Sunday, November 13, 2016

Sweeteners in the Real World: 12% Increase in GLP-1 and Non-Significant Effects on Insulin W/ Diet Soda From Well-Known Brands and Seltzer + NNS Control - Implications?

Drinking diet soda now and then is certainly not going to harm you.
Everyone knows that "[n]on-nutritive sweeteners (NNS), especially in form of diet soda, have been linked to metabolic derangements (e.g. obesity and diabetes) in epidemiologic studies" (Sylvetsky. 2016). What only SuppVersity readers know is that experimental evidence to prove that the associations between obesity and artificial sweetener consumption from epidemiological studies is not the result of reverse causation, i.e. obese / overweight individuals gravitating towards the consumption of non-nutritively sweetened drinks in the false belief that this alone would help them to lose weight does not exist.
You can learn more about sweeteners at the SuppVersity

Aspartame & Your Microbiome - Not a Problem?

Will Artificial Sweeteners Spike Your Insulin?

Sweetened Drinks Beat Water as Dieting Aid

Chronic Sweeten-er Intake Won't Effect Microbiome

Sucralose Tricks 'Ur Energy Gauge - Implications?

Sweeteners In- crease Sweet- ness Threshold
Against that background, it is nice to see that a recent study by researchers from the NIH, the US National Institutes of Health in Bethesda, did something no previous study has done (which only proves that the currently available real = experimental science on artificial sweeteners is insufficient). In said study, the authors "aimed to test acute metabolic effects of NNS in isolation (water or seltzer) and in diet sodas" (Sylvetsky. 2016).
Table 1: Overview of the ingredients of the "real world" treatment with  Diet Rite Cola™ and Diet Mountain Dew™ - the ingredients I highlighted in yellow may be responsible for their ill effect on insulin (see bottom line).
To do just that, they conducted a four-period, cross-over study at the National Institutes of Health Clinical Center (Bethesda, Maryland). Over the course of the study, thirty healthy adults consumed 355 mL water with 0 mg, 68 mg, 170 mg, and 250 mg sucralose. So far that's nothing other researchers have not done already. What makes the study unique, though, is that the researchers also tested the effects of two different popular diet sodas (ingredients see Table 1) against an unflavored (only sweetened) seltzer control:
  • 355 mL caffeine-free Diet Rite Cola™, Diet Mountain Dew™, which contained 68 mg sucralose and 41 mg acesulfame-potassium (Diet Rite) and 18 mg sucralose, 18 mg acesulfame-potassium and 57 mg aspartame (Mountain Dew), respectively 
  • seltzer water with NNS (containing 68 mg sucralose and 41 mg acesulfame-potassium, equivalent to Diet Rite Cola™) 
The 355 ml test drinks were ingested in a randomized order, prior to an oral glucose tolerance tests. In addition to the blood glucose levels, the scientists tested, recorded and analyzed the "satiety hormone" GLP-1, the insulin response regulator GIP, as well as insulin itself and the levels of the insulin-controlling C-peptide for 130 min.
Figure 1: Serial data from OGTTs. Active (a) glucagon-like-peptide 1 (GLP-1) and (b) gastric inhibitory
peptide (GIP) response to the four different treatments (Sylvetsky. 2016)
In conjunction with the likewise measured glucose absorption rates, the rate of  gastric emptying, and the subjective hunger and satiety ratings of their subjects the scientists did thus generate an, as of now, unique dataset that shows:
  • The diet sodas augment active GLP-1, while the sweetened seltzer has no effect on the GLP-1 response to the oral glucose test; with a sign. different between treatments (!)
  • All four treatments left gastric emptying and satiety measures unaffected.
  • All NNS treatments (Diet Rite, Mountain Dew and sweetened seltzer) increased the total amount of insulin that was produced after the oral glucose test, but this difference is (a) not statistically significant (p = 0.53; see Figure 3) and, more importantly, (b) so small that it is almost certainly physiological irrelevant - at least in the short run.
  • None of the NNS treatments affected the subject's glucose levels significantly. 
Now the first thing that's interesting about these results is that the pre-study with water + sucralose showed no effect of the beverages on the insulin response - a result that implies that (a) the often-criticized sweetener sucralose has no effect on glucose metabolism (not news) and that (b) the observed albeit non-significant differences of the insulin response in the NNS vs. control trial must be due to another sweetener (check out Figure 2, differences in insulin occurred with all three NNS treatments, and not just with the sodas).
Figure 2: The acute glucose, insulin and C-peptide levels didn't differ significantly (p < 0.05)  at any point in the
130 minutes after the ingestion of the four test drinks (Sylvetsky. 2016).
The question is thus: What is it that messes with our insulin response in commercial diet sodas if it's not the allegedly bad sucralose, which has been acquitted in this and previous studies (more evidence I, II, III)? Well, since all three treatments had this effect and only Diet Mountain Dew™ contained aspartame, it's also not possible that the even more derided, yet likewise (metabolically) harmless aspartame of which previous studies even suggest that it lowers insulin - at least during workouts (aspartame's anti-insolinogenic effects) - is the culprit.
Figure 3: Glucose, GLP-1 and insulin AUC in the post-absorptive period (Sylvetsky. 2016); a statistically significant treatment effect was only observed for GLP-1 in the Diet Rite group (p < 0.05).
This leaves us with only one option to explain the (non-significant) difference between the test beverages' effects on insulin: acesulfame-K. The latter is not just a artificial sweetener that was present in all three non-nutritively sweetened (NSS) beverage, but also the only artificial sweetener of which studies indicate that it has insulinogenic effects (Liang. 1987a,b; Malaisse. 1998) - at low doses in vitro, and at higher doses in vivo. In fact, the...
Figure 4: Effect of acesulfame-K on insulin in rats.
"[...] effect of Acesulfame K on insulin secretion [see Figure 4] was similar to that observed by injecting or infusing [both right into the cervical vein and thus bypassing any potential effects of digestion and/or the interaction with sweet receptors in the gut] the same doses of glucose (150 mg/kg) body weight for injection and 20 mg/kg body weight/min for infusion), except that no hyperglycemia was observed with Acesulfame K (Loang. 1987)
In view of the exorbitant amount of acesulfame-K (human equivalent: 16mg/kg) Liang et al. (1987b) had to inject right into the bloodstream of their rats to induce significant in vivo effects on insulin, it appears more than questionable, though, if the oral ingestion 18mg and 41mg of acesulfame-K could, in fact, be responsible for the non-significant increase in insulin observed in the study at hand - and that's in spite of the more recent revelation that acesulfame-K has a disproportionately potent effect on the sweet taste receptors (Dotson. 2008).

So, if it's not a specific sweetener, what is it that triggers the differences - esp. in GLP-1?

Water or Diet Soda - What's the Better Diet Beverage? Study Confirms Fake Sweetness Promotes Weight & Waist Loss, Decreases Hunger, Blood Pressure, Cholesterol & Trigs | more
For insulin, the non-significantly different  response (as physiologically insignificant as it may be) could be triggered by a hitherto not fully understood interaction with sweet taste receptors during the sweetener preload that may influence the response to the subsequent oral glucose tolerance test.

The comparatively large, treatment effect on GLP-1, on the other hand, can't be ascribed solely to the interaction with sweet taste receptors. After all, there's a visible difference between the three NNS treatments, Diet Rite Cola™ (+12%), Diet Mountain Dew™ (+11%) and seltzer + NNS (+3%), with only Diet Rite Cola™ producing statistically significant and Diet Rite™ and Diet Mountain Dew™ potentially relevant increases of the incretin hormone GLP-1 (Waldrop. 2016).
The take-home messages from the study at hand are thus: (a) There's no physiologically relevant acute effect of any of the tested artificial sweeteners and/or their combination on insulin and/or the glycemic response to a standardized glucose load. (b) There's some evidence of a modulatory effect of soda on the GLP-1  response of healthy individuals. The latter may be triggered by either the "taste associated with diet soda" or, which is IMHO more likely, "the effect of other ingredients" (Sylvetsky. 2016). As far as other ingredients are concerned, any of the highlighted ingredients in Table 1 is a potential candidate. After all, both Diet Rite™ and Mountain Due™ produced a relevant increase in GLP-1 (~11-12%), while the otherwise additive-free seltzer with added sweeteners did not. Since none of them produced a measurable reduction in gastric emptying time and the rate of glucose absorption, both highly welcome side effects of the currently used GLP-1 drugs for diabetics (those increase GLP-1 10x more, though) the physiological relevance of the occasional Diet Rite™ or Mountain Dew™ diet soda or any other non-nutritively sweetened beverage is yet almost certainly zero.

Another Suggested Read: "If You Want to Lose Weight and Stave it Off, You'd Better Not Drink Water Instead of Artificially Sweetened Beverages" | read the full SuppVersity Article
This does not mean, however, that we can forget / ignore that the study at hand is an acute response study. So, in spite of the fact that (a) the acute effects on insulin are not statistically significant and (b) both the effects on insulin and GLP-1 are physiologically irrelevant if they occur once in a while. The downstream effect of the chronic ingestion of these or other diet sodas may be different. Accordingly, long-term studies to evaluate, whether the chronic ingestion of any of the sweeteners alone and/or in combination can produce potentially relevant health effects (positive or negative) are still warranted; and I say that even though I know and have written about the existing experimental (vs. epidemiological) evidence from human (not rodent) trials which shows that artificial sweeteners can promote body fat loss and maintenance - not the opposite | Comment!
  • Dotson, Cedrick D., et al. "Bitter taste receptors influence glucose homeostasis." PloS one 3.12 (2008): e3974.
  • Liang, Yin, et al. "The effect of artificial sweetener on insulin secretionII. Stimulation of insulin release from isolated rat islets by Acesulfame K (in vitro experiments)." Hormone and metabolic research 19.07 (1987a): 285-289.
  • Liang, Yin, et al. "The Effect of Artificial Sweetener on Insulin Secretion 1. The Effect of Acesulfame K on Insulin Secretion in the Rat (Studies In Vivo)." Hormone and metabolic research 19.06 (1987b): 233-238.
  • Malaisse, Willy J., et al. "Effects of artificial sweeteners on insulin release and cationic fluxes in rat pancreatic islets." Cellular signaling 10.10 (1998): 727-733.
  • Sylvetsky, et al. "Hormonal responses to non-nutritive sweeteners in water and diet soda." Nutrition & Metabolism 13:71 (2016).
  • Waldrop, Greer, et al. "Incretin-based Therapy in Type 2 Diabetes: An Evidence Based Systematic Review and Meta-analysis." Journal of Diabetes and its Complications (2016).