Friday, December 31, 2010

Today I got out of bed feeling like an old man -- specifically, a stiff and sore and irritable old man. This was one more proof of my Unfamiliar Exercise Theorem: any specific physical activity which you're not accustomed to doing will make you hurt the next day. It doesn't matter that you're in the habit of doing more strenuous activities which don't make you hurt the next day -- whenever you do anything unfamiliar, you are going to hurt.

For example, yesterday I did not get out of bed feeling stiff and sore and irritable, even though I'd done an 8 mile trail-run the day before that. I'm used to doing long trail-runs, after all. Feeling confident in my physical resources, I then went to Sugarloaf Ridge State Park and did a 6-mile hike -- which seemingly should have been easier on me than the run had been, as the distance was shorter and I was walking rather than running. But this morning it seemed that the hike had been much, much harder on me than the run had been.

Why should the hike have been harder on me than the run? Probably because I do a lot more running than hiking. Also, it probably matters that I was hiking to the top of a 2729-foot peak, so I did more climbing than I do even on a steep trail-run. Even so, I have the impression that hiking simply uses different muscles -- or uses them differently -- than running over the same terrain. If I had run to the top of that mountain, maybe I wouldn't have felt so sore this morning.

The hike yesterday and the run the day before were both made possible by a break between Pacific storms. As the next storm was rolling in today, I exercised at the gym instead of outdoors. Fortunately, there's a hot tub at the gym, so I spent 15 minutes in hot water after I worked out, and now my limbs and muscles feel a lot better.

I don't mean to suggest that unfamiliar forms of exercise should be avoided. I don't regret doing the hike yesterday. If feeling sore afterwards is the price of participating in activities which aren't routine for me, then that's the price I'll have to pay. I'm never going to be in the habit of hiking routinely, but I like to do it when I get the chance -- especially when the landcape is looking green and brilliant.

Because of the recent storms, I needed to be willing to tolerate a lot of mud and flowing water on the trails. I decided I was willing.

The climbing was hard work, but I always enjoy the way the view gets better and better as you get higher and higher.

At the top, I was startled by how much I could see in every direction. To the west, the Sonoma Valley, and the wineries in the Kenwood area...

...and to the east, the wineries of the Napa Valley. A phenomenon which geologists don't yet fully understand causes the wine which comes from that side of the ridge to be more expensive. 

To the south, I could see San Pablo Bay, and beyond it (if I squinted hard enough) the skyline of downtown San Francisco, 60 miles away.

One of the reasons I like to spend time outdoors, off the beaten path (and particularly on top of mountain ridges with good views) is that I think it's always a good idea for humans to do things which give them a sense of perspective.

It's awfully easy for us to see ourselves as bigger and more important than we actually are in the larger scheme of things. It's also awfully easy for us to see ourselves as very much separate from nature, when the truth is that we are part of nature and always will be.

It seems to me that a sense of perspective is one of the things which a lot of diabetes patients lack -- particularly the angry "Why Me?" crowd, who seem to have been getting through life (up to the point of their diagnosis) on the assumption that they enjoyed a personal exemption from the laws of nature. ("Sure, other people get diseases, and that's fine -- but why should I of all people have to get a disease? Isn't it understood that things like that only happen to other people?")

Alas, none of us are as special as we would like to think. It's a big world, and a big universe beyond it, and the role which we play in it personally is neither large nor extraordinary. That doesn't mean we're unimportant, but it does mean that we shouldn't count too much on receiving special treatment on account of our exceptional brilliance and charm. The things that can happen to living organisms in general can happen to us, too.

We have to cope with that, with as much dignity as we can muster. And we need to keep in mind that there is no more undignified statement a human being can make than "Why me?".

Why not me? That's the question we should be asking ourselves.

Another thing that has given me a sense of persopective is the photography of Thierry Legault, the French amateur astronomer who specializes in capturing "transits" (things passing in front of the sun and moon, as viewed from earth). When astronomers talk about transits, they usually mean a transit of a planet, such as Venus (which sometimes passes between us and the sun). But Legault likes to capture man-made objects doing the same thing:

I didn't realize this until recently, but if you have the right kind of software you can calculate the moment when a spacecraft is going to pass across the visible disk of the sun or moon. You have to act fast, though! Most spacecraft shoot across the sky so fast that you have less than a second to click the shutter before they have crossed the disk of the sun or moon entirely, and become invisible. Timing is everything.

Here is Legault's image of the International Space Station, caught against the background of the sun:

And here is the Space Shuttle doing the same thing:

And here is Legault's luckiest shot to date: a rendezvous between he Space Shuttle and the International Space Station.

And Legault isn't even using some kind of huge telescope unavailable to amateurs. He's using very basic equipment, and a small telescope. His secret is knowing when to take the picture. I've seen a video of these vehicles flying across the disk of the sun, and I know for a fact that I couldn't hit the shutter fast enough to capture a picture like that.

Oh well; here's my wish for all of us in the year 2011: that we can develop, and maintain, a sense of perspective about the issues we need to deal with as diabetes patients.

Friday, December 24, 2010

I'll be taking a holiday break from blogging for a little while.

Enjoy the holidays -- but not too much!

Thursday, December 23, 2010

My company shuts down for a week at this time of year, and the last day before the shutdown (which happened to be today) is a half-day. People only need to be there until noon, and it's not a terribly serious workday -- a lot of departments have some kind of Christmas breakfast event. And a small group of us become strolling players, to play Christmas music at work on this day, in various locations around the work site. It's a tradition that goes back 15 years. It was my idea, if I'm remembering this correctly; anyway, we've done it every year since then, although the personnel involved have changed over the years as employees come and go. 

This year I thought the music went exceptionally well, and was exceptionally well-received, despite the loss of our French-horn player, who just became a father slightly ahead of schedule, and was home dealing with things of greater interest to him than the low harmony line for Have Yourself A Merry Little Christmas.

As usual, our performances today started with an appearance at the Christmas breakfast event for a division of the company that I used to work for. It's a waffle breakfast, and they insisted on feeding waffles to the musicians after we were done playing. (It was more or less a re-enactment of the Christmas Drama I described in my blog post of December 21st, except that I caved in and ate the waffle they gave me -- which, as I'm sure you can guess, is why I ended up doing such a long and challenging run today, to make up for it.)

After that, we just strolled around from building to building, playing three or four Christmas tunes at each location and then moving on. We also played "elevator music" -- that is, we used the freight elevators to get from one floor to another, and we played tunes in the elevator so that, when we arrived on the next floor, we made a fairly dramatic appearance as the elevator doors opened.

After all the rain we've been having lately, it was dry today -- which is nice, if you're planning to carrying your fiddle from one building to another to play Christmas tunes. It's also a nice opportunity to go for a run outdoors. After my half-day at work was over, I headed down to the state park to do some trail-running. By then it had warmed up enough that I could get away with wearing shorts rather than long pants. (Running is easier in coastal California -- no snow to deal with.)

Contrary to what would be the normal expectation elsewhere, this is a greener-than-average time of year in these parts. The rain is bringing the grasses and mosses to life.

I like running in the spooky dark woods, and then catching glimpses of the sunnier world nearby.

When I made it up the hill to Lake Ilsanjo, I found that it was overlfowing...

...and flooding right across the trail I was trying to run across.

In a way, it was nice to have all choice eliminated -- instead of trying to maneuver my way around puddles, I had no choice but to step into water a couple of inches deep. It actually wasn't uncomfortable -- this water wasn't snowmelt, it was just rainwater that had run down the hillsides and overfilled the lake. It cleaned the mud off my shoes without giving me frostbite.

I like being in nature, so long as nature isn't trying too hard to bite me in the ass.

We're part of nature whether we want to be or not, so we might as well learn how to appreciate and enjoy it. And, as the saying goes, don't forget to stop and smell the Egrets.

Wednesday, December 22, 2010

Today the Food and Drug Administration announced a very large recall of defective glucose test strips made by Abott, sold under these brand names:

Up to 359 million test strips may be affected by the recall. The strips have a defect which causes them to absorb a smaller volume of blood than is required for an accurate measurement; as a result, tests made with these strips tend to give false low readings. (How low? No one is saying, but it must be very low indeed, considering how loose the accuracy specs for glucose meters are.) For the details, see the FDA's press release.

To test one of your strips for accuracy at home, wave it in the air a few inches above a Starbucks 7-Layer Frappucino for ten seconds:

Then plug into your meter without applying any blood to it. It should read at least 100 and no more than 120.

Tomorrow's my last day at work before the company shuts down for the holidays. It's also the day that five of us are going to stroll around from office to office playing Christmas music. (There were supposed to be six of us rather than five, but I found out yesterday that our French horn player had just became a father, and I suspect he's going to be otherwise occupied tomorrow.) We wanted to squeeze in one last practice today, and we had to do that at lunchtime, so I couldn't do my usual mid-day run.

Instead I went to the gym in the evening. There I ran into my least favorite kind of crowd to share a gym with: a bunch of swaggering athletic guys in their late teens, talking and laughing at each other very, very loudly. The sound of rowdiness always makes me feel like it would be a lot wiser to be someplace else. Nothing happened, other than a great deal of unpleasant noise, but I don't find it easy to concentrate on my workout while I'm waiting for a gang of guys who aren't quite in control of their behavior to lose it completely.

If it's true, as Woody Allen said, that 80% of success is showing up, I think it's just as true that 80% of maturity is shutting up. America needs to start using its Indoor Voice. Maybe that should be our New Year's resolution.

Here's an interesting research finding: insulin modulates the sense of smell. Researchers at the University of Tubingen in Germany studied the sensitivity to smell of healthy volunteers, and found that the ability to detect odors declines markedly when the insulin level in the blood rises. This effect may be designed by nature to prevent overeating by masking food odors (and therefore reducing appetite after one has had enough to eat).

The study didn't look at people with Type 2 diabetes; I wonder if loss of sensitivity to insulin changes this pattern, so that insulin-resistant people remain hungry longer and eat more.

My own appetite has never been easily discouraged, but it was especially robust around the time of my diagnosis. My appetitate became absolutely insatiable then -- while eating one large meal, I'd already be thinking ahead to the next one.

I don't know if the sense of smell had anything to do with it, but insulin is thought to regulate appetite in a variety of ways, and it makes sense that loss of sensitivity to insulin would cause appetite to become unregulated one way or another.

This issue obviously is relevant to the famous correlation between Type 2 diabetes and obesity. Does the correlation arise because increasing body fat causes a reduction in sensitivity to insulin? Or does it arise because impaired sensitivity to insulin increases appetite and thus leads to an increase in body fat? Almost certainly both are true, but a lot of people have rather recklessly staked out a political position to the effect that only one of these can be true.

In other words, either diabetes patients are totally to blame for their condition, because they let themselves get fat and this made them lose their sensitivity to insulin, or diabetes patients bear no responsibility at all for their condition, because they wouldn't have got fat if their loss of insulin sensitivity hadn't made it happen.

My strong suspicion (which I cannot prove to be right) is that both of these effects play into each other and create a vicious cycle: the more weight you gain, the more insulin sensitivity you lose -- and the more insulin sensitivity you lose, the more your appetite and weight get out of control. Some people, who don't carry any of the genes that promote insulin resistance, don't have to confront this issue -- they may get fat and they may not, but insulin sensitivity doesn't play a role in the matter. Some people do carry genes that promote insulin resistance, in which case insulin sensitivity is bound to be a big issue for them.

I try not to waste a lot of time imagining what might have been different in my life if I had been born with different genes. Who knows? And who cares, for that matter? It's not as if you can take your genes back to the shop and exchange them for a different set. (And suppose you could: who knows if the replacement genes they gave you would be any better? I'd rather have the genes for diabetes than the genes for Huntington's disease!)

Well, anyway, I guess the lesson of the new research from Tubingen is that we now have more evidence than we did before that insulin regulates appetite, at least in healthy young volunteers -- and might not do such a good job of it in you and me.

Tuesday, December 21, 2010

We managed to find an interval when it wasn't raining to go for a run today, which was nice. Now it's pouring, but I'm warm and safe indoors, which is even nicer.

The really heavy rain in California is in the south; up here I'm not hearing any flood warnings yet. I can afford to look at the Pacific storms rolling in as nothing more serious than an inconvenience to my running schedule. 

A special feature today!

"Aren't you going to have some of my Christmas cookies?"

"No, thank you -- they look good, but no."

"Oh, come on, have some!"

"Really, no thanks."

"But I made them myself from scratch."

"Really, no thanks."

"But it's Christmas. You know, Christmas. This is Christmas party."

"Merry Christmas, but I don't think I need to be having any cookies right now."

"What have you got against Christmas?"

"I don't have anything against Christmas -- I just don't need to be having any cookies right now."

"Are you on a diet?"

"Well, sort of."

"You're not that fat, really. I mean, compared to a lot of people..."

"Thank you, but I still need to watch what I eat."

"Not at Christmas!"

"Why not at Christmas?"

"Because it's Christmas. Nobody watches what they eat at Christmas."

"But I still need to watch what I eat."

"Why do you need to watch what you eat, when nobody else does?"

"I just need to be careful about what I eat."

"Are you a vegan?"

"No, I just need to be careful about what I eat."

"Are you kosher?"

"No, I just need to be careful about what I eat."

"Are you anorexic?"

"No, I just need to be careful about what I eat."

"Do you have a food allergy?"

"No. Well, I have an allergy to shellfish, but..."

"There aren't any shellfish in my Christmas cookies."

"I'm very pleased to hear that there aren't any shellfish in your Christmas cookies, but I still need to be careful about what I eat."


"I just do."

"Did you have weight-loss surgery?"


"Did you have your stomach stapled up so tight that you can't eat anything bigger than a raisin?"

"I told you, no."

"Are you going through chemotherapy?"


"Have you got some kind of a disease?"

"Look, I have a health problem that can get out of control if I'm not careful about what I eat."

"Did you have your gall bladder removed?"


"What kind of a health problem is it, that you can't say what it is? Doesn't it have a name? Is this really your way of saying that you're an alcoholic or you're a drug addict or you're insane?"

"No. Look, it's called diabetes, okay? Diabetes."



"You don't say! My aunt had diabetes. I don't really remember her because he died when I was only two. And my next door neighbor has it, but we don't see so much of him now that he can't walk any more."

"Thank you for mentioning these details, because nothing cheers me up at a party more than hearing about people being disabled and killed by the same disease I've got."

"There's no call to be sarcastic. If you need cheering up, why don't you eat some Christmas cookies like a normal person?"

"Because they make my blood sugar go up."

"Don't be silly. Everything you eat makes your blood sugar go up."

"Not as much as Christmas cookies make my blood sugar go up."

"There isn't that much sugar in these cookies. Actually, there's more flour and oatmeal than there is sugar."

"But flour and oatmeal are made out of starch."

"Exactly -- so they're not made of sugar."

"But starch is made of sugar."

"No it isn't. I checked the nutritional label."

"The nutritional label makes a distinction between sugar and other carbohydrates that is completely meaningless in practical terms."

"If it was sugar they'd have to call it that on the label. So it's not sugar."

"But it turns into sugar as soon as you swallow it."

"How can that be? One thing can't turn into another."

"Of course it can. Starch is just bunch of sugar molecules chained together. In your stomach they become unchained, and break up into sugar. In fact, they start doing that in your mouth. Starch breaks up into sugar so easily that it doesn't require stomach acid to break it down. Even saliva is strong enough to break it down, so it starts breaking down while you chew it."

"Now you're being really gross. Why do we have to be talking about your saliva at a Christmas party?"

"We wouldn't be talking about my saliva at a Christmas party if you had backed off when I said I didn't want a cookie."

"How was I to know you were going to bring all this stuff up?"

"How was I to know you were going to assume that 'no thank you, I don't believe I'll have a cookie' is just another way of saying  'please start interrogating me about why I'm not having a cookie until we end up talking about my saliva'?"

"Well, one cookie is not going to kill you."

"If I keep eating cookies, eventually one cookie will kill me. I just won't know which cookie was the one that did it."

"That's ridiculous. You can't die from a cookie."

"How do you know?"

"I just know."

"If we're going to be betting someone's life on an untested assumption, could we please make it your life rather than mine?"

"Well! Sounds like someone doesn't have the Christmas spirit."

"I guess the question is which one of us that is."

--- CURTAIN ---

Monday, December 20, 2010

Great, now it's official -- I've gained gained 5 pounds since Thanksgiving. Perhaps this is not all that unusual (I've read that the average American gains 7 pounds during the holiday season) but I was hoping I could avoid doing it this year. Well, I failed miserably at that. I'm just glad I'm doing okay on the blood-sugar front.

My glucose results today were surprisingly good, considering that I went to another party yesterday. And not only did I go to the party, I went straight from there to an Irish music session at a pub in the same neighborhood (the Plough and Stars in San Francisco).

Not that I consumed anything significant at the pub. I ordered a pint of Guiness, to pay the rent so to speak, but I didn't even drink half of it. I'd had wine at the party, and I was going to have a long drive to get home. So, while I was at the pub, I concentrated on playing instead of drinking. I kept as busy as I could. As you can see, the camera shutter could hardly capture my flying fingers on the fiddle...

How did I manage to go to another party and still get good glucose results the next day? Well, for starters, I was comparatively restrained about carbohydrate intake (if not calorie intake) at the party. But I probably got a lot of help from the lingering effects of a challenging workout the day before.

I'd done a tough trail-run on Saturday (about 13 miles, on very hilly terrain). It was part of the marathon training program -- even though I ended the run feeling very unsure that I want to continue the training and do the Napa Marathon this year. I'm still reserving judgment for a little while on that decision. I didn't enjoy Saturday's run very much, and it didn't make me look forward to doing longer and longer weekend runs for the next 8 weeks or so. But I was running in the rain, and it felt so miserable that imy discomfort may have given me a bad attitude about the whole enterprise. It rained on me when I ran today, too, but today the run was much shorter and it was only raining hard for about five minutes of it, so it didn't bother me that much. On Saturday it rained a lot more than that, and I was slipping and sliding over extremely muddy trails. I can't claim that I had a great time.

The one morale-building aspect of the long, lonely, rainy, muddy trail-run on Saturday was something that happened about half-way through the run, when I was in one of the remotest corners of the state park, and I was seeing so few other people on the trails that I felt almost creepily alone. I was coming down a steep hill when I encountered a couple of mountain-bikers climbing up it, and one of them greated me by name in a Japanese accent. I looked up in amazement. They were coworkers of mine, who used to be my running buddies when we all worked in the same building. One of them had moved back to his native Japan, and the other had taken a job with another company. Now they were both working at my company again, but at a different facility north of town, and I hadn't talked to either of them in a few months. Now I ran into them in the middle of the woods, seven miles from the nearest road.

It was like A Sign From Above, telling me that I was doing the right thing. That it was OK for me to be here. That I didn't need to feel weird and masochistic about being out here exercising in bad weather on muddy trails. That sane, healthy people did this kind of thing, and they weren't being babies about it, so why should I?

We stopped and had a little chat, and then we went our separate ways. Six miles later, when I finally completed the run and got back to my car, I ran into them again -- because they were parked next to me and they finished their ride just as I was finishing my run. We laughed at the coincidence of timing, and once again I felt somehow encouraged and validated by my chance encounter with them.

But marathon training is a huge time-commitment, so I'm still not sure I'm ready to go through it again. I guess it will depend (as it has always has depended in the past) on what my friends can talk me into.

As for why my blood pressure is up tonight, I don't have a ready explanation.

I found an "issue ad" put out by the sugar industry in the 1970s, explaining that sugary foods can't make people fat, because if they could, children would get fat, and we know that doesn't happen. I realize the ad is not exactly readable in this form, so I've transcribed the text below the picture.

If sugar is so fattening, how come so many kids are thin?

Next time you pass a bunch of kids, take a look. Kids eat and drink more things made with sugar than anybody. But how many fat kids do you see?

The fact is, if you constantly take in more food than your body needs, you'll probably get fat. If you eat a balanced diet in moderation, you probably won't. And sugar in moderation has a place in a balanced diet?

For kids, eating or drinking something with sugar in it can mean a new supply of body fuel.  Fuel that can be used in not too many minutes. There's a useful psychological effect, too. The good natural sweetness of sugar is like a little reward that promotes a sense of satisfaction and well-being.

The thing is, good nutrition comes from a balanced diet. And a balanced diet means the right amounts and right kinds of protein, vitamins, minerals, fats, and carbohydrates. Now, what's one important carbohydrate? Sugar.

Sugar. It's not just good flavor; it's good food.

I guess that settles that. Sugar is good food, and we should all eat lots of it in order to maintain dietary balance.

Friday, December 17, 2010

Well, today's results are a bit of a case study, aren't they? A fasting result of 101 (when it was only 79 the day before), and later in the day a post-prandial result of 94. What's going on here? Shouldn't the fasting test have been lower, and the post-prandial test higher?

The results may seem odd, but both of them can be explained -- and the explanation partly relates to the issues I was writing about yesterday.

Yesterday was my rest day from exercise, and it also happened to be the day of a big Christmas party (when those two things coincide, I've got to expect that the next day's fasting test will be higher than usual). The particular circumstances of the Christmas party are worth mentioning. I was a bit distracted, because I was playing music at the party, and for the first hour or so I was too busy to have anything to eat. And by the time I did eat, I was feeling very hungry and impatient. There were multiple buffet tables, and there were also waiters walking around with trays offering you various hot appetizers (crab cakes, meatballs, deep-fried pouches of something gold-colored with cheese in the middle). And there was a free bar, where they were giving generous pours of good wine (I chose the Acacia Pinot Noir, an old favorite of mine which I am seldom offered for free). After my hasty tour of the bar and buffet tables (eating who knows how much of who knows what) I went back to the side room to play more music, while music-loving Good Samaritans refilled my wine glass.

A naive person might think that, because I then took a post-prandial glucose test and got a startlingly low 89 as a result, my system was doing a wonderfully efficient job of processing the party food I'd eaten, even though yesterday was a non-exercise day and my insulin sensitivity was presumably not as high as it could have been. But, as I explained yesterday, this low post-prandial result probably said more about what I drank than about what I ate. My liver was still processing the alcohol from the wine, and so long as it was busy doing that, it wasn't making its usual contribution to my blood sugar. Therefore, my blood sugar was brought artificially low -- but only temporarily! Late last night, after my liver was finished with processing the alcohol out of my system, it must have decided to make up for lost time, and release a generous supply of glucose into my bloodstream. Hence the elevated fasting result this morning.

But then came lunchtime, and (for the first time this week) I was able to make my usual lunchtime run. Partly because of the elevated fasting test this morning, I decided to go on a very tough, hilly run -- even though it was raining and my running buddies declined to go with me. After that, I had a low-carb lunch. And, lo and behold, my glucose an hour after eating was 94 -- 7 points lower than my fasting result had been.

All told, my experience over the last 24 hours illustrate how various factors affect blood sugar. Eating a bunch of party food, on a non-exercise day, leads to elevated glucose -- but only after a temporary, artificial low created by alcohol consumed at the party. Then, after an elevated fasting test the next morning, hard exercise and a low-carb lunch brings glucose back down again. All of it makes sense if you think about it.

Thursday, December 16, 2010

Okay, let's talk about something of interest to many people with diabetes.

A reader who is in the pre-diabetes stage (and would rather not move on the the next) was asking me about alcohol, and specifically about whether I have found alcohol to have a problematic impact on my blood sugar. I gave him a very brief answer, to the effect that I love wine and I haven't found it to affect my blood sugar adversely. However, it's a very complicated topic which needs to be discussed in more detail than that. I wouldn't want anyone to quote me as having said that everyone with diabetes can drink as much as they like and not worry about it.

Leaving diabetes out of the discussion for the moment, alcohol has a very strange and paradoxical relationship with human health. Because it tends to be addictive (or at least heaps o' fun), people often consume too much of it. This has been found to increase the risk of various health problems, including hypertension, liver disease, depression, suicide, violence, and vehicle accidents.

Because of these problems, you would naturally expect that if we plot human mortality rate against alcohol intake, the result would be a staright-line graph, going up and to the right. In other words, the people who drink the most are going to be dying at the highest rate, and the people who don't drink at all will be dying at the lowest rate -- because alcohol is bad for you, and that's that.

It sounds perfectly reasonable. So reasonable that it almost seems like it wouldn't be worth the bother of looking at the data to confirm that this is true. And yet, when researchers did start looking at the data (especially the data for cardiovascular disease), they got a bit of a shock. Plotting mortality against alcohol intake doesn't give you a rising straight line; instead it gives you the famous "U-shaped curve":

Heavy drinkers do have an elevated mortality rate, just as we would have expected -- but the people with the lowest mortality rate are those who drink moderately, not the people who don't drink at all. Apparently we can't reduce the issue to a simple  "alcohol = bad" equation. Seemingly alcohol has some kind of protective effect (particularly on the heart), so long as you don't overdo it, and researchers have been expending considerable effort trying to figure out exactly what that protective effect is.

Those of us who love wine have greatly enjoyed reading about the U-shaped curve, and its presumed contribution to the "French Paradox" (that is, the unexpectedly low rate of cardiovascular disease in France and some other Mediterranean countries). But not everyone has been quite so ready to accept the cheery implications of the U-shaped curve.

In 1989, the authors of the British Regional Heart Study argued that the U-shaped curve (or at least the left side of it) is misleading. Their reasoning was that the ranks of those who drank very little or not at all included "sick quitters" -- people who used to drink heavily, but no longer do so because it had caused them serious health problems. The reason for the elevated mortality on the left side of the "U" is that it includes a bunch of people who ruined their health with heavy drinking and then had to give it up. If we're going to compare people's health with their drinking habits, we have to include their former drinking habits, not just their current ones.

Fair enough, but later studies did exactly that, and found that the U-shaped curve was still there. The people with the lowest death rate are moderate drinkers, and the difference is mainly accounted for by their comparatively low rate of cardiovascular disease.

Of course, this only raises further questions about why moderate drinking should reduce the rate of cardiovascular disease (and why heavy drinking doesn't reduce it). The answer seems to be that alcohol tends to reduce the buildup of cholesterol placques on arterial walls. How does it do that? No one is sure, but alcohol does seem to elevate HDL ("good") cholesterol, which functions as an artery cleanser. But wouldn't this work for heavy drinkers, too? Actually, it does -- alcoholics have the cleanest arteries going, but it doesn't do them any good, because heavy drinking has other effects (such as driving up blood pressure) which more than cancel out the artery-cleansing benefits. Moderate drinkers get the desirable effects of alcohol, without getting the undesirable ones. Because the undesirable effects are stronger than the desirable ones, drinking so heavily that you get both at once is a losing proposition.

What scares doctors away from recommending moderate drinking to people who don't drink currently is the addiction potential. If you advise teetotalers to start drinking moderately, there is no way to be sure they won't go too far, become alcoholics, and end up increasing their mortality risk instead of reducing it. So, don't expect doctors to start defining Minimum Daily Requirements for alcohol anytime soon. However, there does seem to be a consensus that moderate drinking is not bad for you so long as it stays moderate.

However, I should mention another health issue: alcohol is a calorie-dense substance. It's nearly as calorie-dense as fat, actually. Alcohol does not tend to help you in terms of weight control.

Okay, that's the situation for the general population. What about the diabetic population? Well, if you have diabetes, the issue of alcohol and health gets more complicated (just as every other issue in life gets more complicated if you have diabetes).

I should mention at this point a fact which many people with diabetes are unaware of: your digestive tract is not the only part of you that releases glucose into the bloodstream. Even if you didn't eat anything at all for a couple of days, there would still be glucose flowing into your blood -- from your liver. The body can't rely on the digestive system to maintain your blood sugar level, because you're not constantly eating. To protect you from low blood sugar, the liver is constantly doling out a minimum ration of glucose, and if your blood sugar actually does go low, your liver releases an extra burst of glucose to correct the problem (often it overcorrects the problem, if you have diabetes, because the burst of glucose it releases is too big for you to handle). Anyway, the liver plays a big role in maintaining your blood sugar level.

Okay, what happens when you drink alcohol? Well, the body has to go to work on processing it, so that it doesn't build up to toxic levels in the blood. And what part of your body does that processing? You guessed it -- your liver. And why is that fact pertinent to the subject of alcohol and diabetes? Because, while your liver is processing the alcohol, it drops other tasks. And one of the tasks it drops is the task of doling out your minimum ration of glucose. Your pancreas, however, does not stop doling out insulin just because your liver has stopped doling out glucose. In other words, alcohol deactivates one function without deactivating the other, so the two functions are now out of balance, and you may have more alcohol leaving your blood than entering it. The result is that drinking alcohol is likely to drive your blood sugar downward.

This makes it sound as if alcohol must be entirely beneficial for anyone with diabetes. (Who needs insulin? Let's get drunk instead!) Alas, it's not that simple. The glucose-lowering impact of alcohol is not consistent or predictable (drinking can sometimes drive blood sugar up instead of down, especially in the case of carbohydrate-rich drinks such as beer), and in any case the impact is temporary. You really can't use alcohol as a diabetes medication. Even worse, alcohol can make it a lot harder to use insulin or other glucose-lowering medications safely.

Suppose you're an insulin user. One night at dinner you test your blood sugar, so that you can figure out how big a dose of insulin to give yourself. The test reads pretty high, so you give yourself a pretty big dose. But a test is only a snapshot of a moving target; you don't realize that, when you tested your blood sugar and found it high, it was actually dropping steeply (because of the beer you were drinking with dinner). Now the insulin you injected will push your blood sugar down even faster than it was already falling. The alcohol and insulin magnify each other's impacts, and the result could be an episode of severe hypoglycemia. If you've heard anyone say that people with diabetes can't or shouldn't drink, this is the issue they're thinking of.

If you have diabetes (especially if you're taking insulin or oral diabetes drugs which drive blood sugar downward), you have to do enough testing on yourself to find out how alcohol affects you, and develop safe guidelines for yourself.

If you're managing your diabetes without any glucose-lowering medications, alcohol-induced hypoglycemia is less likely to be a problem for you. It hasn't been a problem for me. The hypoglycemic episodes I've experienced since my diagnosis in early 2001 have not been frequent or severe, and not one has ever occurred while I was drinking. I can't be sure it will never become a problem for me, but so far it hasn't. So I guess that's one more advantage of the unmedicated approach.

Alcohol does tend to drive my blood sugar downward; I've done enough testing to establish that. But the effect isn't so dramatic as to be any kind of a problem. 

I should admit that I went to a big Christmas party today, and I was drinking wine there -- and that this could be part of the reason why my glucose result after dinner was a remarkably low 89 (even though this was a non-exercise day). Still, it was a low-carb dinner, so I don't think the wine would have had to make a very dramatic impact to give me that kind of result. I do sometimes get very low post-prandial results -- without any help from alcohol.

Wednesday, December 15, 2010

I think I discovered a diet trick (and not a moment too soon!).

The cafeteria at work was serving a Chinese stir-fry meal today, and as usual my attempt to stop them from giving me too much rice with it was not fully successful (the concept of small serving sizes is an unfamiliar one to most people these days). So, once again, I was confronted with the Asian Restaurant Problem: how to disentangle the food you thought you were purchasing from the foothills of starch over which it has been distributed.

Then it dawned on me that the Asian Restaurant Problem has a simple Asian Restaurant Solution: do not use a fork. Chopsticks were available, and I used those instead. This wouldn't have been much of a solution if I had grown up in Asia, and had become so dexterous with chopsticks that I could shovel platefuls of rice into my mouth with the greatest of ease. But as I am only an awkward American, eating with chopsticks is a rather slow and inefficient process for me, and it tends to work better for bite-sized items such as shrimp and sliced vegetables than it does for grains of rice. In effect, the meal was screened by my method of eating it. A lot of the rice simply fell through the cracks, so to speak.

I eat too fast anyway, even when I am trying to make a conscious effort to slow down. Maybe I should eat with chopsticks more often, just as a way of eating slower and more consciously. Of course, if I ate with chopsticks all the time, I'd get better at it and the advantage would gradually disappear. Later I'd have to find another way to handicap myself. Perhaps the next phase would be to eat using a pipe wrench -- in my left hand.

I suppose it's worth asking why the impulse to eat in a hurry is so strong in me. I'm not sure I can give an explanation of this that would make sense to anyone, but part of it is that I don't tend to see meals, even excellent meals, as sensual experiences to be savored and lingered over. (I would make an extremely bad Frenchman, clearly.) I tend to see a meal as the solution to a problem.

You see, I don't like being hungry. In fact, I find it irritating and distracting to be hungry. When I'm hungry I can't concentrate on anything else. So, as far as I'm concerned, all time spent in a hungry state is time wasted. Why prolong the irritation? I don't want to take my time over a meal any more than I want to take my time over turning off a ringing alarm clock, or brushing a spider off my leg, or emptying a full bladder. I'm not unaware of the pleasant relief that arrives when the alarm clock is silenced, the spider is removed, or the bladder is emptied -- but that doesn't mean I want to make the relief all the sweeter by delaying it. I just want to solve the problem and move on to something more interesting.

I'm not saying this is the way people ought to be, of course. I'm simply admitting that this is the way I am. I realize that being this way creates problems for me. When you think of your appetite as an irritant which must be neutralized, as speedily as possible, whenever it manifests itself, this habit of mind does not exactly make it easy for you to stay focused on sensible food choices or portion control. That's why I wonder if my impatient approach to eating is a common characteristic among people who have trouble controlling their weight. It even makes me wonder whether the real reason people gain weight at this time of year is not so much the Christmas cookies but rather the distractions. It's a busy time of year; we're thinking about a lot of things besides dietary discipline.

I've gained a few pounds lately, and although holiday parties did contribute to that, impatience may have contributed more. Another contributor is that lately, in order to correct a rising trend in my fasting tests, I've shifted my diet more in the direction of high-fat, low-carb. It's worked in terms of glucose control (79 this morning, for the third time in a week), but whenever I do this my weight tends to go up. It's ironic, given that low-carb diets were originally promoted as way to control body weight, and for some people it seems to work exactly as advertised. I don't know why it doesn't work that way for me, but it seems not to.

My own experiments with diet over the past 10 years have shown that, for me, a low-fat vegetarian approach is great in terms of weight control but iffy in terms of glucose control, while a high-fat, low-carb, carnivorous approach is great in terms of glucose control but iffy in terms of weight control. (If there's a perfect compromise between these two approaches which is great for glucose control and weight control, I haven't hit on it yet.) These generalizations have been too consistently true in my case for me to doubt their validity -- however, it does not follow that they would be valid for everyone else. The reports I hear about different people's response to different kinds of diets forces me to admit that what is true for me is not going to be true for everyone.

However, not everyone is as willing as I am to concede that different people will respond differently to the same diet (and that it is therefore useless to inisist that one's own preferred diet should be adopted universally). That's why my daily blog posts no longer include a report on what I ate. When I used to do that, some people interpreted such reports as a prescription. They assumed I was telling them everyone should eat what I eat, and in consequence I received some hateful feedback about the harm I was doing to innocent people by "telling" them to eat things which would drive their blood sugar into the stratosphere. I hadn't been aware that admitting you ate a bowl of oatmeal for breakfast one day was tantamount to instructing all people with diabetes to eat a bowl of oatmeal for breakfast every day, but apparently it can be interpreted that way by someone who wants to start a feud with you on the subject.

For whatever reason, people really like to get angry about food. (Wheat toast? Them's fightin' words in this town, buckaroo!) So, I don't want to share my daily menus with the world anymore. Hoever, if I were doing so, I would currently be reporting more fat and less carbohydrate than formerly (I haven't had a bowl of oatmeal in rather a long time). But, as I've already admitted, this approach has been far more successful in terms of glucose control than in terms of weight control. So I don't promise to stay with this approach. The experiment continues...

Tomorrow there's a big Christmas party at work, and I'm playing a lot of music at that party (for the first hour I'll be playing Christmas carols with six other employees, and later on I'll be playing a bunch of Irish tunes with a coworker who plays mandolin and tin whistle).

To rehearse the tunes for these performances, I've had to use my lunch hour, which means that I've had to sacrifice my lunchtime run every day this week, and work out in the evening instead. I don't like working out in the gym half as well as running outdoors, and tonight I just couldn't face the gym, so I dressed warm, put on a reflective vest and a headlamp, and headed out into the cold and spooky darkness to do a 4-mile run in my rather poorly-lit neighborhood. No harm came of it, and I was glad I did that rather than going to the gym.

It's hard for me to start a run under those circumstances, because I live at the end of a very dark, woodsy private road, and even with a headlamp on it feels really creepy to go running down that unlit path before I get to the main road. But like most things that feel spooky, it also feels exciting, and it's a nice sensation when you come back from a run on a cold winter night, and have a hot shower, and put on a fuzzy bathrobe, and pour yourself a glass of red wine, and sit down to write a blog confessing how weird and irrational your relationship with food is. None of this would feel half as comfortable if you didn't go do a spooky run in the dark first.

Some words of wisdom from SMBC:

Tuesday, December 14, 2010

  • Fasting Glucose: 85.
  • Glucose 1 hour after dinner: 132.
  • Weight:  180.
  • Blood pressure, resting pulse: 121/72, 50.
  • Exercise:  gym workout (weight-lifting and elliptical machine).

I've been reading with great interest about William Hamman, the United Airlines pilot from Milwaukee who has been grounded because it was discovered that he doesn't have a medical degree. Perhaps you're thinking that airline pilots are not required to have medical degrees. True, but most airline pilots don't take a side job at a hospital by pretending to be a cardiologist. 

Apparently, when you're an airline pilot and you're caught telling outrageous lies about your qualifications, it raises questions about your character and judgment -- even though the particular lies you were telling don't relate to aviation.

The FAA says he earned his wings fair and square. Where he comes up short is in the area of medical qualifications. He did attend medical school briefly, but never completed any degree in the field, nor did he have any clinical experience as a cardiologist. His background in medicine was pure invention.

You might well ask how an airline pilot could get away with lying (for 20 years!) about having a medical degree. Well, here's where we get into a bit of a gray area. Hamman does not seem to have taken care of any patients. His primary role in the medical field was that of giving training seminars to physicians, and the subjects of these seminars were things which he knew about from his aviation experience: using simulators for training, and fostering emergency teamwork among doctors (on the model of emergency teamwork among flight crews).

The interesting thing about Hamman's fraud was that it wasn't necessary. He seemingly could have done exactly what he was doing by presenting himself to physicians as a pilot whose expertise in simulation and crisis-management had great relevance in their field as well. Indeed, when the scandal broke, Hamman sought permission from the American Medical Association to go ahead with an already-scheduled training seminar. The AMA initially agreed to this, provided that he was identified on the program as "Captain Hamman" rather than "Doctor Hamman" (but then they changed their minds when it began to dawn on them just how bad this story looked).

For whatever reason, it wasn't enough for Hamman to say "I'm not a doctor, but I've learned some things which you doctors may find very useful". He wanted to be a doctor, or at least to be regarded as a doctor. Perhaps he thought doctors wouldn't attend his seminars if they didn't think he was one of their own. Or perhaps he thought his apparent double career (a cardiologist who flies for United!) would make him seem a unique and fascinating personality. Of course, having a double career was a convenient way to manage his deception. He had told reporters in the past that he "couldn't handle a full-time cardiology practice" along with the demands of being a pilot. So, he was free to pursue his career as an educator and a researcher at William Beaumont Hospital, without actually taking care of patients (which, of course, might have created an awkward situation or two).

The Hamman scandal is causing a great deal of embarrassment to doctors who had worked with him and failed to notice that he was a fake. I can see how that could happen, though -- if he wasn't involved in actual patient care, and was careful to talk only about the things he wanted to talk about, such a deception could be prolonged fairly easily for quite a long time. What finally brought down the curtain was that Hamman became too ambitious. He applied for a research grant, and it turned out that the grant review process involved a check on his medical qualifications. I guess he had been assuming that, if the hospital hadn't done a background check after all these years, they were never going to.

I feel as if I'm more or less in the same situation as Hamman -- or at least in the situation he should have been in. I have no qualifications to give medical advice to anyone, and yet, here I am, handing out what ammounts to medical advice. I don't feel it's wrong for me to do this. After all, I'm sharing practical information based on my own experience, in case it is useful to others -- which I'm sure is what Hamman thought he was doing. The only difference is that I state explicitly, on my home page, that I'm not a doctor and that people should seek their doctor's advice on whether or not my advice makes sense in their particular case. Hamman took a different approach: he lied, and said he was a doctor.

You could argue that there's no need for either of us to be sharing advice or information with others. Shouldn't people just forget about us, and consult real doctors? Well, not so fast! How much time do most people get with a real doctor? Real doctors don't have the time to go into all the issues that people need to learn about. And how many real doctors can speak from personal experience? Maybe there are gaps that need to be filled by people who aren't doctors.

I think people (certainly people with diabetes) need other resources of health information besides whatever they can glean from a brief conversation with a primary care physician which occurs infrequently and cannot easily be repeated. So I think that it's fine for me (and for other people on the web who write about diabetes) to be sharing their experiences and opinions. I even think it's fine that their experiences and opinions may conflict sharply with mine. Readers are free to take in all the information that's out there, and decide what information seems to be applicable to their own experience and what information seems to be inapplicable.

Pretending to be a doctor, though -- I can't get behind that one.

Monday, December 13, 2010

  • Fasting Glucose: 79.
  • Glucose 2 hours after lunch: 100.
  • Weight:  180.
  • Blood pressure, resting pulse: 117/76, 54.
  • Exercise:  4-mile treadmill run; yoga class.

Playing Christmas music over the past week or two has brought back a few odd memories. One of them concerns those two mysterious (and apparently long-vanished) kingdoms of the Near East, known as Ori and Tar.

As a child I wondered how it could have come about that there were three kings of Ori and Tar, not two. How did they work that out? My understanding of royal dynasties was not very deep, but I was pretty sure that one king per country was the normal ratio, and that trouble tended to erupt whenever more than one person at a time was trying to claim the throne.

The song never really explained how the three kings worked out their shared kingship. Maybe Ori had two kings and Tar had one, or the other way around. Or did they all take turns being king of each country, with one of the three on vacation when it wasn't his turn to sit on either throne? I ultimately decided that all three kings were somehow kings of both Ori and Tar, at the same time. I didn't know how they made such an odd system work, but presumably they weren't called "wise men" for nothing.

The strange thing is that I liked the song better for the mysteriousness of the details. Not only did it mention two countries that I never heard mentioned in any other context, and couldn't find on modern maps, it also left it up to me to figure out the political implications of the peculiar two-country, three-monarch system mentioned in it. Also, I every much I liked the exotic sound of the place-names "Ori" and "Tar", which perfectly reflected the minor-key darkness of the melody.

I'm reporting all this mainly for whatever humor value it may have -- but it probably also says something about me. Any child who would expend that much mental effort trying to figure out one puzzling line from a song was probably a child destined to grow up to to try to figure out code-division multiple-access cell phone standards -- and then move on to figuring out something really complicated, such as diabetes.

Last week, when I was first seriously tempted to train for the Napa Marathon in March, I decided that if I was going to do that, I needed to start my marathon training program over the weekend -- which meant doing a 12-mile run. I wasn't sure I was up for that, but that was what the training schedule would call for.

So did I rise to the occasion? Well, not exactly. I ran 15 miles over the weekend, but it was a busy weekend and I didn't have time for a really long run on either day. So I ended up doing 6 miles on Saturday and 9 miles on Sunday. It's not really the same thing as doing 12 miles in one go. However, I think I could easily have done 12 miles if I'd had more time. When I ended the 9-miler on Sunday, I wasn't feeling exhausted or sore, I was just feeling eager to get home and take a shower so that I could show up on time for an event I was going to later in the day.

At least it was sunny. On Saturday, anyway. The trails may have been a bit wet and muddy, but the air was fresh and clear.

As always, I was sharing the trails with a lot of mountain-bikers. This gives me an odd set of mixed feelings. To some extent, I envy them. Mountain-biking always looks cooler and more fun than trail-running.

On the other hand, my own (rather limited) experience of mountain-biking tells me that it riding over rough trails gives you a lot of shock-trauma to your wrists (which has consequences, if you're a fiddler), and also gives you a lot of shock-trauma in other regions of the body (which has consequences, if you have a Y chromosome). So maybe it's just as well that I'm not doing anything more dangerous than running on those trails. Not that you can't hurt yourself running (I've taken my share of falls), but at least there's a limit to how fast I'm going to be going when I hit the ground.

My schedule is strange at work this week, and I probably won't be able to go running at lunchtime before Friday. Oh well -- that's what the gym is for.

One of the consequences of my strange schedule this week was that I couldn't do a 1-hour post-prandial test, as I usually do. So I made it a 2-hour test. I'm sure I would have tested higher at the 1-hour point, but if I was down to 100 at the 2-hour point, it probably wasn't too bad after 1 hour. That's my speculation and I'm sticking to it.

Friday, December 10, 2010

  • Fasting Glucose: 83.
  • Glucose 1 hour after lunch: 104.
  • Weight:  180.
  • Blood pressure, resting pulse: 111/68, 52.
  • Exercise: 4.6 mile run at lunchtime; weight-training workout in the evening.

If the report from my internet service provider is to be believed, someone recently found this site by entering the following text into a search engine:

pregnant diabetic urine ants underwear

I sure would like to know what could have motivated anyone to enter that particular search string. There almost has to be an interesting story there, and now I'll probably never get to hear it.

Of course, you could argue that I shouldn't want to hear from the sort of person who would do a web search on "pregnant diabetic urine ants underwear". I admit I would be nervous about taking their phone calls. But e-mail? That much I could probably handle. I'm guessing I won't hear from them, though. Whatever they were looking for, I probably failed to provide it, so they moved on.

Another search string which brought someone to my site was "redhead trailer glistening spamku", and I'm slightly ashamed to say that I actually understand that one. Maybe it's best if I don't say anything further about it at this point.

Mother nature played favorites today. When we went running at lunch, the rain held off until the very end. But my running buddies finished ahead of me, and they were safely back under the overhanging roof a couple of minutes before I caught up to them there. And as soon as they were under shelter, it really started coming down, so that I was the only one who finished the run dripping wet. I guess nature was telling me I should be trying harder to keep up. I do try to keep up, honest I do. That's the part I can never get nature to understand.

I took a bit of a risk at lunch. The only thing in the company cafeteria that looked good to me was the baked fish, but when I asked for that they served it to me on top of a big mound of rice (they had run out of the vegetables that were supposed to be the rice alternative). I tend to avoid rice most of the time, but I couldn't see how just the baked fish with nothing else was going to make a satisfying meal. So, I decided to eat some of the rice, but not too much. I think I ate about a third of the rice they put on my plate. I had just done a good run, so I thought I could get away with it. It turned out I could: only 104 an hour after the meal!

But I don't think I'm going to make eating rice a regular practice. The fact that you get away with something one day doesn't mean you'll get away with it the next time you try. If I try eating a comparable amount of rice tomorrow, I may get a much higher result afterwards. These things are never entirely predictable.

I had some iskiate (chia seeds in water) before the run; I wonder if that was part of the reason for the low result after lunch. There's a lot of soluable fiber in chia seeds, and many people report that it has a moderating effect post-prandial glucose.

Here's an cartoon which is probably not intended to be anything more than surreal black comedy...

...and yet, in its own strange way, I think it is raising an issue which a lot of diabetes patients think about (silently, no doubt) at this time of year: "Is it possible that my relatives and friends want me to be diabetic, or want my diabetes to get worse? And if so, why?".

There is something weird about the group psychology of the holiday season which causes people to encourage overindulgence in others. It is the time of year we not only tease people into having dessert, we tease them into trying each of several desserts. And we don't cut them any slack just because they come up with some feeble excuse about being on a diet or having diabetes.

Probably the actual explanation for this is simply that we feel an urge to create an anything-goes atmosphere at holiday parties, so that we ourselves benefit from the same permission to overindulge which we are aggressively granting to others. I don't think most people are actually acting with hostile intent (conscious or unconscious) towards the diabetes patients that they interact with at this time of year. But we should keep in mind that a lot of diabetes patients, especially those who are new at this game, feel cornered and even tormented whenever someone is coming at them with a plate of fudge and not taking no for an answer. Let's all resolve to back off a little quicker next time.

When someone says "no, thank you", it is supposed to signal the end of a discussion rather than the beginning of one. If you didn't learn that lesson in the workshop on sexual-harassment awareness at the office, please learn it by Christmas!

Thursday, December 9, 2010

  • Fasting Glucose: 79.
  • Glucose 1 hour after lunch: 117.
  • Weight:  180.
  • Blood pressure, resting pulse: 126/78, 51.
  • Exercise: Yoga class only.

I happen to work in the test and measurement industry, and I can assure you that making a measurement is not nearly as straightforward an operation as we like to think it is. The type of measurement I'm involved with has to do with testing cell phones, but roughly similar problems arise whenever you measure anything at all.

Take, for example, a type of measurement I'm sure you make on a regular basis, even though you may do it without a great deal of enthusiasm: measuring your body weight. How do you do that? "Simple!" you're thinking. "Just step on the bathroom scale, look at the number, get depressed, and then try to forget the whole thing". Okay, but what exactly is happening when you step on that scale, and how does that give us meaningful information about our body weight?

Presumably, when you step on the scale, your weight applies pressure to some kind of spring or strain gauge or pressure sensor concealed within the device. Somehow or other, the degree of pressure which your body weight is applying to the surface of the scale gets translated into a number. A needle on a dial spins, and points to "193". Or, a 165-millivolt output from a pressure sensor gets translated into a digital value, and a microchip consults its little lookup table and says "Okay, 165 millivolts is equivalent to 193 pounds", and it displays that number on the digital readout.

So, for every value of actual weight that you can put on the scale, a result is displayed. And what is the relationship (theoretically, at least) between the actual weight and the measured weight? A nice, straight line, of course!


The scale is, of course, accurate, and it's equally accurate at all weights. A 140-pound person steps on the scale and it says "140". A 260-pound person steps on the scale and it says "260".

Yeah, right.

Now let us consider the real world, where there is never going to be a perfect 1-to-1 relationship between actual weight and measured weight, because of the peculiar characteristics of the pressure sensor. Instead of a straight line, you've got some kind of irregular curve:

The scale is more accurate at some weights than others. And perhaps I'm being too generous when I draw a curve which intersects three times with the line representing theoretical perfect measurement. Maybe your scale never intersects with that line at all; at every possible weight you can measure, it's off by at least a pound. (Or five.)

And let's take another factor into account: temperature. Most measurement equipment is affected by temperature changes, after all. Perhaps the curve for this scale is different in the winter than it is in the summer.

Problems of this sort arise whenever you make any kind of measurement. High-precision measurement is a huge challenge. You have to do a lot of careful calibration and error-correction to eliminate known sources of measurement error -- and no matter how hard you work at that, you can never get it perfect, because there are always tiny sources of error which we don't know enough about to be able to correct.

A lot of diabetes patients feel shocked and even angry when they find out how loose the standards of accuracy are for home glucose meters. Why don't they make them more accurate than that, we wonder? (Probably because measuring blood glucose concentration quickly, accurately, and affordably is not half as easy as diabetes patients imagine it to be.)

But even diabetes patients who are willing to accept that home test kits are inevitably going to have accuracy problems nevertheless expect lab results to be absolutely correct. So why is there still so much uncertainty hovering over the Hemoglobin A1c test?

Well, for starters, let's acknowledge that measuring the percentage of glycated hemoglobin is a daunting technical challenge. Unless you had some kind of magical microscope with which you could go into someone's red blood cells and sort hemoglobin molecules into two piles (glycated ones here, clean ones there), how exactly would you go about measuring this? Hemoglobin is only one constituent among many in the blood, and glycated hemoglobin is not the only kind of hemoglobin. Separating out one element from a complex chemical mixture and measuring it is not so easy!

Two basic methods have been used to measure glycated hemoglobin: chromatography (in which the chemical components of a complex mixture such as blood are filtered through a medium which spreads them out into a kind of spectrum), and immunoassay (in which antibodies that are highly specific to a particular compound are allowed to bind to it in order to find out how much binding actually happens). But there are various techniques, and types of equipment, which a lab might use in order to implement either of these approaches to measuring glycated hemoglobin.

For a long time after A1c testing begin, in the late 1970s, there was a chaotic situation in which every lab seemed to use a different process, and results could not be meaningfully compared because there was too much variation between one process and another. Clearly, something needed to be done to regularize all this, and the result was the National Glycohemoglobin Standardization Program. Supposedly, all you need to know is that your lab complies with the standards of that organization (and these days, about three-fourths of labs do comply).

Does this mean NGSP-compliant labs all do their A1c testing the same way, using the same equipment and the same procedures? Unfortunately, no it doesn't. The fact that two methods of testing A1c both meet NGSP accuracy standards does not mean that those two methods yield exactly the same results; it may only mean that the differences between them are not large enough for NGSP to consider them clinically problematic.

This has been a source of confusion for me, because the medical lab I go to seems to change its method of A1c testing on a regular basis. Every time I get a result back from them, their quoted reference range shifts up or down a few points. This makes it hard for me to interpret small changes in the results from one test to the next. If the number goes down a little, am I really doing better? If it goes up a little, am I really doing worse? Or is the difference merely an artifact of changes to the way they measure A1c? My test in September showed what looked like a disturbing uptick from 5.3 to 5.6, which is hard to reconcile with my mostly low glucose test results. Was this for real? According to their reference range, 5.6 is still considered a normal value, but it's not as good as 5.3, and I worried that it meant things were headed in the wrong direction. Or did it simply meant that their current test method reads higher than the last one did?

I checked their web site to see what method they use for their A1c tests currently. It turns out that they now use what they call a Roche Tina-quant assay. A web search on that peculiar name led me to an article in Clinical Chemistry which pointed out a minor flaw in this type of test: at levels below 6%, this test reports the results as somewhat higher than they actually are.

However, the authors of the article offered the opinion that it doesn't really matter, because the error isn't large enough to make normal people appear diabetic. So long as the patient is in the normal range, what does he care whether he's reading a few points higher than his actual value? Well, obviously I care. But the authors are pretty confident in saying that the error is not clinically significant.

Which means that the definition of "clinically significant" is "someone besides Tom cares about this".

Wednesday, December 8, 2010

  • Fasting Glucose: 80.
  • Glucose 1 hour after dinner: 118.
  • Weight:  179.
  • Blood pressure, resting pulse: 120/71, 54.
  • Exercise: 4.1 mile run.

Not quite so nice a day for running, this time. My running buddies don't like running in the rain, so they opted to work out at the gym tonight instead. That wasn't going to be practical for me, as I had something else to do after work, so I resigned myself to a run in the rain.

It really wasn't that bad, to be honest. It wasn't raining hard. But you feel kind of lonely out there when you're exercising outdoors and no on else is. (Well, I did see one other guy running, but runners and walkers and cyclists are usually abundant in this neighborhood, and when they disappear it feels sad to be there -- as if you're the one person who didn't know the costume-party was canceled.)

Tomorrow the weather is supposed to be better, but I can't run at lunchtime because we're using the lunch hour tomorrow for another Christmas-music rehearsal at the office. Well, I'll have to hit the gym in the evening, then. My life is all about exercise fallback plans. The trick is to act on your fallback plans, instead of giving up on your daily workout as soon some inconvenient development gets in the way of it. I think that's how people get out of the exercise habit. Once you've decided that excuses are acceptable, you begin to realize that there's an excuse almost every day.

Once you've decided that excuses are not acceptable, you begin to realize that there is room for exercise in almost every day, if we try hard enough to fit it in. We fit meals in, after all -- usually three of them!

One of the special difficulties faced by recently-diagnosed diabetes patients is that they bring some very false expectations to the problem of diabetes management:

  • It should be a simple, straightforward task.
  • It shouldn't be difficult to accomplish.
  • All the experts should agree about how to do it properly.

That is how they think it ought to be: no difficulties, no controversies, and no mysteries. There should be one simple, foolproof approach to managing diabetes, which is recommended by everyone and works for everyone.

Boy do they get mad, when they find out that's not how it is! They don't just feel disappointed or frustrated -- they feel cheated. They did what the doctor said to do (some of it, anyway) and their blood sugar didn't drop down to normal! And now they're hearing advice from other experts which doesn't seem to agree with what the doctor said! How can they possibly be expected to figure all this out? It's a rigged game! They can't win! They might as well just give up here and now!

I wonder how outraged they become when they discover that other tasks in life are complicated, difficult, or controversial. Do they feel equally cheated when they find out that there aren't isn't a simple, straightforward, uncontroversial approach to planning a career, raising children, or deciding how to vote? When it turns out that learning to play the guitar isn't as easy as they had expected it to be, do they exclaim furiously that it should be easy?

Well, unfortunately, diabetes is not the simple, straightforward, controvery-free topic we might like it to be. It's complicated, it's hard to understand, and people disagree about it (even people who have been to medical school, amazingly enough).

If scientists had been raising diabetes patients in cages for the last 200 years, we would know more about what causes the disease, what alleviates it, how much it varies from patient to patient, and what is the best method (or methods) of treating it. But scientists have not been doing that, so the situation is a bit more difficult to figure out. People are sifting through the imperfect evidence we have, and reaching their own conclusions about what it means. It shouldn't be too surprising that they don't always reach the same conclusions, or that what they recommend usually turns out to work better for some people than for others.

Yes, managing diabetes is a complicated and difficult task, and intelligent people of good will can disagree about it. But the same is true of other tasks in life which we don't give up on as soon as we find out that they're not easy.

Tuesday, December 7, 2010

  • Fasting Glucose: 81.
  • Glucose 1 hour after lunch: 107.
  • Weight:  180.
  • Blood pressure, resting pulse: 110/73, 53.
  • Exercise: 4.1 mile run.

Not a bad day for a run!

In fact, it was so nice a day that I decided to take my camera with me, and capture some images of our 4-mile loop around the neighborhood.

In much of the country that would look like a summertime picture, but it takes rain to make things green, and we don't get that in the summer here (we're a desert then). If you want to see a lush landscape in northern California, you'd best go there in the winter and wait for a sunny day.

But December is many seasons in one. We also got to kick up some autumn leaves...

...and we even saw some fall colors.

Here are my running buddies, pulling ahead of me on Stagecoach Road to charge up our most dreaded hill. (If stagecoaches ever really did use that road, I pity the horses.)

After the run we took some pictures of each other to see if the camera can capture the effect of endorphins.

Well, sort of.

We've been running together at lunchtime for about six years now. Perhaps you're wondering why we aren't sick of it yet. I'm not sure I can explain why we're not, but we're not. I think we're so used to it -- and so dependent on it, really -- that by now it would feel very strange and even disturbing not to do it. Running regularly has been helpful to us; it's seen us through stressful times.

Another runner at work just had gall bladder surgery three weeks ago, and she's already running again. In fact, she did a ten-miler on the weekend just past. She's training for the Napa marathon in March. I suppose most people, hearing her talk about all that, would say to themselves "She's out of her mind!", but I was impressed rather than horrified. In fact, the thought that came to mind was "Maybe I should do the Napa marathon this year, too!".

I had thought I was done with marathoning, and maybe I am, but I haven't done one in a couple of years, and I'm starting to feel the itch to give it one more try. And why? Because someone I know has already registered for it, and someone else I know is considering it.

I guess this is why I need exercise buddies. I can, and do, exercise alone, but when you're in regular contact with other people who work out a lot, it inspires you, energizes you, and reinforces your willpower. Most people who don't exercise regularly also don't have any regular contact with the sort of people who, not long after abdominal surgery, are impatient to get back to doing 10-mile runs. I do have regular contact with such people, and it makes exercise easier -- simply because it makes exercise normal.

I sometimes think that the most important thing newly diagnosed diabetes need to do is to find someone to work out with. Staying on track with exercise when you're always doing it alone is mighty difficult; staying on track when somebody is expecting to meet you at the trailhead on Saturday morning comes a little more naturally.

I think exercise buddies are crucial to the whole enterprise of diabetes management, so if you haven't got any, do what you can to find some. Finding compatible exercise partners is almost as tricky as finding compatible musical partners, so you have to invest some effort in it. But it's worth it!

Monday, December 6, 2010

  • Fasting Glucose: 88.
  • Glucose 90 minutes after lunch: 92.
  • Weight:  180.
  • Blood pressure, resting pulse: 120/78, 50.
  • Exercise: 5.7 mile run.

My post-prandial test result of 92 today may seem improbably low, but I expected it to be low, for three reasons. First, it was a pretty low-carb lunch. Second, I had a very hard run before I ate. Third, I tested at 90 minutes rather than 60. Because I peak around 60 minutes after a meal, a test at 90 minutes can be expected to give me a lower result (this isn't true for everyone -- you have to experiment to find out if it's true for you). 

The reason I made it 90 minutes today rather than 60 is that I was having a meeting with my boss which ended up lasting lasting considerably longer than planned. I didn't feel I could interrupt him to say "come on, let's wrap this up -- I have to go back to my cubicle in Building 4 and test my blood". 

Perhaps you're wondering why I didn't feel I could do that. I guess it's an issue worth discussing.

I could dispose of the issue simply enough, by pointing out that my boss was giving me good news rather than bad -- and when your boss is saying nice things to you, that is no time to be discouraging him. But to explain it that way would be dodging the question rather than answering it, as I wouldn't have felt able to interrupt the meeting no matter how it was going. So what's going on here?

Unlike many people, I never tried to keep my diabetes diagnosis a secret at work. I guess I figured they'd find out about it somehow, whether I told them or not, and it would look better if they found it out from me. As John LeCarre once said (although he was talking about intelligence officers), "put it on the record before the record puts it on you".

But if I don't make my diabetes a secret at work, neither do I allow it to seem like a problem at work. I don't want anyone to be thinking aloud such thoughts as "Well, we really ought to send Tom to Scotland for the project handover next month. But can he handle that? With his diabetes, I mean?". I don't want them to think of me as someone who is so limited by a disease that he can't do things other employees routinely do. In principle, it's reasonable to expect companies to accommodate the personal problems of their employees, but as a practical matter, we all know it's a lot better for your career if the company does not need to make accommodations which are inconvenient to the usual operations of the business.

There is really only one diabetes accommodation which I want the company to grant me: enough flexibility in my schedule to make daily exercise a practical possibility, on most days if not on all days. Fortunately, this is not a difficult thing to negotiate where I work. Lots of people there work out at lunchtime. They run, they go for bike rides, they play soccer, they participate in exercise classes, they lift weights. My boss and his boss are both very athletic, and they work out at lunchtime on a regular basis. One of my running buddies at work used to be my boss (he talked me into running my first marathon, in 2005, and training for it with him). The company actually encourages this sort of thing -- largely because they are trying to drive down their health-insurance costs, but also because they have noticed that employees who stay fit also tend to stay productive. The work site has locker rooms, showers, ball fields, exercise equipment, and even exercise classes. Of course, when meeting schedules make things difficult, you sometimes have to work out at a different time than you had planned, or do a shorter workout than you had planned, or skip it altogether and go the gym in the evening. But most of the time I can manage to fit a workout in during the lunch break. And I realize what a big advantage that gives me over the average working person, who goes crazy trying to find room in his daily schedule to work out.

Because I have that advantage, I don't want to push my luck by asking for any other advantages. I don't want my coworkers to feel that they have to make any extra "room" for me to deal with diabetes. My diabetes is virtually invisible at work, even though it's not a secret at work. Strictly speaking, they know if have it. Practically speaking, they can't remember that I have it.

I guess I want to keep it that way.

Did you know that the word "gullible" isn't in the dictionary?

Well, actually, it is in the dictionary, but telling someone that it isn't makes a good prank. When they come back to you with a dictionary to prove you wrong, you get to laugh at them for having taken the claim seriously enough to check it out. What makes it a good prank is that they can't win: they're gullible if they accept your claim, but they're also gullible (in a different way) if they make an effort to challenge it.

Like most jokes, this one has a very serious issue at its core: people who try to be skeptical never know if they're being skeptical enough, or if they're being skeptical about the right thing.

Pranksters, magicians, and even joke-tellers all mislead us temporarily in order to show us how easily we are misled. If this seems a cruel lesson for them to want to teach us, bear in mind that the world is full of people who mislead us without revealing that they are doing so, without teaching us anything, and without having any goal other than exploitation in mind. Of all the people who take an interest in human gullibility, the ones who expose it are much less of a threat than the ones who quietly take advantage of it. So it seems to me, anyway. Some people see it very differently.

A lot of people take a kill-the-messenger approach to this issue: if a skeptic points out that a claim is not credible, they resent the skeptic, not the claimant. It's just mean, they think, to raise doubts about something that people would like to take as true. They see skeptics as killjoys, out to spoil everything that's magical and beautiful in life. Why can't they just keep quiet, and leave well enough alone?

The trouble with the view that skeptics should leave well enough alone is that the situation is not always "well enough" to be left alone. Sometimes, admittedly, it doesn't make much practical difference whether you believe a bogus claim or not. (The people who believe in mental telepathy, seances, and alien abductions seem to go about the business of their lives pretty much the same way anyone else does, so what does it matter if they are wrong about these things?) However, when people start using a bogus claim in making decisions about how to spend their money or how to take care of their health, you can no longer argue that it doesn't matter if they're being deceived.

Deception does sometimes have legitimate uses. If modern medicine hasn't found any useful remedy for your back pain, you might as well get what value you can out of the placebo effect. Go for whatever herbal remedy someone has done a good job of selling to you, and if you believe in that remedy enough, the mysterious brain/body interface may actually bring about some changes which will make you feel better. If that happens, it hardly matters that no one has been able to demonstrate that the remedy ever does better than a placebo. The placebo effect is preferable to no effect.

However, Il don't think people suffering from incurable medical problems should be charged exorbitantly for the placebo effect, and I don't think there is anything "mean" about exposing such an exploitive enterprise for what it is.

A particularly good example of such exposure is this one, from Randall Munroe of (a master of squeezing a lot of complex thought into a compact and accessible presentation). Here he evaluates a collection of "crazy phenomena" in economic terms:

If there's a good counter-argument to this, I haven't yet heard it!

Friday, December 3, 2010

  • Fasting Glucose: 79.
  • Glucose 1 hour after lunch: 99.
  • Weight:  178.
  • Blood pressure, resting pulse: 121/79, 49.
  • Exercise: 5.2 mile run.

A much lower fasting  glucose result today, despite my having taking a rest day from exercise the day before. Sometimes you just can't predict these things.

But I did run today, and it was a pretty hard run, with loads of hill-climbing. Maybe that's why my post-prandial test after lunch was only 99 today, even though lunch included a small piece of fruit this time. Yesterday, when I didn't run, a lower-carb lunch yielded a result of 127. But was it just the exercise, or was something else contributing to the difference?

The wild card in my life in recent weeks has been a cold virus, or a series of them, coming and going in an annoyingly persistent series. I've never been terribly sick during any of this, and it hasn't prevented me from working (or working out), but even a low-level viral infection affects your insulin sensitivity. Anyway, I'm feeling much better -- much freer of the virus -- today than I have felt in a while. Maybe I'm finally virus-free and that's why my glucose numbers are down.

I live near the top of a steep 900-foot hill, which means that every time I drive anywhere at all, I find myself plunging down a kind of paved ski-slope. This is a little hard on your brakes, as you can imagine. When I took my car in to the dealer for routine maintenance today, I fully expected that they would tell me it was time for brake-pad replacements. In fact, I was foolish enough to mention the possibility when I dropped off my car this morning, and ask them how much it would cost if it turned out to be necessary. I even authorized them in advance to go ahead with it if it was called for. As you can see, I am the very definition of a bad consumer; I might as well have said "If you're trying to make quota for the month, and you'd like to cheat someone to improve your numbers, I'm here to help".

In the event, when they called me with their findings later in the day, they said the brake pads still had considerable life left in them, so it wasn't necessary to replace them yet. A happy ending! We need to savor those when they occur.

My car is a Honda Civic, and they have a system which gives you dashboard messages to tell you when it's time to come in for routine maintenance, and what kind of maintenance is needed. The result is that I'm taking better care of this car than I ever have of other cars I've owned.

The strange thing is that I tend to think of all this as medical rather than mechanical. I'm not sure I can express my feelings about this any clearer, so maybe I should just leave it at that.

But I do wish we all had little messages that came up on our skin from time to time, and told us what needs checking out. Uh-oh -- the little code that's flashing on my wrist says that it's time somebody had a look at my... I mean someone with a medical degree, of course.

Can you tell which of these is Michelangelo's actual sculpture of David, and which one is the fake?

You guessed it! The fake is the one on the right. Michelangelo's model was on the slender side, apparently. I don't know if his model was actually typical of the population of Florence in 1504, though. Maybe the fake on the right would have been more like the average person Michelangelo would have seen in the street. I tend to doubt that, but I can't prove it. My impression is that the figure on the right is more typical today than would have been the case in previous generations. Some people deny this, though. We only think people are fatter these days!

I don't know who did the fake David above; it was used without attribution on the cover of a scholarly review article which, on the whole, seemed to be sympathetic to Paul Campos's claims that obesity is a "fake problem" -- that people aren't really getting fatter, or if they are it doesn't make any difference to their health.

Campos is the author of The Obesity Myth. I haven't been very impressed by Campos's argumentative style, from what I've seen of it. He seems to be one of those people who wants to have it both ways, rejecting data in one context and then using the very same data to support his points in another context. To me his approach seems opportunistic and lacking in candor.

For example, he decries the Body Mass Index as a useless indicator of obesity, since it often classifies very muscular people as obese. Then he turns around and uses the BMI when it suits him, as a way of claiming that obesity isn't linked to health problems. Well, which is it? Either BMI really does track obesity, in which case it can be used to evaluate the health impact of obesity, or it doesn't track obesity, in which it can't be used to evaluate the health impact of obesity. Don't tell me that BMI is meaningless when it can't be used to support your views, and meaningful when it can.

The thing is, even if the CDC is wrong to use BMI in measuring the obesity rate in America, there still has to be some reason why the obesity rate, using that measure, has been going up in recent years. So why is it going up?

If the rate isn't going up because people are gaining body fat, then it must be going up because people have been doing a lot more body-building lately -- especially in the deep south. Not having been to Mississippi, I don't know whether or not a body-building craze is under way there. If it isn't, though, it may well be that people are getting fatter there, not stronger.

But even if people are getting fatter, Campos would argue, it doesn't really matter, because obesity doesn't really have a harmful effect on health. He also claims that there would still be no point in urging people to lose weight, even if it did improve their health, because he thinks people can't do it.

These claims sound to me like dispatches from another planet. To Campos tell it, I'm only imaging that I lost 70 pounds (because that can't be done), and I'm only imagining that doing so helped me get my blood sugar under control (because losing weight doesn't improve your health).

Okay -- it was all a dream. I can live with that. Sometimes I would really like to.

Thursday, December 2, 2010

  • Fasting Glucose: 93.
  • Glucose 1 hour after lunch: 127.
  • Weight:  178.
  • Blood pressure, resting pulse: 118/73, 59.
  • Exercise: Yoga-class only.

It occurred to me that I haven't had a day off from exercise in slightly over a week, so when I was unable to run at lunchtime, I decided that maybe that was okay. I went to my yoga class in the evening, but that was it (and yoga's not really exercise), so this counts as my rest day.

The reason I wasn't able to run at lunchtime is that I needed to devote my lunch break to a different purpose: our first rehearsal for the office Christmas music. It's a tradition we've been carrying on for the past 15 years: on the last day before the Christmas break, a group of us stroll around from building to building at the work site, playing three or four Christmas tunes in each location before pushing on to the next one. Here is part of our ensemble from last year (a sad year, as one of our key players had lost her job in a layoff and couldn't be with us): 

Christmas music presents a special challenge to musicians: for eleven months of the year, you never play this stuff -- so, no matter how familiar the tunes may seem, you're always rusty on them when you take them up again in December. Hence the urgent need to hold rehearsals. If we were going to be playing Irish reels, which I play all year long, it wouldn't occur to me to do anything other than wing it. That is not the way I feel about Santa Claus Is Coming To Town. Also, we have to work out the arrangements collectively each year, based on who is participating (this year we'll have two violins, two guitars, a flute, and a French horn).

I have to say that today's rehearsal went exceptionally well, though. A new member of our group, playing with us for the first time today, blended in perfectly. The music sounded really good, especially for a first rehearsal. The music even had some nostalgic power for me, which was a relief. I was beginning to think I'd become immune to the Christmas mood -- and when that happens it is probably a sign that you're getting old and boring, or that your liver is failing, or something like that.

A recent study, reported in the November 24 issue of the Journal of the American Medical Association, looked at the effect of exercise on Hemoglobin A1c test results in patients with Type 2 diabetes, and concluded that exercise was effective in reducing HbA1c, but only when patients practiced both aerobic exercise and resistance training. (Resistance training means strength-training, with weights or weight machines or other muscle-building exercises.) Patients who practiced only aerobic exercise, or only resistance training, saw no signficant reduction in their HbA1c results (although they benefitted in other ways -- for example, reduction of fat mass in those who practiced resistance training).

The patients who used both forms of exercise saw an average reduction in their HbA1c score of 0.3 (the maximum reduction in the study was 0.6).

When I was originally diagnosed, my doctor advised me to do both kinds of exercise, but after some years went by, I found that I seemed to be able to get good results with aerobic exercise alone. Because I liked aerobic exercise a lot better than I liked resistance training, I ditched the resistance training, and focused my exercise program on running.

In September, however, I was disappointed by an increase in my A1c test from 5.3 to 5.6, and I started looking for areas in which I could improve my routine. In October I started doing resistance training again.

It takes a while to make much progress with resistance training, but at the gym last night I found that I was able to move the weight up a notch on several of the machines I was using. I also found that I could do 16 rather than 13 of those "good-form" pushups I discussed on November 19 (that is, pushups in which you actually lower yourself far enough to touch the floor -- which seems to me a cruel requirement, but nobody asked me).

Although I've never liked resistance training, I've learned to like other things that I disliked before, because I knew that I had to do them whether I liked them or not, and it would be easier to do them if I found a way to enjoy them. Maybe I can accomplish the same thing with resistance training.

Wednesday, December 1, 2010

  • Fasting Glucose: 91.
  • Glucose 1 hour after lunch: 135.
  • Weight:  179.
  • Blood pressure, resting pulse: 123/71, 61.
  • Exercise: 5.3-mile run at lunchtime; weight-training workout in the evening.

I stand corrected! Yesterday I said it was absurd to be "angry" at experiencing a misfortune (such as diabetes) which nobody inflicted on you. How can you be angry over something if there's no one for you to be angry at? That was the way I saw it, anyway. But today The Onion revealed some new information on the subject: 

I guess I'll have to reconsider my stance on this issue!

I usually eat lunch immediately after running; today, because of a schedule problem at work, I ate lunch first, and didn't get to run until later in the afternoon.  So, would my post-prandial glucose test result be higher than usual, since it wouldn't reflect any of the benefits of my daily workout? Yes, apparently so. Perhaps 10 to 15 points higher than I would usually see from a similar lunch. 135, for a 1-hour post-prandial test, is within my doctor's guidelines, but I believe 120 is closer to a truly "normal" level, and I try to hit that target when I can. Exercise clearly makes it easier to do that.

It may be that I'm unconsciously letting myself eat more carbs at lunch, just because I know they'll have less impact right after a run than they would hours later, at dinner.

Deborah Nolan, a professor of statistics at the University of California at Berkeley, likes to begin a semester by giving her class an odd assignment. She divides them into two teams. One team will toss a coin 100 times and record the results in order. The other team will merely imagine that they are tossing a coin 100 times, and will record what they imagine the results might have been if they had used a real coin. While all this is going on, she leaves the room. The results from both teams are written out on the blackboard, but not identified. When she comes back, she will look at the results on the blackboard, and identify which of the two sets of results was produced by actual coin-tossing.

And every time she does this, she gets it right.

So how does she do it? How does she look at a random sequence of heads/tails outcomes and determine whether or not there was an actual coin involved? Well, I'll come to that shortly, but the fact that she's able to do it at all is enough to tell us that randomness is a lot more complicated than we tend to think. Or at least it's a lot different than we tend to think. We humans tend to have all sorts of false ideas about randomness, and this makes it hard for us to understand processes in which there is a random element at work.

I tried the experiment myself tonight, recording a hundred coin-toss results and also inventing a series of a hundred imaginary coin-toss results.

If it's not immediately obvious to you which series is the truly random one, I can safely say that Deborah Nolan would be able to spot it in an instant.

What does random really mean, anyway? It's usually defined as "without a plan or pattern". Mathematically, though, it has to be defined a little more strictly: a process is random if the possible outcomes are governed by the laws of probability and nothing else. Tossing a coin is a random process because the two possible outcomes are equally likely.

If, in a supposedly random process, something is causing one outcome to be likelier than another, the process is not truly random (as in that movie scene where WC Fields is dealing cards, and someone asks him "Is this a game of chance?", to which he replies "Not the way I play it"). But, if there is no cheating involved, tossing a coin really does give you random results. That being the case, how could Deborah Nolan possibly tell, just by looking at a sequence of heads/tails results, whether it is real or invented?

If it seems impossible that a genuinely random process could be distinguished from an imitation of a random process, that is because people tend to make a lot of false assumptions about what a random process is, and what random results look like. Results that are governed by the laws of probability have identifiable characteristics. Random results are recognizable as random results if you know what to look for. But most of us don't know what to look for, and in fact we're usually looking for the wrong thing.

Suppose you have just tossed a coin six times, and it came up heads each of those six times. At this point, what is the probability that, on the seventh toss, you will get heads again? "Extremely low!", you are no doubt thinking. But the actual probability is 50%. The coin, you see, isn't aware of its recent history. It doesn't know that it came up heads the last six times, and it isn't suffering from the anxious feeling that it really ought to do something different this time. It doesn't feel that it's "due" to come up tails. So the probability of heads remains 50%. But, to the human observer, this just feels wrong. We feel it would be far more appropriate for the coin to come up tails this time. And, because we feel it would be more appropriate, we conclude that it would also be more likely -- which is certainly not the case.

Because the coin doesn't feel uncomfortable about repeating the same result several times, the outcome of a series of 100 coin tosses is almost certainly going to include "streaks", in which the same result is repeated many times in a row. This bunching-up of a particular type of result is one of the normal, expected features of any random process. But, for whatever reason, people don't feel comfortable with this.

People expect a random process to be characterized by frequent alternation between possible outcomes -- not streaks. So, when we try to imagine tossing a coin 100 times, we can't bring ourselves to include the kind of streaks that would actually occur in a random sequence. We might write down heads 3 times in a row, but we're uncomfortable with 4, and 5 is out of the question -- that would just be too weird. Or so we think!

It's obvious, in the picture below, that my imagined series was the one on the left: it includes no streaks longer than 3 heads or 3 tails. Deborah Nolan's students always do the same thing with their imaginary results that I did with mine; she can easily spot the fake results by the absence of long streaks. The genuine results are another matter entirely. My actual coin tosses included 7 streaks, one of them so long as to be nearly incredible.

14 heads in a row! Believe me, I felt very uncomfortable when that happened, but it did happen, and I certainly wasn't trying to make it happen. (Deborah Nolan's students also get uncomfortable durings streaks -- it's as if they're doing something wrong, but they can't figure out what it is.) Admittedly, 14 heads in a row is quite an exceptionally long streak, but 8 to 10 is pretty common, and students making up an imaginary set of results never dare to go that far. So, the difference between the real results and the fake ones is usually glaring.

Anyone crafting a set of imaginary coin-toss results is sure to feel, when he gets done with it, that there is no pattern to the numbers he's made up. But of course there is a pattern, and it's a pretty easy pattern for a statistics professor to spot: the expected streaks are missing.

If Nolan had asked her students to draw 100 dots on a piece of paper, placing them randomly, it would be pretty easy for her to tell that the dot pattern wasn't truly random -- because she wouldn't see clusters of dots (with a few large voids in between them) on the paper. The students would tend to distribute the dots more evenly on the paper; clustering the dots would have seemed strange or artificial to them. But in random data, clustering is the rule rather than the exception.

We find it very hard to accept that random data involves clustering. In fact, we tend to get it backwards: we assume that, if we find clustering in data, it means the data isn't random. We think we've found a pattern! (To a statistician, the absence of clutering would be considered a pattern; clustering itself is what we should expect to find, unless it gets too extreme to be plausible.)

This issue starts to become pertinent to public health when there is a scare about a "cancer cluster" in your town. Multiple cases of a fairly uncommon malignancy in one fairly confined geographic region? That can't just occur randomly -- something's got to be going on here! There must be some kind of toxin in the groundwater! Well, maybe. It's worth investigating, to see if we can find any cancer-causing pollutant or other problem in the area. But it isn't true to say that it can't just occur randomly. Of course it can occur randomly, and we'd even expect it to. So let's not get too hysterical about the situation, especially if we can't find any other factor which might be causing the problem. If something could be a random, and you can't explain it any other way, it probably is random.

Of course, pure randomness is something you usually have to contrive artificially, by tossing coins or rolling dice or pulling numbers out of a hat. Most of what goes on in nature is a combination of random and non-random factors, but people tend to overlook that fact. They figure that, if any part of the picture is random, then it's all random -- it's mere chaos. This kind of mistaken assumption doesn't make it any easier to grasp scientific issues, believe me.

A seemingly educated person told me recently that he didn't see how evolution could be real, because organisms are so complicated, and "how could that happen just randomly?". But nobody ever said that it did happen just randomly -- certainly not Darwin. For heaven's sake, Darwin's claim to fame is that he figured out the part of the process that isn't random!

Gene combinations and mutations are random, all right (at least within the limited confines of how much genes can change and still remain functional), but that only means that offspring vary. What isn't random is that some of those variable offspring have better odds of surviving than others. Therefore, some of those genetic variations build up in the gene pool and some don't. It's like tossing gravel into a screen box and shaking it. The rocks vary randomly in size, and you shake the box with random motions -- but the screen lets small rocks through and not big ones, so what falls out of the box is not random at all. The "screen" that Darwin discovered is natural selection (meaning that circumstances favor the survival of some offspring over others -- which has the same practical impact as an animal breeder choosing desired traits). By now, Darwinian evolutionary theory has been confirmed so well by the DNA evidence that there's really no remaining possibility of its being refuted. But, because people are so confused or uncomfortable about any process which includes a random element, they see Darwinian evolution as "just random" and therefore preposterous.

The thing is, very little happens in this world that doesn't have some random element involved in it. You can claim that weather is "just random" (and, I suppose, that weather is therefore preposterous and can't be happening). But is weather entirely random? You can't find out what the weather is going to be like on January 5th. But does that mean that you know nothing about how the weather in January is going to compare with the weather in July? Often, the minor variations are unpredictable but the larger trends are pretty clear. 

Glucose meter readings include a certain amount of random variability. If your actual blood sugar is 115, and you test 5 times in rapid succession, your results might be 121, 110, 113, 116, and 114. Five different answers, and not one of them is right! Why doesn't the meter just say 115 every time, like it ought to?

You could conclude from this that your meter is giving you worthless data, so you might as well ignore it and stop testing. But that would be rash. If the meter's results were "just random", they'd probably be more like 255, 129, 67, 422, and 13. Instead, they have a pattern -- they hover pretty closely arround the correct value. The five results are reasonably close to 115, and they average out to 115. All things considered, I'd say the results are meaningful and useful. It would be nice if they were consistently and absolutely accurate, but as long as the errors aren't too big, it's a lot better to have the test data than not to have it.  

It's too bad that we humans have such a poor ability to understand randomness, and such an unreasonable fear of it. If we were more at ease with randomness, we'd be a lot less likely to be hoodwinked by casinos, we'd have a lot less trouble understanding science, and we'd be a lot less likely to ignore useful information on the grounds that it's "just random".

Older Posts:

November 2010
October 2010
September 2010
August 2010
July 2010
June 2010
May 2010
April 2010
March 2010
February 2010
January 2010
December 2009
November 2009
October 2009
September 2009
August 2009
July 2009
June 2009
May 2009
April 2009
March 2009
February 2009
January 2009
December 2008
November 2008
October 2008
September 2008
August 2008
July 2008
June 2008
May 2008
April 2008
March 2008
Jan/Feb 2008