The BBC Announced!
The BBC announced–and continue to announce all over the world–that red meat increases bowel cancer by 20%. Now lets look at the statistics they refer to:
- “For every 10,000 people in the study who ate 21g a day of red and processed meat, 40 were diagnosed with bowel cancer
- The comparable figure for those who ate 76g a day, was 48″
And they have a bunch of bacon on the cover… this one:
Image from the article. In case you are not aware, bacon’s first classification is that of processed meat.
To summarize: 40 people eating less red meat out of 10,000 get bowel cancer and 48 increased red-meat-eating people out of 10,000 get bowel cancer. The suggestion is that this is a 20% increased rate in bowel cancer rate as a result of eating more red meat; so the suggestion is that red meat causes cancer.
Mind you, BBC is not the only one announcing this fake news. here are some other news stations/papers that announced the same thing: The Guardian, The Times, iTV, Independent, MedicalExpress, CNN, etc.
The original study can be found here.
High School Statistics
Although the title on my blog is “Clueless Doctors”, in this instance I should rename my blog to be “Clueless Scientists” and “Clueless Newsflash”.
Do news anchors complete high school education? More importantly, do scientists that suggest such statistical conclusion have the high school education? What on earth are they doing as scientists if they cannot even calculate a percentage properly? 40 people out of 10,000 versus 48 people out of 10,000 is not 40 people versus 48 people! Man! Where did they go to school?
Please read the following statement, this was on BBC this morning, and then answer the 5 questions I ask:
“40 people out of 10,000 who eat meat get bowel cancer. 48 people out of 10,000 who eat read meat get bowel cancer. Therefore red meat increases bowel cancer rate by 20%”
- Is the above statement correct in your opinion?
- if yes, why so?
- if not, why not?
- are you going to stop eating red meat?
- have you any idea of the difference between red meat and other meat, like say poultry?
The answers are amazing (read some on my FB pages and feel free to place your answer there as well). Nearly no one dares to approach the statistics–clearly they don’t understand. Some specifically conclude that it is indeed 20% increase!
So what does this study actually mean by the 20% increase?
The Actual Outcome
The increase is not between 40 and 48 people but 40/10,000 (that is 0.004) people and 48/10,000 (0.0048) people. This is a major error in transferring assumptions. Assuming that is 40 people didn’t get bowel cancer but 48 did, we assume the difference is the 8 people, which is 20%. But this is not correct! The original numbers are given per 10,000 people; it is a ratio and not a number. The actual difference is 0.0008 people.
Given that we are looking at 40 people out of 10,000 people versus 48 people out of 10,000 people, the true increase is from 0.40% to 0.48%, which is a whopping 0.08% increase–spelled out: a zero point zero 8 percent increase–marginal and insignificant, close to zero.
Some people asked me to derive this:
40 out of 10,000 people is 40/10,000=0.004–this is the ratio of the people getting bowel cancer from the non-red-meat-eating group. Not 40 but 0.0040
48 out of 10,000 people is 48/10,000=0.0048–this is the ratio of the people getting bowel cancer from the red-meat-eating group. Not 48 but 0.0048
The simplest way to look at this is that 8 was the difference between the red-meat-eater and the non-red-meat-eater groups but each contained 10,000 people, so the actual percentage is 8/10,000=0.0008 or 0.08%.
Could the 20% ever be correct?
If we compared 48 people who never ever ate red meat (ever in their lives) with 48 people who ate only red meat (all through their lives) and 40 got bowel cancer in the non-red-meat-eater group while all 48 in the red-meat-eater group got bowel cancer, then the increase in bowel cancer rate is 20%. Note, we still cannot tell if that was from red meat so we cannot jump to causal conclusion! That is because if they have eaten other things–even if the same other things–their genetics, location, etc., are all factor in diseases.
When can we tell if red meat caused the bowel cancer?
If we had in group one 48 people eat nothing at all other than non-red-meat all their lives (meaning no veggies, fruits, seeds, nuts, coke, juices, grains, whatever.. nothing. just non-red meat) and in group two 48 ate nothing all their lives only red meat (meaning no veggies, fruits, seeds, nuts, coke, juices, grains, whatever.. nothing. just red meat), and they were genetically completely identical and lived in the same room all their lives, interacted with absolutely no one, and only 40 ended up with bowel cancer in group one but all 48 in group 2, then we can say that there was a 20% increase in bowel cancer from red meat.
What do People Think is the Outcome?
People are greatly confused by statements like “20% increase”, and rightly so, because it is seldom used correctly. It is also hard to relate to for the novice in statistics and if the error is huge, as it is in this article, it amplifies the confusion. In this case, as I have shown, the 20% increase is really 0.08% increase, so miniscule.
Placing the information into context, we find that the 0.08% increase in bowel cancer risk–supposedly associated with and not caused by red and processed meat-eating–is preferable to some other daily risks we all take.
For example, we have much higher odds of ending up with skin cancer just by living:
“the lifetime risk of getting melanoma is about 2.6% (1 in 38) for whites, 0.1% (1 in 1,000) for blacks, and 0.58% (1 in 172) for Hispanics” (here)
Or the odds of ending up with breast cancer:
“The absolute risk of developing breast cancer during a particular decade of life is lower than 1 in 8
The younger you are, the lower the risk. For example:
- If your current age is 20, the probability of developing invasive breast cancer in the next 10 years is .06%, or 1 in 1,732. This means that 1 in 1,732 women in this age group can expect to develop breast cancer. Put another way, your odds of developing breast cancer if you are in this age range are 1 in 1,732.
- If your current age is 30, the probability of developing invasive breast cancer in the next 10 years is .44%, or 1 in 228.
- If your current age is 40, the probability of developing invasive breast cancer in the next 10 years is 1.45%, or 1 in 69.
- If your current age is 50, the probability of developing invasive breast cancer in the next 10 years is 2.31%, or 1 in 43.
- If your current age is 60, the probability of developing invasive breast cancer in the next 10 years is 3.49%, or 1 in 29.
- If your current age is 70, the probability of developing invasive breast cancer in the next 10 years is 3.84%, or 1 in 26.” (here)
Was this study even a research?
It appears that this study also made some other major errors that are hard to relate to. This “research” was not based on any experiment, and so it is not even a research. It was data-mining of result in databases of people whose diet was not controlled at all and so no one knows the true cause of their bowel cancer–for all we know the bowel cancer may have been caused by high carbs, lots of junk and processed foods, or anything else. Absolutely nothing can be determined from data that was collected from food surveys; one can only observe associations. Associations are not causation.
With such misleading statistics, making people believing anything, one wonders of the ulterior motive. I often wonder if news reporters are really this dumb making announcements like this to deliberately mislead the people… they seem to be doing a good job at getting regular people totally confused. And the scientists? In all honesty: they are not scientists. If they have a PhD, this would be the ideal time to question the merits by which they have earned them.