Tuesday, October 12, 2010

Statistics, Reporting, and Flaws: Caution about Jaundice-Autism Link

Recently Danish researchers published in Pediatrics a link between Jaundice and Autism in infants, based on statistical data gathered.  The data suggests a link between children who are born with Jaundice (a yellowing of the skin at birth due to red blood cells dying) and those children with that condition who were subsequently diagnosed with Autism.  The research found that 67% of those children born with jaundice subsequently developed autism. Sounds pretty convincing, right?  Let's look at the numbers first.


Statistics is a simple comparison of data to determine relationships, whether or not any real relationships exist.  In this case, the statistical relationship suggests causation, but that is not what is happening (as I'm sure the researchers would agree).


Currently, one in every 78 boys are diagnosed with Autism in the US and in the UK.  That's almost 25% of the population of boys.  Children in general come in at 1 in every 98 (though there could be more girls who are going undiagnosed, according to other surveys run), which makes it comfortably within the 1% of all children.


Now, according to the Children's Liver Disease Foundation's pamphlet on Jaundice in the new born baby, 90% of all new born babies get jaundice.  So figuring that out of every 100 babies, 90 will have jaundice, and out of those same 100 babies, one will have Autism, it doesn't quite match the research. 


So, why doesn't it match?  Is there a problem with the research, or a problem with the numbers?  The statistics were taken from a baby population of over 733,000.  It represents the results from that statistical demographic, which could in some way apply to the world population as a whole.  Instead, it's quite possible it is one of two things:


1.  There is a link, and Autism rates will be going up based on diagnosis.  But the research is not calling this certain, but rather posing a question that needs to be answered with more research. 


2.  It's just a coincidence that happened to show within this demographic, and is not indicative of the overall population of children with jaundice, or children with Autism.  I'm leaning to this conclusion, if for no other reason than the disparity of numbers and lack of additional research.


Ultimately, it just means the medical community has a lot more research to do on Autism and biological links to the condition.  If you had a baby that got jaundice within days of being born, don't panic.  While it never hurts to learn more about Autism from reputable sources, your child's jaundice is not necessarily indicative of your child developing Autism.  Look beyond the news hype, and look for the facts that will eventually be shown in additional clinical research.


No comments: