More on that LC v. LF Study: Older study compares actual diets!

Yeah ... I'm off on a tangent again ;-)  In looking at that new LC v. LF study blogged on here and here, I was reminded of a few studies I've looked into previously that actually increase the protein in the diets.  That post is in the works, but in reviewing one study it hit me -- the macro percents were exactly what the recent Swedish study intended to compare!  

This study from Gannon & Nuttall fits the bill perfectly:  Effect of a High-Protein, Low-Carbohydrate Diet on Blood Glucose Control in People With Type 2 Diabetes.  I blogged on this long ago, here.

{Aside:  Gannon & Nuttall are the researchers behind the Dreamfields pasta study one of my detractors used to blast the company and its product (I agree with that, BTW). But that study has since been withdrawn so another one of my detractors would feel justified in calling them liars.  Sigh.}

G&N (or N&G depending on the publication) call their reduced carbohydrate diets LoBAG for Low BioAvailable Glucose.  The LoBAG-20 diet in this study is the lowest carbohydrate diet they've studied (to date, to the best of my knowledge).  The carb/protein/fat percents of the diets were:  


Control:         55/15/30             
LoBAG-20:   20/30/50

Therefore the LoBAG-20 is exactly "high fat" diet from the Swedish study, and the Control diet meets the Swedish recommended "low fat" diet ratios.  

Study participants were otherwise healthy mildly diabetic men   who were untreated by medication who had been weight stable for the three months prior to the study.  Averages:   age = 63, weight = 217, BMI = 31, length of diabetes = 46 mo  (characteristics of each subject are here).  This was, however, not intended to be a weight loss study, and they were counseled not to alter activity level during the study.   In another similarity, two weeks before the study, usual caloric intake was assessed with a three-day food log that included one weekend day.  Now this was a small study -- only 11 subjects started and 8 completed the study -- but it was a crossover study where each participant received both the control and the study diets with a washout period in between.    Because of dropouts, only 2 received the control diet first while 6 received LoBAG first ... significant?  Don't really think so.   The only thing I note that I wish they'd adjusted for, given that fiber has been shown to influence glucose homeostasis, that the LoBAG contained 50% more fiber than the CD.  

In any case, this was what I would call a semi-metabolic ward study.  For the most part they were free-living, but they spent several days in a medical facility and all foods were provided for them during the course of the study.   Each diet phase lasted 5 weeks with a 5 week washout between.  I note that although it was not intended, and the average caloric level of the administered diets was a rather high 2825 cal, and despite adjusting levels every 2-3 days to maintain weight, the men lost an average of 4 lbs during the 5 week diet legs.  Yeah ... I gotta say it, despite the LoBAG containing almost 150g carb and Control a whopping almost 400g carb, these men lost weight.  TWICHOO be damned and Mark ya gotta change that carbohydrate curve!!   

The LC diet wins!!   Surprised?  I'm not.  But even consuming  an average of 142g carb/day (and I note they emphasized that the carbs were "starchy") we see a reduction in %tGHb* from:

9.8 ± 0.5  to  7.6 ± 0.3% 
a change of 2.2%. 

(%tGHb = total glycohemoglobin, according to this Nuttall paper, changes in %tGHb and HbA1c are equivalent/interchangeable).  While the change appears to be leveling off somewhat, it is reasonable to postulate continued reductions over a longer period.  

Now, I'm not going to go all into various suppositions here.  This study demonstrates that what most would consider moderate carbohydrate restriction, dramatically improved glucose homeostasis.  It does not, however, tell us anything about ketogenic diets.  In fact, the researchers measured urinary ketones and there were no differences between diets.  In addition to the glycated hemoglobin, we have 24 hour glucose and insulin profiles, shown below.  (A = control  B = LoBAG)

     FIG. 3.

I note that there is substantial improvement in the FBG for the LoBAG diet.  The lower postprandial glucose levels and insulin "spikes" are to be expected for the considerably reduced glycemic load.  But I also note that fasting insulin levels are unchanged (with either diet).  Therefore, the lower FBG of the LoBAG group can likely be attributed to improved hepatic insulin sensitivity.  What's interesting is if we translate the LoBAG curve up to the CD curve to match up the fasting levels (e.g. baselines), the postprandial glucose spikes account for 8 hrs effect on top of a 24 hour baseline glucose level of a 40-45 mg/dL differential.

In a subsequent study and analysis, these researchers discuss the possible role of IGF-1 in all of this.
We obtained evidence in previous studies that increasing the protein content of the diet resulted in an increase in IGF-I concentration. The current data provide additional support. Increasing the protein content of the diet from 15 to 30% resulted in an ∼35% increase in IGF-I regardless of whether the carbohydrate content was 40% , 20% , or 30% as in the present study. Thus the dietary protein-induced increase in IGF-I is independent of the amount of dietary carbohydrate and fat.
I think the protein content is an unfortunately often overlooked and even downplayed part of low carbohydrate diets.  In the current keto-frenzy amongst dedicated low carbers, protein gets demonized, particularly when the longevity and cancer gangs get into it.  And 210g protein for the LoBAG group ... Wowza!  That's a lot of protein!  (But at just over 2800 cal/day, that's a lot of food too).

In any case, I thought I would share this with you today.  The results of what have been observed (in albeit a small cohort) implementing a diet that is likely FAR more feasible for many people and actually in compliance with the purported dietary interventions.  It's also quite Perfect Health Diet like!  I only imagine the results would be improved were weight loss a goal and a caloric deficit prescribed.  

Comments

ProudDaddy said…
Do I remember correctly that the LoBAG30 diet results were quite similar to LoBAG20?
ProudDaddy said…
Since A1C is generally considered to reflect about a three-month average of glucose levels, my limited math says the results would HAVE to continue downard for quite some time. Do I err?
CarbSane said…
I believe so PDaddy, they discuss this in the article but I'm preoccupied with something else at the moment.
CarbSane said…
Yep - 1.7% considered "similar"
Karen said…
PDaddy I dont think so. If you are talking about them going back to eating like they ate before the glucose would attach theirselves to the red blood cells with rapidity. It may show the same a 3 month A1c test but go up again at a 6 month. If thats what you are asking
ProudDaddy said…
We engineers ('cept Evelyn) sometimes aren't as clear as we should be. What I was alluding to was the fact that in an unchanging environment, A1C won't reach equilibrium until about 3 months have passed. Since the curve was still tending down at 5 weeks in the study, continuing the diet, assuming blood glucose levels weren't rising, should see A1C continuing to decline. It would be like a moving average.
Karen said…
PDaddy, sounds right to me!! sorry. I misunderstood. :)
LeonRover said…
Yes Proud, it is conceptually equivalent to a continuously weighted moving average, with the proviso that absent readings of the inputs to a calculation, there is no estimate of the weighting function.

In addition, while the ADA published - some years ago - a back estimating linear equation for the average equivalent BG, it has little clinical relevance to those who concienciously perform daily BG levels for active daily monitoring of CHO intake. When my stepson became T1, we quickly found that he became more rigourous in keeping to his recommended intakes two weeks before having his A1c test with the endo.

(Not an engineer, but operate from an engineering - and economic forecasting - POV.)