Friday, January 27, 2012

Of meat, cancer, adultery and statistics

More confirmation of the Stratford Theory of Numbers from the BBC which reports:
A link between eating processed meat, such as bacon or sausages, and pancreatic cancer has been suggested by researchers in Sweden.
They said eating an extra 50g of processed meat, approximately one sausage, every day would increase a person's risk by 19%.
Pancreatic cancer is very nasty but the lifetime risk for men is one in 77 or 1.3%. So for the one-a-day sausage-eating community the risk would seem to be 19% more than that, or 1.5%. Hmm. If you ate just one or two sausages a week it might be 1.35%. Change of lifestyle indicated? No, me either.

Recent reading included Panic on a Plate: how society developed an eating disorder by Rob Lyons (Societas, 2011). In the chapter “A smorgasbord of panics”, Lyons debunks the “five a day” fruit and veg policy – a UK professor of food policy (I am not making this up) explains that five was chosen because “it was considered a nice round sum” – then discusses the suggested link between meat-eating and cancer (here is a condensed version) and points out the flimsiness of the evidence:
The studies in question suggest that the people who eat the most processed and red meat have an increased risk of somewhere between 15 per cent and 45 per cent. But given the crude nature of these studies – often based on a questionnaire of a whole variety of different eat­ing and other personal habits in the first week of the study followed by checks on disease status some years later – there is room for all sorts of confounding factors to creep in. Are the kinds of people who eat a lot of processed meat dif­ferent in other ways, too, like being poorer?
That’s why the US National Cancer Institute declared in 1994 that “in epidemiological research, [increases in risk of less than 100 per cent] are considered small and are usually difficult to interpret. Such increases may be due to chance, statistical bias, or the effects of confounding factors that are sometimes not evident”. In other words, if a particular factor does not at least double the risk of a particular disease or condition occurring, treat that result with a pinch of salt. The link between lung cancer and active smoking, for example, is unambiguous, with increases in risk more in the order of 2000 per cent, not 15 per cent. In such circumstances, correla­tion seems a good indication of causation. The red/pro­cessed meat link to cancer is nothing like as strong.
That figure of a 100% increase in risk being the threshold for concern is worth remembering. Speaking of statistical bollocks, the good Oil linked to a superb demolition of that spectacularly stupid story in the NZ Herald based on a press release from an infidelity website – that’s right, you can go online to find someone with whom you can cheat on your spouse/partner. It was headlined “Infidelity: it’s a right-wing, meat-eaters’ thing” and claimed that “If your partner supports National, has a PC, drinks Coke, eats meat, has a tattoo, smokes and is a Christian, be warned – they could be a cheater.”

Over at the excellent StatsChat, statistician Thomas Lumley pulls this apart under the title “Unfaithful to the data, too”:
The proportion of National supporters in the election was 47%, among website members it’s 33%, so National supporters are substantially less likely to be members of the website than supporters of other parties. The proportion identifying as Christian among website members is very similar to the proportion in the 2006 census. 79% of website users are on PC (vs Mac). Again that’s a lower proportion of PCs than in the population of NZ computers (the Herald said 10% were Macs in July 2010, and for Aus+NZ combined, IDC now says 15%) but one explanation is that Macs have more of the home market than the business market.  More members drinking Coke vs Pepsi is also not surprising — I couldn’t find population figures, but Coke dominates the NZ cola market.
The story doesn’t say, but we can also be pretty confident that the website members are more likely to be Pakeha than Maori, more likely to be accountants than statisticians, and more likely to have a pet cat than a pet camel.
Cactus Kate has boldly gone where I wouldn’t and checked the cheaters’ website out. To no one’s surprise what she found a) was repulsive and b) confirms that New Zealanders have but one degree of separation.

So here is Dean Martin (because he was a member of the Rat Pack) with Hank Williams’s “Your Cheating Heart”:


Paul said...

I ate a large bowl of coleslaw this morning; a silly thing to do, as it turned out.

Stephen Stratford said...

Not necessarily silly but maybe unnecessary. Depends what time this morning, really.