Correlation and causation

[ema-adama]

I know that there are some people who say that just because something is linked in time, does not mean that there is a cause and effect relationship. Dr. Sears wrote very powerfully in his Vaccine Book of a baby who came in for the vaccines and something delayed the adminisration of the vaccines and in that delay the baby had convulsions. He was struck by how if the vaccines had been administered minutes before, everyone would have been convinced that the vaccine had harmed the baby as the convulsion followed the vaccine.

So, I was wondering about how you can 'proove' if something is related just because two events are related in time.

I have had a really hard time trying to figure out what role vaccines have had in the decrease in disease that is seen in graphs. And different graphs that show that disease was already in decline before vaccines were introduced.

Is there more to the idea that vaccines created the decrease in disease other than correlation equals causation?

[Marnica]

Originally Posted by ema-adama

I have had a really hard time trying to figure out what role vaccines have had in the decrease in disease that is seen in graphs. And different graphs that show that disease was already in decline before vaccines were introduced.

Is there more to the idea that vaccines created the decrease in disease other than correlation equals causation?

wonderful point! my humble opinion is NO!

[carriebft]

Well, I could go on and on on this one. But just take your example of these graphs. Look at the measles vaccine-- HUGE drop once the vaccine is taken up, with the death numbers also indicating a huge drop. But we also can see how the vaccine prevents measles in large studies as well as by simply looking at the outbreaks happening-- almost exclusively under and non vaccinated people getting it.

Large epidemiological studies can also pinpoint causes or dismiss causes (though obviously the null hypothesis cannot be proven completely).

Another factor is that not all vaccine preventable diseases are diseases of environment-- of "the dirty"-- kwim? Rotavirus is so contagious that it rarely discriminates, but obviously it will be more common in areas with poor hygiene. But due to that highly contagious nature, even in developed countries it is very prevalent-- so it will be easier to see the vaccine's place in its reduction.

So, in short, I think for some vaccines the data is easy-- the question of how well they work is obvious. For others, it is not so obvious.


edit: forgot to add the evidence of immunity caused by the vaccines-- some are better than others at creating an immune response, so those will be easier to connect as the "cause" vs something like the influenza vaccine which is not so successful.

[mysticmomma]

I think she was talking more about vax reactions than vax efficacy.

[carriebft]

From the OP:

I have had a really hard time trying to figure out what role vaccines have had in the decrease in disease that is seen in graphs. And different graphs that show that disease was already in decline before vaccines were introduced.

Is there more to the idea that vaccines created the decrease in disease other than correlation equals causation?

[mysticmomma]

I should try reading the whole thing next time!

Can I blame it on the baby? It ate my brain.

[ema-adama]

OK - I was not clear enough.

I had in my mind the image of polio specifically. I think polio the disease and polio vaccination is central to the huge confidence in mass vaccination policy. And of course smallpox.

I would have to agree that disease like rota virus and hib are in another catergory, and not part of the graphs I was looking at to understand how vaccines became such an integral part of our culture that we accept them without a 2nd thought for the most part.

So, with the disease polio, I do not understand how anyone can proove one way or the other just what caused the huge spike in incidence and then the dramatic reduction in reporting. It seems to me that the correlation equals causation is the strongest aurgument that can be put forward for the reduction of the disease in the 50's.

And this is just passing musing for me, I really have not put much thought into it at all - it just struck me that there is the possibility that the very people who choose to ignore the risks of vaccines based on the correlation not equalling causation aurgument, may indeed be relying on just that aurgument to build the foundation for mass vaccination against disease.

[carriebft]

well, we have more than just vague "look at the numbers before and after" when it comes to polio. OPV creates immunity in the gut and we can see that, besides the other stuff such as the numbers in places that still need OPV to get that gut immunity necessary to stop transmission in endemic areas.

[Fyrestorm]

But, we also saw a change in diagnostic criteria in polio after the vaccine.

[ema-adama]

OK, but it is not the OPV that gets the credit for reducing the incidence of polio in the US. It is the IPV that was lisceneded in 1955 and was used until 1963 when the OPV became the vaccine of choice. Why? IPV is not as effective as OPV at confering immunity in the gut. But it is interesting that despite an initial mass vaccination campaign in 1955 with the IPV, another one was needed with the new OPV in 1963.

Is there anything else that might impact on the health of the population at large? Why is it assumed it was the vaccine? Like a parent who sees their 'normal' healthy child fall ill after a vaccine?

It is a similar story with tetanus. It is assumed that tetanus vaccine has saved humanity from certain death from tetanus. Again, I see the potential for this claim to be correlation equals causation. Although on a much grander scale.

[ema-adama]

Originally Posted by Fyrestorm

But, we also saw a change in diagnostic criteria in polio after the vaccine.

Yes, that helped too.

[orangewallflower]

Ema-edama,
Without specific reference to vaccines (I am just not an expert) I can tell you that what moves science from correlation to causation is understanding the mechanism of causation involved. For example, it might be argued that children who nursed as babies are on average a little more intelligent than babies who were bottle fed, on first glance can look like simple correlation without causation. (Parents who are committed to breastfeeding are probably committed to many more things that help a child's intellectual development.) But a good study will control for all those other differences AND there needs to be an understood mechanism for the causation. In the case of breastfeeding the mechanism is the good fats in breastmilk help myelinate the neurons in the brain. It is something that is now observable. Without this understanding, even in a controlled study, the relationship between breastmilk and intelligence would just be correlation.

[carmel23]

To help make the charts make some sense, it helps to listen to this TED talk by Larry Brilliant. It really shows how reported cases isn't always so clear-- apparently many folks in India didn't want to report, because of a cultural belief. So just because cases weren't reported, doesn't mean they weren't there.

Here is how the statistics were made, before the internet:

http://www.ted.com/talks/larry_brill...pandemics.html

There are very graphic images, so it might be best to watch it without kids around (mine would be very disturbed by the images).

[ammiga]

In addition to the change in the diagnostic criteria when the polio vax was introduced, there was also a major change in water purification systems and in the use of DDT.

I have read many studies that claim many reported cases of paralytic polio were actually DDT poisoning.

With the change in water purification, our world saw the decline of many many diseases. Although somehow, vaccines seem to take all the credit.

[ema-adama]

Thanks for the explanation on how science moves towards understanding correlation. As far as I know there is no way science can 'proove' the correlation of the decrease in disease in the 50's. There are too many factors that cannot be controlled - were not thought to be significant enough to be controlled.

And what is interesting is that despite a very strong suspicion or conviction on the part of parents that vaccines played a role in their child's poor health, science is not really doing very much to try and understand the correlation between serious adverse events that are linked in time with vaccinations.

Which makes me think that it is a little strange that correlation can equal causation when it suits, like in the reduction of disease, but it certainly can't equal causation when something as horrific as the death of a child or the permanent disability of a child is involved.

[carriebft]

I am not getting what you are saying. Are you saying that the only thing that can show vaccines helped reduce the incidence of certain diseases is correlation? We know the mechanism, we can see the mechanism...unless you really don't believe in germ theory or something like that...how is that not more proof than just correlation?

[bbrandonsmom]

Emaadma-You're getting at there were to many factors in that time when the vaccine was introduced to effectively say it was the vaccine correct, that caused the decline?

No one can say it was any one thing. If a lot of the diseases were on the decline anyhow w/o the vaccine, couldn't it be safe to say that w/o the introduction of them, the others would have ended as well? We can't go back in time, so we will not know.

[MyLilPwny]

Originally Posted by carriebft

We know the mechanism, we can see the mechanism...unless you really don't believe in germ theory or something like that...how is that not more proof than just correlation?

Because the Germ Theory is not proof. If the Germ Theory is actually false, then the whole premise behind vaccines is wrong. I'm not sure if you've seen this thread already, but in post #9, I linked to several sources that disprove the Germ Theory: http://www.mothering.com/discussions....php?t=1135209 and I can easily get more if those weren't enough.

[orangewallflower]

Well, actually, Germ Theory is a proven theory. It is important to remember the difference between hypothesis (a statement that one seeks to prove or disprove through experimentation) and theory (an explanation for a phenomenon or set of phenomena that has a great deal of evidence to support it.) Germ Theory was once a question because the germs were not visible. With a basic microscope, bacteria are visible. With an electron microscope viruses are visible. There is really no doubt that bacteria and viruses exist.

[Mirzam]

Originally Posted by orangewallflower

Well, actually, Germ Theory is a proven theory. It is important to remember the difference between hypothesis (a statement that one seeks to prove or disprove through experimentation) and theory (an explanation for a phenomenon or set of phenomena that has a great deal of evidence to support it.) Germ Theory was once a question because the germs were not visible. With a basic microscope, bacteria are visible. With an electron microscope viruses are visible. There is really no doubt that bacteria and viruses exist.

1. This does not mean they cause disease.

2. There are legitimate doubts that viruses actually exist.