How do we know something causes cancer? Aren’t there essentially infinite variables it would be impossible to control for?
In: 40
Yes, but you can remove their Influence by looking at huge sample sizes. The law of the large numbers will make it unlikely you’re wrong UNLESS there is something cancer causing that heavily correlates with the thing you’re investigating.
For example:
In a sample of 10,000 users of cigarettes there was a 50,000% increase in lung cancer compared to non smokers. What kind of thing would be common to all those smokers that could cause such a big outlier if it isn’t the tobacco? Is there a shampoo that could cause lung cancer that only smokers use? Possible but very unlikely.
What you can’t do is a detailed analysis when things correlate too much. For example many weed smokers also smoke tobacco, so it’s very difficult to find out how much cancer smoking weed causes
Cancer is uncontrolled cell growth. Cell division is regulated by a number of proteins in the body. When they stop working, there are no checks being made and the cell divides without control. The cause of this is DNA mutation as DNA code for the proteins. Change the DNA = change the protein.
Scientists can observe if and how chemicals interact with DNA. Some ways would be to introduce mice/rats to the suspected chemical and check for cancer after time had passed.
Many things have been discovered to cause cancer this way, For example, radiation is simply high energy waves that literally shake the DNA to pieces.
BTW, there are two terms used for this:
Mutagenic – causes mutations in DNA
Carcinogenic – causes cancer
There generally isn’t a single study you can do to establish if something causes cancer. What you have to do is put together a variety of evidence.
Theoretical understanding, do we have an idea of why a substance might cause cancer? Do we have a mechanism in mind?
Mechanistic evidence, so can we do studies showing that that substance causes cancer in cells in “test tubes”?
Animal experiments. Can we give a rat the substance and see if it causes cancer. Here we can control for everything else, so it’s just that one thing that is different.
Longitudinal studies. Do people who consume more of that substance have higher levels of cancer.
So if we know a substance would cause cancer, and we have done RCT in animals showing it causes cancer while controlling for everything else, and we have correlational studies in humans linking it to cancer then it’s reasonable to assume that substance causes cancer.
Are you talking about cigarettes, or burned toast?
There are decades of studies that show that smoking leads to an increase in the chance of developing lung cancer.
Things like burned toast are often trotted out as causing cancer because journalists can’t read studies.
There are a few things that are pretty definitive, such as the link between cigarette smoking and lung cancer, but one of the reasons you have people arguing all the time about what is healthy and what is not is that it’s extremely difficult to control all the outside variables.
I remember for years the experts telling everyone that eating breakfast was healthy. Then they did a few studies, and found people who ate breakfast were healthier. They never considered that saying something is healthy will naturally attract more health conscious people to that thing.
Nowadays, I don’t know many people who still rave about how you have to eat breakfast to be healthy. Coffee and eggs also used to be linked to negative health outcomes.
The argument now seems to be regarding eating meat. However, you run into the same issue. Those regularly consuming highly processed meat products are probably not the most health conscious to begin with.
One way is choosing a random sample of test subjects.
Suppose you had 1,000 lab rats, and you divided the group in half, randomly assigning each rat to the control group or the test group. Keep everything the same for both groups, except for the one thing you are testing (like adding tobacco to their diet). If one of the groups has a much higher cancer rate then the other, it’s likely to be the thing you tested. It probably won’t be a genetic predisposition in one group of rats, because about half of those rats should be in each group.
It is actually more complicated, of course, but that’s one good practice in choosing test groups.
Ames test. This is the test developed specifically to test for the mutagenic properties of various chemicals. In essence you expose bacteria to a chemical in question. Now bacteria are grown on a media (gel like substance) that normally does not support their growth. However if the chemical is carcinogenic (cancer causing) it will mutate bacterial DNA and, more likely than not, it will allow bacteria to acquire a phenotype that will allow them to live on that media.
Colonies present = substance is likely to be carcinogenic
No colonies present = unlikely to cause mutation and this cancer.
Plus, like other people said, lots of other evidence.
Edit: spelling
Latest Answers