To keep it simple these are the points we need to consider:
- The IARC classification system isn’t based on strength of risk, rather (and somewhat confusingly), strength of evidence. The reason that processed meat is classified next to smoking and asbestos is due to available data on the relationship between these and cancer (which we can argue about later). The media outlets that are running headlines saying processed meat is as dangerous as smoking are missing the point entirely. Just to give some perspective, sunlight is also classified as a class 1 carcinogen….
- This classification is based on a large meta-analysis of observational studies. This is important. The stories about eating 50g extra red meat a day increasing risk of colo-rectal cancer by 18% don’t understand the science. Unless you perform an intervention study where you actually feed people that extra 50g, control other variables, and compare to a control group over time you can’t make the claim. This is data from studies where individuals are asked to complete a food frequency questionnaire and are then followed for a period of time. Can you remember what you ate last week? What about last year? Safe to say the FFQs aren’t particularly reliable.
- We next need to consider what ‘processed meat’ actually means. In studies such as this pizza, sausage rolls, bacon sandwiches, pasties etc. are all classed as processed meat. This data isn’t based on health conscious people such as you & I having some pastured bacon without the bap for breakfast or some deli meat as a snack, instead it covers the carb and veg oil heavy processed food which we all know isn’t good for us.
- Now, let’s go back to our old friends, absolute and relative risk. The media, as ever, reported 18% increased likelihood of cancer, which is the relative risk. If we take some figures on colorectal cancer risk from Cancer Research UK we can see, that in 2011, the risk of CRC was 43 in 100,000. Assuming the accuracy of the WHO data, if we were to eat 50g of processed/red meat a day this absolute risk would go up to 51 in 100,000. That’s an extra 8 incidences in 100,000. Not exactly terrifying is it? Can you see why we need to ignore relative risk on these things!
- And finally… Correlation vs causation. Unless you perform a randomised control trial on something you can’t even enter a conversation about causality. This is correlational data. All it tells us is that people who eat more processed meat seem to have a (very slightly) higher incidence of CRC. It doesn’t tell us that the meat causes CRC. If we keep our heads and consider this we shouldn’t be surprised. Where do we think the average person is eating their processed meat? Greggs or an independent local deli?
So to summarise. This is a classification based on strength of evidence, not strength of risk. It is weak, observational data, that isn’t able to comment on causality. The correlational increase demonstrated is tiny considering absolute risk. And it fails to take into account the differences between manufactured gunk and traditionally cured ‘processed’ meats.
So if you, like me, eat a wholefoods based, lower carb diet. You exercise, sleep well and limit classic ‘unhealthy’ lifestyle behaviours then you can pretty much ignore this report and carry on as you were. Nice to know eh.