(Source: Aspiring Gentleman)
A recent study used meta-analysis to determine that meat is a carcinogen -- but how much danger are we really in?

In addition to hastening the (seemingly) inevitable onslaught of global warming, triggering the occasional allergic reaction, helping sedentary souls to pack on the pounds, and by the accounts of some perpetuating a culture of animal abuseThe World Health Organization's (WHO) International Agency for Research on Cancer (IARC) has blamed red meat for yet another evil -- cancer.  That's the key finding of a new paper by WHO leaders that was published in Lancet, one of the most prestigious medical journals.

I. Tepest in a Teapot

But wait before you cry foul, I'm not here to demonize your plate of bacon.  Far be it from me to do so.  In fact I'm here to point out why the fear this story provokes may, in fact, be overstated.

From a disclousure standpoint I feel that I should state that I am a vegetarian (but not a vegan).  It might come as a surprise though that when it comes to this kind of study I feel mostly a dispassionate objectivity.  I have no vested interest, as I'm really not the type of vegetarian who tries to tell you what to eat.  I eat what I eat because my family has a genetic history of hypercholesterolemia (basically, diet independent high cholesterol) cholesterol and a vegetarian diet (or at least a low meat one) helps to keep it in check.  But I digress.

And hence, while it might amaze some, this vegetarian is stepping up with a plateful of science to defend you red-blooded red meat eaters.

Red meat
[Image Source: plos-one]

What draws my interest to the IARC's classification is that it is producing a high degree of hyperbole in the media.  Quartz's Akshat Rathi and Deena Shanker, for instance summarize in a subheading for social media:

Bacon's as bad as booze, x-rays, and air pollution.

Here it is in the page code (nice ASCII banner, btw!) in case they swap it out for something a bit less frisky:

Quartz red meat

As someone with a graduate background in chemistry and physiology, that raises my eyebrow.  And a cursory read of the piece itself reveals that from a pedantic standpoint the headline is at best only partially accurate.  It is true indeed that cured bacon -- which does dominate the majority of America's consumption -- has been classified by the WHO as a Group 1 carcinogen.

Shocklingly Delicious
Bacon on the 'ol fry pan?  Watch out for dem carcinogens. [Image Source: Shockingly Delicious]

Other Group 1 carcinogens include: So what do booze, x-rays, and air pollution have in common?  They're all manmade and they're all well known to have health risks.  Now if I were trying to have a bit of tongue-planted-firmly-in-cheek sort of fun with that list it would be far more glorious to declare:

Bacon's as bad as birth control, wood dust, sunlight, and mineral oils.

I'm guessing the average person wouldn't think of those as carcinogens and would be a bit perplexed.  But that's the fun, isn't it?

Brith Control
Birth control pills are also considered modestly carcinogenic. [Image Source:]

It appears however, that the authors (or the editorial staff) over at Quartz were having a different, more obvious kind of fun with their readership.  Different strokes, I guess.

But either way the headline is in part wrong, as I said, because not all bacon is processed.  Bacon (generally) refers to a specific cut from a specific treif (hebrew for "not kosher!") animal.  But it does not always refer to curing with nitrites and nitrates to artificially extend shelf life.  Even if you don't get all fancy pantsy and get organic bacon there's plenty of "natural" bacon brands that are uncured [example].

Nitrates vs. uncured
[Image Source: RealFoodCo.]

Why cure?  Well first, uncured bacon is generally processed with sea salt, which can be more expensive both from a materials and time perspective versus chemical curing.  Cost aside, the main difference is shelf life.  wiseGeek claims that uncured bacon lasts for roughly a week, while cured bacon can last several months.

II. Food Poisoning Risks Encourage Processing

Here's where I get to play devil's advocate to for my meat eating friends.

Let us consider the primary cost if we switched to all uncured meat -- food poisoning.  Food poisoning kills over 3,000 people annually in the U.S. and causes 128,000 Americans to be hospitalized annually according to the Centers for Disease Control and Prevention (CDC).

Salmonella and other nasty bugs can quickly take up residence in uncured meats. [Image Source: CDC]

Now let's consider the cost of eating that processed meat.  Processed meats, according to metanalyses (numerical/statistical combination of the results of multiple studies) by the WHO, raise cancer risks.  The WHO's analysis on unprocessed mammal meats, which it refers to as "red meats", found a similar but lesser risk.  Hence while the processed meats earned a Group 1 rating, the red meats earned a Group 2A status.

Other Group 2A carcinogens include: Of course, the risk isn't the same for all types of cancer.  Both kinds of meat primarily increase the risk of cancers of the digestive tract, most notably colon cancer.  The report states that for every...
  • 100 g of red meat, your risk (of colorectal cancer) is raised 17%
  • 50 g of processed meat, your risk is raised 18%
Colon cancer
An artist's depiction of colorectal cancer. [Image Source: QuantumBooks

First, I'll spare you some confusion and state that a "rasher" of bacon" means different things in different regions.  But in the U.S. generally it's referring to a cut that is further subdivided into "slices" or "strips".  4 average cut strips weigh approximately 1 oz. or 28 g [partial source].  So that takes you a quarter of the way to the above stated risk.

Quartz claims:

By one estimate, US men eat an average of 110 g of red meat a day. Women eat an average of 70 g.

Here we see some pretty interesting rounding going on.  The source -- the North American Meat Institute (NAMI) -- gives the following numbers:
  • 55% of meat consumed by Americans is red meat
  • Men consume 6.9 oz. daily of meats of all kinds daily
  • Women consume 4.4 oz.  " " " daily
  • Men consume 3.8 oz. / 0.2375 lb / 107 g of red meat daily
  • Women consume 2.4 oz. / 0.15 lb / 68 g " "

Pig cuts
[Image Source: Quality Beef Co.]

For the men, I suppose that you could round to 2 sig figs and arrive at a similar number.  But for women a simple analysis gives 68 g (preserving the sig figs in the percent red meat and the oz. per day).  Maybe I'm being pedantic, but to me that's not professional napkin math -- it's more like middle school "almost" math.  Remember tiny errors have led at least one multimillion dollar space probe to crash at the American taxpayer's expense.  So math is important!

Also notice the other kinds of meat have 3 significant digits which implies that 55% really means 55.0%.  So 55.0% and 6.9 / 4.4 oz for men and women are our inputs.  Now let us assume that implies a 20% confidence in the smallest digit.  So +/-0.2% and +/- 0.2 oz.  Propogation of uncertainty...

Propogation of uncertainty

....for multiplication based on my crude estimates gives:
  • Men: 3.795*((0.2/55.0)2 + (0.2/6.9)2)0.5 = +/- 0.1 oz. (+/- 3 g)
  • Women: 2.420*((0.2/55.0)2 + (0.2/4.4)2)0.5 = +/- 0.1 oz. (+/- 3 g)
I find it interesting just how many people in scientific fields -- not just news writers -- are ignorant of statistics, particularly propogation of errors.  You see this as often propogation of errors gives more digits than the good old fashioned "sig figs" you were taught in middle school, high school, etc.

If you don't know what you're doing I would argue it's better not to round off when unit converting, as you'll at least lie close to the average.  in this case it would be troubling to think there's less that a 20 percent confidence in either estimate.  So really a (well trained) scientist would write:

By one estimate, US men eat an average of appr. 108 +/- 3 g a day.  Women eat an average of appr. 69 +/- 3 g a day (note: uncertainty is estimated).

...rather than...

By one estimate, US men eat an average of 110 g of red meat a day. Women eat an average of 70 g.

Interesting rounding, I should add.

IV. On the Importance of Proper Rounding

Aside from costly space program mistakes, it's worth being pedantic here because again we see the chosen skew -- trending towards sensationalism.  Granted the difference isn't huge, but if we're talking cancer risks, every little bit counts, right?

Second, the numbers may be a bit high, which wouldn't be shocking given that they're published by an industry trade group.  Conveniently collated by The Wall Street Journal (WSJ) is a wealth of pertintent data on 2010-2012 consumption collected by the U.S. Department of Agriculture (USDA).

USDA results
[via: WSJ]

Granted, it doesn't subdivide by genders but it gives the following figures
  • Americans annually consume on average (from 2012, a leap year):
    • 71.2 lb (+/-0.2 lb seems reasonable) of red meat 
    • 54.1 lb (+/-0.2 lb) of poultry a day
  • Adjusting for vegetarians @ 5 +/- 0.2% of pop. at any given time (orig. x (1/0.95)):
    • Red Meat: 75 +/- 4 lb per year
    • Poultry: 57 +/- 4 lb per year
...and finally that gives:
  • Red Meat: 0.2048 + 0.0008 lb per day, on average (92.9 +/- 0.4 g)
  • Poultry: 0.1556 + 0.0007 lb " " " (70.6 g +/- 0.3 g)
Meat per capita

[via: WSJ]

So not considering fish:
  • The USDA's 2012 red meat to poultry ratio is roughly 1.28:1.
  • The NAMI ratio is 1.49:1
That makes me suspicious of NAMI's numbers.  But let's first check our percentages with some crude estimation.  Using NAMI's numbers for fish, and the USDA's numbers of red meat vs. poultry we get:
  • Both genders percent of daily meat:
    • Red Meat: 52.2 +/- 0.3% (vs. 55.0% NAMI)
    • Poultry: 39.6 +/- 0.3% (vs. 36.8% NAMI)
USDA sales
[via: WSJ]

Clearly the level of disagreement between those values is more substantial than my estimate of uncertainty

I have a sneaking suspicion that the same trend we see in income vs. meat consumption is accentuated for red meat, which tends to be more expensive than poultry.  Hence I have a sneaking suspicion that NAMI -- who doesn't explicitly mention what year its daily weight estimates come from -- is using data from the peak consumption years from 2002-2007.  If I'm right that means that the NAMI data isn't the most accurate and is skewing consumption a bit higher.

V. Is a Meat by Any Other Name as Sweet?

Still, in the grand scheme of things one of Quartz's points in piece does hold true -- for men that still places the average American male at a roughly 15% elevated risk if they eat no unprocessed meat but hit the average red meat consumption.  Assuming that half of poultry consumption is processed that works out to (70.6*0.5/50*18 + 92.9*0.5/100*17+92.9*0.5/50*18)*(2*4.4/(4.4+6.9))...
Meat percent

For men and women consuming the average amount of meat for Americans, with at least 50% of the meat coming from a processed source the risk is:
  • ~46% higher risk of colorectal cancer for men
  • ~29% higher risk of colorectal cancer for women
Yikes.  Statistically, not only did Quartz inflate the red meat consumption, but it overlooks the glaring fact that much of the poultry consumption is processed (e.g. turkey cold cuts).  Remember USDA "safe" levels are nitrates are based on weight [source; PDF], so for the same weight unit of processed poultry vs. processed meat, in theory the carcinogen risk should be similar if the preservatives are primarily to blame (which appears to be true).  In reality the metanalysis's commentary re: red meat suggests the risk of processed red meat is likely somewhat higher, perhaps 20 or 30%, but not double.  Still that means that the poultry processing can't be discounted.

But wait... canned vegetables are ALSO oft processed with nitrates/nitrites at similar ppm [source].  So why are meats being picked on?

canned vegetables
Similar artificial additives are found in may canned vegetables, as well. [Image Source: TVe Jundiai]

Clearly there's a lot of moving variables here, so meat lovers can take comfort in that there's likely some skew in the relative risks.  Still, let us assume there's some risk.  Does that mean that the Quartz's assessment:

[The WHO] delivered a damning report linking meat consumption and cancer, in particular a strong link between processed meats and colorectal cancers.

To some extent yes, to some extent no.  The biggest flaw I see in this whole rigamarole is that most of the studies used involved Americans -- a population who tend also to come up short in the exercise department.  The tastier cuts of red meat are clearly higher in fat than your average piece of poultry.

That leads us to the most critical questions -- questions that have largely been untread by other writers: I would hypothesize that the answer on the obesity angle is yes.

It's tempting to suggest hormones are to blame and indeed at least one recent study published by the American Society of Clinical Oncology (ASCO) says as much.  Still that doesn't definitvely explain the difference between red meat and poultry, as poultry are reportedly seeing much higher levels of hormones.

The study I mention above reports near zero levels of the two kinds of estrogen -- estradiol-17β (E2) and estrone (E1) -- in the meat of beef and some chicken meats.  But chicken fat saw very high levels (E2: 0.021 ng/g ; E1: 0.055-0.066 ng/g), as did some chicken meats (0.016 ng/g).  Beef fats were found to sometimes contain up to (E2: 0.014 ng/g; E1: 0.008 ng/g).
Chicken fat card
[Image Source: Listal]

Clearly the large differences in chicken estrogen suggest that some of this may be due to hormonal treatments to the livestock.  However, the gap between chickens and cows is high enough to suggest that chickens may simply naturally have a higher concentration of estrogen in their fat, as well.
  • A regular dose birth control pill contains roughly 30-35 µg of estrogen [source; PDF]
  • That's the equivalent to roughly 345-400 kg of chicken fat
  • The USDA estimated that roughly 71 g of poultry are consumed per day
  • And journal articles note that chicken is anywhere from 3-14% fat by weight
So (71 * 0.086 * 30 * 0.08% + 71 * 0.016 * 30 * 0.92%)...

Do the math and you find that the average American may be getting as much as 46 ng (~0.05 µg) of estrogen per month from poultry.

VI. Chicken Limbo

While that roughly 1/600th the dose of a birth control pill there's some suggestion that perhaps the extra estrogen may subtly disrupt hormonal systems leading to elevated risk of hormonal cancers.

Chicken surprise
Surprise! Some poultry may actually be worse for you then red meat as it is processed with the same amount of chemicals per unit weight. [Image Source:]

Again, if this emerging concern proves accurate that would again indicate the relative danger from red meat is perhaps being overstated, as poultry -- even uncured -- after all, poses a much higher hormonal risk.

At the same time this indicates pretty conclusively that the source of the baseline risk in uncured red meat lies in cholesterol and fat.  For those who don't have the genetic predisposition to high cholesterol (like I do) this may be a wholly fixable issue.  Eat uncured red meat whenever possible and exercise.
exercise chick
[Image Source: Gamers Graveyard]

I would argue that given the overly sedentary demographic the study overlooks my question about potential benefits of red meat.  For those without cholesterol risk factors, ethics and religion aside, red meat is one of the better protein sources.  While the higher levels of fat make it a health risk for the more sedentary, for runners and other high intensity athletes, the extra protein and energy may actually help to fuel workouts and reduce injury.

Michael Phelps
If you're Michael Phelps, cold cuts fuel your workout recovery.  If you're Jared Fogle you get cold cuts as the prison lunch. [Image Source: WebMD]

In the long run red meat may actually help combat many forms of cancer, as exercise and a healthy body mass has been generally tied to lower cancer rates, as well as better health in general.

That said, I will add that meat is not the perfect protein.  While it has all the essential amino acids (unlike soy) it lacks the nonessential acids in premade form.  For that reason eggs are probably a better protein option for athletes, although meat is a close second.

Eggs wide
Eggs are the most complete protein. [Image Source: Kellie Coates Gilbert]

Soy is a decent protein source, but it suffers from the same estrogen concerns as chicken.  Soy protein has roughly 102 µg of isoflavones which can be converted in the body to phytoestrogens.  Granted, it's not as cut an dry as that... 1 g of soy powder clearly isn't like taking a birth control pill.  

Isoflavones have some health benefits.  Ultimately they are transformed by the body into the female sex hormone, estrogen. [Image Souce: Skin Care by Louisa]

Likely the phytoestrogens -- slightly different chemically see slightly lower binding affinities.  And likely much of them are sequester or broken down prior to being ever converted.  Hence the impact of 1 µg of isoflavones is likely some orders of magnitude lower than 1 µg of E1/E2.  Still, some small amount is converted so whether its 10 ng E1/E2 per 1 µg isoflavone or 1 ng E1/E2 per 1 µg, you're still talking about a slightly higher risk than with chicken -- the highest risk meat.

sex hormones

Sex hormones of the human body. [Image Source: Wikimedia Commons]

Finally, we must consider the potential causation and bigger picture with regardds to processed meats.

Clearly nitrates/nitrites are carcinogenic.  But considering that they're also found in many poultry and vegetable sources, villainizing red meat for containing similar levels of this cancer-causer seems to be missing the more serious issue.  Now meat is more dense than vegetable food sources, so per volume you probably perhaps an order of magnitude less nitrites in canned vegetables, etc.  But the processed poultry angle is being particularly ignored by the media in their rush to judgement.

Then again, perhaps the food poisoning aspect gives cause for pause.  Again, annually (from the CDC):
If we switched to all uncured food how much higher might the incidences of severe food poisoning creep?  One might estimate as high as 5-10 fold.  That could take the death toll closer to that of colorectal cancer.  So reducing the risk of the cancer 10-15% at the expense of raising food poisoning by up to an order of magnitude may not save lives.  Or to put it more succinctly cutting nitrites out of commercial meats may not save lives.

If there's one thing the study does show clearly its that nitrites and nitrates -- be it in processed poultry or red meat -- should be avoided when possible as they raise your cancer risks modestly.  However, in locations where hygiene may be the more pressing concern (say the local sandwich shop), the cold cuts may carry a lower risk when you factor in the risks of food poisoning related illness and death.

Sensationalism sells
[Image Source: Radio Dialogue]

To summarize:
  • Quartz made modest mistakes in its math.
  • It picked overly high statistics.
  • It missed the poultry angle.
  • It offered no expert commentary on potential causative routes.
  • It cherry-picked its analogies versus more inocuous seeming ones (e.g. "sunlight").
  • It called the report "damning" when all things considered it's more "troubling" or "food for thought".
But if the Quartz writers had some concious or subconcious beef with the beef -- or red meat in general -- they aren't alone.  As Americans love their red meat, questioning that love provokes strong reactions and clicks.  My criticism isn't meant to pick on the Quartz commentary in particular -- rather I'd say I expect that commentary to be reflective of the general kind of writeup we see on the topic.

One of the better moments of the Quartz piece is seen nestled amidst the minor discrepancies.  University of Oxford Cancer Research UK director Tim Key is a voice of reason remarking:

This decision doesn’t mean you need to stop eating any red and processed meat. But if you eat lots of it you may want to think about cutting down

Indeed, that's how I perceived the study as well.  It is neither daming nor relieving.  It's somewhere in between.


[Image Source: Nature Reviews|Drug Discovery]

Nitrates and nitrites are believed to be responsible for the elevated cancer rates in individuals that eat a lot of preprocessed meat -- even of the lean variety.

But as I've laid out here there's plenty of risks that seem healthy enough -- mineral oil, wood dust, sunlight, etc. -- and there's plenty of seemingly healthy alternatives with risks of their own -- namely the hormonal issues with poultry and soy, nitrate/nitrite levels in processed poultry and canned vegetables.  It's nice that the WHO is bold enough to state that red meat -- in America's current lifestyle at least -- may adversely impact our health.

But it's equally important to identify where that harm likely lies and moreover to view the big picture.  That big picture is that life is a risk.  As a vegetarian with a very clear cut risk from red meat, I say the good news for my meat lover friends is that the risks here are far more murky and clearcut.  I would argue they're far from damning.

this is life

Go ahead and chow down, meat fans -- these statistics are interesting but paint too incomplete a picture to be used as justification for radical diet change.  Even the seemingly most crystalline perfect criticism -- the nitrites/nitrates issue is found to be flawed when examined in the context of food poisoning, a competitve risk.  Thus I would be inclined to say anyone making overly bold statements disclaiming this study as "rubbish" or overstating its findings as "damning" share something in common -- they either:
  • Haven't done their research
  • Are failing to grasp the science/math required to put this all in context
  • Have a vested interest in promoting or damaging meat sales or processing

The answer, my meat eating friends, is that we have a lot of intereresting statistics but a murky big picture.  And THAT is the honest truth from a scientist's vantage point.

Source: Quartz

"There's no chance that the iPhone is going to get any significant market share. No chance." -- Microsoft CEO Steve Ballmer

Most Popular ArticlesiSO 11 is Almost Ready:
September 19, 2017, 7:48 AM
We will Get 5G Phones in 2019
September 19, 2017, 6:49 AM
Intel Coffee Lake release date
September 10, 2017, 6:07 AM
Google Pixel 2XL will reportedly cost less.
September 20, 2017, 6:17 AM
Uber faces ban in London – Loses License next month.
September 22, 2017, 7:02 AM

Latest Blog Posts

Copyright 2017 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki