no, that makes no sense. How does having identical genetics negate problem of different lifestyles?

Even the authors agree with awestley:

> Additionally, having a tattoo, especially among adolescents, has been suggested as an indicator of risky lifestyle highly associated with e.g. smoking [4] and alcohol consumption [46] – both risk factors of certain cancer types. Hence, evidence of an association between tattoo ink exposure and occurrence of cancer may be confounded by other health-related lifestyle factors. We intend to exploit the remainder of the information gathered in the survey in the future.

Indeed, it does seem like a big problem that should really be accounted in the initial analysis.

Twin studies can help control for things like genetics and environmental factors in childhood (exposures, socio-economic factors, etc). They can't control for lifestyle choices made by adults. So, if a person has tattoos, are they more likely to smoke? To have sun exposure? To drink heavily? All of these factors would need to be addressed to see how confounding they are to having a tattoo.

You could also look at tattoo coverage, as in how much of the body is covered in ink? Would a small tattoo on the shoulder have as much risk as a full back tattoo? There are a lot of extra confounders here that could be better explored, but it gets difficult to get a full dataset. However, given their survey data, they should have more analysis options with more time.

The lack of confounder analysis is a bit surprising, but perhaps the paper was long enough already.