Why Do Sweet Foods Taste Good? The Importance of a Simple Observation

Stuart King writes:

I was very hungry today at dinner and the thought of sweet food wasn’t appealing at all, but after filling up on some rice, chicken and coconut cream curry I immediately had ice cream and chocolate slice [= what Americans call a brownie], which had had no appeal 15 minutes or so before!

An everyday observation that anyone can make. Studies have shown what Stuart noticed: When you are hungry sweet foods are unappealing. This is why dessert is eaten after the rest of the meal.

The main way that psychologists explain an experimental effect — choose between explanations — is by finding out what makes the effect larger or smaller. For example, discovery of what makes learning more or less (what increases or decreases the effect of one learning trial) is the main way psychologists have chosen among different theories of learning. Different theories predict different interactions.

Why do we like sweet foods? The usual answers are that sweet foods are a “good source of energy” and they provide “quick energy”. But these explanations do nothing to explain what Stuart noticed. If sugar is a good (= better than average) source of energy, we should eat it before other foods (average sources of energy) when we are hungry (hunger signals lack of energy). The opposite is true. You may not want to call it a “contradiction” but there is no doubt the conventional view does not explain what Stuart noticed. Of course many nutrition experts, such as Weston Price, are/were entirely sure sugar is unhealthy.

As a tool for choosing among theories, Stuart’s observation is especially good because (a) it is very large (sweets go from unappealing to appealing) and (b) paradoxical (eating calories should make all calorie sources less appealing).

If you have been reading this blog, you know I explain Stuart’s observation by assuming that we need sugar in the evening to sleep well. Sugar (sucrose, fructose, glucose) eaten in the evening increases blood glucose, which increases glycogen. During sleep, glycogen becomes glucose, which the brain needs to work properly. Evolution shaped us to like sweet foods after a meal so that we will eat them closer to when we sleep. (The value of replenishing glycogen close to bedtime also explains why we eat sweet foods after dinner more than after breakfast or lunch.)

I can’t think of another case where what experts say is so out of line with what’s easily observed. For example, I’m sure cholesterol doesn’t cause heart disease, but there is no everyday observation that supports my belief.

I can’t think of another case where what experts say is so out of line with what’s easily observed. For example, I’m sure cholesterol doesn’t cause heart disease, but there is no everyday observation that supports my belief.

If sugar is helpful for sleep, why is it associated with diabetes? My guess is that sugar is almost always consumed in foods that taste exactly the same each time — what in The Shangri-La Diet I called ditto foods. For example, soft drinks. Ditto foods with sugar, because they have a strong precise CS (smell) and a strong fast US (calorie signal), produce an especially strong smell-calorie association. Such an association raises the body fat set point, thus causing obesity. Obesity causes diabetes. It’s also possible that eating sugar during the day — at the wrong time — hurts sleep. Maybe sugar during the day raises insulin and thus reduces the conversion of sugar to glycogen. Less glycogen causes bad sleep, bad sleep causes diabetes. My blood sugar levels clearly improved when I started eating sweets in the evening — opposite to what the sugar-diabetes link would predict.

35 Replies to “Why Do Sweet Foods Taste Good? The Importance of a Simple Observation”

  1. I never ate sweets at night but USED to crave chocolate. Not anymore. I have been liberated from my addiction to chocolate. How? I stopped eating it and 60 days later I lost the craving.

    The good news is that if you cut out all the added fats, sugars, and salt from your food your taste buds WILL adjust. It takes time. 2 or 3 weeks is NOT enough for most folks. You need 30-90 days. After that you’ll actually learn to taste subtle flavors that were overpowered with grease, salt & sugar before. Give yourself some time and your taste buds will adjust. You’ll be amazed at how good “real food” really does taste and how satisfying it can be! I did not believe it would happen for me but it did.

  2. Our taste for very sweet would be based, I assume, on dried fruits and honey. Like most people, I only like something that sweet when I am not hungry. However, I only want fruit (more sweet than starch and less sweet than honey) when I am hot and thirsty.

    I wonder if people who like sweet breakfasts are those who don’t have sweets at night and if they started having dessert after dinner they would stop wanting a sweet breakfast.

  3. This is a good example of using selective evidence. While what he says here is real, there are also lots of times when you crave sweets on an empty stomach. I went through a period where I dieted hard and then binged, and my binges would often start with sweet food and I would often crave sweets. I start the day with a sweet food, and want nothing else in the morning. The ubiquity of sweet food for breakfast as the main meal around the world, which is unheard of for lunch of dinner, also suggests otherwise. (breakfast is not always or necessarily sweet – some people eat bacon and eggs – but its the ONLY meal where it is common for the main meal to be entirely sweet. Waffles with maple syrup is considered normal for breakfast but if that was your lunch or, worse, your dinner, you’d be looked at as very eccentric)

    Basically, what he and you are doing is just fitting the facts to your theory.

  4. Agree with George here. Both as a kid and now I loved sweets for breakfast – waffles, pancakes, oatmeal with honey or maple syrup, and of course, regular cereal (loaded with sugar, check the ingredients list sometime). In direct contradiction to this theory.

    And what do people eat for snacks? When they come home and want something to eat as fast as possible, what do they eat? If it’s not fruit (my own standard after-school snack), it tends to be starchy carbohydrates. (Hold a saltine cracker in your mouth for less than a minute and you can start tasting the sugar as it’s broken down.)

    Plus, does the mechanical theory make any sense here? How does eating an ounce or two of sweets before bedtime make a hill of beans’ difference to metabolism 8 hours later? Is one’s body so dysfunctional it cannot do some basic regulation of blood sugar while asleep? And it’s not like one is exercising vigorously while asleep.

    > I can’t think of another case where what experts say is so out of line with what’s easily observed.

    I look forward to hard data, perhaps even randomized data – I realize it will be a great sacrifice to decide whether to have a late-night snack based on a coin flip and then to analyze the data, but posterity will thank you!

  5. Seth:

    “Obesity causes diabetes.”

    Not in my opinion. In fact you may have it pretty much backwards. There is a correlation, and that’s it. And we know what that means, right? Virtually nothing.

    “Most overweight people never develop type 2 diabetes, and many people with type 2 diabetes are at a normal weight or only moderately overweight.”

    Until you can explain to everyone (and back it up with data to support your claim) why the vast majority of obese people DON’T develop T2D, saying that obesity causes cancer is not too different from saying that cholesterol causes heart disease . We both know how that worked out, right?


    I think Dr. Attia is on to something:

    What if we’re wrong about diabetes?


  6. Here is an article/interview with Lewis Cantley, PhD in biophysical chemistry , with commentary on the actions of glucose/fructose/artificial sweeteners, the action on the body/brain, including a thought about why fructose turns readily to fat (found through Mark’s Daily Apple).


    Seth: Thanks, I saw that. When did biologists become climate scientists? (who feel no need to make correct predictions)

  7. Seth, some interesting thoughts in the last few sentences of the post. Hmm…

    gwern, blood-sugar management seems to be open to great variability between individuals. I’ve noticed a big difference on my sleep patterns based on blood sugar once I started experimenting. I used to find it impossible to stay awake at 5:00 p.m. I assumed it was some wacky circadian glitch. Once I went paleo (low-carb) that stopped. Just stopped. The honey at bedtime trick (and no other sweets/starches during the day) made a huge difference for my kid. But other people have no trouble staying awake at 5:00 p.m. even if they eat a spaghetti sandwich for lunch, and most other kids sleep through the night just fine regardless of what they ate all day.

  8. “sweets” didn’t really exist until recently. there was fruit, when it was in season. very little fruit kept long out of season. apples and pears. honey i’d think was very scarce – when did people learn to keep bees? so i doubt that evolution has much to say about this.

    1. “i doubt that evolution has much to say about this.”

      okay, how do you explain these generalizations: 1. we like the taste of sugar. 2. we don’t like it when we are hungry, only when we are not hungry? 3. (bonus question) we like it more in the evening than earlier in the day?

      I don’t think it’s a stretch to suggest that evolution shaped when we eat fruit.

  9. I have noticed that I really want something sweet after dinner at night, but by the time two hours has past I no longer want it (mostly). Not sure how this would fit your theory.

    And after years of on again, off again, occasions of no sugars and no chocolates I can tell Shelley that one can have totally lost all cravings and then for no apparent reason the cravings return without apparent trigger and without being self indulgent. So don’t be too self righteous about it all. If your cravings stay away then be thankful.

  10. “we don’t like it [sugar] when we are hungry, only when we are not hungry”

    I just don’t understand how anyone can say this. Am I missing something? I LOVE sugar when I am hungry.

    It’s fascinating to me that the field of nutrition seems to produce a long line of highly intelligent people who make bold claims that are contradicted by readily available and extremely obvious information – Ancel Keys, Taubes, Guyenet, Seth, etc. I am convinced the field of nutrition functions as a channel for displaced religious sentiments. Something about nutrition seems to recruit mental habits that usually play a role in religion.

    1. “how can anyone say this?”

      Elizabeth Capaldi said it based on research she did — data she collected. Perhaps you missed the part where I said studies showed this.

  11. Seth, perhaps this is beyond the scope of this post, but do you have any theories about how we sated this hunger for sugars at night in the deep past? I can’t see anyone raiding a beehive and then saving the contents until after dinner. Gathering and saving fruit until after dark, perhaps? Does anyone know if any contemporary hunter gatherer tribes save sweet carbs until near bed time? It would be a nice addition to the theory if we could identify such behaviours. Likewise, not being able to identify such behaviours would be an interesting thing, why would evolution offer such low-hanging fruit and yet we’re unable to find HGs taking advantage of it?

    1. I think people gathered fruit and saved the fruit to eat until late in the evening. They were hungry when they gathered the fruit; it wouldn’t have tasted good at the time.

  12. I try to do a thought experiment when I read something like this. My principle is that whatever is supposed to be healthful should not be contradictory to the environment we had during the palaeolithic times. As we are probably much more adapted to that environment than we are to the modern times. My reasoning for this principle is that our environment has changed too fast for us to be able to adapt to the changes. This does not mean that we have not changed, and there is ample that we have changed, but the changes are not universal, and they do not negate the adaptations to the Palaeolithic times. This in a nutshell is the Palaeolithic Principle.

    Testing Sugar in the night theory with the Palaeolithic Principle, below.

    In the palaeolithic people, would get sugar from fruits or honey. They would gather them and eat them during their cook-out in the evening, or eat them while they were on hunt or gathering. This means that they were eating sugary stuff either before sleeping or during exercises. They might have also been eating them for breakfast as left-overs. The least sugar would be eaten during the breakfast, as it would be a premium item, and unless you got a lot by luck, nothing would be left over.

    I think the logic seems very plausible. And maybe beneficial. A little in the breakfast may not hurt.

    In any case, without some amount of hard work everyday (or at least alternate days), you are not going to be healthy, whatever you do.

    1. Surely hunger was very common long ago. If hunger gets rid of or at least reduces our sweet tooth, that is a stunning fact. Having a sweet tooth makes sense — because sugar supplies energy — so not having one…how strange!

  13. Research has shown that we like sugar ONLY when we are not hungry?

    Wow, I shudder for the state of scientific research. ONLY when we are hungry, not even just it’s more common when we are not hungry?

    Okay, I’m outta here. This place is a joke. Pure nonsense, and an unwillingngess to back off from nonsensical claims.

    Seth: The research showed people liked sweets much less when they were hungry. I was trying to convey it was a big effect. I see I misled you.

  14. Just a personal observation – I will eat sugar when I am hungry but I realize I don’t find it satisfying. If I’m not careful I will look for more sweet foods (the idea of pay attention and don’t eat mindlessly) and end up feeling slightly sick but if I eat something not sweet the hunger is more satisfied. I think that is a lot of the problem with our modern society, we often don’t pay attention to what our bodies are telling us and sweet foods are way too easily available everywhere we turn.

  15. I agree that honey just before bedtime improves my sleep. I’d like to know why.

    But jumping from that personal observation to the theory that people eat sweets after supper so as to improve their sleep ignores other lines of thought.

    1) Many people eat their last meal hours before they go to sleep. The honey effect, for me, depends upon honey ingestion within 1 hour of bedtime, not a dessert immediately after supper.

    2) People eat sugar throughout the day, when they need to stay awake. One line of proof includes inspecting the menus of breakfast chains such as Dennys, IHOP, and Perkins. Another line of proof: stand in a fast food chain during lunch, the kind where people buy a cup and fill their drink at will from the dispenser. What percentage of customers select a sugared soft drink? From my limited personal experience, I remember it as being a large percentage.

    3) Using soft drinks as a measurable product of how Americans consume lots of sugar, using a search engine to look for its consumption before bedtime yields few results. Searches such as: “bedtime coke”, “bedtime cola”, “bedtime soda”.

    4) Then again, Americans do love their ice cream. The search phrase, “ice cream before bedtime” yields a wild mixture of positive and negative search hits. Apparently, if you eat ice cream before bedtime, either you lose weight and sleep better, or you gain weight and have nightmares.

    If I remember correctly, the speculation that desserts improve sleep originated due to a dinner party where dessert was served, thus, it was concluded, it was the dessert that improved sleep. However, there are at least two alternate theories.

    The first, which is the one I prefer, is that a dinner party is a minor stress event which typically ends well. It is stressful because people might not like your jokes and stories, or you may be stuck talking with a bore. Also, you have to be ‘on stage’ for a long period of time. It’s similar to going on a vacation and exploring a bunch of new places.

    The second explanation is that a dinner party simulates paleo life; a day of hunting/gathering, followed by sitting around the fire talking with companions and hearing and telling stories. What is similar in modern life? Many people, after a day of work, watch TV, where they see friendly faces and hear stories. A dinner party is a more intense expression because one not only consumes but tells stories.

    1. “I agree that honey just before bedtime improves my sleep. I’d like to know why.”

      I think the idea that “honey before bedtime improves sleep” is far from proven (although the idea that “sweets in the evening improve my own sleep” is proven beyond doubt). It is also unclear to me how much weight (a lot? none? moderate amount?) to put on what people tell me about their sleep. For example, you tell me that honey before bedtime has improved your sleep. How much weight should I put on that?

      By looking into associated phenomena, such as the notion that when we are hungry, we don’t want sweets, I learn more about both issues (1. is it generally true that bedtime honey improves sleep? 2. Should I trust what people tell me about their sleep?) If the answer to both #1 and #2 is yes, then the idea that we don’t like sweets when we’re hungry makes sense. If the answer to both #1 and #2 is no, the idea that we don’t like sweets when we’re hungry is not explained.

      I agree with you, this post is no help in understanding why bedtime honey improves sleep.

  16. Why do people fuss so much about paleo life? The population has grown so much since that it’s easy to believe that we’ve evolved a long way from then.

  17. dearieme:

    “Why do people fuss so much about paleo life? The population has grown so much since that it’s easy to believe that we’ve evolved a long way from then.”

    I’m not sure that I understand your question. But the people who “fuss so much” about paleo are usually the people who think eating a 100% plant-based diet is the ONLY way to eat. Which is preposterous, of course. Man evolved while eating animals, animal fat, fish, insects, snakes, etc., and certainly veggies and fruits (when in season) for millions of years. That’s why we are omnivores, not carnivores or herbivores. And we’ve thrived.

    On the other hand, it’s only been about 10,000 years (since the beginning of agriculture) that we’ve eaten things like grains, especially wheat, year-round fruit, etc. Which is basically a blink of the eye, in terms of evolution. Many of us think our health has suffered for it. Others don’t agree. Thus the “fuss.”

    Also, it seems to be a sort of religious experience for one side, saying and implying that those of us who choose to eat animals and animal fat, like we have for millions of years, are now “unethical.” You can probably imagine how some us would respond to that. Hint: Not very well.

  18. I think an interesting question to ask would be: Do people who eat deserts after dinner eat big breakfasts, smaller breakfasts, or no breakfast? And do people who eat big breakfasts tend to eat desert after dinner?

    When I am hungry but not stressed I crave meat and/or fat. When I’m hungry and stressed I tend to still crave meat and/or fat but less of it and only two things as “desert” satisfy my “sweet” craving: yogurt or dairy (e.g. milk) or something sweet.

    Another question I’d be interested in knowing the answer to is do people who eat more full fat yogurt crave less sugar?

  19. Just a few thoughts:
    1. i think we crave sweets after meals because we didn’t get sugar during the meal, not because we naturally prefer it after a meal. In other words, if you have steak and potatoes, you may crave dessert. If you have a Morrocan meal with meat, rice, figs and honey, you may not crave sweets after a meal. (This would be easy to test).
    2. No one craves just sugar. They want dessert, which is fat, starch and sugar together, so i see sugar as part of a balanced diet.
    3. Almost all starches and meats are more palatable when eaten with sugar. Breakfast cereals are loaded with added suga;, meats like barbaque are covered in sugary sauces. I mean is there anyone when hungry who likes plain oatmeal??? No, we love it with brown sugar and fruit or honey.
    4. There is a lot of discussions of what Paleo ancestors ate. First, I’ve been following the Paleo diet for over 10 yrs and it changes every few years. It was high protein and veggies, then they added more saturated fats, then they added more tubers like sweet potatoes. For a diet that hasn’t changed in a million years, it changes about every 2 yrs with a new recommendation of how we misunderstood something.
    5. Which brings me to #5, Paleo people say we would not have had access to sugar but traditional peoples have collected honey throughout history. Honey can be stored indefinitely, so i kind of roll my eyes when a Paleo advocate says that we should only eat honey if you’ve gone through the trouble to climb a rope ladder, beat off beats, recover the honey and sprint away from the bees. Those same people never say you can only eat meat if you have run for miles, tracked an animal, killed it, skinned it, and carried it back to your tribe. My point is just that i think we have to be careful creating narratives for why we should or should not do something.
    6. Finally, we don’t live a Paleo life. We struggle under a 10 hour cognitively demanding job with more chronic stress which has been shown to damage the brain, given sugars role in providing energy for the brain (mostly through the liver), its makes sense that we might crave (and therefore need) more sugar than we had during Paleo times.

  20. Seth Roberts, I believe your theory how about sugar improves is correct, but I do not quite agree with the explanation for diabetes. To me, a simpler explanation would be that our livers need a certain amount of micronutrients to digest properly high amounts of sugar, including magnesium, complex B, conpenzyme q10 and choline (choline is needed in liver to properly export triglyclerydes produced by excessive ingestion of alcohol and sugar, see this post made by Chris Masterjohnson: http://blog.cholesterol-and-health.com/2010/12/meeting-choline-requirement-eggs-organs.html).

    So the brain prefers to ingest sugar after a meal because half of it is glucose, who can provide immediate energy for brain, and the another half, fructose, is metabolyzed exclusively in liver, providing a stable source of glucose for a brain whose body cannot store high amounts of glucose. My proposal to fix your theory is that sugar (and honey) helps improve not only sleep, but our health and brain function as well. But if we ingest insufficient ammount of those mentioned micronutrients and high amounts of sugar, we’ll develop metabolyc syndrome and inexorably DM II. Of course, like traveling to Rome, there are many ways to develop DM II, not just this one.

  21. Strauss, i agree with your points, and its a reason why i think there is confusion on the issue. Fructose is incorrectly seen as bad, but its really the refined sugars and their lack of nutrients that cause problems. Honey and fruit have a ton of nutrients that accompany the sugars and help them metabolize. Second, the glucose and fructose are free and not bound together like in sucrose.
    Third, i agree that the primariy benefit of sugar is supporting the brain. With only ~2% of body mass and yet consuming 20% of the calories, keeping the brain healthy with a steady supply of fuel seems paramount.

  22. “Seth: The research showed people liked sweets much less when they were hungry. I was trying to convey it was a big effect. I see I misled you.”

    Fair enough, Seth. I still think you’re ignoring lots of readily available evidence. You’ve ignored the point about breakfast, which as the name implies is breaking a long fast, being the most commonly eaten sweet meal. That alone poses a serious challenge to your theory. That the ONE meal eaten after a long fast every day in all cultures around the world tends to be sweet, or is often sweet, and that this isn’t the case with lunch or dinner, massively does not support your theory.

    There there is the well known fact that heavy dieters almost always binge on sweet foods – they don’t binge on steak and bacon and eggs, but donuts and ice cream and chocolate. Again, massive un-evidence.

    Sorry, it’s a nice theory but it just doesn’t add up. Do the right thing and move on. Truth is more important than our pet theories.

    Seth: You might want to read the research I mentioned, by Elizabeth Capaldi.

  23. David, I’m quoting this because I like it:

    “4. There is a lot of discussions of what Paleo ancestors ate. First, I’ve been following the Paleo diet for over 10 yrs and it changes every few years. It was high protein and veggies, then they added more saturated fats, then they added more tubers like sweet potatoes. For a diet that hasn’t changed in a million years, it changes about every 2 yrs with a new recommendation of how we misunderstood something.”

    As for the rest, honey is easy to store, but I don’t have a feeling for when people invented a way (ceramics?) to store it. Could they have used animal bladders? If so, I don’t think there’d be a fossil record.

    It’s true that people don’t want straight sugar for dessert. Hard candy is for snacks, not dessert. Still, dessert varies a lot in it’ fat/starch/protein/sugar ratio. Your basic cake is sugar/refined flour/a little egg and dairy. Cheese cake is mostly cheese and sugar. Let to themselves (no theory about nutrition), people don’t use whole grains in dessert so far as I know.

    I believe we need more investigation of dessert.

  24. I have been experiementing with a ketogenic diet. Recently my blood sugars upon rising have increased to well over 100. Within a couple of hours upon rising my blood sugars drop to 90. That was not the case before I started the diet,

  25. continuing…

    as my blood sugars were normally around 90 upon rising. I have been reading about the Somogyi Effect and Dawn Phenomenon for diabetics as a possible clue to why my bs have shifted. I am beginning to think the SE and DE are not mutually exculusive to active diabetics. From this post, and the one about honey before bed, I am wondering if being in a state of nutritional ketosis might just only be beneficial during the day. This thought comes about because last night I just went crazy with eating sugar before bed, and I awoke in very good spirits after having a wonderful dream. For the last several months this has not been the case. I usually had very uncomfortable dreams. I think for me my hours of rest have increased to 8 hours due to the time change. During the winter I usually will awake at 3am and stay awake for a couple of hours.

    Today I am going to experiment with eating a ketosis diet through out the day and tonight I am going to have honey.

    1. “I am beginning to think the SE and DE are not mutually exclusive to active diabetics.”

      I agree. I found this in Wikipedia:

      “Clinical studies indicate that a high fasting glucose in the morning is more likely because the insulin given on the previous evening fails to last long enough.”

      I had/have a high fasting glucose in the morning but don’t take insulin (and am not diabetic).
      So the explanation of the high glucose in the morning, at least in my case, cannot be my insulin didn’t last long enough.

  26. I tried the honey last night, and it did not work. I actually awoke at 12am in a panic. This also happens when I have fruit. I am thinking the fructose in both is the culprit.

  27. Humans are born loving sweet, fatty and salty foods. Scientists think it is evolutionary because we didn’t get much of this stuff when we first evolved so people who ate this whenever they found it lived longer and passed on liking this in their genes to their offspring. The human brain takes a lot of energy to keep it going, as well as our body, and snacking on a few plants just won’t do. Humans who ate sweet things (fruits, etc.) and fats (meats, nuts) got more nutrition and lived longer.

Comments are closed.