Missed Distinctions

post by jimrandomh · 2009-04-11T03:15:40.595Z · LW · GW · Legacy · 13 comments

Contents

13 comments

When we lump unlike things together, it confuses us and opens holes in our theories. I'm not normally one to read about diets, dieting advice, or anything of that sort, but in today's article about the Shangri-La Diet, I saw an important distinction that no one's talked about. Something Eliezer said in the comments struck me as odd:

a skipped meal you wouldn't notice would have me dizzy when I stand up

And a few posts later,

I can starve or think, not both at the same time.

Reading these, I thought, that's not what being hungry like feels like for me. But while being hungry doesn't feel like that, those descriptions were nonetheless familiar. And then it hit me.

He wasn't describing the symptoms of hunger. He was describing the symptoms of hypoglycemia, more commonly known as low blood sugar. Blood sugar is one of the main systems responsible for regulating appetite, so for most people, having low blood sugar and being hungry are one and the same. The main focus of the Atkins diet, for example, is reducing swings in blood sugar, thereby reducing appetite. The Shangri-La diet seems like it would have a similar effect.

Being diabetic (the kind caused by immunology, not the kind caused by diet), I monitor and control my blood sugar, so I have ample opportunities to observe how it affects my eating habits and how I feel. Like most insulin-dependent diabetics, I have been trained with a fairly detailed model of blood sugar, how it's affected by food and insulin, and procedures to follow if it's too high or low. The standard procedure for low blood sugar, taught to all diabetics, is to test blood sugar, eat exactly 15g (60 calories) of sugar, wait 15 minutes, then test again. In practice, I have sometimes responded to hypoglycemia, not with 15g of sugar as the procedure specifies, but with 15g of sugar, immediately followed by a thousand plus calories of binge eating - basically, as much food as I could shove down in the time between when I first started eating, and when my blood sugar returned to normal (about ten minutes). This behavior is common among people on diets stricter than they can handle. For me, someone not on a diet, with a mostly full stomach, it's downright odd. Or is it?

Being hungry is not the same as having low blood sugar. Hypoglycaemia feels like extreme hunger (plus a few other symptoms), but while extreme hunger takes a lot of food to get rid of, it only takes 60 calories and 15 minutes to completely eliminate hypoglycaemia. If you're hungry, you ought to suppress it. If you're hypoglycaemic, on the other hand, you need to deal with it swiftly, and in a controlled manner. What happens if you don't? As a diabetic, this, too, is in my training. The pancreas will release glucagon, a hormone that causes the liver to release stored sugar into the bloodstream. Getting rid of stored energy is good for a dieter, right? Well, in this case, no it isn't; the sugar stored in the liver would have been released the next time you exercised. Rather than burning fat, you're burning short-term energy reserves, so that when you do make it to the gym, you'll hit a wall more quickly. And of course, while your blood sugar is low and you aren't eating, you can't focus and you quickly burn through willpower.

Today's best diets prevent low blood sugar entirely, rendering the hunger vs. hypoglycaemia distinction moot. However, if you can't tell the difference between hunger and hypoglycaemia, then you can't tell whether it's your diet failing or your willpower. Blood sugar test kits are affordable and don't require a prescription, and once you know what low blood sugar feels like, you won't need the kit anymore. There is much more to dieting than just controlling blood sugar, of course, but we do know that blood sugar is important. So why has no one proposed the Prick Your Finger diet? Why do none of the popular diets involve measuring blood sugar at all, ever?

Mild hypoglycaemia feels like a caffeine overdose without the energy: irritability, palpitations, and tingling in the extremities. It is a distinctly alien feeling, and includes an urgent desire for food. Only sugar can eliminate it; fat, protein and complex carbohydrates will not help at all, and should be avoided. Severe hypoglycaemia produces other symptoms, but can only be produced using medication.

13 comments

Comments sorted by top scores.

comment by Chase_Johnson · 2009-04-12T22:26:56.733Z · LW(p) · GW(p)

Been lurking since LW launched. First time posting. Hi guys.

This article is interesting and relevant to me, because I regularly experience distinct de-motivation and lack of willpower, which I believe is the result of blood sugar regulation issues. I wonder how much variation in motivation and productivity in life is caused by inadequate handling of one's personal biochemistry. I know I can be productive as hell one day, and completely unable to focus the next.

comment by PhilGoetz · 2009-04-11T04:19:03.594Z · LW(p) · GW(p)

So why has no one proposed the Prick Your Finger diet? Why do none of the popular diets involve measuring blood sugar at all, ever?

Excellent question.

Replies from: David_Gerard, mattnewport
comment by David_Gerard · 2011-04-11T20:02:51.151Z · LW(p) · GW(p)

And of course, Tim Ferriss writes of not merely doing this, but attaching a continuous blood sugar monitor to his body.

comment by mattnewport · 2009-04-11T08:30:07.514Z · LW(p) · GW(p)

The G.I Diet is based around controlling blood glucose response (G.I. stands for Glycemic Index and refers to the measured blood glucose response for different foods). I'm not sure if measuring blood sugar levels on a frequent basis is ever suggested as a useful part of the diet but I believe they suggest getting your own insulin response measured.

Replies from: jimrandomh
comment by jimrandomh · 2009-04-11T11:46:44.639Z · LW(p) · GW(p)

Amazon book search is quite handy; it confirmed for me that the GI diet does not promote or even mention measuring blood sugar yourself. The premise seems to be that foods with high glycemic index (that is, foods which cause a large blood sugar response in a representative test subject) are bad, but that you should use the book and web site to determine which foods those are, rather than measuring anything yourself. It definitely doesn't make the hunger vs. hypoglycemia distinction anywhere.

(Aside: Blood sugar levels aren't something that you can usefully "go get measured"; you have to get a test kit and administer the test yourself, because the point is not to measure one number but to see how it changes in response to the various foods in your diet and learn what those changes feel like. Monitors are often given away for free or almost-free as a loss leader for the test strips, which cost about $1 each and can only be used once.)

Replies from: mattnewport
comment by mattnewport · 2009-04-11T22:37:25.417Z · LW(p) · GW(p)

My understanding is that getting your insulin response measured involves several hours at a medical facility after a period of fasting. You consume a glucose solution and then your insulin levels (and blood sugar?) are measured at intervals over a period of several hours. The data obtained is used as a diagnostic in identifying insulin resistance which can be a precursor to type 2 diabetes I believe. The idea of getting the test done is to establish whether you are already on a possible path to something like metabolic syndrome. Measuring blood sugar levels yourself as part of monitoring the diet progress sounds like a good idea, but I'm not a nutritionist.

Replies from: jimrandomh
comment by jimrandomh · 2009-04-11T23:53:12.836Z · LW(p) · GW(p)

The test you describe is called a glucose tolerance test. You'll spend a few hours in the hospital, and get back a result that says "you don't have diabetes". Whoopty doo. What you want is to find out whether your blood sugar ever goes low, if so when, and what it feels like. That test can't be done in a hospital, because being in a hospital means not following your usual routine.

comment by Eliezer Yudkowsky (Eliezer_Yudkowsky) · 2009-04-11T13:21:56.017Z · LW(p) · GW(p)

I was ambiguous over whether to promote this, but it does make a nice illustration of the fallacy of compression. Hopefully the upvotes indicate that others agree with me about the high post quality. But we probably want to compress further dietary discussion into the existing posts, lest LW wander off-topic.

Replies from: billswift, Cyan
comment by billswift · 2009-04-11T15:50:57.645Z · LW(p) · GW(p)

I don't think it's necessarily off-topic, since reducing distractions and difficulties to thought arising from physical body processes should also help make us Less Wrong in our thinking.

comment by Cyan · 2009-04-11T15:31:13.774Z · LW(p) · GW(p)

I upvoted it for the brilliant-and-yet-obvious-in-hindsight suggestion of monitoring blood sugar. Seth Roberts would approve, I think.

comment by JulianMorrison · 2009-04-11T12:54:53.379Z · LW(p) · GW(p)

What does it feel like, going the other way and having too much sugar at once?

Replies from: jimrandomh
comment by jimrandomh · 2009-04-11T15:01:20.115Z · LW(p) · GW(p)

Slightly elevated blood sugar (120-160mg/dL, what you'd get from an ordinary amount of junk food) is imperceptible, even with training. Moderately elevated blood sugar (160-200mg/dL, what you'd get from a thanksgiving dinner) causes slight lethargy. Highly elevated blood sugar (200-250mg/dL, what you'd get from eating a bag of cookies quickly) causes full lethargy, nausea, and excess urination leading to dehydration. All of these can be produced psychosomatically, so a test kit is required. The interesting symptoms don't start until your blood sugar gets higher than normal biochemistry will allow it to go, and stays there. However, frequent or chronic high blood sugar causes slight damage to every organ, causing increased risk for a long list of things including heart disease.

Replies from: ciphergoth
comment by Paul Crowley (ciphergoth) · 2009-04-13T10:49:40.870Z · LW(p) · GW(p)

For those like me used to BG meters that measure in mmol/l (I have a diabetic partner) the rule is to divide by 18:

  • 120 mg/dL ≈ 6.7 mmol/l
  • 160 mg/dL ≈ 8.9 mmol/l
  • 200 mg/dL ≈ 11.1 mmol/l
  • 250 mg/dL ≈ 13.9 mmol/l

since glucose is about 180 daltons, but then there's a factor of ten in the conversion from decilitres to litres