dlo

What would make this more useful?

"Oversimplified models of biochemistry" - when insulin binds to an insulin receptor on a fat cell, the fat cell activates LPL, translocates GLUT4 to the cell membrane, and inhibits HSL. What else am I missing?

"No measure of insulin" -- insulin is not as easy to measure using off-the-shelf devices. If one meal leads to higher blood glucose levels than another, especially if it is vastly higher, it is a reasonable working assumption that it elicits a greater insulin response.

"No control of meal time" -- both meals were dinners.

"No control of fasting time" -- I ate the same amount during the day prior to each meal. Activity levels were also similar.

Re: widely varying blood glucose levels, in my own experience, it is very rare to see a meal cause a blood glucose concentration reach 150 mg/dL. This is very, very high amongst my meals and only happens with meals like this.

The following meal would most certainly be nowhere near 150 mg/dL no matter the time of day: vegetables (kale, broccoli, cauliflower), black beans, sardines, and avocado.

Statement: “... slow or fast, your body still had to break down and consume that burger.”

Response: You are saying that the digestion rate of the burger doesn’t matter, because the entire burger is still processed by the body. But the way your body processes the burger is different depending on the digestion rate.

A couple of things can happen after a given amount of glucose enters the digestive system.

Consider when the digestive system releases the glucose into the blood rapidly, much more rapidly than cells are able to absorb it. The pancreas releases a lot of insulin. Muscle and fat cells both significantly increase their blood glucose absorption rate.

Alternatively, consider when the digestive system releases the glucose into the blood gradually. Cells are able to absorb more of the blood glucose as it is coming in, so blood glucose levels do not rise as much. There is some insulin signaling, but there is less of it. High-need muscle cells already have increased their blood glucose absorption rate due to increased energy requirements, while insulin signaling to fat cells is minimized.

In the alternate scenario, high-need cells muscle cells, while fat cell growth is limited.

Statement: "not eat the burger in the first place."

Cutting out something I like is not a sustainable way to eat healthily. Food intake is not just regulated by energy homeostasis; the reward system, e.g. the mesolimbic pathway, is also potent.

Statement: "improve your insulin resistance/glucose tolerance"

Agreed.

There are two ways to avoid excessive insulin signaling. One, you can decrease the rate at which glucose enters the blood. Two, you can increase the rate at which glucose leaves the blood. My article talks about one, and you seem to be an advocate for increasing insulin sensitivity, which is advocating for two. I think most people should do both.

This is a good idea. Thanks for the suggestion.

Re: your comment on HbA1c, I will just leave this paper here and comment later.

"A Simple Meal Plan of 'Eating Vegetable Before Carbohydrate' Was More Effective For Achieving Glycemic Control than an Exchange-based Meal Plan in Japanese Patients with Type 2 Diabetes"

http://search.informit.com.au/documentSummary;dn=12049472055...

Thanks for your comment, Aaron. I will address your points in order.

You start by questioning the background and motives of my colleagues. While I have thoughts on how to respond to this, I choose to stick to discussing the facts.

Then you represent the goal of my experiment in the following way: “can I stuff enough kale in my mouth to slow down the release of insulin, and therefore reduce glucose.” This is not accurate. My goal was to see if I could influence my digestive system to release glucose into the blood more gradually so as to minimize insulin signaling to fat cells. More explicitly, the amount of glucose is about the same in the two meals; it was the amount of insulin signaling by my body that changed.

I will add additional detail that I do not mention in my blog post. Cells that have high need for glucose do not need much insulin signaling. In response to their need, they already have placed more glucose transporters on their cell membrane. Gratuitous insulin signaling causes fat cells to absorb glucose that could have been absorbed by a high-need muscle cell.

Then you say, “Here’s a few other things that reducing insulin does....” From your wording, particularly in your use of “other,” you seem to agree that less insulin signaling means less fat cell growth and that your concern is primarily with its “side effects.” I will address amino acids and potassium in a later post. For now, I will address your point about brain signaling.

Actually, too much insulin signaling interferes with leptin signaling, making you respond less to leptin signaling. Leptin is the satiety hormone that is released by your fat cells. The neurons in your hypothalamus that have receptors for insulin also have receptors for leptin, and unfortunately for us, they have commonalities in their signal transduction pathways. So a lot of insulin signaling interferes with leptin signaling. This makes leptin signaling less effective and makes you want to eat more in spite of having enough body fat. (See the paper, “The First Law of Thermodynamics Revisited” by Robert Lustig of UCSF.)

You state firmly, “Fear not glucose and insulin. Fear insulin resistance and glucose intolerance.” You do not state however how you would advise someone to avoid insulin resistance and glucose intolerance. My view is that slowing down the digestion rate of your meals, such as I have done here, is part of the strategy for doing so.

Unfortunately, I have to run now, but I will address your remaining points later. I am also happy to write another blog post with ample references to primary research literature.

The experiment with sautéed kale and spinach is worth a try. If it is made too soft though, it may not slow down digestion much.

I hypothesize that a mix of coarse vegetables, for example, one that includes broccoli, kale, red cabbage, and cauliflower, would slow down the digestion rate even more than 10 oz of kale. Replacing the kale with dino kale would make this mix even better.

I didn't drink soda with the meal, simply because I don't normally drink soda with this meal, but an experiment with a caffeinated beverage would be easy enough to do.

One of the things we're working on is quantifying how different things, such as caffeine intake, affect your metabolic rate. This is the subject of an upcoming blog post.

Re: the experiment being bothersome... yes, it was!

I sampled my blood glucose every ten minutes, which might have been unnecessarily frequent. Having to interrupt myself every ten minutes to do a pinprick prevented me from doing anything that required concentration. I basically found myself watching YouTube videos for two hours.

I've heard of people doing continuous blood glucose monitoring, but this may be a bit heavyweight. I really only have to do this a few times for a few different meals to get an idea of what is needed by my body.

I just ate the kale raw from a box. The brand was Taylor Farms, i.e.

http://www.taylorfarms.com/products/organic-salads/baby-kale...

This is an unreasonable amount of kale :-). I'm going to need to figure out a better meal:kale ratio or use a different vegetable side.

Actually, you're right. They're minutes. I started sampling my blood glucose when I started eating. Let me go ahead and add annotations to some of those data points, especially the data points that correspond to the end of the meal.

This is a cool paper. I'm saddened that there is not much discussion on it. It may be too jargon-laced. The above WaPo article takes away too much, and in my opinion, makes the topic less interesting.

If there is interest, I can take a stab at writing an explanation of this result for the technical person, such as those here on HN, who is uninitiated to the field.

Great job! It seems really interesting. Re: work, I sent you an email.

This is terrific work.

On a related note, Atomwise has been doing this for over a year. And they've been able to make predictions that have been tested in the real world in areas such as multiple sclerosis, C. Difficile, and Ebola.

http://www.atomwise.com/

Full disclosure: They are my YC batch mates.

Thanks for the suggestion!

I would revisit this. Financial aid at Stanford is extremely generous... I was a beneficiary myself. Someone in that office will help you figure this out. I am still in the Bay Area, so if you need someone on the ground, let me know.

To be honest, your $20K loan sounds like an error. Alumni like myself do not donate back to Stanford so that things like this happen.

EDIT: Added the final paragraph.

I would say go back to Stanford if that is an option. You'll grow a lot there. I don't know what you mean by "change," but if you need pointers on how to handle the schoolwork or on how to take advantage of the resources you would have at your disposal, just let me know here and I'll get in touch.

Scott, you seem to have a surprisingly good handle on a fundamental problem with our healthcare system. Let's catch up offline :-).

- Your old co-worker

Are there papers in top-tier journals and conferences whose primary authors are still in high school? I would be curious to read these!

EDIT: spelling

I believe you are making the assumption that doctors are pocketing this money (or at least much of it). I am not saying that you are wrong, but I am saying that you need to identify this as an assumption and re-examine it.

(As it turns out, you are wrong, but I don't have time to go into it right now :-P.)

Does anyone have experience using Chronon?

http://chrononsystems.com/

It is billed as "DVR for Java."

It depends on what you're working on. If you're working at the cutting edge, you will find that Google often doesn't have the answers you need... simply because they aren't out there.

Wow, that's ambitious, but I never put it past a Ph.D. to do something like this. Let us know if you actually succeed!

It seems to me that there are a lot of efforts around making the front-end easier to understand. I think there needs to be at least one project that touches upon the back-end in a way that doesn't trivialize it.

There is not even one similar project for the back-end that I am aware of. And it's a shame, because I think the back-end is way cooler than the front-end.

She links to to an article that propagates said myth. Click on "arbitrary technical puzzles".

Great work!

Might I offer a suggestion? I would argue that most of the work that goes into a compiler is not the front-end functionality but rather the "middle-" and back-end functionality, i.e. making the generated code fast. It would be very insightful to those outside the compilers community as to what it's actually like to work on one to go over the basic optimizations, such as global data-flow analysis (reaching definitions, live variables, partial redundancy elimination, constant propagation, etc.), loop unrolling, and so on.

It's really cool how you're able to complete homework while you're watching video. Can you get into more detail about how you're able to do this?

My roadblocks to doing this would be that I wouldn't be able to pay attention to what's going on while thinking about a homework problem, that some things need to turn over in my head before I understand it at the level needed to apply it, and that solutions often take a long time to type up, especially proofs and programming assignments.

Haha, yes, I agree. 2:30 would probably be run at the pinnacle of one's amateur career -- a personal best rather than a typical time.

I do want to point out somewhat pedantically that it takes being in better-than-2:30 shape to run 2:30 on the Boston course, a notoriously slow course. There are runners who ran more slowly than 2:30 in Boston when they could have run 2:30 elsewhere.

2:30 is an eternity away from 2:03:38.

I am not thinking of dabblers; many people will do marathons after having barely trained. The goal for these people is just to finish.

I think 2:30 is achievable for a healthy, moderately talented man who is not merely dabbling: someone who is training correctly and has put in a few years of sustained, injury-free effort.

2:30 is still very much an amateur time. For example, I think that even though 2:15 or faster is professional-caliber it is still nowhere near being competitive.

Triathlon training really isn't just tacking together the training you would do for all three sports; this would be impossible. Instead, it is doing the same amount of training as you would do if you were doing one of the three sports exclusively. But in the same amount of training, you are training to perform three sports rather than one.

This necessarily means that a triathlete is not performing to his/her ability in any single sport. For example, a good time for the marathon portion of an Ironman is 2:40-2:50, but a competitive marathon time is 2:05 or faster. Just to add additional perspective, a good marathon time for an amateur runner is 2:30, faster than what a topnotch triathlete would be expected to pull off.

I wholeheartedly agree that exercise increases the ability to focus at work. I would even go as far to say that people who are not doing it are leaving some latent ability at the table. This is also true of sleep and nutrition, not just exercise.

I do want to remark briefly on what is admittedly a minor point that you made, namely that triathlons are highly recommended "... even if just doing sprint/olympic/half distance." It is a common misconception that the longer the distance, the more difficult or meritorious it is to complete. However, this is not true: It is not the case that one races an Ironman at the same intensity that s/he would for, say, a sprint-distance triathlon; rather, in order to last the entire duration of the longer distance, s/he is racing less intensely. (A person who performs a sprint-distance triathlon at the intensity s/he can hold for an entire Ironman is not racing the distance.)

Let's go to running for a concrete example. A topnotch overall pace in the marathon is 4:50/mile. But a miler will run a single mile in 3:50, a whole minute faster than marathon pace. (Note that these are typical competitive times rather than world-record times.)

With all this said, I would be remiss not to say that I agree with the overall sentiment that one should take baby steps rather than brashly pursuing unreasonable goals.

In response to your final question, I haven't done triathlons since college. At the moment, I am working on getting into shape to finally break 5:00 for the mile run. Next up is breaking 1:00 in the 100-meter swim. I'm still figuring out what a reasonable goal is for cycling... maybe 5 miles in 10 min.?

After I have all the pieces, I'll put them together and start doing triathlons again :-).

Indeed. But the converse is also untrue: Not everything that people post on the Internet has a basis in reality. I know many people who went to other universities, and many do not appear to have a similar concept: For example, there are probably few superficially serene ducks at MIT, because many of them seem to wear their hard work on their sleeves; and students at Cal have nicknamed their school "Berzerkeley".

I am not saying that this term or even this phenomenon is strictly unique to Stanford, but I simply have never observed "duck syndrome" used in this sense in any other context. This includes the Internet, personal conversations, and so on. And now, it includes this Hacker News thread, because aside from Drbble's vaguely plausible remark no one has yet come forth that their institution employed this term as well.

My take on it is that, just as everyone has their own mascot, everyone has their own vocabulary, which probably includes their own coinage for this experience if they indeed have this experience to describe.

Out of curiosity, I typed "duck syndrome" into Google. The only relevant links on the first page pertain to "duck syndrome" at Stanford. Maybe the term is beginning to catch on elsewhere?