This is our first article in PLOS ONE and our first publication in an open access journal. Also, it is our first attempt at open science. Interested readers can download our data and analyses from Figshare (linked here and in the article itself). I increasingly feel that scientists like me who conduct research using public funds have an obligation to make our articles and data publicly available.

This paper also represents a turning point for me in terms of my attitude to scientific publishing. We originally submitted this manuscript in 2012 to the journal Neuropsychologia, where it was rejected because some of our results were statistically non-significant. Rejecting papers on the basis of ‘imperfect’ results is harmful to science because it enforces publication bias and pushes authors to engage in a host of questionable research practices to generate outcomes that are neat and eye-catching. With some ‘finessing’ of the analyses, we could probably have published our paper in a more ‘traditional’ outlet. But we decided to play a straight bat, and when we were penalised for doing so I realised on a very personal level that there was something deeply wrong with our publishing culture. As a consequence I severed my relationship with Neuropsychologia.

A short time later, I was contacted by Sergio Della Sala, the Editor-in-chief of Cortex, who read my open letter to Neuropsychologia. Sergio very kindly offered me an associate editorship (never let it be said that blogging is a waste of time!) and together we built the Registered Reports initiative. Our hope is that this new option for authors will help reform the incentive structure of academic publishing. Since then we’ve been part of a growing movement for change, alongside Perspectives on Psychological Science and their outstanding Registered Replications project, the Open Science Framework, and a special issue at Frontiers in Cognition which has adopted a variant of the Cortex pre-registration model.

In early June this year, Marcus Munafò and I, together with more than 80 of our colleagues, published an article in the Guardian calling for Registered Reports to be offered by journals across the life sciences. I’m delighted to report that the journal Attention, Perception and Psychophysics and two other academic journals are now on the verge of launching their own Registered Reports projects.

My small part in this reform traces back to having this manuscript rejected by Neuropsychologia editor Jennifer Coull in September 2012. So, in a very true sense, I owe Jennifer a debt of gratitude for giving me the kick in the butt I needed. Sometimes rock bottom can be a great launching pad.


1 comment:

  1. I'm also working in the area of neglect and pseudoneglect. In the spirit of open science and method sharing have you come across the 'bumping task'? It's a really good real-world indicator of pseudoneglect where pen and paper tasks can fall down in my opinion. It's from a 2007 paper published by Nicholls and colleagues, entitled "Things that go bump in the right". Really elegant study.