Some Spiked Leptinade with My Science Krispies Please!

Well, Jimmy Moore is out with another installment of his crusade against "safe starches".  Yeah, I know, I know.  He's all about learning and helping people get to the truth, moving the debate forward constructively, and above all else protecting people from potentially bad advice such as that a major staple macronutrient for 99.99% of the human population for at least the past 10 millenia and more can actually be "safe".  Sorry, but I call things as I see them, baaayybee!   

It's another long piece, and I just can't stomach reading much of Jimmy's self-delusions anymore.  But I am interested in this whole notion that somehow we're all somewhere on a diabetic spectrum and our carb-induced post-prandial glucose spikes are lining our rat poison sprinkled paths to an early grave.  The post includes a long response by Dr. Ron Rosedale, pastor at the Church of the Not-Too-Late-in-the-Day Spiked Leptinade Drinkers.  

I'm going to address a few of the cited studies and Rosedale's basic claims.  (I'll be breaking this up in parts).  

The crux of the ‘safe starches’ argument is that no harm will come of this and it is, in fact, healthy. It is acknowledged that blood glucose will elevate after eating ‘safe starches’, but will generally stay below 140 mg/dl that Jaminet says is perfectly safe. Is it?? The science writing (below) is on the wall and the answer is…clearly no.

It is important to realize that this answer is no; there is no safe intake of sugar, nor a threshold level of blood sugar below which no harm will come, and I will shortly devote a fair amount of time to show this.

Now it would have been nice for Rosedale to include links to his science ... you know, just to make it a bit easier for the curious to go check out the studies in their entirety.  But by not doing so only those pesky blogger types like moi will likely bother. 

First up we have: Is There a Glycemic Threshold for Mortality Risk? Balkau et.al. This study looked "the 23-year mortality data from the Paris Prospective Study of the 7,018 men, aged 44–55 years, who were not known as diabetic at the baseline examination". Rosedale quotes (emphasis his):

“…the lowest observed death rates were in the intervals centered on 5.5 mmol/l [99mg/dl] for fasting glucose and 5.0 mmol/l [90 mg/dl] for 2-h glucose.

CONCLUSIONS: In the Paris Prospective Study, there were no clear thresholds for fasting or 2-h glucose concentrations above which mortality sharply increased; in the upper levels of the glucose distributions, the risk of death progressively increased with increasing fasting and 2-h glucose concentrations.”

The first thing that jumps out at me, before even tracking down and looking at the study was that first quote:

 Lowest observed death rates 

centered on FBG of 99mg/dL

Note this is ALL causes of death.  Umm ... maybe there's something to the "we're all diabetic" meme after all, because last time I checked, that FBG level was just a measly 1 mg/dL shy of prediabetic!!  Or for the teaspoon crowd we might as well just forget the teaspoons, this is but a grain of sugar if that!  The conclusions are interesting, but the absence of a clear threshold where a fit curve "takes off" does not mean that there is no safe level of glucose ever.  Sheesh!  I also think this paper is the source of the the "shallow U" I've been referring to seeing somewhere (I still think it was Ned's Health Correlator blog).  Here's some worth-more-than-1000-words plots for you, for all deaths (top) and cardiac deaths (bottom):

From the conclusions: 

It is noteworthy that the relationship between glucose concentration and mortality followed a  J-shaped curve, except for fasting glucose concentration and death from CHD, which had a positive and linear relationship. The observed and the modeled relative risks were higher at very low glucose levels ... The modeled relationship between fasting glucose concentration and CHD mortality was linear, although a higher death rate was observed in the lowest fasting glucose class; the number of CHD deaths may have been too low to show a significant curvilinear relationship.

... When the 308 men with newly diagnosed diabetes were excluded from the analyses, there were still positive and curvilinear relationships between glucose concentrations and both all-cause and CHD mortality, except for fasting glucose concentration and CHD mortality, for which there was no statistically significant relationship.

Those J-shapes are tricky, especially when the leading side of the J appears more pronounced in the observed data vs. the model, and the bottom of the J appears awfully flat.  If we're talking health and life, CHD is only part of the equation.  One might even say that mild hyperglycemia is protective from the results of this study, that perhaps there's no clear threshold below which mortality risk doesn't also increase.  *I* didn't pick this study to make a point, Rosedale did.

Death rates were higher for the men with the lowest glucose concentrations ... [the discussion lists suspected cirrhosis and lower smoking rates as possible confounders] ... The curvilinear relationship between 2-h glucose concentrations and both all cause and CHD mortality remained when the men with these low 2-h glucose concentrations were removed from the analysis, whereas there was no longer a statistically significant relationship between fasting glucose concentration and CHD.

... There is no doubt that the increasing all-cause and CHD mortality associated with increasing concentrations of both fasting and 2-h glucose is in part due to the worsening risk factor profile with increasing glucose concentrations. The hazards ratios decreased after adjustment was made for other risk factors, but the fasting glucose levels still carried a statistically significant for all causes of death, as did the 2-h glucose concentration for both all cause and CHD mortality. 

In the Paris Prospective Study, there was no clear concentration of either fasting or 2-h plasma glucose above which the risk of death or death by CHD sharply increased. 

OK ... perhaps one of you dear readers can help me out here, but a diagnosis of diabetes here in the US begins with an FBG of 126 mg/dL or higher = 6.93 mmol/L.  So when they talk about excluding diabetics is that just looking in the center there?  Clearly above around the 7 mmol mark the curves seem to "take off".  Are the slight differences significant in the normal range?  Is lower better?  Or does being on the high end of normal mean that many of those men progressed to full-on diabetes in the two-decades following?

I believe Rosedale and I are in agreement that the hyperglycemia of diabetes is not the disease but the symptom.  He blames leptin resistance due to leptin spikes caused by carbohydrate consumption.  And yet he seems to be making the seminal error of equating the symptom with the cause.  Let's apply a uniform standard here, shall we?  It is pretty indisputable that LDL concentration correlates with CVD risk -- this is seen in an overwhelming number of studies --  but is not seemingly "causing" it by "clogging arteries" as they say -- rather it's a symptom of metabolic dysregulation.  Well, so, too it likely is with blood glucose levels.   The FBG and 2-hr post glucose-challenge levels are more markers for insulin resistance -- the underlying pathology of fasting hypeglycemia (liver IR) and impaired glucose tolerance (skeletal IR).  It is not dietary carbs per se that cause this no matter how many times the Taubesians try to repeat the mantra.  The dietary instigators of IR are, if anything, fat and fructose, not glucose which promotes insulin sensitivity.

So what we're looking at in this Paris Study is men starting in middle age (44-55) and their IR markers at that age.  And oddly enough, the fewest deaths from all causes occurred within 23 years in the group that had FBG's on the threshold of "prediabetes" diagnosis.  Is it possible that mild hepatic IR is protective after about age 50 for men?  I'm not saying this, just throwing out hypotheticals.

Bottom line:  This is "science writing" number one that fails to support Rosedale's contention that there is no safe level of BG "spike".

To be continued ...

In the meantime, enjoy a hump-day Happy Hour cocktail on me!  Cheers!!