After many mind-sapping years of trawling through the morass of health scare stories, I formulated a number of laws, one of which was the Law of Beneficial Developments: The intensity of the scaremongering attack on any new development is proportional to the level of benefit that it endows. Unbelievably, the Chlorine Scare has …
Great addition to the staff
I've been reading Numberwatch for many years and it's been a valuable addition to my reading mix. Like the always brilliant Lewis Page, John Brignell is a welcome addition to the staff of El. Reg.
Lovin' the time of cholera?
No doubt certain more radical environmentalist can regard calmly the rise in mortality that we could expect from giving up chlorination, but it might be stressful for the rest of us.
Headline: Dihydrogen Monoxide found in all cancer cells
Ban this dangerous substance before more innocents die.
For further detail, visit the campaign site at http://www.dhmo.org
This warning brought to you by the Minstri uv Ejukashun an Sines
If (as an analogy) you perform a random poll of 2000 adults and ask a question to which only 4 respond in a specific way, the claim that you survey'd 2000 people is still valid.
Why are you arguing that the study size was only 80, if it was in reality 400,000 of which only 80 were not healthy (in the specified way)?
Offer every household a choice of chlorinated or non-chlorinated water.
After a while, the non-chlorinated argument will just sort of go away...
Would the author care to explain...
why p=0.05 is described as an "abysmal standard of significance". While I concur with his loathing of data combing ('cos, as we all know, correlation doesn't imply causation), there's nothing wrong with stating your thesis, setting up your study and then seeing if the observed numbers match the prediction with small p; traditionally, we call this process "science" (although any true Popperian would agree that only if the thesis is disprovable does it have any value).
Anyone got the appropriate papers to hand and can tell me what Doll's significance was? I assume we're all adult enough to have got over the "no, smoking doesn't cause cancer, it's - erm - something else that smokers do and non-smokers don't" theory so beloved of Forest in the late 70s and 80s...
RE: Great addition to the staff
Second that, welcome.
Nice to read an analysis by somebody who is clearly an expert in this area. I sometimes think that the misuse of statistical techniques is the true curse of our times. The use by politicans and sociologists is bad enough (a closely overlapped group for anybody who has read the History Man), but the apperance of professional epidemologists is a modern pestilence. Of course there have been some stunning success stories (like the off-quoted one linking smoking with lung cancer). However, the attempt to use such techniques to find incredibly rare events is fundamentally prone to errors due to all sorts of real-world issues, factors they don't take into account, interpetation of results, data source errors. Even if the real risk changes in absolutely numbers is tiny, they can be dressed up into alarmist talk like 200% increases. So my chance might go from one in a million, to three in a million. Big deal - there are lots more real world risks that I can influence (like not riding a motorbike).
Of course the real truth here is that epidimologists have to find something, or they can't justify their academic salaries. If they can't generate scare stories, get in the media and wake up a few journalists or politicians in need of a cheap story then people might ask why they exist.
For those who want to read more about the malign effects of epidimology on medicine and "lifestyle" campaigns, then I'd recommend reading The Rise and Fall of Modern Medicine. It's not infallible but it puts a lot of this into perspective.
It's not as if this stuff can just be ignored. Channel 4 news recently ran a story supported by a "nutritionalist" describing a wekeend fried brekfast as "binging on fat" on the basis that this one meal could cause you to exceed your entire day's saturated fat allowance. This stuff is all based on government guideline numbers, often produced under political pressure, with next to no scientific evidence to support the numbers (admitted as such by those who came up with the guideline amounts on daily alcohol units). A dose of common sense is required here, but the opportunity to politicise policy decision making based on this soprt of junk is open-ended.
As for Peru, then I can personaly testify as to the effects of having rampant bacteria in your dietary intake having sucumbed twice to some such pestilence in the space of a three week trip. A bit of chorinated drinking water would have been the least of my troubles.
And don't start me on the subject of Greenpeace junk science...
So DDT wasn't bad?
I didn't know that.
Imagine that the question were 'Are spurs supporters more likely to commit murder than Arsenal supporters?'.
Now, go find 1 million people - and ask if they are Spurs or Arsenal supporters, and say you get 10 of each.
Then you ask if they are murderers - and you get two Spurs supporter and one Arsenal supporter.
Now, does that mean that you have a sample size of 1,000,000 people to say that Spurs supporters are twice as likely to commit murder than Arsenal supporters?
When you're looking at the affects of <x> on <y>, your sample size is the people that <y> applies to, not everyone that you could have asked.
Public Health Hazards!
Chlorine is nothing! There is a more insidious killer that is lurking in every home in the free world! Dihydrogen Monoxide threatens our wives and children each and every day of the week! Look it up!
We need immediate legislation to remove this poisonous substance from sale everywhere!!!!!!!
Junk science is everywhere!
The US government spends vast sums on "research" designed to find harmful effects of cannabis, dammit, no matter how long it takes.
Meanwhile in the UK, bicycle helmets are a scam: they actually make most cyclists less safe, but there's money in selling them and they shift the blame for accidents onto the person who got hurt.
A welcome addition to the Reg, indeed, Sir.
My favourite stats+ignorance=news
My favourite has to be the one I came across whilst doing Knowledge Management for a agro-chemical business. An attack on an already unpopular weed killer, still on sale in many countries today...
The headline read "Four Killed By Chemical X In South America!"
What the article didn't bother to print was the full story...
One day a helicopter was spraying chemical X across an illegal cannabis plantation in south america, it was known to kill cannabis plants almost instantly which is why the US and the local government were using it.
First the gardener guy appeared, shouting the odds, then two locals appeared and started shooting up at the helicopter. The helicopter crashed on all three of them and the pilot was killed in the crash too.
Hence the headline "Four Killed By Chemical X In South America!"
Not exactly the fault of the chemical company...
Paris, cause you would, wouldn't you...
The original study is shows no correlation
I'm looking at the paper from the journal Environmental Health which is the basis of the Torygraph news article.
Out of almost 400,000 births, only 43 suffered from anencephalus, so we're dealing with a very tiny sample to start with.
If we look more closely, we find there is zero correlation between the number of cases and the level of trihalomethanes in the mothers' drinking water.
The same is true of cleft palate and the other birth defects in the study.
The authors are clearly not statisticians, and Roger Highfield no longer deserves the title of science correspondent.
when apples appear on trees the weather starts getting colder. kill all apple trees and have summer all year round!
Was there an IT angle to this?
RE: So DDT wasn't bad?
No, it wasn't.
DDT should still be used to spray the insides of human dwellings in mosquito prone areas, as it has been proved to be a life saver and perfectly safe for the environment when not used to excess in areas where it is likely to quickly end up in aquatic systems.
Measure of exposure.
"Then there is the measure of exposure itself. How much of the dreaded fluid did the pregnant women drink? How did the boffins distinguish between a thirsty mother in a low dosage area and a non-thirsty mother in a high dosage area?"
And did they account for regular attendance at aqua-natal classes or other pool-based activities...?
that depends on what question you are asking and whether a negative answer has any bearing on your study.
Lets say you ask the 2000 people in your example 'how long did you have to wait for your hip operation on the NHS' and 2 said 'a year', one said 5 months and the other 1997 said 'I havent had a hip operation'.
It is true that you asked 2000 people, but only the answers of the three that matched the critera for your question are relevant, so your sample would be 3, not 2000.
it was used to battle maleria, but due to residential resistance it's no longer as effective.1 million deaths a year due to maleria - jee I bet that those 60 million dead people were pleased that rich folk didn't have to worry so much about little timmy eating some ddt on his crops.
And as with most of these things it's the poor people that die becouse they don't have a choice, not the rich ones who can afford quality health care, hygene, medicine, vacination and food (and the ones who spear head resistance to these things.)
Re: Public Health Hazards!
>What are some uses of Dihydrogen Monoxide?
They should update this to include interrogation techniques in Gitmo.
"The abysmal standard of significance in modern epidemiology is a one in 20 chance of the result having occurred by accident. But if you look at ten different diseases, this standard means that the probability of at least one crossing a given threshold of risk level becomes 40 per cent, which should be adjusted for, but isn’t."
Very true. Unfortunately this is far beyond the general populations understanding of statistics. More worryingly, it seems to be beyond many scientists understanding of statistics!
And Next ...
Could some good professor please turn their attention to road accident statistics?
I (and also Jeremy Clarkson, apparently) am sick to death of every road accident where "speed was a factor" being used as another reason to lower speed limits. Of course speed was a factor - it's always a factor - if the car was stationary, there'd be no accident .. and there'd be no point, either! The essential factor, so often not mentioned, is that the driver made a serious error of judgement ... so education and driver skill improvement should be the appropriate response, not lowered limits, speed cameras and harsher penalties.
And then there's the road design, sign posting, fatigue, mobile phones, etc., etc., etc.
OK, sorry I'm off topic, but there haven't been any driving/traffic stories that I've seen recently and I needed to vent.
That would be the lethal strain of cannabis then.
Roger Highfield ...
... is the new editor of New Scientist.
"Roger is a formidable force in science journalism. He has immense knowledge and wisdom and is brimming with new ideas," said Jeremy Webb, New Scientist's editor-in-chief. The magazine is right at the centre of all these efforts and we need a strong, creative editor to lead it."
I think the chlorine article demonstrates what Webb means by "creative" - makestuffup.
My subscription lapsed a few years ago - I won't be renewing it.
Yes it was bad... but not so bad as nothing!
Nice catch, that is indeed a addition to the uses/dangers of Dihydrogen Monoxide
@ Darkside - Junk Science
So, in a rant on "Junk Science" you make the claim "bicycle helmets are a scam: they actually make most cyclists less safe".
I think you may have difficulty showing, scientifically, quite how a plastic helmet physically makes someone a poorer cyclist. Call me MR Presumptuous, but I would hypothesise that the causality is a purely psychological one and not the fault of the poor, defenceless helmet.
But thank you for giving me a good laugh on an, otherwise, entirely dull Thursday afternoon. Junk Science indeed.
@Darkside Re:Bicycle Helmets
I know my cycle helmets have saved my life (or at least prevented a nasty cracked head) twice during bicycle accidents. The first time was on a cycle path when someone stepped out in front of me and in avoiding them I came off and totally trashed my helmet, the second was when the new chain on the bike snapped as I accelerated away from traffic lights and I went over the handlebars.
I personally think you would find it quite hard to prove cycling without a helmet is safer than cycling with one!
One thing I can't get over is why not wear a helmet? It isn't exactly onerous to remember to put it on before you go out? But the thing that gets me the most is where you see parents cycling with their kids, the kids with helmets and the parents without... I mean do the parents think they are immune to accidents?
Because even Paris knows a helmet when she sees one...
The reason why 0.05 is an abismal standard of significance is that if you do a data trawl looking for correlations at maybe 100 different possible causes then a one in twenty level of significance is virtually certain to show a few positive examples. That's how many epidemologist work - rather than starting of with a potential causative model which they want to test, they effectively use the same analysis on a whole range of ills. Even if this particular study wasn't looking at 100 different illnesses, then at any one time there are probably thousands of such studies being worked on. It's inevitable that some wil show a correlation so the result is the same. Large numbers of false positives.
The best that can be said about a one in twenty test for significance is that it may be sufficient to trigger of a further set of studies to see if some form of causative link (or at least non-randomly occuring correlation) is present. This can be very difficult to do, especially on very rare illnesses. Certainly this sort of study is not sufficient for an alarmist headline which states that exposure to chlorinated water doubles the risk.
There have been many occasions where corrations have been desmonstrated with no conceivable causative mechanism, simply by tramping through huge numbers of possibilities (as a method of debunking this sort of approach I should add). That doesn't stop the generation of stupid "hidden message" type books about how predictions of past events can be found, retrospectively I should add, in the Bible, works of Shakespeare, you name it. This fallacy doesn't just apply to cases like this - false genetic and fingerprint matches have been found the same way (and cases goine to court). People have been jailed due to misapplication of statistics, sometimes by so-called experts whose methodologies have been shown to be wrong (and anyone reasonably conversant with statistics - which mostly does not include lawyers) could have seen the error.
What I also note from some of the responses to this article is that a considerable number of people have got no real sense of how statistics work, but that doesn't stop them having an opinion. That includes sarcasm and ridiculopus comparisons to the lung cancer/smoking study. The correlation factors here are just many orders of magnitude different.
I dare say that the average Register reader is more numerate than the average member of the population. The truth is that most people can't do statistics (and the inability is virtually universal among politicans and national newspaper journalists). What they tend to do is get blinded by numbers thrown around by "experts" as somehow definitive when they are not. Well I say that, but most people I know are now so confused with the conflicting messages that they just don't take any seriously. The problem is that fanatics who do, more dealers in faith than logic.
more Chlorine scare!
I'm tracking the EC plans to allow the so called (and rightly scary) "American Chlorine Chicken" into the EU market. This hybrid monster seems to be some sort of beast that is raised purely in terms of good economics, then bathed in Chlorine dioxide to annihilate the inevitable nasties that crowd its surfaces, then posted off to Europe by a slow boat. Once in the EU will probably be stamped as, "countryside nature's best" or similar, sprayed with limonene to dampen the nerve gas olfactory remnants and sold to customers at the price of Bourg-en-Bresse bantams?
More info available by searching the francophone web with "Poulets américains au chlore"
for the record, I've converted my swimming pool at home to use 14 tons of water with around a hundred kilos of dissolved NaCl. Galvanically electrolysing with 14amps for three hours each night. Pool doesn;t smell of Cl2, tastes mildly of salt, and hasn't suffered death by algael bloom as it has in previous years using mega 'double' & 'triple' Cl expensive stock pool chemicals. Just don't use Iodised salt!
100kilos of salt can be had for around 8 quid (for the whole season!)
Mine's the smock with the bleached, tattered holes in it
What I remember seeing was:
- a Cyclist wearing a helmet appears more "professional" to the driver.
- the driver then gives the cyclist less room during an overtake.
- less room = more chance of contact between large 1+ tonne object and small <120kg cyclist
There is also probably an element of "... i'm wearing a helmet so I'm safer from damage, and can take more chances..." effect on the cyclist side.
@ Peter - It angle (or close)
see this article - some bored "researcher" determined they needed some extra exposure in the media -
so Cell phones are (again) a cancer danger -
YET - MRI is not even though the electromagnetic fields involved are 1000s (or more) times as strong.
a commentary on those dreadful "how clean is your kitchen/room/small intestine" where you see these cretins taking swabs from dirty kitchens and then exclaiming "my lord, you have ecoli growing on your plates! and you have children.. you filthy man (cause it's *always* a man), buy loads of cleaning products and sterelise your environment!".
Yet, I can't help but notice they never mention that the person is yet to fall ill to these bugs and, in fact is likely be in posession of a healthier immune system and a reduced incidence of allergies.
Of course, there's always limits to this - IE total squalor - but the abuse of bleach and cleaning products in the home (and their constant pitching of them through programs such as those mentioned above) is a massiively ignored issue because everyone is obsessed with cleaning rather than living in a slightly dirty environment (IE given a big clean only once a month or so and simply maintained with little wipedowns in between) and promoting the use of their bodies own defense mechanisms and the reinforcement they get through contact with outside particles.
@ James Bassett
The fault would not lie with the helmet, but rather with motorists who are (supposedly) more likely to give a cyclist adequate space if they are not wearing a helmet. From what i remember this wasn't from a scientific study as such but from the experiences of one cyclist. I could be wrong though.
@ James Bassett
Nice strawman you've got there. Where did it say anything about 'physically makes someone a poorer cyclist'? Causality is no less real for being psychological.
Actually, though, a helmet can physically make a cyclist less safe (though it would be a stretch to say 'most cyclists'). Damaged or weakened helmets (of which there are a substantial number on the road) can concentrate an impact to a greater degree than it would have been with no protection, and so cause more severe injuries.
Christine Houghton said: "Since its creation, chlorine has been a chemical catastrophe. It is either chlorine or us."
Chlorine is an element, comes as a gas in its native form. It was formed along with all the other elements when the universe was created.
I expect this sort of rubbish from creationists, but this is by inference insulting the sky fairy.
A good extension would be call the un-chlorinated water "God Water" and the chlorinated water "Darwin Water". This would fix a number of problems.
The usual form of the bicycle helmet argument is that because helmets are needed, people perceive cycling as dangerous and so do it less and discourage their children from doing it. This puts more cars on the road as a result, which increases the hazard for those who do still cycle, so therefore the helmets do make cycling more hazardous. There's also less exercise done because less people cycle, so a few more people will die for reasons associated with that.
Therefore, cycle helmets are a bad thing for normal road use.
(simple, isn't it?)
@ Darkside - Junk Science
Well, I for one am still here due to junk science.
If it wasn't for my cycle helmet It would have been my bare head smacking into the side of the kerb (at a fairly respectable 15mph if witnesses are to be believed). One cracked cycle helmet, one living stef.
Keep up with the junk science, and with the helmets too!
Can't really see the point of this article
I mean, it's not as if anyone is really going to go out and ban chlorine in our water supply on the strength of one weak study is it? The most likely (and probably most appropriate) reaction is just to ignore it. The literature isn't short of papers making somewhat extravagant claims based on weak evidence in any discipline, so where's the news angle? A weak study is not unusual, it's not important and it's only marginally interesting. By making a fuss about it you only increase the likelihood that some idiot will be convinced by it.
@Dave - Leave me out of this
"Sky fairy" eh? How about I open up a can of pestilence on your ass?
Gotta Love the Ecoloons
One invited to a barbecue was pontificating the current salmonella in tomatoes and lettuce scare (his attention was drawn by the lack of tomatoes and lettuce in our burger customising kit that day). He wondered aloud why it was that we had these problems with fresh tomatoes and not with the ketchup we had supplied.
I looked up from the grill and said (with a secret glee) "Don'cha know? Ketchup is treated with gamma radiation when it's bottled to kill off all lurking biologicals"
When this was confirmed by a couple of our other guests, Mr Green (47), product of his US upbringing and former fan of Messrs Heinz's product made a big show of rejecting the ketchup.
I then asked him if he was feeling well.
He said he was.
I sprang the trap. "Well, they've been zapping the ketchup since the 1950s. A lifetime of eating irradiated ketchup can't have done you much harm, can they?" and loaded up my own sandwich.
Anyone rejecting sane water treatment should be made to live well-away from the rest of us. I don't want to catch any one of the plagues these loons will re-introduce to the population because of some half-arsed new-age "knowledge" they picked up in a newspaper or website.
Did you know we have whooping cough, *whooping cough* for christ sake, in some neighbourhoods of New York? The fools that live there think that because an enlightened immunisation plan has all-but eradicated the disease in the wealthy west there is no need to vaccinate. Idiots. Worse. Dangerous idiots.
I despair of the human race sometimes.
I'm missing something, obviously
but isn't this kind of investigation valuable, even if the sample sizes are small, as it can flag up important issues?
This particular one doesn't seem to fall into the "important" category mind, but I'm sure all the thalidomide babies would have appreciated searches like this a bit sooner....
the mistake that's made is to put too much emphasis on %age increases rather than absolute increases (1 extra death per mil is pretty much irrelevant even if that's a 500% increase, if elsewhere it's saving 10 per mil).
Is increasingly being pushed aside in favour of correlating statistics, particularly in medical `research´? The reasons probably being that it is a cheap way of doing it and it's not even a requirement to actually see a patient. I subscribe to amongst others an online mag called Science Daily.
Literally every day in it there are new so called discoveries to show that one thing or another can cause this or that disease/condition, all this research is carried out by statistical analysis of records with seemingly no empirical science to back it up. So, hundreds of published papers are relatively meaningless, they serve only as indicators of trends in numbers of patients suffering from similar things, mathematically. I guess the thing is driven by the current trend in the sciences that you are a nobody unless you have published a decent number of papers.
Oh and as far as wearing a cycle helmet is concerned.
I will wear one when they start to design them to look cool and not make the wearer look like a complete twat!
Clean and safe
I used to work in the water industry* and the whole point about using Chlorine to treat tap water is this: Chlorine is added to the water near the end of the treatment process to kill the nasties and then most of it is removed again before the water leaves the treatment plant. What is left keeps the water safe all the way to your tap. If you're worried about it, why not filter the water before you drink it or even just leave it in a jug for an hour or two for the remaining Chorine to escape?
*It was 6 months before someone told me what the windsock at the treatment plant was for - in the event of a Chlorine gas leak, you run in the opposite direction to where the sock is pointing!
Lies, Damned lies, and Statistics!
All evil. We (the public) are easily duped by these numbers (aren't they all made up anyway!). Most of this is due to the "dumbing down" of our education.
I grew up in the 50's and given all that has been discovered as being "bad" it is a wonder I made it at all. Fortunately we had nice clean chlorinated water and it had fluoride as well, so I have descent teeth.
Now for the silly fact of the day: If you grow up in a house with lots of people (who just happen to share a bathroom) you have a stronger immune system. No, you don't have to be slobs, but the variety of germs that everyone brings to the bathroom induces a good immune system. If you are "too clean" your immune system doesn't get "trained", and you have things like asthma later in life.
If you start out life in a bubble, you are destined to live all of it that way.
Remember: Life is a terminal disease!
RE: So DDT wasn't bad?
DDT wasn't bad. The gross over application of DDT was bad. And even then it wasn't nearly as bad as the environmentalists and the media would have you think. That stuff is VERY effective in what it does and it really doesn't take much to work wonders.
It's kind of strange to see the exact same FUD that was applied to DDT being used today against CO2.
Something never change.
so we're all gonna die again
Just stay in bed, kids. If you get up you might die. Next up: Breathing causes greenhouse gases, so don't do it! Think of the children!
Chlorine was nor produced when the universe was created, it was formed after this event, see
- Analysis Who is the mystery sixth member of LulzSec?
- Analysis Hey, Teflon Ballmer. Look, isn't it time? You know, time to quit?
- Murdoch Facebook gloat: You're like my $580m, 'CRAPPY' MySpace
- Tablet? Laptop? HP does the splits with Tegra-based SlateBook x2
- NASA signs off on sampling mission to Earth-threatening asteroid