Wednesday 8 November 2017

Eating more meat MAY increase your risk of stomach cancer by 4 thousandths of one percent. Yikes!


Thus raising the risk by 4 thousandths or one percent.  0.004%.
From 0.0211% to 0.0248%. Yikes!
Is the above not the most blatant misrepresentation?
LATER (10 Nov): I changed the headline of this post, reporting the results of research on the relationship between processed meats and stomach cancer rates. They had headlined it with the much more sexy, newsworthy, accurate but misleading. eating processed meats increases cancer risk by 18%.
If you increase something tiny by a tiny percentage point it’s still tiny. That’s the nub of it. That’s the real risk of stomach cancer from meat eating: tiny. 

Full disclosure: I've used all meat and not just processed meat (PM) in my analysis here.  That's because finding world figures for PM are, for me at least, hard to find.  And for sure there's going to be increased risk from PM vs all meats, if only because of the salt content. To balance that, though, I've used figures for only the top 20 countries for stomach cancer rates.  If we took all countries' rates, they would be lower, so this balances out.  In any case, I'm after a illustrative pattern here, not exactitude.  I'd be surprised if the real figures were very different from these.  And the sum of that it this: the risk of stomach cancer increase through normal use of PM is tiny and insignificant.
********
I got talking with a friend recently about "The China Study".
I recalled the book, and think I may even have it somewhere.
It hit the world's bookshelves with a great thwack in 2005 and was an immediate hit.
Its a book about diet; specifically in China.
It looks at the diets of tens of thousands of Chinese in different counties.  The difference in their health and longevity are assumed to be due to diet, since all other factors -- ethnicity, smoking and drinking rates, etc -- are taken to be the same.
Its strongest conclusion was that the more plant-based diet, the better health and longevity the population.
Bill Clinton bought into it, just after his heart operation, and became vegan.
I took issue with some of the study; but before getting to that, a related matter.
Around a year ago, a study by the American Institute for Cancer Research and the World Cancer Research Fund purported to show the danger of eating processed meats (PM) like salami and ham.
The study showed, it was breathlessly reported, that if you ate a lot of PMs you were 18% more likely to get stomach cancer.
Many people vowed they would henceforth avoid PMs.
My own first thinking about this was to have a look at the stats.
Specifically for stomach cancer.
What is the rate of stomach cancer in our societies and what does a 18% increase mean?
The short answer is:
NOT MUCH.
The average rate of stomach cancer in the 20 worst countries in the world: 21.09 per 100,000.
That's 0.0211%. (rounded to the nearest 10,000th percent)
If that increases by 18%, the rate of stomach cancer becomes 24.9 per 100,000.
That's 0.0248%.
From 0.0211% to 0.0248%, is the true story. We're talking hundredths of a percent here.
If you put it that way, does that cause you any worries, any stomach pains?
I'm guessing not.
Yet the organisations that reported the findings decided to report: 18% increase.
Everyone saw that headline figure, freaked out, and sales of PMs plummetted.
I'm not sure whether they were being duplicitous or naive, but I'm going for duplicitous.
After all, to report the incidence of stomach cancer may (and yes, it was only "may") go up from 0.0211% to 0.0248% is not very earth-shattering.
I did my own bit of excel spreadsheeting.


Here are a few charts I generated, to show how the figures can be represented or misrepresented:


Chart 1 shows the relationship between meat consumption (not just PM), and stomach cancer rates, for the 20 countries in the world with the worst stomach cancer rates.
The correlation coefficient is 0.26, which is below "moderate", tending towards "none".  That is, there is marginal to no correlation.  The trendline above reflects this in the moderate slope: you can double your meat consumption, from, say 40 kg/year to 80 kg and the result may be an increase from 21 to 23 per 100,000 or 0.021% to 0.023%.  That is, an insignificant increase in risk.


Chart 2 shows the same info as that above, but this time I've removed the three Outliers:, Japan, Korea and Mongolia.  Why do that? Well, because an outlier is usually an outlier because of different factors than those affecting the overall correlation.  In the case of those three countries, because of high rates of smoking and salt intake, both of which identified in the study as factors in stomach cancers.  In other words, if you live in one of the 17 worst countries in the world for stomach cancer (20 minus the three outliers) and you eat more meat,  it will have exactly zero effect on the rate of stomach cancer.
The correlation above = zero.


Chart 3: is just to highlight the percentage rate of stomach cancers, in percent instead of per 100,000 terms.  And changes the meat consumption from Kg per year to grams per day.  Feel better?


Chart 4: is the same as the Charts 1 & 2, but with the axes inverted and with no trendline. When you just eyeball it, you're going to judge, I'd guess, that there's not much effect on stomach cancer by the amount of meat eaten.

Note: that I'm using figures for meat (not just PM) consumption, because I couldn't find figures for PM.  But meat is still a proxy for processed meat consumption. Moreover, the organisations doing the study have admitted that they used virtually unrealistic figures for PM consumption, something like salami and ham eaten in large amounts every day of the year.

In sum: I think that the whole business of PMs give you stomach cancer is a crock. Yes, there's an increased risk. But the risk increase is tiny, insignificant.

So how does all that relate to the China Study?

I'll get to that in a later post.....