Science fascinates many people including me. I love the complexity of it, the technology behind it, the sheer enormity of it. The hospital and laboratory scenes of everyone's favourite unpronounceable physician, Dr Hfuhruhurr, kinda sum up science: flashing lights, bubbling test tubes, 'the cranial screwtop method' and brains [and the words 'get that cat out of here']. During my very earliest days in research, one area of brain science was always a marvel to me: conducting and interpreting electroencephalograms (EEGs).
I have never been physically present during an EEG but have had the opportunity to look as some of the raw data produced and it fascinated me. What is perhaps so interesting is that a person, suitably trained, can look at the various squiggles of an EEG and determine the presence of an epileptic seizure and also, depending on the activity of the EEG, diagnose different types of seizure even pointing to differing epilepsy-related conditions. I know of at least one physician who is skilled in this dark art. Yes, the Force is indeed strong with them.
I talked about epilepsy and autism in a previous post, although not specifically discussing the use of EEGs. When looking at the literature about autism and EEGs, there are quite a few interesting points to note about things like the comorbidity of abnormal EEG results and the presence of epilepsy in some cases of autism. Also are the various suggestions that EEGs might serve as a potential objective early marker of autism and possibly a few other things. Mmm, we will see.
The study at the centre of this post was published in BMC Medicine by Bosl and colleagues* a few months back (full-text available here). The suggestion was that a statistical algorithm applied to EEG results might be able to classify very young children as either typical developers or as being at-risk for developing an autism spectrum condition. The results whilst small-scale made a few media waves along the lines of 'it will change the field, if this works' (comment made by one of the study authors).
Of course autism, and the various issues around it, is never so straight forward. Guaranteed that with every giant claim about diagnosis, treatment, aetiology, etc. an opposing view will always emerge, and like a classical opera, scientific battle breaks out ('no you can't, yes I can'). In the case of Bosl and colleagues it was this paper** which appeared in the same journal a few weeks later, questioning some of the claims suggested. The main crux of the criticism seems to be about the use of high-risk at population levels vs. high-risk at an individual level.
Fair's fair there is an author response to the criticism of their paper, available full-text here. They do offer some explanation for why they chose to record and report their results as they did. Credit also to them in highlighting why the media took so much interest in the story: novel medical research + lots of public interest + practical implications = a good story.
Its not just autism that is getting the EEG-biomarker treatment; so is Chronic Fatigue Syndrome (CFS) with the publication of this paper (again full-text) by Duffy and colleagues*** published in BMC Neurology. There hasn't been time for a formal scientific criticism of Duffy's paper but I'm sure it will come eventually. I have to say that the sample size included in this recent CFS paper is pretty big: 70 patients with well-defined CFS, 24 with depression, 148 with general fatigue and 390 asymptomatic controls. With a large sample size comes greater study power and hence greater predictive value. Indeed Duffy and colleagues are talking about nearly 90% of unmedicated women with CFS and over 90% of controls being correctly classified using their method.
As an outsider looking in on the world of EEG use and research, I have to say that I am intrigued by the possibilities of EEG-derived biomarkers even if only as part of a suite of investigations. Aside from the fact that EEGs are fairly inexpensive to do and importantly not invasive like the recent data on motor skills, I would like to think that the electrical activity of our brains, or rather their patterns of activity, are just waiting to be decoded in much the same way as the analysis of any other genetic or biological system might hide various secrets about our human condition. At the very least, EEGs might provide a window on the effectiveness of interventions targeting the brain as exampled here with the antidepressant treatment response.
To end, Dr Hfuhruhurr imparts his pearls of medical wisdom.
* Bosl W. et al. EEG complexity as a biomarker for autism spectrum disorder risk. BMC Medicine. February 2011.
** Griffin R. & Westbury C. Infant EEG activity as a biomarker for autism: a promising approach or a false promise? BMC Medicine. May 2011.
*** Duffy FH. et al. EEG spectral coherence data distinguish chronic fatigue syndrome patients from healthy controls and depressed patients-A case control study. BMC Neurology. July 2011.