Saturday, September 1, 2018

Sucralose revisited

Does sucralose (Splenda) have effects that were not reported when it was approved by the FDA in 1998? This research suggests that it does. A summary can be found here.

The developers originally claimed that most of the sucralose was not absorbed from the intestine but was secreted unchanged in the feces. And they said the small amount that was absorbed was secreted unchanged in the urine.

But this research shows that, at least in rats, some of the sucralose is metabolized to produce two new products that haven't been described before. They also show that some of the sucralose accumulates in fatty tissue.

The researcher say that the reason their findings differ from those used to gain FDA approval because they used more sensitive methods.

No one knows what the effect of having sucralose or its metabolites in fatty tissue is, so the researchers suggest that sucralose should be reexamined by the FDA. This is a long process, so in the meantime what should consumers do?

One commenter said that the amount of sucralose given to the rats is the equivalent of a 150-lb person drinking about 90 Diet Cokes in one day, suggesting that this research is irrelevant to real life. However accumulating a tiny amount in the fat tissue every day could build up over time in a person consuming a lot of sucralose.

Sucralose seems to be passed into breast milk, and because growing infants are more susceptible than adults to the effects of any toxins, it would be prudent to find alternative sweeteners during pregnancy and nursing.

Today, as I pointed out recently, we have a choice of many sweeteners. We really don't know the long-term effects of consuming a lot of any of them, even those extracted from the stevia plant. So one approach would be to give up sweet things altogether. Most people wouldn't want to do that.

Another approach would be to eat small amounts of different sweeteners, rather than choosing only one. And downing huge amounts of diet sodas is not a great idea regardless of how they are sweetened.

We should remember that this is only one study, done in rats, and it needs to be replicated before the results are certain. In the meantime, we should be aware of the possible detrimental results from yet-another sweetener.

Friday, August 31, 2018

Are low-carb diets dangerous?

The popular press is having a field day with this studywith headlines like "Is your low-carb diet killing you?"

Well, my exhaustive analysis is as follows:


If you want more details, Zoe Harcomb has done a thorough review.

Tuesday, August 28, 2018


When I was diagnosed with type 2 diabetes in 1996, the number of artificial sweeteners was limited. There was saccharin, of course, which had been discovered in 1878 and offered for sale in the 1880s. But studies in rodents showed that it could cause bladder cancer when used in huge amounts, so some people were wary of it.

Cyclamate, aspartame, and sucralose became popular in the 1960s and 1970s. Then it was discovered that huge amounts of cyclamate produced bladder cancer in rodents, and in 1970 it was banned in the United States. It is still sold in Canada. Both aspartame and sucralose are still on the market in the United States.

Aspartame was controversial, with internet groups claiming all sorts of horrid side effects, although the FDA said it was safe. In 2002 a related sweetener, neotame, was approved.

Another sweetener, acesulfame K, had a slightly bitter taste and was primarily used mixed with other sweeteners.

Today there are a myriad of artificial sweeteners. Stevia, which comes from a South American plant, has been used there for more than 1500 years. One can use the whole leaves, and one summer summer I grew stevia and used some of the leaves, but some people complain of a licorice flavor. The whole or ground leaves have not been certified by the FDA as GRAS (generally recognized as safe) for food products, although they approved it as a supplement..

However, companies have also isolated the sweet compounds from the stevia leaf, mostly one called rebA, or rebaudioside A. These purified compounds have been FDA approved for use in food. RebA can have a bitter taste if one uses too much.

All of these sweeteners are vastly sweeter than table sugar, or sucrose. For example, saccharin is 200 times sweeter, and neotame is more than 7000 times sweeter than sucrose. This means one has to use tiny amount, difficult to measure. And for this reason many companies cut the sweetness of the compounds with substances like maltodextrin. Some even use glucose. When one is on a low-carb diet, the addition of these bulking agents is problematic.

Another problem with these compounds is that because you need so little of them, they won't reduce the freezing point of frozen desserts as sucrose does. So if you try to make ice cream with saccharin or stevia, it will become rock hard when you put it in the freezer. Fran McCullough in her excellent book The Low Carb Cookbook, published in 1997, has several pages devoted too making ice cream that won't turn into icebergs. None of the suggestions like using alcohol or gelatin worked for me,

But today we have more options, with granular sweeteners that are measured just like table sugar. A popular combination is erythritol combined with a sweeter compound.

Erythritol is a sugar alcohol, but unlike sorbitol and similar sugar alcohols, it doesn't cause gas. This is because it is absorbed into the bloodstream and excreted in the urine rather than going down the intestinal tract to be fermented in the colon to produce gas. It's not quite as sweet as sucrose, so most products combine it with one of the supersweet substance mentioned above.

One popular granular sweetener is Truvia, which consists of erythritol and stevia. Another one combines erythritol and lo han guo, which comes from a southeast Asian plant. Like RebA, the extract has been clasified GRAS. A third, Swerve, combines erythritol and "oligosaccharides." They all seem to contain "natural flavorings," without saying what they are.

All three of these products are now available at my local supermarket.

Two new granular sugars have also come on the market. These are allulose and tagatose. Both are isomers of fructose. An isomer is a compound with the same chemical formula but with a slightly different arrangement of atoms so the enzymes that process them may not recognize them.

Allulose is now available at my supermarket, but you can also get it and tagatose online. These product are not cheap. A pound of sucrose costs about 50 cents. A pound of allulose is about $12.

Despite the price, I decided to try the allulose and tagatose. First I tested the effect on BG, using tiny amounts (1 teaspoon) so I wouldn't spike too high with the control (sucrose). Just 1 teaspoon of sucrose made me go up 20 points on an empty stomach first thing in the morning, and I was back to 80 in an hour. One teaspoon of tagatose made me go up 6 points, as did allulose. As my BG usually goes up after I get up even if I don't eat, it's not clear if this is meter variation, effect of getting up, or the allulose/tagatose, but it doesn't look as if these sugars have a big effect. I don't want to test larger amounts.

I didn't test Truvia or Swerve.

Then I made some ice cream. Eureka! The ice cream was delicious and didn't get rock hard in the freezer. Then I made some crustless cheesecake. Same thing. At last, despite the price, we can occasionally have a traditional dessert with very little effect on blood glucose levels.

Today liquid sweeteners are also more easily available. I didn't use to buy sucralose at the grocery store because of the carby bulking agents it contained. A liquid version was available online, but not at my grocery store. Now the local grocery store carries both liquid sucralose (Splenda Zero) and a stevia version (also Splenda Zero but in a green bottle). Sweet Leaf has offered liquid stevia for some time. Of course the liquids contain preservatives, so if you're sensitive to those, this wouldn't be an answer.

Who knows what new products will emerge in the future or what horrendous side effects will be attributed to the existing sweeteners. Ideally, it would be good to not eat anything sweet so one would lose the taste for sugars. But this isn't an ideal world, and I do like whole-milk yogurt with some flavoring and some sweetener. At least we now have more choice.

Odd Logic about Milk

A study presented at ESC Congress 2018 in Munich has concluded that current dietary guidelines are wrong. With the exception of milk, they said there's no relation between dairy consumption and heart disease, and in fact dairy protects against both total mortality and mortality from cerebrovascular causes.

But what is really odd is the final sentence of the press release (the results of the meta-analysis haven't been published yet).

"And given the evidence that milk increases the risk of CHD, it is advisable to drink fat-free or low-fat milk." Huh? A complex food increases the risk of CHD and the authors assume it's the fat in the milk that is causing the problem?

This is an example of researchers having a preconceived notion and then interpreting their results accordingly. They're assuming fat is bad, and thus if milk has deleterious results, it must be because of the fat. This is despite the fact that cheese, which has more fat and less lactose than milk, is protective.

It's faulty logic like this, as well as reliance on inaccurate Food Frequency Questionnaires,  that has made me ignore most nutritional studies.

Monday, August 6, 2018

Self-Monitoring Saves Money in Finland

"Physicians continue to recommend routine self-monitoring of blood glucose for patients with non-insulin treated type 2 diabetes, in spite of its lack of effectiveness (italics added), because they believe it drives the lifestyle changes needed to improve glycemic control." This is from a recent press release from the American Academy of Family Physicians.

This suggests that the "lack of effectiveness" is a fact, not their opinion, despite the fact that they go on to say that there are both proponents and opponents of self-monitoring of blood glucose (SMBG).

I've discussed this before here  and here. Basically, SMBG does little good if patients are given a meter and told to test once a day, usually fasting or if they're told to test more often, usually before meals but not after eating but they're not told what to do with the results.

 Motivated patients have used their meters to determine which foods make their blood glucose (BG) levels increase the most, and they then eliminate these foods or greatly reduce their consumption. When they do, they find great improvement in their measures of control such as the hemoglobin A1c test. However, such success stories are annecdotal, not formal studies.

One common criticism of SMBG is the cost. Thus it is encouraging to see a Finnish study showing that SMBG along with an electronic feedback system reduced total costs by almost 60%. The study was done in a rural area, and the researchers said that reducing travel costs contributed about 20% to the total savings. So the combination of SMBG and electronic feedback reduced total costs about 40%.

The travel costs are the focus of this article, so there's no discusion of the SMBG or the electronic feedback. Nevertheless, it's encouraging to see that some health care systems recognize the importance of SMBG in treating people with type 2 diabetes.

Thursday, July 26, 2018

Nondiabetic blood glucose levels

The internet is buzzing with articles saying that nondiabetics can have high blood glucose (BG) levels after meals but they don't know it. Big news. Most people with diabetes already know that, because most have tested "nondiabetic" friends and relations and have seen some readings that are higher than is considered normal.

I found that some friends would go up to over 160 mg/dL after meals. But unlike most people with diabetes, they'd then return to baseline in a couple of hours. I confess I didn't do exhaustive testing because most people I know don't especially like to have their fingers pricked.

One friend seemed to get diabetes complications without having been diagnosed; for example he had frozen shoulder and trigger finger, both more common in people with diabetes. So I tested him. His fasting number was 71. An hour after a breakfast of white bread, jam, and honey,  his BG was 101, hardly a diabetic number. As it took him about 5 minutes to get up the courage to prick his finger, I didn't do more testing.

But of course all this is annecdotal. Some researchers have also tested people who are considered nondiabetic. Some time ago (2006) a Swedish researcher named Christiansen hooked a bunch of people up with continuous glucose monitors (CGMs) and looked at what their BG levels did over the course of a day. After a high-carbohydrate breakfast, the BG levels of some of them went quite high, although the average was only about 120. There was a lot of variation, some people's BG  hardly budging and others going high. I hope you can see this here (click on thumbnails). The one that looks like tangled shoelaces shows the individual variation. I can't get the lecture to play anymore, or get the URL of single slides.

Here's a German study of nondiabetics.

But the recent research that is causing so much comment was done by Michael Snyder's group at Stanford. Snyder is the geneticist who had his own DNA studied and in the process discovered that an infection had triggered a sharp rise in his BG levels, leading to a diagnosis of type 2 diabetes. Because he was diagnosed at such an early stage, he was able to reverse the condition with diet and exercise, although it took six months for his BG levels to return to normal.

This new research hooked people, mostly nondiabetic according to standard tests, up to CGMs and studied what happened with their BG levels for 2.5 hours in their normal environment. They found that people tended to fall into three clusters, which they named glucotypes. Some had relatively flat curves, which they called L, some had "severe" curves (S), and some had in-between or moderate ones (M), as shown here.

 From Hall et al., PLOS Biology  

They said that recent evidence suggests that glucose variability, more than fasting BG or hemoglobin A1c, predicts the development of cardiovascular disease. And they found that more than 25% of the "normoglycemic" individuals were in the S category.

 Then they studied how a smaller group of people would react to standard meals: cornflakes and milk, a peanut butter sandwich, or a PROBAR protein bar (an odd choice in my opinion; why not use real food?). Oddly, the "protein bar" had less protein than the other meals. They discovered what most people with diabetes already know: a breakfast of cereal and milk will make your BG soar. In this study that happened even with many of the people who were not supposed to be diabetic. Christiansen showed the same thing.

This study produced a lot of data. How much can be used in the real world is unclear. For example, they suggest that using CGMs in people before they're diagnosed with full-blown diabetes would help identify those at the highest risk, so they could make changes before it was too late. This is true. But insurance often won't pay for CGMs even for people with diabetes. Are they apt to pay for them in people without a diagnosis? Maybe some day, but probably not now.

And how many people warned that they were on the path to diabetes would actually do something about it? We all know that overeating tends to make people put on weight. But when offered the choice of cheesecake or an apple, how many people choose the apple?

The data do show clearly how different people vary in their insulin resistance and BG control. An earlier study showed how different people vary in their BG responses to different carbohydrate foods, sometimes showing opposite responses to rice vs bread, for example.

Insulin release is biphasic, and most people with type 2 diabetes lack the rapid phase 1 response that knocks down the BG before it gets too high, but they still have the phase 2 response that kicks in later and lasts as long as the food is being digested. This study did not differentiate (which would not be easy to do).

If you want to slog through the full text, it's available. But I warn you, it's slow going. There is lots of complex statistics, much of the supporting information is in supplementary material, and the figure legends are sometimes on different pages from the relevant text, so you have to go back and forth. But I suppose going from page to page is a lot easier than doing this exhaustive research, so I shouldn't whine.

One interesting sentence in the article is ". . .  American Diabetes Association dietary recommendations are based mainly on reduction of carbohydrate content." When I was diagnosed in 1996 we were told to increase our intake of carbohydrates, and a standard ADA breakfast if you were hospitalized was orange juice, toast and jam, cereal, and skim milk. A friend was told to add raisins to her oatmeal "to get the carb count up." Luckily, I never listened to the ADA and went on a low-carb diet about 6 months after diagnosis.

The essence of this research is that type 2 diabetes is complex, there are different manifestations of the disease, and it would be nice if we could all have the tools to decipher them. Someday we will, but likely not in the near future.