Oy vey. Statistician/statistical geneticist here. The second article is too harsh on the first. The point is subtle and hard to express: carcinogens set the odds, but importance of the part of the risk due to "biological luck," the part that comes from the mutation chain, means whether you individually are affected is random in a way that makes it hard to bother protecting yourself.
Let's say (arbitrary numbers) having poisonium in the food supply increases the lifetime population rate of cancer of the thingamajig from 0.11% to 0.12%. In the US, banning poisonium will save 300 million * (0.12%-0.11%) = 30,000 people, very predictable and important. However, if you personally go on a poisonium-free diet, you decrease your personal risk by 0.01%. You chances go from about 1 in 1000 to about 1 in 1000... so maybe if it involves any effort, you should spend your time doing something else to improve your health.
It's different if the carcinogens targeted specific vulnerable groups. If there was one gene that makes you instantly get cancer at the first whiff of poisonium, we would screen everyone for that gene. The difference between the 0.12 and 0.11 would be due that gene. Everyone with that gene would be put on poisonium-free diets. Everyone else can eat all the poisonium they want, and would still have 0.11 risk. (This is more or less the situation with the active ingredient in aspartame sweetener, though the disease is not cancer)
I don't like the use of the word "luck" in these descriptions, since both genetic and mutation chain randomness are a kind of luck. Also, the studies give quality numbers across many cancers, but the basic concept has been a mainstream model of cancer risk for a long time.
Absolutely nothing about the data collected related to different human populations just different types of tissue.
Further, I don't see anything that relates to % of biomass of the tissue just total cell divisions. Presumably the skin, lungs, a digestive track tissues are at high risks because they compromise a lot of tissue, it divides rapidly, and it's not protected from environmental factors ex: sunlight/smoking/spicy food. However, I suspect the paper digs into things a little further to account for such factors.
Yeah, I'm talking about the different populations thing to illustrate an extreme example about the 'luck' concept. It's not directly related to the study.
Not sure I understand the % biomass issue. The tissues you're referring to - epithelial - are the ones doing the bulk of the cell division in adults, and the ones where you'd typically get the (early stage) cancers. Something like smoking increases your risk in two ways. You get a higher mutation rate (chemicals enter your cells and mess with DNA replication). You also increase the number of 'lung cells' you need to clean out and replace, causing more cell division. Obesity (and size in general) has a direct theoretical effect: more cells = more opportunities to divide = more risk events. Caloric restriction presumably is the opposite of that: do nothing and eat nothing, don't have your cells get replaced, and you reduce the number of risk events.
I think the deeper question in cancer math, if you get the single-gene level biologists and population level statisticians to talk to each other, is how much of the cancer is due to mechanics (number of cells, division counts) that are out of our control, mechanics that we can affect, genetics that are out of our control, and how much is genetics that are fragile but susceptible to prevention and early treatment. That's the sweetspot; that's what we ultimately want to find - genetic markers of things that are going to break but can be patched or protected or fixed.
That's why we want accurate models of risk - in different kinds of tissues - in the first place.
It seems like more cells = more risks of mutation = more risk of cancer. So, I would presume it's something that needs to be accounted for in their model.
However, based on the wording that seems to be total cell divisions and I don't know enough about rates of cell division or cancer to tell from that chart. In the end a gall bladder weighs less than 1/4th of a pound where the small intestine is ~3.5 pounds so it hardly seems like they can just ignore it.
Though, if it's really just cell divisions and not organ weight that matters then that's a much stronger finding IMO.
PS: And no I am not paying 20$ to read the actual article. Though if it's free somewhere I would like to read it.
The fact that species with more cells don't get more cancer is known as Peto's paradox. Perhaps it is also true at the tissue level. See http://en.wikipedia.org/wiki/Peto%27s_paradox
Bit of a tangent, but can you elaborate on that aspartame comment? I've found it difficult to research aspartame risks online, since the vast majority of what comes up appears to be unsubstantiated FUD, and the official sources (FDA etc.) don't appear to recognize any risk whatsoever. There does seem to be a general public perception that aspartame is "bad for you" in some way though.
As to the other "generally bad for you" ideas, they're not solidly substantiated. The idea that artificial sweeteners somehow lead you to eat more by messing up feedback regulation is actively studied, but I wouldn't say established.
Phenylalanine is one of the amino acids that you always need and have around in some quantities. But if you get too high a concentration in the brain, it becomes too easy to make the neurotransmitter Phenethylamine. That's got a list of possible effects for different people - could be a stimulant, could be a migraine trigger - but, especially since it doesn't do the same thing for everybody, there is no consensus for what it does. Nor have I seen good numbers on how much is too much.
I was quite surprised to find this in one of the most prestigious journals in the world. Vogelstein is one of the biggest names in cancer research, and no doubt that was the most important factor in its publication.
Some problems:
1. They leave out breast cancer and prostate cancer, two of the most common cancers out there. Most of the dots on their plot relate to a tiny proportion of cancers that people get. Furthermore, breast cancer has a very low mutation rate, which also casts doubt on their hypothesis about accummulation of somatic mutations on stem cells.
2. I ran their data using just the number of stem cells and the number of cells versus incidence. The correlations are still good, about 0.5 for just cell mass and 0.67 for number of stem cells. It's hardly a revelation that a larger more cellular organ is more likely to get cancer.
3. For glioblastoma, a type of brain tumour, their data says that there are 0 stem cell divisions, because it is thought there is no cell renewal in the brain (which isn't even true, but anyway). This contradicts the whole hypothesis, but they still include it in there.
4. The lifetime incidence of various cancers is age dependent. For example sarcomas almost always occur in the young, medulloblastoma is vanishingly rare in the elderly, testicular cancer occurs in young men. Their hypothesis ignores all this, because their model only allows for the acquisition of mutations over time, which should give a unimodal relationship between age and cancer incidence.
5. There are cancers that have almost no mutations. Gene copy number changes are just as important if not more so than mutations. There is very little evidence that copy number changes accumulate over time in stem cells in a random fashion required by this model.
I don't think this warrants much attention in the end. This can join the 100,000 other unverified (or unverifiable correlations) in the cancer literature.
I want to reiterate the articles recommendation of the book "Bad Science" by Ben Goldacre [0]. I learned a lot from it and found it enjoyable--and it's available in Audiobook form.
Here's the deal - we know it's 'chance' (not really luck) which part of your DNA is damaged by any given environmental effect. However, some DNA is more important than others. In general, a little damage is not a problem and is easily 'taken care of' by the cell which caries that DNA - either by literally repairing the DNA, silencing that DNA's function, killing itself, or asking other cells to help kill it.
We know quite clearly that cancer is (generally) caused by a set of mutations - not necessarily in an order, but some orders are not successful. There are four or five genes which keep social order amongst the other genes. If you silence all of these, you get cancer. Chance has it's role in the roll of dice for which DNA gets damaged, but all the other parameters can be changed too - how many sided the dice are, how often the dice get rolled, and whether all the cells in the same tissue have correlated dice-rolls.
Cells that deviate from what they're supposed to do are either (in order), repaired, silenced, voluntarily commit suicide, or are killed. There are proteins (genes) that are the final judges for each of these processes - and have 'go, no-go' power. Only if all of these judges are killed do you get a cell that can do anything it wants - like replicate uncontrollably to the detriment of the host ('cancer'). Thus the statistics of getting cancer roughly follow the idea that you have to get random DNA modifications of those exact 5 genes, in a single cell. Lots of things can increase your random modification rate (UV, smoke, radiation, etc). Some of these things correlate though - and again, what hurts one cell, might hurt its neighbor just as bad. They're not entirely independent events. For example, losing your DNA repair machinery (this is what HPV does - it silences your DNA repair machinery) amps up the baseline mutation rate and makes further mutations more likely (dependent correlations then arise).
The Brca gene that has caused so much controversy in patent law (whether a test for its existence could be patented) and indicates whether a person might or might be susceptible to breast cancer, is the master repair technician of the cell. In people who have this gene in working order, the Brca gene signs off on whether the cell is in need of repair. But if the Brca is not it working order, cells that are in need of repair might not get it, and instead are allowed to more freely operate under non-optimal internal conditions. If you are missing or have a mutated version of Brca, you are missing one of the checkpoint processes.
So again, we quite clearly know of a handful of genes which do most of the master regulation of a cell's job - and if these jobs go unfulfilled - by having their blueprints be damaged by the environment - you have fewer and fewer mechanisms to prevent that single cell from runaway growth.
And most in the U.S. are deficient:
'Magnesium intakes for ~56% of adults in the United States are below the Estimated Average Requirement (EAR), the current measure of micronutrient inadequacy (the RDA is set at 2 standard deviations above the EAR).'
The Ames paper is worth the read through. Including this one on magnesium, 'In a study of 4,035 men followed for 18 years, the highest quartile with serum magnesium at baseline compared with the lowest had a 40% decrease in all-cause mortality and cardiovascular disease and a 50% decrease in cancer deaths'. He goes through other micronutrients as well.
You would be correct. Errors are not the same as cancer.
Cancer is generally a cell that can replicate uncontrollably on its own (most cells would die if extracted from their very particular environment). Your body has trillions of cells. Many have mistakes, errors and other issues. Most of the trillions of times the errors are very minor and in no way cause uncontrollable cancer - merely a less efficient cell. Only a countable number of times in trillions of cells dividing every day for 80 years are those errors precisely of the kind that would permit uncontrollable replication - cancer.
Aging is different yet. Copying of dna has a Turing issue. It's really hard to copy the end of a Turing tape when you have to hold that tape to copy it. So dna keeps roughly 60 replications worth of dead space at the end of each strand of dna. So even if the replication machinery drops a page or two at the back of the book full of blueprints, it doesn't matter so much. But after 60 divisions or so (a human's lifespan), the replication starts eating into the actual useful blueprints - dramatically increasing the chance of failing to copy an essential gene. These blank pages are called 'telomeres' and the Nobel prize was recently awarded for their discovery.
Aging is a lot of things, but it's not generalized cellular breakdown. Aging is more associated with the gradual failure of larger systems and frameworks in the body in ways that evolution has had no incentive to effect a repair strategy for.
i.e. wrinkles are associated with the loosening of the collagen support framework for the skin. This isn't cellular breakdown - it's just a system which there was never any point evolving a repair mechanism for beyond childhood.
the original paper is kind of confusing itself. Why would they essentially do a 1 variable regression? why not do a multiple regression with lifetime cancer risk as y variable and both number of cell divisions and smoking as x variables?
Exactly. The same old debate between statistical truth and deterministic uncertainty. What bothers me is that, meanwhile, the many people who have tried to get funding for years or even generations to study deep, complex mechanisms involved in DNA replication (on histones for example) didn't get much. Of course replication will eventually, on large numbers, begin to fail. Microprocessors do too. But would we say that when instructions get contaminated it's bad luck?
Wait, is this an article from the Guardian making fun of their own article on the same study earlier on ? See http://www.theguardian.com/society/2015/jan/01/two-thirds-ca... ? If that's the case, that's a novel way of doing journalism: publish crap first, then sell more paper by doing a critique of your crap.
Why do you assert that there is malice involved? That just makes no sense. What’s with this reflexive impulse to always, always, always loudly proclaim malice?
I just don’t get where this weird worldview comes from. If there is no malice involved then this is not a big deal. Wow, people working for the same publication nevertheless criticise each other. Oh my god, what a concept!
"If that's the case, that's a novel way of doing journalism: publish crap first, then sell more paper by doing a critique of your crap."
I wouldn't assume it was an intentional strategy. Papers and magazines have never been shy about critiquing themselves. (At least, the better ones haven't.) Back in the day, it wasn't uncommon to see an editor at the New York Times take a staff reporter or columnist to task in a sidebar, or even in a followup story. And the role of ombudsman used to mean something. (Whereas today, it's a very underappreciated and little-understood role.)
Second, we shouldn't necessarily presume that writers at the same paper are coordinating with each other. Say what you will about that as a business practice. But it's a fact of life at many publications, with distributed staffs, section-specific editors, shrinking numbers of full-timers, and ever-increasing numbers of independent contractors. I wouldn't attribute to malice what can better be explained by uncoordinated action.
[EDIT: This comment got a downvote, and I'm guessing it's because someone assumed I'm defending the Guardian here. Just to be clear, I'm not. When I say that people aren't always coordinating, I'm not offering that as a defense of lack of coordination. I'm offering it as a possible explanation. Clearly, a lack of coordination is a bad thing.]
>Don't they do peer reviews before publishing articles to the eyes of millions ?
No.
The closest journalism gets to peer review is fact-checking. I know of no newspaper or website that does formal fact-checking; the only times I recall being formally factchecked were when I've written for monthlies like Playboy, Wired magazine, etc. Here's a famous example of where fact-checking would have been useful: http://en.wikipedia.org/wiki/Stephen_Glass#The_New_Republic_...
At other news organizations there is, in theory, informal fact-checking. I've done that as an editor and reporter, and I've had that done to me by my editors and, more frequently, by the good folks working on copy desks. This is not necessarily a rigorous process, but it can flag more obvious problems, hyperlinks to the wrong web pages, information stated as fact that is no longer accurate, etc.
As <jonnathanson> points out nearby, some of these built-in safeguards have been eroded by the journalism crunch. Others have been removed by the perceived need to increase publishing speed. I know of one news organization that had 3-4 humans involved in reviewing a story circa 2002 before it could be published on the web. By 2010-2012, however, reporters were clicking the "publish" button in the CMS themselves for almost all stories -- and then sending a request to the copy desk for an edit after the story was live.
[Disclosure: I worked as a technology journalist, albeit one with a technical background, before leaving CNET/CBS last year to found http://recent.io/]
In theory, this is the function of the editor -- be it a section editor, a managing editor, or someone else who bears the title "editor." In practice, editors have been very hard hit by the journalism crunch. There are still plenty of fantastic editors out there; they're just not getting hired, kept on, or appropriately compensated for their contribution anymore. Editors' heads are often the first on the guillotine when the cost-cutting rounds begin. The editors who survive the purge are often wildly overstretched. Others are extremely green--college kids, hired for a third of what an experienced editor gets paid, under the misguided assumption that anyone with a journalism degree can successfully fulfill the responsibilities of an editor.
(For what it's worth: I don't know anyone on the Guardian's staff, and am making no specific accusations here.)
In journalism there is frequently the case that when a story plays out, the coverage then becomes about the story itself (will try to find a reference if I can). (This one hasn't played out yet so by the play book he really needed to wait a bit longer.)
I think you just see something like "one site - many people writing - no global review".
The article you link seems to be syndicated from "Press Association" (??) and was probably added by a scientifically illiterate journalist, editor or somesuch.
The OP here was written by actual scientists who have a blog-type of thing on TheGuardian.com.
Your link does not go to a Guardian article, it goes to a Press Association (PA) article published on the Guardian's website, and on thousands of newspaper websites.
Naturally the way one lives has an impact whether potentially he/she will get cancer. Smoking, drinking, eating certain types of food (manufactured wrong) etc. all have direct relation on cancer. It's never just luck who happens to get these horrible diseases.
Some of my friends get this argument all the time from their less mathematical business-manager bosses. "Sure, probability applies to playing cards, they're random, but it doesn't apply to business_process_X, because there things happen for reasons!" (I've even heard "sure, probability applies to baseball players, but it doesn't apply to website visitors"!)
Pretty much everything that we use probability to model has underlying causes and reasons, if we could only detect and measure them (at least until you get down to quantum mechanics), but probability can still be a useful way to model problems where either we don't fully understand the underlying causes, or the underlying causes are just too complicated and tangled to directly model.
There are over 200 different types of cancer. Not all of these are caused by contact with carcinogens. So, for these cancers it's pretty much "luck" whether you get them or not. Not smoking; sensible drinking; etc will not prevet you from getting them.
I think that's a bit unfair. Sure, we can play semantics and claim it's technically incorrect to say "smoking causes cancer" because if it did everyone who smokes would get cancer, and that isn't the case.
However, we are aware of quite a few known, and many probable, carcinogens.[1][2] Among women, breast cancer comprises 60% of alcohol-attributable cancers.[3] And to support the claim that certain foods, when prepared in certain ways, can become carcinogenic take a look at the research on Heterocyclic amines,[4] and acrylamide.[5]
Waiting for a strong causal link, especially when there is a high level of statistical relationships, can mean we might be waiting too long to take serious action.
If repeated studies show strong associations, like say between cigarettes and cancer, it would make sense to take some steps to prevent the potential cause.
Well, carcinogen is one those things that has been known for decades causing cancer. And yes, human-being can make the difference whether they have those unnatural carcinogens or not.
There are plenty of "natural" carcinogens too. Avoiding carcinogens may reduce your cancer risk, restricting yourself to things that are "natural" won't.
Whatever the official cause of cancer it will never be food additives, processed or poor quality foods, prescription drugs, air pollution, contaminants in the water supply, contaminated vaccines (I'm a believer in vaccines so don't go there!), new car smells, pesticide residues in food and clothing, fire retardant chemicals, xenoestrogens, secondary cancers caused by chemotherapy or radiation... it will just be bad luck.
Let's say (arbitrary numbers) having poisonium in the food supply increases the lifetime population rate of cancer of the thingamajig from 0.11% to 0.12%. In the US, banning poisonium will save 300 million * (0.12%-0.11%) = 30,000 people, very predictable and important. However, if you personally go on a poisonium-free diet, you decrease your personal risk by 0.01%. You chances go from about 1 in 1000 to about 1 in 1000... so maybe if it involves any effort, you should spend your time doing something else to improve your health.
It's different if the carcinogens targeted specific vulnerable groups. If there was one gene that makes you instantly get cancer at the first whiff of poisonium, we would screen everyone for that gene. The difference between the 0.12 and 0.11 would be due that gene. Everyone with that gene would be put on poisonium-free diets. Everyone else can eat all the poisonium they want, and would still have 0.11 risk. (This is more or less the situation with the active ingredient in aspartame sweetener, though the disease is not cancer)
I don't like the use of the word "luck" in these descriptions, since both genetic and mutation chain randomness are a kind of luck. Also, the studies give quality numbers across many cancers, but the basic concept has been a mainstream model of cancer risk for a long time.