Monday, May 2, 2011


At the beginning of every change of the season some scientists publish a few papers in which they try to scare the pants off people. This time its “your calcium tablets may cause heart problems”.

I hate to say it again but it must be said: If the people who publish newspapers and magazines would read the actual reports in the medical journals they would not publish such hogwash.

This Spring’s publication is one that combines a number of studies which show no significant increase risk of heart attack or stroke. This is called “meta-analysis” which is another word for lumping a bunch of studies, some of which are poor, and seeing if anything comes out of it.

What happens when you do this is the number of people in the “meta-analysis” becomes large and this increases the likelihood that the average for all will become statistically significant. In this case the relative risk was barely significant (RR= 1.15 Confidence Interval; 1.03-1.27). Remember if the lower CI number is 1.0 the RR is not significant. So in this case, the RR is of marginal significance and would be considered by most to be of no consequence.

Remember if the relative risk is small the real or absolute risk (the only one that matters) is going to be exceedingly small.

This is very bad science and should not be published in any scientific journal. However, when it does get published the authors call the news media and indicate that the risk of heart and stroke problems “may be increased” by calcium tablets. Then the news media print headlines like “Calcium Tablets increase Risk of Heart Disease”.

Saturday, February 19, 2011


People tend to laugh when a scientist proclaims that if the length of your index finger is longer than your ring finger your risk of prostate cancer is lower than other men. But that was recently in the news. Can it be true? What about the data? Does it substantiate such a conclusion?

Well sort of, but not very convincingly. For one thing the data are taken from men who judged their own finger lengths from pictures which showed examples of various finger lengths, and then they returned this information to the investigators by mail.

Can you imagine Grandpa looking at those finger lengths examples and exclaiming “Mine is bigger than his.” or “What did cousin Homer put on his report? I can beat him any day.” The possible sources for inaccurate measurements are endless. Feel free to make up your own and after you check out your own finger lengths send them to me for publication.

Although the whole idea of finger length and risk of prostate cancer seems almost frivolous there is some basis for the belief that there might be a relationship. It turns out that the growth and pattern of the fingers and gonads (testes and ovaries) are controlled by the same genes and that the male hormone, testosterone, may control finger length.

Therefore, it is possible that in the male fetus the testes may produce relatively high testosterone and this in turn results in a short index finger. Whereas, if testosterone were relatively low the index finger would be longer. This guess work is based on some very squeaky data from another laboratory which attempts to show that fetal testosterone is correlated with length of the index finger. These data are shown on a scatter plot which looks like Wild Bill shot it with his 12 gauge shotgun.

So, my conclusion is there may be something to the finger story but better studies need to be done before any valid conclusions can be drawn.

Monday, February 14, 2011


As I explained in my blog on cholesterol and heart disease (posted on 8/14/2010), although dietary and blood levels of cholesterol were declared the enemy years ago it is now clear that for the vast majority of people cholesterol is not the enemy. Other highly touted villains in our diet like total saturated fat and animal fat have also bitten the dust. The latest epidemiologist’s horse by the name of trans fats has stumbled and has falling.


These findings have not deterred the statin drug industries who make Crestor, Lipitor etc. from continuing to propagate the false claims that cholesterol is your enemy and if you take our drugs you can lower your blood cholesterol and decrease your chances of heart disease. There is truth in these claims because these statin drugs will lower blood levels of cholesterol and they do lower the risk of heart disease, BUT it is not because of the lowered cholesterol. IF NOT, --WHAT?


So if the positive effects of statins on heart disease are not due to lowered cholesterol, then what are they due to? Recently it has become clear that statins block the inflammatory responses associated with heart disease. These inflammatory responses decrease the opening in the arteries and results in plaque formation. Some investigators believe that these responses occur as a result of infection by bacteria.

These inflammatory responses are spelled out in a little more detail below:

1. Endothelial malfunction.

The endothelium is the lining of the arteries where plaques are formed. Plaques are nodular accumulation of a soft, flaky, yellowish material and in larger plaques the center is composed of macrophages nearest the lumen of the artery. All of these things can be viewed as improper function of the endothelial lining due to an inflammatory response which is inhibited by statins.

2. Reduce remodeling of the arteries and heart muscle.

The smooth muscle cells in arteries increase in number during inflammatory episodes and this contributes to narrowing of the blood vessels and reduced blood flow. Also statins decrease production of smooth muscle cells seen after transplants.

3. Inhibit vascular inflammation and stabilize atherosclerotic plaques.

Statins decrease immune activation and exert anti-inflammatory
effects on the vascular wall by decreasing the number of inflammatory
cells in atherosclerotic plaques. They do this by decreasing the expression
of adhesion molecules so blood cells do not stick to the lining of the artery


You can bet that the drug companies are clamoring to come up with a
statin-like drug that is even better than the current statins at inhibiting inflammatory responses. But, it is not easy because statins all act the same way: 1. By blocking the synthesis of cholesterol and 2. By blocking the synthesis of compounds which act as inflammatory agents. Both of these actions have undesirable consequences.

Many people do not know that “Cholesterol is Us” as Pogo would have put it. We could not exist without it. The cells in our bodies can and do make their own cholesterol. It is a major component in our membranes and without it we would all be just be a blob on the floor. So it is obvious that blocking the making of cholesterol by the liver or any other organs in the body has many potential dangers. Some of these show up as reported side effects.

The clinical trials which test these drugs usually indicate the side effects are rare; however, in the real world that we live in side effects occur more often. These include muscle pain, muscle tissue loss, liver damage, nerve damage and memory loss to name a few. It is clear that these are dangerous drugs and prudent use is warranted.

The blocking of inflammatory responses appear to be the good side of the statin-coin; however, inflammatory responses are important and necessary in healing of wounds and infections. So blocking these may lead to troubles which are still lurking in the shadows.

Wednesday, January 26, 2011


Even as kids growing up in the 1930-40s we knew smoking wasn’t good for you. In fact, we said if you smoked a cigarette you were driving another nail in your coffin. Sure enough after years of research it was decided that smoking does cause lung cancer and other problems. But, the question for today is does it increase breast cancer?

In a new paper just published the answer is: Yes and No and Maybe (Xue et al 2011 Arch Intern Med 171 125). As usual, this epidemiological study presents marginal data some of which they say show a “modest” increase risk of breast cancer associated with smoking.

For example: for women who never smoked the absolute risk of breast cancer was 0.28% compared to those who smoked, 0.30% or an increase of 0.02%. Or put another way the odds are 28/10,000 if you don’t smoke and 30/10,000 if you do. Such a small difference cannot be considered significant, even though they say it is.

Other subgroup studies indicated some more marginal, just barely significant data. The more interesting subgroup study suggested that smoking before menopause was associated marginally with a slight increase incidence of breast cancer, whereas smoking during the postmenopausal period was associated with a consistent and significant decrease.

The most interesting data which the authors fail to discuss is in a subgroup consisting of women who smoked and took postmenopausal hormone therapy: The surprising answer: Whether they smoked or not there was no increase in breast cancer. How could the authors not mention that this finding contradicts the prevailing and incorrect assumption that postmenopausal hormone treatment increases the incidence of breast cancer??? Could it be that they are embarrassed by the possibility that they were wrong about this important point???

At any rate, the paper is just another example of the kind of epidemiological data which should never have been published. All of you smokers out there need not worry about breast cancer, that is the least of your worries.