A recent article in Science magazine (1) generated a raft of headlines shouting “Most cancer types ‘just bad luck’” with subheadings like “Most types of cancer can be put down to bad luck rather than risk factors such as smoking” (bit.ly/bbccancer). It seems that in some news outlets, the topline message being pitched was that modifiable risk factors like drinking, smoking and overeating aren’t really that important – now that cancer is mostly just bad luck. The importance of modifiable risk factors was mentioned later in the article – but at a point where most people have stopped reading (bit.ly/DMcancer;bit.ly/poynterinst). Not ideal.
So what did this “shock new study” (bit.ly/dailymirrorshock) actually find? The study’s authors wanted to understand why cancer risk varies between different tissues in the body – the average lifetime risk of lung cancer is around 1 in 14 (bit.ly/CRCUK), but for uveal melanoma, it’s closer to 1 in 13,000 (2). Their hypothesis: cancer risk is related to the number of stem cell divisions required to maintain a tissue. The adult eye undergo very few divisions; intestinal epithelia undergoes an awful lot. They examined the incidence of cancer occurrence across 31 tissue types, in order to gain an understanding of the relative contribution of unforced errors in cell division to cancer. What they found was pretty uncontroversial: greater proliferation rates increase the risk of developing cancer; and that ~65 percent of cancers were explained by this association – hence the headlines, although hardly good copy, and the fairly wide confidence intervals (95% CI; 39 to 85 percent) never made it into the newspapers… People can’t cheat chance, but the remaining 35 percent (or anywhere between the CI’s 15 to 61 percent range) can be influenced with those unpopular lifestyle interventions listed above. As PZ Myers put it: “If you were going off to the casino to shoot craps, you wouldn’t sneer at something that gave you a 10 percent edge on the table?” (bit.ly/pzmyers).
Of course “Study shows that a complex interaction of random molecular events, genetic propensity, environment and lifestyle” isn’t exactly a catchy headline. When it comes to healthcare stories, news outlets take a very “black and white” view: X is either the elixir of life, or out to kill you, as illustrated by the Daily Mail Oncological Ontology project (bit.ly/DMOOP). Perhaps a few shades of grey (about 50 should do it) would restore a little reason to the debate.
