Sunday, September 23, 2007

Fruit and vegetables, last post (almost)

I've posted twice on the links between fruit and vegetables and oxidative damage to lipids, protein and DNA. This then begs the question as to why diets high in fruit and vegetables, on an epidemiological basis, are associated with less chronic diseases than diets deficient in fruit and vegetables. There seems to be a paradox here, fruit and veggies are repeatedly associated with oxidative damage, avoiding them is repeatedly beneficial, yet overall they look good for health when studied at the population level.

So, epidemiology finds fruit and veg are associated with better health. They shouldn't be. But epidemiology shows only shows association, never causality. Only interventions studies do that. On the basis of controlled intervention studies fruit and vegetable consumption shouldn't improve health, so there must be confounding factors here.

You have to ask what the confounding factors might be. Is there anything about the sort of person who eats lots of fruit and veg, compared to those who don't, which might over ride the damage done by the vegetables?

Well, some factors come to mind.

Poverty has to be the first. Poor people do worst on a health basis on just about every measurement possible. They also don't buy much in the way of vegetables. I remember one study showing that poverty is associated with very poor outcomes in diabetes. The conclusion was that we should spend lots of money educating poor people to spend their limited resources on vegetables. You can guess what I think about that. The obvious solution is to give MONEY to those in poverty. When they are rich enough, their health might improve. They might even be able to tolerate eating vegetables with impunity. Of course the link between poverty and ill health is epidemiological. I haven't seen an intervention trial where a large group of impoverished people were give £100,000 each per year and the placebo group given a similar notional value in Enron shares. I'm waiting for that one.

The second consideration is to ask what vegetables might displace from the diet. That is, what do people eat instead of vegetables. Chances are it is sugar or high fructose corn syrup that forms a big chunk. In the fruit and vegetable wash out study a typical daily diet is given. The diet, including total calories, was very strictly controlled. The depletion diet was high in carbohydrate, but mostly from potatoes, rye bread and white bread. The only fructose in the menu was from sucrose in carrots (which are loaded) and in the cake. So this is a pretty low sucrose diet. Eating sucrose and high fructose corn syrup is like getting all the bad aspects of fruit and vegetables but none of the vitamins needed to process the sugars.

Vegetables are bad, but refined sugar is probably worse.

It is possible.

Peter

9 comments:

Stephan Guyenet said...

I appreciate your thoughts on this question. I have a possibility. I've always suspected that the reason vegetables are healthy is because when you eat a salad or a saute, you're essentially replacing carbohydrate in your meal with fat. This is because most vegetables are pretty low in carb, and are doused in fat before serving (salad dressing, butter, olive oil etc.).

The calories from a salad or a vegetable saute (excluding starchy tubers/roots) are almost exclusively from fat.

Stephan Guyenet said...

And by the way, as someone who studies the mechanisms of aging, my view is that the oxidative damage theory is far from proven. Just because the antioxidant capacity of your blood is going this way or that, it really doesn't allow us to predict anything about health outcomes at this point.

For all the talk you see in the media about how great antioxidants are, I'm not aware of a single convincing study showing that antioxidants improve a health outcome in humans (without vitamin deficiencies).

Peter said...

Now there's a thought. The big point with the WHEL and PPT studies was their fat restriction, so you would expect them to loose this benefit. And to be absolutely fair I don't think that there was any increase in overall mortality, certainly with the WHEL study. So if I'm completely honest I think vegetables (excl roots) are pretty neutral. They're what we both eat routinely, as a vehicle for fat. Could well be the same effect on a smaller scale in general. I think the fructose in fruit is a different matter. It's such an effective glycating agent that the only reason we can eat fruit at all is the liver's constant vigilance.

A couple of points came out of Dr Bernstein's site forum. Acidic foods/dressings blunt postprandial hyperglycaemia. This appears to be a pretty consistent observation and speaks in favour of acidic dressings, particularly olive oil based. The other was that just about the only effective antioxidant was alpha lipoic acid, used for neuropathy amelioration. Most posters found this effective, but I notice that our body excretes the stuff asap, on a par with green tea extract...

On the mechanism of aging, you're obviously eating reduced carb yourself. Do you have any thoughts on insulin as the determinant of "breed young, die young" vs aging being the build up of chronic glycation damage? Or is it all just the luck of the draw with telomere repeats?

Peter

Stephan Guyenet said...

I see what you mean about the WHEL study. I have to say, my diet would be pretty boring without vegetables. I wonder if leafy greens are any good? High nutrient density, low carb. Plus our closest relatives eat them like they're going out of style.

I'm interested in the alpha-lipoic acid thing. My suspicion with "antioxidants" that are associated with improvement of some measure in a biological system, is that the effect isn't due to the antioxidant activity of the molecule. For example, curcumin was first thought to help in Alzheimer's models because it's a powerful antioxidant, but now they're finding it has specific effects on inflammation and cell death pathways that seem to be independent of this activity.

As far as the insulin/aging connection, I don't know the answer for sure, but there are some tantalizing clues. First of all, I don't want to present myself as an expert. I'm a grad student whose work intersects with aging mechanisms, and I also have a personal interest in it.

I'll start by saying that I don't believe in the "wearing-out" theory of aging, which is still alive and well even in professional circles. I think aging is a controlled process that is selected for in evolution. For example, not every cell ages. I have cancer cells in my lab that have been happily dividing orders of magnitude beyond their Hayflick limit with no signs of senescence. These cells have reproduced to the point where they have outlived the person in whom they originated, and probably created thousands of times more biomass than this same person did. This is simply because they accumulated the wrong two or three mutations. On a cell culture level, you can activate or block the expression of certain genes that dramatically affect the onset of senescence. You can even rejuvenate senescent cells. How this applies exactly to whole organisms (with stem cells etc) is unknown, but these observations are pretty telling in my opinion.

The second big blow to the "wearing out" theory is the fact that we, as a species, do not age. Our germline is the same one that began with the very first organism ever to appear on Earth, and a process as slow as evolution seems to be enough to keep us from degrading over time as a species.

Also, naked mole rats live 10X longer than their closest genetic relatives, accumulate lots of oxidative damage, and seem to tolerate it well.

That having been said, I do believe it is possible to have a system in which "wearing out" is permitted to occur in a genetically controlled manner. For example, you might have a gene that senses when you hit 70, turns on (or off) and slowly shuts down your capacity to defend yourself from damage.

Back to insulin. I have a couple of thing to say about it. First, mutations in the IGF1 pathway can extend lifespan in a variety of organisms. They also come with deleterious effects, but that could be simply because our methods are crude. Second, I've heard that low insulin is one of the only things that predicts extreme longevity.

And about telomeres, I don't believe they're causal in organism-scale aging.

Peter said...

Hi Sasquatch,

The alphalipoic acid marketing site is here. The Bernstein site threads on Insulow are gathered together here. I'm still suspicous that if it is so good for us then our bodies shouldn't dump is so quickly, but then type 2 diabetics have a "broken" metabolism, so even pharmacology may well have a roll...

Re longevity: At my age I like to keep my fasting insulin just below the bottom of the lab ref range and budge it as little as possible, so far so good! It would be nice to see my youngest son grow up if at all possible. My father just made 60, by a few days.

Peter

Stephan Guyenet said...

Hi Peter,

I'm sure you will make it past 60.

If you don't mind me asking, do you take insulow? I'm suspicious of any kind of supplement, personally. Is there any data on alpha-lipoic acid and all-cause mortality in humans?

Since you're interested in keeping your insulin low, I'm sure you're familiar with intermittent fasting. It has dramatic effects on the fasting insulin level. I've been doing weekly 24-hr fasts lately, along with restricting carbs. If you haven't seen it yet, here's a neat study on the metabolic effects of alternate-day fasting on nonobese people:

http://www.ajcn.org/cgi/content/full/81/1/69

Peter said...

Hi sasquatch,

Nice paper. I'm familiar with the concept of intermittent fasting, much of it from Dr Bernstein's site. If I can remember the units correctly the Kwasniewski diet is giving me a fasting insulin of 2.5mU/l, so on a par with the women's value on a fasting day after 3 weeks in that study. The Kitavans averaged 4.0mu/l. The nice thing about Dr K's approach is no hunger. I wonder if this would have occured in the study if the eating days had been carb restricted and high fat? The longest I've fasted for is 48h (cooking for the kids through this) and it was effortless. It seems to me that I eat very few of my carbs through the day, eggs and cream account for most daytime food. I wonder if this is the equivalent of 1 meal a day with a 24h "carb fast" between each meal? An "insulin fast" too...

Peter

Stephan Guyenet said...

Hi Peter,

I think you, Gary Taubes and I would agree about diet composition and hunger. One of the things Taubes talks about in his book is that hunger, in the most common sense of the word, is the result of carbohydrate consumption elevating insulin.

When people are on very low-carb diets or when they have been fasting (both of which have similar metabolic effects), they experience hunger differently. I've found this to be true of myself. 24-hr fasts are fairly easy for me now. I even do strenuous workouts at the end of my fasts sometimes and have plenty of energy.

Interestingly, you can fast for days without being very hungry, but if you're eating a calorie restricted high-carb diet, you'll be ravenous. I guess it's the difference between burning fat, which your body has plenty of even if you're not eating, and burning carb, which you can't store much of. Maybe that's an oversimplification.

So yes, I agree with you that the diet would probably have been more tolerable in that study if the subjects had been eating fewer carbs.

Peter said...

Oh, forgot to say, no I don't take Insulow. Occasional Vitamin C, vitamin D 10,000iu daily in the Winter and fish oil 5g/d year round. That's it. I think the Vit C is fairly irrelevant if you LC. Mostly I use it as a paracetamol substitute for colds.

Peter