Fasting Insulin & Weight Loss II


I finally found a study that demonstrates a correlation between fasting insulin levels and weight loss.  In this study, they manipulated fasting insulin levels through macronutrient composition an diets administered at various caloric levels.   Below is the scatter plot of the data for all subjects:  

Horizontal axis = change in fasting insulin level between time points
Vertical axis = change in weight 


See the correlation?
Neither do I.

So many studies these days begin with correlative analyses of various variables obtained in various studies.  But if the investigator sees a scatter plot like above - a blind person throwing darts at a target might yield a tighter cluster! - they move on.  Because there really is nothing to see here!

It is bad enough that if the plot had looked more like the one below, one would see all manner of causation assigned to the correlation seen in the scatter plot below.   


At least we see a first quadrant trend towards a positive correlation (increase in horizontal axis → corresponds to increase on the vertical axis ↑).  But when there's not even a correlation to be seen on the scatterplot from my study.  It's so "all over the map" as to require +/- on both axes, or all 4 quadrants with a smattering of points to be had in each.

Yes.  This is the Grey & Kipnis data re-plotted (roughly estimated from poor quality plots but makes no difference!) .   




Comments

Christian said…
"See the correlation?"

Why should there be a correlation?
Stargazey said…
Random thoughts:

1. Although these women (13-30 years old) were obese, judging from their fasting blood glucose values and their insulin AUCs, they were probably not insulin resistant.

2. During the course of the study, all but one maintained weight on an isocaloric diet and all lost weight steadily on a 1500 calorie diet. In people who do not have insulin resistance, this is not surprising. (What's surprising is that all but one of them appear to have complied precisely with the study protocol regarding food intake.)

3. On the isocaloric diet, women who did not lose weight, nevertheless decreased their plasma insulin about 50% during the low-CHO phase of the study.

4. The contrasts were not quite as stark in the hypocaloric patients, but it is interesting that insulin levels generally increased from the low-CHO to the second high-CHO phase, even though more weight had been lost by the second high-CHO phase.

5. This suggests, as the authors state in their introduction, that the hyperinsulinemia of obesity may be a consequence of dietary factors rather than exclusively a secondary response to insulin antagonism.
CarbSane,
Don't know if you caught the footnotes in Taubes' "Why We Get Fat" chapter 11, A Primer on the Regulation of Fat:
"A hormone...known as acylation stimulating protein [ASP] is almost assuredly an INSIGNIFICANT exception. It is secreted by the fat tissue itself, a process that is regulated at least in part by insulin."
-What do you make of that?

So a few days ago I had a Dr. visit and found out that I weighed in at 5'9", 130 lbs. In high-school I wrestled at 152 lbs, so that puts me at least 20 lbs. less than what might be considered by "natural"(?) weight.
I find my experience contradicts Taubes' argument that weight-loss is not determined by choice of calorie restriction, and only regulation of insulin/carbohydrates is what matters. Yes some are more predisposed to higher BF% than others, but at the end of the day, it's the choice of the individual whether they are going to eat that last 200 calories and go over, or NOT eat that, and be in negative energy balance.
CarbSane said…
@Christian: There should be if, as some claim, fasting insulin levels have any bearing on weight loss.
CarbSane said…
Wow BE, you just caught Gary being disingenuous red handed again! He didn't acknowledge even the existence of any other hormone in his Jimmy interview. No wonder I get so confused!
CarbSane said…
@Stargazey: Given this was a met ward study, I'm not surprised by the compliance with protocol. Elevated FBG is due to IR of the liver and being young they may not yet be exhibiting this yet. I have some thoughts on why the basal insulin levels varied so with these extreme diets (75% carb for 1500 cal is almost 300g, and for the biggest eater of the maintaining group, isocaloric diet would have contained almost 800g carb!). I'm going to have to blog on this separately. This study is what I'm mulling over in conjunction with G&K: http://carbsanity.blogspot.com/2011/01/hyperinsulinemia-hyperglycemia-and.html
Basically I think these ladies had a degree of peripheral IR and perhaps impaired insulin disposal to where postprandial insulin contributed to basal levels.
Bill said…
Hi CS,
was just listening to your podcast interview with Jimmy Moore and at one point you mentioned: time and time again you saw failure of a meal to suppress NEFA was one of the 'earliest' indicators of insulin resistance (the “fat fails first”). I’m very interested in why you said “earliest.” This would be most obviously demonstrated in a study where subjects exhibited impaired NEFA suppression despite normal glycemia in response to a glucose tolerance test, but I haven’t been able to locate any such studies. Would you please direct me to a study that supports "earliest?"
Thanks, Bill
CarbSane said…
Hi Bill, I was referring to articles like the Frayn one in this post, and the wording in the article in this post. Then there's LynMarie's Fat Fails First? post and references. These are but a handful of study after study that I read where the progression of pathologic insulin resistance was traced to a breakdown in fat tissue function - most specifically excessive NEFA release and/or failure to trap dietary fatty acids.
Bill said…
Hi CS, thanks for the quick reply, there was a lot of great information in the links you provided (and refs therein). But what about a study that shows impaired NEFA suppression despite normal glycemia during an OGTT (in predisposed individuals?). Wouldn’t this be the best (or only) type of study that could really address the question?
Anonymous said…
Nice correlation. Thanks for the great resource!! Keep it up.
CarbSane said…
Hi Bill, I'm thinking that the statements so prevalent in the literature are based on clinical observations.
CarbSane said…
Welcome calotren, and Thanks!
Bill said…
how would arrange the following events in a timeline?

High fasting glucose
High fasting NEFA
Impaired glucose tolerance
Impaired FFA suppression
Obesity
CarbSane said…
In most:

obesity
impaired ffa suppression/high fasting nefa
impaired glucose tolerance/high fasting glucose

It seems that the FFA disturbances occur before the *pathologic* glycemic problems. This is not to say that a single high fat or high sugar meal can't elicit a transient IR and IGT.

It also seems the IGT/ppFFA may manifest themselves a bit sooner as the person is in the overfed "full-up" state.

In some with low fat cell buffering capacity, "obesity" = enough excess weight to exceed capacity.
TheFountainHead said…
William, I do have a straight answer for you although it won't be paraphrased in the answer you asked for.

You seem to understand the concept of "Five Scientists and the Elephant". Neuroendocrinology and cellular biology is fabulous, however much terminology gets people confused when they don't understand how one part plays into the entire scheme.

First look at INFLAMMATION. Not caused by fat, but by an imbalance between T-cells due to naive T-reg cells in the thymus gland. This can be caused by FFA, it can be caused by sex hormones shrinking the gland, etc.

When inflammation occurs, the TGF-beta 1 (in TNF alpha) blocks inositol. Inositol is naturally produced in the body and it's responsible for suppressing insulin levels to normal ranges. When the inositol is blocked, then insulin goes up.

In people such as myself with endocrine issues, TGF-beta 1 increases follastatin. Follastatin enables activin to increase insulin levels.

I'm harping on insulin spikes, not just out of concern for obesity. Insulin blocks IGFBP-1, which is produced by the liver. IGFBP-1 is positively correlated with progestorone development in women and in women it protects us from breast cancer. And in men, IGFBP-1 protects them from prostate cancer.

People harp on both at risk groups for being too obese, insulin is the connection between their insulin resistance/easy weight gain and cancer risks.

BTW, like I mentioned before.. I do have an endocrine issue that involves insulin spikes. All I did was burn off the sugar with cardio exercise and avoid high fructose corn syrup and my glucose is normal. I encourage it to everyone, it's nice to have an option to reduce it without the side effects.