[montecarl]

The tricky part is that it isn't clear what "balanced" means. Our bodies can survive on vastly different allocations of the three macronutrients. Carbohydrates are not required at all to my knowledge, while some amino acids and fatty acids are essential and cannot be synthesized from other foodstuffs.

Its possible to get most of your calories from carbohydrates, protein, or fat. The question is what is the right balance for optimum health? How much does this value change across different people (and possibly at different stages of life)? If you have epilepsy then a ketogenic diet might well be best. How about for other people?

What macronutient profile is "balanced"? What do you even base it on? Should it be 33% of each? Should we eat protein and fat in just a little in excess of what we need (to get the essential fatty acids and amino acids) and get the rest from carbohydrates? I don't think we have clear answers to those questions yet. Although, I think we have ruled out some diets as unhealthy (e.g. eating a lot of refined carbohydrates can cause diabetes). Since eating too many refined carbohydrates is bad does that rule them out as a primary calorie source? Probably not, but we need more research.

Personal Note: I've eaten a low carb, high fat, moderate protein diet for about 4 years. All my blood markers (cholesterol, triglycerides, etc) have improved significantly over that time period. Additionally, I used to get incredibly hungry all the time and feel bad if a meal was delayed. Now I can go much longer without eating and still feel well.

[gnaritas]

You don't have to be able to define balance, to point out unbalanced, and any diet largely cutting out one of the 3 is unbalanced.

> I've eaten a low carb, high fat, moderate protein diet for about 4 years. All my blood markers (cholesterol, triglycerides, etc) have improved significantly over that time period.

That's like saying I've been smoking for 4 years and I don't have cancer so smoking must not have any long term adverse effects; bad logic and insufficient sample size.

[dragonwriter]

> You don't have to be able to define balance, to point out unbalanced

Yes, you actually have to be able to define what range of mixes counts as "balanced" before you can label anything as outside of that range. (And, moreover, if you want to credibly assign significance to that label, you probably also need to be able to provide evidence that your definition of "balanced" corresponds to a range outside of which there are serious negative consequences.)

[gnaritas]

> Yes, you actually have to be able to define what range of mixes counts as "balanced" before you can label anything as outside of that range.

No, you don't; for example, eliminating one is not balanced now matter what numbers you assign to balance.