Learn more about Peter Attia’s digital program
Once you realize how harmful sugar is (by sugar, of course, I mean sucrose and high fructose corn syrup or HFCS, primarily, but also the whole cast of characters out there like cane sugar, beet sugar, dextrose, corn syrup solids, and others that masquerade as sugar), you inevitably want to understand the impact of substituting non-sugar sweeteners for sugar, should you still desire a sweet taste.
Want more content like this? Check out my article on replacing sugar with allulose and my Ask Me Anything (AMA) podcast on sugar and sugar substitutes.
If you’re not yet convinced sugar is a toxin, it’s probably worth checking out my post, Sugar 101, and the accompanying lecture by Dr. Lustig. Sugar is, tragically, more prevalent in our diets today than we realize – our intake of sugar today is about 400% of what it was in 1970. And it’s not just in the “obvious” places, like candy bars and soda drinks, where sugar is showing up, either. It’s in salad dressings, pasta sauces, cereals, “healthy” sports bars and drinks, low-fat “healthy” yogurt, and most lunch meats, just to name a few places sugar sneaks into our diet.
I know some people have an aversion to aspartame (i.e., Nutrasweet, Equal) over sucrose (i.e., table sugar, sucrose, or HFCS). In other words they think Coke is “better” that Diet Coke because it uses “real” sugar instead of “fake” sugar. If you find yourself in this camp, but you’re now realizing “real” sugar is a toxin, this poses a bit of a dilemma.
There are two things I think about when considering the switch from sugar to non-sugar substitute sweeteners:
- Are non-sugar sweeteners more or less chronically harmful than sugar?
- What are the immediate metabolic impacts of consuming these products, relative to sugar?
Let’s address these questions in order.
Question 1: Are artificial (i.e., non-sugar or substitute) sweeteners more chronically harmful than sucrose/HFCS?
There’s no shortage of fear out there that consuming aspartame, sucralose, or other non-sugar substitute sweeteners will lead to chronic diseases like cancer or heart disease. However, there is no credible evidence of this in humans. One can actually make a convincing case that no substance ingested by humans has been more thoroughly tested by the FDA than aspartame. The former Commissioner of the FDA noted, “Few compounds have withstood such detailed testing and repeated, close scrutiny, and the process through which aspartame has gone should provide the public with additional confidence of its safety.” While it might be the case that you can harm a rat with aspartame, it seems you need to force the rat to eat its bodyweight in aspartame every day for a year to do so (I’m being a bit facetious, but you get the idea). In fact, even water would be harmful to us in the quantities required to render aspartame harmful if we extrapolate from rat studies.
Since its invention/discovery in 1965, there is not a single well-documented case of chronic harm to a human from ingesting aspartame, and prior to its approval for human consumption in the early 1980’s it had been studied in approximately 100 independent studies. A possible exception to this might be in the rare person with phenylketonuria (PKU). Such folks lack an enzyme required to metabolize a breakdown product of aspartame.
So, aside from the rare person with PKU, does this mean aspartame is 100% harmless? Not necessarily. 100% harmless is a pretty high bar. “Harmless,” using air travel as an analogy, is not getting on an airplane at all. Consuming aspartame is more like getting on a commercial airplane – statistically speaking you are very safe, but something bad could happen that we’re not aware of yet. Consuming sugar in the amounts we typically do, by contrast, is downright harmful. “Harmful,” by the air travel analogy, is not only getting on an airplane but skydiving with a poorly-packed parachute – you might make it, but you’re really taking a chance.
As far as other non-sugar substitute sweeteners go (e.g., sucralose, saccharin, stevia, xylitol), the same logic holds except that we don’t have quite as much data on them because most of them (see figure, below, for the most popular ones) haven’t been on our tables quite as long as aspartame. However, to date there are no data linking these substances to the diseases people tend to erroneously link them to in casual conversation.
Question 2: What are the metabolic differences between sugar and non-sugar substitute sweeteners?
The metabolic effects of table sugar (sucrose) and high fructose corn syrup (HFCS) are well understood, so I won’t review them again. If you want a quick review of sugar and why it’s probably as chronically harmful as tobacco, see my previous post on the topic. Also, Dr. Lustig and his colleagues last week published a paper in the journal Nature titled, The toxic truth about sugar, which you may want to check out if you have a subscription to Nature. The press picked this up in spades, also, and here is one such story.
So how do non-sugar substitute sweeteners compare to sucrose/HFCS in the acute or immediate metabolic phase? Most non-sugar sweeteners (e.g., aspartame, saccharin, sucralose, stevia) are much more potent in their sweetness relative to sucrose, and therefore require a fraction of the amount to give the same “sweetness” as sucrose. So for these sweeteners, only a fraction of the substance is required for equal sweetness. That’s why when you look at a can of Diet Coke it has no calories in it. The amount of aspartame that’s used is so small (given its sweetness), it doesn’t even add a calorie worth of energy. Hence, we consume a fraction of them, relative to “real” sugar to get the same sweetness.
Other non-sugar substitute sweeteners, such as alcohol sugars (e.g., xylitol, sorbitol), are not actually sweeter than sucrose, but they have very different metabolic and digestive properties. Furthermore, one actually uses similar amounts of these sweeteners, relative to sucrose (e.g., substituting an alcohol sugar in the place of sucrose occurs at about a one-to-one ratio). In other words, when consuming alcohol sugars you actually ingest non-zero calories of them. This is why you’ll note non-zero amounts of them when you look at the ingredient labels of foods containing them. Even a piece of gum sweetened with alcohol sugars contains 1 to 2 grams per piece. While an excess of alcohol sugars can cause gastrointestinal distress (e.g., if you overdo it on these you can get diarrhea), in most people they do not cause secretion of insulin from the pancreas due to their distinct chemical structure (see figure of their structures, above).
The same is true for the first group of non-sugar substitute sweeteners I mentioned (e.g., aspartame, saccharin, sucralose), with respect to the lack of insulin response. In addition to studies confirming this, I’ve also documented this in myself for xylitol (my personal favorite), aspartame (Equal), and sucralose (Splenda). I cannot speak to the other substitute non-sugar sweeteners in myself, but these three compounds seem to pass through my digestive tract without ever alerting my pancreas (i.e., without stimulating insulin). When I consume these non-sugar sweeteners neither my blood glucose nor insulin levels rise.
I should point out that some people have noted/suggested a cephalic insulin response to non-sugar substitute sweeteners. A cephalic insulin response occurs when the pancreas begins to secrete insulin before the “meal” actually gets into the bloodstream – the usual step required for the pancreas to secrete insulin. In other words, the anticipation of the meal leads to the release of insulin. This has been documented in humans, and a few studies have attempted to elucidate the mechanism indirectly by using various drugs to attempt to block this response. Furthermore, some have suggested that you can still experience the harmful effects of regular soda while consuming an equal amount of diet soda. It’s not clear to me this is true. First, this hypothesis has never been studied rigorously (i.e., prospectively and with random assignment in a controlled setting). Second, if there is some cephalic insulin response to non-sugar sweeteners, it is probably significantly less than that of sugar in both magnitude and duration, based on the studies I’ve read. To reiterate a common theme – this phenomenon is probably minimal in most people but significant in others. When I work with people who seem to be doing everything “right” but can’t seem to make improvements (e.g., fat loss), I will usually suggest removing all non-sugar substitute sweeteners to test this hypothesis.
Lastly, there has been some recent discussion about how diet soda may cause even more harm than regular soda. A few observational studies have commented on this, including a study released last week in the Journal of General Internal Medicine. Due to time and space, I’m not going to comment broadly on this paper in this post (though I will write a great deal more about this sort of study in the future). I do want to make one very important point that is true of virtually every study of this nature: it is impossible to make a correct inference without doing a prospective, random-assignment, controlled trial.
While the authors of this study acknowledge that “further study is warranted,” the lay press picks up the title of this paper: Diet soft drink consumption is associated with an increased risk of vascular events in the Northern Manhattan study, and fails to ask any questions. While I am not trying to be overly critical of the study authors (whom I do not know, either personally or by reputation), I am actually quite critical of the press that like to report on bumper-sticker messages without reading the fine print. Most people (including many policy makers, who are bombarded with this sort of bumper-sticker information) tend to form their opinions based on this sort of information.
A table from this study (Table 2) is shown below. It’s a bit hard to read unless you click on it, which I’d suggest you do to see what I’m talking about. The group of 163 people who consumed diet soft drinks daily had worse clinical outcomes than the group of 1,948 people who consumed no more than 1 diet soft drink per month (all self-reported). That is, the people who drank more diet soda were more likely to have a vascular “event” (on a per person basis). Seems pretty bad, right? Is drinking diet soda actually causing this?
Well, let’s double-click on this question. Note that the people who were consuming daily diet soda (relative to those not) also had a few other factors not working in their favor including higher blood pressure, higher circulating triglycerides, a higher rate of diabetes, higher BMI, lower HDL-C, higher pre-existing vascular disease, a higher rate of metabolic syndrome, and a higher rate of previous cardiac surgery just to name a few. And in many of these factors the difference between the groups was very large (e.g., diabetes, history of peripheral vascular disease). You will also note that both groups of diet soft drink consumers reported between 1,500 and 1,700 calories per day, below the national average, suggesting yet another problem with this sort of study — self-reporting.
The authors try to correct for this obvious shortcoming by employing a statistical technique called “adjustment.” This means you try to “strip out” the differences between groups and see if the effect still holds. I do not want to turn this into a detailed post on elaborate statistics (a topic I greatly enjoy), but it’s REALLY important that you understand why this is a dangerous way to conduct science. The reason this is “dangerous” is that it only proves an association exists, not that there is any causal link between drinking diet soda and getting cardiovascular disease.
If anyone is really interested in the details of this, I actually reached out to my thesis adviser who did his Ph.D. in applied statistics at Berkeley at the age of 20 (and is one of the smartest people I’ve ever met). Just to make sure I hadn’t gone too far off the reservation, I asked him his view. Here was his response:
Do I buy the analysis? The unadjusted data (Table 2) is pretty hard to disagree with. The adjusted results are open to debate, because the conclusions depend on the form of the model that the authors proposed. A Cox proportional hazards model assumes that the predictors enter linearly into the model. This form is chosen for computational and mathematical convenience, not because anybody has a convincing argument for why the model is correct. But it’s the most popular model out there for survival analysis.
Remember also that the study is observational, not a controlled experiment. The authors can’t conclude that the diet soda is the cause of the greater disposition toward a vascular event, only that there is an association.
My translation: Taken as-is (i.e., unadjusted) there is no way to draw any conclusion from this. After statistical adjustment, you might be able to make the case that it’s worth looking into things further, but as it stands there is only an association between consuming diet soda and having a vascular event.
Want another way to think about? Think of a simple (and silly) example: I once read (I’m not making this up) that people with red cars are more likely to get into car accidents. Let’s assume this is true (though I can’t confirm it). Does it mean owning a red car causes you to have a higher chance of getting into a car accident? Or is it more likely that someone who buys a red car may drive in such way that they are more likely to have an accident? My guess is, even if this correlation between car color and accident frequency were true, there is no causal information contained within it.
Yes, it is possible the reporting of all behavior (e.g., intake) was accurate, and yes it is possible that, even when adjusting for these pre-existing differences between the groups, the outcome would have been the same. But is it likely? It is very hard for me to believe this.
There is a reason I refer (only half-jokingly) to observational epidemiology studies as “scientific weapons of mass destruction.” If you remember nothing else I write or say, please remember this: Never confuse association with causation. If we want to definitively know the answer to this question, we need to design a prospective, well-controlled, random-assignment experiment.
So what is the upshot of all of this?
I would argue (along with a legion of others) that once you eliminate sugar from your diet, your cravings for sugar actually vanish. So the question, rather than, Is it ok to consume sugar substitutes?, may actually be: Why do we need things to be sweet in the first place?
I think this is a personal choice, and something worthy of self-experimentation. I know many people who have eliminated everything sweet (both sugar and non-sugar sweeteners) from their diet, and within weeks they completely lose the desire or craving for sweet foods. Others (like me) still like the occasional taste of sweet things. One of my favorite snacks is home-made whip cream (heavy cream whipped with a touch of xylitol). But the point is this: despite occasionally consuming sugar substitutes I’ve really shed my pathologic need for sweet things. There was a day when I needed something sweet with every meal. That’s no longer true. I go days without ingesting anything sweet and don’t miss it. Other days, I feel like having some xylitol-enriched whip cream, or drinking a Diet Dr. Pepper, and I do so. Would I be better off without them? Maybe. But now we’re well past first-, second-, and third-order terms. (For a refresher on the concept of “ordered terms,” check out my post on irisin).
In summary, if you must drink a sweet beverage (or add sweetener to your coffee or tea) you are better off using a substitute for sugar than you are using sugar. But if you want to be really sure, and you want to kick the habit of needing a sweet taste, you’re probably better off avoiding substitute sweeteners altogether. If you want to be 100% safe, drink water. Just don’t make it bottled water (though that’s a whole other story). And don’t fly in airplanes or drive in cars, either.
Photo by Brooke Lark on Unsplash
Peter,
First off, wonderful blog. Your insight & expertise combined with what another commenter called your superb writer’s tone makes for easy to digest (sorry bad pun 😉 ) reading.
First question: Perhaps I missed it somewhere, and I’m quite tired (11:33pm EST and been up since 6am) but how would we go about getting our blood serum insulin levels checked? Is there an at-home device like a blood glucose meter or does it have to be done by a blood lab with a MD’s prescription?
Second question(sorry its a long one): I have recently learned read Gary Taubes books. I began with “Why We Get Fat” and then read “Good Calories, Bad Calories”. Over the last 4 months, with my own body and blood lipid profile labs, I have seen the positive results that come from changing from the Standard American Diet to low-carb diet. My body fat has decreased, my LDL was elevated & now only slightly elevated and my triglycerides are within normal ranges. So would say that I certainly have a good understanding of both the concept and the application of the statement “Carbs drives insulin which drives getting fat”.
That said, I am concerned about my own mother. She is 61 years old and has metabolic syndrome. She is overweight & has been diagnosed as borderline Type-2 diabetic. She was on metformin, but is now off it and can manage without it via diet (avoiding high GI carbs). She is post-menopausal and is on hormone replacement therapy to alleviate the emotional/mental effects of being post-menopausal. Her blood pressure was high but has resisted the doctor’s attempts at putting her on another drug. She had her gallbladder removed about 13 years ago and has to take nexium for heartburn flare-ups after eating too much fats, most often the culprit is beef fat.
She has never been a big fan of junk food but had cut out junk (soda, crackers, cookies, chips, icecream) within the past 5 years. She has also read “Why We Get Fat” and for the past 2 months, she has begun following carb-restriction ideas and no longer eats traditional sources of carbs (bread, pasta, grains, fruit, potatoes). Her observation is that she has not seen any amount of on-the-scale body weight change either up or down. Her clothing is not any tighter or looser.
I am reminded of a certain paragraph in “Good Calories, Bad Calories”.
Chapter 22 THE CARBOHYDRATE HYPOTHESIS, II:
INSULIN:
"All other hormones will work to release fatty acids from the fat tissue, but the ability of these hormones to accomplish this job is suppressed almost entirely by the effect of insulin and blood sugar. These hormones can mobilize fat from the adipose tissue only when insulin levels are low - during starvation, or when the diet being consumed is lacking in carbohydrates. (If insulin levels are high, that implies that there is plenty of carbohydrate fuel available.) In fact, virtually anything that increases the secretion of insulin will also suppress the secretion of hormones that release fat from the fat tissue."So with this knowledge, do you know if Type-2 diabetics have an easier or harder time losing the body fat on a low-carb diet compared to someone who is simply overweight but not insulin resistant? Perhaps I assume too much and perhaps my little knowledge on this topic might be a dangerous thing, but I gather that as a Type-2 diabetic, upon seeing glucose, that her pancreas creates too much insulin (hyperinsulinemia ?). I have no conclusive proof (blood labs, etc.) that would show consistent elevated levels insulin in the blood which would inhibit lypolysis.
I have to guess that she is eating about 30g carbs per day and probably ranges from 20g/day to 40g/day during an average week. Is it possible that some people had achieve a lypolytic or ketogenic state at about 30g carbs/day but do Type-2 diabetics have to shoot for a lower number? Atkins diet book recommends for Phase 1 induction about 20g/day and I as understand it that person must stay below that to be in ketosis.
But could it be that even at 30g carbs/day that it is too much carbs because the pancreas puts out too much insulin? I am wondering if it could be that some Type-2 diabetic’s blood serum levels of insulin remain elevated for long periods of time, much longer than that of a normal person, and this results in no body fat loss. Thus they must be even more strict about their carb-restriction.
Again, thank you for the blog and we all look forward to your work with Mr. Taubes on NuSI.
James
James, to your first question, unfortunately insulin can not be check with a home device the way glucose and ketones can, but it can be easily checked at a lab. I won’t bore you with the reason but it has to do with the chemical technique of quantifying insulin, which is quite difficult, given that it’s a large polypeptide (vs a small molecule like glucose).
To your second question, it’s really hard to answer a specific question about your mom, but I will say that from both the scientific literature and my personal experience, even the most “brittle” diabetic patient (e.g., one on several medications, plus >100 units/day of insulin) can reverse this trend with steady and consistent carbohydrate removal from their diet. Some folks take longer than others. But remember, insulin is a very short-lived hormone. So what insulin you secrete in response to a given meal is gone later. The question is what insulin response are you generating, and over time this can be reversed.
To me it always comes back to getting more data. It’s probably worth asking her doctor to do some tests to see if she is improving her insulin resistance, such as HOMA-IR and OGTT.
I wanted to share my thoughts on this because in my own past experience with going low-carb (~30 grams carbs/day), after the first couple weeks, I stopped losing weight. I was clueless and angry. I now realize that just as much as my body gets fat from carbs, especially refined ones, I am also sensitive to protein.
I weigh myself everyday because I want to ensure that I am in ketosis and losing weight (at 330 lbs, it works out to about 1 pound a day loss). Some days, if I eat too much meat, I have actually gained 3 to 5 pounds overnight (sucks!). Whatever part of that is water-versus-fat, I don’t care. Bottom line is that the effect is observable.
So, I am now making an effort at keeping both my carbs low, but also my protein. Yes, it is annoying because I really don’t yet know what to eat, sometimes, but I muddle through it.
Remember, though, despite less carbs and less protein, I am enjoying my butter, cream, and my (Julia Child-based) hollandaise sauce over my eggs.
There is a lot of trial-by-error here. I wish it weren’t, but it is. Please accept my experience as just that, as I don’t mean to pontificate anything, but all I can say is that I am losing weight.
To the poster who observed that excess protein made him gain: the body can convert protein it doesn’t need for enzymes and structural repair to glucose through the process of gluconeogenesis. And we know what happens when the body has excess glucose.
Hi James:
You can go chat with the post-menopausal in-remission-T2D ladies over on the Atkins forum. As a group they all have your answers (and in fact, some advise against an OTTG and in favor of a HOMA). Here is the collected wisdom they report over there:
“Is it possible that some people had achieve a lypolytic or ketogenic state at about 30g carbs/day but do Type-2 diabetics have to shoot for a lower number?”
Yes. Often. Some, even the renowned Dr. Mary Vernon, live on as little as 12 net carbs a day in order to keep their issues in check. This the trade-off for total remission of T2D. But as a result, they eat well with plenty of green leafy veggies and no more drugs of any kind.
“Atkins diet book recommends for Phase 1 induction about 20g/day and I as understand it that person must stay below that to be in ketosis.”
This is a misconception. Atkins follows the science (of Dr. Phinney) and argues that most people are in ketosis anywhere below 50 total carbs a day, and the active male may be in ketosis at up to 80 grams a day. Too much bother is placed on ketosis for women. Your mom should focus on finding the net carb level that puts her T2D into remission. And then she should stay there! 🙂
“But could it be that even at 30g carbs/day that it is too much carbs because the pancreas puts out too much insulin? I am wondering if it could be that some Type-2 diabetic’s blood serum levels of insulin remain elevated for long periods of time, much longer than that of a normal person, and this results in no body fat loss. Thus they must be even more strict about their carb-restriction.”
Absolutely yes. Totally yes. In fact, weight loss is very slow for post-menopausal women overall. Life and metabolism are different after menopause. Most women are actually poorly educated about the full range of metabolic changes after menopause. It’s a different country, and most doctors who aren’t specialists in post-menopause are also undereducated, frankly. Many will act as if post-menopausal women are suddenly men, but no no no!
T2D is a triple-whammy. It just is, ‘cuz life ain’t fair. So you just have to keep at it, focus on keeping the net carb level at a place where the T2D stays in remission. Some of the ladies on the Atkins board who were quite ill report it taking up to 18-24 months before their insulin resistance fully disappeared. Some then were able to “up” their carbs – but only to 20 net a day, with the majority coming from veggies.
Your mom may just want to spot check a couple of days of calories to ensure she’s around 1800-ish, just to ensure the appetite suppression of low-carb is working for her. But after that she shouldn’t worry about calories at all – focus relentlessly on getting off all meds. There are a few women who even after a year still require a bit of metformin, and a couple on the Atkins forum added the lowest dose of Januvia. These are women who were however 300 lbs+ and severely ill with T2D. Still, even this group reports on the Atkins forum that they have markedly improved their health, have normal sugar numbers, and all are below 200 lbs. They also all exercise for the benefits not of weight loss but for blood chemistry improvement.
The question that rages in this group is “normal weight.” If you’ve been in this group, followed a very strick 12-net-carb Atkins for 2 years or more, you may not reach that pre-menopausal goal weight. You may never be 130 lbs again. You may have to be content at 170.
It’s a problem that the low-carb community offers very little information for this group of women, even tho’ ironically they may be one of the largest segments on the diet. I do wish more focus was offered overall on women of all ages who live low-carb. Our bodies are different than men’s and we really do have a different experience, esp. after age 30 or 35.
Neurosurgeon Russell Blaylock has been a very articulate critic of aspartame. A few examples of his arguments:
https://www.wnho.net/blaylock_blasts_aspartame.htm
https://www.naturalnews.com/020550.html
Putting aside comparisons with sugar toxicity, I wonder if you find Blaylock’s concerns credible.
Hi Bill, yes, I’ve seen much (but not all — there is so much of this work out there) of this work. Certainly there is a chance that Dr. Blaylock’s concerns are valid, but I have yet to see any data to convince me. You probably know by now that I’m pretty dismissive of animal data, and also longitudinal observational data (like the study I present in this post).
I am not suggesting that Dr. Blaylock is not credible or that he’s out there selling books on the topic just to make a buck. I can’t speak to that, given that I don’t know him personally. I would give him the benefit of the doubt, though, as I do most people, so I am sure he this insight reflects his honest assessment of the data. It’s all a question of individual risk tolerance.
I still stand behind my assertion 100% that you are better off drinking a Diet Coke than Coke but, as I said, we don’t yet have enough data to “prove” you can drink an unlimited amount of Diet Coke with no harm. It’s a tough one.
Hi Peter, gee I really hate to see your comment about Dr. Blaylock: “or that he’s just out there selling books on the topic to make a buck.” I don’t think you are intentionally being unkind, Peter, because you’re not that sort of person – but be careful! Once YOU write a book, or make videos, or whatever you decide on, people will say that about you too!!!!!
I’ve even seen it written about Mr. Taubes for heaven sake – don’t people realize he’s writes for a living! But seriously, there is NOTHING wrong with wanting to make money from one’s expertise, while helping people at the same time.
Anyway great post, thanks.
Barbara, I think you may have missed my point. I wrote (or at least I hope I wrote), that I DO NOT think he’s out there trying to make a buck…though, now you’ve got me worried I was typing too quickly. Completely agree, nothing wrong with making money.
Thanks. Great post! Have you ever noticed that when you sweeten whipped cream with a bit of stevia, the leftover doesn’t collapse and separate in the ‘frig like whipped cream sweetened with sugar? 🙂
Xylitol has some antibacterial properties:
https://www.biomedcentral.com/1471-2180/8/45
Marilyn, I’ve never made it with sugar (back when I was a sugar-eater I would not have DARED to eat whipped cream)…
Marilyn, you have LEFT OVERS! Wow.
Correct. the mechanism by which Xylitol (toxic to canines)retards dental caries is this: the mouth flora take in the 5-carbon sugar, can’t process, & die off. Xylitol has no effect on the gut flora, BUT sucraclose does. There is a massive die-off of gut bacteria. The gut bacteria make a lot of your neurotransmitters (serotonin) for one, & hormones.
Peter, u really blew this whole article- it needs to be redone. Just google mercola & sucralose to get the references!
Other than that, take care & happy swimming!
“Peter, u really blew this whole article- it needs to be redone.” Really? I wasn’t writing about gut flora, so I’m a bit confused by your point.
Dr Attia. i like u lot. we may be realted- my grandmother was an Attiyeh. But u r really blowing it on Splenda. The name of this article is “What are the side effects of … sugar substitutes.” They have a side effect on flora. In Finland during WWII, there was a shortage of hops, so they started using a 5 carbon wood alcohol sugar. The data came in that they were having fewer cavities. The bacteria causing the cavities ingested the Xylitol, could not digest it & died. A side effect of changing the flora too much is that yeasts can run rampant. Xylitol does not effect the gut flora, where serotonin is made, and is very active in the immune system, b/ sucralose does! This is a side effect of a sweetner. No? https://articles.mercola.com/sites/articles/archive/2000/12/03/sucralose-dangers.aspx
looking forward to your post on superstarch- the NFL teams have been using it, I understand. Take Care!
Fair point, if we believe this side-effect of sucralose is clinically relevant for fat accumulation (which is my implied “side effect”). Your point is taken, though. Thanks for sharing.
Blend a few drops of culinary lavender oil or ground dried lavender flowers into your Stevia, Splenda etc, and allow to sit in a jar for the flavors to meld for a few days or longer. Use some in your next batch of whipped cream, or try it in puddings or tapiocas, and try to tell me you have not gone to Heaven. Enjoy!
Excellent!
Thanks Peter for an outstanding blog,,,Gary Taubes turned me on and I’m astounded by the great insight and prose from a true professional in the field. Gary needs all the help he can get in spreading the word.
Thanks so much, but it’s safe to say we still need all the help we can get! Thank you for your support.
Great blog write up, I really appreciate the information. I am an admitted Diet Soda addict – I have been living low-carb for over 15 years, but diet soda is a big vice. So these subjects are always personally concerning to me. There was a study that came out last year ago showing diet soda having an effect on kidney cells:
Clin J Am Soc Nephrol. 2011 Jan;6(1):160-6. Epub 2010 Sep 30.
Associations of sugar and artificially sweetened soda with albuminuria and kidney function decline in women.
Lin J, Curhan GC.
This study alone was enough for me to kick the habit for about 8 months. Have you read anything new on aspartame and kidney function? Because my addiction is back, and I drink a lot of diet soda. Thanks for the great blog!
I actually have not read this study, but I will try to check it out when I can climb out from the burial ground of papers I’m drowning in. What was the effect on you when you gave up the habit? What changed?
I really didn’t notice much difference. Initially, I felt like I was dragging because of the lack of caffeine, but that only last 3 weeks or so. I also drank a lot of iced tea to compensate.
I too was once “hooked” on diet soda. I have not had one in over 10 years. Here is a substitute I really enjoy, especially appropriate to mention on this thread about non-sugar sweeteners.
Flavored sparkling water, (Poland Spring is one brand, and there are others,) with stevia (I buy the liquid Whole Foods brand,)is excellent!
Enjoy!
Hello,
I would highly recommend “Zevia” soda. It has zero calories, is made with stevia, comes in 12 flavors, and is incredibly good. For what it’s worth, it is also completely natural. You can buy it by the case (one flavor or a variety pack) on Amazon, and other places, I’m sure.
i believe if you check zevia’s nutrition info you will find that it contains a significant amount of carbs. sadly, it is sweeted with sugar alcohols in addition to the stevia. frowny face.
another incredibly informative post. thank you much for your efforts.
i agree with the relative safety of most artificial sweeteners as you laid out… but there seems to be some mounting evidence against Splenda (google: Splenda Duke study).
and as you know, i am having a great deal of trouble with fat loss despite a near perfect ketogenic diet, and exercise. the latest casualty has been my morning cup of Peets (i get it shipped in from CA!). I tried drinking it black but almost gagged, so i just gave it up completely. Another testament to the side-effects of a ketogenic diet, I find myself not even missing the coffee in the morning, whereas I used to not be able to function without it. Basically, I sleep deeper now and am completely rested when i wake up at 6am. never though that would happen!
thanks again…
Hi Adam, thanks for a great point. If I recall, I think the Duke study was in rats or mice, correct? I’m not suggesting that EVERYTHING studied in rats or mice is not important, but they have to be taken with the proverbial grain of salt. Also, keep in mind the dose response. A lot of the negative effects of things we eat are relative (like carbohydrates!). All that said, I think your approach is completely logical and makes perfect sense. Start changing things around and measuring the outcomes. Keep up the great work.
It’s my understanding that stevia has a long track record (thousands of years) of safe use. So, even though stevia’s technially a sugar substitute, I think it should be considered as separate from wholly synthetic alternatives, e.g., sugar alcohols, aspartame, sucralose. For example, I’ve never heard of any research showing that stevia produces a cephalic insulin response.
Jake, the same case can be made for xylitol, which occurs naturally in fruit. However, it’s still important to think about it for 2 reasons: 1) the process of isolation and purification can add toxins, potentially, or otherwise change the composition in a deleterious manner, and 2) lots of natural things are actually harmful (both acutely or chronically), so just because something is natural, does not ensure it is good for you. Fructose may be the best example of this.
I agree. I tried stevia for the first time (from Trader Joes) which is in a white powder form. I also wondered about the potential hazards of a natural plant now a highly refined form.
I have to admit I really am NOT crazy about the taste, as it is both bitter and and sweet; yet, it works well in “stronger” dishes (like Thai beef salad with lime juice and fish sauce). It is okay with whipped cream over blueberries, but it is a different taste.
I am a little “scared” of the names like xylitol, erithrytol, etc., because they sound so “chemical”. Intellectually, I realize how ridiculous that sounds, but I find that it’s just best for me to avoid all this stuff entirely. I don’t have any sweet tooth now and I don’t want to “toy around” with tempting myself, as I know how easy it can be to start making some of these low-carb recipes and slide back into some bad habits.
“the process of isolation and purification can add toxins, potentially…”
This statement indicates that, to the best of your knowledge, this problem is still theoretical. Regarding fructose, the problem is in the dose (as you pointed out to another commenter). If people ate fructose in the amounts our paleolithic ancestors did, our livers could handle it quite adequately, and we wouldn’t have an epidemic of NAFLD.
Since, as you point out, it takes less stevia than sugar to get the same sweetness effect, moderate consumption seems like a third-order (or lower) contributor to appetite and insulin regulation.
If you are open to stevia safety-wise, I suggest experimenting with it in different drinks and dishes. With some it combines deliciously, (as with flavored sparkling water, which I mentioned above,) with some it tastes terrible.
Eddie, maybe I’ll give it a try, but right now the only thing I like the taste of is xylitol. Sucralose is ok. I don’t really consume much of either, though.
How does one offer to be of assistance with the NuSci project? I thought I saw something on your blog about it, but now I can’t find it.
I continue to tell anyone who will listen (very few people)about your and GT’s work. I wonder if it will start another flamewar if I share the article from Nature? The fallout from the NY Times article went on for weeks!
Alexandra, please send me a personal email/comment in the next month or so as we get closer to it. Hope to be launching in May or so. Thanks so much for your interest in helping. LOTS to do.
Will do!
Hi Peter,
I came across the website sugarstacks a few years ago, and I thought I’d pass it on. The bloggers use sugar cubes to visualize how much sugar is in a given product. https://www.sugarstacks.com/
Thanks Amanda. This is fantastic.
I noticed you left fructose out of the list of sugar sweeteners at the start of the post. Agave nectar is an increasingly popular sweetener and it is almost 100% fructose. It is used to sweeten coconut ice cream that I used to enjoy but now pass-up because of the fructose. Do you have more writing dedicated specifically to addressing fructose? It is my understanding that it doesn’t provoke much of an insulin response, but I assume (I know the old trope about assumptions) you don’t like it in the diet. If so, why are the alcohol sugars any better?
Edmund, I’ve written quite a bit about fructose previously. Most of the harm from sugar is actually from the fructose. If you haven’t already, look at the Sugar 101 post (and definitely watch the video). You’ll learn everything you need to know about fructose. As far as sugar alcohols versus fructose, sugar alcohols are typically far less digestible (i.e., GI losses), and don’t undergo the same metabolic pathways, but you’re right, in this case the insulin spike is not the issue.
Dear Peter
As I’m sure you must be aware, sucrose is 1/2 glucose, 1/2 fructose. It is the glucose that causes the insulin response when sugar is consumed.
Dextrose is the commercial name for glucose. And Dextrose shows up as a major component in Equal and other “artificial” sweeteners, such as “Stevia in the Raw”. This is BAD NEWS and the reason I do not use such sweeteners.
I think you should shine some light on this use of dextrose.
Thanks,
Steve Pehnec
Steve, yes fully aware of the 50/50 split in sucrose and 55/45 in HFCS and the “semantic swap” of glucose to dextrose. And you are correct, some commercial brands of non-sugar sweeteners (such as Equal) actually contain dextrose! All the more reason to read the labels. Dextrose is not found (at least in any great amount) in most diet beverages, though.
Thank you!
I’ve lost a great deal of weight (50 lbs) being low carb for the last 8 months, however I seem to have stalled with about 10-15 lbs to go. The one thing I have not been able to give up is Diet Pepsi. I drink 3-5 cans a day. Other than that, I’m very nearly zero carb with my diet. Maybe it is what’s stalling me. I feel like a need to have it, and maybe that need is tied to insulin in some Pavlovian way, and that insulin is what’s keeping those few lbs on.
Mike, I would definitely suggest you see if you can reduce that (ideally, to near zero). I’ll be very interested in hearing about your results.
Thanks! I’ve been looking forward to this post.
Peter – here’s a question and non-scientific observation.
What is your take on sugar (or artifical sweetener) and “feeling hungry.”
My experience: I have been off all sugar and non-sugar sweeteners for nearly 5 months. Water & coffee with a dab of heavy cream. Very low-carb diet, in full ketosis at least 5 days a week. NEVER get hunger pangs. Two weeks ago, I “craved” a DIET Dr. Pepper soda. I bought one and drank it.
For the remainder of the day – I was starving! In and out of the kitchen opening the fridge, etc. My mind & body was under the control of something else, and I didn’t enjoy the feeling at all. Luckily I didn’t snack on anything, but it was a definitely unusual and abnormal day.
What is your take on that? Can I draw any correlations to that? Or just pure coincidence? I don’t want to experiment with that again, but I will if you suggest it.
Awesome question and very conflicting data. Suggests there is probably some truth to this hypothesis, but that (as usual!) individual variation plays a role. I think you’ve already got an idea of which side you may be on, but it’s good to know that you can test it.
What I’m saying basically is – that since going low carb, my between meal cravings for snacks virtually vanished.
After drinking the diet soda, something screwed with my body, and I became a different person – heavy hunger pangs… urge to snack on something…
If the artificial sweeteners don’t have an effect on insulin like you suggest – what is causing that feeling of hunger, desire to eat?
My comment about non-sugar sweeteners, to be completely accurate, is that many of them do not produce an insulin response in most people. Still likely in some people, with some compounds. On the hunger question, it’s not entirely clear. Hunger is a physiologic response that has not been completely figured out. Certainly possible it’s one of the other hormones (besides insulin) that plays a role, such as leptin. All that said, focus on what matters: YOU have figured out that these things are bad for you. Very important to know.
Thank you!
Again, group experience from the Atkins board – those formerly T2D or borderline report this often. Thus the Atkins community will tell you to just stay away from all sweeteners. And they argue it’s a brain thing – you have been addicted to sugar, and the artificial sweetener revived your brain’s addictive response, with predictable results. 🙂 It’s basically based on Lustig’s remarks at Jumpstart MD.
Hope this helps.
For the people thinking of giving up diet pop, I would suggest you do it asap. I used to drink 2-4 cans a day, then one day I decided to just give it up. The results were, I got past a weight loss plateau and I craved sweets a LOT less. I just don’t miss drinking that type of stuff anymore and it will be 3 years in June. I still crave sweets, but definitely a lot less than before.
BTW, great work Peter. I am surprised with your findings in this post, because I always thought the results of the sweeteners did cause a insulin spike. Good to know now though, but I am not going back to diet soda…I am not going back. 🙂
Fred, really glad to hear it. You did some troubleshooting, had a hypothesis, tested it, fixed it. Great example for us all.
First of all thank you for this great website and the effort you put into helping people. You’re doing amazing work and making a huge difference in people’s lives.
My experience with Diet Coke is the same as Fred’s, the previous commenter. I would just start having all sorts of craves after drinking it, and definitely be eating more on that day and the one after. Based on that, I do believe there are people (e.g. me) for which aspartame causes that insulin spike.
On the other hand, there is this fragment in your article which indicates you might be the holder of some piece of information I am missing. You said: “When I work with people who seem to be doing everything “right” but can’t seem to make improvements (e.g., fat loss), I will usually suggest removing all non-sugar substitute sweeteners to test this hypothesis.”
I think I am doing everything right – I have an average of 20g of carbs per day, between 1 and 1.5g of proteins per kg of weight, rest comes from fat but never ever go above 1800 cal per day (from carbs, fat and protein). I average at about 1600. This might seem very little, but I do weight 53kg and I’ve been getting on with that kind of number for years. A bit of history: I’ve been on low carb for years (no sweets, no pasta, bread, potatoes, etc, in the fruits department just berries and the occasional small orange etc) and I oscillate between 51 and 54kg. Again, this might seem very little, but I’m one of those persons with really thin bones. One year ago I had my body fat percentage measured and it was 28%, so there’s plenty of stuff to shed off.
Since the start of the new year, I decided to go for the full ketosis… and it sort of worked, in the sense that I lost about 2.5kg, but in the past 10 days I actually gained about 1.2kg back… I do still eat 100g of berries every 10 days or so and have the occasional square of 90% cocoa chocolate (never more than 2 a day)… but other than that I can’t really think of anything else which I eat and contains sugar. Besides the small weight loss, the effects are great, I do feel so much better – no cravings – and most importantly a much much clearer mind (memory and concentration).
Though, this sort of went away in the past days and I actually felt very hungry (which felt quite surprising) and, as mentioned, gained weight, what else should I be paying attention to?
Christina, thanks so much for your comment about the blog. I don’t disagree with you and Fred. Personally, though I don’t drink more than 1 diet coke or diet dr. pepper every 2 weeks or so, I don’t notice the effects many of you have noted. I completely accept that many folks experience them, though, which is why I always come back to the idea of self-experimentation. I like what you’ve done, which is try to tie cause to effect for your eating habits.
With respect to your question, it’s really tough, if not impossible, for me to troubleshoot this way. For me to provide meaningful guidance (as I do for folks I coach) I need lots of time and data. A question I would ask you is if you are able to confirm the extent to which you are in ketosis (by blood measurement of beta-hydroxybutyrate)? I find this helpful when I want to assess the impact of different foods I eat.
You’re welcome but again you’re the person who deserves the thank you :).
Troubleshooting wise, no, no, I wasn’t looking for targeted guidance, it was just a case of me falling down the me-me-Me! well :). And insisted on the details to ensure I communicate properly that I am doing things right.
I wanted to know if there are general things that might affect fat loss like, for example (I don’t know if these are actually true, I’m just trying to illustrate a point), coffee consumption or having one big meal instead of 3 smaller ones, things of that sort, that you might have noticed in other people.
About ketosis – I didn’t have any blood tests, I thought I was in it when I stopped having craves and feeling hungry and now I think I am out of it as I have some craves and feel quite hungry at times.
I followed your advice and looked back on my food journal and have identified few suspects: coffee consumption increase, protein consumption increase (I’ve been on the upper limit lately), I’ll change those and conclude based on what happens. Trial and error, I guess that’s the way.
I will reply to my own thread, hoping this will benefit some of the readers:
1) measured ketone levels with Ketostix – I am above 0.8 but less than 1.6 (in g/L).
2) removed most of the cheese and all coffee from the diet, increased meat consumption, introduced virgin coconut oil and 1 little lettuce with every meal => almost got rid of cravings and obtained GI regularity, better sleep. I blame almost all the problems on cheese, but because I’ve done too much at once, can’t be sure.
3) I’m retaining lots of water. I do have about 1 teaspoon of salt per day. I tried increasing, but that seemed to make matters worse: felt lightheaded and weak and got really bloated. I still get headaches every now and then. So there’s still stuff to correct in this area. Not sure how, yet.
On the other hand, no weight loss yet. Actually, I am now one pound heavier since my last post. (On a humorous note, I need to go out of my way to get into some jeans that were ok a month ago). I just hope it’s the water and there’s a way to fix it. Still, I will keep on tweaking on this for another month and decide then.
Great article, Peter!
Quick N=1 thought:
I’ve been off and on ketogenic diets for the better part of 10 years and have found them to be the only consistent way for me to lose fat and keep it off while at the same time feeling great.
Over the last few years I was finding it harder and harder to enter ketosis and stay there. It was really frustrating bc. I was doing everything exactly as I had done in the past and was being quite diligent about it.
As a last ditch effort, I cut out Diet Coke from my diet (I had been averaging 1-4 per day) and it turned out to be the magic bullet I was looking for. Entering and maintaining ketosis became as simple as it had been many years ago. It was hard to do at first bc. I really love that stuff!
In my online research, I have found a number of similar stories on low-carb blogs and the like. In terms of possible causes, the best that I can glean is that:
1: My insulin or blood sugar was ever so slightly raised by the consistent doses of aspartame…enough to make ketosis harder to enter and maintain.
2: The relatively large amount of citric acid in in Diet Coke was repeatedly messing with the natural KREBS cycle in my body and preventing the induction of ketosis from taking hold.
My best guess would be #2. This is firstly because I must defer to the research regarding aspartame (which you have very thoroughly laid out above) and secondly, I have noticed after a few rounds of ketogenic dieting recently that once I have maintained strong ketosis for a few weeks in a row, drinking a Diet Coke from time to time doesn’t seem to throw me out anymore. I just really need to get through those 2-3 weeks of adjustment to ketosis.
Anyway, just my two cents. Thanks for all of the great posts and please keep them coming!
Excellent insight and observations, Ben. Thanks so much for sharing them with us.
Pet peeve: if you have 400% more of something, you have 5 times what you started with. Current sugar intake is 4 times the intake compared to the 1970’s; consequently, we have 300% more of it.
Good blog. enjoy the work you put into writing your entries.
If we consumed X in 1970 and we consume 4X today is this not a 400% increase? That is, X times 400% = 4X, right?
David is right semantically. The increase is what is additional. We are consuming 400% of what we consumed in 1970, that is not the same as a 400% increase.
Jack, David – I stand corrected! I meant to imply that we consume 400% of what we consumed in 1970. Thanks for correcting.
sorry – didn’t mean to distract and take away from your excellent blog. But since you responded . . . if we consume 100 of something, and now consume 200 of that thing, our consumption increased by 100%, or 2x. The intake has not increased by 200%, but rather the total intake is now 2x. Continuing with this, if we now consume 400, it’s a 300% increase (not a 400% increase)
Hi David,
I really appreciate your explanation of percent versus “2x”, as I now “get it”!!!
Thanks for this terrific post and this terrific blog. I am on a LCHF diet since early September, and have lost 64 lbs, just over halfway to my target weight of 184 (corresponding to BMI of 25). More importantly, I am no longer taking Novolog, Lantus, nor Avandia and am able to get the best bg control in many years with only Metformin.
One thing I have not given up is artificial sweeteners, though many people continue to encourage me to do so. When possible I use the EZ-Sweetz liquid version of Splenda so I can avoid the carbs in whatever filler they put in the powdered Splenda packs. It never affects my blood sugar no matter when I have it and I don’t know where people get this idea. I do not use the sugar alcohols, though, because they create an er, explosive problem with me.
Who knows, maybe there will prove to be something wrong with the sweetener, but I find it no more likely than something be proven wrong with any number of things the self-appointed experts who tell me what I shouldn’t eat, are themselves eating.
Jack, first off CONGRATULATIONS on some exceptional progress, not just on the fat loss, but on the insulin requirement. I think your instincts are right to try weaning off the artificial sweeteners (rather than substituting the laxatives).