Hair Dye Isn’t Giving Everyone Cancer

Why you probably don’t have to worry about dyeing your hair and breast cancer

Image for post
Image for post

Hair dye is in some ways the hero of modern cosmetics. Even a relatively short time ago, the main options were to go grey with age or look very odd, but now you can easily change your hair colour almost at whim*. It’s cheap, easy to do, and there’s some evidence that most women have tried it at least once in their lives.

And, according to the news, it’s also a terrifying chemical nightmare that’s giving you cancer.

Image for post
Image for post
Seriously, there have been a lot of scary headlines Source: Google

For anyone who’s ever dyed their hair — which includes somewhere between 50–75% of adult women — this must’ve been a scary week. But while the headlines and stories are frightening, in reality the science isn’t nearly as worrying as you might’ve heard.

Chances are you can keep dying your hair. It isn’t giving you cancer after all.

All of the media stories are reporting on the same study, that was recently published in the International Journal of Cancer, where scientists looked at the association between hair dyeing behaviour and future breast cancer risk. Basically, they asked a large group of women how often they dyed their hair, and followed up about 10 years later. Compared to the women who dyed their hair the least, women who dyed their hair a lot had an increased risk of breast cancer.

And from this, the headlines were born.

Image for post
Image for post
Pictured: Cancer, probably. Source: Unsplash

And on face value, the study sounds very robust. It was a large group of people — more than 50,000 women — and the follow-up was very good. The scientists even controlled for a range of factors that could influence both hair dyeing behaviour and breast cancer risk, in what’s known as controlling for confounders, which makes the results a bit less likely to be wrong.

But, on closer examination, it turns out that the results really aren’t that worrying after all.

The first thing to note is that this was an observational study. What this means is that, even in the best case scenario, it’s pretty hard to know whether the relationship is causal or correlational. In other words, it’s quite challenging to know if hair dye causes breast cancer or is just associated with it.

For example, this study had no direct control for socio-economic status. It’s not unlikely that people who dye their hair more often are more well off than people who do it less, and we know that there’s a relationship between breast cancer diagnosis and wealth (i.e. rich women are more likely to have their breast cancer picked up, but less likely to die of it). It’s entirely possible that this study simply shows a fact we already know — poor women don’t have their breast cancer well diagnosed, using ability to buy professional hair dye as a proxy for poverty.

Image for post
Image for post
Pictured: More expensive than going grey Source: Pexels

This is what’s known as residual confounding, and it’s a big problem for observational studies. While it’s possible that the scientists managed to control for all of the potential factors that could get in the way of their analysis, it’s much more likely that there are still some issues remaining. If we look at women who dye their hair compared to those who don’t, we see important differences in age, education level, number of children, and a range of other factors that can influence breast cancer risk.

In other words, it might not be hair dye causing all of those cancers.

There are a number of indications that this is true in the study itself. For one thing, there was no consistent biological gradient. This means that, in some cases, women who dyed their hair more were less likely to get breast cancer than those who dyed their hair occasionally. The relationship was also statistically quite weak, with some results only just meeting the usual threshold for statistical significance**.

There was also an issue with how the study define exposure to hair dye. They asked women once, at the beginning of the study, how often they remembered dying their hair. Then, they assumed this stayed constant for the next decade, which is a big issue. It’s not unlikely that women change their hair dyeing behaviour over the years, to do it more or less, which makes these results quite likely to be spurious. Women might also misremember, or lie about how often they dye their hair, which again could bias the results.

Even worse than these issues was the difference in absolute and relative risk. While the headlines only reported the 45% increase that seemed the scariest, the absolute increase in risk was only about 0.05%. To put it another way, for every 10,000 women who never dye their hair, about 72 get breast cancer every year. For women who dye their hair every month, this rate goes up to about 77 in 10,000.

Quite a bit less scary than the headlines when I put it like that, don’t you think?

In cases like this, it’s always good to go to prior research and see what it says. In this case, the answer is basically “not all that much”. There’s some signal of risk associated with hair dyes and certain cancers, but the reality is that this link is quite weak and unlikely to be causal. The scientists theorized that the link they found in their study might be to do with formaldehyde, but that’s very unlikely given that you probably get more formaldehyde from eating pears than dyeing your hair.

Image for post
Image for post
Pictured: More poisonous than hair dye Source: Pexels

So, the headlines were wrong. There’s really no reason to believe that hair dyes directly cause breast cancer.

There may be a risk associated with being the sort of person who dyes your hair a lot and getting a breast cancer diagnosis, but realistically that’s not the sort of thing you can change anyway.

If you’re worried about your health, as ever it’s a good idea to talk to your doctor.

Just don’t stress out too much about dyeing your hair.

It probably isn’t giving you cancer after all.

If you enjoyed, follow me on Medium, Twitter or Facebook!

You can now listen to Gid on the Sensationalist Science podcast for your weekly dose of scientific shenanigans and media muddling:

*Unless, like me, you started going bald at age 20, in which case you can only wish that you hair was going grey.

**Note: we can argue about frequentist statistics and the arbitrariness of p-values, but absent of a biological gradient, having several p-values at 0.05 (with confidence intervals that cross 1) makes the relationships a bit more tenuous in my opinion.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store