I have a PhD in cancer cell biology. That doesn’t mean I’m an expert in the disease of cancer. I studied molecular mechanisms that are dysfunctional in cancer, but they may or may not be the cause of the particular cancers in which we see that dysfunction. Sometimes, an incidental mechanism gets messed up once cancer has already developed, and that gives the cancer cells another advantage over non-cancer cells in the body.
Cancer is a tricky beast for lots of reasons, not least because it is not one disease. It is hundreds of diseases, existing in many millions of people across the world. When we want to look for patterns in the profiles of people who are prone to certain types of cancer, or even how those cancers progress in certain people, we have to be extremely careful and sensitive to the nuance around the patterns we identify – and we have to be even more careful in how we share that research with the wider public.
Case in point: a few weeks ago, a friend who is a passionate ultramarathon runner dropped me a message linking a new paper, asking me the question “Dr Alice am I going to die??” Now I know my friend well enough to know that her framing of the question was intended as a joke, of the ‘make something serious and scary sound flippant and trivial’ variety. But I also know she can be a ruminator, and that this sort of research will live in her brain until she can get a handle on what it means.
Taking a step back
The paper in question is titled “Risk of pre-cancerous advanced adenomas of the colon in long distance runners”, published in the Journal of Clinical Oncology in June this year. In fact, it isn’t a paper at all; it is an abstract for a poster that was presented at a conference. This isn’t unusual, often researchers attending a conference will showcase some of their research, perhaps in a short talk or even a poster. In this case, the poster abstract was published at the ASCO conference – The American Society for Clinical Oncology, whose flagship journal is the Journal of Clinical Oncology.
While not unusual, it does mean that the abstract includes overall findings, but not the data or full results, so we can’t go and look for ourselves and come to our own conclusions. And we have to bear in mind that it has not been peer reviewed. That’s fine, I’m sure these researchers are preparing the manuscript for peer review as we speak. This is how the academic system works, and sometimes that system takes time.
Peer review is an important process – researchers are human and humans are subject to bias. They might have gone into this research with a particular mindset, and made slight errors in their interpretation. Peer-review allows someone not involved in the research to look it over and give the researchers feedback. That can sometimes help them strengthen their position – they might go off and do another experiment that solidifies their result and gives them even more confidence in their data. But it might also tone down the result if the researchers have neglected to think about something that might have confounded their finding, and help them add a little more nuance to their interpretation.
Results in context
The researchers noticed a number of ultramarathon runners present with advanced colorectal (bowel) cancer at their cancer centre. This is preliminary research, based on an observation. This sort of observation could be influenced by all sorts of things – not least, the cancer centre that’s central to this research is in America, a place where access routes to medical treatment can lead people to rely on recommendations for specific clinics. The ultramarathon-runner community is very close-knit, so it makes sense that, when you experience symptoms, you might ask your friends for recommendations. That’s one possible explanation for why multiple people from a certain demographic might present with symptoms of a particular disease at one clinic, without it necessarily meaning their demographic is at higher risk of the disease.
The researchers designed a preliminary study to interrogate this further. Preliminary studies are often about seeing whether there is something interesting enough to spend the money and time involved in doing a big, comprehensive study. They recruited long-distance runners who had completed at least two registered ultramarathons (50km or more), or at least five registered marathons (26.2 miles or 42km). They excluded people with known or suspected inflammatory bowel disease, as well as some other conditions that might skew the data, then asked participants to fill out a questionnaire on their diet, bowel habits and long-distance running history.
The studied included 100 participants, with a median age of 42.5 years, who were given a screening colonoscopy, and any polyps (small growths in the bowel) found were reviewed by a panel of gastroenterologists, pathologists and oncologists to check if they met the criteria for advanced adenoma – a pre-cancerous condition. They found that 15 of the 100 participants had advanced adenomas, compared to historical benchmark data in people with average risk aged 40-49, which shows an expected advanced adenoma rate of just over 1 in 100.
Advanced adenomas aren’t yet cancer, but they have some features that mean they’re more likely to progress into cancer later. This does not mean that those 15 participants will definitely go on to develop cancer – a study published in JAMA earlier this year suggests that people with advanced adenomas have a 5.5% chance of developing bowel cancer in the next 20 years, compared to 1.8% for people with no polyps and 3.9% for those with non-advanced adenomas. So, the 15 runners had an increased risk, and it’s a notable one, it’s definitely worth identifying and removing advanced adenomas, but we still need to be cautious in interpreting these data.
An increase from 1.2% in the average-risk population to 15% in the ultramarathon runners is a huge increase, especially in this age group – usually, we start to see polyps increasing in the over-50s. However, that data is often collected from hospitals routinely conducting colonoscopies on patients who are under investigation for some sort of gut problem that calls for the colonoscopy.
Simply looking at historical benchmark data is not an adequate control for this study – and the researchers confirm this themselves, with lead author Dr Timothy Cannon telling Fox News: “We did not have a control group, so this raises the question as to whether or not there is some bias.”
Colorectal/bowel cancer, which is the third most common cancer worldwide after breast and lung cancer, doesn’t produce symptoms right away, and when it does they can be easy to dismiss. It’s even easier for a population like ultramarathon runners to dismiss issues (if the ultramarathon runner I know is anything to go by), as they can be incredibly reluctant to stop running, even when injured.
Looking at the research into ultramarathon runners compared to sub-ultramarathon runners, or recreational runners, it’s probably true that there are some traits that are more common in those who run at ultra distance. In particular, ultramarathon runners score highly on scale get around setbacks. They don’t give up when something goes awry, they might just adjust their goals slightly.
Running very long distances is associated with gut symptoms. A notable proportion of runners, even at ‘only’ marathon level, report symptoms such as bloating, abdominal cramps, rectal incontinence, and gastrointestinal bleeding.
In fact, there is a rare condition called exercise-induced ischemic colitis, where blood flow is diverted to the muscles and away from the bowel, which can cause severe pain and damage to the guts, causing bleeding. As this is a relatively well-known phenomenon, bloody stools (a red-flag symptom for suspected bowel cancer) might even be dismissed as normal by regular runners, and runners may also be more likely to be dismissed by doctors instead of following up with further investigation. This might be further compounded by the perception that they are an extremely low risk group for bowel cancer, since this cancer is more common in more sedentary people.
There is no evidence that ischemic colitis is associated with an increased risk of cancer, but there is evidence that running at high intensity for a long period of time can cause intestinal damage. Speaking to BBC Science Focus, Dr Cannon said: “When you have so many cells dying and then growing back in a disorderly and unregulated way, there are lots of opportunities for DNA replication errors.”
The research matters, but so does the coverage
This is where this research is important. Running at ultra distance is an extreme sport, and it is going to have a physical impact on your body. That’s to be expected, and no ultramarathon runner will be ignorant to the significant risks they subject themselves to, or the strain their hobby has on their body. Finding an increased risk of certain health conditions isn’t going to surprise anyone.
The coverage of this preliminary study, too, has been reasonable – the articles on this topic are incredibly well balanced. They highlight that the research is preliminary, hypothesis generating, not controlled and not peer reviewed. However, most of the stories are presented as a quirky interesting finding about ultramarathon runners, and this misses an opportunity. Coverage of this study could have helped highlight that colorectal cancer is hard to catch early, and there are some symptoms that are cause for concern, including bloody diarrhoea, which the NHS says requires an urgent GP appointment.
This could also have been an opportunity to talk more about screening – which is recommended from age 45 in the USA, where this study was conducted, and from age 50 in the UK. In 2017, screening uptake for bowel cancer in the UK was lower than the other national screening programmes (cervical cancer and breast cancer) but this screening is known to reduce deaths. Screening in the UK is incredibly simple – a test kit arrives through the post with instructions on how to collect a stool sample, and can be returned via the post. For most people, it’s very straightforward.
While most of the articles on this study were reasonably balanced, their headlines have a lot to answer for. The New York Times article asked “Are Marathons and Extreme Running Linked to Colon Cancer?”, Fox News said “Long-distance running tied to higher risk of colon cancer, early study suggests”, health.com had “Marathon Runners Face Unexpected Colon Cancer Risk, New Study Suggests” and even the BBC’s Science Focus went with “Long-distance runners may face higher colon cancer risk, early findings suggest”. While the headlines are true, the issue is in how people read and digest information.
The New York Times heading is probably the best – they pose a question and they link it to the extreme endurance running, while still getting the salient information across. The others become an issue when people don’t then read the article. I’d speculate that most people don’t, especially if it doesn’t directly affect them, because they are not long-distance runners.
A study by Penn State University last year looking at 35 million public Facebook posts found that 75% of those links shared on Facebook were shared without clicking through the link to read the article. Many will have taken away the notion that running causes cancer, not making the connection to the extreme sport of ultramarathon running or even that the study doesn’t prove a causal link. Headline writing has to change.
Ultimately, this is a well-conducted preliminary study, which has been reasonably well reported, but people could be left thinking that ultramarathon runners – or even just regular-distance runners – are at greater risk of cancer, or that those forms of running cause cancer. Most of the reports said it was an early study, or it wasn’t controlled, or it wasn’t peer-reviewed yet, but that doesn’t mean that people understand what that means. We can’t be sure there isn’t something else at play underlying this finding.
People will change their behaviour long before we get the follow up studies – and the follow up studies won’t be nearly so well reported. And if that behaviour change is that ultramarathon runners and their doctors take bowel symptoms more seriously and people attend screening when it’s offered to them, that’s great; but if the behaviour change is that people avoid an exercise they might enjoy out of fear, but don’t then find an alternative, then we could be inadvertently contributing to an increased risk of ill-health.
