Do Tattoos Cause Cancer?

Science helps keep us safe, but it's not infallible.

When we hear this, we tend to think of cases where scientists failed to spot some hazard. For instance, in the late 1950s, thalidomide was prescribed for morning sickness, because rodent studies had found no signs of toxicity, and drugs weren't expected to cross the placental barrier.

Tragically, scientists were wrong on both counts. People aren't rodents, and some drugs do pass through the placenta. Thousands of babies were harmed by thalidomide before the drug was taken off the market in 1961.

There are also cases where scientists raise false alarms.

One of the most notorious examples is the 1998 Lancet study claiming that the MMR vaccine causes autism. The paper has been retracted and wholly discredited (flawed methods, fraudulent data, financial conflicts of interest, etc.), and no other data links the MMR vaccine to any form of Autism Spectrum Disorder. Unfortunately, some damage was done, and this study continues to be cited by anti-vaccination activists.

False alarms don't necessarily reflect bad intent. They may not even reflect weak methods. Sometimes data is just quirky, and the risks of some established practice are mistakenly exaggerated. Hold onto that thought for a moment.

In this newsletter I'll be discussing a new study claiming that tattoos increase the risk of cancer  – specifically, lymphoma, a group of cancers that originate in lymphocytes (a type of white blood cell essential to immune system response). To what extent can we trust this study? Is it credible, a false alarm, or something in between?

Why this is important

Tattoo inks are not regulated by the FDA, and some of them do contain carcinogenic chemicals. (See here for an excellent discussion of their physiological impact.) Meanwhile, according to a 2023 Pew survey, 32% of Americans have at least one tattoo, and another 6% say they're very or extremely likely to get one. The findings of the new study have been widely covered in the news and social media these past two weeks. If the study is credible, we can thank scientific and statistical methods developed over the past century for revealing some health risks of an ancient practice. If the study is a false alarm, a lot of folks may be growing needlessly anxious.

The new study

This study, published in the June 2024 issue of eClinicalMedicine, was conducted by Dr. Christel Nielsen and colleagues at the University of Lund in Sweden.

The researchers used a case-control method, meaning that they sampled people already diagnosed with lymphoma (cases) as well as a similar group who were cancer-free (controls).

Would having a tattoo increase a person's risk of ending up with lymphoma? The researchers addressed this question by means of something called incidence rate ratios (IRRs). Here's the basic idea: People get tattooed, years pass, and each year some of them develop lymphoma. But, as time passes, some people who've never been tattooed also develop the disease. IRRs tell us whether tattooed people are developing the disease at higher rates over time.

As in many countries, Swedish law requires all primary cancers to be reported to a national registry. This allowed the researchers access to data on every Swedish person diagnosed with lymphoma between 2007 and 2017. The final sample consisted of 1,398 of these individuals and a matching group of 4,193 cancer-free controls who were nearly identical in distributions of gender, age, years of education, marital status, and smoking behavior (current smoker, former smoker, or never smoked). About a fifth of each group had tattoos, and the groups were quite similar in age of first tattoo, length of time since first tattoo, and characteristics of tattoos such as number, size, and color.

Main finding

Overall, having at least one tattoo increased the risk of developing lymphoma by 21%. This is the finding that made the news. The Lund researchers – and many journalists – concluded that tattoos may cause lymphoma. End of story.

Actually, this is only the beginning of the story. And not just because we're only looking at one study.

It turns out that much the rest of the data undermines the main finding. One by one, the results of additional analyses call into question, or at least fail to support, the overall link between tattoos and subsequent lymphoma. (Not to be snarky, but it's as if this article itself suffers from a sort of autoimmune disorder, where the body ends up attacking itself.)

Some context

Earlier this week I reached out to Dr. Andrea Love with some of my concerns about the data.

Among other things, Dr. Love is a microbiologist and immunologist, the Executive Director of the American Lyme Disease Foundation, and the founder of Immunologic, a science and health communication organization, where she writes that

"A scientific expert has a duty to serve as a credible source for evidence-based data. More than that though, scientists must...promote scientific literacy by ensuring that reliable information is accessible for the general public."

Dr. Love practices what she preaches. She sent me a lengthy, easily-understood email that provided even more information that I'd asked for, including some important context for Nielsen and colleagues' main finding:

"[W]hen talking about lymphomas broadly, the overall RISK of them is quite low. Lymphomas – broadly – have an incidence of 5 in 100,000 persons – which is far less common than many other things that people think nothing about taking risks for: not wearing seatbelts, smoking, not wearing bike helmets, etc."

This is a reminder that if tattoos increase the risk of lymphoma by 21%, the actual extent of risk for tattooed people is still tiny.

Here's an analogy:  I happen to be a runner, and I run when it rains. My risk of being struck by lightning may be greater than average. But even if my risk is 21% higher, I'm still extremely unlikely to ever get fried.

Conflicting data

Now for some of the data reported by Nielsen and colleagues that contradicts or fails to support their main finding.

1. If tattoos are truly carcinogenic, we'd expect that the more ink a person has in their skin, the greater their risk of lymphoma.

(Since tattoo inks vary in carcinogenic content, here's a more careful phrasing: To support the argument that tattoos cause lymphoma, it would help to show that people with more tattooed skin are at greater risk. If that's not found, the argument becomes weaker.)

In fact, Nielsen and colleagues found no association between total area of tattooed skin and risk of lymphoma. However, they did find the highest risk among people with the smallest tattoos (i.e., a single tattoo smaller than the size of one palm).

Does this mean that if you have one small tattoo, you should get more in order to lower your risk of lymphoma? No. What it seems to mean is that the data are messy and unreliable, possibly owing to differences in composition of tattoo ink and other issues (see #3 below).

2. If tattoos are truly carcinogenic, we'd expect that the longer the time since getting tattooed, the greater the risk of lymphoma.

Nielsen and colleagues actually found the highest risk within the first two years of getting a tattoo. Risk began to increase again beginning a total of 11 years later.

This is especially problematic, as Dr. Love noted:

"If one wanted to make a claim that tattoo inks cause lymphoma, there would absolutely need to be a clearly defined relationship, and I don't see that from these data....I would expect a time relationship at the very least – individuals with tattoos for longer periods of time *should* have a tighter relationship to cancer.... Even with lymphomas, there is a process of cumulative mutations that need to occur in order for cancer to develop."

I also reached out to Dr. Nicolas Kluger, an expert at University of Helsinki who has published a review showing no associations between tattoos and skin cancer. Dr. Kluger was similarly unpersuaded by the new study, for much the same reasons as Dr. Love. He viewed the overall association as "coincidental", and he commented that if tattoos truly increased the risk of lymphoma we should expect to see effects of amount of tattooing and time elapsed, neither of which emerged in the Lund team's data.

3. If tattoos are truly carcinogenic, we'd expect that tattoos themselves, independent of anything else, increase the risk of lymphoma.

The researchers themselves acknowledge concerns here, at least on general principle: Even if we could trust the data, we can't be sure that tattoos per se are the culprit. Perhaps people who get tattoos are more inclined to other behaviors that elevate their risk of cancer. Nielsen and colleagues attempted to rule out some of those other behaviors (e.g., smoking), but you can't control for everything. Even smoking wasn't particularly well-controlled for, because the researchers didn't record any details about smoking behavior (e.g., number of years as a smoker) among those who currently or formerly smoked.

4. If tattoos are truly carcinogenic, we'd expect to see evidence for the physical mechanisms linking tattoos to lymphoma.

Since the Lund team didn't know the composition of each person's tattoo(s), they couldn't link carcinogen levels to a higher risk of lymphoma. Instead, all they could show is a weak (and, as I've noted, implausible) association between having tattoos and being at risk.

In Dr. Love's view, one of the biggest issues with the study is that

"all types of lymphoma are lumped together, when each type of lymphoma reflects very distinct etiologic and pathologic processes. Unfortunately, that really limits any sort of interpretation you can make from these data."

The researchers' perspective

In the new article, the Lund team acknowledged some of the problems I described here (and some that I didn't), but they weren't able to do more than to speculate about the causes. I reached out to the lead author for clarification but haven't heard back. In public statements, and in an article for The Conversation, she is, understandably, positive about the findings while acknowledging that more research is needed.

Some hypotheticals

The new study shows no broad link between tattoos and cancer. Dr. Love notes that in individual cases, tattooing might – in theory – increase the risk of some sorts of lymphoma. Here are her explanations of what might – in theory – happen in such cases:

"Theoretically, could some people have been tattooed with inks containing high levels of potentially carcinogenic substances? Sure, absolutely. That's one of the issues with tattoo inks and the lack of regulation of them. The *hypothetical* mechanisms could be: 1) contaminants of carcinogenic nature - such as lead, arsenic, cadmium; 2) inflammatory immune response as a result of attempts to digest/break down ink particles, particularly in the lymph nodes, leading to a state of persistent lymphocyte activation and proliferation; 3) genotoxicity - ink chemicals causing DNA damage directly which, over time, can cause mutations and aberrant proliferation, 4) disruption of lymphatic system/function."

In the end though, Dr. Love doesn't see tattoos as increasing the risk of lymphoma overall, in part because immune response to tattoos is primarily retained in the skin. More details can be found in her newsletter here.

Media coverage

The new study appears to be a false alarm, as it provides no basis for concluding that tattoos increase general risk for lymphoma. Rather, the findings are consistent with what appear to be the only two prior studies on the topic, one cited by the Lund researchers, the other presented at a 2023 conference and reporting the same null result.

Unfortunately, this is not necessarily the message the public is hearing.

CNN and Yahoo do a fairly good job of summarizing the strengths and limitations of the study, interviewing independent experts, and offering readers guidance about potential health risks.

Other outlets, like People, Newsweek, The Conversation, and News Medical cover the study without mentioning a single concern, including any of those acknowledged by the researchers themselves. (Newsweek does toss in one vague qualification: "These findings are purely associative, and more work needs to be done to verify them.") U.S. News even claims that "there's prior science backing up the tattoo-lymphoma link", which seems flat wrong given that the only other studies on the topic show no association. Treatment of the study on Reddit and other social media sites is even less accurate.

Here's a detail that's not very consequential in itself, but strikes me as symptomatic of a broader problem: Every one of the news reports I just cited (except for The Conversation piece and the U.S. News article) got the sample size wrong. These reports indicate a total sample of over 11,000, including 2,983 lymphoma patients. In fact, these are the numbers of people who received questionnaires asking about demographics, smoking behavior, tattoos, etc. Only about half of those folks actually responded. All of the data analyzed in this study came from 1,398 individuals diagnosed with lymphoma, plus 4,193 controls. (The actual numbers in some analyses were even smaller owing to missing data.)

This mistake hints that a study that may cause people to worry needlessly is not being examined closely enough. Accidental mistakes about sample size are no big deal in this case, but a careful reading of the paper would reveal important caveats, such as the tiny increase in risk that's at stake, if you could trust the data, plus the fact that you can't trust the data in the first place, for the reasons I've described here. I've only pointed out a few of the problems; it's unfortunate that some media reports didn't cover any of them. A false alarm is being sounded.

Final thoughts

Getting tattooed is not risk-free. Allergic reactions, infections, and diseases transmitted by contaminated needles are among the nasty consequences for some people. Some tattoos can actually conceal visible signs of skin cancer.

At the same time, nothing we do is risk-free. We make cost-benefit decisions every day, and to the extent that we can manage risk, we tend to choose the levels we're comfortable with. If you want a tattoo, you should get one. (Check out this useful guide first.) The point of this newsletter is that there's no reason at the moment to worry that tattoos increase the risk of lymphoma. What we need, if anything, is FDA regulation of the chemicals in tattoo inks.

Ordinarily I like having the last word, but Dr. Love summed up it so well I'll just step aside and share what she wrote:

"Ultimately, I don't believe that this paper should cause alarm in people who have tattoos or who want to get tattoos (and I say that as someone with 6 tattoos myself) - although I would love to see a world in which inks are regulated!"

Thanks for reading!