A dramatic increase leads to a simple question: why?

From the year 2000 until around 2016, the incidence of breast cancer among young women – those under age 50, rose steadily, if slowly. And then, this happened.

I look at a lot of graphs in my line of work, and it’s not too often that one actually makes me say “what the hell” out loud. But this one did. Why are young women all of a sudden more likely to get breast cancer?

The graph comes from this paper, appearing in JAMA Network Open.

Researchers from Washington University in St. Louis utilized SEER registries to conduct their analyses. SEER is a public database from the National Cancer Institute with coverage of 27% of the US population, and a long-track record of statistical backbone to translate the data from SEER to numbers that are representative of the population at large.

From 2000 to 2019, more than 200,000 young women were diagnosed with primary invasive breast cancer in the dataset, and I’ve already given you the top-line results. Of course, when you see a graph like this the next question really needs to be “why”?

Fortunately, the SEER dataset contains a lot more information than simply whether someone was diagnosed with cancer. In the case of breast cancer, there is information about the patient’s demographics, the hormone status of the cancer, the stage, and so on. Using those additional data points can help the authors, and us, start to formulate some hypotheses as to what is happening here.

Let’s start with something a bit tricky about this kind of data.  We see an uptick in new breast cancer diagnoses among young women in recent years. We need to tease that uptick apart a bit.  It could be that it is the YEAR that is the key factor here – in other words, is it simply that more women are getting breast cancer since 2016 and so more young women are getting breast cancer since 2016 – these are known as period effects.

OR – is there something unique to young women? Something about their environmental exposures that put them at higher risk than they would have been had they been born in some other time? These are known as cohort effects.

The researchers teased these two effects apart as you can see here and conclude, well, it’s both.

The rising incidence of breast cancer in young women is due both to the general increased incidence over time, and the unique risk of being born in the late 70s to late 80s.

Stage of cancer at diagnosis can give us some more insight into what is happening. These results are pretty interesting. These higher cancer rates are due primarily to stage 1 and stage 4 cancers, not stage 2 and stage 3 cancers.

The rising incidence of stage 1 cancers could reflect better detection, though many of the women in this cohort would not have been old enough to quality for screening mammograms. That said, increased awareness about genetic risk and family history might be leading younger women to get screened, picking up more early cancers. Additionally, much of the increased incidence was with estrogen-receptor positive tumors, which might reflect the fact that women in this cohort are tending to have fewer children, and children later in life.

So why the rise in stage 4? Well – precisely because younger women are not recommended to get screening mammograms, those who detect a lump on their own are likely to be at a more advanced stage. But I’m not sure why that would be changing recently. The authors argue that an increase in overweight and obesity in the country might be to blame here- prior studies have shown that higher BMI is associated with higher stage at breast cancer diagnosis.

Of course, we can speculate as to multiple other causes as well. Environmental toxins, pollution, hormone exposures and so on. Figuring this out will be the work of multiple other studies. In the meantime, though, we should remember that the landscape of cancer is continuously changing. And that means we need to adapt to it. If these trends continue, national agencies may need to reconsider their guidelines for when screening mammography should begin – at least in some groups of young women.

A version of this commentary first appeared on Medscape.com