Dr. David Perlmutter states that one must eat a high fat/low carb diet in order to protect his brain: http://www.youtube.com/watch?v=qJIJ0gyCoRY. I was wondering if there is any research that shows that there is a causal link between high carbohydrate intake and diseases such as Alzheimer's and dementia. How can we trust these studies if the experimental designs behind them are faulty? Is any diet right? Should we apply Ockham's Razor and state that one can eat anything in moderation?

  • You cannot eat everything in moderation; you can eat good things with some negative consequences in moderation though. But I understand what you were trying to say. :) Mar 25, 2014 at 4:01
  • Occam's Razor, btw.
    – JohnP
    Mar 25, 2014 at 15:18

2 Answers 2


The only studies to my knowledge that have any correlation between carbohydrates and Alzheimers are epidemiological. In short they use survey data to find areas of study where more controlled studies should be performed to see if there is a true cause and effect relationship. The problem is that many people assume that correlation is the same thing as causation, and the distinction is not made clear by the media.

Dr. Wiel has some thoughts on the epidemiological study. In that, he highlights the fact that there are no studies that provide a direct link or causation between carbs and Alzheimers.

  • Alzheimers is influenced by high systemic inflammation
  • Some carbs (particularly sugar and high fructose corn syrup) increase inflammation dietarily
  • There are other foods that increase inflammation that do not carry high carbs (from Self Nutrition Data).
  • Sleep deprivation has a stronger link to Alzheimers based on a a controlled rat study. Sleep flushes the metabolic waste in the brain, including the proteins associated with Alzheimers.

Similar studies have been used to try and promote whatever pet doctrine or diet that someone wants to sell. It is very important to understand what kind of study is being performed that the person is referencing, and whether the supposed mechanisms are the conjecture of the researchers or the result of a focused and controlled study. Many times epidemiology has provided a strong correlation, and then further controlled studies both with rats/mice and with humans have proven that there is no causal relationship.

A more infamous example is the "lipid theory" which is the foundation of the USDA dietary recommendations to this day. The study which started it all was poorly done even from epidemiological standards, and data was thrown out to support the researcher's hypothesis. Several studies have come along later showing higher fat diets do not have a causal relationship with heart disease. The long story short: dietary fat does not have a causal relationship but body fat does.

The bottom line:

If you eat the correct amount of food, with a proper mix of macro nutrients to support healthy metabolic function, your body will be able to take care of itself just fine--even if you indulge in sugar laden foods occasionally.

NOTE: carbs are neither angels nor demons. Due to the "lipid theory" being proven inaccurate in controlled studies, there are a number of people who want to throw the pendulum the opposite direction. There are a number of epidemiological studies trying to vilify carbs that were once held as the number one thing you should eat. They are looking at insulin response or affect on growth hormone, both of which are useful when training as an athlete. Just in case you stumble across it: insulin and human growth hormone do not cause cancer, but they can assist it spreading faster. If you already have cancer, you may want to reduce carb intake. If you do not, then you should eat to give your body the energy it needs to do it's job.

  • Does you consider Dr. Perlmutter's advice to avoid grains misinterpretation of the studies and do you thing he is to "extremist" in his views?
    – phil12
    Mar 26, 2014 at 2:14
  • I think he's making too much out of epidemiological studies. For sedentary people there is nothing wrong with following his advice, and it may or may not work for them. For any athlete, the demands for carbohydrates are much higher as the body needs to replenish glycogen in the muscles. Not every doctor's advice applies to you specifically. Since grains are effective at providing carbohydrates, it is silly to simply exclude them from an athlete's diet. Focusing on high fiber grains controls the insulin response, and helps keep you satisfied longer. Mar 26, 2014 at 13:02
  • Having tried strict low-carb I agree that that isn't a good diet if you want to improve your power stats. For me personally, I've found that complementing the basic strict no-grains low-carb diet with a decent hand full of pre-workout berries provides all the carbs I need to keep improving my power stats while (AFAICT) still reaping all other low-carb benefits.
    – Pibara
    Mar 26, 2014 at 19:00

I don't think that, lacking definite proof of causality, basing advice on strong correlations would qualify in any way as bad advice or as "extremist".

We know there are strong correlations at a populations level between eating relatively 'much' sugar, 'much' refined grains and relatively little lipids and many diseases. We don't have any data indicating that eating very low levels of grains and sugars are correlated with any negative consequences.

Correlations can be misleading and arn't definite proof, its the difference between being a suspect and being a convict. But the idea is that its not unwise to incarcerate the suspect while you continue to look for definite proof.

At this moment, given the available data, reducing sugar and grain intake seems like a safe bet.

Not the answer you're looking for? Browse other questions tagged or ask your own question.