Low Carb Diets II

The new Dietary Guidelines for Americans 2005 is specifically targeted towards preventing obesity. The previous food guide pyramid from 1992 emphasized a reduction in saturated fats to prevent heart disease. With an estimated 300 million worldwide expected to be afflicted with Type 2 diabetes in the near future, mostly due to obesity, we have a new priority. The guidelines recommend a diet that is high in fruits and vegetables, low in fat, and low in sugar. The latter recommendation was especially contentious with the food industry.

As expected, low carbohydrate diets like the Atkins diet were not held in high regard. However, there really isn't overly strong evidence that they are unhealthy and in some cases they have been found to actually lower blood lipid levels! As a weight loss measure, low carb diets do seem to work although keeping that weight off is another matter as I discussed previously.

The Atkins idea is that carbohydrates make you fat. When you eat carbohydrates they get broken down to glucose (except fructose) which then triggers an insulin response. Insulin allows muscles to uptake and burn glucose (instead of fat). At the same time it suppresses the release of free fatty acids from adipocytes (fat cells). Thus when insulin levels are low, the muscles mostly burn fat (except during exercise). Glucose is thus spared for the brain which cannot burn fat. This much is true. However, Atkins also claims that when insulin levels drop after a meal, you get a strong hunger response. The data is not so clear on this point.

To make up for the lack of carbohydrates, you must eat more protein and a lot more fat. Thus low carb diets are high fat diets. The traditional Inuit diet is an Atkins diet. When you begin a low carb diet, the first thing that happens is that you get depleted of your glycogen (which is the body's only store of carbohydrates). This is accompanied by a loss of water so you lose a lot of weight quickly. Your body then goes into a state of ketosis where the liver makes ketones. The brain only burns glucose or ketones. No one knows if maintaining ketosis for prolonged periods of time is detrimental. The fact that the Inuit did it for generations probably means it's okay.

There may be other reasons for why the Atkins diet works. For one, the diet does limit calories. Secondly, there is some data that shows that restricting food choices can result in eating less. Thirdly, until recently, there were very few snacks that are low in carbs. Finally, the quick weight loss in glycogen and water may motivate people to stay on the diet.
However, people do eventually give up and the weight inevitably returns after two years.

I personally think, from a health point of view, that controlling the total amount of calories consumed is more important than the composition of the diet. If you are in energy balance, you will basically burn everything you eat so it doesn't really matter if it is mostly fat or carbs. When you overeat, you are going to store that extra energy as fat. The data shows that losing just a little bit of weight can greatly reduce insulin resistance which is a precursor to Type 2 diabetes. Exercise also seems to confer benefits that go beyond the extra calories burned. So although eating lots of fruits and vegetables is probably good for you, if eating pork rinds helps you lose weight, then stick with that.

Oxygen

In the debate over global warming, little has been said about its affect on oxygen balance. Atmospheric oxygen serves two crucial purposes for sustaining life on earth. In the upper atmosphere it takes the form of ozone and blocks ultraviolet radiation. Without this layer, no life on the surface of the earth could exist. The second purpose is as an oxidant for respiring life forms. Most of the oxygen on earth is sequestered in rocks like silicon dioxide. Atmospheric oxygen comes exclusively as a waste product of photosynthesis of plants and prokaryotic organisms. The bulk of atmospheric oxygen comes from photosynthesizing bacteria near the surface of the oceans.

The air is comprised of 78% nitrogen, 21% O₂ and the rest as trace gases such as argon and carbon dioxide. The oceans also contain a large supply of dissolved carbon dioxide and oxygen. As water warms, it can hold a smaller amount of oxygen. Thus, as the oceans warm due to global warming, they are actually outgassing some oxygen (Keeling and Garcia, PNAS 99:7848-7853, 2002) . At the same time, oxygen in the atmosphere is also slightly decreasing due to combustion of fossil fuels.

Right now the decrease of oxygen in the ocean is 0.7 µmol per kg per decade. For comparison , the concentration of oxygen in the ocean is on the order of a few hundred µmol per kg. So the decrease is small but not insignificant. Given that one hypothesis for the extinction event 250 million years ago known as the Great Dying (see my previous entry) was due to a lack of oxygen in the oceans, I think that this is of some concern. We simply don't know enough to predict the consequences of global warming. However, I really think we need to take it seriously. Affecting the carbon balance could have implications beyond the melting of glaciers and the increase of hurricanes. It could affect the oxygen that we breathe.

Gender Differences?

As the sides line up on the Larry Summers debate/debacle, more data is coming in. One of Summers suggestions was that women were less likely to want to make the sacrifice required for an intense academic job. An indication would be that the attrition rate for women would be higher throughout the training process. However, as reported in the New York Times today, after earning a bachelor's degree in physics, American women are just as adept as men in climbing the academic ladder. A full report can be obtained from the AIP website. Currently women make up 22% of bachelor's degrees and 18% of PhD degrees awarded in physics . In astronomy the numbers are much higher - in 2003, women earned 43% of bachelor's degrees and 26% of PhDs. There also seems to be a huge surge in women in the last five years into both fields. Women comprise of 10% of physics faculty now and their numbers have been slowly increasing over the past few years.

The data shows that the main source of attrition of women from physics is between high school and college where half of the students in high school physics are females but fewer than a quarter of the degrees go to women. This supports Summers assertion that the problem may arise early. However, Summers then contended that this early gender disparity may be biological although he conceded that there is no evidence and challenged the audience to do the studies to test the hypothesis. As pointed out in the comments section of my previous post, Brad De Long shows data indicating that social factors may swamp out any biological effects.

My stance, which I've argued in Steve Hsu's blog, is that even if there are biological differences, they are small and this knowledge is not useful and is potentially damaging. The data clearly shows that policies enacted over the past several years are having some effect. The percentage of women in science and engineering is increasing. We should continue to eliminate social barriers and wait another generation or two. If disparities still exist then perhaps we can start looking for other reasons.

Science Job Market

Steve Hsu has been arguing in his blog that young people should not be encouraged to pursue a career in science. He feels that they would find greater success if not satisfaction in a more lucrative field like finance, engineering, law or medicine. His argument is based on the premise that there are a plethora of well educated and well trained people from eastern europe, the far east and south asia that are willing to endure more hardship and take less pay for the same job. In the end, the top American students may waste 10 years pursuing an academic job that may never materialize. The various merits of this argument is debated in the comments to Steve's posts. I do not deny that competition for academic jobs in science is fierce but why should it be less so in these other fields?

Law and especially medicine restrict the supply by limiting the number of students admitted to their respective professional schools and making it extremely difficult to practice in the US if one is trained in another country. Thus, in terms of these fields, our hypothetical top student interested in science would need to out compete other Americans to attain coveted slots in these schools. If they succeeded (and that is not guaranteed by any means) in getting admitted, employment as a physician or lawyer is reasonably certain if they went to a top rated school. From what I've seen, insurance companies have been declaring war on physician salaries for the past several years and law does not seem to be a guarantee of a high salary or job security.

Engineering and related discplines like computer science certainly have more options in industry than say physics or mathematics. However, it may not have many more opportunities than say chemistry or biology. Additionally, the same cohort of foreign students competing for jobs in science should also be competing for these fields. We have all heard about software outsourcing lately and I'm certain it will only get worse.

Finance is the area that Steve was really referring to in terms of missed opportunities for the top science student. Between ten and fifteen years ago, a wave of physicists entered finance. Many have done exceedingly well and have retired or could retire. However, I think that well is drying up as well. I don't see why competition from foreigners won't be any less fierce, if not fiercer here. The brightest students in other countries will probably follow the same advice Steve is giving. Correct me if I'm wrong but I bet that purely quantitative jobs are probably not as lucrative as they were in the hay days. The only way to really succeed in finance is to advance into a leadership position and being able to do that requires a skill set that is independent of those that are optimal for science.

So I think the bottom line is that if you are a truly spectacular student then you could probably be successful in medicine, law, engineering or finance. On the other hand you probably would be successful in science as well. Steve's argument hinges on the fact that it takes a long time in science before you know if you have what it takes. Given that the average American will change careers five times in their life, I'm not sure if that is not true elsewhere as well. I think the bottom line is that we now live in a global economy and this will affect everyone. As they say, there is no free lunch.

Smallest multicellular organism

I've been searching for the smallest multicellular organism and there do not seem to be any adult creatures with fewer than a thousand cells. For some reason there seems to be no evolutionary advantage for being say an organism of two cells or three hundred cells. This seems to also hold true for colonies of cells like sponges or algae. No one seems to have an explanation for why this would be true.

The diversity among organisms on the order of a few thousand cells is immense. On the one hand we have the nematode worm C. elegans which has 959 somatic cells with a nervous system of about 300 neurons. It has muscles and a metabolic system that operates surprisingly like humans. It reproduces sexually with sperm and egg. It's genome has 100 million base pairs encoding an estimated 17,800 genes.

On the other hand we have Trichoplax adhaerens which is a candidate for the simplest multicellular organism. It is the only species in the phylum placozoa. Trichoplax is comprised of a few thousand cells that differentiate into four types. It has no neural or muscular systems. It basically looks and acts like a large amoeba. It reproduces by binary fission or sometimes by budding although sexual reproduction may be involved like yeast. It has the smallest genome of any known animal at 50 million base pairs which is only a factor of two smaller than the nematode.

Both animals are about the same size - a few millimetres in length - and both have roughly the same number of cells but they have employed drastically different strategies for survival. So it seems that the constraint on minimum number of cells in an animal is not one of limited strategies. Perhaps is is a result of a constraint of molecular biology or cellular physiology.

Misuse of language

The way we misuse phrases and quotations may tell us something about how our brains work. I give three common examples. The first is "Now is the winter of our discontent", which comes from Shakespeare's play Richard III. The line refers to an ebb in discontent (for the house of York in Richard III's case). Winter is meant to be a modifier of discontent. However, people almost universally use discontent as a modifier of winter as in there is an entire season of discontent and now we're in the winter part of it. Even Steinbeck used it as the title of his book to indicate a state of disaffection. Some rephrase it to "Winter of Discontent" which could be thought of as a clever ironic pun of Shakespeare. However, if you think about it "Winter of our discontent" makes most sense when used as Shakespeare did.

Now why is it misused? I think part of it is laziness. Most people have never read Richard III. However, it may also be evidence that our memory for language is an attractor (Hopfield) neural network. In this idea, memories are attractors in a dynamical system each with a basin of attraction. "Winter of our discontent" is remembered as a complete phrase separate from its component words. If it was incorrectly associated with disenchantment the first time we encountered it, then it may be frozen as such. So the next time we wanted to express the sentiment (and who is more poetic and profound than Shakespeare) we would call up this quote without an examination of the true meaning.

The second misquote is "If music be the food of love, play on" from Shakespeare's Twelfth Night. Most people use this quote as a positive statement about music. However, it actually is a negative statement about love. In the play, Orsino is heartbroken and wants to hear so much music that he will be come nauseated and no longer have any interest in love. In this case, I can see why it would be misused the way it has. Music is a surrogate for love so let's have more of it.

The third oft misused phrase and one that irks me the most is "begging the question". This term, which refers to a circular argument that presumes the truth of an assumption, is commonly used as a substitute for "raises the question". I've heard many hyper-educated people misuse this phrase. People will say "The fact that X happens begs the question...". I think in this case, people have heard this term in their youth but never understood what it really meant. However, it remained lodged in their memory and is now retrieved whenever a situation calls upon a particular question. In time I think this new and incorrect usage will eventually dominate and become accepted. This is too bad because now more than ever we really need to challenge those that "beg the question".

Principle of Least Action

I've always been amazed at how much faith physicists have in the principle of least action, which says that all classical objects travel along a path that minimizes the action. In quantum mechanics, particles will take all possible paths but the amplitude is weighted by the classical action along the path (so the most likely path will be the one of least action). Steve Hsu discussed the path integral, which sums over the paths, in his blog recently.

From what I've seen as a spectator, "Theories of Everything" have fiddled with symmetries, spatial and temporal dimensions, been comprised of strings or branes or what have you but they've all been based on the premise that the quantum amplitude is giving by a sum over paths weighted by some action. I'll be the first to admit that this approach does work amazingly well for all the energy scales that we have tested thus far. Quantum electrodynamics is the most accurate theory we have.

However, we should not forget that least action and it's quantum cousin are assumptions. They were discovered empirically. Yes, in order to make any progress we must make assumptions but sometimes these assumptions are so internalized I don't think people even realize they are making them. The greatest breakthroughs often involve questioning our basic assumptions. Relativity, quantum mechanics, plate tectonics, democracy, and so forth all arose as fundamental challenges to the prevailing status quo. So even if M theory turns out to be the right theory some day it still won't tell us why we have the principle of least action.

What is life?

It is notoriously difficult to find a satisfactory definition of life. No matter what definition you come up with you can always find a counter-example. I've come to believe that the reason is that you can't separate a life form from its environment. To go even a little further, I don't think life can be separated from the underlying dynamical laws from which it arises. Thus, the fact that macroscopic life forms on earth tend to be discrete and identifiable is incidental. We can come up with ad hoc rules to define them but there will always be exceptions. One could imagine an ocean of self-catalyzing molecules (i.e. RNA world) that reproduce, exchange information and behave as a giant life form. Morris Hirsch and Freeman Dyson both have suggested that you could have electromagnetic life forms living in a plasma cloud of charged particles.

So my definition (which I'm sure is still flawed), is that life is a universe comprising of a set of dynamical laws acting on a set of constituent particles in which entropy can decrease in a localized region that is much larger than the scale of the particles and for timescales longer than the average interaction time of these particles. Implicit in this definition is that the localized region is in a state of quasi-equilibrium so that entropy can be defined. Thus a set of non-interacting particles would not be alive because entropy lowering fluctuations only last as long as the particle interaction time (or particle transit time in this case).

I have no idea what would be a minimal set of laws or rules that could support life. Maybe life-possible rules fall into Steve Wolfram's select class of cellular automata (CA) that exhibit nontrivial complexity. Wolfram's argument is that you cannot predict a priori what a set of CA rules will do. There is no short cut to deducing what patterns can arise without actually iterating the rules. Thus, life itself is the ultimate grand experiment. I've always found great solace in Andrei Linde's multiverse idea. Universes with different physical laws pop in and out of existence, and no one can predict if life is possible in a given universe before it exists.

Evolution and Ernst Mayr

The theory of evolution has been in the news quite a bit lately. This past week marked the passing of evolutionary biologist Ernst Mayr who has been described as one of the most influential scientists of the 20th century. Mayr is most noted for his 1942 book - Systematics and the Origin of Species, where he outlined his theory of allopatric speciation. This view, which was actually rejected by Darwin, held that speciation could only take place if populations were geographically separated. It was Mayr that introduced the currently accepted definition of a species as a population of interbreeding organisms. It is only through spatial separation that evolution would lead to enough alterations that would prevent interbreeding.

I don't think the general public realizes how important evolution is to modern biology and medicine. The reason why experiments on bacteria, yeast, worms, flies, and mice have any impact on human physiology is that we share a large portion of the same genetic material. For example, 50% of the genome of the tiny nematode worm C. elegans can be found in humans. Evolution gives us a framework for understanding why this is so and where to look next.

November's issue of National Geographic has an illuminating article on Darwin. A sobering statistic is that in polls taken in 1982, 1993, 1997, and 1999, the creationist stance - that God alone and not evolution produced humans - had never drawn less than 44 percent of the US population. Only 37 percent of Americans were satisfied with allowing room for both God and Darwin (a view that is compatible with Roman Catholic dogma).

I think we have a public relations problem here. Perhaps, instead of battling directly with the creationist movement, we need to point out how useful the concept of evolution can be. We can show them how biologists use ideas from natural selection to develop the therapies that could save their lives. Maybe I'm treading on dangerous ground but maybe we should argue that evolution may not be in contradiction with the bible. Evolution is about dynamics and not necessarily initial conditions. Who knows, the world could have been set in motion 4000 years ago with the entire cosmological, fossil and historical record in place for us to discover. Creationists cannot argue that we do manipulate domestic animal breeds and that antibiotic resistance is a real thing implying that evolution is currently taking place. Surely, a beneficent God would have given us the means to help ourselves. Organizing life around evolution would be one of God's gifts.

The Great Dying

Throughout history, there have been many periods of mass extinctions. The most famous is the one that happened 65 million years ago between the Cretaceous and Tertiary epochs when the dinosaurs became extinct. The accepted reason is that an asteroid or comet about 10 km in diameter landed in the Yucatan Peninsula in Mexico. The ejected debris from the impact blocked the sun for several months cooling the earth sufficiently to kill of half of all life forms on earth. It also allowed mammals to rise up and eventually dominate the earth (i.e. us).

"The Great Dying" happened 250 million years ago between the Permian and Triassic epochs when ninety percent of all marine life and three quarters of all land life vanished. New evidence published in two articles in this week's Science, postulates that this extinction was not consistent with an impact. Ward et al. showed that the extinction happened gradually over 10 million years with a final push of 10,00 years. Grice et al. find evidence that the extinction was due to a combination of low oxygen and sulfide toxicity associated with a disruption of the carbon cycle and sulfur cycle. They show that the sulfide came from bacterial reduction of sulfate. The source of the sulfate is not known but may have come from volcanic eruptions.

In interviews, Ward has postulated that global warming due to massive eruptions of volcanoes filling the air with greenhouse gases leading to a runaway greenhouse effect where trapped methane in the oceans is released by the warming temperatures is the culprit. This is a plausible hypothesis but there is no solid evidence yet as far as I can see.

SNPs of Music

I've been attending a meeting on obesity and diabetes this week (which explains the paucity of entries). A whimsical thought came to my mind as I listened to one of the many talks related to gene expression - the number of unique human beings is finite. Our genome is about 3 billion base pairs long. However, individuals differ in something like 3 million of these bases in what are called Single Nucleotide Polymorphisms (SNPs). Thus there are at most 4 to the power of 3 million unique individuals! Since two of every three SNPs, involves replacing cytosine (C) with thymine (T), this number is reduced somewhat. This is a very big number but it is still finite.

A much smaller number is the number of different pop songs. In most cases, pop songs consist of no more than 4 or 5 melodic phrases of 5 to 10 notes. George Harrison was sued for plagarizing his song "My Sweet Lord" for repeating 8 notes of the song "He's So Fine". Usually a song is identified by its ``hook'', which is a single melodic phrase such as "We All Live in a Yellow Submarine" or "My Sweet Lord ". Almost all pop songs are based on a diatonic scale in a minor or major key which consists of 7 notes. Pop songs are almost always in 4/4 time. In order to be singable, notes cannot differ too much in time or range. So let's say each note can be either a whole note, half note, quarter note or eighth note and in one octave. That gives 28 possibilities per note. For a 10 note phrase this is 28^10 different hooks, not an astronomically large number. However, a hook is usually identified by fewer than 10 notes and the timing of the notes can vary by a certain degree and still seem the same to most people. The rules of harmony also restrict transitions to a smaller set of notes. So maybe there are just 10 possibilities per note and since absolute key and tempo doesn't matter the first note is fixed. For a five note hook, there may only be 10^4 different combinations. In other words, there are only about ten thousand different possible pop songs! No wonder they are starting to sound the same to me.