More "Crashing" Diabetes

I found this study tweeted by a follower, @JuhanaHarju, and the ever helpful MM was able to provide me with the full text.

Background: Caloric restriction in obese diabetic patients quickly improves glucose control, independently from weight loss. However, the early effects of a very-low-calorie diet (VLCD) on insulin sensitivity and insulin secretion in morbidly obese patients with type 2 diabetes are still unclear.
Objective: The objective was to study the relative contributions of insulin sensitivity, insulin secretion, or both to improvement in glucose metabolism, after 1 wk of caloric restriction, in severely obese diabetic patients.
Design: Hyperglycemic clamps were performed in 14 severely obese (BMI, in kg/m2: >40) patients with type 2 diabetes in good glucose control (glycated hemoglobin < 7.5%) before and after 7 d of a VLCD (400 kcal/d).
Results: The VLCD caused a 3.22 ± 0.56% weight loss (P < 0.001), 42.0% of which was fat loss, accompanied by decreases in fasting plasma glucose (P < 0.05) and triglycerides (P < 0.01). In parallel, the Disposition Index, which measures the body's capability to dispose of glucose load, increased from 59.0 ± 6.3 to 75.5 ± 6.3 mL⋅ min−1 ⋅ m−2 body surface area (P < 0.01), because of improvements in indexes of both first- and second-phase insulin secretion (P < 0.02), but with no changes in insulin sensitivity (P = 0.33).
Conclusion: The marked improvement in metabolic profile, observed in severely obese patients with type 2 diabetes after a 7-d VLCD, was primarily due to the amelioration of β cell function, whereas no contribution of insulin sensitivity was shown

I've blogged on this topic before:  The Diabetes "Crash" Cure & Pancreatic Fat.   Late last spring, the greater low carb community was all abuzz about that study, where 11 obese diabetics went on a 500 cal/day diet for 8 weeks, lost a modicum of weight and ... saw their diabetes reversed!  What was more puzzling to some, was that even though they returned to a "normal" diet for an additional three months, and even regained some weight,  7 of the 11 remained diabetes free.  The authors of that study were prompted to do their experiment based on the widespread reports of gastric bypass surgery reversing diabetes in short order.  They wanted to see if perhaps it was the rather drastic caloric restriction post-op (prior to major weight loss) that was responsible for the observations separate and apart from the surgical changes.

This study took things a step further.  They took 14 severely obese -- average BMI 44.8, 250 lbs -- patients -- 7M/7F, average age about 60, average duration of diabetes almost 5 yrs, diet or oral meds only (suspended prior to the intervention, no insulin) -- and restricted their intake to only 400 cal/day for one week.  One week you say?  That's right.  Just one week.   This wasn't a weight loss study (they did lose almost 8 lbs total, about 3.5 lbs fat), or any attempt to look at long range effects.   They focused on ß-cell function and measured insulin secretion.  I would note that the subjects were hyperinsulinemic before the study (fasting insulin 76 pmol/L) and remained  so after one week (69 pmol/L - difference not statistically significant from baseline).  The reduction in fasting glucose (7.66 to 6.9 mmol/L) and triglycerides (2.51 to 1.75 mmol/L) was statistically significant.   Before and after the intervention (VLCD = very low calorie diet), insulin response to a hyperglycemic "clamp" was measured (*).

To review, T2's tend to have elevated basal insulin secretion, but lack the AIR = acute insulin response to a glucose challenge -- in other words, they are incapable of mounting the appropriate posprandial insulin spike to an insulinogenic load.  The methods used here are designed to assess the AIR.  The AIR improved significantly -- from -0.63 to 38.2 pmol/L.  Here is the table of all parameters assessed and the "Disposition Index" data for all participants and the mean line (with the error bars).  

The authors posit three potential reasons for the observed improvements in DI: 
  1. improved ß-cell function 
  2. increased sensitivity to insulin
  3. decreased insulin clearance

From the discussion:
We detected signals of improved ß-cell function after the VLCD, whereas the changes in SI and insulin clearance were minimal.  Note that VLCD caused detectable reductions in body weight, fat mass, and triglycerides. Nevertheless, SI was unaffected, apparently in contrast with the improvement of hepatic SI reported by Lim et al (9).  Differences in study subjects (extremely obese compared with overweight/obese patients), in experimental design (hyperglycemic clamp compared with isoglycemic insulin clamp), and in the role played by glucose toxicity (28) might underlie this apparent discrepancy (9).
The Lim et al study referenced above is that "Diabetes Crash Diet" study.  What this study showed is that, in more severely obese subjects, the imposition of a dramatic caloric deficit rather rapidly results in improved insulin secretion comparable to postprandial GSIS (glucose stimulated insulin secretion).  
In our study, we explored first-phase and second-phase insulin secretory responses with both traditional indexes and model based assessments, the latter of which measure insulin secretion rates, not concentrations, and ß-cell sensitivities to glucose, independently of changes in insulin clearance. Although statistical significance was achieved only for AIR and 2ndISR, the general scenario is consistent with improvements in both first and second-phase insulin secretion brought about by the VLCD in only 7 d.   In contrast, no effects were detected in fasting insulin secretion rates, as reflected by insulin and C-peptide concentrations before the hyperglycemic clamp.
Improved ß-cell function despite no effect on the basal situation.   So ... these participants were mounting a better insulin response in the matter of a week, despite their pancreata pumping out roughly as much fat-trapping insidious hormone in the fasted state.  How did they lose over three pounds of body fat doing that?  Can't use the "well their diet was so low calorie it was low in carbs so their insulin went down" -- oh ... and the reminder here is that when these severely obese became diabetic, they no longer mount a considerable postprandial insulin response -- yet they are basal hyperinsulinemic.  How are carbohydrates causing this again??  Can somebody please find me a physicist to explain this to me?  But I digress ...
On the whole, our results are consistent with and extend to morbidly obese diabetic patients the recent findings of Lim et al (9), which were obtained in patients with a BMI of >30, in whom the dramatic improvement in glucose control caused by the VLCD was primarily due to quick improvements in ß-cellfunction and liver SI, with no evident role of peripheral SI.
To sum up:

  • 1 week VLCD in BMI>40 diabetics improved GSIS without improvements in either peripheral or hepatic (liver) sensitivity to insulin
  • 8 weeks VLCD in BMI>30 diabetics dramatically improved glucose homeostasis by improvement in GSIS and hepatic sensitivity to insulin, but without improvements in peripheral SI 

The study authors address the GBP observations:
Among themechanisms underlying remission of type 2 diabetes (12) after bariatric surgery, the potential role played by sudden negative energy balance per se has recently gained renewed attention (29, 30). Very recent data support the hypothesis that caloric restriction is a mediator of early metabolic improvements, after Roux-en-Y gastric bypass (31), via an improvement of hepatic SI, with a consequent reduction in hepatic glucose production, at least in diabetic patients. Our data support the idea that bariatric surgery may improve ß-cell function in patients with type 2 diabetes, also by inducing a negative caloric balance.
In summary, we have reported evidence that short-term caloric restriction per se improves glucose control and b cell function in morbidly obese patients with type 2 diabetes, ie, a class of potential candidates for bariatric surgery. Further research exploring the actual beneficial role of acute caloric restriction per se after bariatric surgery procedures is therefore warranted.

I agree!  Perhaps the true crime of likes of Hope Warshaw, other ADA minions and every doctor besides Wheat Belly is not educating patients and counseling them strongly enough to go on a VLCD.  I gotta say folks, these studies are compelling.  If I were diagnosed, this would be the first thing I'd do is try this.  Instead of packaged liquid formula diets, however, I'd opt for a Lyle McDonald style whole foods PSMF with some supplementation.  Eek ... who'lda thunk Kimmer might just have been on to something "healthy" after all?    Is some "tough love" really the best way to "baby your pancreas"?

Some other thoughts:

Studies like this where individual results are shown, are instructive and remindful (is that a word?) for us that most studies report means, and "significance" referring to a statistical test.  There was a significant statistical improvement in the mean DI, was it significant physiologically?  Perhaps, perhaps not.   But the other thing we do note is that three of the 14 had reduced DI -- are they oddballs or outliers?  No.  But they are the people on the internet who might be reading the study (or worse, a media report) saying to themselves "well I ate 400 cal/day for a week and it didn't improve my insulin secretion so these scientists are quacks for even suggesting such a ridiculously dangerous diet in the first place".  The improvements were quite varied as well.  Two had rather more dramatic improvements than 3-4 others while the remaining 5 or 6 had only slight improvements.  This is not at all surprising to me, and it would be interesting to see how well these improvements correlated with length of diagnosis and/or who among them were those managing their disease through diet alone or requiring oral meds.   

I would like to focus on something else for a moment, however.  A something that the "alternative nutrition" rebels would normally embrace were this the results of a very low carb ad libitum diet (likely around 13-1500 cal/day), but will probably dismiss in this context.  The results indicate that their ß-cells clearly are "not dead yet".   Given that these are severely obese, older diabetics (the women likely postmenopausal) who were not newly diagnosed ... the mere fact that their insulin responses improved *at all* should be encouraging, let alone how rapidly it did.  As with the GPB data, this offers hope ... or it should.  It also demonstrates that the conventional wisdom that by the time you're diagnosed with diabetes your pancreas is all but on its deathbed may well be wrong.  I note that in this case, the conventional wisdom is often stated as "fact" by many in LC circles.   Why the pessimism amongst the great healers?  Will they celebrate?  

So even if only the two most remarkable cases saw improvements worth "writing home about", going on such a diet costs nothing, really, and doesn't involve taking more pills or injections.  Why not embrace this route?  As mentioned in the study discussion, these participants were candidates for GBP surgery -- perhaps the "dangerous" crash diet is an option that needs serious reconsideration.  After all, the surgery is not without considerable risk, and the same dangers associated with the severe calorie restriction will be visited upon the GBP patient for a period of weeks following the surgery.  I so often hear in LC circles the dismay over anyone undergoing GBP because in most cases you have to (a) lose some weight first, and (b) eat what is essentially an LC diet anyway.  So why don't all these folks just go LC and be done with it?   Perhaps where doctors go wrong with their obese patients diagnosed with T2 diabetes is that they don't push a diet that's stringent enough?  

Yes, yes ... it's unsustainable.  It doesn't have to be, apparently.  Imagine transitioning to a more moderate plan, low carb if you like, once the VLCD has done it's trick.  

Today, I stand in awe of my pancreas.

* Both at baseline and at the end of VLCD, a hyperglycemic insulin clamp study was performed in all patients, as previously described (18). All studies were carried out at 0800 after a 12-h overnight fast, while the subjects were lying in bed, and lasted 180 min. In all subjects, 2 intravenous catheters were inserted into an antecubital vein (retrogradely) and into a wrist vein for substance infusion and sampling of arterialized blood, respectively, according to the hot box technique (19). After a 60-min period to establish baseline (260 to 0 min), a hyperglycemic glucose clamp was carried out at zero time for the following 120 min (18). Plasma glucose was measured (glucometer Ascensia Breeze 2; Bayer) at bedside every 2–5 min as needed and was clamped at +7.0 mmol/L (+126 mg/dL). Under these conditions of constant hyperglycemia, the normal b cell secretory response is biphasic with an early burst of insulin release within the first 10 min (first phase), followed by a later monotonically increasing hormone release (second phase) (18). Blood samples for glucose, C-peptide, and insulin measurements were drawn every 2.5 min from 0 to 15 min and every 15 min from 15 to 120 min.

18. DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol 1979;237(suppl 3):E214–23.

19. Copeland KC, Kenney F, Nair K. Heated dorsal hand vein sampling for metabolic studies: a reappraisal. Am J Physiol 1992;263:E1010–4.


bentleyj74 said…
I'm standing with you in awe. I'm seeing a lot of evidence that contrary to the view of the body as a machine that is so easily and permanently "broken".

I'm shocked you didn't title this one "The quick and the dead" though. How often does an opportunity like that land in your lap?
ProudDaddy said…
My doctor called me in to discuss my fasting blood glucose about the same time as the news hit about the original VLCD study (Liu, 2011). I then used my own 8-week version of a 600-cal diet to drop 20 pounds (mostly fat? -- no workout strength loss), and to get off metformin. Some words about safety. A significant percentage of VLCDers after 4 weeks will get gallstones, which can usually be prevented by ensuring 7+ grams of daily fat. As for the sudden heart failures of the 70s "liquid protein" diets, none occurred during the first 8 weeks (and the protein was rather deficient).

A VLCD diet for diabetes is not brand new (Wing, 1991, e.g.), but I think the Liu study was the first to measure pancreatic and hepatic TG and note that the effect occurs long before any significant weight loss. As to said weight loss, in 2008 Lefarrere compared a 1000 cal/day diet vs GBS and extended the dieting until equal weight loss occurred. The results were positive, but not nearly as much as the 600 cal/day study.

Finally, I was so disappointed in Jenny Ruhl's comments about the Liu study, that I removed her from my RSS list. I'm not even going to dignify the "It was the low carbs" crowd by pointing out all the glaring contradictions. Those so inclined don't seem to listen anyway.
Dawn said…
So these studies always make me wonder: is the same approach good for diabetics with no weight to lose? I keep thinking that diabetes in the absence of obesity must be a different beast, one way or the other.
ProudDaddy said…
I sometimes wonder if the better question is whether T2D occurs very often when hepatic lipids aren't elevated. As I understand it, even the very skinny can have high liver TG levels. Perhaps CarbSane can elighten us.
"the reminder here is that when these severely obese became diabetic, they no longer mount a considerable postprandial insulin response -- yet they are basal hyperinsulinemic" - So just by cutting calories sufficiently, the energy intake was sufficient for starting to turn the condition around? That's amazing.

So those "Calories Don't Count" Crickets are more off the mark that I realized. If eating way too many calories kicks off everything from diabetes to, I would guess, conditions like metabolic syndrome - thanks to supersize portions, constant snacking & lack of exercise - in some ways it seems to me that those calories are the only thing that count.
Anonymous said…
If the average BMI is 44.8 and the average weight is 250, that would make the average height less than 5'2" with half men and half women. Check those numbers.
Kindke said…
what exactly is it about fasting or VLCD that allows this improvement in beta cell function?
Sue said…
Maybe it's the FFA that get used up on a VLC diet and then didn't interfere with the beta cells.
KitavanEater said…

I was wondering if you have read Hans' interpretation of diabetes? Does it make sense to you? I always found his long articles hard to interpret because of all the studies used.
CarbSane said…
Welcome to this side of the Asylum ProudDaddy! Hope you're well :)

I am going to have to look for/at the Lefarrere article -- link? That was one thing missing from my puzzle when I blogged on Liu -- was it the weight loss, the calorie deficit, or the dramatic calorie deficit. The evidence seems to be mounting that there's something to the last one. Many low carbers in induction likely come pretty darned close w/o realizing it unless they are listening to current "knowledge". You can't put much dressing on a cup or two of greens. Bacon is not as caloric as folks think and when you eat 2 eggs with 2 slices of bacon or even 3 eggs with 3 slices, it's not as much as folks think. Also eat when hungry in those early days had me skipping lots of meals. I wasn't very hungry.

It's too bad about Jenny as I've listened to both of her podcasts with Jimmy and found a lot of what she has to say to be well grounded and very good. Then she goes off on how diabetes research is OK but diet research is total crap (how she makes a distinction I don't know, as there is a ton of overlap) and tells her readers on a whole page devoted to it, that they didn't eat their way to diabetes. Well, diet and activity level DO play into it for the overwhelming majority of T2's. I understand with her rare genetic form of diabetes she suffered with the medical establishment and a proper diagnosis, but even the title of her book and such (what your doctor won't tell you) is unnecessarily derogatory -- as if your doc is hiding some secret??
CarbSane said…
Certainly there are other mechanisms. I think PD may be onto something, that it is more the ectopic -- specifically but not limited to hepatic -- fat accumulation that may be the common link. I think we need less lumping of all conditions where hyperglycemia -- fasting and/or pp -- manifests under so few umbrellas all with the term diabetes in them.
Kindke said…
Which FFA would those be?
CarbSane said…
Interesting thought there FTD ... right? The common meme is "you can get the same thing just eating LC and a ton of calories" -- but (a) in the studies (like Westman), they aren't eating a ton of calories, and (b) the documented results fall far short of the miraculous claims made on the net. Drop the calories like a stone? Drain the "sludge" ...
CarbSane said…
Hey 26, If you click on the table it should enlarge. I converted kg to lbs and checked my math there. It could be that with a sample of only 14, you have just a couple short 300 lbs folks and it throws off the average BMI.
CarbSane said…
Hi KE, Yes, I've read it. Hans has graciously provided me with two of the big review references -- Frayn & Boden's 2011 manifestos. There's a lot there and I am a huge fan of Hans' as he often picks up on different nuances than I do. I've come to the conclusion that I'll have to roll out something in installments or it will never get published ;)
Sue said…
I thought excess FFA can interfere with beta cell function and with a very low calorie dieting the excess FFA is cleared and beta cell function improves.
CarbSane said…
Sue's on it! See my latest post.
ProudDaddy said…
Thanks for the welcome! I am indeed VERY well, thanks significantly (p<.05) to people like you.

You would think that someone who used to write systems software for a living, would know how to embed a link, but alas, I still don't. I'm sure you've found the Laferre study by now, but here's the reference and link anyway:

Effect of Weight Loss by Gastric Bypass Surgery Versus Hypocaloric Diet on Glucose and Incretin Levels in Patients with Type 2 Diabetes, Laferre et al., JCEM, July 2008

Regards, Wayne
ProudDaddy said…
Indeed. I think I, for example, would be diagnosed has having Impaired Fasting Glucose (or some such), as an Oral Glucose Tolerance Test hardly bumps my glucose level at all (<120) before plummeting to fainting levels. Guess this means my beta cells are still quite functional, but my stupid liver insists on pumping out too much glucose (probably from glycogen, not gluconeogenisis).
Frosty said…
Wayne, I'm interested in the 600 kcal VLC you came up with as I know someone that is having difficulty getting her fasting BS down (especially in the morning). Do you have a blog or something that lists what you ate and other details of your program?
ProudDaddy said…
I'm not a blogger, just a old dude interested in living to see his first grandchild. I therefore suggest one is best off listening to people like Paul Jaminet and Evelyn as to something as drastic as a Very Low Calorie Diet. I believe the jury may still be out on whether it might permanently lower one's metabolism (as in yo-yo dieting). I also strongly advise anyone to read the original Lim study that Evelyn covered so well. (The full text is available free).

That said, I'll tell you a little bit about what I did. I am very parsimonious (my wife says, cheap), and since the Optifast balanced liquid diet formula isn't available in the US and Medifast is quite expensive, I used whey and casein protein powders like bodybuilders use with a few grams of coconut oil. This resulted in a higher protein content than the Lim study used. I also did (and still do) serious strength training twice a week, with some biking and sprinting intervals thrown in. I figure that any diet that causes loss of muscle mass is an exercise in futility.

So, read the study, be aware of the dangers, and then decide how to proceed.
Sanjeev said…
you eat Kitavans? Is the meat really cheap or something?

How do you avoid Kuru?
ProudDaddy said…
I also wonder why more isn't being said about glucagon. Beta cells and insulin get all the headlines while the poor alpha cells are practically ignored. One also doesn't hear much about the fact that certain amino acids can be more insulinotropic than sugar. I have read that the reason bodybuilders' blood sugar levels don't drop like a stone is that glucagon is also stimulated. Given this, and my understanding that metformin acts mostly by suppressing glycogenolysis (glucagon's stimulative role), my "diabetes" might possibly be due to overactive alpha cells. Has such a theory ever even been proposed?
Anonymous said…
@KE, thanks for the link.

Hi Evelyn, I was looking into crash dieting for the pancreas (considering my interpretation of diabetes) and found that you already looked into it quite a bit. Thanks for the analysis.
@Sanjeev yeah, they actually have fresh Kitavans on sale this week at the Whole Foods by my place. Tastes like chicken.
CarbSane said…
Hi there! I plan to link to your post when I discuss some of those studies. Have you looked more into the adipose tissue glycogen thing? I'd be very interested in what you've found if you did.
bentleyj74 said…
Get out. Whole foods has SALES!!!?
bentleyj74 said…
Aw shucks. KE doesn't have a blog. :(
KitavanEater said…
Should have thought through my name. I meant to say that I eat a high carbohydrate paleolithic diet. Wish Whole Foods had sales!!
Sanjeev said…
Your name's completely clear KE ...
if you lurk here long you'll see I rarely pass up a chance to make a joke.
Anonymous said…
Helpful information. Lucky me I found your site by accident, and I am shocked why this coincidence did not happened earlier!

I bookmarked it.

Feel free to surf to my website :: diets that really work