Cancer and Height
By Simon Harvey, Director of Life Sciences
There’s recently been a lot of media coverage of work that has identified height as a risk factor for a number of cancers. Now it’s been many years since I’ve looked seriously at the scientific literature around cancer risk, but it’s been known that height is a risk factor for cancer, particularly breast cancer, for a long time. That this story was so widely covered therefore seems a bit surprising. What is the deal? Is this a new discovery, a confirmation of an old discovery or it is just that cancer, dementia and heart disease, the big three killers, are always newsworthy?
A good starting point for any science story it to determine what has actually been found. So, what is the science?
This story is based on a preliminary report presented at the European Society for Paediatric Endocrinology conference. Unlike what will happen when the study is finally published, which is probably at least a year or so away, this means that it is not possible to look at the results in detail. The abstract of the presentation is however and the results are quite consistent with previous work (which is a good thing). Taken together with the very large sample size of the study (another good thing), it is likely that what has been presented here will match what we eventually see in print – that increased height increases cancer risk.
More specifically, the results of the study are presented as hazard ratios (HRs). These HRs, which are adjusted based on other known risk factors such as education and income, represent the ratio of risk between two groupings. Overall, the study showed an increased risk of developing cancer and specifically highlighted an increased risk of breast cancer and melanoma, with HRs varying from 1.11 (total cancers in men) to 1.32 (melanoma in women). Now this finding has been presented in the media as a percentage, for example that an HR of 1.20 indicates that ““
Does this mean a 20% chance of developing breast cancer if you’re tall? Thankfully not. Formally, the higher HR means that a taller woman who has not yet developed cancer has a higher chance of doing so at the next point in time compared to a shorter woman. This does not give us any information about the absolute risk. For example, a comparison based on two groups with absolute risks of 5% and 6% would give an HR of 1.20, the same HR as you would have if the groups had absolute risks of 50% and 60%.
Importantly, these HRs also don’t tell us anything beyond the risk of developing cancer, presumably over the entire duration of the study. This means that we’ll have to wait for publication of the results to see if there are any interesting things about the age distribution of cancer incidence and the outcomes.
So is this newsworthy? Yes, it’s the biggest study of the sort and it tells us something important about the risk of developing a disease that will affect many us either directly or indirectly. It is however perhaps unfortunate that the story has spread now without the detail and nuance that the real analysis will bring.