Posted on 01/08/2007 8:08:10 PM PST by Pharmboy
One study shows that milk can help people lose weight. Another shows that tomato juice might prevent cancer and a third shows benefits to fizzy sodas.
But consumers should take those studies with a grain of salt, researchers reported on Monday. If a study was industry-funded, it was far more likely to have a positive finding than if it was paid for by the government or an independent group, the researchers found.
"We are not singling out any industry or any particular study," said lead researcher Dr. David Ludwig of Children's Hospital Boston and Harvard University.
"Our first look shows evidence strongly suggestive of bias," Ludwig said in a telephone interview.
The study, published by the Public Library of Science online journal PLoS Medicine, echoes other findings that show industry-funded research on drugs is more likely to be favorable to the drugs than independent research.
Ludwig's team reviewed 111 studies on soft drinks, juice and milk that were published between 1999 and 2003.
"We chose beverages because they represent an area of nutrition that's very controversial, that's relevant to children, and involves a part of the food industry that is highly profitable and where research findings could have direct financial implications," Ludwig said.
Studies funded entirely by industry were four times to eight times more likely to be favorable to the financial interests of the sponsors than those paid for by other groups, the researchers found.
Of the 22 studies clearly identified as funded by companies or industry groups, just three, or 13.6 percent, had findings that were unfavorable to the beverage studied.
More than 38 percent of the independently funded studies were negative, the researchers found.
This "raises serious concerns that some food industries may distort the scientific record on diet and health," Martijn Katan, professor of nutrition at Vrije Universiteit in Amsterdam, wrote in a commentary in the same journal.
Ludwig said the studies could be set up differently if they are funded by industry. Or it could be that sponsors choose not to publish studies that turn out unfavorable to their product, he said.
Researchers funded by industry may do rigorous work, but may choose to ask certain questions more likely to produce a result favorable to the product, Ludwig said.
"I don't blame researchers for this problem. I think that most are highly ethical and dedicated to science," Ludwig said.
He said the problem is that the government does not spend much money studying nutrition.
"Industry money becomes difficult to resist," he said. "Imagine ... you are facing the choice of accepting industry money or closing up shop."
Ludwig's study was paid for by his hospital and by the Charles H. Hood Foundation, a childhood health philanthropy.
Ya think?
Who wrote this article, Captain Obvious?
"He said the problem is that the government does not spend much money studying nutrition."
Bull!
I know, we're supposed to look to government for everything.
The REAL problem besides looking to Uncle Sam for everything is that people are DISHONEST and SELF SERVING.
If I remember correctly, 90% of comparative drug studies find that the drug made by the sponsor of the study is the most effective. Scientists whose results turn out otherwise tend not to get funded for another study. The number of games that can be played with data is infinite, even without any actual cheating.
A good friend of mine was an animal technician in a lab where these types of studies were done. In one case they compiled the effects of the drug on the monkeys for 57 days and ran their comparisons. Yup, the drug was safe and effective.
The published study didn't mention that about half the monkeys died within a month of the end of the study period.
One caveat. "Independent" or government-funded studies can be just as biased as those funded by industry. A lot of these are "hits" on particular products or industries, as we saw in the infamous EPA secondhand smoke studies, and those compiling the data know exactly which side of the bread their transfat-free margarine is on.
Conversely, I would suggest the problem is that people studying nutrition don't work for free.
You are so right. Look at Global warming they are going to say what either or depending on who gave them the funding.
Except, of course, when it's Federal money that goes to study global warming—which money will be granted if and only if the scientists find a crisis to justify the expense.
For some reason, scientists on government grants here and in Europe have been finding that we are in the most horrendous climate crisis, which requires immediate action and much more study . . .
To even get published most studies need to show "significant" findings. IOW, if you have ten different studies and only one shows significance, which do you think the Journals will publish? It affects all Professional areas.
How to Lie With Statistics (Paperback)
http://www.amazon.com/How-Lie-Statistics-Darrell-Huff/dp/0393310728
Great book. Required reading for one of my college courses. It's very short but after reading that, you'll never believe any reports again.
Likely they just keep running these trials and studies until they get the statistical sampling they want and publish THAT one. I can't believe that they had to do a study on this, though. What next? A study to study the study on studies?
If these people were so concerned about obtaining objective results from scientific studies, they should do them without getting paid for their work.
The problem with government funded research is that the research grants are authorized and overseen by bureaucrats who will be using the results to argue for more regulatory authority and larger budgets based on the results.
File this under DUH!
Who has the 'Master of the Obvious' pic? It's needs posting.
Well...there are still people who believe in the purity of science. This revelation will upset some applecarts.
Suuuree, just look how accurate all those second hand smoke and global warming studies funded by the government or independent groups are
...unless it's a study done by the non partisan, left wing activists at our University (which studies show are the most intelligent and accepting amongst us).
On the other hand if the makeres of V8 Juice have an in housse study showing that V8 Juice reduces the chances of getting cancer then it is likley that Makers of V8 juice will fund a study to show the effects of V8 juice on cancer.
Corporations tend to pay for studies to lend crediblity to reports of benefits of products that they already know about.
On the other hand Government studies tend to be done when the government has no idea if the results will be positive or negative.
News organizations tend to believe that other corporations are no more moral and honest than they are. But that is not true. A corporation that tells lies in its products will suffer severly in the courts. A news organization that lies in its products is protected by the Constitution.
Yeah, remember how accurate the alar scare was. The "non-partisan" leftist groups come up with so much junk science, it's ridiculous. John Stossel regularly exposes their dishonest claims.
And if the money comes from left wing ideologues or biased grant makers and universities, we can just rely on it? Right?
I'll judge each study on its merits, not its sponsors.
And BMS had a blockbuster for high bllod pressure that failed in studies.
The Real fifi is one smart dudette...you are ALWAYS most welcome on any thread I start.
"If these people were so concerned about obtaining objective results from scientific studies, they should do them without getting paid for their work."
Oh, so that's WHY YOU believe the GOVERNMENT should be doing these studies?
Truly, you make no sense at all.
The government has NO BUSINESS in these studies. They need to remain in the private sector. What is in short supply out there are HONEST people who LACK SELF SERVING AGENDAS.
I agree completely. This is classic holier-than-thou stuff from academics, who are also capable of what they accuse everyone else of doing.
I love the line about how studies funded by industry could be "set up differently." We're supposed to read something sinister into that. Of course, different could be better! And to the extent it means worse, it should be obvious to this learned individual that studies set up by governmental and academic entities can also be "set up differently."
The problem is, John isn't on every station every night, so these myths have a growth pattern all their own.
I could write my own study to say what I want, but the problem arises when some bureaucrat quotes from it and the media runs with it night after night. If you saw that I wrote it, you might say, "who the hell is he?", but when Katie Couric and Pelosi run it over and over, then you have to affirmatively disprove it to be taken seriously. Some will never believe you are wrong once it's out there in the ether. Ergo, global warming, eggs are bad for you, silicone boobs cause cancer, etc. Silicone boobs were put into bankruptcy, but now all the sudden, they are ok again? Saccharine causes cancer, but I just put some in my tea? Second hand smoke is a killer, but the airline stewardesses lost their case in court because there was not one believable study to back them up.
Knowing there is not one credible case of second hand smoke causing cancer, ask someone if they believe it does or doesn't. They've known it for years, they just can't come up with one case.
That's a link to the article in question.
Does Industry Sponsorship Undermine the Integrity of Nutrition Research?
Thanks, pharmboy
All studies are biased...including this one.
Where's that Master of the Obvious graphic?
I have the 1954 edition; it's an interesting book, but there are some statistical issues it doesn't cover.
One which is rather technical, but nonetheless significant, is that two factors may be correlated with each other in subsets of the population without being correlated in the population as a whole. Unfortunately, mathematically expressing this requires eight variables:
a/b > c/d and A/B > C/D does not imply (A+a)/(B+b) > (C+c)/(D+d)Understanding this principle can go a long way toward unravelling issues of causation and correlation, though perhaps it's a bit too technical for the book.
Another technique of nonsense I haven't seen covered is what I call statistical homeopathy, in which vanishingly-small causal relationships imply increasingly-huge causal powers.
For example, some people in the area of a factory get sick; their sickness is blamed on emissions from a factory which is then forced to install polution-control equipment. This helps somewhat, but occasionally people--even many miles away from the factory--still get sick. Of course, people sometimes get sick for a variety of causes; the more remote the connection of a sick person to a supposed cause, the more likely it is that the sickness was caused by something else. Sickness in someone who should be too far away from the factory to have significant exposure is actually a data point against the theory that the factory is causing all the harm.
To a statistical homeopathist, however, every time a person gets sick is a data point proving the factory's culpability. The more remote the data point, the stronger the proof. If someone 5,000 miles away from the factory got sick, that wouldn't be seen as proof of the fact that people sometimes get sick for causes other than the factory; rather, it would be seen as proof that the factory's polution is so potent that it affects people 5,000 miles away.
You mean all those studies funded by Exxon that throw doubt on the human link to climate change despite overwheleming evidence to the contrary might be skewed? /sarcasm
Works both ways...agendas are agendas.
Of course, we'll never see a study showing us how studies funded by government are far more likely to find the need for more government control than those funded by private industry...
Government too is notorious for funding studies with an idea of what the outcome should be, and pulling the financial rug out from under those who come up with different findings. The latest example is Global Warming - I hear repeatedly that as a researcher you can get funding to study the right ways to mitigate GW, or to show the future extent of GW, but if your research doesn't come up with the right answers (GW is increasing and caused by humans, mass transit and increasing urbanization would help mitigate, reduce energy use by all citizens, reduce reliance on privately owned automobiles) your funding is either not renewed or canceled.
Not at all. Just pointing out the fallacy of thinking that government-sponsored research is free from being influenced by the self-interest of the sponsor.
Bureaucracies have a vested interest in finding "problems" to solve. It's their bread and butter.
I remember an old commercial for Camel cigarettes that stated that 9 out of 10 doctors switched to Camels.
It was found out later that 8 switched back to women. ;-O
What caused climate change before there were humans?
And another issue that was not raised, namely that of aggressive, ambitious researchers who falsify data in order to advance their careers. The ones we catch are the proverbial tip o' the iceberg...
"Not at all. Just pointing out the fallacy of thinking that government-sponsored research is free from being influenced by the self-interest of the sponsor.
Bureaucracies have a vested interest in finding "problems" to solve. It's their bread and butter."
True ... my apologies ... I over looked that nasty fact.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.