It really was only a matter of time until a longitudinal study on a massive scale, such as this one in the European Journal of Clinical Nutrition, came out to back up what common sense told us was unhealthy. That being, that artificial, gross manipulations of diet aren't great for you. In fact, they may just shorten your life span a bit. This study was published online in advance at the end of 2006, but didn't come to my attention until right this minute (its now free online).
The study, a collaboration between researchers in Greece and at Harvard, attempts to evaluate the effects of a low carb-high protein diet on mortality (likelihood of death.) To this effect they followed 22,944 healthy adults from 1993 to 2003 and assessed diet through validated questionnaires. These participants were distributed into 'deciles' or categories based on their diet.
The quick-and-dirty explanation of the results is that higher intakes of carbohydrates was associated with a REDUCTION in mortality, however differences in protein intake did not affect mortality. The most powerful predictor of higher mortality was the combination of a low carb-high protein diet, which was significantly correlated with increases in mortality from cancer and heart disease.
Now these results shouldn't be interpreted to mean that all short-term applications of a low carb/high protein diet are unhealthy, only that for protracted periods of time such a diet has a negative effect.
Comments
Sensible diet and exercise for the win.
I have the pictures to prove it too. ;-)
Posted by: Boosterz | May 15, 2007 10:01 AM
Interesting...I wish they seperated out sugar intake from the rest of carbs.
Posted by: Cameron | May 15, 2007 10:56 AM
Interesting - why does 'common sense' tell us that eating foods that have recently been introduced to the human diet (Grains, potatoes, rice, etc.) over foods present for at least a million years (Meat, eggs, fish, etc) would be healthy?
Of course, the answer is that a lots of meats have quite 'artificial' fatty acid profiles, correlations between high protein intake and high processed food/trans fat intake, etc.. the apparent lack of control for total calorie intake dosen't help.
And excluding people with diabeties is, whilst understandable, excluding those who are least well adapted to a high carbohydrate intake, pretty much by definition.
Posted by: Andrew Dodds | May 15, 2007 11:18 AM
A.Dodds: Interesting - why does 'common sense' tell us that eating foods that have recently been introduced to the human diet (Grains, potatoes, rice, etc.) over foods present for at least a million years (Meat, eggs, fish, etc) would be healthy?
It would do well to emphasize the gatherer in hunter-gatherer. Look at our closest relatives, the great apes - and you will see that plant foods are far more important in total calorie intake. Human neural development may need slightly larger amounts of protein and fat in the diet, but does not mean that we had to be mighty hunters. Note that tubers may have been a major food source in many arreas along with fruits, when available.
And excluding people with diabeties is, whilst understandable, excluding those who are least well adapted to a high carbohydrate intake, pretty much by definition.
Not so. The most effective diet for Type 2 diabetics is a vegan - low fat, high carbohydrate, low glycemic index reference. It's the type of carbs that counts much more than the amount of carbs. In addition, higher fat generally leads to higher serum lipids which increases insulin resistance.
Posted by: natural cynic | May 15, 2007 2:28 PM
Did you actually read the article? The results are w/ respect to low carb / high protein diets. There's a pretty big functional difference between this and low carb / high fat, and a world of difference between the different sorts of high fat diets.
I've known people who eat bacon and eggs for breakfast every morning in accordance with one flavor of high fat diet; needless to say, there are consequences to this.
Posted by: Shane | May 15, 2007 3:04 PM
How much of the great apes diet consists of refined (manmade) carbohydrates, vegetable oils or added sugars? As these 3 things comprise about 63% of calories in the standard American diet, I don't think it's a valid comparison. As for "the most effective diet for type 2 Diabetics" you might want to look at the 56 week study that was published recently in the journal, Molecular and Cellular Biochemistry- "Beneficial effects of ketogenic diet in obese Diabetic subjects" which in summary showed the following: weight loss of 24.4 kg, total cholesterol down 28%, LDL down 33%, HDL up 63%, Triglycerides down 78%, and Fasting blood glucose down 50%.
Posted by: Michael | May 15, 2007 8:06 PM
The major point about type 2 diabetics is that anything that induces a hypocaloric state is going to improve their lipid profiles and insulin sensitivity - high fat, low fat, high exercise with no change in # of calories consumed, grapefruit+chicken breast+megavitimin+Nutri-drinks+..., whatever. Everything works if it is hypocaloric. The problem is always going to be staying with a hypocaloric situation. The point about Barnard's study is that it shown that % carbs in the diet is not be the problem - it's the source of the carbs. The most promising research into the ultimate causes of type 2 diabetes/metabolic syndrome is in the PPAR's which appear to control insulin receptor action along with cholesterol and other blood lipids.
The real take home lesson from the Greek study in an ad libitum feeeding situation in a free-living population with an essentially isocaloric diet [or a normal social situation]. In that case, the low fat diet is safer. And plenty of other studies have demonstrated that the source of carbs counts heavily - high glycemic index vs. low glycemic index. That, along with differences in fat source - oils from traditional sources [olives, nuts and fish] is better than fat from fatty meat and processed vegetable oils.
Posted by: natural cynic | May 16, 2007 3:49 AM
There is a gigantic problem with this study, which may completely invalidate the conclusions. This study was based on a self-selected diet. The implicit assumption (which may be completely wrong), was that health outcomes were a consequence of the diet. It may be that the (unconcious) choice of diet was dictated by underlying medical conditions that may be responsible for the increased mortality.
There was a recent JAMA meta-review that showed that supplemental antioxidants increased mortality.
There have been numerous very robust and very strong effects showing that green leafy vegetables in a self selected diet are associated with good health and low mortality. My reconciliation of these two seemingly disparate observations is that there is an "oxidative stress setpoint", which is regulated by physiology, and dietary choice is part of that regulatory control system.
It may be that the balance between dietary carbohydrate and dietary protein is also regulated by underlying physiology, a "carbohydrate protein setpoint". That underlying physiology may be what is responsible for the increased mortality observed.
If the "setpoint" hypothesis is correct, eating a diet that is incompatable with your "setpoint" would likely cause worse health (as observed in the case of the supplemental antioxidants above).
Posted by: daedalus2u | May 16, 2007 9:22 AM
daedalus2u,
Your hypothesis is that some or all of the hp/lc eaters chose their diet because they were already at high risk for mortality from cardiovascular disease or cancer, and that the choice is unconscious, perhaps instinctive.
This study focused mainly on subjects with no pre-existing coronary artery disease. But they did compare their results to those of subjects with one or the other of these disease. In both cases, researchers found that the hp/lc diet correlated significantly with increased mortality.
So, for Greeks with coronary artery disease or diabetes, the hp/lc diet was a bad choice, just as it was for those without these diseases.
The researchers did not compare the distribution of hp/lc index scores for the three cohorts (non-disease, CAD, and diabetic), so they do not address your self-selection concern.
Posted by: Family Nutritionist | May 16, 2007 1:04 PM
I am not suggesting that they were only "high risk", they may have already had the diseases in sub-clinical stages. They did exclude individuals with clinical disease, but those diseases don't happen "all of a sudden", they are chronic conditions that (no doubt) get progressively worse until they cross an arbitrary defined threshold and become clinically significant.
I think the only way to determine the true effect of diet is to make the study prospective, and force people to eat the diet that is under study. That is going to be a very hard study to do, particularly if it is multi-year. Maybe in prisons? But I don't see any other way of getting around the self-selected part.
Posted by: daedalus2u | May 16, 2007 10:44 PM
Human are still evolving according to their enviroment and cutlure. Agriculture societies like those in China or Europe will produce carbon-hydrate dependent metabolism. So this european study might fit people of agriculture society. But hunter-gather society depends on total different food which often has very little carb. Such people include Mongolian or eskimos who are almost pure carnivores (meat-eater). For these people, high-protein and low carb food might be most healthy ones which fit their metabolism better.
Have you noticed that people of hunter and gatherer society tend to have higher incidence of diabetes when they live with bread or rice which are typical agriculture food?
Posted by: AG | May 22, 2007 3:11 PM
Thank you for bringing such nice posts. Your blog is always fascinating to read.
Posted by: Emily | June 6, 2007 8:35 AM