Further results from the thimerosal-autism study

14 Sep

The recent study on thimerosal and autism was extensive. Much data and many results were included in two technical reports (nearly 400 pages total, volume 1 and volume 2). I haven’t had the time to read them thoroughly yet, but I did catch some interesting pieces of information.

The authors give the ASD prevalence (cases/1000) as a function of HMO and year of birth:

This is pretty flat. No huge increase seen over that 6 year time span. No evidence for an “epidemic”. The prevalence is about 1.1%. This is consistent with the current value quoted by the CDC.

There are remarkably a big variation by parents educational status (63% kids born to mothers who graduated from college). However, since the researchers are working with kids who are enrolled in an HMO, it seems likely this is due to some bias. I.e. the HMO participants could be better educated (and higher income) than the average.

There are indications that increasing parental age is related to increased autism risk. 36% of mothers were in the 30-34 age bracket, 35% were in the 35+ age bracket. So, about 70% of mothers were over 30 years of age. They list father’s age with different brackets, but 79% were 30 years old or older at the time of birth of their child. Again, could be some bias due to HMO membership, but a large fraction of the parents were older.

Some children (both ASD and controls) received no vaccines. Many received vaccines but no thimerosal–i.e. all their vaccines were thimerosal free.

The use of prenatal vitamins is given as having an increased risk of autism, but the odds ratio is not given.

Being a first-born or second-born child has a significant increased risk for autism (hazard ratios of about 1.6). They may discuss this, but I don’t know right now how much of this is due to parents of autistic kids deciding not to have larger families.

One of the stranger results–there is an increased risk of autism when mothers were using prenatal vitamins with folic acid. The hazard ratio was 2.3 with a p value of 0.0176

Pica and childhood lead exposures had very high hazard ratios: 3.7. This is a good case where it is worth asking if this is causal–does pica cause autism or, as is more likely, does autism cause pica and, with it, higher lead exposures.

Poverty was slightly protective for autism (hazard ratio of 0.92). Again, one has to question if this is real or whether poverty just results in a greater likelihood that a child’s autism will be misdiagnosed.

The authors list coexisting conditions for the autistic, ASD and control children:

Epilepsy is much higher at about 5%, compared to 1.6% for controls. Reports of the prevalence of epilepsy amongst autistics are often much higher, though.

Developmental delay (by parent report) is at about 18% for autistics, compared to about 0.66% in controls.

Gi disorder prevalence is about 2% amongst autistics. It is the same (or slightly higher) for controls. This is very interesting given the anecdotal reports of a high prevalence of GI disorders amongst autistics. I suspect this will form some of the complaints about this study–some will say they aren’t looking at the correct population and that a specific study on autism/regression/GI complaints needs to be done.

Cases (those with ASD) were more likely to get thimerosal free HepB and HIB vaccines.

Infants in this study do not get flu vaccines (near zero). Unless that habit has changed dramatically in the past few years–and that most doctors are giving infants flu vaccines with thimerosal–flu vaccines are not likely to be a reason for the continued climb in autism prevalence.

There is a lot more information there. If/when I get the chance to give the reports a more thorough read I’ll post what I find.

16 Responses to “Further results from the thimerosal-autism study”

  1. ANB September 14, 2010 at 19:45 #

    One of the stranger results—there is an increased risk of autism when mothers were using prenatal vitamins with folic acid. The hazard ratio was 2.3 with a p value of 0.0176

    Isn’t folic acid recommended for expectant mother’s over 35?

  2. Sullivan September 14, 2010 at 20:35 #

    ANB,

    yes, there is a strong recommendation for all mothers to take folic acid to prevent neural tube defects such as spina bifida.

  3. _Arthur September 14, 2010 at 20:45 #

    “One of the stranger results—there is an increased risk of autism when mothers were using prenatal vitamins with folic acid. The hazard ratio was 2.3 with a p value of 0.0176”

    There you have it:
    VITAMINS CAUSE AUTISM !!!!

    How silly were they wasting their angst obsessing about mercury additives.

    Now they will try a cure, chelate kids and feed them only white rice, to deplete their vitamins, to make them *normal* kids again.

  4. passionlessDrone September 14, 2010 at 21:30 #

    Hi Sullivan –

    One of the stranger results—there is an increased risk of autism when mothers were using prenatal vitamins with folic acid. The hazard ratio was 2.3 with a p value of 0.0176

    Wow, that is pretty crazy!

    IIRC, the potential for folic acid supplementation to have an effect was part of a discussion on this blog a few months ago.

    https://leftbrainrightbrain.co.uk/2010/01/autism-clusters-found-areas-with-high-incidence-of-autistic-children/

    MARBLES will be taking a closer look at this in particular.

    Thanks for posting this. Very nicely done.

    – pD

    • Sullivan September 14, 2010 at 22:38 #

      pD–

      you have a good memory.

      To respond to a comment you left elsewhere, there was a discussion that the CDC “tipped their hand”. It was on this blog.

  5. passionlessDrone September 14, 2010 at 21:42 #

    Hi ANB –

    Isn’t folic acid recommended for expectant mother’s over 35?

    It has also been supplemented into grains in many industrialized countries, including the US for the prevention of spina bifida and associated problems. Not every women knows she is pregnant when the formative events are happening. And, lots of women do know they are pregnant, but don’t take a prenatal vitamin.

    Spina bifida and anencephaly before and after folic acid mandate–United States, 1995-1996 and 1999-2000

    This report summarizes the results of that analysis, which indicate that the estimated number of NTD-affected pregnancies in the United States declined from 4,000 in 1995-1996 to 3,000 in 1999-2000. This decline in NTD-affected pregnancies highlights the partial success of the U.S. folic acid fortification program as a public health strategy. To reduce further the number of NTD-affected pregnancies, all women capable of becoming pregnant should follow the USPHS recommendation and consume 400 microg of folic acid every day.

  6. les September 14, 2010 at 22:17 #

    It is increased PATERNAL Age that is associated with an increased risk of autism and other problems in offspring.

  7. CuriousCommentor September 14, 2010 at 22:28 #

    “This is pretty flat. No huge increase seen over that 6 year time span. No evidence for an “epidemic”. The prevalence is about 1.1%. This is consistent with the current value quoted by the CDC.”

    Yes, but look at the individual HMOs and compare the rates. Interesting that for a few of them, the rate of diagnosis was considerably less. I can’t help but think that this may either be bias or an anomaly.

    “There are remarkably a big variation by parents educational status (63% kids born to mothers who graduated from college). However, since the researchers are working with kids who are enrolled in an HMO, it seems likely this is due to some bias. I.e. the HMO participants could be better educated (and higher income) than the average.

    There are indications that increasing parental age is related to increased autism risk. 36% of mothers were in the 30-34 age bracket, 35% were in the 35+ age bracket. So, about 70% of mothers were over 30 years of age. They list father’s age with different brackets, but 79% were 30 years old or older at the time of birth of their child. Again, could be some bias due to HMO membership, but a large fraction of the parents were older.”

    I agree; I tend to think this is likely bias. Most professionals tend to have children later in life, so their HMO membership would reflect this fact.

    “Being a first-born or second-born child has a significant increased risk for autism (hazard ratios of about 1.6). They may discuss this, but I don’t know right now how much of this is due to parents of autistic kids deciding not to have larger families.”

    Interesting. I can’t seem to find that reference. Is there a higher increase for first or second-born?

    “Poverty was slightly protective for autism (hazard ratio of 0.92). Again, one has to question if this is real or whether poverty just results in a greater likelihood that a child’s autism will be misdiagnosed.”

    Agreed. Quality of care would certainly determine a proper diagnosis.

    Good analysis. I look forward to reading your assessment of the rest of the data.

    • Sullivan September 14, 2010 at 22:49 #

      CuriousCommentor,

      HMO-A only had 30 cases total. That is more than a factor of 10 fewer than the next biggest HMO. I am not surprised that there is a lot of scatter in the prevalence by year for that HMO.

  8. Sullivan September 14, 2010 at 23:02 #

    CuriousCommenter–

    and, note that HMO-C had an increase (about 50%) over the time span, but that the other HMO’s had different behaviors.

  9. passionlessDrone September 14, 2010 at 23:37 #

    Hi Sullivan –

    Thank you. I was pretty sure it was here that I saw that thread, but didn’t have the time or energy to hunt for it at the time.

    – pD

  10. Neuroskeptic September 15, 2010 at 16:55 #

    “Developmental delay (by parent report) is at about 18% for autistics, compared to about 0.66% in controls.”

    This struck me as odd, because this implies that 82% of the ASD kids were “high functioning” – but I’ve seen it claimed that developmental delay and some degree of mental retardation are the norm in autism, and HFA is rare; likewise the 5% rate of epilepsy is, as you say, much lower than other estimates.

    To be honest, I suspect what this indicates is that the older estimates of the prevalence of epilepsy & mental retardation are outdated. But it is a bit odd.

    • Sullivan September 15, 2010 at 17:49 #

      Neuroskeptic,

      When one looks at the California Department of Developmental Services (CDDS) data, one finds that those served under the autism label have a much lower percentage with intellectual disability and/or epilepsy than in the past.

      Even with that, 5% for epilepsy seems low.

  11. daedalus2u September 15, 2010 at 19:10 #

    Neural tube defects happen around the same time (3-4 weeks post cconception) as when thalidomide causes autism. Valproate is a little later.

    Inhibitors of nitric oxide synthase cause neural tube defects too. Folate rescues animals given NOS inhibitors from neural tube defects.

    Thalidomide is a NOS inhibitor. NOS inhibitors produce neuronal hyperplasia (larger brains with more neurons) as is seen in autism. There is substantial cross-talk between NO and folate (in both directions).

    There has been some published research showing a negative correlation between in utero exposure to smoking and ASDs. This study had a pretty low incidence of tobacco use and they coded it only as nicotine exposure. I suspect the effect of smoking (if there is one) is via the carbon monoxide in tobacco smoke which has a lot of cross-talk with NO pathways. Differential incidence of smoking with income might reflect some of that difference too.

    Smoking is associated with a reduced incidence of neural tube defects.

  12. Billy Cresp September 15, 2010 at 19:30 #

    I don’t completely get what the odds ratios mean. Especially the odds ratio for an increase in 1 unit of exposure. I wonder how they calculate that, but especially wonder what levels are being referred to as being increased from by 1 unit. I don’t think there would be equivalence for increases of 1 unit, starting from different numbers.

  13. Roger Kulp September 16, 2010 at 04:42 #

    Folate cycle mutations,MTHFR in particular,but other genes as well,and the circulatory and folate metabolism disorders they can cause are somewhat common in autism.I know,my mother and I both have these mutations,as well as an assortment of some of the problems they can cause.I,of course,have a lot more,and more severe ones than her.

    As I like to say,I believe high homocysteines,and low folic acid in the brain,was responsible for much of my autism related problems,like my eloping.

    As with women who have stillbirths,or neural tube defects,women who have MTHFR polymorphisms,especially of the C677T gene,can take all of the folic acid they want,and still have a baby with autism,or any of these other problems.If the mother cannot metabolize it,it’s pretty much the same as not taking it.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: