Wait, cell phones do cause brain cancer?

[Tsukasa Buddha]

Professor Khurana – a top neurosurgeon who has received 14 awards over the past 16 years, has published more than three dozen scientific papers – reviewed more than 100 studies on the effects of mobile phones. He has put the results on a brain surgery website, and a paper based on the research is currently being peer-reviewed for publication in a scientific journal.

He admits that mobiles can save lives in emergencies, but concludes that "there is a significant and increasing body of evidence for a link between mobile phone usage and certain brain tumours". He believes this will be "definitively proven" in the next decade.

Noting that malignant brain tumours represent "a life-ending diagnosis", he adds: "We are currently experiencing a reactively unchecked and dangerous situation." He fears that "unless the industry and governments take immediate and decisive steps", the incidence of malignant brain tumours and associated death rate will be observed to rise globally within a decade from now, by which time it may be far too late to intervene medically.

Linky. (http://www.independent.co.uk/life-style/health-and-wellbeing/health-news/mobile-phones-more-dangerous-than-smoking-or-asbestos-802602.html?r=RSS)

Of course, there is doubt presented at the end:

Late last week, the Mobile Operators Association dismissed Khurana's study as "a selective discussion of scientific literature by one individual". It believes he "does not present a balanced analysis" of the published science, and "reaches opposite conclusions to the WHO and more than 30 other independent expert scientific reviews".

[Sefarst]

Of course, there is doubt presented at the end:
...By the Mobile Operators Association, a lobby group for cell phone operators.

[casebro]

But on the other hand, Professor Khurana did a Mehta study- no new data, only a re-hash. Lots of room for selection bias there. And other possibilities.

Plus, who has looked up Professor Khurana's biases? What were the subjects of his other papers? He's a neurosurgeon, but that doesn't make him an epidemiologist. Or an electronics engineer.

[krelnik]

Mark Hoofnagle took this one completely apart this morning:

http://scienceblogs.com/denialism/2008/03/cell_phones_and_cancer_real_or.php

Basically, this so called study is total junk. He has all the details.

[Walrus32]

He's probably got a very lucrative deal to shill for the tort lawyers.

[MattusMaximus]

In answer to the thread title... No, cell phones do not cause cancer :rolleyes:

The misguided belief that they might be increasing the risk of cancer is due to a fundamental misunderstanding of the physics of the electromagnetic spectrum (http://en.wikipedia.org/wiki/Electromagnetic_spectrum). The radiations that cause cancer are on the low wavelength, high frequency end of the spectrum - ultra-violet light, x-rays, gamma rays, etc. The stuff on the low frequency end - radio and microwaves - where cell phones and similar technologies operate don't cause cancer. The reason is basically because the low frequency, long wavelength radiations are "too fat" to get down into our DNA and rip it apart.

Here's more info on this topic: http://skepdic.com/emf.html

[Tubbythin]

In answer to the thread title... No, cell phones do not cause cancer :rolleyes:

The misguided belief that they might be increasing the risk of cancer is due to a fundamental misunderstanding of the physics of the electromagnetic spectrum (http://en.wikipedia.org/wiki/Electromagnetic_spectrum). The radiations that cause cancer are on the low wavelength, high frequency end of the spectrum - ultra-violet light, x-rays, gamma rays, etc. The stuff on the low frequency end - radio and microwaves - where cell phones and similar technologies operate don't cause cancer. The reason is basically because the low frequency, long wavelength radiations are "too fat" to get down into our DNA and rip it apart.

Here's more info on this topic: http://skepdic.com/emf.html

[Devil's advocate]
Isn't the current "theory" more to do with local thermal heating by the radiation rather than ionization taking place?[/Devil's advocate]

[Gord_in_Toronto]

I think the evidence points to the "fact" that people predisposed to brain cancer are also predisposed to using cell phones. Well, it there were any evidence in the first place. ;)

[MattusMaximus]

[Devil's advocate]
Isn't the current "theory" more to do with local thermal heating by the radiation rather than ionization taking place?[/Devil's advocate]


So those proponents of the current "theory" would need to explain exactly how "cooking" equates to cancer-causing ionizing radiation. I'm not going to hold my breath waiting for a well-reasoned response, though.

Hmm, that's funny. I never knew that when I was heating up my dinner in the microwave that I was giving the food cancer ;)

[zosima]

I mean if your brain got to the point of char, that would probably generate a few free radicals...but then again at that point I imagine cancer would be the least of your worries.

But forget about cancer, I just wonder what the consequence of spontaneously heating someone's brain has on someone's train of thought. If it does heat the brain I don't think it does much, if it did using a cell phone would cause fever, inflammation. In other words your body would start flipping out as it is likely to do when experiencing brain trauma....but as I understood it, microwave radiation doesn't penetrate further than a couple millimeters.

I remember reading about how a company was thinking about marketing microwaves as an efficient way to heat a house, but people were too freaked out to go for it. The benefit is that less energy is lost to heating air than to heating people and changes in temperature are felt instantly.

But I figure if the military can use them for crowd control:
http://www.jwharrison.com/blog/2006/09/15/military-to-test-microwave-weapons-as-crowd-control-in-the-us/
There should be no reason we can't heat our houses with microwaves.

[portlandatheist]

Dr. Steven Novella also covered this topic. (http://www.sciencebasedmedicine.org/?p=84) he seems to be more along the lines of the jury is still out and there may be a link.
I am very doubtful. I do get the heeby jeeby's when standing underneath high tension powerlines but I think that is just psychological. I doubt these people are concerned about a flash light or a indoor lighting causing cancer.

[Ivor the Engineer]

Dr. Steven Novella also covered this topic. (http://www.sciencebasedmedicine.org/?p=84) he seems to be more along the lines of the jury is still out and there may be a link.
I am very doubtful. I do get the heeby jeeby's when standing underneath high tension powerlines but I think that is just psychological. I doubt these people are concerned about a flash light or a indoor lighting causing cancer.

...or the heating from the sun.

[Tubbythin]

So those proponents of the current "theory" would need to explain exactly how "cooking" equates to cancer-causing ionizing radiation. I'm not going to hold my breath waiting for a well-reasoned response, though.


They'd have to explain how mobile phone "cooking" leads to malignant DNA mutations.

[sol invictus]

[Devil's advocate]
Isn't the current "theory" more to do with local thermal heating by the radiation rather than ionization taking place?[/Devil's advocate]

Hot baths cause cancer?

[Tubbythin]

Hot baths cause cancer?

Note the use of "devils' advocate" and the speechmarks around the word theory. All I was suggesting was that if Mattus said what he said to a "mobile phones cause cancer" proponent they may say it had nothing to do with ionizing radiation and was something to do with heating of the brain. I.e. that Maximus may be dismissing the wrong argument. At no point did I say that said "theory" was credible.

[sol invictus]

I was merely pointing out that the devil might want to seek new legal representation.

[Gord_in_Toronto]

Hot baths cause cancer?

Sure. People take hot baths. People get cancer. QED. :D

[Gord_in_Toronto]

Hot baths cause cancer?

Sure. People take hot baths. People get cancer. QED. :D

[Soapy Sam]

Mobile phones give me headaches. I hear voices in my head. I want to kill.

This is not due to microwaves, but stupidity waves.

"Hello? I'm on the train!"

I know, lackwit. So am I. The guy with the aggressive attitude and the screwdriver, in the seat opposite.

European proposals to permit these instruments of Satan on aircraft is the last straw. I predict people will start dying due to cellphone use much sooner than Prof. Wotsit's most pessimistic prediction.

[Gord_in_Toronto]

Hot baths cause cancer?

Sure. People take hot baths. People get cancer. QED. :D

[Tubbythin]

I was merely pointing out that the devil might want to seek new legal representation.

Sure. Like someone who believes the theory they are defending. And someone with a better idea of what he's defending in the first place.

[JJM]

It is unlikely; but, unknown http://www.sciencebasedmedicine.org/?p=84

[BillyJoe]

Sure. People take hot baths. People get cancer. QED. :D


Hell, it wasn't that funny the first time. :cool:

[Gord_in_Toronto]

Hell, it wasn't that funny the first time. :cool:

Sorry about the double triple post.

I sure gets weaker with each telling doesn't it? :o

[BillyJoe]

An interesting quote from the comments section of the following blog:http://www.sciencebasedmedicine.org/?p=84Comment by Wallace Sampson:
The cell phone brain tumor claim began with a talk show call-in listener whose wife developed a brain tumor and who got into his head that a cause must be found, and the only one nearby was their cell phone. That was it. There never was a study or an objective observation of any association.
Ignore material like that and one can easily get lost in the indeterminacy of pseudoscience. Public money spent by the billions chasing non-causes of non-events.


Is that really an accurate description of the origins of the cell phone/brain cancer link, does anyone know?

[BillyJoe]

http://journal.media-culture.org.au/0106/cell.php


It begins in August 1988 in Florida when David Reynard gave his wife Susan the gift of a cellphone. Seven months later, a medical scan revealed a tumour in Susan's brain. She claimed that as a result of being bombarded by radiation from the cellphone, the damaged cells either caused her tumour or accelerated the growth of an existing tumour.

In April 1992, Susan launched a lawsuit against the phone's manufacturer, the company that provided the cellular service, and the retail store that sold the phone. A month after filing the lawsuit, Susan died of brain cancer.

In January 1993, David Reynard was interviewed on the highly-rated CNN show Larry King Live. The interview sent shockwaves through the telecommunications industry.


5 years later, there was a $800 million lawsuit against Motorola:

http://www.fda.gov/fdac/features/2000/600_phone.html

Convinced that a nine-year cell phone habit led to his brain cancer, neurologist Chris Newman, M.D., has filed an $800 million lawsuit in Baltimore against his cell phone's maker and several other telecommunications companies. His suit comes five years after the dismissal, for lack of evidence, of a lawsuit filed in Florida by David Reynard, who alleged that a cell phone was responsible for his wife's fatal brain cancer.

In Newman's case, his lawyer has said, "it's really not a question at all" whether the cancer is cell phone-related. The evidence, she says: Newman's own doctors made the connection between his long-time cell phone use and his tumor, which is positioned in "the exact anatomical location where the radiation from the cell phone emitted into his skull."



:rolleyes:

[BillyJoe]

So...

The new meta-analysis shows no link for the first 10 years but a small link after 10 years, but the case that triggered off the whole controversy was that of a brain tumour in a woman who had been using a cellphone for mere 8 months!

[BillyJoe]

There is also the myth that the brain tumour developed more often on the side of cellphone use.

This origin of this "myth" was a case control study looking retrospectively at a population of cellphones users. It obviously had to rely on the self-reporting of those with brain tumours as to which side of the head they used their cellphones. Of course they already knew which side their cancer was on! Therefore reporting bias obviously came into play as the following summary from that report indicates:

http://www.mthr.org.uk/documents/MTHR_report_2007.pdf (page 9 of 64)

Well it won't allow me to quote from the pdf file so I will summarise:

There did not appear to be any increase in the incidence of brain tumours amongst cellophone users but those who had brain tumours were more likely to report that they used the cellphone on the same side as the tumour. This of course means that the use of cellphones on one side of the head reduced the number of tumours on the opposite side of the head!

[MattusMaximus]

Hot baths cause cancer?


Oh hell, I hope not. If that's true, then I'll be dead before I'm 40...

[MattusMaximus]

Note the use of "devils' advocate" and the speechmarks around the word theory. All I was suggesting was that if Mattus said what he said to a "mobile phones cause cancer" proponent they may say it had nothing to do with ionizing radiation and was something to do with heating of the brain. I.e. that Maximus may be dismissing the wrong argument. At no point did I say that said "theory" was credible.


Tubbythin makes a good point (I caught it Tubby, btw ;) )

Basically, if a "cell phones cause cancer" proponent got pinned on the whole ionizing radiation argument then switched to the "cooking" argument they'd be engaging in a classic pseudoscience maneuver called "moving the goalposts."

And I feel that if you kept knocking out the supports from their "theory" then eventually that old stand-by would pop up... the Conspiracy!!! <insert spooky music here>

[BillyJoe]

...or they'd go back to the ionising radiation theory again just as if the previous demolition job on it hadn't occurred.

[zosima]

Tubbythin makes a good point (I caught it Tubby, btw ;) )

Basically, if a "cell phones cause cancer" proponent got pinned on the whole ionizing radiation argument then switched to the "cooking" argument they'd be engaging in a classic pseudoscience maneuver called "moving the goalposts."

And I feel that if you kept knocking out the supports from their "theory" then eventually that old stand-by would pop up... the Conspiracy!!! <insert spooky music here>

Cheer!

Seriously though, microwave radiation doesn't penetrate the skin, so you would get an increase in skin cancer not brain cancer. The fact that the claim is brain cancer shows that this is based around people's fears not scientific fact.

[BillyJoe]

...yet, despite the billions of dollars spent so far, there are plans to continue spending billions of dollars on epidemiological studies to see if there is a link beyond ten years.

[blutoski]

Tubbythin makes a good point (I caught it Tubby, btw ;) )

Basically, if a "cell phones cause cancer" proponent got pinned on the whole ionizing radiation argument then switched to the "cooking" argument they'd be engaging in a classic pseudoscience maneuver called "moving the goalposts."

I think it's more an example of ad hoc hypothesis. Goes like this:

Pseudoscience investigatory steps:

A may or may not be true
B is a plausible mechanism for A
B is not supported by evidence
OK: C is a plausible mechanism for A
C is not supported by evidence
OK: possibly D, E or F explains A. Who knows. Anyway, I just want to reiterate that there's no reason to return to square one and question A.
Gallileo!
Hey! I just thought of something... B is a plausible mechanism for A


We saw this with the autism/vaccine link theories. As soon as one was demonstrated to be insufficient (MMR), another would be presented (Thimerosal), and when that fails to show a good correlation, then they propose that maybe it only happens when there's a genetic predisposition *and* vaccines.

The point is that disproving hypothesis A is seen as an occasion to announce a new explanation (to be tested at a later time), rather than to finally bury hypothesis A.

[MattusMaximus]

We saw this with the autism/vaccine link theories. As soon as one was demonstrated to be insufficient (MMR), another would be presented (Thimerosal), and when that fails to show a good correlation, then they propose that maybe it only happens when there's a genetic predisposition *and* vaccines.

The point is that disproving hypothesis A is seen as an occasion to announce a new explanation (to be tested at a later time), rather than to finally bury hypothesis A.


It's worse than those god-damn zombies in the George Romero flicks. Except these suckers don't go down even after you shoot them in the head! :boggled:

[zosima]

Gallileo! lol....

I think it's more an example of ad hoc hypothesis. Goes like this:

Pseudoscience investigatory steps:

A may or may not be true
B is a plausible mechanism for A
B is not supported by evidence
OK: C is a plausible mechanism for A
C is not supported by evidence
OK: possibly D, E or F explains A. Who knows. Anyway, I just want to reiterate that there's no reason to return to square one and question A.
Gallileo!
Hey! I just thought of something... B is a plausible mechanism for A


We saw this with the autism/vaccine link theories. As soon as one was demonstrated to be insufficient (MMR), another would be presented (Thimerosal), and when that fails to show a good correlation, then they propose that maybe it only happens when there's a genetic predisposition *and* vaccines.

The point is that disproving hypothesis A is seen as an occasion to announce a new explanation (to be tested at a later time), rather than to finally bury hypothesis A.

:)

[ponderingturtle]

The interesting thing is that hard science people seem to take a stronger stand about the ridiculusness of this than doctors. Is it that doctors have no understanding about the differences between ionizing radiation and low energy radiation such as a cell phone would produce?

Sure it seems that each photon is not going to have anywhere near enough energy to cause any kind of chemical reaction so you are just getting heating, with a physics background that makes sense to me. Is it that doctors don't have the physics and chemistry background to understand this or what?

[MattusMaximus]

Sure it seems that each photon is not going to have anywhere near enough energy to cause any kind of chemical reaction so you are just getting heating, with a physics background that makes sense to me. Is it that doctors don't have the physics and chemistry background to understand this or what?


I think the answer is simple... even some doctors are idiots. Just ask Gregory House :)

[Slimething]

Brains cause brain cancer. Perfect correlation!

[BillyJoe]

Is it that doctors don't have the physics and chemistry background to understand this or what?
I think the answer is simple... even some doctors are idiots. Just ask Gregory House :)


What you don't realise is that it's scientists who are taking this thing seriously.
The largest study on this subject was undertaken by this group:

http://www.mthr.org.uk/documents/MTHR_report_2007.pdf
(see page 52)

The pdf file won't allow me to copy and paste but we have here a:

- professor of physics
- radiocommunications expert
- psychologist
- mathematician
- physicist
- cognitive neuroscientist X2
- theoretical physicist X2
- electrical engineer
- biochemist
- neurophysiologist

Not a single medical doctor amongst them.
So there we go.

[ectoplasm]

I ran across a report at bioinitiative.org about ELF and RF radiation and the health hazards associated with it, and the public standards for allowable level of exposure to that radiation. There is a 38 minute video, which I didn't watch and a lengthy report.

BioInitiative Report:
A Rationale for a Biologically-based Public Exposure Standard
for Electromagnetic Fields (ELF and RF)

Afew highlights:

4. Brain Tumors and Acoustic Neuromas
in Cell Phone and Cordless Phone Users (Meta-Analysis)

Glioma and Acoustic Neuroma

Hardell et al., (2007) reported in a meta-analysis statistically significant increased risk for glioma with exposure of 10 years or greater in persons using cell phones. Risks were estimated to be 1.2 (0.8 – 1.9) for all use; but when ipsilateral use was assessed (mainly on same side of head) it increased the risk of glioma to 2.0 (1.2 – 3.4) for 10 years and greater use.

For acoustic neuromas, Hardell et al., (2007) reported the increased risk with 10 years or more of exposure to a cell phone at 1.3 (0.6 – 2.8) but this risk increased to 2.4 (1.1 – 5.3) with ipsilateral use (mainly on the same side of the head). There is a consistent pattern of increased risk for brain tumors (glioma) and acoustic neuromas at 10 years and greater exposure to cell phones.

The meta-analysis by Lakhola et al., (2006) reported that brain tumor risk was 1.3 (0.99 – 1.9) for ipsilateral use of a cell phone, but no data was given for exposures at 10 years or greater (all exposures were of shorter duration).

The meta-analysis by Kan et al., (2007) reported “no overall risk” but found elevated risk of brain tumors (RR = 1.25, CI 1.01 – 1.54) > 10 years, reinforcing the findings of other pooled estimates of risk. No estimates of increased risk with ipsilateral use were provided, which would have likely increased reported risks.



They also cover quite a few other health hazards, childhood leukemia, breast cancer, alzheimer's, and summarize the evidence of links to EMF and RF radiation.

Here is part of their conclusion:

We cannot afford ‘business as usual” any longer. It is time that planning for new power lines and for new homes, schools and other habitable spaces around them is done with routine provision for low-ELF environments . The business-as-usual deployment of new wireless technologies is likely to be risky and harder to change if society does not make some educated decisions about limits soon. Research must continue to define what levels of RF related to new wireless technologies are acceptable; but more research should not prevent or delay substantive changes today that might save money, lives and societal disruption tomorrow.

there is much more, but I didn't want to make this post too long. I'd be interested in what others think of it.

[BillyJoe]

I'd be interested in what others think of it.


I think the answers to your questions are already contained in this thread, or maybe I read it in that 64 page report I linked to before. Anyway...

Hardell et al., (2007) reported in a meta-analysis statistically significant increased risk for glioma with exposure of 10 years or greater in persons using cell phones. Risks were estimated to be 1.2 (0.8 – 1.9) for all use;


If the risk was 1.0, that would mean that there was no risk posed by cellphones. But the risk was actually 1.2. What does that mean? Well, what it does not mean is that there was necessarily an increased risk. The 95% confidence interval is quoted as 0.8 - 1.9. If the 95% confidence interval contains 1.0 (which it does), then the result is not statistically significant.
In other words, and despite what the author says, 1.2(0.8-1.9) is not a statistically significant result. This meta-analysis does not support a link between >10 years cellphone use and glioma.

...but when ipsilateral use was assessed (mainly on same side of head) it increased the risk of glioma to 2.0 (1.2 – 3.4) for 10 years and greater use.



By contrast 2.0(1.2-3.4) IS a statistically significant result because the 95% confidence interval (1.2 - 3.4) does not contain 1.0. Therefore there is a statistically significant positive link between >10 years ipsilateral cellphone use and glioma.
But remember that there is not a statistically significant link between >10 years (ipsilateral + contralateral) cellphone use and glioma. Therefore, there MUST a statistically significant negative link between >10 years contralateral cellphone use and glioma.
What this means is that cellphone use increased the risk for ipsilateral glioma and decreased the risk of contralateral cellphone use. I hope you agree that that is a pretty unlikely scenario. In fact, a much more reasonable explanation for that result is reporting bias. In other words, cellphone users who have glioma already know which side their tumour is on and therefore tend to report that they mostly used their cellphones on that side.

For acoustic neuromas, Hardell et al., (2007) reported the increased risk with 10 years or more of exposure to a cell phone at 1.3 (0.6 – 2.8)


Same analysis: No statistically significant effect of >10 years cellphone use and acoustic neuroma.

but this risk increased to 2.4 (1.1 – 5.3) with ipsilateral use (mainly on the same side of the head).


Here the result is just barely significant, with 1.0 falling just outside the 95% confidence interval. Remember also that 5% of results would lie outside the 95% confidence interval even if there is actually no risk of cellphone use on acoustic neuroma.

There is a consistent pattern of increased risk for brain tumors (glioma) and acoustic neuromas at 10 years and greater exposure to cell phones.



I would say that this conclusion is not supported by the results.

The meta-analysis by Lakhola et al., (2006) reported that brain tumor risk was 1.3 (0.99 – 1.9) for ipsilateral use of a cell phone, but no data was given for exposures at 10 years or greater (all exposures were of shorter duration).



Again, this is not a statistically significant result.

The meta-analysis by Kan et al., (2007) reported “no overall risk” but found elevated risk of brain tumors (RR = 1.25, CI 1.01 – 1.54) > 10 years, reinforcing the findings of other pooled estimates of risk.


Risk = 1.25 with 95% confidence interval of 1.01 - 1.54.
To call this result statistically significant (which they do not do) with 1.0 falling so closely outside the 95% CI of 1.01 - 1.54 with the barest of margins, would beggar belief.

No estimates of increased risk with ipsilateral use were provided, which would have likely increased reported risks.


...and would have pointed towards reporting bias as the likely cause of that result.


BillyJoe

[BillyJoe]

Here are that organisation's misleadingly public statements about the meta-analysis report:
http://www.bioinitiative.org/report/docs/section_1.pdf (page 10)

Meta-analysis report:

Hardell et al., (2007) reported in a meta-analysis statistically significant increased risk for glioma with exposure of 10 years or greater in persons using cell phones. Risks were estimated to be 1.2 (0.8 – 1.9) for all use;

Public statement:
For brain tumors, people who have used a cell phone for 10 years or longer have a 20% increase in risk (when the cell phone is used on both sides of the head).



Meta-analysis report:

...but when ipsilateral use was assessed (mainly on same side of head) it increased the risk of glioma to 2.0(1.2 – 3.4) for 10 years and greater use.

Public statement:

For people who have used a cell phone for 10 years or longer predominantly on one side of the head, there is a 200% increased risk of a brain tumor.



Meta-analysis report:

For acoustic neuromas, Hardell et al., (2007) reported the increased risk with 10 years or more of exposure to a cell phone at 1.3 (0.6 – 2.8)

Public statement:

For acoustic neuromas, there is a 30% increased risk with cell phone use at ten years and longer;




Meta-analysis report:


but this risk increased to 2.4 (1.1 – 5.3) with ipsilateral use (mainly on the same side of the head).



Public statement:



...and a 240% increased risk of acoustic neuroma when the cell phone is used mainly on one side of the head.






I think there may a vested interest operating here.

[Ivor the Engineer]

The figures quoted in the public statements were the point estimates. How is it misleading to quote them?

Given that many researchers don't understand what a 95% CI means, do you really want to try to explain the concept to the general public?:boggled:

[BillyJoe]

The figures quoted in the public statements were the point estimates. How is it misleading to quote them?


These point estimates are misleading because they do not give an accurate interpretation of the results obtained from these studies. In fact, they give a false interpretation. The data do NOT support a link between cellphones and brain tumours but the results are given in such a way as to make it look as if they do support a link.

Given that many researchers don't understand what a 95% CI means, do you really want to try to explain the concept to the general public?:boggled:


You don't need to. An accurate summary of the data is as follows:
These studies do not show any statistically significant link between cellphones and brain cancer.

[Ivor the Engineer]

<snip>

You don't need to. An accurate summary of the data is as follows:
These studies do not show any statistically significant link between cellphones and brain cancer.

Many of the results were statistically significant. You appear to want your cake and eat it when you say:

Risk = 1.25 with 95% confidence interval of 1.01 - 1.54.
To call this result statistically significant (which they do not do) with 1.0 falling so closely outside the 95% CI of 1.01 - 1.54 with the barest of margins, would beggar belief.

To call this result statistically significant (at a level of confidence of 95%) is accurate. It is probably not of much practical significance, unless repeated studies find results which are consistent.

The biggest problem with these (observational) studies are confounders, especially when the effects are small.

[ectoplasm]

I remember reading somewhere that the first study that showed an association between lung cancer and smoking had a risk factor of 24. (I tried to find the reference again and I couldn't). It basically jumped out at you and stared you in the face.

We are talking about a risk factor here with last example given, of 1.25.

What risk factor is considered epidemiologicaly significant? I have read 2.0 and above, also 1.50 and above. I am assuming it would depend on the type of trial and the number of confounding factors and if they could be controlled - if it had a large number of participants, was adequately randomized and blinded, then a lower risk factor could be considered significant. Am I correct in saying this?

Does anybody have any thoughts on how you know if the risk factor is significant for the type of trial ?

[BillyJoe]

Many of the results were statistically significant. You appear to want your cake and eat it when you say:

To call this result statistically significant (at a level of confidence of 95%) is accurate. It is probably not of much practical significance, unless repeated studies find results which are consistent.

The biggest problem with these (observational) studies are confounders, especially when the effects are small.


I don't think we are actually disagreeing much. ;)

[Ivor the Engineer]

I don't think we are actually disagreeing much. ;)

What is funny is experiments have show using a mobile phone while driving increases your chance of having an accident by four times, yet many (most?) members of the public are not particularly concerned by this.

http://www.ncbi.nlm.nih.gov/pubmed/16854702?ordinalpos=1&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsP anel.Pubmed_DiscoveryPanel.Pubmed_Discovery_RA

It seems people are only worried about "invisible rays", or, more generally, aspects of a technology which they do not have understanding of and/or control over.

[fls]

I remember reading somewhere that the first study that showed an association between lung cancer and smoking had a risk factor of 24. (I tried to find the reference again and I couldn't). It basically jumped out at you and stared you in the face.

We are talking about a risk factor here with last example given, of 1.25.

What risk factor is considered epidemiologicaly significant? I have read 2.0 and above, also 1.50 and above. I am assuming it would depend on the type of trial and the number of confounding factors and if they could be controlled - if it had a large number of participants, was adequately randomized and blinded, then a lower risk factor could be considered significant. Am I correct in saying this?

Does anybody have any thoughts on how you know if the risk factor is significant for the type of trial ?

Most epidemiological studies are not experiments, but are observations (a large group is observed and measured but the researchers don't control the exposure), and therefore cannot be randomized and blinded. It is difficult to infer causation under these circumstances. In particular, it is difficult to infer causation from a single study or from a collection of the same kinds of studies, unless the association is very strong - on the order of 10, 20, 100 times the risk. Relative risks under 2 are too weak to suggest causation under these circumstances. Even a risk of 2 is too weak, but it at least gives reason to study the connection further and look for supporting evidence. By themselves, low relative risks, even if statistically significant, are too easily spurious or due to confounding variables (a third characteristic that is associated with the other two variables but is not part of a chain of causation - for example, people who smoke tend to drink more coffee and they tend to get lung cancer, but that doesn't mean that coffee causes lung cancer, despite the association).

The Bradford-Hill criteria are generally used to infer causation under these circumstances.

The Bradford Hill Criteria
1: Strength of Association. The stronger the relationship between the independent variable and the dependent variable, the less likely it is that the relationship is due to an extraneous variable.
2: Temporality. It is logically necessary for a cause to precede an effect in time.
3: Consistency. Multiple observations, of an association, with different people under different circumstances and with different measurement instruments increase the credibility of a finding.
4: Theoretical Plausibility. It is easier to accept an association as causal when there is a rational and theoretical basis for such a conclusion.
5: Coherence. A cause-and-effect interpretation for an association is clearest when it does not conflict with what is known about the variables under study and when there are no plausible competing theories or rival hypotheses. In other words, the association must be coherent with other knowledge.
6: Specificity in the causes. In the ideal situation, the effect has only one cause. In other words, showing that an outcome is best predicted by one primary factor adds credibility to a causal claim.
7: Dose Response Relationship. There should be a direct relationship between the risk factor (i.e., the independent variable) and people’s status on the disease variable (i.e., the dependent variable).
8: Experimental Evidence. Any related research that is based on experiments will make a causal inference more plausible.
9: Analogy. Sometimes a commonly accepted phenomenon in one area can be applied to another area.

However, if causation can be established based on these criteria, even an increase in relative risk of 1 percent may be significant, depending upon the extent of the exposure and the frequency of the outcome.

Linda

[Slimething]

What is funny is experiments have show using a mobile phone while driving increases your chance of having an accident by four times, yet many (most?) members of the public are not particularly concerned by this.

http://www.ncbi.nlm.nih.gov/pubmed/16854702?ordinalpos=1&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsP anel.Pubmed_DiscoveryPanel.Pubmed_Discovery_RA

It seems people are only worried about "invisible rays", or, more generally, aspects of a technology which they do not have understanding of and/or control over.

Great point, ItE. I credit this illogic more to human reaction to assumed (volutnary) risk vs involuntary risk. People will get willingly drive an automobile, one of the riskiest activities there is, without a second thought but tell them that they have an infinitesimally low chance of dying of cancer due to radiation exposure while flying and you'll get anger. "Why doesnt the gub'ment do something about it?"

[BillyJoe]

The problem is that "the general public" - which includes most of us* - do not know how to assess, interpret, and compare probability.

*If you buy lotto tickets, for example, you are one of these people. ;)