[ExI] Old Nutrition Studies

Harvey Newstrom mail at harveynewstrom.com
Sun Aug 23 15:44:13 UTC 2015


The China study is old.  It started in 1983 and measured variables of diet known at the time.  We know a lot more now.  Nobody should be taking their advice from such an old study.  The question of whether this study was "right" or "wrong" is irrelevant.  We know a lot more nuance about nutrition now than we did then.  So arguing whether this study was right or wrong misses the point about whether we should eat meat, fat, cholesterol, or turn vegetarian.

Here are some examples of what has changed since the China study and other similar older studies.

1. Early statistics seemed to show that meat causes cancer.  People who ate meat get cancer at statistically higher rates than people who didn't eat meat.  So it made sense to avoid meat.

However, correlation does not mean causation.  Later studies subdivided variables to try to isolate how and why meat causes cancer.  It turns out that cooking method was one of the big factors in this.  Charring meat with heat creates carcinogens.  If we cook meat without charring it, we avoid the creation of most of these carcinogens.

Does that mean the original studies were wrong and meat did not cause cancer?  I say no.  As a whole, most meat really was causing cancer.  But does this mean that the study was right and should we stop eating meat to avoid cancer?  Again, no.  We now know how to cook meat to avoid these charring induced carcinogens.  Does that mean reality changed or the studies keep flip-flopping?  No, we learn more and more and keep further refining our nutritional understanding.  This is not the same as being wrong or changing our theories all the time.  We are actually continually improving our theories so that they get better and better over time.

2.  Early statistics seemed to show that cholesterol clogs arteries leading to heart attacks.  People who had higher blood cholesterol levels get clogged arteries and have heart attacks at statistically higher rates than people who don't.  So it made sense to avoid dietary cholesterol.

However, correlation does not mean causation.  Later studies subdivided variables even further.  It turns out that dietary cholesterol is very small and does not raise blood cholesterol as much as other dietary factors that raise cholesterol.  Eating cholesterol containing foods such as eggs and shellfish turns out to not raise blood cholesterol very much.  However, eating lots of saturated fats and refined carbohydrates turns out to skyrocket cholesterol and related compounds like triglycerides.

We also know that there are different kinds of cholesterol, HDL (good) cholesterol and LDL (bad) cholesterol and other kinds of cholesterol.  Most people have the bad kind, so most early studies found the bad results.  But some people who measured "high cholesterol" actually had low bad cholesterol and high good cholesterol.  Such people actually had better results they higher their good cholesterol got.  So any old study (such as the China study) that simply measures "cholesterol" without distinguishing the good from the bad is practically useless.  We don't know what they were measuring.  We don't know if differences were in the good or bad kind.  We don't correlate which changes caused which results.

Does that mean the original studies were wrong and cholesterol did not cause heart attacks?  I say no.  As a whole, most high cholesterol readings were the bad cholesterol which really was causing heart attacks.  But does this mean that the study was right and should we stop eating cholesterol or use blood cholesterol levels to predict heart attacks?  Again, no.  We now know that total cholesterol does not indicate good or bad predictions.  We now look at different cholesterol levels and the ratios between them for better measures.  Again, we are not changing the theories about cholesterol.  We are further refining our nutritional understanding of how different cholesterols effect the cardiovascular system.

3. There are similar examples with eating a high fat diet.  We now know that there are good fats and bad fats.  Early studies did not distinguish between good fats and bad fats.  We now know that some fats should be decreased in the diet while others should be increased.  It is too simplistic to argue whether "fat" is good or bad.

4. A related example of refining our knowledge occurred with butter vs. margarine.  Scientists were correct when they statistically correlated saturated fat with heart attacks.  They suggested that people switch from butter to vegetable oils.  So many people did.  And many people started having more problems than before.  Were the scientists wrong?  Not really.  We now know that most vegetable oil margarines are hydrogenated, making them even more super-saturated than the saturated fats.  They also induced more trans fats, which is also a very bad dietary fat.  So while scientists recommended that people switch to vegetable oils to eat less saturated far, people actually switched to hydrogenated vegetable oils which were even more saturated fat than before.  While it was mistakenly believed that people were eating less saturated fat, they were actually eating much more saturated fat.  We now know a lot more about the whole spectrum of fats.  And we know that hydrogenated oils in margarines are worse than saturated fats in the same way that saturated fats are worse than natural vegetable oils.

5. There are older studies that found that the autopsied brains of people who ate soy had more dementia indications than people who did not eat soy.  However it was later found that the people who ate soy were living longer.  So what they really found was that older people have more dementia indications than younger people.  When correlated for age, people who eat soy do not have more dementia indications than people who do not eat soy.

6. There was a recent theory that too much estrogen might cause hormone problems for males and that soy contains a pseudo-estrogen.  So people stopped drinking soy milk and switched back to cow's milk.  Then it was realized that cow's milk has a magnitude more times real estrogen than soy has pseudo estrogens.  If people really wanted to avoid estrogen, they should stop drinking cow's milk and switch to soy milk.

My point with these various examples is that science works over time and keeps getting better and better.  But it's not perfect.  We cannot argue whether an older study is right or wrong.  The answer is almost always more nuanced and has to be subdivided into other better questions about what was right and what was wrong.  Ideas about increasing or decreasing a single food group (meat, grains, dairy) or a single macronutrient (fat, protein, carbs) are usually too simplistic.  The more we study these questions, the more they fragment into many more precise questions until the original question becomes meaningless.

Are carbs good or bad?  Is fat good or bad?  Is meat good or bad?  Is vegetarianism better or worse?  Is veganism better or worse?  All of these are invalid questions.  Depending what kinds of carbs, fats, meats, vegetarian diet or vegan diet you choose, you can answer these questions either way.

And the answer to the question about whether any particular study is good or bad should be approach by looking at further refined studies.  It almost never can be answered by deciding if the study is good (and accepting all of its conclusions) or deciding that a study is bad (and rejecting all of its conclusions).  Both of these extreme edge-case positions is almost always wrong.

--
Harvey Newstrom   www.HarveyNewstrom.com






More information about the extropy-chat mailing list