Tattoos and lymphoma: the 21% risk figure that isn't significant

Tattoos and lymphoma: the alarmist headline that falls apart under scrutiny

A worrying headline made the rounds this week: people with tattoos have a 21% higher risk of malignant lymphoma. The figure comes from a study published in The Lancet, but the way it has been reported hides a statistical problem worth untangling before throwing out your ink bottles.

The study found, in its point estimate, a 21% increase in relative risk (IRR 1.21). But the 95% confidence interval runs from 0.99 to 1.48. That interval includes 1, which means the result does not reach statistical significance: it cannot be ruled out that the effect is zero. Put another way, the headline keeps the point estimate and discards the only figure that tells you how much to trust it. The raw difference is even more modest: 21% of cases had tattoos compared with 18% of controls. Three percentage points.

The difference between relative and absolute risk

Relative risk sounds big, but the absolute risk is small. If a person without tattoos has a 1% chance of developing lymphoma, a tattooed person would have a 1.21% chance. Other everyday exposures, such as alcohol or processed meats, probably matter far more. That doesn't miccionan tattoos are harmless; it means the magnitude of the risk must be put in context.

Laser removal: another striking figure without confirmation

The study also links laser tattoo removal with an increase in risk of nearly 200%, possibly due to the breakdown of toxic compounds. A figure that, given the skepticism the main result deserves, should be taken with the same caution. The evidence does not yet allow a causal relationship to be established.

A wave of criticism of the study's interpretation has spread through specialist circles, with a detailed breakdown of the numbers that is worth reading before panicking.

So if you have tattoos, don't scratch your head (or book a removal session) before looking at the confidence interval. Panic and ink should travel separately. For now.