Monday, May 9, 2016

Experimental Diet: Postscript

Three more things I have to mention about my recently concluded experimental extreme low-carb diet.

1. I speculated that the weirdly low reading of 75 on the first morning after I started the diet might have been due to having taken an omeprazole pill the day before. So I took one again and measured my sugar the next morning and it was around 100, the same as during the rest of the diet. So the omeprazole had nothing to do with it.

2. Originally I sort of wanted to get into ketosis, i.e. using ketone bodies for energy instead of glucose. I did notice some changes like headaches, a funny taste in my mouth and changes to my urine, that are said to be associated with ketosis, but none of these things are objective.

3. I briefly mentioned that some populations seem to be able to live on a high-carb diet without getting fat and I speculated that there could be genetic differences at work. For example, the Japanese can afford pretty much all the rice they want these days, but still they don't get fat. Well, today I read this preprint which lists the top 16 traits found in English populations that have been selected for over the past 2000 years. Four of them have to do with glucose metabolism.  The authors don't interpret exactly what is going on, and I am no biologist, but I can see that increases in the fasting insulin level, which is what helps the body tolerate big spikes in blood glucose, was the fourth-strongest trait.   This is a population evolving to deal with a diet high in refined carbohydrates. 

Evolution may be good for the species but it is hell on individuals. For the insulin trait to be selected, someone with poor glucose tolerance has to die before reproducing, or be unable to reproduce. Wouldn't it be easier to just stop eating so much sugar?

Wednesday, May 4, 2016

Experimental Diet: The End

I had to work late the last two days, so I didn't have time to write a post. But I finished the diet two days ago.

The one remaining question was whether 45 grams of sugar would give me a bigger blood glucose spike than 45 grams of whole wheat carbohydrates. These would both have to be my first meal of the day so as to have the same baseline.

If you look at my previous posts, I get about a 50-point spike in blood sugar an hour after my experimental diet breakfast of plain shredded wheat in milk (45 g complex carbs). On Monday, I just drank two big glasses of orange juice (45 g sugar) for breakfast.

After the juice I got the same 50-point spike, from 93 to 142 (close enough). That means as far as blood sugar is concerned, whole wheat complex carbs are the same as sugar, gram for gram. My guess is that the only reason whole wheat is "better" is that you're inclined to eat a lot less of it than if you ate something sugary.

I did one other measurement of interest. On Monday I experienced my usual drowsiness after lunch. Is it due to low blood sugar? Nope. I measured 124 while I was really drowsy. Maybe the drowsiness is because a lot of blood goes to my digestive tract after lunch; I really don't know. The feeling is similar to the "loopy" feelings I reported during my late-afternoon lows, but those are more likely to be because of blood sugar.

Wrapping up, the experimental diet reduced my waking blood glucose from 108 (averaged over three days) to 96 (averaged over nine days), more than a 10% drop. And there is a clear decreasing trend, so I could have taken it lower if I'd stuck to the diet.

I have three final plots. The first one is my daily waking values, which as we now know are nowhere near my daily lows. You can see the downward trend after I started the experimental diet as well as the oddball low value on the first day of the diet.

The next plot is all the measurements on a continuous time scale. I connected the points with a line if I didn't eat in between measurement; otherwise I just left them as isolated points. You can see the three days I measured after-breakfast spikes, as well as my afternoon lows in the 70s. Waking time is shaded yellow in this plot.

Finally, I plotted all the data versus time since waking. You can see the big breakfast spike, the smaller lunch spike, the smaller-still dinner spike and the fact that the overnight values are higher than the daily lows. Again, waking time is shaded yellow.

Will I change my diet based on this experiment? Yeah, I think I will. I am going to cut down on sugary snacks and desserts, cut down on starches (pasta, noodles and bread), and eat more protein, mainly ethically sourced meat. I'm also going to stop worrying about butterfat. My kids like whole milk, so that's what they're going to get. And I like ice cream, so I'll keep eating that (in moderation) and not mess around with "light" ice cream or frozen yogurt. In fact, regular ice cream should be better because the fat lowers the glycemic index compared to fake ice cream.

Sunday, May 1, 2016

Experimental Diet Days 7/8

The first key remaining question in this experiment is: would I get the same 50-some point spike in blood sugar if I ate shredded wheat at noon instead of at 7:30 am? Data show some obvious glucose release by the liver during the night which puts my level typically in the high 90s on waking, so a 50-point spike at breakfast means I hit almost 150. But my waking value is more than 20 points above my low for the day which hovers in the mid-70s.

If I got less than a 50-point spike, then it could be that the liver was shutting down any remaining glucose output so as to blunt the effect of a midday intake of carbs. If I got more than a 50-point spike, then it could be that the liver becomes more active in order to push my peak sugar to some set level - the total spike would have to be about 75 points to hit my morning peak of 150.  Or maybe the liver can release other substances into the bloodstream or the digestive tract to control how much carbohydrate is broken down into glucose and absorbed into the bloodstream. Carbs are digested into glucose in the small intestine (the more refined, the higher in the intestine); they get absorbed as sugars.
The answer turns out to be that I got the very same 50-point spike from the cereal at noon as at 7:30 am.  I measured 74 at noon, just before eating. Then I ate the cereal, waited until 1 pm, and measured 124. There's your 50 points.

This suggests that the liver releases glucose from its stored glycogen on its own schedule, regardless of food intake.  I'm assuming here that glycogen breakdown is the mechanism active overnight and for a couple hours after I wake up. The liver has another mechanism, gluconeogenesis, in which glucose is made from non-carbs, but it's my understanding that doesn't become active until your glycogen stores are gone. Despite my zero-sugar experimental diet, I think I'm still taking in enough total carbs to keep my glycogen stores nonzero.