Sunday, January 20, 2008

Getting to the Heart of Heart Disease

BusinessWeek recently ran a nice article on the actual benefits of cholesterol drugs. They discuss the scientific flimsiness of the entire lipid hypothesis as well. It's interesting (and good) to see the mainstream press picking up on this discussion, because despite appearances, there really has never been anything close to scientific consensus as to the role of fats and cholesterol in causing heart disease.

The behavior of people is remarkably consistent. We previously saw in the context of the AIDS/HIV hypothesis how the "pro" group simply heaped scorn and anger on any critics, rather than presenting the supposedly "overwhelming" scientific evidence supporting their position. Those in favor of the lipid hypothesis engage in similar behavior, as we see in this excerpt from the BusinessWeek article:

Add it all together, and "current evidence supports ignoring LDL cholesterol altogether," says the University of Michigan's Hayward. In a country where cholesterol lowering is usually seen as a matter of life and death, these are fighting words. A prominent heart disease physician and statin booster fumed at a recent meeting that "Hayward should be held accountable in a court of law for doing things to kill people," Hayward recounts. NECP's Cleeman adds that, in his view, the evidence against Hayward is overwhelming.

Great, except that one supposes in order to prevail in a court of law, you'd actually have to supply said evidence. I don't know whether the evidence exists in the AIDS/HIV case, but I'm about 99.999% sure it ain't there for the lipid hypothesis.

I'm not going to go over the case against the lipid hypothesis in detail, since it has been covered in many other places. Here's a few books and links you can check if interested:

Taubes' book dismantles the whole business in detail, providing a lot of scientific and sociological insight. Kendrick's book is less scientifically dense and an easier read, and quite funny. Try dropping some of the info you learn on your doctor next time he recommends a cholesterol test (or statin treatment). You'll likely witness a great example of cognitive dissonance, much as Nigel Tufnel did when asked to explain why his guitar amplifier "goes to 11".

What I want to talk about today is a sort of mental exercise I undertook after reading some of the above material, along with the Robert Sapolsky's most excellent book Why Zebras Don't Get Ulcers. The idea is to start with physical phenomena that are likely directly causative for heart disease, then try to find whatever might cause those things, and keep working backward until we hit causes (like diet or infectious agent) which originate outside of the body. We can then make a graph of sorts, where the various phenomena are connected by arrows to show how they are related (an arrow from one box to another means that the thing in the first box potentially causes the second).

Click on the image at the left to see what I came up with in the case of heart disease. I'll be the first to admit this is far from complete, but it should at least be thought-provoking. It proves nothing per se, but the kind of logical relationships shown here should at least be reflected in any theory of the cause of heart disease, as I believe the relationships shown are well-established in the scientific literature (and if you have different information or something to add, please post a comment).

Okay, enough disclaimers. What does this mess mean? The red boxes are supposed to represent the direct causes of heart disease, i.e. those conditions which directly lead to plaque formation. I picked these because it seems fairly clear at this point that arterial plaques form as part of the body's response to injuries of the endothelium. The endothelium is a very thin layer of cells which line the entire circulatory system, and amongst many crucial functions, it controls the passage of material in and out of the blood. Even a tiny injury requires repair, and it is thought that the "patch" is actually the beginning of an atherosclerotic plaque.

I don't know if there exists any direct evidence for this phenomenon, but the indirect evidence is pretty compelling. First of all, plaques only form in arteries, where blood pressure is high. Indeed if you transplant a vein to replace a damaged artery, the newly transplanted vein will itself begin to show plaque formation. The plaques themselves are localized. People often think that "clogged arteries" is just junk building up over time, much like a clogged drain. But the plaques form in local spots, preferentially in areas where the blood pressure is the highest, and/or the blood is forced to change direction, like where arteries fork off.

When your skin gets a little cut or nick, a blood clot or scab forms. The skin grows back underneath the scab, which eventually falls off. But this wouldn't be a good idea for repairing blood vessels. If a scab formed and then fell off when the underlying endothelium had grown back, that scab would likely travel down the artery until it got lodged, causing stroke or heart attack. It's also important that the integrity of the endothelium is preserved, so whatever patches up the injury should be water resistant, so material does not leak across at the injury site. hat is thought to happen is that various blood proteins quickly rush to the injury site, forming a water-resistant patch. One of the main players is low-density lipoprotein, or LDL, the so-called "bad cholesterol". Special enthothelial progenitor cells, which are always circulating in the blood, then grab hold and form a new endothelium on top of the patch. In a healthy individual, the plaque may then be slowly absorbed by the body.

There are two possibilities by which a plaque (technically called an atheroma) might grow. The first is that the endothelium which overlies the existing plaque sustains an injury, causing another patch and new endothelium to be laid down on the first one. The second can actually occur from the inside. LDL is prone to oxidation, as are certain kinds of lipids that LDL typically carries, namely cholesterol and unsaturated fats. Oxidized stuff will provoke an immune reaction, with white blood cells "eating" the oxidized LDL. What happens next is less clear, but for some reason it seems a lot of these white blood cells die before leaving the plaque, causing a buildup of "foam cells". If a plaque filled with all of this junk undergoes a major rupture, spilling its contents into the blood, you get an immediate strong clotting reaction, potentially leading to heart attack or stroke (most likely further up the artery as it narrows).

So we see the two main phenomena at the root of plaque formation and subsequent "bad" outcome (heart attack or stroke) are injury to the endothelium and inflammation in the form of immune response and clotting. Thus any conditions which promote endothelial injury will cause more plaque to form. Any conditions promoting inflammation can cause the plaque to grow faster, or a more severe clot to occur when the plaque ruptures.

What promotes endothelial injury? High blood pressure is a biggie, no doubt, and high blood pressure is generally a sign of systemic inflammation. Of course anything that damages the integrity of the endothelial cells also likely promotes damage. This would include glycation, a process where sugars damage proteins (fructose is one of the major culprits promoting glycation damage), or oxidation damage from free radicals in the blood. And we keep working back from those causes until we start hitting the root exogenous (from outside the body, basically stemming from "lifestyle") causes, and we've got the headache-inducing graph I posted above.

Now as I said, that picture is probably incomplete. For example, I later learned about B-vitamins and homocysteine, and for simplicity I left out the effect of cigarettes. But what I find interesting is that for all of the internal complexities, everything ultimately traces back to a few root causes, really two, if you ignore cigarettes or other drugs: psychosocial stress and crappy food.

It's worth talking a bit about stress. We usually think of stress like "my job is really stressing me out." But stress is fundamentally a hormonal response, governed by the master glands of the sympathetic endocrine system: the hypothalamus, pituitary, and adrenal glands. Together, these form what is known as the "HPA axis", and activation of the HPA axis (regardless of the source) fundamentally defines the stress response.

The stress response is evolution's way of dealing with major crises that potentially threaten the survival of the organism. Suppose a cave bear decided to snack on one of your distant ancestors. Physical responses like increased heart rate and blood pressure, clotting response, immune response, etc. could come in handy, assuming said ancestor avoided actually being eaten. Most of the physical reactions caused by stress are detrimental over the long term; the system is designed to act in emergencies only, where the presumed trade-off is between bad (e.g. short-term increase in arterial damage) and worse (death). This point is illustrated nicely by the title of Sapolsky's book Why Zebras Don't Get Ulcers: zebras don't get ulcers because they only "stress out" when required, like when being chased by a lion. Otherwise, they're cool.

Humans, on the other hand, seem to be very adept at ramping up the stress response in the absence of a life-threatening situation. The major modern source of stress is termed psychosocial, originating from many sources, like unhappiness in a job, divorce, money troubles, etc. The problem with psychosocial stress is that it tends to occur over long time periods, and causes the same physiological responses that your distant ancestor experienced while trying to avoid being cave bear food (this actually suggests one way of dealing with stress: compare your problem with that of being eaten alive, and adjust accordingly). The negative effects of stress are then visited upon the body of an extended time period, and the damage starts to pile up.

Oddly enough, certain foods can actually reduce stress in the short term, but increase it in the longer term. Insulin is the "master" metabolic hormone, and comes from the parasympathetic endocrine system. The sympathetic (HPA) and parasympathetic systems are generally antagonistic: when the sympathetic system is activated, the parasympathetic system is inhibited, and vice versa. During the stress response, insulin is inhibited, allowing the blood levels of energy-giving macronutrients (glucose, fats, and proteins) to increase, making them available as necessary for, say, leg muscles so you can run away from a hungry bear. Interestingly, it works the other way as well. If you eat food that raises your insulin, the HPA axis will be suppressed. In short, sugary food reduces stress, which nicely explains why most "comfort foods" are high in refined carbohydrates.

The problem is that all of the insulin that gets released when you eat said comfort foods ultimately drives much of the energy nutrients out of the blood and into the cells, resulting in insulin-induced hypoglycemia, or low blood sugar. Your body interprets this state as starvation, and as you might imagine, starvation induces stress (see this article for more info). So now you've not only got your original psychosocial stress, but additional stress from eating junk food, which then induces you to eat more junk food to get a temporary fix, and around and around you go. And to make matters worse, all of that sugar you're dumping into your blood is promoting glycation damage of the endothelial cells, so it's a double-whammy.

One final piece falls into place from junk food as well: bad fats. Junk food tends to be replete in vegetable oils. Most vegetable oils are polyunsaturated, which makes them prone to oxidation. Ever notice how the top of your bottle of "healthy" vegetable oil gets sticky? That's the result of oxidation. Fats are generally transported in the blood via LDL, and if that fat in the LDL is oxidized, it can in turn oxidize the LDL, and we've already seen what kind of mess that creates. Further, most of these vegetable oils are high in omega-6 fatty acids, which the body tends to use to create pro-inflammatory hormones. They are deficient in omega-3 fatty acids, which the body uses to make anti-inflammatory hormones, and high insulin further inhibits conversion of any vegetable omega-3's (alpha linolenic acid) into the forms required by the body. So the net effect is to further increase the inflammation already present from chronic stress, as well as increase the problems associated with oxidation.

Alright, that's a lot of information. Let me just leave with one final observation. You might note in the picture that there are some boxes that don't have arrows leading out of them, namely the following:

  • Increased LDL/decreased HDL
  • Increased triglycerides
  • Obesity

These are all identified as "risk factors" for heart disease, yet the causal connection is weak, at best. For example, you might argue that increased serum LDL implies that more LDL will be deposited in the arterial patch; but if the rate of endothelial damage is low, it's pretty much irrelevant, since you won't be making many of these patches in the first place. It seems to me (and this is strictly my opinion) that the traditional risk factors are symptoms of the underlying cause. It is well-known that chronic stress increases LDL and decreases HDL. Triglycerides are formed by the liver from excess blood sugar, i.e. when you eat refined carbohydrates. And everything that is known about fat storage at the molecular and cellular level indicates that obesity arises from some combination of stress and consumption of refined carbohydrates. If this is true, then current treatment recommendations like low-fat diets and statins do nothing but mask the symptoms while the underlying disease continues to progress for exactly the reasons that caused it in the first place.

So relax, and have a steak.


GreatnessBlog said...

I've done a little informal poking around since our last discussion about HIV/AIDS and frankly, I can't imagine why you are skeptical that HIV causes AIDS. There appears to be plenty of evidence not only of the virus, but of the virus's timeline without medication and with various medications. We can inject a baboon with a quite similar virus -- HIV-2 and, lo and behold, they get all of the symptoms of AIDS. People who test positive for HIV who aren't treated are so close to 100% likely to get AIDS that it's hard to imagine some unrelated mechanism involved.

The reaction of some scientists to those who question the "hypothesis" is perhaps rude and counterproductive, but young-Earth creationists probably get a similar response when they question paleontologists or cosmologists. That doesn't mean they are right.

I think this combining of the HIV/AIDS hypothesis with the lipid hypothesis is bad for your cause. Imagine how you would feel if you found out that Taubes wrote a book strongly arguing that the Earth is literally flat or that automobile collisions don't cause fatalities.

Dave said...

Hi GreatnessBlog. Thanks for the link and information. That's exactly the kind of thing I was looking for. I will check it out and let you know what I think.

I agree with your point that scientists often badly react to those who disagree with the prevailing hypothesis. Whether said hypothesis is "right" or not, it is usually still a case of the defending scientist holding a dogmatic belief in that hypothesis. It's bad science either way.

I don't have a "cause", other than to encourage people to obtain good information relating to health issues (just as you've done) so that they can make decisions in their own best interest. If the evidence for AIDS/HIV (or the lipid hypothesis, for that matter) does lean heavily one way or another, people should be able to discover said evidence, rather than rely on "experts" with questionable motivations. Questioning hypotheses is never bad in science, as long as it done in a rational way (creationists generally fall into the irrational category, as they are driven by their own dogma). Dogmatic acceptance of a hypothesis is always bad in science.

I actually wouldn't feel strongly if somebody published the books you suggest. After evaluating the evidence, I would try to make a judgment as to which hypothesis is correct. There's no reason to get emotionally attached to hypotheses - their merit lies strictly in the weight of the evidence which supports them, including their predictive power.

Dave said...

Hi GreatnessBlog. Good stuff in the link you sent. I'm glad to see that effort was made to answer the "dissidents". I must admit I was a little underwhelmed by a lot of the evidence presented. For example, the baboon thing is interesting, except that as far as I know (anybody know differently?) the HIV virus cannot be isolated and purified. So said baboons were not only injected with HIV, but also with other stuff. It is my understanding (albeit limited) that injecting animals with random proteins can cause a wide range of immune reactions. I'd also like to see these results replicated on a larger number of animals. For instance, chimpanzees will generate HIV antibodies when injected with the virus - do they also progress to AIDS, and ultimately death? For that matter, did the baboons die, or did they recover? I can't get at the full text of the article, maybe somebody who can will summarize it for us.

Even if chimpanzees don't get AIDS or the baboons did recover, it doesn't necessarily mean that HIV doesn't cause AIDS. Chimpanzees and baboons are not humans. But it would be nice to know the outcome of that experiment.

The site you linked is dated 1995. I'm hoping further progress has been made in the ensuing 13 years. Maybe somebody can point it out? There still seem to be an awful lot of people who come across as sane and responsible scientists asking tough questions. This 2007 interview with Andrew Maniotis is a good example.

One key question I have involves the mechanism of a "slow" virus, of which HIV is supposed to be one type. The slow virus idea is that somehow your body gets infected by a virus and generates antibodies to that virus. The virus becomes latent for some period, perhaps many years, and then causes a disease state. This is supposedly not the same as "latent" viruses like herpes, which can hide from the immune system and pop up during episodes of immunosuppression. Can anybody point me to a good explanation? Thanks.

Dave said...

Here is an article detailing recent research linking stress and heart disease.