Saturday, July 2, 2011

Warped Logic

In science, it's good to have an open mind, because we're always discovering new things that change the "facts" we were once so certain of.

But too many scientists get stuck in a groove, and they can't be budged from the current dogma, even when the evidence doesn't support their beliefs.

This is apparent in the eternal debate about the best diet. One problem is that people are different, they interpret diets differently, they keep track of what they're actually eating differently, they have different exercise patterns, they take different drugs, and so on and so on. But focus on "evidence-based medicine" means physicians won't believe anything unless it's been proved in a double-blinded controlled randomized trial.

And because such trials involve large groups of people, some of them usually respond one way and others respond in another way, and only statistical analysis will show whether the intervention worked on average. It says nothing about how the intervention will work on any individual patient in the future.

Nevertheless, what bothers me is the tendency of the scientists doing these trials to interpret the results in the light of their own biases. One such ploy when your intervention didn't work is to suggest that you didn't intervene hard enough.

For example, when a study of l0w-fat diets resulted in no benefit, the researchers said maybe the fat content wasn't low enough, that the study should be repeated with even less fat in the diet. It didn't seem to occur to them that perhaps lowering the fat content of the diet wasn't helpful.

A recent study shows the same type of reasoning. This study showed that adding moderate exercise (walking) to diet in people with type 2 diabetes resulted in no benefit for hemoglobin A1c. The first explanation by the lead author of the study was that "the activity chosen, walking, was suboptimal."

In other words, if the exercise you used shows no benefit, maybe more exercise will help.

A lot of studies have shown that exercise doesn't contribute to weight loss. Just Google "exercise, weight loss, doesn't help" for a smorgasbord of articles. Often, exercise just makes you hungrier. Other studies have shown that diet and exercise do work better for overall fitness than either diet or exercise alone.

Exercise helps the cardiovascular system and is certainly a good thing to get. (So why am I sitting here typing instead of finishing the wood stacking I started this morning? Answer: I'm human, just like you.)

I'm not suggesting that exercise is bad. What bothers me is the knee-jerk reaction of some science investigators. "My study doesn't support my hypothesis, so maybe the study wasn't done right" instead of "My study doesn't support my hypothesis, so maybe the hypothesis is wrong."

If everyone thought like this, we'd never make any progress.

Luckily, there are always a few brave souls who dare to defy the current dogma. They're usually laughed at when they start, and some of them give up. Some persist. And they're the ones who end up with the Nobel Prizes.






6 comments:

  1. My motto is, I'm an individual, not a statistic. As an outlier, this is REALLY important, because what works for someone else may NOT work for me. Evidence-based medicine is a great concept, but if, in reality, it reduces all of us to statistics, it does us a real disservice. In the medical world, experience should matter, too.

    ReplyDelete
  2. I agree, Natalie. One way to judge a good physician is whether he or she will treat you as the individual that you are.

    ReplyDelete
  3. If we could only all find physicians who would treat us as individuals.

    One of my former docs had an aunt with osteoporosis, so he seemed to focus more on that than take seriously my concerns about my glucose. "Maybe you're just sensitive to carbs," I was told. I guess!! I kept pushing the issue, was eventually diagnosed as pre-diabetic (high post-prandials), and started following Bernstein.

    The down side of LC/HF for me (4 years now) has been a substantial increase in lipids. After cutting out eggs and greatly reducing SF this spring, I've lost weight I couldn't afford to lose. So here we go again--how can I gain it back?? A CDE dietitian said I should eat more carbs and take more meds!!

    The stress is not helping, and I think my brain needs more ketones.

    I really enjoy your blog, Gretchen. You're down to earth.

    ReplyDelete
  4. "Maybe you're just sensitive to carbs," I was told.

    What an idiotic comment! I'm glad it's a former doc.

    I think the CDE is correct that eating more carbs would probably make you gain some weight. If you're prediabetic, you might be able to tolerate a few more slow carbs.

    But if you're not overweight and need to gain, you may be one of the non-type 2s. Is there any way you could see an endocrinologist who supports LC diets?

    Most people on LC see improvement in lipids. But I think it's not universal. The "Life Without Bread" LC doctor says that people over 65 are less likely to see improvement in lipids on LC than younger people.

    When you say "increase in lipids" you didn't specify which ones. Increase in HDL is good.

    I think you really need to find someone who can look at your whole medical record to figure out what's going on with you. Even an MD can't prescribe on the Internet, and I'm not an MD.

    Thanks for your comments. Let us know how you do.

    ReplyDelete
  5. No, no, I didn't expect you to diagnose or prescribe. My "How can I gain weight" question was rhetorical. I guess I was simply venting my frustration with the medical & dietary professionals I've encountered.

    I don't think I'm a non-Type 2, based on some early tests and everything I've read. I'll check out the Life Without Bread doc you mentioned.

    Re my lipids. My triglycerides have always been low & still are, though this year they jumped from the 30s to 79.

    Both my LDL and HDL increased since I started LC/HF in early 2007. In mid-2006 when I was eating many more carbs but much much lower fat (and was diagnosed with pre-D just 4 months later), HDL was 88 and LDL 87. In 2008, HDL was 99 and LDL 170.This Feb, HDL was 146 & LDL 314--yikes!

    So I'm a bit of a puzzle. : )

    ReplyDelete
  6. I agree it's a puzzle. Slighty changing lipid levels wouldn't be surprising. But this is a huge change. You could have your particle sizes measured. If they're light and fluffy, they're good.

    Or if you could afford it, you could see Bernstein, who might have an explanation.

    BTW, I double-checked Life Without Bread book, and although it does show that "older" people don't improve cholesterol, the cutoff is lower than I thought, not 65.

    If you can Search Inside the Book on Amazon, look for Figure 6.5 in the Lutz book. Keep in mind it shows only total cholesterol, not HDL and LDL.

    ReplyDelete