The Endless Supply Of Fake Scares And Statistical Scams
Back in May I wrote about the mortal blow suffered by the anti-salt campaign when the Institute of Medicine decided to do a real study and apply some actual statistical analysis to the results. Turns out that there is no evidence at all that a low salt diet is good for you, and some that it might even be bad. Meanwhile, a long list of power hungry busy-bodies -- the Federal Departments of Agriculture and HHS, the WHO, the American Heart Association, Mayor Bloomberg -- got exposed for trying to use a plausible hypothesis and some fake math to seize a little more control over your life. I'm still waiting for any of them to apologize.
Well, salt is of course just one example. If you start looking around, there is an endless supply of these fake scares and statistical scams. And the essential ingredients are always the same: a plausible hypothesis, little or no evidence, fake math, and power-hungry busy-bodies looking to get more control over your life. In looking at these things, it doesn't matter if you don't know much math and never took statistics, because there is only one thing that you need to know to tell the real from the fake: Relative Risk of less than 3 is meaningless. If the Relative Risk is more than 3, they will tell you, because this will universally and immediately be recognized as a significant result. If they aren't trumpeting a Relative Risk of over 3, then they don't have it and the whole thing is a scam. It's easy!
Let's take a couple of examples for today. First, the low fat diet. The idea that we need to reduce fat in the diet is by this point so completely ingrained in our culture that it is second nature. You can't go to the supermarket without seeing dozens of low fat and reduced fat items, all playing off the multi-decade browbeating of the consumer into believing that lowering fat will improve heart health and help keep weight under control. The anti-dietary fat campaign is everywhere! And certainly, it starts from a seemingly plausible hypothesis -- "fat" makes you "fat"! Get it? Now go to the Dietary Guidelines for Americans put out by the Center for Nutrition Policy and Promotion of the Department of Agriculture. This is the Official Word of the Federal Government -- surely they know what they are talking about! Click the link for Chapter 3 - Foods and Food Components to Reduce. You will find that the first four pages of that chapter are about the need to reduce salt in your diet (uh-oh!), immediately followed by the next three pages on the need to reduce fats. The section on fats is filled with seemingly authoritative statements like this one:
A strong body of evidence indicates that higher intake of most dietary saturated fatty acids is associated with higher levels of blood total cholesterol and low-density lipoprotein (LDL) cholesterol. Higher total and LDL cholesterol levels are risk factors for cardiovascular disease.
A "strong body of evidence" you say? Can you please tell us the Relative Risk? Of course not. Actually, I have read three full-length books on this subject, and if there's one thing that's clear it's that the attempt to demonstrate a link between dietary fat and cardiovascular disease has yielded zero, zilch and nothing. They're making it up.
Actually it's worse than that. Starting in 1993 the government actually commissioned and paid for one of those massive studies specifically intended to gather evidence on whether there were any health benefits to a low fat diet. It was called the Women's Health Initiative Dietary Modification Trial, and it enlisted almost 50,000 women between the ages of 50 and 79. This was one of those randomized studies -- 19,541 assigned to follow a low fat diet, and 29,294 assigned to continue the usual diet. The results were reported in the Journal of the American Medical Association in 2006. Here is a summary report on the study from the Harvard School of Public Health. And the results?
The results . . . showed no benefits for a low-fat diet. Women assigned to this eating strategy did not appear to gain protection against breast cancer, colorectal cancer, or cardiovascular disease. And after eight years, their weights were generally the same as those of women following their usual diets.
So, kindly explain why the Dietary Guidelines, most recently issued in 2010, continue to contain the advice to reduce fats? The answer is that these people do not give up power willingly, and do not apologize for being wrong, even spectacularly wrong. All I can say is, I hope you aren't wasting your time trying to reduce the fat in your diet. Let's have a steak for dinner!
Now how about example 2 for today -- "environmental tobacco smoke," aka ETS, aka second-hand smoke. Surely we all know that that stuff is dangerous. The EPA has told us so!
I should start by mentioning that I am no fan of tobacco smoke, and I found the days of thick tobacco smoke in bars, restaurants, airplanes, and offices quite unpleasant. The evidence that smoking is harmful to health of the smoker is strong. That of course gives the zealots their plausible hypothesis that second-hand smoke must also be harmful to health. But is there real evidence that ETS is harmful to health?
The answer is, not that I can find. Go to the basic EPA page for the public on this subject, and what you get is mumbo jumbo authoritative statements just like in the Dietary Guidelines. "Exposure to secondhand smoke can cause asthma in children who have not previously exhibited symptoms." "Infants and children younger than 6 who are regularly exposed to secondhand smoke are at increased risk of lower respiratory track [sic!] infections." But no evidence. There is a link to a 2006 Surgeon General report. That report, it turns out, does contain long lists dozens of studies of subjects exposed to ETS and their development of certain medical conditions. And they do have a column in their charts to report Relative Risk. And for the large, large majority of the studies the result in that column is "NR" - no result. And for the rest of the dozens of studies, the results are almost universally a RR of well under 2. I managed to find one such study with a RR of just over 3 and a total of 29 subjects.
Frankly, this is insulting to our intelligence.
Unfortunately, an awful lot of what you think you know is one sort or another of fake scare or statistical scam.