It’s important to understand the context for what’s being said. “given the low doses received by the majority of exposed individuals, any increase in cancer incidence or mortality will be difficult to detect in epidemiological studies.”

Detecting an increase in cancer deaths from a specific source when cancers represent 1 in 6 deaths worldwide is extremely difficult. If cancer killed 100 people per year then detecting an increase of ~X00 extra deaths per year would be much simpler. On the other hand try to isolate ~1 extra cancer death per 10,000 deaths is a pipe dream.