A New Decade of ME Research: The 11th Invest in ME International ME Conference 2016
Mark Berry presents the first in a series of articles on the 11th Invest in ME International ME Conference in London ...
Discuss the article on the Forums.

Biomarkers for autism discovered

Discussion in 'Other Health News and Research' started by fla, Mar 23, 2012.

  1. fla

    fla Senior Member

    Messages:
    234
    Likes:
    45
    Montreal, Canada
    Firestormm and Enid like this.
  2. adreno

    adreno PR activist

    Messages:
    4,841
    Likes:
    11,074
    Very interesting, thanks.
     
  3. Enid

    Enid Senior Member

    Messages:
    3,309
    Likes:
    862
    UK
    :victory: Stunning news - thanks fla - and an ME test with another immune related protein signature - let's hope. (And treat Autism more effectively in the meantime).
     
  4. Merry

    Merry Senior Member

    Messages:
    1,358
    Likes:
    624
    Columbus, Ohio, USA
    In my family the autism story and the ME story are connected, so this is very good news indeed.
     
  5. oceanblue

    oceanblue Guest

    Messages:
    1,174
    Likes:
    362
    UK
    Sadly, this is the usual story with no replication on an independent sample. The thresholds used are chosen to maximise the difference between patients and controls in this particular sample which artificially inflates the effectiveness of the test; until the results are replicated using the same thresholds on an independent sample it really doesn't mean much. Worse still, this isn't one biomarker but a combination of 3, which means 3 thresholds to tweak to maximise the difference between patients and samples; replication on an independent sample is key.

    In any case, the ROC (AUC) for all subjects of 0.81 is nothing to write home about; the ROC of 0.92 for those without sample hemolysis (indicating a damages sample) is more impressive, but is based on only 49 subjects.

    Given that ASD is a heterogenous condition, much bigger independent samples are needed. Hopefully they will do this, at which point they may be able to back up the claim that:
    Meanwhile, I'm sure it will grab a lot of headlines.

    Sorry if I sound irritated, but I am a little fed up with researchers jumping up and down about biomarkers everytime they make an interesting finding. I suppose it's good to know it isn't restricted to CFS research.
     
    alex3619 and Firestormm like this.

See more popular forum discussions.

Share This Page