On Fri, 12 Jan 2001, Irene wrote:

> > I was
> > able to lose weight finally, although I've never reached a low
> > body fat level, but only by a very restrictive regimen.
>
> Sorry, I don't remember -- Todd, are you still on the Anchell diet?

Not strictly, and I have gained weight as a consequence.  It is a
very monotonous diet, and I cycle on and off it.  I really don't
know what the secret of that diet is, but I suspect that Rachel
got it right when she suggested that the monotony is the key.

> Well, there are these two studies, although their study groups were
> restricted to diabetics who do have a different blood chemistry from
> non-diabetics. In these studies, though, a high-fat, low-carb diet *did*
> improve cholesterol levels.
>
> http://www.pslgroup.com/dg/10786a.htm

Yes, but please note that this is also a restricted-calorie diet,
at 1800 cal/day.  There is evidence that suggests that people on
lowcarb diets in general will see lower cholesterol when they are
losing weight, but then the cholesterol goes back up when they
stabilize.  This is also consistent with Dr. Lutz's observations
(Life Without Bread).

> At this point, it's all still a game of statistics. Whichever
> numbers you'd like to make important, you can find some kind of support in
> the journals.

I completely agree.  In the end one must decide what to pay
attention to.  I don't presume to tell newcomers here how they
should think about cholesterol, so I just point out that this
kind of diet may well have an effect that many would consider
adverse.  It's better to know that at the outset, rather than
being disillusioned later, as I was.

> One number that stands out to me, is the highly significant number of people
> who have their first heart attack with all of their cholesterol levels
> falling in the 'desirable' range -- I think it's somewhere around 40%, but
> I'd have to look it up.

According to McCulley, it's 60%.

> This is an issue where there is no clear-cut guide at this point. I don't
> think I'd be comfortable completely ignoring cholesterol levels that were
> off the charts (though I must admit, I've never had my cholesterol checked).
> I have seen a lot of medical data that would support the TG:HDL ratio as
> being more important than a simple Total Cholesterol or LDL count.

The reason why I favor this ratio is that there are several lines
of evidence to support it (The Eades discuss this in Protein
Power Lifeplan).  First, there is Gaziano's 1997 research
indicating that this ratio had more predictive power than other
ratios.  And what I like about his study is that it was a
*prospective* study, rather than a fishing expedition.  Gaziano
took the blood samples at random and just stored them for a
number of years (5, I think).  After the 5 years he looked at who
had heart attacks and who didn't, and the TG:HDL ratio had by far
the highest correlation.  The only problem is that after 5 years
the absolute number of heart attacks was fairly small, so the
statistics may be off.  It needs to be done again with a larger
population, for a longer time.

Second, there is the research that suggests that it is not LDL in
general that is atherogenic, but the small, dense fraction of it.
This is not measured in routine lab tests, but is known to be
strongly correlated with TG.  So if TG is low, so it the small
dense LDL, whatever the total LDL count may be.  Finally, I think
the high LDL, low TG profile is fairly uncommon, except in
lowcarbers; most people with elevated LDL also have elevated TG,
which would tend to explain why there is any correlation at all
between heart disease and LDL levels.

That's my story and I'm sticking to it.

Todd Moody
[log in to unmask]