… or was I just unlucky?
It is a natural question for anyone to ask when he first discovers that he has cancer. As a scientist, my first inclintion was to look for some data to help me to understand whether anything I did could have contributed to my cancer’s development. It doesn’t take long to find out that there is a whole list of risk factors for oesophageal cancer. From a Wikipedia page I found a long list of factors that increase the risk of getting oesophageal cancer, of which the only relevant ones were my sex (I’m male, and men are much more likely than women to get this cancer) and the fact that I was overweight (until the cancer got going). One of the factors mentioned is age, but only old age, so in that sense my age reduced my risk of developing cancer.
This is not the end of the story, however. There are some factors associated with reduced risk of contracting oesophageal cancer. From the same page I found that the fact that I occasionally used ibuprofen worked in my favour, as did the fact that I like both brocolli and coffee, maybe even pizza! Unfortunately, I don’t have an ulcer, which is a shame, as Helicobacter pylori (the bacterium associated with ulcers) may have a protective effect.
All very well, I thought, but there are no numbers. How harmful was it that I was overweight (technically obese, as I was more than a quarter heavier than my ideal weight)? And did my diet otherwise help? Let’s try to find some numbers.
The Cancer Research UK site was helpful here. One page told me that obesity is associated with a risk factor of roughly 2. What does that mean? Simply put, it means that I was twice as likely as a non-obese person to develop cancer. Wikipedia has more details on risk factors if you’re interested.
How likely is a normal person to develop cancer? Once again the Cancer Research UK site provides the information. Of every 100,000 males aged between 45 and 49 years, 6.8 of them are expected to develop oesophageal cancer. In other words, I had a probability, in the absence of any risk factors, of 0.0068% of developing oesophageal cancer. If you leave out all the positive things I did, such as drinking coffee and eating a wide variety of vegetables and pizza, and only consider the fact that I ate too much and was therefore overweight, I had a probability of 0.0136% of developing cancer.
I think the conclusion is clear: I was quite simply unlucky. It would not make any sense at all to say that my obesity caused my cancer. Otherwise, how could I explain the remaining 99.9864% of probability? Similarly, it probably made no practical difference that I ate and drank all those “good” things, as the associated probabilities were equally tiny.
What does this tell me about risk factors? Let me try to illustrate with a question. How would you feel if you were told that being obese increased your chances of developing oesophageal cancer by 0.0068%? Or put another way, that you were increasing your chances to 0.0136%? Would you worry about it? No, neither would I. But if you were told that you were doubling your chances of developing the cancer? That would be far more likely to be worrisome, wouldn’t it? These differences are statistically significant. That means you can reliably measure the difference; it doesn’t mean that the difference is important. And there, I think, is the reason that people talk about risk factors. The measurable difference is something you can talk about and get worried about. I have noticed frequently that news programmes on the radio or television often mention increased risk factors, but they never say what the absolute risks are. Maybe that’s because those risks are not uniform across the population, which would make it difficult to summarize them in a single, easily understood number. (Consider that as a relatively young man, my risk of developing my cancer was a lot lower than that of a retired man, for example.) But I don’t think it would be too hard to show a graph of the relevant distributions. Scare stories attract an audience, don’t they? And they’re easier to write, just less informative.
There. I’ve said enough. Time for others to have a say.
{ 5 } Comments
I agree that the risks of getting a specific cancer are very small, especially when it is rare, but what about cancer in general? We are told 1 in 3.
Apparently we all produce mini-tumours everyday in various parts of our body, and many factors probably play a role in allowing them to grow further or not. Undeniably it is always bad luck, some risk factors might be smoking, obesity, food, stress, environment, genes,… the problem is simply too complex (like trying to predict the weather beyond a few days!).
That’s true, I suppose, although it is a different question from the one I was trying to answer. It seemed to me that it is quite common for people to ask why they got cancer. If you Google [why OR how did I get cancer] you will see that you get a huge number of exact hits. People are clearly interested in understanding what happened to them, and whether they could have prevented it. Probably pointless, I know, but it’s something that matters to them. I wanted to address that, and lay to rest any fears that it might have been their fault — or my own in my case. In fact I don’t think I managed really to answer my own question, so there will probably be a follow-up post later today, in which I try to determine to what extent my obesity actually explains my cancer. At the moment, I think the maths agrees with my intuition, but I’d like to check all my assumptions and reasoning before going public.
As for the mini-tumours, I can believe it, although I haven’t heard that before. The human immune system is remarkable, I know. Many of us try to live healthy lives, taking the advice that we are given by the “authorities”. The thing that made me rant was that the advice is often couched in terms of risk factors, without giving the raw numbers, which makes it impossible to assess the effects of ignoring the advice. Would I have eaten less if I had been told at age 25 that I was doubling my chances of oesophageal cancer from tiny to twice-tiny? I can’t think that I would. How many people would give up something they enjoy as much as I enjoy good food on the basis of such a tiny probability? It’s just not rational. And then the effort to try to change my behaviour would have been pointless. In someone else’s case, the risk factor and the absolute risk might have been much higher, and the effort would be worthwhile. A heavy smoker, for example, is much more likely to get lung cancer than a non-smoker. That’s why it’s actually worth trying to get people to stop smoking. I was at far greater risk of heart disease from my obesity than cancer. The cancer was just extremely unlucky, although the heart disease would likely have killed me just as dead!
Phew! Long comment. This subject does seem to make me rant, doesn’t it? đ
I realise this is not your original question, but presumably if one were investigating the effect of obesity, you would want to sum across all known risk factors. So you first work out your starting risk of all possible problems (different cancer types, heart disease, etc). This must be reaching into an actual percent or two. Then you work out the risk of all those same problems given obesity, sum again. Then you have two risk factors, large enough to be worth considering (although even an increase of 1% to 2% may not be enough for most people to worry about).
David Colquhoun has an excellent long article about risk factors (http://www.dcscience.net/?p=1435) which sticks in my mind for his (very paraphrased) conclusion: “The question of how diet and other âlifestyle interventionsâ affect health is fascinating to everyone… The first great success was cigarette smoking but that is almost the only major success. Very few similar magic bullets have come to light after decades of searching (asbestos and mesothelioma, or UV radiation and skin cancer count as successes).” He says that almost every risk (of the “X causes cancer” type that so regularly appears in news outlets) is not really worth worrying about: “The National Cancer Institute (Nelson, 2002) advises that relative risks less than 2 should be âviewed with cautionâ… In fact hardly any of the relative risks reported in the WCRF report (2007) reach this level”. Anyway, I’ll stop quoting it, but I recommend anyone interested should read it.
First of all my mother language is NOT English, so I’ll make my best to express myself.
I USED TO BE A SCIENTIST.
According to my opinion Science has let us all down. So, I as painful as it is for me I am not proud of Science any longer.
My own common sense and my scientific training are my weapons now. As result of my own personal investigation I believe toxicity to be behind all cancers. That is: toxicity in water, food, air and soil. I do pay lots of attention on what I eat and the water I drink and have moved to a place with very-very clean water.
I believe all industrial products to be toxic, including bottled water and I also believe that cancer can be cured ONLY if we isolate ourselves from all toxicity. I can understand cancer prevention ONLY in terms of a toxicity- free life.
Very hard to achieve.
That’s an interesting point of view, Miranda, but I’m not at all sure I can agree with you. Science is behind the treatment that I have received which is making me feel so much better. Science is responsible for the advances that lie behind the development of the computer on which both you and I typed our messages, and also the network that allowed each of us to see the other’s message. There may well be a lot of toxicity around, but as a scientist, I’m sure you understand that you would have to present a more convincing body of evidence that toxicity is behind cancer development for me to agree with you.
As for paying attention to what one eats and drinks, you should know that I took — and still do take — a huge amount of responsibility for my food. Almost everything I cook is locally produced and fresh, and I cook it from scratch. I eat very little processed food. To suggest that “toxicity” is responsible for any cancer is simply too weak an argument for me even to begin to think about accepting it.