Friday, November 15, 2013
I'm Alive, But Too Busy To Post
What happened is that I have been working on the precursors to diabetes for a couple of years (medically). To some extent I was doing this with SuperDoc, to some extent with PNat. PNat just does the helpful analysis that hints at where to look.
Anyway, I had gotten to the point where I could consistently roll myself in and out of diabetes (I wasn't going to experiment on other people this way), but it had nothing to do with all the things you read about in the standard medical literature. Doc concluded at one point that I must have some weird sort of cancer, and I pointed out to him that I was doing this more than 20 years ago, and no way could I be living if that were true.
So we met at the pass and then doc was so fascinated about how I had diabetes but didn't that I think he concluded I must be a super genius at managing blood sugars (he should have known better, but I guess the truth was too weird). Anyway, I told him that I could do this with other people. I knew this because I HAVE been doing it with other people rather consistently.
So he started having me work with newly diagnosed severe diabetics, and sure enough, I can roll them back to an extent. They still have something wrong with them, but they are technically not diabetics. Also their bloodwork turns miraculously beautiful and their organ function jumps way up. Including pancreatic function.
So then I figured out how to all the way reverse my condition. So then doc had me work with this guy in heart failure, and lalala, it appears that he is abruptly coming out of cardiomyopathy. So now we are trying to do the full rollback with patients to see if we can.
This is extremely complicated both procedurally and medically. It appears to be an approach that hasn't been taken before, and according to PNat, this is also implicated in the development of several other odd diseases, especially auto-immune.
So SuperDoc is running in aphasic excited circles, both of us are staying up all night reading, and if we hadn't made an agreement on Wednesday that we needed to sleep I wouldn't even be posting this.
At the very minimum it appears we may have some protocols for metabolic disease that will be hugely helpful to many patients.
So to cut a long story rather short, SuperDoc says we are setting up a clinic just to do this and research, so we can develop the whole thing. I have agreed to be part of it, because the protocol is individual for each patient.
This has to do with the ATP cycle. But then, what doesn't in a living organism?
If you haven't already, check out the resistant starch threads over at FreeTheAnimal. And if you haven't heard of Peter at Hyperlipid, he has dreams about mitochondria and ATP.
Cheers & good luck!
Resistant starch, high fat intake and the reformers in the paleosphere are more about moderating one's carb intake such that the pancreas isn't called upon for any heavy lifting in the first place.
IOW, if you don't need insulin, then it doesn't matter so much that you can't make insulin. An asymptotic diabetic if you will.
That sounds amazing. Hoping to hear more about it sometime.
I've often noted the similarities between medicine, economics, and the engineering of complex electro-mechanical systems. It's interesting to see you put that sort of expertise to such good use.
This wouldn't include Crohn's disease by any chance would it?
The commonality is small vessel dysfunction, by which I mean to include the capillaries, which leads to low circulatory function, which can result in chronic infections and inflammation. This can result in auto-immune disease or something that seems functionally like it.
I do know of something, maybe two somethings, that might be of help, but I have seen more people injure themselves through random use of such things than I can count.
If you or a close one are fighting with that, email me at maxedoutmama2 at gmail.com, and I will try to give you some things that you can take to a doctor which may be of use, but I don't want to randomly kill people with hope.
Two weeks ago he was being treated with high doses of insulin in the hospital with blood sugars above 300.
We've already taken him off all diabetes meds except metformin. Four weeks ago SuperDoc told me that this guy's only hope might be a heart transplant but he didn't think he could qualify, and he asked me to do my thing with the guy in the hopes of just stabilizing him long enough and building a compliance record to get him onto a transplant program waiting list.
Our clinical successes have been mounting, but this one I think is the one that caused SuperDoc to run in aphasic circles.
It really is a mathematical analysis. I worked out a method for decomposing the blood sugar curve and then making very small tests to see what was controlling inflection points. It's almost exactly the methodology I use for economic and interest rate analysis, which is also the type of analysis you would use to analyze a very complex circuit with multiple functions that kick in and out with voltage changes.
The problem is that it is extremely time-intensive - it takes me at least 20 to 30 hours on one patient, and with this guy I have already spent 45 and will get to a hundred or so.
And SuperDoc and I are just learning to talk to each other. I come to him with stuff like "okay, there's a vector pushing this here up at this point". You can imagine the blank stares. I knew he could solve it if I gave him any meaningful info, so I basically used PNat to do an associational analysis of a mass of medical literature in the same way I used to use PNat to analyze the county samples for controlling vectors.
As a result of that analysis, I came up with one possibility that then cured me of what is basically relapsing remitting multiple sclerosis without the auto-immune positives, but then we used the same thing to correct reactive hypoglycemia, which is the primary puzzle I've been working on for two years.
So all of a sudden we were in new territory. Well, we are not going to get the doctors to go to engineering school, and we are not going to get the engineers to go to medical school, so I think that in the end this will have to be an AI program.
SuperDoc is an incredibly careful and detailed clinician. I wouldn't be comfortable doing this otherwise.
May the wind be at your back for this new journey.
But your real threshold of fame will be a link by Glen Reynolds, the BlogFather.
Seriously, the best of luck. If you can get people out from under the ACA, it will be a great service.
I have a strikingly bad attitude. We move from one debacle to another.
WSJ - I am most certainly analyzing the feedback behaviors, but not so that I can control it, but so that I can shift the system to better equilibriums. Actually I have been doing this for diabetics for more than ten years. It works startlingly well. Mostly I end up removing a lot of medication, but the real goal is to get a flat curve, and then slowly move that down.
Now I am embarrassed to admit that I never knew that doctors didn't know how to do this. I just figured they mostly didn't bother, but according to SuperDoc, this is indeed something new and different.
However the implications of the way that I can move people's curves down are that some of the fundamental theories about diabetes are flat wrong.
About half of the people I end up doing this with wind up not having H1ACs at diabetic levels and not needing medication within a year, so an apparent healing is occurring.
As far as I can tell, what we are really doing is changing electrical impedances across either capillary membranes or cell membranes.
The time frame involved suggests that nerves are healing and that this is to some extent a signalling problem.
But it would be different for every patient depending on a number of factors, and I think we need to try a hundred to get a firmer grasp. There should be failures, and those would tell us more than the successes at this point.
Congratulations on the interesting work! If it saves lives, it's definitely worthwhile.
Is he on a beta blocker? That can cause abnormal blood sugar curves in many people. I'm not saying to come off it, but when you use those medications you need to check blood sugar curves, especially if the person seems to be reacting to more sugar in the diet.
Have you tried that? Get one of those cheap drugstore monitors and a package of 25 strips and test before he eats dinner, then every half hour after until the blood sugar has been dropping for two measurements. Do that for a couple of different types of meals.
One of the first symptoms of abnormal blood sugar metabolism are shifts in curves.
The Free The Animal blog is a bit wide ranging. Fortunately, the author is good about using tags and keeping the posts themselves on a topic. This should be helpful in cutting through what might be otherwise uninteresting clutter: Resistant Starch
You mention MS, I'm guessing you've heard of Terry Wahls? (This is... interesting... upon checking the spelling of her name with 'teh goog,' I see "quack" comes up as the third auto-complete. Clearly she's onto something important. As Gandhi might note she's made it past "ignore" and is now in the "laugh" stage. :) )
Last as an engineer that did want to be an MD, and knowing a few engineers that did go on to become MD's, it's not that engineers don't want to be doctors, it's that MD's don't want engineers. The MD Guild is set up quite nicely as it is, thank you very much, and they don't see any reason to change it. My riff on (vast majority, albeit not all) doctors is: Doctors are not scientists, they aren't even engineers. They are mechanics, and not very good ones.
You have piqued my interest. Do you have any links or sources where I could read related material?
I have a sister with a PhD working at Harvard Med (she is looking to get out, she is not a physician) in continuing education and would be fascinated by what you and SuperDoc are doing. Any links or sources would be even more useful for her.
Sounds like you're working to get O3:O6 ratios back in line? That's another big one among the Paleosphere. The typical diet these days can have an O3:O6 ratio of 1:20 all the way to 1:50. If memory serves, traditional diets would have those around 1:8. If you're trying to repair a lifetime of poor eating, I've seen some people suggest supplementing with fish oil to get it to 2:1 the other way.
For otherwise healthy folk, the recommendation is don't overdo the O3 (I'd seen some people reporting taking 1 to as much as 3 TBSP daily), but moreso watch the O6 intake.
As for me and mine, we stay away as much as possible from the S's: Soy, Safflower, and Sunflower, and take about 500mg daily or so of high-EPA fish oil formulation. The EPA seems to really help my son's (and my wife's back when she was preg & then nursing) mood. DHA might be better, but there are suspicions that it oxidizes too readily, causing anything you can actually buy to be ineffective. We're young and healthy, so as they say, YMMV.
For folks reading along and curious, I think Robb Wolf & Mark Sisson might be good guys to get started with. I was most partial to Kurt Harris, but he hung up his keyboard a while back. The archives are still there and worth reading. A lot of the advice hasn't changed, just percolated out to a broader public.
I lost a lot of weight, I started walking briskly 45 minutes a day, I adjusted my carbohydrate intake to 40% of total calories, I started paying attention to glycemic index, and I started taking a modest dose of metform (1 gram per day).
A1C presently 5.5%.
I'm not sure which of those things makes a difference. Would prefer not to experiment on myself. I wonder, though, fi anything I'm doing correlates with anything you do.
Heck, it's probably time to hang up the economic spurs anyway. If the Census Bureau is faking the labor statistics for political gain....well, GIGO.
The rapidity of your results makes me think this is more than just a diet. If you can do this with hypoglycemia, I might know someone(s) who would be interested.
The only way I know of to do the kind of analysis you describe is to set up N-dimensional ODE's and run them with a variable-step numerical solver, then check the results. Runge-Kutta and friends.
IMHO, AI is over-rated for parameter identification. I see people throw AI at a problem all the time, to uncertain results. I've made a fair amount of money cleaning up these sorts of "silver bullet" solutions. It does sort of work if you use AI to help discover the parameters of arbitrary ODE's, but it's no substitute for having known parameters with physical descriptions. AI tends to find local minima that don't pass the giggle test.
Perhaps you've got too many parameters to be able to quickly identify them directly through phsyical tests. But it's not necessary to actually UNDERSTAND calculus in order to USE calculus...
Sometimes the reactive hypoglycemia is a very easy fix. Send me an email.
The program would be to guide the person through the process that I do, because no doctor is going to have the time. Then it would tell the patient how to test dietary changes, and have them enter the results. Then it would spit out suggestions for a medication/balancing regime. Doctors would have to do the medication/risk balancing part.
My next challenge is a woman who's on over 200 units of insulin a day, plus other medication.
I'll see if I can check his blood sugar. Right now, he's worried about his blood pressure, although I tell him it is a good thing that he doesn't need the meds any more.
Basically true, I think...you do have to have a conceptual-level understanding (what is a derivative, what is an integral), but you don't have to understand how to analytically derive the solutions to equations (and most of the ones that matter in practice aren't analytically-solvable, anyhow.)
I've seen US Navy training manuals from WWII and shortly thereafter in which they did a great job of explaining to the target audience (most of whom were probably only high-school graduates, if that) how the mechanical fire-control computers of the time worked: this required one to grasp the concepts of integration and differentiation.