Saturday, April 30, 2016

Experimental Diet Day 6: Light at the End of the Tunnel

This morning I woke up at 1 am and read 94 mg/dl. That was, again, surprisingly high considering I hadn't eaten in six hours and that my daytime low bottoms out in the 70s. It suggests that my sugar never goes below about 90 when I'm sleeping. I guess my liver is not just an early riser, but positively nocturnal.

My regular fasting measurement (which was at 9:00 this morning because it's the weekend and that's when I get up) was 96, which is 1 lower than yesterday. Taken as a whole, the last five days' fasting values support a slow downward trend, which damn well ought to be happening considering I haven't eaten so much as a gram of refined sugar in 7 days.

The plot below shows that my sugars are higher when I'm sleeping than when I'm awake. That's either atypical, or medicine has the wrong definition of typical.


Blood glucose over the past two days or so,with waking and sleeping periods marked

I only have two more questions I want to answer. (1) If I eat my regular breakfast of plain shredded wheat and milk at noon instead of in the morning, would I get the same 50-some point blood sugar spike that I got in the morning? (I'd be starting from about 70 instead of the high 90s, so would I end up at 120 or would I shoot all the way up to 150 like I did in the morning? Or would something different happen?, and (2) If I eat a morning meal consisting of 40 grams of sugar in half a cup of milk, would I get the same blood sugar spike that I got when I ate 40 grams of plain shredded wheat in half a cup of milk? In other words, do the foul-tasting carbs in the shredded wheat get rapidly and completely digested into sugar, like my Monday spike suggests? Because if they do, then from a sugar point of view there's no benefit to eating plain shredded wheat versus frosted mini-wheats. That would be good news if true.

I'll do question (1) tomorrow and question (2) on Monday, and that will end the experiment. As a final note, I wasn't trying to lose weight with this weird diet, so I didn't measure it carefully, but I stepped on the scale this morning and it looks like I lost about 5 pounds in 7 days. Interesting but probably not sustainable for the same reason the first week's weight loss rate on any diet is unsustainable.

Friday, April 29, 2016

Experimental Diet: Day 5

Yesterday I found that my "fasting" glucose was actually higher than it was the rest of the day, except right after meals. So this morning I woke up at 3 a.m. to try to catch the absolute low point, before my liver started putting out glucose in preparation for my normal waking time.

I got 110 mg/dl, even higher than my previous morning's fasting value!

This experiment has been full of surprises, which is why we measure instead of guessing. Apparently my liver is a really, really early riser. So tonight I'm going to wake up and measure at 1 a.m.

My fasting glucose this morning was 98, not significantly different from the last couple of days. I didn't take any other readings until 5:30, when I got 75. I was definitely feeling loopy so I drank a little orange juice right after. 70 is considered hypoglycemic and is where diabetics are counseled to eat something quick so as not to lose consciousness.

The interesting thing about the later measurement is that I didn't eat lunch until 2:30 because I worked through lunch and left at 2:00 to take my boys go-karting (they're on spring break this week.) So the 5:30 measurement was only 3 hours after a rather substantial (but still low-carb) lunch of three slices of leftover cranberry-apple stuffed pork roast. I even got a little sugar from the stuffing. But I was back down to 75 in three hours. The hormonal regulatory system really wants to maintain the glucose level according to the time of day and fights any dose of carbs from a meal.

So far, here are my conclusions:

1. The morning fasting value is probably the best one for measuring long-term trends in an individual, because it occurs under the most uniform conditions. But I bet it varies a lot from person to person even after identical fasting times. They should probably not publish a normal range for that value because someone could easily have a late-rising liver and get consistently "normal" values, but then it skyrockets after they eat breakfast and the liver finally gets going.

2. Even a drastic change in diet produces slow changes in blood glucose. My fasting value averaged 108 over three days of my normal diet but when my diet dropped from ~100 grams/day of sugar to near zero, it's averaged 97, only a 10% drop.

3. The fasting value should absolutely not be considered a daily minimum. My fasting value of yesterday was 97 but my daily average during waking hours was only 91, and the lowest I measured was only 70!

4. I tend to get tired after lunch and this could be due to low blood sugar. So I need to consider eating a smaller or later breakfast and a bigger lunch in order to smooth things out. I've never been a big breakfast eater and it evidently has to do with not needing food because of the big glucose release I get in the morning.

People who study this for a living probably know all this stuff already, but it's more fun to find it out directly, by measuring. The diet will go on to Day 6...

Thursday, April 28, 2016

Experimental Diet Day 4: Big Data

Today I took a blood sugar reading when I woke up, and every hour afterward. (Well, almost every hour. I was at work and things got in the way.) Fasting (morning) sugar was 97 mg/dl, trending down but still significantly higher than the Day 1 teaser of 75.



8-hour fasting values (taken first thing in the morning before eating)


On the next reading, an hour after breakfast, I had a big spike to 149, which is certainly in the normal range but higher than I expected. One thing I didn't mention about this diet is that I still eat cereal for breakfast because otherwise my bowels would grind to a halt. But instead of my usual Frosted Mini-Wheats, I've switched to plain shredded wheat. It is 40 grams of carbs. The nutrition label shows zero sugar and I can believe it; the stuff tastes like cardboard. There's a little sugar naturally present in the milk I had to add to choke down the shredded wheat.

The train arrived on schedule this morning (refer to previous post for meaning).

The after-breakfast spike decayed steadily to about 80 within 3.5 hours, which is interesting because it's significantly lower than my "fasting" value of 97. After lunch, I got hardly any bounce at all - it peaked at 86 after two hours and then decayed to 70 over the next three hours. There were only 7 grams of carbs and 3 of sugar in my lunch (roast beef and green beans), so I guess it makes sense that there wouldn't be a big spike like with breakfast.

When I hit 70 at about 5 pm, I felt a little loopy and had to eat an apple because I needed to drive home from work without passing out. 70 is 27 points below my "fasting" value of just this morning! That was certainly unexpected. One explanation for the high "fasting" value is that your liver wakes up before you do, and starts pumping out stored glucose to get you ready to wake up. So tonight I'm going to wake up and take a reading at 3 a.m. because that should be the absolute low point, before the liver has started to rev up.


Hourly blood glucose readings with meals labeled at time and by total carbohydrate content

From this I conclude that my blood sugar depends on more than just carbohydrate intake. Something else is regulating the level, and it takes a big dose of carbs (like my 46 gram cereal breakfast) to overwhelm it. And even then, the effect is short-lived. Probably the most interesting thing is that the 47 g of carbs (including some sugar from the apple) had a much smaller effect in the late afternoon than did the cereal in the morning. 46 g of carbs in the morning gave me a 54-point jump, but 47 g of carbs including the apple sugar only gave me an 23-point jump in the late afternoon. Probably the cereal was just adding to the liver's generation of glucose from glycogen that was stored the day before. But those glycogen stores have got to be getting pretty low at this point. The readings at 3 am and 7:30 am tomorrow should be interesting.

Wednesday, April 27, 2016

Experimental Diet Day 3

Unfortunately, yesterday's glucose bounce-back was not a fluke. I measured 101 this morning. I didn't take any omeprazole yesterday morning so that probably isn't the cause, although I'm still curious about its effects.

One hypothesis is that the liver and possibly some other organs can make glucose out of non-carbohydrates - it's called gluconeogenesis. Once you stop taking in sugar, gluconeogenesis kicks in, but according to my wife the doctor it might not kick in right away. So the 75 reading after day 1 might have been before my gluconeogenesis got going. [EDIT: A different mechanism - breakdown of stored glycogen - happens in the liver far before gluconeogenesis. It would be this glycogen breakdown that would be responsible for short-term, overnight glucose production.]

But gluconeogenesis is a costly process and if you continue not to eat carbs, it eventually runs out of steam. Then your body gets energy from ketone bodies instead of glucose. That's the point when your blood glucose is supposed to drop and stay down. So it may take a few more days.

Another possible explanation is that I did eat a very small amount of carbs the night of Days 2 and 3. It wasn't much - three or four tortilla chips - but it could have thrown things off because it was so late.

I also learned that the experimental diet I'm eating is well known to medicine and is called the ketogenic diet. It's a treatment for epilepsy, but nobody really knows why it works.

The experimental diet is harder to stick to than I thought it would be. Slim Jims are kind of a guilty pleasure and I thought I'd enjoy them as a regular snack for a while, but I really don't want anything to do with them. I am definitely craving sugar. A guy retired from our lab today and they brought in one of those big, fancy sheet cakes with about an inch of frosting on it. I almost didn't go into the room, and when the tributes were over and the cake was cut, I had to leave.

The ethics of this diet are interesting.  I stated before that I was a kinda-sorta vegetarian for a few years because I wasn't convinced that it is OK to kill an animal for food when there is other food available. And I really didn't like factory farms - I still buy free-range chicken and grass-fed beef when I can get it. But what if not eating meat is harming my health or even shortening my life? If so, there's a tradeoff that must be made, that I never considered before.

I was also curious as to whether I'd feel any different on this diet. Objectively, two things happened (warning: bodily functions.) One is that the very regular schedule by which I, as Borat would put it, make dirt, is altered. You know when you're waiting for a train that is usually on time, but for some reason it skips a couple of stops? That's what is happening. The other thing is that when I pee, on my normal diet it foams up a lot. This is supposed to be due to proteins in the urine. It foams up markedly less now. Not sure what to make of that.

Journalist Borat Sagdiyev

I got a new package of test strips and lancets so I am ready for the hourly testing tomorrow. Tonight I will eat nothing after dinner so as not to interfere with my "fasting" glucose test tomorrow morning.

Tuesday, April 26, 2016

I'm Doing a Medical Experiment on Myself

I read the book Good Calories, Bad Calories by Gary Taubes. It's a review of nutritional science by a smart and careful nonscientist that says the conventional dietary advice of the last 40 years is totally wrong. The gist is that the human body is not evolved to handle refined sugar and starch, and if you eat a lot of it your metabolism goes haywire and starts laying down fat, and causing additional problems unrelated to fatness. A corollary is that saturated fat and dietary cholesterol aren't bad for you. The obvious conclusion is that reduced-calorie, high-carb, low-fat diets are not only unpleasant, they don't work. If you want to limit something, it should be sugar and carbs, and even people in studies who were allowed to eat all the meat they wanted (but nothing else) lost weight.



It was a pretty compelling read. I was an on-and-off vegetarian for several years on ethical grounds, but I was assuming (based on all dietary advice except the Atkins Diet) that vegetarianism was at least harmless and possibly much healthier. Well, I've had to put that assumption in the trash can. Not that Taubes's book is perfect. I didn't think he addressed the question of why the human body evolved to be not just indifferent to, but intolerant of, simple carbs, nor did he address evidence against, like certain tribes that live on carbs but don't get fat. (Authors have to be very careful about touching that sort of thing because it implies a group can genetically adapt to the available diet in only a few thousand years, which implies that race has biological significance...) But taken as a whole, sugar seems to be really bad for you, and if you cut down on sugar or carbs, you pretty much have to eat more meat. This was actually standard advice up to about 1970, as funny as that sounds today.

Eating too many carbs screws up the natural insulin response to elevated blood sugar. I got curious as to how responsive my blood sugar is to my diet. So I got a blood sugar meter and took three days of baseline readings on my normal diet. These are taken first thing in the morning on an empty stomach. The meter was cheap and easy to use, and the lancets are so tiny you hardly feel them. Here's what I got:

100, 113, 110 mg/dl

Anything over 99 is considered pre-diabetic! (That is, it indicates you might be becoming insulin-resistant, which is Type 2 diabetes. More than 126 is diabetic.) Now, I did eat about three pints of ice cream over those three days, which I don't do every weekend, but I probably do it once a month. We scientists must be brutally honest. I'm maybe 10 pounds overweight but generally healthy.

Then I started an experimental diet cutting out sugar and carbs as much as I could. I went to the grocery store and bought a bunch of meat, cheese and eggs. I even bought Slim Jims to snack on instead of chips or pretzels. I got only green leafy vegetables, not potatoes or anything starchy like that. This would not be sustainable as a permanent diet - it's just a temporary experiment to see what would happen.

On this experimental diet my sugar consumption dropped from about 150 grams/day to near zero, and my total carbs decreased from 400 grams/day to about 100.

After one day of this sugar-free diet, my glucose dropped to 75 mg/dl!

I got all excited and figured that with two more days of low readings, I could prove that my glucose is strongly and rapidly influenced by my diet, and call off the experiment to return to a more moderate (but still reduced-sugar) diet. Well...

On Day 2, my glucose bounced back to 115 mg/dl. This was a surprise. I can only conjecture that on the morning of Day 1, I took an omeprazole tablet (that I use for longstanding acid reflux) and it delayed my digestion such that I didn't get a fasting glucose value on Day 2 but something more like a post-prandial value. (Post-prandial is the term doctors use for "after eating" when they don't want you to know what they're talking about.) I also got a mean headache on the evening of Day 2, but it went away and I started feeling pretty good.

So it's going to take several more days to see what the real trend is. The human body is a strange machine and nothing is clean and simple. My plan is to take a reading every hour some day, to see how it goes up and down. I'd like to do this on one day when I take the omeprazole and on one day when I don't, for comparison. But I only have 5 more lancets left and I'll have to find a place where I can buy more. Stay tuned!