Can the MMR or other vac ever cause autism?

[StarlightMcKenzie]

www.ncbi.nlm.nih.gov/m/pubmed/25114790/

Catherina, are you happy with that estimate of 22 black autistic children?

Noblegiraffe, Hooker takes exactly the same data as Destafano. This is because he uses the data from Destefano. It is the same data. The numbers are 624 case children and 1824 controls.

Hooker did the same as Destefano and examined the numbers according to subgroup criteria, and the numbers as a whole. Both studies found the same thing - an association in the timing of MMR administration and the likelihood of an autism diagnosis in African American boys. DeStefano decided that this was due to "health seeking behaviour" and they dismissed the effect. They also adjusted their inclusion criteria in order to dilute the statistical significance.

Hooker has reported the effect and suggested that it merits further investigation (by which one would imagine would involve bigger numbers). He also highlights the fact that other research has suggested that autism is more common in African Americans than it is in other populations.

Of course this merits further investigation - and one of the reasons it does is because the numbers in the original study are quite small for an epidemiological study. The scientifically honest thing to do is run a bigger study - not tout your study as being solid evidence to the safety of MMR for all populations. Or use such a study as solid evidence in court cases. Or adjust your inclusion criteria in order to dilute the strength of a potential phenomenon.

The problem here is not the numbers, but what was done with them. In the original study, scientific curiosity should have been piqued and they should have examined the subgroup in bigger numbers. I mean that's what the CDC should have done if they were serious about autism research. There was a potential clue in the Destefano data - and they buried it rather than further examine it. They then used this shaky study to declare that MMR safe for all, which is utterly inexplicable (or should be to the intellectually honest).

The whole point of the MMR controversy is that some scientists think that SUBGROUPS may be vulnerable. Here a subgroup may have been identified. But if you are going to claim that the numbers are too small to have identified a subgroup, you must, in all honesty also say that the numbers are too small to have ruled out any subgroups. And yet MMR defenders have used Destefano to do exactly that . For years.

Plus it is scientifically very dubious to draw the conclusion that these kids had had MMR because they were autistic/accessing special healthcare rather than considering that the relationship could equally be the other way round. In fact one could say that it isn't scientific at all - it is political.

Hooker has reported the effect and suggested that it merits further investigation

That's not what I've seen. That's what you might hope for. Instead what I've seen is Hooker has reported this and tried to whip up media hysteria, an appalling video with him and Wakefield where the CDC were called all sorts of names, worse than Hitler etc. Where, as I mentioned, Hooker said Thompson was culpable for the permanent damage of a significant proportion of the population. I've read that Hooker recorded conversations with Thompson without his permission - dishonest and possibly illegal. I've read that Thompson was named as whistleblower despite expecting anonymity.

Hooker is not some scientist simply calmly crunching the numbers is he?

The legality of recording conversations depends on which state they're in.

Noble, you seem to be criticising the sample size, do you know how that figure quoted earlier was calculated? Did you read the link?

I'm talking about the actual science. I'm talking about what Hooker concluded in his paper. If you want to talk about politics then I think we also need to examine how the CDC, the IOM, the DoJ and other institutions have conducted themselves...

The science is that Wakefield et al have been saying for years that they suspect a subgroup effect and a temporal affect. Destefano stumbled upon exactly that - a subgroup with a temporal association. And the chose to bury it and explain it away with an interpretation that they gave no scientific basis for.

I agree with you that the video is hyperbolic. However, if Wakefield and Hooker are correct the CDC is responsible for a massive public health scandal that has caused untold damage to children and their families. They would also be responsible for standing in the way of scientific progress and understanding of a condition which causes terrible suffering and distress to a great many people. They may have hampered understanding of potential treatments or ways of identifying and protecting vulnerable populations. And that would be unforgivable, so I understand the anger of Hooker who has a son with autism and Wakefield who sees countless numbers of distressed families seeking medical help for their devastated children.

Hooker did the same as Destefano and examined the numbers according to subgroup criteria, and the numbers as a whole. Both studies found the same thing - an association in the timing of MMR administration and the likelihood of an autism diagnosis in African American boys. DeStefano decided that this was due to "health seeking behaviour" and they dismissed the effect. They also adjusted their inclusion criteria in order to dilute the statistical significance.

Hooker apparently used a different and less strong statistical technique. The health seeking behaviour comment was (I believe, not having access to the paper), not restricted to black boys.
From what I understand, the birth certificate subgroup gave access to more data so a more detailed analysis could be done. When more details were brought in it showed that being a black boy wasn't as important as thought - it wasn't a race thing, other stuff was coming in that was associated with being a black boy which explained the original figures. That's not covering up, that's fine-tuning your analysis.

mrs let's hope he was in a state where it was only morally dubious then.

No, I don't know about the accuracy of the 22 figure. Do you know why not? Because Hooker didn't include sample sizes in his report.

Given that the first thing you should do when reading a report like this is check the sample size, it's completely bizarre that he didn't give them. Unless he didn't want people to know and be open and transparent about his analysis?

And none of us can really know what is going on between Hooker, Wakefield and Thompson.

What we do know is that Thompson chose to reveal information to Hooker.

Did he really expect Hooker to do nothing with something that is potentially scientific, political and medical ethics dynamite? To sit on information that affects the general public in deeply profound ways? I doubt it somehow. And if he did, I think it would be hugely narcissistic to the point of being sociopathic. Perhaps that what happens to you when you work for the CDC.

Thompson may well be saying that he didn't intend to go public under legal advice. The poor sod is up against the CDC for crying out loud.

Noblegiraffe, might I suggest that you actually read Hooker's paper rather than just what biased bloggers say about it?

www.ncbi.nlm.nih.gov/pmc/articles/PMC4128611/

So you're just assuming that the sample size is too small then? Is that because you want to be able to dismiss it? I would have hoped that people would approach studies with a more open mind but it seems that anything that suggests further investigation into the MMR is immediately met with resistance. Why is that? Doesn't it make sense to try to identify people who may react badly to certain vaccines?

I've read Hooker's paper, it was in the OP. Still can't find how many black boys there were.

And you might try to separate the science of the paper from the hysterical shitstorm, but when the person on the crusade is the guy who wrote the paper, 'conflict of interest' doesn't quite cover it.

So you're just assuming that the sample size is too small then?

No, I know there was an issue with the sample size because in the paper Hooker states he had to cut off a particular analysis at 31 months instead of 36 otherwise one of his groups would have fewer than 5 children in it. As I said upthread.

What issue are you talking about?

The Hooker paper says this:

A final analysis was completed on African American children in the “birth certificate” cohort, excluding low birth weight individuals (Table 4). Results were obtained at 18 months, 24 months and 31 months, rather than 36 months, as there were insufficient cases (less than 5) at the 36 month mark to carry out the analysis. Even at 31 months, there were insufficient cases of African American females. Thus, these results are not included in the analysis.

So, in your opinion, the fact that one particular analysis that excluded low birth weight had fewer than 5 cases at 36 months means that there are issues with the sample size overall? How exactly?

Because if they had issues with a group containing fewer than 5 children, and for girls the situation was even worse, then that doesn't fill me with confidence about the other groups. If they were short on kids at the upper age range, does it suggest to you that the other groups were full to bursting? It also means that they were prepared to do analysis on groups containing just 5 kids.

It would be nice to know, right? But for some reason the figures aren't there.

And the analysis which excludes low birthweight is important, because low birthweight is linked to autism, and it seemed low birthweight was also more common among African American children.

"This article has been removed from the public domain because of serious concerns about the validity of its conclusions. The journal and publisher believe that its continued availability may not be in the public interest. Definitive editorial action will be pending further investigation."

That should be the end of it, I hope.

"it seemed low birthweight was also more common among African American children."

Which is probably why, when they excluded low birth weight children, that they were left with insufficient cases.

The link between low birth weight and autism is probably what encouraged them to exclude that group to see if there was still an effect.

Their result:

Even excluding low birth weight individuals, a relationship was seen between first MMR age and autism incidence at 31 months for African American males (RR = 2.64, 95% CI: 1.08-6.46, p = 0.0280).

I think you are making assumptions about group size and you seem to want any reason to dismiss this. You were quite happy to accept the figure that Catherina gave without even looking into where it came from.

Which article was that Lweji?

Noblegiraffe, the reason I suggested that you read the study is because you said this Hooker apparently used a different and less strong statistical technique.

When Hooker states the following in "Statistical analyses";

"The Pearson’s chi -squared test contained in the SAS® software was utilized for current statistical analyses, and a two-sided p-value < 0.05 was considered statistically significant. This is in contrast to the original Destefano et al. [14] (CDC) study, where a case–control study design was used, where 3 control children were matched to each case child, and analyzed using conditional logistic regression dichotomized for the three age cut-offs at 18, 24 and 36 months. Pearson’s chi-squared is, in general, a more conservative analysis and therefore chosen for the present study. However, our results were also confirmed using a conditional logistic regression design similar to the Destefano et al."

Which led me to think that you hadn't read the paper, also I have seen various bloggers make the claim that Hooker used a 'weak' technique (although they don't actually seem to explain why the technique is weak).

To be honest I think all this scrabbling around for ways in which to smear/criticize Hooker is just diversion tactics. Thompson said that they found the same result - and that they buried it. So we have two papers, both using the same data and both finding a subgroup effect that they consider statistically significant. One lot of scientists decided (with no scientific basis) that they could explain away the statistically significant phenomenon by deciding that it was due to lifestyle. Another scientist thinks that that isn't good enough and that more actual science is needed rather than shaky unjustified untested hypothesizing. And yes, we can say that a weakness of both studies is the low numbers - that should lead us to demand a bigger study, NOT to demand that Hooker's work is ignored or to smear him (for using CDC data!).

Thompson says that they omitted data and changed their inclusion criteria in order to dilute their findings. This is a very very serious matter. But let's watch all the bloggers and shills tear Hooker apart for using the same data as DeStefano whilst they fail to question therefore the validity of DeStefano. If Hooker/DeStefano are too weak to identify a subgroup, they are also too weak to rule out subgroups. Therefore, logically, DeStefano is a white elephant of a study and should not have been used in the Omnibus Proceedings or as a reference for subsequent studies or to reassure the general public as to the safety of MMR for all.

Which article was that Lweji?

The article in the OP, that is the cause of this whole thread

My link to it still works. I downloaded the full article as a pdf as well.

The link between low birth weight and autism is probably what encouraged them to exclude that group to see if there was still an effect.

Obviously.

Which is probably why, when they excluded low birth weight children, that they were left with insufficient cases.

Indeed. And thus any statistical analysis run on such a small group should be treated with extreme caution (and not, for example, be used to compare people to Hitler).

I think you are making assumptions about group size and you seem to want any reason to dismiss this. You were quite happy to accept the figure that Catherina gave without even looking into where it came from.

I think that when I've referred to it, I've said something like 'if the figure upthread is correct'. Catherina gave a link which I've been reading, but it's massive and I've not come across the 22 figure yet, hence my caveat. Re assumptions about group size - there was a group of a few hundred children. They narrowed it down to African American children. Then it was narrowed further to African American boys. Then it was narrowed further to African American boys who weren't low birthweight. Then they said that a particular subgroup of that group had fewer than 5 children in it. Do you think they were working with a massive sample? And why didn't Hooker say how many, to validate his analysis if it was?

I'm not 100% sure on whether MMr or any vaccination could cause autism. I do know that no vaccination will be 100% safe and effective on every single person who receives it. Nothing is 100% safe and effective.

DS had his 1st MMR and several months later started slipping away from us, eventually after almost a year of tests diagnosed with autism at 3. We rejected the 2nd MMR as a precaution due to Wakefield and also rejected it for our younger children. I just couldn't vouch for its safeness.

DS has just turned 15 and after being almost completely dry for a year and developing so well at school, integrating himself with his peers he received another vaccination, either the Dip/tet/polio or men C. Well the change in him has been catastrophic. He wets himself and doesn't notice. He is back in pads fulltime. He is reluctant to do anything by himself. He needs constant supervision to do simple tasks. He has regressed so much it is heart breaking. The only thing we can relate it to is the injection and he is now going through investigation to find the cause. I suspect it will be to rule out vaccine damage.

Starlight Yes I do think vaccines could exacerbate autism

Noble, they excluded the groups that had insufficient cases. They did not run statistical analysis on them because there were too few cases and they said that themselves (see the italic text above). The smallest expected frequency should be 5 when using the Pearson Chi-Square test.

You are assuming that because one particular sample was small when low birthweight was excluded (after acknowledging that low birthweight was more common in African American children i.e. the group being studied) that the whole sample is too small.

Sorry to hear about your son Debs. That must be so hard for you after he was doing so well. :(