Getting the science right – hair dye & chemical straightening products as a case study
In this post, we'll dig into the facts around hair dye and chemical hair straighteners.
While it's easy to get caught up in clickbait headlines about the latest scientific breakthroughs that can fill us with fear and anxiety, it's essential to take a step back and explore the information critically and from the source.
In this post, we'll dig into the facts surrounding that latest scientific discovery that's led to sensational headlines around the risks associated with hair dye and chemical hair straightening products.
What does the news say?
You may have seen the headlines, "Study Links Hair Dye and Chemical Straighteners to Breast Cancer, With Risk From Dye Highest for Black Women," "Permanent hair dye, chemical hair straightening could increase risk of breast cancer, study finds," "Black women are over 6 times more likely to get breast cancer from hair dye and relaxers- New study finds", "Hair dyes linked to higher risk of breast cancer: study," and "Is coloring hair safe? Dye, straighteners may increase breast cancer risk, study finds".
These are a few headlines of the many stories making their way through our newsfeeds over the past couple of weeks. But, do these headlines reflect the absolute risk? Do the majority of these articles explain the scientific nuance?
We found the reporting quality varied dramatically depending on the news source. While some sources explained the research and the risks fully, other sources provided little useful information other than reinforcing the headlines that hair dye and chemical hair straightening products cause cancer.
Most concerning, it appears that the articles shared more frequently contained less scientific rigor and more on adverse health outcomes.
So, what are the facts?
We've looked at the study and found that, frustratingly, many news headlines and stories did not fully represent the science or the researchers' conclusions. Unfortunately, this isn't uncommon in the world of science journalism.
What makes this case particularly poignant is that the limitations of the study are specifically called out by the study's authors, yet were not discussed in many of the journalism and blog posts around this topic. The study's shortcomings are significant in understanding the actual rise in cancer risk associated with the use of hair dye and chemical hair straightening products.
The authors identified vital shortcomings, including:
- The chemical composition of the products used by these women is unknown.
- This study required women to recall information from the prior year using a survey, which can lead to unintentional errors. Additionally, the information about the frequency of use is solely limited to the 12-month period prior to study enrollment.
- The study population is a high-risk population, meaning they already have a 2-6-fold increased likelihood of developing breast cancer compared to the general population, which can limit researchers' abilities to generalize the findings.
What do the shortcomings mean?
The author-identified shortcomings help put research into context and provide insights into future research directions.
Why do we need to know the chemical composition of the products?
The chemical composition of the products that participants used is critical to understanding the inherent risk of the product. If researchers knew the ingredients in the products, they could further test those ingredients to discover how they interact in the body.
Because we do not know the chemicals that participants were exposed to, researchers cannot directly link the potentially harmful ingredients to the adverse health outcomes. This also means that if the harmful ingredients have not been identified, they cannot be removed from products.
What's the problem with participant recall?
A major problem with epidemiological studies that require participants to recall information that happened in the prior year is that people can easily forget or misremember, in this case, the time or frequency they were exposed to a product.
The study also focused on a 1-year only timeline, which limits the known frequency of exposure. By not taking into consideration the frequency of use during the average 8-year monitoring timeline, researchers were not able to see longer-term exposure and frequency patterns, thus limiting the known associations and correlations.
Moreover, not knowing the frequency of use, the authors lack information about the exposure of the participants to these products. In the absence of information about exposure, it is hard to draw associations to adverse health effects caused by the discussed products.
Why does a high-risk population matter?
This is perhaps one of the most critical aspects of the study as the participants are more likely to develop breast cancer. Specifically, the sister study only looked at individuals with a sister already diagnosed with breast cancer. It's well established that when one sister has breast cancer, the risks of another sister(s) developing cancer increases by 2-6 times of the general population.
Indeed, as the investigators acknowledged, results from a vulnerable population make it difficult for researchers to generalize finds to the larger population because the participants are already at an elevated risk.
The risk many news articles miss
While many articles did rightly explain that it's too early to draw too many conclusions about the health implications of hair dye and chemical hair straighteners, publications did not explain the absolute risk.
Saying a risk increases by a certain percentage doesn't help the reader much if they don't know their risk before using those products and after using those products. Stating that the risk increases, while factual, can introduce undue anxiety, and our goal should be to inform the public rather than cause alarm.
It is important to understand relative risk and absolute risk. In the most eye-catching stories, the relative risk was highlighted, for example, one source states, "black women participants who used permanent dye were associated with a 45 percent higher breast cancer risks."
While this is technically true, it's only focusing on relative risk. While 45% sounds dangerously high, that is relative to the background breast cancer risk for this high-risk population which is 0.6% 1. So, a 45% increase to a 0.6% risk is 0.87% risk. While it's still an increase, it remains below 1 percent and is not dangerously high compared to other lifestyle factors such as cigarette smoking and obesity.
Nearly universally, we found many publications did not explain the change in risk, nor did they frame the risk in a way people can easily understand.
If the readers do not easily conceptualize the risk, people cannot easily make informed decisions. Saying, "the risk increased by 45 percent" or "seven percent" while factual, doesn't put the risk into context, making it difficult for people to make sound decisions.
The findings many news articles miss
While many news articles focused on the sensationalized finds, they did not include information that showed no trend of adverse health impacts. For example, the study did not find a statistically significant correlation between the “years of personal use” of permanent hair dye products and the “incident breast cancer.”
Readers need to see the full story and that requires sharing all of the relevant results, not only the eye-catching results but also the results that show no adverse effect.
What does this mean for the scientific community as a whole?
We think it's important to call out the facts and help demystify the science and, in some cases, calm fears. We need to do our best to continue highlighting study shortcomings, vetting our research through the lens of the public and the press, and be willing to work with journalists and even bloggers to make certain they get our research right.
In our opinions, if publications had added more context, including the absolute risk, the stories would have succeeded in informing the public rather than instilling fear.
1. Calculated with data from the paper, https://doi.org/10.1002/ijc.32738