Correlation does not mean causation. Perhaps a lot of shift workers happen to be nurses and they work in an environment where it is more likely to contract cancer (from x-rays or whatever). I'm using this as an example.
Also, I don't think it's "made up by a historian", it's more like a "historian commented on common behavior of our ancestors".
I have a friend who works at the American Cancer Society. They constantly have to refute the "[insert something here] causes cancer" myths. They do some pretty insanely large and in-depth studies on these things. They've even investigated the "your couch gives you cancer" myth.
In the end, they say that smoking is by far the biggest risk factor for cancer. After that, there's a "three-legged stool" of obesity + sedentary lifestyle + poor diet, meaning those things compound one another. Genes also have a measurable influence on your likelihood of getting cancer.
After that, very few things significantly increase your risk of cancer. A lot of it is just random. Most of the studies you see linking cancer to something are bullshit. My friend says you should be skeptical of anything shorter than 10 years and with fewer than 100,000 participants, and even then there are many other factors that go into gold-standard research.
Also, I don't think it's "made up by a historian", it's more like a "historian commented on common behavior of our ancestors".