Brad Handley has commissioned a telephone polling company to perform a telephone poll:

Generation Rescue commissioned an independent opinion research firm, SurveyUSA of Verona NJ, to conduct a telephone survey in nine counties in California and Oregon. Counties were selected by Generation Rescue. Interviews were successfully completed in 11,817 households with one or more children age 4 to 17. From those 11,817 households, data on 17,674 children was gathered. Of the 17,674 children inventoried, 991 were described as being completely unvaccinated. For each unvaccinated child, a heath battery was administered.

Oooh – exciting!

The results are damning apparently….

We surveyed over 9,000 boys in California and Oregon and found that vaccinated boys had a 155% greater chance of having a neurological disorder like ADHD or autism than unvaccinated boys

Woah, what? Like autism…? And what the hell has ADHD got to do with anything? Oh right, right – I remember, Generation Rescue redesigned their site when they couldn’t make their old message of:

Autism is treatable. It’s reversible. It’s nothing more than mercury poisoning,” said JB Handley, founder of Generation Rescue.

stick. Now its more than just mercury and its more than just autism. Hey – if you can’t make one idea work, expand it and pretend you’ve always meant that. In this survey, applicants were asked about ADD, ADHD, Aspergers, PDD-NOS, Autism, Asthma and Juvenile Diabetes. Nothing like muddying the water to make things clearer, right Brad?

On the Generation Rescue page I link to above, Generation Rescue have kindly provided their source data but in closed access PDF’s. How helpful. Never mind, I turned all the aggregate data into an Excel file and had a bit of a look myself. UPDATE: All Generation Rescue Survey data is now available in Excel.

Now, my issue with Generation Rescue is solely to do with autism and vaccines. I really don’t care about their newly found interest in asthma or juvenile diabetes. Lets see what they say about their autism results:

Vaccinated boys were 61% more likely to have autism

Well, thats one way to look at it. Another way is to look at it properly. In the spreadsheet I created using Generation Rescue raw data the following was found.

Number of boys with Aspergers
Unvaccinated: 2% of total
Partially vaccinated: 3% of total
Fully vaccinated: 2%
Fully and Partially combined: 2%

Conclusion: you are 1% more likely to have Aspergers if you have been partially vaccinated. If you are fully vaccinated your chance of being Aspergers is exactly the same as if you were unvaccinated.

Number of boys with PDD-NOS
Unvaccinated: 1% of total
Partially vaccinated: 2% of total
Fully vaccinated: 1%
Fully and Partially combined: 1%

Conclusion: you are 1% more likely to have PDD-NOS if you have been partially vaccinated. If you are fully vaccinated your chance of being PDD-NOS is exactly the same as if you were unvaccinated.

Number of boys with Autism
Unvaccinated: 2% of total
Partially vaccinated: 7% of total
Fully vaccinated: 3%
Fully and Partially combined: 4%

Conclusion: you are 5% more likely to have autism if you have been partially vaccinated. If you are fully vaccinated your chance of being autistic is 1% greater than if you were unvaccinated.

Number of boys with all ASD’s
Unvaccinated: 4% of total
Partially vaccinated: 8% of total
Fully vaccinated: 5%
Fully and Partially combined: 5%

Conclusion: you are 4% more likely to have an ASD if you have been partially vaccinated. If you are fully vaccinated your chance of having an ASD is 1% greater than if you were unvaccinated.

These figures are laughable. 4% more likely? And that’s if your son has been partially vaccinated! If he’s been fully vaccinated the percentage increase drops to 1%. The figures for girls are even worse.

Number of girls with Aspergers
Unvaccinated: 1% of total
Partially vaccinated: 1% of total
Fully vaccinated: 0%
Fully and Partially combined: 0%

Conclusion: you are no more likely to have Aspergers if you have been partially vaccinated. If you are fully vaccinated your chance of being Aspergers is 1% less than if you were unvaccinated.

Number of girls with PDD-NOS
Unvaccinated: 2% of total
Partially vaccinated: 1% of total
Fully vaccinated: 0%
Fully and Partially combined: 0%

Conclusion: you are 1% more likely to have PDD-NOS if you are unvaccinated. If you are fully vaccinated your chance of being PDD-NOS is 2% less than if you were unvaccinated.

Number of girls with Autism
Unvaccinated: 1% of total
Partially vaccinated: 2% of total
Fully vaccinated: 1%
Fully and Partially combined: 1%

Conclusion: you are 1% more likely to have autism if you have been partially vaccinated. If you are fully vaccinated your chance of being autistic is no greater than if you were unvaccinated.

Number of girls with all ASD’s
Unvaccinated: 3% of total
Partially vaccinated: 3% of total
Fully vaccinated: 1%
Fully and Partially combined: 1%

Conclusion: you are no more likely to have an ASD if you have been partially vaccinated. If you are fully vaccinated your chance of having an ASD is 2% less than if you were unvaccinated.

My goodness, this is awful for Generation Rescue. Finally, we’ll look at girls and boys together:

Number of boys and girls with Aspergers
Unvaccinated: 1% of total
Partially vaccinated: 2% of total
Fully vaccinated: 1%
Fully and Partially combined: 2%

Conclusion: you are 1% more likely to have Aspergers if you have been partially vaccinated than unvaccinated. If you are fully vaccinated your chance of being Aspergers is no greater than if you were unvaccinated.

Number of boys and girls with PDD-NOS
Unvaccinated: 2% of total
Partially vaccinated: 2% of total
Fully vaccinated: 1%
Fully and Partially combined: 1%

Conclusion: you are 1% more likely to have PDD-NOS if you are unvaccinated. If you are fully vaccinated your chance of being PDD-NOS is 1% less than if you were unvaccinated.

Number of boys and girls with Autism
Unvaccinated: 2% of total
Partially vaccinated: 4% of total
Fully vaccinated: 2%
Fully and Partially combined: 2%

Conclusion: you are 2% more likely to have autism if you have been partially vaccinated. If you are fully vaccinated your chance of being autistic is no greater than if you were unvaccinated.

Number of boys and girls with all ASD’s
Unvaccinated: 4% of total
Partially vaccinated: 6% of total
Fully vaccinated: 3%
Fully and Partially combined: 3%

Conclusion:you are 2% more likely to have an ASD if you have been partially vaccinated. If you are fully vaccinated your chance of being autistic is 1% less than if you were unvaccinated.

There’s no getting away from this. This is a disaster for Generation Rescue and the whole ‘vaccines cause autism’ debacle. Generation Rescue’s data indicates that you are ‘safer’ from autism if you fully vaccinate than partially vaccinate. It also indicates that across the spectrum of autism, you are only 1% more likely to be autistic if you have had any sort of vaccination as oppose to no vaccinations at all – and thats only if you are male. If you are a girl you chances of being on the spectrum are less if you have been vaccinated! Across both boys and girls, your chances of being on the spectrum are less if you have received all vaccinations.

Thank you Brad, thank you very, very much.

Elsewhere


Orac
Prometheus

Comments

88 Responses to “Generation Rescue Survey Results”

  1. bethduckie on June 26th, 2007 11:17:26

    Oh my….

  2. Tom on June 26th, 2007 11:24:32

    Thanks for finding the truth behind the GR lies.

    I see that U.S. Representative Carol Maloney of New York reintroduced her bill, the “Comprehensive Comparative Study of Vaccinated and Unvaccinated Populations Act of 2007” (H.R. 2832). She did this on the very same day GR released their publicity stunt.

    Maloney has no idea how much she’s been lied to.

  3. Brian Deer on June 26th, 2007 11:32:21

    As a professor of medical statistics once said to me: “You have to remember that half the population is of below average intelligence.”

    I’m sure Generation Rescue would think it was more.

  4. bones on June 26th, 2007 12:09:12

    Makes Kirby’s Room-5 diatribe seem like genius. The EoH gang is idolizing JB for this. That’s what it’s all about, after all, isn’t it? Who gets the most recognition.

  5. notmercury on June 26th, 2007 13:22:46

    Conclusion: you are 5% more likely to have autism if you have been partially vaccinated.

    Or, once a child has been diagnosed as autistic the parents will read up on autism on the internet and stop vaccinating.

  6. mumkeepingsane on June 26th, 2007 13:37:15

    But doesn’t even one un-vaccinated child with autsim kinda wreck the whole theory?

    Oh, and when I found out my child had autism I went online, read the ‘vaccination theory’, realized it was unbelievably unsupportable and continuted vaccinating my son.

  7. Kev on June 26th, 2007 13:40:51

    “Or, once a child has been diagnosed as autistic the parents will read up on autism on the internet and stop vaccinating.”

    Exactly.

  8. notmercury on June 26th, 2007 14:12:41

    mumkeepingsane: But doesn’t even one un-vaccinated child with autsim kinda wreck the whole theory?

    One would think though I should have specified that they may stop vaccinating their autistic child hence the partially vaccinated status.

  9. Joseph on June 26th, 2007 14:16:39

    So comparing 2% to 3% they conclude it’s 60% more likely! That’s one of the most hilarious things I’ve seen in a long time. I’m ROTFL. Did they teach Brad about margin of error at Stanford?

    BTW, with “155% more likely” they of course mean “1.5 times”. It’s like with the “6000% increase” which really means “60 times”. It’s a pattern with GR apparently.

    If you consider likely confounds, e.g. distrust in healthcare, I think the results of this survey are meaningless, and Brad probably realizes that. He was certainly expecting to come up with something a lot more substantial. I’m sure he’s quite disappointed.

    In fairness, I should note there are obvious confounds in the other direction, and I expected the results to be more dismal and surprising.

  10. kristina on June 26th, 2007 14:20:49

    SurveyUSA notes that its research “can be completed within days”——it seems that this fast turnaround is a key selling point. Among the “private market research work” that they do are customer satisfaction surveys, brand positioning, talent testing, coincidental research…... (full list here). They’ve done research (they say) for Merck too.

  11. Joseph on June 26th, 2007 14:29:51

    BTW, aren’t these rates of autism kind of high, or am I interpreting that in the wrong way? How did they select the surveyed population?

    There does seem to be a possible trend in differences between boys and girls. I bet it’s because the confounds apply differently to boys vs. girls. But watch for the Geiers using that to suggest vaccines interact with testosterone somehow.

  12. culvercitycynic on June 26th, 2007 16:03:50

    How did they select the surveyed population?

    By selecting specific counties within specific Regional Centers, I wonder?

  13. Prometheus on June 26th, 2007 16:54:36

    I just ran the GR survey numbers on autism, comparing autism + PDD in unvaccinated and fully vaccinate boys and found that the difference is not statistically significant (p=0.2 for one-tailed and 0.4 for two-tailed distribution).

    But, this shouldn’t come as a surprise.

    Prometheus

  14. MercuryDad on June 26th, 2007 17:38:17

    Hey Prometheus:

    Whatever you do, don’t mention any numbers that may actually implicate vaccines. And, any interest in correcting Kevin on his math up above, or are you going to let that slide?

    All vaccinated boys, compared to unvaccinated boys: – Vaccinated boys were 155% more likely to have a neurological disorder (RR 2.55, p<0.001) – Vaccinated boys were 224% more likely to have ADHD (RR 3.24, p<0.001) – Vaccinated boys were 61% more likely to have autism (RR 1.61, p<0.05)

    Older vaccinated boys, ages 11-17 (about half the boys surveyed), compared to older unvaccinated boys: – Vaccinated boys were 158% more likely to have a neurological disorder (RR 2.58, p=0.001) – Vaccinated boys were 317% more likely to have ADHD (RR 4.17, p<0.02)
    (Note: older children may be a more reliable indicator because many children are not diagnosed until they are 6-8 years old, and we captured data beginning at age 4.)

  15. Kev on June 26th, 2007 17:55:13

    What’s this ‘neurological disorder’ stuff Brad? I’m talking about autism . If you think older kids are more significant, lets look at your data:

    Ages 11-17, all kids:
    Aspergers (unvaccinated): 1%
    Aspergers (part vaccinated): 2%
    Aspergers (full vaccinated): 2%
    Aspergers (comb part + full): 2%

    Thats a difference of 1%.

    PDD-NOS (unvaccinated): 1%
    PDD-NOS (part vaccinated): 2%
    PDD-NOS (full vaccinated): 1%
    PDD-NOS: 1%

    Thats no difference – except for part vaccinated who are subject to bias.

    autism (unvaccinated): 2%
    autism (part vaccinated): 4%
    autism (full vaccinated): 2%
    autism (comb part + full): 2%

    Thats no difference – except for part vaccinated who are subject to bias.

    ASD (unvaccinated): 3%
    ASD (part vaccinated): 5%
    ASD (full vaccinated): 3%
    ASD (comb part + full): 3%

    Thats no difference – except for part vaccinated who are subject to bias.

    So Brad – when we drop this silly ‘neurological disorder’ crapola and concentrate on what we know you’re really interested in – autism – there are no differences between unvaccinated kids and fully vaccinated kids . Do you see that?

  16. MercuryDad on June 26th, 2007 18:06:35

    Kevin:

    80% of the diagnoses are with the boys, and you are quoting only boys and girls. You have no interest in dealing with the clear differences the survey highlighted, which is fine with us, but looking at this by gender when 80% of the cases are with the boys is both reasonable and appropriate.

    Of course, doing so wouldn’t help your case so I’m sure you won’t bother.

    MercuryDad

  17. clone3g on June 26th, 2007 18:08:06

    Holy crap, did Brad hire 2.1 Geiers (RR 2.55, p 0.001) to do the number crunching for him?

  18. Kev on June 26th, 2007 18:19:36

    OK, so we have to now look not at younger kids and not at girls. What next? Star sign?

    Ages 11-17, boys:
    Aspergers (unvaccinated): 2%
    Aspergers (part vaccinated): 4%
    Aspergers (full vaccinated): 3%
    Aspergers (comb part + full): 3%

    Thats a difference of 1%.

    PDD-NOS (unvaccinated): 1%
    PDD-NOS (part vaccinated): 2%
    PDD-NOS (full vaccinated): 1%
    PDD-NOS (comb part + full): 1%

    Thats no difference – except for part vaccinated who are subject to bias.

    autism (unvaccinated): 2%
    autism (part vaccinated): 6%
    autism (full vaccinated): 3%
    autism (comb part + full): 3%

    Thats a difference of 1%.

    ASD (unvaccinated): 3%
    ASD (part vaccinated): 8%
    ASD (full vaccinated): 4%
    ASD (comb part + full): 5%

    Thats a difference of 1% for full vacc, 2% for full and part and 5% for part vacc (who are subject to bias).

    I really don’t know what to say to you Brad. You’re quibbling about the difference between 1% and 2% for ASD. If you really think these data are in any way significant then, well, good luck!

  19. Rich on June 26th, 2007 18:26:42

    I can’t believe they paid $200,000 for this. For that kind of money, you could do an entire prevalence study in India, screening up to 45,000 kids and doing ADOSs and ADIRs and really get some rates where our knowledge of autism is virtually nothing.

  20. Rich on June 26th, 2007 18:33:44

    I just hope this organization doesn’t claim that this is a “prevalence” study or try to use words like “risk ratio,” because having read their data and analysis, this is certainly not an epidemiological study. It would be crucial, of course, to know what their participation rate is. How many people hung up? People that are going to stay on the line through something like this are going to be people who have children with issues or are concerned about their kids having problems. Enormous problems with selection bias. The study will be completely ignored by scientists. But they will probably get the ears of congressmen, etc. and further decrease vaccination rates. In the time it took me to write this comment, 5-10 probably died somewhere in the world from Measles. I would not want to be responsible for decreasing vaccination rates.

  21. Ms. Clark on June 26th, 2007 18:38:18

    oooooh,

    so when does this get published in a peer reviewed journal and not just on a pharma sponserd news webstie (PharmaLive?)

    Without knowing how slanted the questions were and without knowing how many people hung up on the surveyors because their kids had no problems at all, how can we know the significance of the numbers. If only 5 people with unvaxed kids claimed to have children who were totally free of ADHD etc, and 800 people claimed to have vaxed and partially vaxed kids who had ADHD then the results are meaningless.

    People who don’t vaccinate their kids are probably less likely to see out diagnoses of anything for their kids. They might not even don’t have regular doctors because regular doctors sometimes won’t accept antivax patitents.

    Also, are the kids unvaxed because they have an genetic disorder or something that makes them sensitive to vaccines (DeGeorge syndrome?), or are allergic to eggs or something?

    Did Brad’s surveyors ask why parents didn’t vaccinate? Fear of “the man”? Fear of the medical system? Did their kid get an ASD dx from “the man” and did the parent then rediagnose the kid as “an indigo” or “a crystal”?

    Maybe Generation Rescue will become heroic in the lives of ADHD kids now who will be getting EDTA suppositories by the gross and vinegar and garlic IV’s by the gallon, and GR will just leave autism alone.

  22. caseofthevapours on June 26th, 2007 18:41:29

    A prevalence study in India would have been interesting and helpful as well. But Brad has stated more than once that this is the group he’s most-interested in advocating for:

    White. Upper income. Married. College degree. Concerned about vaccine safety and do their own thinking rather than having a doctor do it for them

  23. passionlessDrone on June 26th, 2007 18:57:46

    Hi Kev –
    You are performing some very curious operations with percentage values. Why?

    For example, lets evaluate the alcohol content of beer and wine.

    beer: 5%
    wine: 10%

    Tonight I may drink 100 ounces of beer. Tomorrow I may drink 100 ounces of wine. How much drunker will I get on the second night? 5% or 100%?

    You may also be interested in knowing that Excel is capable of calculating numbers to an area to the right of the decimal point. When comparing differences between small numbers, this is frequently useful.

    I’m not here to defend a phone study; they can be done very poorly by accident or on purpose. But that is no reason to pile on with more bogus information.

    -pD

  24. Joseph on June 26th, 2007 19:04:01

    I think there’s a clear reason why the results for “neurological disorders” like ADHD are more pronounced than those of autism. ADHD is a diagnosis that is easier to “miss” (using curebie terminology). A parent who is against vaccination, anti-science and anti-medical establishment, is also less likely to see a psychiatrist about having a child labeled with ADHD. It would be fairly easy to get away with that too for something like ADHD, despite pressure from schools and so forth. (Note that ADHD diagnoses are given very liberaly in the US, with about 10% of children diagnosed).

  25. Ms. Clark on June 26th, 2007 19:25:44

    So is Brad’s final message that it doesn’t matter if you vaccinate or not, your kid has the same chance of having “a neurological disorder”, but whatever you do, don’t partially follow the vaccine schedule?

    Leaving a kid unvaccinated is far too dangerous an option, unless there is a medical reason not to vaccinate (even AIDS babies get vaccines in Zimbabwe) . The kids who can’t be vaccinated are depending on heard immunity even more than others are. Even Brad says kids should be vaccinated. So I guess the outcome of Brad’s study is fully vaccinate your kids according to the CDC schedule?

    Brad, you have a new baby in the house don’t you? Are you vaccinating that baby? Inquiring minds want to know.

  26. Kev on June 26th, 2007 19:45:19

    pD – download the data, or get it from Brad. I’m following the same percentage/decimal conventions as GR are. No tricks. If GR round it up or down then so do I.

    A 0 is a 0 no matter how many ways you try and retro fit it to an agenda.

  27. Prometheus on June 26th, 2007 20:01:15

    I can only wonder what statistical methods were used to come up with the p-values MercuryDad quoted. I wonder if we’ll ever know.

    Fortunately, since the raw data is available, we can do our OWN statistics, which – as I mentioned above – shows that the differences are not statistically significant.

    Of course, big issue is the composition of the survey itself, including the nature of the questions, the number of people contacted to get the number of responses (response rate) and the definitions of “autism” and ADD/ADHD used.

    Has anyone seen the accompanying full-page newspaper ad? I’ve heard about it from friends, but haven’t seen it myself. Reportedly, it comes across as being broadly anti-vaccination and doesn’t mention mercury even once.

    Prometheus

  28. MercuryDad on June 26th, 2007 20:02:59

    Criticism: Parents who do not vaccinate their children are less likely to seek an ND diagnosis, which explains the difference in prevalence you found.

    Response: We think the data disproves this, because we found no meaningful difference in prevalence for NDs between vaccinated and unvaccinated girls. If this was simply an issue of parent behavior, the girls would have shown wide discrepancies in prevalence, too, and they did not.

  29. Kev on June 26th, 2007 20:28:50
  30. Prometheus on June 26th, 2007 20:33:13

    Darn “less than” sign tripped me up! See the corrected comment below.

    Prometheus

  31. Prometheus on June 26th, 2007 20:33:59

    I ran the GR numbers again, looking just at boys, comparing “unvaccinated” with “any vaccinations” and lumping all ASD’s together and got a p=0.54 (not significant) by Chi Square and p=0.27 (not significant) by treating them as two independent proportions.

    If you take ALL “neurologic disorders” (they only report ASD and Add/ADHD) in boys, THEN you can get p less than 0.001. This, of course, begs the question of what – if any – connection there might be between autism and ADD/ADHD (apart from being over-used and poorly-defined diagnoses).

    Interestingly, with girls, the ASD-only comparison (“unvaccinated” vs “any vaccinations”) showed a statistically significant difference, where it wasn’t significant in the boys. This, apparently, wasn’t in line with their pre-conceived notions and so didn’t get reported.

    Unfortunately, this is just a telephone survey, where people were asked questions by people who don’t know very much about autism OR ADD/ADHD.

    It is full of sound and fury, signifying nothing.

    Prometheus

  32. Joseph on June 26th, 2007 20:39:20

    It is believed (although not proven either way) that the difference in ASD prevalence between boys and girls is largely cultural.

    The same cultural bias that presumably exists in seeking the diagnosis of autism for boys vs. girls could be at play in the survey results, Brad.

    And like I said, there’s at least one counfound that works in the other direction, so we could be looking at a sort of equilibrium of confounds in the results for girls. (Actually, girls seem to be at a slightly higher risk of ASD when not vaccinated, right? Are you going to mention this on the GR website, Brad?)

    Prometheus: “Has anyone seen the accompanying full-page newspaper ad?”

    That actually answers the question about the relatively high rates found by the survey. It also suggests other likely confounds.

  33. bones on June 26th, 2007 21:03:52

    If able, I recommend watching ABC Nightline News, 11:30 (EST, in the States). Tonight’s report features the OAP hearing, Geiers and Rev Sykes.

    Nothing like a good comedy.

  34. Margaret Romao Toigo on June 26th, 2007 21:19:36

    Regarding the ABC Nightline story, there is a related print article on the ABC News web site.

  35. Junior on June 26th, 2007 23:10:43

    “Spending time with the Geiers leaves one with the feeling that it is something other than money that motivates them to swim outside the mainstream. Perhaps it is defiance or ego or the love of people like Lisa Sykes, who trusts the Geiers and hope for a positive outcome.”

    From the ABC News story linked above. Get out the barf bags. Can someone contact Kathleen Seidel, in the comments section there is a link to email the segment producers if you have additional information about a story.

  36. Phil on June 27th, 2007 00:12:40

    MercuryDad/Brad,

    Give it up. Kev has owned you lock stock and barrel and you just won’t admit it.

    Get off the mercury bandwagon – you’re making a fool of yourself just like the lawyers for the petitioners in the Omnibus case are. Autism is not reversible and the sooner you admit that the better.

  37. Rich on June 27th, 2007 00:42:26

    so to follow up on Joseph: if you take the “rates” (I hate to legitimize such a “study” with the word “rates) of ASDs in this phone survey, doesn’t it come out to about 10%? So is this bizarre organization saying that 10% of kids have an ASD?

    That’s way more than the CDC’s 1 in 150.

  38. Bartholomew Cubbins on June 27th, 2007 01:30:22

    The good that could have come from this kind of money is maddening. What could have been: advocacy, disability rights, transition training, etc.

    This little phone tree was an ill-conceived venture and any attempt to defend it is simply circling the BMWs in an effort to calm donors.

  39. not by the hair of my chinny chin caplan on June 27th, 2007 02:26:45

    Mercury Dad,

    “Neurological disorder”? First rate buffoonery my friend. What do you do for an encore?

  40. JScarry on June 27th, 2007 03:10:54

    Kevin.
    Speaking as a retired economist who played with numbers all day, you are probably correct in thinking that the study has flaws but your interpretation is very, very wrong.
    Generally, if one sub-group has a 2% incidence and another has a 4% incidence then the second is 100% more likely than the first to be affected. Likewise, the first group is 50% as likely as the second to be affected.
    You should get some help from a statistician and then repost.

  41. Matt on June 27th, 2007 03:45:13

    Hello…a phone survey…sure…no, my parents are not at home… vaccinations..nope..I hate them… neurological disorder…sure, I have one….

    HeHe, hey Bevis, he said neurological disorder!

    Matt

  42. Interverbal on June 27th, 2007 03:45:59

    JScarry,

    These data should not be described as “incidence” data.

    Also, I have a question for you, that has three possible answers:

    When one compares 2 groups where the sample size of one group is only 6% of the size of the other; in which direction does this create bias?

    a)Towards the larger group?
    b)The smaller group?
    c)No problem here, it is a well designed study in this regard?

    It is carefully explained to the public what is meant by “100% more likely”. Especially since stating whopping percentages, sounds a lot more impressive then saying: “We failed to get a statistically significant score using a basic Chi-Square analysis”.

  43. preston on June 27th, 2007 04:19:11

    For example, lets evaluate the alcohol content of beer and wine.

    beer: 5%
    wine: 10%

    Tonight I may drink 100 ounces of beer. Tomorrow I may drink 100 ounces of wine. How much drunker will I get on the second night? 5% or 100%?

    Neither.

    Assuming you weigh about 200lbs, are male, and consume the beverages on both days over the course of about 4 hours, and that you define “drunker” as the difference in blood alcohol level, the answer is probably closer to 300%.

    Although day two’s blood alcohol level is 300% of day one, the difference between the two is about .23%

    When it comes to blood alcohol level, that’s a huge difference, but the only reason that difference is meaniful, is because the real physiological and behavioral effects in terms of “drunker” as a direct result of alcohol consumption are known.

    Real physiological and behavioral effects in terms of “autistic” as a direct result of vaccination have not been scientifically shown to exist (here’s the part where all the idiots say, “yeah, but they haven’t been shown to not exist”). Brad can shout for real studies (and he probably will) all he wants. But, for now, it sure looks like the methods of this collosal piece of waste point to the results being essentially worthless. Maybe it is a decent expression of the relative fear induced in parents who stop vaccinating at the “first sign” and internet signon. GR and their ilk could be having an effect. They seem to be reaching about 1%, which sounds about right when you consider that the total OAP cases probably represent about 1% of their respective age groups for autism spectrum disorders (5,000 is 1% of 500,000).

    It might grow difficult for Handley to find new stones to turn over as he continues to flush real $$$ down the toilet, and as the veins in his forehead begin to rupture.

  44. Kev on June 27th, 2007 05:46:34

    “Generally, if one sub-group has a 2% incidence and another has a 4% incidence then the second is 100% more likely than the first to be affected. Likewise, the first group is 50% as likely as the second to be affected.”

    Cool – if thats how you want to see it then be my guest. I’m posting on how I see it.

    As stated by others, using words like ‘incidence’ dignify this waste of time and money. As it is, I’ll take the Sue M tried and trusted ‘common sense’ point of view when it comes to rubbish studies/data.

  45. David N. Andrews M. Ed. (Distinction) on June 27th, 2007 09:20:51

    “But doesn’t even one un-vaccinated child with autsim kinda wreck the whole theory?”

    Actually, since they have been working on some sort of an induction proof, and since it takes only one counter-example to effectively destroy an induction proof… yup, their theory’s buggered.

    (“My kid was vaccinated and is now autistic, and her kid’s vacinated and is now autistic… and someone else’s kid was vaccinated and is now autistic…” is basically their pitch and it is basically an induction proof).

  46. Junior on June 27th, 2007 12:47:44

    “But doesn’t even one un-vaccinated child with autsim kinda wreck the whole theory?”

    I have the hardest time understanding the parents who did not vaccinate their child and the child is now autistic, but the parents are still firm believers in the mercury/vaccine causation idea. Even though they don’t belong in the club (cult?), they still want to be in it. They even chelate their children. Bizarre.

  47. Joseph on June 27th, 2007 14:15:02

    Kev’s explanation of the numbers is not quite right – that’s true. But there’s also an error in assuming that the risk ratio can be just calculated by dividing the proportions. 3% divided by 2% gives a 1.5 RR, but you can’t assume that’s what the RR is. What matters is the 95% confidence interval, which could be, say, 0.5 to 3.0 or something. As Prometheus explained, there is no statisticial significance in the RRs for autism at least, except an RR less than 1.0 for girls as I understood it. The average RR for boys and girls combined should still be less than 1.0, which is really what Kev’s post points out in the end.

    What’s interesting about this survey, as commented above, is that the fantasy that unvaccinated children don’t develop autism (e.g. the Amish and so forth) is no longer tenable. They are left with arguments about a small increased risk.

    Further, if Brad believes his own survey, I’m not sure if he can, in good conscience, not caution parents about not vaccinating girls. Is he fighting autism or is he fighting vaccines? He needs to ask himself that.

    Though to emphasize, the high “rates” found by the survey point to a significant methodological flaw in it.

  48. Kev on June 27th, 2007 14:43:15

    Jospeph’s absolutely right. I’m well aware I’m not using these numbers in a statistically correct manner. My point with this post is that these data are so far removed from anything approaching significance that anything can be read into them.

  49. 666sigma on June 27th, 2007 15:01:54

    I have not had time to look at the data, but one thing I am certain of from reading your comments. Kev, I’m sorry, but you simply are not qualified to interpret this data. You don’t have the training or educational background to analyze this data.

    And that’s the problem with virtually every study done on this subject, including the so-called reputable studies. The people doing them really are not properly trained to perform these kind of analysis.

    You’ll be happy to know that Excel will do the calculations for you if you know how to use them. Look under Tools for Data Analysis. Have fun.

  50. Kev on June 27th, 2007 15:31:40

    What???? Are you seriously suggesting my media/design related degree isn’t up to understanding stats????

    Like I’ve said at least twice???

    Good grief, what is the world coming too? Next you’ll be suggesting that I’m not a lawyer even though I talk about the Omnibus hearings

    You don’t have to be sorry Siggy. As far as I see it, you don’t need any training to be able to see that 0 + 0 is 0.

  51. passionlessDrone on June 27th, 2007 17:07:01

    Hi Preston –
    “Assuming you weigh about 200lbs, are male, and consume the beverages on both days over the course of about 4 hours, and that you define “drunker” as the difference in blood alcohol level, the answer is probably closer to 300%.

    Although day two’s blood alcohol level is 300% of day one, the difference between the two is about .23%”

    My point is that simply subtracting two percentages and evaluating the difference can lead to misleading ‘conclusions’. Clearly you understood this, yet for whatever reason you have introduced time lines and blood levels. What if I drink all one hundred ounces of beer / wine at once, and instead of drunkedness we measure the number of molecules of alcohol my liver processes?

    “Real physiological and behavioral effects in terms of “autistic” as a direct result of vaccination have not been scientifically shown to exist (here’s the part where all the idiots say, “yeah, but they haven’t been shown to not exist”).”

    Behaviorally, you are correct. I’m curious, could you describe for me an ethical experiment wherein the ‘real behavioral effects in terms of autistics as a direct result of vaccinations’ could be shown to exist? Considering that vaccinations are given starting the day a child is born, and autism is nearly never diagnosed until the child is at least 2, how could such an expirement be implemented? Surely you wouldn’t demand expiremental proof of something that would be impossible to obtain, yet I cannot determine how such an expirement could be carried out. If you would like to argue that it is impossible to perform such a study, how can you simultaneously argue that vaccines have been proven not to be associated with the behaviors associated with autism?

    As far as physiological effects, I’m not sure you are correct. It would seem that neonatal infection by ruebella, and possibly other agents has been shown to increase incidence of autism, has it not? Has not encephalitis at birth been shown to be associated with greater rates of autism? Are there not reported instances of viral encephalitis and/or meningitis later in life resulting in behaviors consistent with autism? Some vaccinations are known to carry with them the side effect of acute encephalitis, do they not? Admittedly, these side effects are reported very rarely, but this is for acute issues, as opposed to sub acute. We now know that autistics show signs of brain inflamation at levels that were not described as encephalitis; the people in the study certainly were not hospitalized, or killed by the inflammation. While I understand that there are different mechanisms of encephalitis, and those described in autism thus far do not correspond exactly to known conditions such as MS, issues such as post mortem collection and the possible differences between initiating processes and those observed during chronic periods cannot rule this possibility out. As far as other physiological issues, most of them simply have not been studied.

    Imagine if everyone smoked Marlboros, Parliments, Kools, and Camels. Some people believed this might be responsible for lung cancer in some cases. Several expirements are performed. In some, people are no longer given Kools; yet their incidence of cancer is seen as contsant. In some expirements, the tar is removed from all cigarettes, yet again, the incidence of cancer does not move. Would these experiments exhonerate cigarettes from cancer? Yet this is an apt analogy towards what is repeatedly declared as the great body of evidence proving that there is no association between vaccinations and autism or other disorders. Why?

    -pD

  52. notmercury on June 27th, 2007 19:38:42

    pD: “I’m not sure you are correct. It would seem that neonatal infection by ruebella, and possibly other agents has been shown to increase incidence of autism, has it not?”

    I’m not aware of any case reports of neonatal infection resulting in autistic like behaviors. Maternal exposure to rubella and things like valproic acid are associated with autism but that’s generally prenatal.

    Of course there is Heller’s Syndrome, first described in 1908, which, conceivably might be exacerbated by immune stimulation including like vaccination, but there is probably a pre-existing and underlying infection to blame.

    There are reports of older people who develop autistic-like symptoms following herpes encephalitis but those would be considered atypical.

    pD: Yet this is an apt analogy towards what is repeatedly declared as the great body of evidence proving that there is no association between vaccinations and autism or other disorders. Why?

    Because groups in the US have made it about the single ingredient thimerosal while groups in the UK have tried to make it all about measles. And, it’s not that there is a great body of evidence proving a lack of association, it’s that there is an extremely small body of evidence to confirm an association.

  53. passionlessDrone on June 27th, 2007 20:31:23

    Hi notmercury –
    “I’m not aware of any case reports of neonatal infection resulting in autistic like behaviors. Maternal exposure to rubella and things like valproic acid are associated with autism but that’s generally prenatal.”

    My terminology was incorrect. I was referring to the study associating maternal exposure to ruebella. None the less, exposure to a virus results in an increase in autism; both in utero and in some rare instances, to older individuals. Even though the exact mechanism of action is not known (though encephalitis certainly would seem one logical choice), the association has survived the chi squared analysis. It seems only logical to assume that infants, with reduced immune capabilities, would be at risk for the same experience.

    (I was not aware of the valporic acid link, very interesting.)

    “Because groups in the US have made it about the single ingredient thimerosal while groups in the UK have tried to make it all about measles.”

    I would tend to agree with the statement that singling out individual ingredients and/or agents as causes of such a heterogeneous condition as autism is misguided.

    “And, it’s not that there is a great body of evidence proving a lack of association, it’s that there is an extremely small body of evidence to confirm an association.”

    It would seem that we are in agreement in one thing, namely that there has been very little research in either direction questioning the association between the total load of the vaccination schedule and development of neurological disorders.

    While not a damning indictment, it is also, hardly a ringing endorsement. What’s more, this is quite a different spin than is usually assigned to the argument of vaccination side effects. How refreshing.

    Take care.

    – pD

  54. Prometheus on June 28th, 2007 00:27:29

    I ran the numbers from the GR survey and a few odd things popped out at me:

    The prevalence of autism in the entire group (all 17,674 kids) was 4.7%.

    That’s right 4.7%!That would be 1 child in 21!

    The latest report from the CDC gave a prevalence of about 0.65% (1 in 154).

    The ADD/ADHD number were also suspect, coming in at 10.6%.

    That’s right, 10.6%! That would be 1 child in 9!

    Clearly, even without looking at their methods, the numbers coming out of this survey are simply garbage.

    You’d think they would have checked this out before paying for a full-page ad.

    For more details, see my blog.

    Prometheus

  55. Cathy on June 28th, 2007 03:52:55

    All the information about the study can be found here http://www.generationrescue.org/survey.html
    including the data in PDF form which I had no trouble accessing, the organisiation that ran the telephone survey and the counties that were surveyed.

    I agree there needs to be more information. What were the reasons for not vaccinating? (eg medical contraindications or safety concerns? this could effect dx’s) Why did the partially vaccinated begin to vaccinate and then stop? (bad side effects perhaps? which would explain the higher rate of dx’s in partially vaccinated)

    And What I also want to know is how do they classify partially-vaccinated? That could mean anywhere between one vaccine and how ever many are on the vaccine schedule at this point in time. And new vaccines are always in the pipeline to be added, so fully vaxed kids today will be considered partially vaxed in 5 yrs time. The 17 year olds would classify as partially vaccinated compared to the 4 year olds … because the 17 year olds would not have had as many vaccines as kids do now. (check history of vaccine schedule 1989-2002 to verify facts) Did the phone survey have a checklist of vaccines? or did they simply ask if they were fully vaxed?

    And what makes you say partially vaccinated figures are biased?

    Maybe fully vaxed parents are biased because they obviously wholeheartedly support the vaccine programme, they would be less likely to want to admit their child is ASD.

    And 1% IS statistically significant. The difference between 1% and 2% would mean the difference between 100/10,000 and 200/10,000 … it is TWICE AS MUCH which is not 1% more than.

    For example; if there was 100/10,000 children with an ASD, 1% more than that would would 101/10,000.

    What is the population of America? What is 1% of that? And what is 2% of that? BIG DIFFERENCE!!

  56. Kev on June 28th, 2007 04:45:44

    “And what makes you say partially vaccinated figures are biased?”

    The very reasons you state above.

    “And 1% IS statistically significant.”

    Statistically significant? If this was a well designed study and the results were data then possibly. In this case? Not even close.

    “What is the population of America? What is 1% of that? And what is 2% of that? BIG DIFFERENCE!!”

    If I have 100 matchsticks, what is 1% of that? And what is 2% of that? TINY DIFFERENCE!!

  57. Ms. Clark on June 28th, 2007 05:13:34

    Why isn’t anyone freaking out that Handley found such a high number of autistic kids? And why are these kids more likely to have “autism” than PDD, nos, it should be twice as many PDD,nos as autism, if they were tapping into a random sample.

    Is Handley going to tell everyone to vax their kids “fully” always because if they partially vax their kids will be more likely to be autistic? Will the Handley baby get all his or her vaccines now? Will Handley be too ashamed to explain his choices for vaccinating that baby? Will he vaccinate according to the CDC schedule or not at all? Vaxing half way is dangerous, obviously.

    Why? Why? Why did they find ANY autistic boys among the unvaxed? The experts on EoHarm Yahoo! group say that it’s not possible to find autism in unvaccinated kids. Shouldn’t the rate be below the predicted level (0.66%) Autism among unvaxed boys is more double the expected rate.

    Handley’s survey numbers would translate to 200 per 10,000 of unvaxed boys in the U.S.
    The rate should be something like 25 per 10,000 if you believe that autism exists in kids with Frag X and so forth.

    Something is VERY wrong here!!

  58. Cathy on June 28th, 2007 05:58:13

    Its not a tiny difference, its twice as many.

    We are not talking about 100 kids here! we are talking about .66% of the population if you go by the 1:150 ratio.

    By “biased” I thought you meant that parents would be biased in one way or another in their answers to the survey.

    The problems I stated related to the discrepancies between the different “amount” of vaccines given that would be classfied as “fully” vaccinated, I dont call that a bias but it would account for the odd results … I think for simplicity sake it would be better to compare vaccinated (at all) to unvaccinated. This would be for a general overview of the trends, not “specific vaccine” causes “specific dissorder” type information. But simply “what are the health outcomes for vaccinated vs unvaccinated children”?

    And your math is all wrong. 2% is not 1% more than 1%. There is a difference of 1%. But 1% of 1% is 0.01%, so 1% more than 1% is 1.01%. 2% is 100% more than 1%.

    When you are talking about a percentage “more than” something, you have to find out “what percentage the difference between them” is of the original amount.

    For example 6 is 20% more than 5.

  59. HN on June 28th, 2007 06:04:01

    The survey was bogus, and has absolutely no basis in reality.

    For another take on it check out:
    http://photoninthedarkness.blo.....thing.html

  60. Kev on June 28th, 2007 07:53:04

    “Its not a tiny difference, its twice as many.”

    Or, one more.

    Cathy, I’ll say it again. If we were discussing anything remotely close to a decent survey then I’d happily agree with you. But we’re not. This survey claims a rate of ASD in its population of 1 in 21. Compare this to the actual population of 1 in 150. That tells you all you need to know about the validity of this debacle. As such, I feel free to take from it whatever I feel like – much like Generation Rescue are doing.

  61. 666sigma on June 28th, 2007 15:51:52

    Kev, this is NOT addition or subtraction. It’s statistics and statistical significance. However, I did find some of your basic math rather amusing as did a few others. Stick to web design.

    I did notice someone point out that the prevelance rates in this study are incredibly high. Now, that would call the study’s methods into question.

    Either way, GR did something I really have not seen elsewhere and that’s lay out its numbers for everyone to see. Even if the study is flawed, I think they won the debate.

    It’s time for the CDC to step up to the plate and do a real study. Better yet, we need a real independent study based on more solid ground than the GR study. What has passed for science on this issue (both sides) is a joke. The CDC is still hiding and where there’s smoke, there’s usually fire.

  62. Kev on June 28th, 2007 16:00:18

    “Kev, this is NOT addition or subtraction. It’s statistics and statistical significance.”

    Seriously – its really not. If it was good data, yes. This? This is just shit.

    But go ahead and explain to me how 0 compared to 0 comes out as statistically significant…?

    “Even if the study is flawed, I think they won the debate.”

    What debate?

  63. notmercury on June 28th, 2007 17:27:00

    “Even if the study is flawed, I think they won the debate.”

    What study?

  64. Steve D on June 28th, 2007 18:31:16

    I just completed an informal survey of my own.
    I have 14 coworkers (who are American-born), so the total staff is 15 including myself. The combined number of children parented by myself and coworkers is 15. All of the adults and children are fully vaccinated, except 2 relatively newly born who have not yet reached an age where the vaccine schedule can be fulfilled. So this leaves a sample size of 28. Of the 28, one child has autism, no other Neurological Disorders are reported (except clearly everyone here except me suffers from Narcissistic Disorder).
    Therefore, the incidence of vaccine-induced autism in my workplace is 1 in 28, or 3.5%! This is horrible! But since this is so much higher than the natioanl average, it must be something else in my workplace. Let’s see – we ship fresh flowers. That’s it! Exposure to beautiful, freshly picked flowers causes autism. Damn, now I have to find a new industry to work in…

  65. Interverbal on June 28th, 2007 18:50:51

    666sigma,

    “Either way, GR did something I really have not seen elsewhere and that’s lay out its numbers for everyone to see. Even if the study is flawed, I think they won the debate.”

    The question was never “can we do this study”, but instead was “can we do it well”. Keep that in mind and review the follwojng facts:

    This study is not peer reviewed.

    This study did not control for any of the six threats to random and systematic statistical error.

    As reported by Prometheus, the data do not show a statistically singificant result as assesed by Chi-Square, one-tailed, or two-tailed tests, when at the 99% confidence interval.

    If the null-hypothesis is that there is no difference in terms of autism, between kids who get vaccines and those who don’t, then we must accept the null-hypothesis. That is simply what the tests show.

    So, how can GR have won any debate here? Will you explain this?

  66. Brian Deer on June 28th, 2007 19:18:19

    They won the debate because there are some people who will have erroneous beliefs reinforced, and some who will be misled. That’s enough for this so-called “Generation Resue”. Any honest organisation would know that they were operating way outside their competence in presenting this “survey”. But they just don’t care. As with the vaccine court and the total absence of discussion about it.

    They just don’t care.

  67. caseofthevapours on June 28th, 2007 19:55:53

    The CDC is still hiding and where there’s smoke, there’s usually fire

    Or sometimes simply dry ice.

  68. David N. Andrews M. Ed. (Distinction) on June 28th, 2007 20:10:32

    “And 1% IS statistically significant. The difference between 1% and 2% would mean the difference between 100/10,000 and 200/10,000 … it is TWICE AS MUCH which is not 1% more than.”

    Wanna name a test of significance that could demonstrate that in regard to THIS so-called study?!

  69. Ms. Clark on June 28th, 2007 21:49:06

    Handley will probably need to take responsiblity for real world babies not getting vaccinated because of his “survey”. Real world babies may die or end up severely disabled because of his “survey”. But probably no one will come and lay the coming deaths and suffering on his doorstep, which is unfortunate, because he’s getting another 15 minutes of fame for his 200K, but he probably won’t get the half hour of infamy he’ll deserve if babies die because parents are frightened off of vaccines because of him and his antivax friends, like Kirby and Katie Wright.

    It’s just the same as few people realize that Andrew Wakefiled (or “Andy” as the Omnibu PSC lawyer called him) is personally accountable for the deaths and non-fatal infections of people by measles.

  70. Cathy on June 29th, 2007 05:29:51

    Ok, so because they’re wrong its ok for you to be wrong to, fair enough. The numbers they got back might be biased for a number of reasons … like how many hang ups did they get? Quite possibly it would be the parents of kids with problems who would most want to participate in the survey. Which would account for the high numbers.

    But whatever the reason, the numbers they got were the numbers they got!! And their interpretation of them is correct … your interpretation is incorrect.

  71. Kev on June 29th, 2007 05:35:36

    Cathy, please try and hear me. I am not attempting to interpret them. That word ‘interpretation’ implies there is something of value there to interpret. There is not.

    I am simply taking their numbers at face value and doing exactly what they are. Choosing which bits of it suit me.

  72. Prometheus on June 29th, 2007 05:50:08

    Cathy,

    To be fair, while Kev’s interpretation of the survey is … novel, GR is also playing fast and loose with the facts.

    One of the things that EVERY scientifically valid survey has to do is prove its validity – prove that it is an accurate reflection of reality.

    After all, the reason we DO surveys and random samples is so that we can determine something (or a number of things) about a population without having to check every member of that population.

    To find out if the sample is an accurate reflection of the population as a whole, we check certain data against what is already known about that population – things like the percentage of males and females – and see if they match what we know about the population.

    When the aggregate autism prevalence (the percentage of the whole sample that had autism/PDD/Asperger’s) exceeded what LARGER surveys have found, and not by a little – nearly TEN times – then we can only conclude that this survey’s sample does not reflect the larger population.

    Since we don’t know what led to this sampling error – and neither do GR or the company they hired to do the survey – there is no way to extract ANY meaningful information from the “data”.

    Now, they can (and will) spin this ten ways from Sunday, but the reality is that their results are meaningless. Their attempts to make some political hay from this meaningless pile of numbers is simply more evidence that they are far more interested in “winning” than in finding the truth.

    Prometheus

  73. passionlessDrone on June 29th, 2007 12:11:26

    Hi Promotheus –
    Granting that the rate discrepancies in the study are very likely bogus; I’m curious on your take as to the underlying question of whether or not there is a large enough population of unvaccinated children, that can be sampled without bias, to conduct a comparison of vaccinated versus unvaccinated children that has statistical power?

    It seems this issue, like many in the autism arena, is frequently treated to broad brush strokes by both sides, when a more nuanced evaluation is necessary to truly understand the issue.

    By way of example, the example of the Amish is frequently held up as an example of a potential unvaccinated population. There are also reports that the Amish do vaccinate. I recently heard an interview with a person who spent some considerable time with an Amish family, (vaccination was not mentioned) and one thing I took from the interview was the fact that there are actually, many different stratifications within the Amish community. The particular sect he spent time with, for example, only showered on Saturdays, something which other Amish found a bit funny. My point being, referring to ‘The Amish’ as a uniform entity makes as much sense as ‘The Autistics’, and thus, it seems likely there are some Amish that do not vaccinate, and some that do.

    Of course, genetic similarities may introduce problems with using the Amish as a single population for which to perform such a study. Other areas, such as Chicago, and apparently, Oregon also, are annectodally valid areas for such participants. It also occurred to me that there is a growing population of children who are genetically diverse, but paradoxically, genetically more likely to have autism; namely the children of parents who have had one autistic child, and did not vaccinate subsequently. It would seem that one of the problems with this population would be selection bias; such parents are likely to believe in a vaccination relationship, and they may be unlikely to enroll their unvaccinated, but autistic children in such a study.

    Anyways, several posters on this board have understandings of the power statistical studies, and potential limiting factors in such studies, than I do. Given that, I’d be interested in knowing the potential problems they would see in identifying a pool of candidates for a true unvaccinated / vaccinated population study.

    Take care.

    – pD

  74. Tom on June 29th, 2007 14:38:00

    The question of whether an unvaccinated vs. vaccinated study COULD be done seems to beg the question as to whether it is scientifically JUSTIFIED.

    The only people demanding this study are those who lack any scientific understanding of the issue. Scientists in the vaccine and immunology disciplines sure don’t seem concerned. The prestigous Institute for Vaccine Safety at Johns Hopkins is not concerned. Paul Ofitt is on record as saying that based on research, an infant could receive 100,000 simultaneous vaccinations. While that may be hyperbolic, the point is that no one in the scientific community is expressing concern.

    In a review article, Pardo and Zimmerman, who authored the post-mortem studies on autistic brain tissue, state that their findings are not consistent with infections. They say they don’t see any evidence of immune suppression and infection leading to the microglia activation and neuroinflammation they found. Kids with atuism aren’t more susceptible to infection. The investigators are not sure whether the response is a helpful, neurotrophic one or even when it began (prenatal, post natal).

    More recent evidence by Pardo’s lab implicates terbutaline, a drug given to prevent stop preterm labor. They’ve even built an animal model using the drug. Zimmerman’s lab finds evidence of maternal-fetal immune interaction. If I remember correctly, Zimmerman was slated to be a witness in for the govt. in the omnibus case. It would seem that speaks very loudly about vaccine concerns.

    If the anti-vax lobby succeeds in undermining vaccine uptake, it may become necessary as a political rather than scientific study.

    And if such a study is undertaken, good luck with that. You’ll have to find all those unvaccinated kids and it won’t be easy. I don’t believe the GR unvaccinated numbers.

    It seems as though you are looking for a single association among a group of people who live off the grid in more ways than one. Are they going to want to participate in a government study? Clinical trial recruitment is very difficult and costly. This study would be very expensive.

    Unlike Dan Olmsred, I have actually bothered to talk to a doctor who sees Amish children at the Clinic for Special Children in Lancaster county PA. The doc reports that the vaccination rate is 70% in Lancaster County and garden variety autism is present. In IL, the vaccination rate is 90% and again autism is present. It is even listed in the diagnostic pages on the IL clinic’s Web site.

    Despite what the anti-vaxers believe, autism is predominately genetic, so you’ll have to account for all those unvaccinated kids with de novo CNVs and other known genes. And then you still have to diagnose for autism as most unvaccinated children likely go undiagnosed.

    I for one hope that reasonable people will allow science to pursue leads and quit pretending that they know better.
    There are interesting studies finding associations. These findings should be the subject of dialog about the future of autism research, not a random event that a bunch of ignorant yet arrogant lay people have made.

    If this study becomes politically necessary, it will cost a fortune and the anti-vaxers won’t be happy with the results. They will resort to their tired old conspiracy theories.

  75. Brian Deer on June 29th, 2007 15:40:10

    Autism makes fools out of experts, and experts out of fools.

  76. passionlessDrone on June 29th, 2007 17:04:28

    Hi Tom –
    The only people demanding this study are those who lack any scientific understanding of the issue. Scientists in the vaccine and immunology disciplines sure don’t seem concerned. The prestigous Institute for Vaccine Safety at Johns Hopkins is not concerned. Paul Ofitt is on record as saying that based on research, an infant could receive 100,000 simultaneous vaccinations. While that may be hyperbolic, the point is that no one in the scientific community is expressing concern

    If I am not mistaken, Mr, Offit is the owner of patents on some vaccines, as well as en employee of Merck. Some people might consider that as a source of information, he has a significant conflict of interest. For example, if I stood to gain from vaccine legislation and said the opposite; wouldn’t this be viewed as a conflict of interest?

    In a review article….

    Do you have a link to this article? I simply read on their website what I stated, that an infection at intiation of disease time could not be ruled out for a variety of reasons.

    Kids with atuism aren’t more susceptible to infection.

    This is a curious statement. I may be mistaken, but I was under the impression that the c4b allele identified as having greater prevalance in autistics would result in decreased functioning of the complement system, which in turn, is responsible for pathogen removal. Further, a great many studies do show that the immune systems of autistics are imbalanced when compared to controls, do they not? If you would like to argue that disruptions to the immune and complement system would not lead to increased succeptibility to infections, please do.

    You may be referring to studies based on hospitalization records. (?) If my son is sick, but I do not take him to the hospital, does this mean he was not sick? If he was sick on vacation and taken to a different hospital, does this mean he was not sick? If my autistic son is sick for four days, and my non autistic daughter is sick with the same bug for 3 days, does this not mean my son with autism was sick for longer? Yet, this type of information is not reflected in this kind of study.

    More recent evidence by Pardo’s lab implicates terbutaline, a drug given to prevent stop preterm labor. They’ve even built an animal model using the drug.

    Considering you seem to think that autism is predominantly genetic pointing this out would seem to be a strange argument. It is interesting none the less.

    If the anti-vax lobby succeeds in undermining vaccine uptake, it may become necessary as a political rather than scientific study.

    Increasing vaccine uptake would seem to be a very good reason to perform the study, regardless of if you feel it is scientifically justified or not.

    It seems as though you are looking for a single association among a group of people who live off the grid in more ways than one. Are they going to want to participate in a government study? Clinical trial recruitment is very difficult and costly. This study would be very expensive.

    If living off the grid showed an association with autism prevalance, this would be a good thing to know, would it not? As far as cost goes, this country spends a few billion a week in Iraq; clearly there is money to go around. If the end result, as you feel it would be, would be surefire evidence of the vaccination schedule safety and the presumed uptake in vaccinations, wouldn’t it be worth it?

    You’ll have to find all those unvaccinated kids and it won’t be easy. I don’t believe the GR unvaccinated numbers.

    Specifics on the complications of such an undertaking was exactly what my post was asking, yet for whatever reason, you have chosen not to address it meaningfully. Go figure.

    doc reports that the vaccination rate is 70% in Lancaster County and garden variety autism is present.

    My question could then be reframed as, ‘If some counties report 30% rates of no vaccination, would this be a sufficient starting point for a population based study?’

    Despite what the anti-vaxers believe, autism is predominately genetic, so you’ll have to account for all those unvaccinated kids with de novo CNVs and other known genes.

    Hm. I am certainly not arguing against a genetic component; however several of the studies that I have read indicate that having particular genes does not raise your likelyhood of getting autism all that much. By way of example, the MET chromosome recently found by Levit increases your risk of autism by 2.27. Here are some quotes from the studies author:

    “This is a relatively common variant, seen in about 47 percent of the population,” Levitt said. “So why doesn’t everybody have autism?”

    That speaks to the environmental and other genetic contributions, Levitt said.

    “Genes create a vulnerability that then gets coupled with some environmental disturbance – but right now, we don’t have any idea what those factors might be.”

    Link

    If you have a better understanding of the level of succeptibility that genes provide than the author of the MET study, your talents are being wasted in the blogosphere.

    If this study becomes politically necessary, it will cost a fortune and the anti-vaxers won’t be happy with the results. They will resort to their tired old conspiracy theories.

    This may be true. However, if the result in increased vaccination rates, wouldn’t it be worth it?

    Take care.

    -pD

  77. Tom on June 29th, 2007 18:50:18

    PD,
    In no particular order:

    The review by Pardo, Vargas and Zimmerman is Immunity, neuroglia and neuroinflammation in autism. Int Rev Psychiatry. 2005 Dec;17(6):485-95

    Autistic kids being no more susceptible to infection is from the literature. See the Pardo Vargas Zimmerman review and below.

    From Zimmerman’s Web site: “However, there is no evidence, as yet, that children with autism have increased susceptibility to infections, or that specific therapies for the immune system can alter their symptoms.

    Dr. Zimmerman and colleagues recently found that rheumatoid arthritis and other autoimmune disorders are more common than expected in the families of children with autism. This leads to speculation that autoimmune disorders might be a sign of genetic susceptibility to Autism. Such a predisposition may act through genes associated with the human lymphocyte antigens, which commonly have specific associations with autoimmune disorders. These genetic effects most likely begin before birth and might be modified by the mother’s, as well as the father’s, genes. This may lead to disruption in normal development of the immune, as well as the nervous, systems in the fetus.”

    As for Offit, the “Big Pharma” card infers that he has been bought and sold. I just don’t buy that.

    Pointing out potential environmental risk factors in autism does not undermine the scientific data that indicates autism is predominately genetic. There are already purely environmental causes known.

    As for genetics, your discussion of MET is but one gene variant. There may well be an environmental trigger for that gene. Or there may be other modifiying genes that protect against MET. MET may be variably expressed in those without autism.

    But the point isn’t about MET. There are lots of genes and de novo CNVs yet to be discovered. The point is that people studying autism tell us that it is predominately genetic. So PD, if you know better, go argue with them. Or YOUR talents are being wasted in the blogosphere.

    Lastly, I never chose to ignore your question about undertaking the study you seem so interested in. That’s your issue not mine. Besides, you asked Prometheus not me.

    Funding decisions are made on the basis of good preliminary data. In this area, the only people suggesting vax vs. unvax has merit are are anti-vaxers who seem to somehow know better than scientists.

    Sometimes, studies are conducted purely due to political decisions. Laetrile in cancer comes to mind. I worked at the NCI when hydrazine sulfate was being pushed. Money was wasted to conduct lab studies.

    It’s a shame that with science spending so tight, truly meritorious science
    will go unfunded to instead go on wild goose chases.

  78. daedalus2u on June 29th, 2007 22:15:25

    PD, the Offit paper is
    http://pediatrics.aappublicati...../109/1/124
    The point (and it is an extremely important point) is that antigens are antigens, and whether you get an antigen from a vaccine, or from exposure to bacteria, viruses, food, spores, pollon, cooties, pets, dirt, etc, if newborns didn’t have an immune system that could respond to tens of thousands of antigens, they would die from massive infection in a few weeks.

    The vast majority of exposure to bacteria is non-harmful. The reason (and the only reason) it is non-harmful is because the immune system can respond to those bacteria before the bacteria can kill the host or cause disease. It is only for a relativly few bacteria and viruses which can kill the host before a sufficient immune response develops that immunization is useful.

  79. notmercury on June 29th, 2007 23:55:34

    pD: “I may be mistaken, but I was under the impression that the c4b allele identified as having greater prevalance in autistics would result in decreased functioning of the complement system”

    Decreased functioning of the complement system would probably not result in increased complement assembly and deposition as reported in some studies.

    If you have anything other than your own chain of logic to indicate an increased susceptibility to infections, I’d be interested.

    I’m not really sure if you are arguing for immune suppression or neuro-inflammation but most autistic children are not immune deficient.

  80. JScarry on June 30th, 2007 20:14:52

    Many of you seem to be confused about the statistical meaning of “bias’ and significance-in addition to making stupid mistakes with arithmetic.
    Bias does not mean that someone has a preconceived notion of the result. In this context it means that the data may be unreliable because of who choose to answer the survey. Real statisticians go to a lot of trouble to make sure that the people answering a survey match the characteristics of the underlying population. If a survey is biased, there generally isn’t anything that you can do to fix the bias so the results are garbage.
    Significance is a statistical test. All it means that there is a 95% chance that the result is not zero. You can have a result that is statistically significant but the effect is small. I think that a lot of the lifestyle results that appear in the popular press are statistically significant, but the effect on extending lifetimes is minimal. The poster who thinks that 1% is not significant is confused about what the 1% represents. When statisticians say that 1% is not significant, they mean that they can’t say that there is an effect at all. If there was an effect then yes, 1% of a population is a large number but they are saying that there is really no effect.
    And Kev, you really should take a statistics course or read some books. Your comments are even worse than the original post.

  81. Interverbal on June 30th, 2007 20:39:19

    JScarry,

    “Significance is a statistical test. All it means that there is a 95% chance that the result is not zero.”

    If alpha is at .05, then that’s true, but 99% and alpha at .01, is also very common.

    “Bias does not mean that someone has a preconceived notion of the result.”

    I think most of us already know this.

    “Real statisticians go to a lot of trouble to make sure that the people answering a survey match the characteristics of the underlying population. If a survey is biased, there generally isn’t anything that you can do to fix the bias so the results are garbage.”

    Given that these data do not match in terms of prevalence, or in Autistic Disorder to PDD-NOS, what should we conclude in this case, based on your statement?

    “I think that a lot of the lifestyle results that appear in the popular press are statistically significant, but the effect on extending lifetimes is minimal.”

    That’s a well known one; and research in treatments for severe injury, is another example of this problem.

    “The poster who thinks that 1% is not significant is confused about what the 1% represents.”

    I am not too sure that 1% is statistically significant in this case. Am I also confused?

  82. Kev on July 1st, 2007 00:07:39

    “And Kev, you really should take a statistics course or read some books. Your comments are even worse than the original post.”

    Once more, for the hard of comprehension I guess.

    This survey is shit. The results of it are shit. My analysis of the results are shit. Do you see a pattern here?

    Or, lets put it another way. Do you think all those antivaxxers are at all bothered about ‘confidence’ or ‘ratio’ or even ‘statistics’? No. They read into this what they want to. Well, so have I.

  83. anonimouse on July 1st, 2007 17:54:21

    Brad Handley is lying, and he’s doing it deliberately.

    Brad purportedly has a degree in economics from Stanford. If you get a degree in economics from Stanford, you’ve taken classes in statistics.

    So unless Handley is stupid and/or didn’t listen in school (a distinct possibilty), he knows how to perform an appropriate statistical analysis of this data. He knows that his data is shit, and it doesn’t say what he says it does.

    Yet he lies anyway.

    He lies because he knows his case is dead, that the theory he staked his personal reputation on is dead, and that the only people that think he has any credibility are the small cadre of delusional minions who think that there is any link between thimerosal, MMR, and vaccines.

    Brad, rather than continuing this farce, go back to doing what you do best, which is do a mediocre job on the boards of second-tier companies. Leave science to people who know what they’re talking about.

  84. JScarry on July 4th, 2007 02:12:16

    Interverbal,
    You could use a 99% confidence interval if you want. That makes false positives less likely. What is more common in situations like this is to see 90% confidence intervals because the study doesn’t really show anything but maybe by stretching the definition of significant they can get something they like.

    “what should we conclude in this case, based on your statement?” You should conclude that spending any time on analyzing the results is a waste of time. The data collection method is so seriously flawed that it is a waste of time to do any analysis of the results.

    Kev, I fail to see how lying helps your cause any. You seem to be arguing that because the anti-vaxers lie it is OK for you to lie too. It’s not. There are no circumstances where is is acceptable. If you are willing to lie about the numbers here, what else are you willing to lie about? You are ignoring the evidence in front of you and insisting that you are right. That’s not how science is done and that’s not how rational people behave.

  85. Kev on July 4th, 2007 07:01:56

    Lying? Hardly. Maybe you could grab a little perspective whilst you’re up there on your high horse amigo ;o)

    I’ll give you one more clue as you still seem to require some education: Have a look at the aggregated 11 – 17 age bracket. Aside from Aspergers Syndrome (which a lot of GR members including Brad don’t think is ‘real’ autism), could you tell me the statistical differences between non vaccinated and fully vaccinated kids?

  86. Interverbal on July 5th, 2007 04:32:53

    “You should conclude that spending any time on analyzing the results is a waste of time.”

    Some of us, do take the time anyway. This is because we can show that even if one does accept the data collection method to be well controlled, the stats still do not show what they are advertised to show.

    I do these sorts of analyses more for the pople who may not agree with us that the analysis is junk. In this case, I can have my cake and eat it too.

    re: the 90% CI. I do not know about other forms of demographics, but I almost always see a 95% CIs in autism epidemiology, with maybe a very occasional 99%.

  87. Cathy on July 14th, 2007 07:56:03

    Are you aware of the new figures coming out of the UK?

    “”On July 8, 2007, researchers from Cambridge University’s Autism Research Center in London released a report estimating that one in every 58 children (not just boys) in the U.K suffers from “some form of autism disorder” compared to previous estimates of one in 100.[1] The reality of that statistic should make one gasp: Nearly two percent of the citizens of the U.K. will become mentally handicapped adults.””

    So perhaps it is the CDC phone survey results that are flawed? Their surveys cover only non-institutionalised cases, could this make a difference?

  88. Kev on July 14th, 2007 08:04:26

    Are you aware that the 1 in 58 figure is rubbish?