Saturday, April 3, 2010

Is Dieting a Sport?

One thing that annoys me (well, OK, a lot of things annoy me; I'm becoming a curmudgeon) is when people approach dieting like a team sport.

You pick your favorite diet, and then you defend that diet come heck or high water. When a scientific paper supporting your diet choice is published, you crow. When a scientific paper supporting some other diet is published, you ignore it.

A couple of recent papers concerning the impact of saturated fat on heart disease illustrate this unscientific approach by some people in the science-discussing community.

In January, a study titled Meta-analysis of prospective cohort studies evaluating the association of saturated fat with cardiovascular disease was published online ahead of print publication. The study concluded that "there is no significant evidence for concluding that dietary saturated fat is associated with an increased risk of CHD or CVD."

The study was pretty much ignored by the mainstream science press, which tends to support the official American Heart Association low-fat approach to heart health. The New York Times didn't mention it. The various popular science summary services like Science Daily and EurekAlert also didn't report on it.

As diabetes blogger David Mendosa wrote, "I couldn't find any mainstream articles about it today. Not one of the four sources that I rely on heavily for leads to new studies has carried a word about this one.
"

But response in the low-carb community was immediate. People on low-carb diets tend to eat a lot of fat, often including a lot of saturated fat. Blog after blog reported on this study, and some of the bloggers made fun of the "low fatters" and patted each other on the back for following the "correct" diet.

More recently, another paper, titled Effects on Coronary Heart Disease of Increasing Polyunsaturated Fat in Place of Saturated Fat: A Systematic Review and Meta-Analysis of Randomized Controlled Trials was published online. This paper concluded that replacing saturated fat with unsaturated fat could reduce the risk of having a coronary heart disease "event" almost 20%.

This study was picked up by the science reporting services like Science Daily, but to date, I haven't seen a single one of the sites or blogs that publicized the "no effect of saturated fat" study mention this other study, and I've been looking.

To be fair, I get the URLs of some lipid blogs from the links in other blogs, and because people tend to link to blogs that agree with them, they do tend to read each other's posts and come to similar conclusions. But I find this business of ignoring the studies you don't agree with sad.

This isn't science. This is religion, or politics . . . or sports. When I was a child, I was a big supporter of the Washington Senators, the team that ended up in the basement year after year. The big excitement was whether they'd end up last, as usual, or perhaps claw their way up to next-to-last. So I know what it's like to root for a loser. You grasp at straws.

For example, Dean Ornish, who advocates an extremely low fat diet to prevent heart disease, was once asked about the fact that when your fat intake is low, your HDL cholesterol, the "good" cholesterol, goes down as well as your LDL, the "bad" cholesterol, so the ratio remains the same or even gets worse.

He said well, maybe when you're not eating fat, you don't need HDL.

But finding the best diet for people with diabetes shouldn't be pursued like this. We need to look at all the evidence, whether it supports our preconceived notions or not.

In fact, these two studies are not that far apart in their conclusions. What the first study said was that there was no significant evidence for linking saturated fat with heart disease. But they hinted at the results of the second study: "More data are needed to elucidate whether CVD risks are likely to be influenced by the specific nutrients used to replace saturated fat."

And the second study concluded that yes, it does matter what you replace the saturated fat with. Replace it with carbohydrate, and people's risk goes up. Replace it with unsaturated fat, and people's risk goes down.


Both studies were meta-analyses, and like many people, I'm not a big fan of meta-analyses, as I discussed here. Nevertheless, they hint at possible relationships.

And I don't think we should sit around throwing darts and this study or that study and maintaining the ideas we've had for decades. What we need to do is to look at all the evidence and try to interpret it in the best way we can given today's scientific and statistical tools. We need to try to find out why different studies seem to give different results and figure out how we can apply those findings to individual patients.

When I was in graduate school, forced to read a little in the history of biology, one thing that struck me was that often when there were two different schools of thought on some topic, it turned out they were both wrong. The answer turned out to be something else, which they couldn't have known because the technology for testing for that thing had not yet been developed.

So it's possible that a similar thing applies to research on dietary fats. Maybe it's not the saturation/unsaturation of the fats that is important but how fresh they are. Maybe it's the degree to which the fats are oxidized, or glycated because of high blood glucose, or modified in some other way that makes the most difference in heart disease.

Maybe the type of fat depends on what you're doing with that fat. Unsaturated fats, expecially the omega-3 fats found in fish, are easily oxidized when warm. This is what causes the "fishy" smell when fish sit around before you cook them. Using fish oil for frying would be a bad idea. The best fats for frying are the saturated fats. But most studies don't ask about how the various fats were used, or how fresh they were.

Maybe we can tolerate any kind of fat when it's not modified by food additives or the many chemical pollutants in our environment. Even organic food and bottled water are not free from contaminants, especially when the water is bottled in plastic.

Maybe we can tolerate almost any kind of fat in limited quantities, but when we overwhelm our metabolism with huge amounts of any kind of fat we'll see our heart health decline.

If it turns out that any type of food does, indeed, affect heart disease, we need to study why that food has that effect. We need to determine if it's eating any of the suspect food or eating a tremendous amount of that food that is important.

We need to abandon more studies designed to prove some preconceived notion (fat is bad, or fat is good) and instead encourage studies that show why different studies appear to give different results. Was it study design? Poor use of statistics? Poor choice of patient populations? Poor choice of endpoints?

You can look at short-term effects or long-term effects. You can lump together all cardiovascular events, including mortality, or you can study only mortality, or you can separate strokes from heart attacks, or you can try to study them all. In the latter case you need gargantuan overall sample sizes to have statistical significance in all the groups. And that means very expensive studies, especially if it's a long-term study.

You can study saturated fat from meat, butter, chicken, and coconut oil or you can study saturated fat from fast-food burgers, luncheon meats, hot dogs, french fries, potato chips, and southern fried chicken. The latter sources are apt to be associated with other behaviors such as eating a lot of processed convenience foods and drinking lots of sodas. So is it the effect of saturated fat that you're measuring or an overall unhealthy eating pattern?

So until we find the best possible diet, what do I think is the best diet for both preventing heart disease and controlling diabetes?

A l0w-carb diet. I've been following a low-carb diet for about 14 years.

But if you start out on a "standard American diet" that is high in both carbs and fats, I think the best approach is to drastically reduce the carbs and not replace them with anything. This way, your percentage of fat will increase; a typical low-carb diet includes about 60% fat. But your calories will go down.

In fact, studies have shown that when most people switch from a typical American diet to a low-carb diet, they reduce calories without thinking about it. This is because a low-carb diet tends to reduce hunger, so you don't want as much food.

Simply losing weight (not that the process itself is simple) improves blood pressure and blood glucose levels in most people. So if you reduce the carbs and don't replace them with a lot of other calories, you're apt to lose weight.

If not replacing the carbs with anything means that you're hungry, you can eat a little extra protein. Or even have a little extra fat. Just don't make a big effort to replace those 1000 calories a day you were eating in the form of bread, mashed potatoes, and doughnuts with something else.

I don't know why, when the press keeps blathering about the "obesity epidemic" the nutrition researchers hone in on replacing fat calories with something else. Do they want to keep people fat?

Come on, people. Let's stop bickering and use our brains and figure out how to make us all as healthy as we can be.


10 comments:

  1. I think the "rah, rah" attitude re: diets is just a way to seek social support. Dieting or maintaining can be difficult and socially isolating.

    "maybe when you're not eating fat, you don't need HDL." Well, if Ornish has some data, it would be nice to see it. He's had plenty of time. I have seen a study linking high HDL to longevity and another associating lower HDL to all cause mortality.

    I basically agree with this post, but I am concerned that seed/grain oils are very new foods to the human diet and require high heat and solvents to produce.

    ReplyDelete
  2. I don't think Ornish has data on that. It's like Helen Nearing ("Living the Good Life"), a vegetarian who, when told she was vitamin B12 deficient, said that if you don't eat meat, maybe you don't need vitamin B12.

    I'm not a big fan of commercial vegetable oils. But allegedly cold-pressed olive oil is a good compromise. Unfortunately, a lot of that is said to have been diluted with other oils.

    ReplyDelete
  3. I agree with your take on various studies. I often read research papers that clearly lay out their bias in the introduction. More often, the results of the study are consistent with the data, but that means very little if the study is poorly designed. I am constantly researching new information and base my decisions based upon the best information i get.

    My favorite example of this is my approach to raising my HDL. I haven't found any "proof" that raising my HDL will reduce my risk of heart disease. In my study sample of one (me), moving to a low carb diet has produced dramatic increases in my HDL. Couple that with niacin and my HDL has more than doubled. Will this help? I don't know for sure because I have't seen any studies demonstrating that doubling HDL does anything beneficial. At best, I'm merely guessing that it will help.

    The key is to critically examine all new information and carefully draw conlusions.

    ReplyDelete
  4. "I am constantly researching new information and base my decisions based upon the best information i get."

    ET, I agree that this is the best approach. We're constantly learning new things, and we have to be open-minded enough to change our beliefs when we get new evidence.

    This is one problem with people who have made their reputation writing books on one type of diet or another. They're sort of wedded to that approach. I'm not a big fan of Dr Whittaker, but I did admire the fact that he admitted that he was wrong when he said people with diabetes should eat a lot of carbohydrate.

    I try to read different points of view, although I don't spend much time reading pronouncements by formal groups like ADA or AHA because, like diet book authors, they are often wedded to a certain approach and can't think outside the box -- or write outside the box -- without risking their careers.

    So keep up with your approach. I think it's the best one.

    ReplyDelete
  5. Thanks for this post Gretchen. I am new to DM (1 year) and am learning as much as I can while keeping other areas of my life afloat. So when I read reports/articles, I do find myself looking for the 'one' that will help me keep my BG down. I need to be reminded to keep an open mind and to look at each one to see what I can learn and not feel if the title sounds like the opposite of what I am doing, that it is 'wrong' or cannot teach me something. I so appreciate your work and that you are willing to take the time to share what you know with those of us who don't have the background that you have.

    Michelle

    ReplyDelete
  6. Thanks, Michelle. I agree that if you read that huckleberries help with BG, there's no harm in trying it. Maybe it would help. Sometimes trying a new food means you eat less of some other food that makes your BGs go up. That's why no diet works for everyone. We all interpret them slightly differently.

    ReplyDelete
  7. Good points Gretchen. We need to keep an open mind.

    Ornish's point regarding HDL makes some sense, particularly when you look at the risk reduction effect of a high HDL in people with no risk factors:

    http://healthcorrelator.blogspot.com/2010/02/what-should-be-my-hdl-cholesterol.html

    The risk reduction is small, perhaps because those people are not developing atheromas, and so there isn't a lot of atheroma-related cholesterol to be picked up by HDL particles.

    And there is also the Kitavans ...

    Elevated blood glucose levels though, that is another matter. There is recent evidence that blood glucose control before age 55 may increase your chances of living beyond 90:

    http://healthcorrelator.blogspot.com/2010/04/blood-glucose-control-before-age-55-may.html

    ReplyDelete
  8. Ned, I don't understand what you're saying. Ornish says HDL doesn't matter, that low HDL is OK, and you've cited a paper showing the benefits of high HDL.

    Yes, the risk reduction is less in healthy people, but it's still there, and many Americans do, in fact, have other health problems. I don't think many people would follow his very stringent diet and exercise plan if they didn't already have signs of heart disease.

    ReplyDelete
  9. I am not suggesting that people should follow his diet.

    What I am saying is that his point about a low HDL is not completely wrong. HDL may be relatively low in some people, who do not have risk factors associated with the formation of atheromas, and that will be okay.

    The exponential decay function on the post for no-risk individuals is practically flat compared with those for people with risk factors. The more risk factors, the more a high HDL matters.

    ReplyDelete
  10. I don't think we know for sure that low HDL is OK in people without other risk factors. The risk is low. With HDL of 25 it's about 75 per 1000 per 10 years. With HDL of 60 it's about 25 per 1000 per 10 years.

    That means with HDL of 60, about 50 more people in every thousand every 10 years would have a coronary event. That's 5 per thousand a year. Pretty small number. But if you're one of those five, it's important.

    For that matter, we don't really know that high HDL is always protective. There are various subgroup sof HDL, and some people say only some are protective. I once did a blog post about a study that showed that high HDL was bad for people with diabetes!

    So all of this is tentative. We just have to do the best we can, and use common sense.

    My point about Ornish was that people tend to defend their dietaryt positions without thoroughly researching them. If he'd looked at studies and made the same point, it wouldn't have bothered me.

    ReplyDelete