I remember one of my mentors in surgical residency made a very important distinction for me. He said, “Peter, never forget what you are getting paid to do, and what you are doing for free.” You see, there are some aspects of being a surgeon that are not particularly enjoyable. The hours are long. Sometimes you’re asked to intervene in a situation where there is no hope, and you feel you may only make things worse. A lawsuit is just around the corner. But there are many aspects of surgery that are pure bliss. Though I’m no longer a surgeon, some of my fondest memories in life stem from moments there. Anastomosing a transplanted kidney into a patient (especially the renal vein anastomosis, which is the easiest to screw up). Endartarectomizing a plaque-filled carotid artery. Telling a patient and their family that you were able to remove the entire tumor in their colon, and that the lymph nodes were free of tumor.
Want to learn more? Check out our article on replacing sugar with allulose and our Ask Me Anything podcast all about sugar and sugar substitutes.
What my mentor was saying to me was that those moments of pure bliss are not what we’re getting paid for. In fact, we’d probably pay to do them! What we’re getting paid for is the time we have to be away from our family. The long hours, smelly call rooms, and bad hospital food. The cost of medical malpractice insurance.
What does this have to do with the toxicity of sugar? Well, nothing actually. But I constantly remind myself of this when I feel my personal stress and anxiety mounting. The past year has been a whirlwind of kinetic energy that makes my days in residency, 80 to 100 hours of work every week, seem tame and almost boring. Most of what I do today is wonderful, but some is not. I fly about 12,000 to 15,000 miles a month (in coach, no less) and spend about 8 to 10 nights a month in hotels. Red-eyes are a regular part of my existence. My day starts between 4:45 and 5:00 am and goes until 11 pm or later. I can’t put in words how much I detest traveling and being away from my family. So, I guess, that is exactly what I get paid to do. (By the way, do not feel sorry for me. I’m pretty sure all of this is self-imposed, but I still hate it.)
So, do you want to know what part of my role at NuSI gives me bliss that rivals the finest moments of surgery? It’s exactly what I did a couple of weeks ago (and, fortunately, something I get to do often). I spent a day with some of the best and most thoughtful scientists talking about their work and how we can make it better.
Can you imagine (assuming you’re as much a geek as I am) getting to pick the brains of the best scientists for hours on end? Finding out why they are obsessed with the questions they ask? What keeps them up at night? What are the challenges they face? What’s preventing them from resolving uncertainty?
I would pay to do this part of my job. This is the bliss described by Joseph Campbell. And this meeting a few weeks ago was a great example of it. This particular meeting focused on sugar research, specifically the metabolic impact of sucrose, high fructose corn syrup (HFCS), and fructose. If you need a quick refresher on the distinctions, this should help. Spending so much time with this group got me thinking about a broader issue, which is actually the focus of this post: Is sugar toxic?
What does ‘toxic’ mean?
Before we dive into the main focus of this post, we need to get crystal clear on our semantics. Too much tomfoolery has already taken place for the simple reason that many people don’t understand the words they use.
For the purpose of rigor, let’s turn to the pharmacology literature for a clear understanding of toxicity. Even though we all have an understanding of what “toxic” means, let’s be sure we’re understanding the nuance. If you troll the medical textbooks you’ll eventually settle on a definition something like this (from Harrison’s Principles of Internal Medicine):
TOXICITY: The degree to which a substance can harm humans or animals. Acute toxicity involves harmful effects in an organism through a single or short-term exposure. Subchronic toxicity is the ability of a toxic substance to cause effects for more than one year but less than the lifetime of the exposed organism. Chronic toxicity is the ability of a substance or mixture of substances to cause harmful effects over an extended period, usually upon repeated or continuous exposure, sometimes lasting for the entire life of the exposed organism.
The first thing you may notice from this definition is that toxicity is actually subdivided into acute, subchronic, and chronic toxicity, based on how long it takes to progress from exposure to insult and the number of exposures necessary to cause insult. This constitutes what I call:
Important point #1 – don’t confuse acute toxicity (what most people think of) with chronic toxicity.
Acute toxicity and the LD50
An example of acute toxicity is acetaminophen (abbreviated APAP, but commonly referred to by its trade name, Tylenol) overdose which, if significant enough, requires a liver transplant within days to prevent death from fulminant liver failure. (As an aside, this is particularly tragic because the liver, unlike the heart, lungs, and kidneys, can’t be supported extracorporeally; so if a person overdoses, and the liver is irreversibly damaged, they will need a liver transplant within days, or they will die.)
The question, of course, is what dose of APAP is toxic? (In this case, the toxicity is liver failure, which results in near-immediate death.) Enter the LD50. LD50 stands for “lethal dose required to kill 50% of the population.” How is this quantified for a given substance, including APAP? Obviously, we don’t do randomized trials of increasing APAP doses in people until we definitively resolve this. Instead, we (I’m using ‘we’ pretty liberally here – obviously I have never done this) do 3 things typically:
- Carry out the above experiment in animals to accurately estimate LD50 (in the animal);
- Mathematically model the best human data available and try to estimate LD50 (in humans);
- Compare the two estimates.
Not surprisingly, the answer to #1 is usually much higher than the answer to #2. In the case of APAP, the LD50 in rats depends on age, but is probably somewhere between 800 and 1,500 mg/kg, suggesting a 75 kg human would expect toxicity (on average) between 60 and 110 gm (120 to 220 extra strength Tylenol tablets!). Is this likely? Let’s go to step #2. Below are data integrating known human toxicity with a mathematical model to estimate LD50, as a function of APAP dose (x-axis) and death versus survival over time (y-axis). As you can see, this analysis suggests LD50 in humans is closer to 20 gm. (These are data from humans who did not undergo liver transplant, except in the case of the yellow triangles.)
I’ve looked at several other models and they all appear to suggest about the same thing, The LD50 of APAP in humans is about 10 to 20 gm (10,000 to 20,000 mg or 20 to 40 Extra Strength Tylenol tablets), and most sources point to the lower end of that range.
So what’s my point of this? My point is that there is a statistical distribution (see figure below) of the toxicity, which is why it’s called LD50 and not “LD” (which would imply everyone would experience toxicity from the same dose). In other words (let’s simplify and ignore weight differences since this is in mg/kg and just assume I’m talking about a 75 kg human), some people will experience toxicity at 6 gm and others not until 16 gm. In the figure above, you can see one person lived, despite a dose of 40 gm (given that he received the antidote early enough) and another at 25 gm, without antidote.
Important point #2 – there is a spectrum of susceptibility to any toxin.
What about chronic toxicity?
Sticking with APAP as our example, it turns out that much lower doses than the LD50, if taken day after day, are also toxic to the liver. How much lower? As you’ll see below, the answer is highly dependent on the timing of these doses and other host factors. In general, though, some authorities suggest repeated daily doses of more than 6 gm are toxic, while repeated doses below 4 gm daily are rarely toxic. The point is that much smaller doses, if taken repeatedly, are still toxic.
Important point #3 – just because a dose does not result in acute toxicity does not mean it can’t or won’t cause chronic toxicity.
Complicating things a bit further…
There is no reason to expect physiology to be simple or binary, so adding one more layer of nuance to this already-longer-than-you-wanted-to-read-explanation is the following point. Factors such as alcohol consumption, underlying liver disease, viral infections, and genetic susceptibility are highly influential in both acute and chronic toxicity from APAP. This shouldn’t be surprising, of course, though it complicates our discussion. Since APAP taxes hepatocytes (liver cells), taking other drugs that do the same, consuming alcohol (uniquely metabolized by hepatocytes), or having underlying liver disease are invariably going to reduce hepatic reserve. So, an individual’s ability to tolerate APAP is highly dependent on both measureable (e.g., cirrhosis) and idiosyncratic variations.
Important point #4 – host factors play a significant role in susceptibility to toxins.
I would be surprised if anyone reading this has not taken or used APAP (i.e., Tylenol or some generic equivalent). In fact, most of us have experienced great relief from it. That does not change any of the points above.
Important point #5 – the term “toxin” does not imply something is “bad” or universally harmful.
What does APAP have to do with sugar?
There must be some reason I’ve gone through all of this, right? After all, the question of sugar’s toxicity is a somewhat polarizing one. On the one hand, folks like Dr. Rob Lustig have argued that fructose is harmful at the doses most people are consuming it today. On the other hand, folks like Dr. James Rippe have argued the opposite. Having read just about every paper and review article on this topic (I think) over the past year, I can say the debate has many facets, which I’ll outline briefly:
- The PRO sugar folks** argue that sugar, while void of any nutritional value, is no more or less harmful than a calorie of any other nature. In other words, it has no unique metabolically harmful consequence.
- Depending on affiliation, some of the PRO sugar folks debate back and forth about the advantages or disadvantages of sucrose (natural sugar from beets, cane, etc. composed of a linked glucose and fructose molecule) and HFCS (synthesized sugar composed of 55% fructose, 45% glucose mixture). (There may be some merit to this discussion, though it would probably qualify as a “higher order” term. To a first or second order approximation, they are biochemically equivalent.) It appears this debate is a convenient way to avoid really confronting question/point #1. The “natural sugar” producers can point at the corn growers, and vice versa, without really confronting the jugular question. Both of these groups (sugar and corn) downplay research on pure fructose (which is pretty rare in nature and even our current environment), which is a valid point, though a distraction from the issue above.
- The ANTI sugar folks argue that sugar is indeed a “unique” macromolecule distinct from other carbohydrates. Whether solely due to the fructose content, the combination of fructose and glucose, and/or the kinetics of the fructose (i.e., the speed with which the fructose requires hepatic attention when not accompanied by fiber) is a matter of debate and speculation, but those in this camp do agree that sugar is not “just” an empty calorie. The toxicity of sugar, they argue, is primarily related to its hepatic metabolism. Specifically, “excess” (see below) ingestion of fructose increases VLDL production which increases apoB or LDL-P due to greater triglyceride load. Additionally, at least at reasonable doses according to most literature, insulin resistance is worsened which amplifies the harm caused by other foods.
- Even among those who don’t subscribe to the idea that sugar is metabolically unique (and harmful), with or without a dose-effect, some argue that fructose consumption impacts subsequent food consumption in a way that glucose does not. In other words, eating sugar may fail to satiate you and/or make you subsequently hungrier. These data are sometimes confounded, as are many data in this area, by the use of pure fructose, rather than the glucose-fructose mixture found in sugar. Furthermore, evidence is emerging that sugar is addictive, much in the same way that a drug like heroin or cocaine might be, as suggested by functional MRI. So, while folks in this camp argue that sugar per se isn’t harmful, it does make you eat more (sugar and non-sugar, alike), and that is the harm.
- Perhaps the largest debate in this area stems from the dose issue. The PRO sugar folks argue that at the doses most Americans consume sugar, there is no harm (even if there is a theoretical harm at very high doses). The ANTI sugar folks argue that there is a dose-dependent (and probably a context-dependent – e.g., the insulin resistant person vs. the insulin sensitive person) deleterious impact of sugar, AND that current consumptive patients are indeed in this zone of toxicity. (This is probably the most comprehensive single review I’ve read on the entire topic, and I’ve discussed it point-by-point with 2 of the 3 authors.)
- Speaking of the dose issue, no area of this debate (in my opinion) has generated so much controversy. How much sugar, defined as added sweetener (so this does not include the fructose found in fruit, for example) do Americans actually consume? This is important, of course, if we want to know how applicable the above studies are to the question at hand. Where to begin? (This topic alone is really a 3-part blog post.) Estimating how much added sweeteners Americans consume is primarily done via two methods:
- Taking the difference between food availability data and waste data (ERS); or
- Using nutritional surveys (NHANES).
One of my colleagues, Clarke Read, looked into this recently. Here is what he found (this was in response to a recent NY Times article suggesting sugar consumption is less than typically reported):
The adjustment to loss rates was done by RTI International in this report to the USDA Economic Research Service (ERS). Section 4-1, which is an example calculation, is most useful. RTI was asked to calculate estimates of loss, not estimates of consumption, and rather than working down from availability data, they in fact used NHANES 2003-2004 data to estimate consumption, then basically compared this to availability numbers (with a few adjustments) to find amounts of loss. These loss percentages calculated from 2003-2004 then became the standard, and all other consumption data was calculated by applying this % loss to the availability data.
This means that all consumption numbers are effectively derived from NHANES data. This is especially relevant for added sugars. Since NHANES data tracks only consumption of foods, not ingredients, this availability-versus-consumption comparison initially leads to a 96% loss of cane and beet sugar (seen on page 95 of the 2011 USDA document — the “4 percent” referred to in the NY Times article), since NHANES data only reflects sugar added to foods directly, rather than used as an ingredient. The judgment of a panel of experts was then used to determine the percentage of available sugar used as an ingredient, which led to their 34% loss estimate for sugars. For HFCS, which is never consumed as a food and always as an ingredient, they simply gave it the same value as honey (15% loss between availability and consumption). The ERS, however, overruled them (as described in the NY Times article — see the end of the Losses at Consumer Level section in the link for ERS evidence) and used the 34% loss estimate instead.
Page 10 of the 2011 USDA document shows who these 6 experts are. The NY Times article asked 2 of these 6 about these sugar estimates, who don’t recall making them, though it’s implied that they simply don’t remember what happened back in 2008. In other words, while the overall trend in sugar data is determined by availability data (since % loss is assumed to be constant over time), the absolute amount consumed on any given year, as estimated by these loss-adjusted numbers, is entirely dependent on this RTI loss estimate which, for added sugars, is almost entirely dependent on an expert’s estimation. All foods that are primarily consumed as foods rather than as ingredients have consumption levels that are based on an extrapolation of 2003-2004 NHANES data.
Translation: this is a complete cluster. If you triangulate between the ERS and NHANES data, you wind up with an estimate of about 90 pounds of added sweetener per person, per year, or about 110 gm per day which, on average, works out to about 15% of total caloric intake. Of course the actual consumption is much more nuanced (what isn’t?), since consumption varies a lot by age, gender, and socioeconomic status. Furthermore, this estimate doesn’t include the fructose in fruit juice or fruit, though the latter probably isn’t nearly as high, or relevant, as the former.
Another very interesting point uncovered by colleague, Clarke, was that in a 2009 paper in The Journal of Nutrition, Dr. James Rippe (one of the leading proponents of sugar) noted the following:
“…fructose, as a component of the vast majority of caloric sweeteners, is seen to be particularly insidious.”
“It has also been shown to increase uric acid levels, which in turn promotes many of the abnormalities seen in the metabolic syndrome including hypertriglyceridemia.”
“There is considerable evidence of a detrimental effect on metabolic health of excess fructose consumption.”
Whether by accidental omission or otherwise, this paper is not listed on Dr. Rippe’s CV on his website.
**Sadly, it’s difficult to really interpret the data objectively from those in the PRO sugar camp because of the conflicts of interest. Most of the PRO sugar scientists are heavily funded by the sugar industry. For those interested in the historical context on science and the sugar industry, you’ll find this article particularly interesting.
Take home messages
What I find frustrating about this debate is that most people yelling and screaming don’t fully define the terms, perhaps because they don’t appreciate them (forgivable) or because they are trying to mislead others (unforgiveable). The wrong question is being asked. “Is sugar toxic?” is a silly question. Why? Because it lacks context. Is water toxic? Is oxygen toxic? These are equally silly questions, I hope you’ll appreciate. Both oxygen and water are essential for life (sugar, by the way, is not). But both oxygen and water are toxic – yes, lethal – at high enough doses.
What did the APAP example teach us? For starters, don’t confuse acute toxicity with chronic toxicity. Let’s posit that no one has died from acute toxicity due to massive sugar ingestion. But, what about chronic toxicity? Can eating a lot of sugar, over a long enough period of time, kill you (presumably, through a metabolic disease like diabetes, Alzheimer’s disease, cancer, or heart disease)?
Even among a healthy population (i.e., people without overt liver disease), toxicity is a distribution function. What’s toxic to one person may not be toxic to the next. This is true of APAP and it’s true of sugar. It’s true of most things I can think of, actually, including tobacco, alcohol, cocaine, and heroin. Ever wonder why “only” about one in six smokers dies of small cell lung cancer? Maybe it’s the same reason some people (e.g., me) get metabolically deranged from even modest doses of sugar, while others (e.g., Jill, my wife) can mainline the stuff and not appear to suffer many adverse effects.
I posit that Jill and I are both outliers on the distribution of susceptibility, probably driven mostly by genetic difference (rather than, say, exercise as we both exercise a lot). So, I offer you a framework to consider this question. I know some of you just want an answer to the question, Is sugar toxic or not? But I hope this slightly more nuanced response can help you figure out what you should be asking: Are you like me? Like Jill? Or like an Average Joe somewhere in between us?
This is what you will need to figure out on your own. You could play it safe, assume you’re like me and eliminate all sugar from your diet (I eat no more than about 5 gm of sugar per day, almost exclusively in 85%+ dark chocolate – so less than 4 pounds per year). But if you have Jill’s genes, maybe this is overkill. (Though, I would argue, and may do so in a later post, that even Jill has noticed a change in her energy levels and a number of biomarkers by reducing her sugar content somewhat over the past 3 years.)
It’s pretty easy (conceptually) to figure out where you are on this spectrum, but it does involve a few deliberate steps:
- Without making any adjustment in your current eating habits (i.e., fight like hell to avoid the Hawthorne effect), record everything you eat for a week and, using a database like this one (or something fancier like Nutritionist Pro), calculate exactly how much sugar you consume.
- Collect blood work (paying special attention to lipoproteins, triglycerides, glucose, and insulin among other things) and other measurements (e.g., DEXA if you want to assess body composition, waist measurement).
- Get intimately familiar with all the places sugar shows up that may seem counter-intuitive (e.g., “healthy” cereals, sauces, salad dressings, bread). To do this experiment, you need not restrict your complex carbohydrate intake, but you’ll have to substitute products without added sugar. For example, before I was in ketosis but beginning to discover my own susceptibility to sugar, I had to make my own spaghetti sauce from scratch rather than pour it out of jar. I had to make steel cut oatmeal rather than eat Quaker oats. I had to buy bread made with zero sugar (at $7 a loaf!) rather than my usual “whole wheat” bread. You get the idea. It takes time, and you should expect to spend a few extra dollars on food. But, it’s actually possible to find foods that contain minimal to zero added sugar.
- With this information in hand, begin the intervention: aim for a reduction of at least 50% from step #1. (In my first experiment I did 6 days per week of zero sugar, and one day of all I wanted. Ultimately, this became too difficult, and it actually became easier to just go zero every single day.)
- Repeat the measurements (i.e., step #2) after about three months. If you’ve seen minimal effect, assuming you were methodical and consistent, you’re probably in the Jill camp. If you’ve lost fat, seen a reduction in your triglycerides, fasting glucose and insulin levels, increased your HDL-C, and decreased your apoB or LDL-P (assuming you were able to measure them), you’re probably in the Peter camp.
Last point I’ll make, as I suspect at least some of you are wondering. How do two genetic outliers treat their genetic hybrid (i.e., our daughter)? I’ve written about this previously. In short, we limit the sugar she eats in our house, but not so much outside of the house (e.g., birthday parties). I estimate she eats about 25% of the sugar a “normal” kid does. There is no doubt she loves it, and even a week ago when we went on a daddy-daughter date, I got her ice cream with sprinkles for dessert (the irony of me carrying a bowl of sprinkle- and Oreo-covered ice cream through a crowded restaurant was not lost on me).
What does amaze me is how it seems to override her senses. That night, she had a big plate of salad, a bowl of soup, and even a large slice of pizza (if you’re wondering, I had 3 large plates of salad with chicken). She claimed to be absolutely full, and I believed it. But when I brought that ice cream out, it was like she had never seen food in her life. She simply devoured it. The best part? When she looked like she was done, I said, “OK sweetie, looks like you’re done, time to get going…” only to have her say, “No daddy! I’m still finishing the chocolate broth!” She literally left not one drop of melted ice cream (“chocolate broth”) or one single sprinkle or one single crumb of Oreo behind.
This does not seem “normal” to me, and for this reason I guess I refuse to accept, personally, that sugar is just a benign empty calorie. But, one day our daughter will have to decide for herself where she lies on the distribution and how much she cares to do anything about it. Until then, we’ll save the chocolate broth for special (and not too common) treats.
So, in response to the question, “Is sugar toxic?” it seems to me the answer is, “yes, sugar is probably chronically toxic to many people.” And so is water. And so is oxygen. My sincere hope, however, is that you now understand that this is probably the wrong question to be asking. The better question is probably “What dose of sugar can I (or my child) safely tolerate to avoid chronic toxicity?” The goal should be to figure out your toxic dose, then stay well below it. (It’s probably not wise to consume 95% of the toxic dose of APAP just because you have a really bad headache.) What makes this important, of course, is that with water and oxygen, the toxic doses are so far out of the range of what we normally consume, it’s not really necessary to expend much mental energy worrying about the toxicity. But with sugar, at least for many of us, the toxic dose is easy to consume, especially in world where sugar resides in almost everything we eat.
You know, I’ve inadvertantly done this experiment on myself and I am, unfortunately, on the “Peter” end of the spectrum. I guess my question for myself is how low to go in carbohydrate consumption without sacrificing nutrients from vegetables? How high in fat? Or does that even matter? How high in protein since it has the potential to cause an insulin response if you eat too much? I don’t know the answers to these questions. Although I could not figure out what Tylenol had to do with sugar toxicity, I got it by the end. Pretty powerful analogy, Peter. I got the point. Thanks.
Julie, there is little reason to limit non-starchy vegetables and low-GI fruit, even in ketosis. By extension, of course, if not trying to be in ketosis, there is little reason to restrict these at all. If you’re in the Peter camp, though, high GI fruit are probably best if minimized.
Have a look at http://www.dietdoctor.com. In regards to a Low Carb High Fat diet, the answer to your nutrition question – how low to go in carbohydrate consumption without sacrificing nutrients from vegetables?
The answer is: NONE. Veggies will still be the quantifiable bulk of your meal, but avoiding or minimising the ratio of certain carb high root veg. (Simply: if it grows over ground – eat it!, if under ground – don’t!) Your protein intake should probably stay pretty much the same as any normal diet allows, but with one simple difference – you don’t need to cut the fat off your steak! The final component – and the most important will be fats. Eat until your full and supplement your meals with butter, oilive oil, full fat dairy and nuts in moderation. In simplification, it’s a Paleo (stoneage) diet with no added crap. If the food you hold in front of you has a ‘contents’ list – avoid it.
Dr Enfeldt (Diet Doc) has a tonne of analogies and own studies on his page, and I warmly recommend all related content. My dad has followed and lost a tonne of weight on LCHF – he’s now approaching 73 years old and is as healthy (and slim) as ever!
The example you use at the end concerning your daughter eating her ice cream when she was already “full” is a great one. This is an anecdotal piece of evidence almost everyone can related to! I know I have been to the point where I couldn’t possible choke down another piece of meat or take another bite of broccoli, only to destroy some sweet treat moments later.
I look forward to your continuing posts on this issue. Perhaps an important thing to address is how people who can’t afford all the blood work can assess their tolerance to sugar…
I’ve been low-carb for a few years now, and experimented with ketosis last year. Long story short, it seems I’ve fixed myself metabolically. Recently I’ve added some sugar back in my diet and have seen a slight increase in lean muscle. However, even on days I eat sugar, my carbohydrate intake is still below 150g. Even with the muscle gain I’ve seen no fat gain (in fact I look leaner) and my waist size has not increased (even with going from about 155lb to 165lb).
I fast most days, and my eating window is usually 6-10 hours long (I never fast for entire day). My metabolism dosen’t seem to be affected by the increase in sugar consumption (from what I can tell), and my muscle gain has plateaued.
Having said all of that, I don’t consume much added sugar. But, before, I was consuming almost none, and eating only 50-75g of carbs per day (on average). Now I eat 75-150g per day. The one thing I guess I can’t even be somewhat certain about, is whether or not this increase in sugar/carbohydrate consumption has increased my LDL-P…
Interesting question. To your first question, absent the “fancy” blood work, even a simple lipid panel and metabolic panel (very cheap) will give lots of info.
Just wanted to say thanks for the work you do. I really enjoy the detailed explanations to some very interesting and complex (for me anyway) topics. I’ve spent hours on your blog reading and rereading your posts trying to wrap my little brain around it all 😉
Your statement “this does not seem normal” to me, really hit home. While I am in a fact a “Jill” and could handle higher levels of sugar I too choose not to because I feel/function better without. I realize I’m eating more for psychological desire than physiological need when I eat sugary foods.
Working as a personal trainer clients often ask “what do you think about eating____?”
I’ve suggested they’ve ask themselves if that food will pass the “Dessert Test?” Essentially would you still eat that food if you were already full? If you you answer Yes, you should NOT eat that food. If you answer NO, it’s probably an okay “treat.” I think it’s a pretty easy way to find those foods that have the affect that “does not seem normal.”
I’m like your daughter, I could be full and still hammer in the ice cream, chocolate broth and all! Thanks again for all the work you do, your cholesterol post is my go to guide when the cholesterol questions are asked!
Nice way to think about it, Luke. Next time you have some ice cream, just think of the broth!
Nice thoughtful article. Your points are well made. Have you given consideration that each person’s “toxic” level might also be influenced by age and hormonal and other imbalances like excess PUFA ingestion, which might be corrected. Lastly, you say you have talked to a lot of pro sugar scientists. Have you talked to Ray Peat, PhD yet, or read his work ?
Yes, Albie, I do believe that for most people there is a trend toward lowering the “LD50” over time (i.e., as we age). Not sure about the role of PUFA, because the data as so conflicting, but it’s a very important idea to be explored.
Another Great article. Thank you Doctor. Interesting that your distributions are severely right-tail skewed (higher “tolerance”). Clearly I wouldn’t expect the distributions to be “normal” however do you think all the tails are in the same direction?
One other question…You mention that the toxicity of sugar is a polarizing topic among scientists. Is there general agreement anywhere? Any common ground on the role of sugar in the obesity issue emerging?
Good points, James. The figures are just schematics, and I think it’s unlikely the are normal. My guts says they skew the way I’ve showed them. I will say this, MOST scientists who really study this are closer to the ANTI camp than the PRO camp, but the devil is always in the details: dose? empty calorie vs. specific metabolic harm? specifics of liquid sugar?
Nice post. I suspect that refined sugar isn’t the healthiest food to eat in quantity (with large variability in effects between individuals), though I’m not convinced that fruit is fattening or metabolically harmful.
I wanted to expand on the example of your daughter eating ice cream. How would she have reacted to a pile of plain sugar cubes on a plate? If it were me as a child, I might have eaten one or two and then lost interest, despite the fact that I would happily wolf down hundreds of calories of ice cream. The point is that the ice cream is a reward/palatability superstimulus because it combines sugar, fat, calorie density, a pleasant flavor, and a pleasant physical sensation. This nexus of palatability/reward factors, often but not always including sugar, characterizes all hyperpalatable foods. Sugar is a big one, but it doesn’t act alone. I’ve tried fat-free ice cream and it doesn’t do it for me.
Stephan, this is a great point. I think “mode of delivery” really matters. Lumps of sugar…maybe not so palatable. Though, I suspect a big cup of a sugar beverage — pure liquid HFCS — would have been gulped down without regard for actual satiation in the case of my daughter. Personally, as we’ve discussed, the combination that seems to override my ability to control myself is: salt + n-6 PUFA + “roasting” + some sweetness. This combination makes salted, roasted cashews a personal heroin. As you know, I have literally consumed an entire Costco tub (nearly 7,000 kcal) in one sitting. And felt sick doing it! To really put a nuance on this topic, different people seem to have different “heroins.” I know many people who could not care less for nuts, but for me… (the other thing I can’t get enough of, by the way, is dill pickles and olives). Clearly salt is an issue for me. But nothing compares to the cashews, and I believe it’s their sweetness combined with saltiness.
So glad that Stephan chimed in here. I’ve often wondered if the palatability superstimulus, as he describes it, in the combination of sugar + fat blows out dopamine and drives craving in a similar manner as ecstasy on serotonin.
Peter, I’m with you on the Costco tub of Cashews. I could eat them all day. Of course, I could eat a box of Vanilla wafers as well, so I just don’t let them in the house. The personal space I’m in is that I strive to eat LCHF but have occasional lapses in judgment (or temptation by sweets). I think I fall prey to boredom as much as anything at times. Sugary foods definitely have a heroin like effect, although the withdrawal probably isn’t as strong. I do get what I call Sugar headaches when I stop eating it. Cheers, Dave
Interesting thoughts Peter. I also find that salted nuts go down much easier than unsalted nuts. I only keep unsalted nuts in the house now, because I’ll stop eating them when I’m full. For the same reason, I keep roasted cocoa nibs in the house but not chocolate.
Also, I wanted to add that although sugar isn’t the only palatability factor, it is a big one, and very deeply ingrained in the reward/hedonic circuitry. Infants (human and rodent) respond to sugar with stereotypical expressions of pleasure. We are deeply hard-wired to like sugar from birth, and there seems to be brain circuitry specifically dedicated to sugar seeking, as opposed to some other pleasurable things that we have to learn to like over time.
My view is that sugar seeking probably performed a valuable function in the ancestral environment and therefore was strongly selected. Sweetness was associated with fruit, and less frequently, honey. Sweet fruit is usually non-toxic and nutritious, but has a relatively low calorie density so it’s intrinsically hard to overeat. The problem is that today, we add concentrated sugar to all sorts of other foods that have properties that would not be associated with sweetness in the wild. We refine and concentrate sugar, then add it to starch and fat, or drink it as a beverage. This takes the brake off consumption because you skirt around the qualities of fruit that intrinsically limit consumption (fiber, low kcal density, moderate palatability).
Stephan, thanks for weighing in again. This makes a lot of sense, and I’m sure most folks can at least personally related to what you are describing.
When I was a young girl and even one not so young, I would eat just plain sugar cubes. I didn’t stop at one or two. I ate more. I also ate spoonfuls of sugar if there were no cubes. I did at times combine shortening with sugar (sort of like lard frosting). While some children would back away from sugar in pure form after a few cubes , not all would. Palatability only answers the question for some not all. And you may not believe that I actually didn’t like the taste after the first cube or teaspoon but continued to eat until I felt a ping and not from feeling sick
Why did I consume vast amounts of sugar this way (in addition to what is present in foods), I don’t know. I have read and thought about, worked on the whys for a very long time. My obsession with sugar began as a toddler. No one in my family ate sugar like I did. Sweets were present always but not in huge amounts. I am the odd one in my family with a need that feels like an addiction. After many years, I have come to the conclusion that there is a biochemical drive that was present at a very young age. Not an emotional one as so many have tried to make it into. Recently I read where a woman traced her sweets problem back to eating a cookie when she was young and her parents were divorcing. I have a hard time with the idea that a five year old can make the connection .
This post is excellent. Gave me some things to think about. Thanks for taking the time to write it.
Thanks so much, Donna, for your input. Sure makes me wonder about individual variation on this specific variation. I know, personally, I found pure sugar very UN-appetizing, but really loved it in liquid form.
Sugar may not act alone, but it doesn’t need much help from other macronutrients in the case of sodas. A bit of citric acid, CO2 and a pleasant flavour (with the appropriate colour) are usually enough. Could you do this with fat? I can’t think of any commercial examples other than dark chocolate, which is an acquired adult taste.
Just piping up in response to Stephen that while my personal kryptonite these days is any form of ice cream, as a girl I, like Donna above, could and did eat whole bowls of sugar cubes. Even now I kind of get a pleasurable chill up my spine remembering how they melted in my mouth. I’d make myself sick eating them…and then continue on.
Great article, Peter. Thank you so much for this lucid and informative approach to a subject that seems to bring out the knee-jerk armchair nutritionist in everyone I know. 😉
Thanks very much, KimBoo. Glad it was helpful.
Remember with cola beverages they include salt, caffeine and acid flavors to offset the sugar so it doesn’t taste especially sweet, and hops to beer to kill the sweet taste. And some fruit flavored sodas contain brominated vegetable oil. Even Whole Foods, (to give the Devil his due) uses unbrominated flour a small improvement in for example, a cinnamon sticky bun.
Very well done. Pro and Anti-sugar people should be happy, and that’s no easy task 😉
Or both pretty upset. Probably an easier task.
Great article, Peter – thanks! I really appreciate the time you put into writing it, in view of your hectic schedule. (Which is also due to the work you are doing on behalf of all us human beings – so thanks again!)
A couple of items of feedback. I have never used any Tylenol – whether because I am British and live in England, or because I have a strong aversion to using any drugs (except caffeine and alcohol) unless it’s absolutely necessary.
Also, if sugar really is addictive, I wonder how come I (and many others) gradually reduced sugar consumption over the years? When I was a teenager, I remember putting 4-5 teaspoonfuls of sugar into a cup of tea! For the past 30 years, the sweetness of the milk I add has been quite enough.
Incidentally, if you can spare a few minutes to look it over, I would like to know your opinion of this article:
I have read and thoroughly enjoyed many of Dr “Theodore Dalrymple”‘s articles and books, and I understand that he is a trained and experienced doctor of medicine. His statements about drug addiction and the relative ease of withdrawal seem plausible, but I don’t know enough to judge their correctness or otherwise.
Not sure, Tom. I can’t speak to Dalrymple’s thesis. A bit out of my scope.
He is quite right. No-one was ever killed by a psychotic episode either. No-one was ever killed by lots of things that doctors like to treat, and that patients want treated.
Now that I’ve read the entire piece, Dalrymple is demented. I certainly wouldn’t trust him with my own care, as he would be perfectly capable of explaining away any pain or dysfunction I suffered as delusional. I’d be OK as long as I didn’t die.
The only research he cherry picks from the entire field is some tests done on prisoners by one researcher in the 1930s. I’m sure that was fun for all concerned. Opioid withdrawal is manageable if you go to considerable effort, use a high-fat, allergen avoiding diet, and take the right supplements (and Ibogaine, better still), and addicts should never be discouraged from doing it, but with sadists like this around, no wonder most junkies relapse.
It’s known as Paracetamol in the UK Tom, as Wikipedia’s entry for Tylenol says :-
” it is commonly known elsewhere in the world by its international nonproprietary name, paracetamol. Like the words “acetaminophen” and “paracetamol”, the brand name “tylenol” is derived from the chemical name for the compound, N-acetyl-para-aminophenol (APAP)”
Cluster, indeed. You’ve made very good sense, as always, of the current state of the art. I have to add my ‘heck yeah’ to the other observations here, both in myself and others, that a sugar fix seems to over ride all common sense and body feedback on food intake. Not only in the moment of indulgence but in the days following, my normal routine of what/how much/when, which is successful in making me not think of food much, is completely disrupted by my brain poking me with a desire for “something”. It’s also good to see the accumulating acknowledgment in the literature of the relationships among sugar, uric acid and metabolic syndrome. I know of two diabetics recently who’ve had gout attacks for the first time and each of their physicians went the low protein route. We needed to send them back to their doctors with research papers on gout and MetSyn.
What an adorable photo of your family. Looking forward to your TED talk. Saw the mini-preview. It was pretty moving.
Thanks, Lorraine. I think the talk will go up on the TEDMED site June 5 or 6, I’m told.
I can relate to those “moments of pure bliss”. After chiropractic school, I decided to go to medical school (currently in my 4th year) and the hospital can be a frustrating place when I can’t give all of the advice I would like to. On a daily basis, I think back to your speech when you gave the example of “catching the eggs before they hit the ground and instead punching the guy who is throwing the eggs” and it makes me laugh every time. But, on the weekends when I see my few patients (as a chiropractor) and I watch the lightbulbs go off and I get to watch their progress unfold, it’s all worth it. That’s what I live for…and hopefully I can experience more of those moments after finishing school. It’s to help those patients in the hospital that need someone to guide them.
Thank you for posting this. The idea of having a ‘threshold’ has always intrigued me. For example, a group of children will get exposed to Strep; some of them won’t be affected, some will get better with or without treatment, some will develop P.A.N.S. In the last group, did they have something predisposing them to having a lower threshold causing them to react differently to the bacteria causing other symptoms? That’s obviously just one simplified example but it’s an interesting concept.
Thank you for your work; this blog definitely inspires me when I’m at the hospital.
Surena, thanks so much for sharing your experience and story. Very exciting path you’re on. Glad to be a part of your journey.
Wow, Peter, just another brilliantly-written, riveting post! I’ve been re-educating myself through your blog (…after 14 years of ‘formal’ academic/medical training, I–and likely many health care providers–still have so much to (un)learn…). I’ve thoroughly enjoyed every word. Keep it up and thanks for sharing your “bliss” moments with the world!
Thanks so much, Mel. Glad it’s helpful.
My personal “heroin” is white chocolate from Choceur with coconut crunchy bits .. super palatable hyper even..
This is the main point with sugar, it is added to everything to make the item addictive, and it works.. same with salt..
SO the toxic issue is now that we know it has this effect, should there be legislation to regulate it, like cigarettes we know the harm, it is not going to happen on its own as the victim is willing and the producer is mendacious. A higher power will need to be invoked to limit the damage. ( I do not mean the Almighty )
I guess the questions I’d ask before I could contemplate this is:
1. How much more scientific work would need to be done to be as clear as possible on the effects?
2. What is the best model for regulation improving a similar problem?
I am totally against the nanny-state solutions to sugar. Vote sugar out of school lunches or food stamps or stop sugar subsidies? Sure. Regulate “sugar” ala Bloomberg or the like? No. Let’s get the science, educate people, and then they can make their choice. The post is telling us you can figure out on your own if your level of sugar consumption is affecting you and act. Look at the great record of the last 30 years and often continued demonization of cholesterol or saturated fat. Think those regulators are always going to get it right? I’m sure many would have regulated butter or cream or eggs under the guise of preventing heart disease Likewise, my Grandfather (born 1910) always spoke about his father and the knowledge that smoking was bad for your health and the $100 he received for not smoking as a youth. This was not a mystery before the surgeon general’s announcement.
Salt is self limiting. This limit can be overcome by adding sugar and sugar can be added by adding salt. In any event the anti salt campaign is likely to do far more harm than good as people will continue to eat processed food with much added salt and skimp on the iodized salt added at the table. (Not even bringing up the recommended level of salt being less than optimal for health.)
The Tylenol example was very enlightening and made me think that perhaps being in nutritional ketosis for to long a period(years ?) might also not be so great for the liver. After all certain cells can use glucose only such as some of the cells in the brain, the red blood cells, cells in the kidney and retina. All these cells from what I understand use about 120G-150G of glucose a day. So, if there is a need for glucose, maybe eating some means less work for the liver, and maybe that is a good thing if the source of the glucose comes from fruit and maybe some glucose starch.
Thanks for your great work
I don’t know…the amount of glucose needed by the RBC and kidney is very small, easily supplied the liver even when in the strictest ketosis. The brain, of course, does very well on ketones. Maybe even better than glucose. But this is getting away from the point…
The trans-sulfuration pathway is upregulated when carbohydrate is restricted, this is what supplies glutathione, which detoxifies tylenol. On the other hand, glucose-6-phosphate is needed to reduce (recycle) this glutathione.
The liver can do lots of things and it’s always making glucose. Most people need less Tylenol on a low-carb diet, we have to factor that in too. NAC is the antidote to Tylenol (called Parvolex in emergency wards, sometimes reduced glutathione is used), and is usually effective if given quickly enough, and for long enough.
This (I am assuming its use is standard practice in the US) would surely influence the human toxicity stats; unless these numbers only represented cases where patients waited too long, or were denied proper care.
I once brainfarted around with the idea that if you put together an LD50 list of foods/macronutrients/etc sorted from most benign you’d probably come close to a reasonably perfect diet, dunno.
As to the “always room for dessert” issue, it is of course most likely ghrelin we’re talking about, it’s a completely different kind of desire – it’s certainly not hunger. Plenty of times now I’ve done self-tests where I overeat fat (like eating 300+ grams of pure fat from a pork belly) and there comes a very specific point where:
a) you no longer feel like eating – the palatability of the previously delicious meal in front of you is suddenly gone
b) in only 2-3 forkfuls of fat later, the mere sight of the food actually makes you nauseous
At this point just the thought of cutting off another piece of fat almost makes me dry heave, but then the idea of having some chocolate is just fine.
They touch on this here: http://www.psychologytoday.com/blog/cravings/200805/why-theres-always-room-dessert
Interesting point, Ash. Whether ghrelin, or some parasympathetic (i.e., vagal) signal, or even something purely in the CNS, is a hotly debated topic. I know Stephan Guyenet and others have been working on this. It would not surprise me if there existed redundancy in the system (unfortunately).
Not that simple. Vitamins A and D for example are mutually protective of high doses of the other. I think K2 factors in there too along with calcium, IIUC K2 helps send calcium to where it is needed and removes it from where it is not needed.
This happens all over the nutrient spectrum, so LD50 are not a valid approach.
And then there is individual variation. Starch is fairly toxic in fairly low doses by SAD standards for many people particularly when they are not exercising.
Oh Baby! Nutrition is hard, if you approach it from a nutrient viewpoint.
Another great article, Peter! I’ve been so inspired by your blog since i discovered it a few weeks ago and you succeeded in motivating me to cut out grains and (my own home-made) baked goods from my diet. It’s only been a week, but I’m down to <150gm a day of carbs and I feel so AWAKE every day. I've struggled with daytime sleepiness since i was a teenager, and I was recently subjected to an overnight sleep study where they found nothing physically wrong with me but gave me a CPAP machine anyways. the CPAP machine made no difference and I still felt like falling asleep every afternoon, so I gave it up and gave up on the medical community. But now, a week into low-carb, I feel like I could take on the world. So I could personally argue that 25+ years of unnecessary sugar did affect my lifestyle quite negatively, although not enough to be the direct cause of my death – unless falling asleep at the wheel counts.
Julia, last line = priceless (if not for the fact that it’s not entirely a joke). So glad to hear about your success.
Great blog post
The reason the word “toxic” bothers me, as applied to sugar, isn’t a question of accuracy. As you explained, refined sugar may well be toxic in some people at some doses, in the sense that it causes some sort of harm. The thing that bothers me about the word toxic is that it’s an emotionally provocative, sensationalist term in this context. Why not just call sugar “harmful” or “unhealthy”? The word “toxic” evokes hemlock and cyanide in most peoples’ minds, even if technically it can be applied to sugar. It’s a calculated use of language to maximize impact by engaging the emotions.
Fair point. My use of the term was based on the rhetoric that seems to be floating around without a clear explanation of what “toxic” means, which is what bothers me. This is why I like the idea (admittedly a bit simple) of the LD50 and the point about Tylenol. Yes, it can be toxic (acute or chronic), but it’s dependent on dose and host susceptibility. But I agree with the point you’re making. I think most people don’t bring the level of understanding to the problem that you might, though.
> The reason the word “toxic” bothers me, as applied to sugar, isn’t a question of accuracy.
So then, would it be accurate to say that it’s more important to you for terminology to be unmoving than it is for it to be accurate?
> As you explained, refined sugar may well be toxic in some people at some doses, in the sense that it causes some sort of harm.
If it may well be toxic in the sense that it causes some sort of harm, then why would you object to the use of a term that may have a motivational impact?
> The thing that bothers me about the word toxic is that it’s an emotionally provocative, sensationalist term in this context.
Have you considered the possibility that there are individuals that are not at all emotionally impacted by the term?
> Why not just call sugar “harmful” or “unhealthy”?
What would you say to people who consider “harmful” and “unhealthy” wildly outrageous characterizations and who would instead prefer the usage of “nonnutritive?”
> The word “toxic” evokes hemlock and cyanide in most peoples’ minds, even if technically it can be applied to sugar.
How did you make the determination of what the term evokes in other people’s minds?
> It’s a calculated use of language to maximize impact by engaging the emotions.
What if it’s a calculated use of language to *minimize* impact being that the word “lethal” was not chosen?
Unfortunaterly strong language is necessary to get people’s attention. Of course, if you’re Dr. Lustig and dealing with a procession of morbidly obese children; it’s not rhetorical at all.
How do you find time to have ice cream with your kid when you are working 100 hours a week and flying 15000 miles a month?
I make the time. Highest priority.
Peter, I can’t thank you enough – enough and them some to negate the negative feedback you sometimes get – for yet another article that encourages me to think for myself and change the sometimes “black and white” way I feel about nutrition in the United States. It’s a bone of contention between my family and myself which can bring out the shouting and screaming you mentioned above. It’s hard for me not to feel extremist about this stuff. Your post and the comments of others (i.e. the “junkie effect” of ice cream) relativized a lot for me.
In the meantime, I dropped the ball on the cholesterol series because I was just feeling mentally lazy. I will promise myself to get back to that and read it. Thank you.
As for the little powerpoint – Carbohydrate Intolerance – I once sent you for a check, it is getting previewed or downloaded weekly and this makes me feel I’m making a little difference. Thank you again.
Thank you thank you thank you a gazillion bazillion times!
Wonderful to hear, Michele.
Finally! A clear explanation about sugar detailing the science and applying some logic (as well as common sense). I can attest to the acute toxicity of sugar myself. In my 20s my vision suddenly changed. Rather than assume it was genetic ( a few of my siblings wore glasses) it was my ophthalmologist that suggested I quit eating sugar. He was right! My vision returned to normal within a month. Anyone familiar with diabetes would attest to the affect of blood glucose on the lens of the eye.
I’m now interested in both the acute and chronic ‘toxicity’ of sugar in oncology. Should I advise our patients with rapidly growing tumors to abstain from all sugars? Can I advise their families to remove sugar from the diet as a preventative measure?
I would really like to see you address these 2 questions with an equally insightful blog.
By the way – I think (hope) you might have finally taken the sting out of the word ‘toxic!’
Interesting, presumably you were diabetic at the time (even if not formally diagnosed)? This is a not uncommon presentation for people with T2D. Fortunately, it is usually reversible if caught early enough and addressed immediately.