Nighttime sugar cravings: what causes them and how to calm them

Antojos de azúcar por la noche: qué los causa y cómo calmarlos

Illustrative image. Does not represent guaranteed results.

It's 9:20 at night. The kitchen is clean, the dishes are put away, dinner was two hours ago, and you're not hungry — not real hunger, not stomach hunger. And yet here you are, standing in front of the pantry with the door open, looking for something you can't even name. Something sweet.

That scene repeats in millions of kitchens every night, and it almost always gets read the same wrong way: as a discipline failure. As if the problem were that at nine at night you become a weaker person than the one you were at ten in the morning.

The evidence says otherwise. It says the craving has a schedule because your appetite has a schedule, that it has a direction because sugar activates different circuits than other foods do, and that most of the fight was decided long before you reached the pantry: in the breakfast you ate, in the snack you didn't eat, and in how much liquid sugar moved through your day without being counted.

This article is about that mechanism. With real numbers, sources at the end, and the honest part about what the science still can't promise you.

The nine o'clock craving is not a character flaw

Let's start by separating two things that get mixed together constantly in everyday conversation: hunger and craving.

Hunger is an energy signal. It rises slowly, it doesn't discriminate much between foods, and it goes away when you eat almost anything. If you're truly hungry, a plate of beans solves it.

A craving is something else. It's specific — you don't want food, you want that flavor — it appears fast, and a plate of beans doesn't touch it. It can show up on a full stomach. In fact, the nine o'clock scene usually happens precisely on a full stomach, and that's what makes it so bewildering to the person living it.

When someone tells you "just don't eat sweets," they're giving you advice designed for hunger and applying it to a craving. That's why it doesn't work. Not because you lack discipline, but because the advice is aimed at the wrong mechanism.

The eating-behavior literature treats cravings as their own phenomenon, with their own triggers: time of day, mood, prior deprivation, cue exposure — seeing the package, smelling the bakery — and the quality of last night's sleep. None of those five triggers is "character."

This matters because it changes the strategy. If the problem were willpower, the solution would be to squeeze harder. If the problem is triggers, the solution is to disarm the triggers, which is a far more concrete and far less exhausting task.

How much sugar is "too much," according to the largest review that exists

In 2023, BMJ published an umbrella review — a review of reviews — on dietary sugar and health. The team examined 73 meta-analyses, drawn from 8,601 unique articles, covering 83 different health outcomes.

It is, as far as we know, the broadest attempt to put all the evidence on sugar on the same table. And it arrived at a numeric recommendation worth keeping handy:

It's worth reading carefully what "free or added sugars" means. It isn't whole fruit. It isn't the lactose in milk. It's the sugars added to foods and drinks during processing or at the table, plus those naturally present in juices, syrups and honeys.

Twenty-five grams is less than most people picture. A 355 mL can of soda contains about 39 grams on its own. In other words: a single can clears the ceiling for the entire day, and you haven't had breakfast yet.

The review also documented adverse associations between sugar consumption and 18 endocrine/metabolic outcomes, 10 cardiovascular, 7 cancer and 10 other outcomes. But we have to be precise here, because the authors themselves are: the quality of that evidence was mostly rated moderate or low. The association with body weight — via sugar-sweetened beverages — and with ectopic fat accumulation — via added sugars — was the one that reached moderate-quality evidence.

Translated: it isn't that "sugar causes everything bad." It's that the most consistent and best-supported signal points to body weight and to fat deposited where it shouldn't be. Which is exactly what we're talking about.

Why the craving shows up at night and not at eleven in the morning

Here a concept enters that almost nobody uses in everyday conversation but that explains a lot: chronotype.

Your chronotype is your biological timing preference — when your body prefers to sleep, wake, move and eat. It isn't a decision. There are early chronotypes ("larks") and late chronotypes ("owls"), with a whole spectrum in between.

A review published in Proceedings of the Nutrition Society gathered the evidence on how chronotype influences eating behavior and appetite control. The findings that matter most here:

• A late chronotype is associated with greater evening energy intake, and that combination may amplify overweight risk and curtail weight-management efforts.
• Late chronotypes tend to have worse diet quality: greater intake of fast food, caffeine and alcohol, and lower intake of fruits and vegetables.
• A late chronotype is also associated with eating-behavior traits that increase susceptibility to overconsumption: disinhibition, food cravings and binge eating.

If you recognize yourself as an "owl" and you've spent years fighting the nine-o'clock window, this isn't a sentence. It's an explanation. And it explains something that tends to generate unfair guilt: the person next to you, with an early chronotype, may be fighting an objectively easier battle at the same clock time.

The honest part about this point

The authors of that review note that whether chronotype influences appetite in response to food intake and exercise is an area of recent interest that has largely been overlooked. Preliminary evidence suggests additive rather than interactive effects between chronotype and meal timing. In plain terms: we know both matter, but we don't yet know well how they multiply each other.

And more studies are needed on the interplay between chronotype, food and exercise timing, and sleep. This is an open field, not a closed manual.

Liquid sugar is its own case

If you're going to change one single thing after reading this, make it this one.

The reference meta-analysis on sugary drinks and weight was published in The American Journal of Clinical Nutrition in 2013. It included 32 original articles: 20 in children (15 cohort studies with 25,745 participants and 5 trials with 2,772) and 12 in adults (7 cohorts with 174,252 participants and 5 trials).

Half a pound a year sounds like nothing. That's exactly the problem: it's so small you don't see it on one week's scale, and you do see it on the scale five years out.

Why does liquid sugar behave differently? The most accepted hypothesis is that liquid calories aren't compensated for well. When you eat 150 calories of solid food, your body tends to adjust — you eat a bit less afterward. When you drink 150 calories, that adjustment is far weaker. The calories come in, but the satiety signal barely notices.

It's worth including the critical voice here, because it exists and it's serious. A review published in Advances in Nutrition examined the same body of evidence and concluded that, although the pattern across observational and experimental studies points to increased risk of weight gain with higher sugary-drink intake, it remains difficult to establish the strength of that association and its independence from other potentially confounding factors. The main reason, according to that author, is the heterogeneity and methodological limitations of the available studies, both observational and experimental.

That same review makes a useful point about artificially sweetened beverages: although some observational studies have suggested they might increase the risk of obesity and cardiometabolic disease, there is no clear mechanism for that pathway and the epidemiological studies are highly inconsistent. An important problem in those studies is reverse causality bias: people who already carry excess weight are more likely to choose diet drinks, which makes the drink look like the cause when it may be the consequence.

In other words: the evidence that sugary drinks push weight upward is solid in its direction though debatable in its magnitude. The evidence that the artificially sweetened substitute is harmful is much weaker than it's often repeated to be. If swapping soda for its no-sugar version is the step you can take today, the evidence gives no strong reason to discourage it.

The snack experiment: how 20% of the dessert switched off

This is probably the most practical study in the whole article, because it tests an intervention you could do tomorrow without buying anything unusual.

In 2020, a team published in The Journal of Nutrition a randomized crossover trial with 39 healthy adults (mean age 26, mean BMI 24.4). Each participant went through three afternoon snack patterns, six days each:

Notice the design: both snacks had the same calories. 240 versus 240. What changed was the composition — one low in sugar and with some protein, the other with 16 grams of sugar.

On day 7 of each pattern, the team standardized breakfast and lunch, gave the snack three hours after lunch, measured appetite and satiety through the afternoon, provided a standardized dinner, and then handed each participant a snack cooler to take home and eat from freely through the evening. They also ran 24-hour continuous glucose monitoring.

The results:

• The low-sugar snack reduced subsequent snacking on desserts by about 20% compared with no snack and compared with the granola bar.
• It reduced appetite measures — hunger, desire to eat, and prospective food consumption — by roughly 70% versus no snack. The granola bar did not.
• Satiety was about 30% greater with both snacks versus no snack, with no difference between the two snacks.
• The low-sugar snack reduced afternoon blood glucose by about 5% compared with the bar.

What matters about this experiment is what it separates: satiety — the feeling of fullness — was the same with both snacks. But appetite and evening dessert snacking were not. Eating something mid-afternoon helps; what you eat mid-afternoon decides whether the effect reaches the evening.

And the limitation, which the authors state: it was an acute study, in young healthy adults, with six days per pattern. They write themselves that long-term trials are warranted. This is not proof that eating hummus will make you lose weight. It's a clean demonstration of a mechanism: the right snack disarms a nighttime trigger.

Today's breakfast decides tonight's craving

The second lever sits fourteen hours before the craving.

A randomized crossover study published in Current Developments in Nutrition compared two patterns for 7 days each in 13 young adults: eating a high-protein breakfast (340 kcal, 30 g protein) versus skipping breakfast.

On day 7 they ran a full 8-hour clinical testing day, measuring hunger, fullness, desire to eat, prospective consumption, related hormones, functional MRI brain scans with food-cue images, and ad libitum evening food intake.

What they found with breakfast versus skipping it:

• Hunger, desire to eat, prospective food consumption and ghrelin — the hormone that drives appetite — all decreased, both in the morning and across the day.
• Fullness increased.
• Activation of the hippocampus, parahippocampus and middle frontal gyrus in response to food images was reduced. Those are regions tied to food-reward memory and executive control; less activation means the picture of the cake "pulled" you less.
• And the datum that concerns us: evening intake of high-carbohydrate and high-fat foods was reduced. Evening sugar intake specifically tended to be reduced, though without reaching statistical significance (p = 0.085).

That last parenthesis is important and we're not going to hide it: the effect on evening sugar was a trend, not a confirmed result. With 13 participants, a study doesn't have the statistical power to confirm much. What was confirmed was the effect on evening carbohydrates and fats, and on ghrelin across the whole day.

The practical reading is the same in both studies, the snack one and the breakfast one: the nine o'clock craving is built during the day. It's built with hours of absent protein, with glucose spikes and drops, and with a ghrelin that never finished coming down. When you arrive at the pantry at nine, you're arriving at the end of a chain that started at seven in the morning.

Seven things you can do this week

Nothing on this list is a treatment. These are trigger changes, ordered by effort-to-benefit ratio.

1. Take out the liquid sugar first

Soda, juice, syrup coffee, sweet tea, store smoothies. It's the intervention with the most direct evidence on weight and the one that involves the least "food sacrifice." Water, sparkling water, unsweetened coffee, unsweetened iced tea. If you need a bridge, the no-sugar version of your drink is a reasonable bridge given the available evidence.

2. Put protein in breakfast

The study used 30 grams in 340 calories. That's three eggs with beans, or Greek yogurt with nuts, or yesterday's chicken. It doesn't have to be a protein shake or anything purchased.

3. Snack mid-afternoon, but low in sugar

The whole point of the experiment was that the high-sugar snack with the same calories did not reduce the evening dessert. Hummus with vegetables, a handful of nuts, cheese, a hard-boiled egg. Three hours after lunch is the window that was studied.

4. Close the kitchen at a fixed time

Not as a moral prohibition, but as a decision made once instead of twenty times a night. The problem window is identified; putting a closing time on it attacks that window directly.

5. Count the hidden sugar for one week, just one

Not forever. One week of reading labels teaches you where your 25 grams live, and that information doesn't leave you. Tomato sauce, dressing, sandwich bread, flavored yogurt, cereal, granola: that's usually where the grams nobody was counting are hiding.

6. Change the environmental trigger

Cue exposure is one of the best-documented craving triggers. If the sweet isn't in the house, the nine o'clock craving runs into a pantry with no answer, and in most cases it burns itself out in a few minutes. That isn't willpower; it's logistics.

7. Protect your sleep and respect your chronotype

If you're a late chronotype, the evening hours work against you biologically. Anchoring your wake time, getting natural light early, and not fighting your internal clock with guilt are more realistic strategies than trying to become someone who falls asleep at nine.

A note of reality: if you work rotating shifts, hold two jobs, or your only quiet hour of the day is after ten at night, several of these points collide with your actual life and that isn't your fault. In that case, start with point 1 and point 6, the two least dependent on schedule.

What changes if you're on a GLP-1 program

Many people arrive at this topic already inside a supervised weight-management program, and the question is fair: does this still matter?

Yes, and for a design reason. The large trials of weight-management medications studied the drug as an adjunct to a lifestyle intervention, not in place of one. Participants in those studies received counseling on diet and physical activity. When you read a result from those trials, you're reading the result of the package.

There's also a practical reason. Many people in treatment describe that physical hunger drops noticeably but the nine o'clock craving — the one that wasn't coming from hunger in the first place — doesn't fully disappear. That makes sense: they're different mechanisms. One responds to satiety signals and the other responds to timing, reward and habit.

So the conversation about triggers doesn't compete with a clinical plan. It accompanies it. And that's why questions about meal timing, protein and sugary drinks are among the most useful things you can bring to your provider at a follow-up visit.

Eligibility is determined by medical evaluation, and who is a candidate for what is a clinical decision that isn't made by reading an article.

What this evidence does NOT say

Let's be explicit, because this topic attracts exaggeration in both directions.

It does not say sugar is addictive like a drug. That debate exists in the literature, it's far from settled, and none of the studies we cite here address it. What is documented is that cravings have identifiable, modifiable triggers.

It does not say that removing sugar will make you lose weight. The sugary-drink meta-analysis measured effects of fractions of a kilogram per year. That's a real, cumulative effect — not a transformation.

It does not say fruit is the problem. The 25-gram recommendation is about free and added sugars. Whole fruit, with its fiber and its matrix, isn't in that category.

It does not say these studies are large. The snack study had 39 participants. The breakfast study, 13. Both were short. They show mechanisms with good experimental design, not population rules.

And it does not say how much you will lose. Individual results vary and are not guaranteed.

Who is this NOT for?

This article isn't for you if your relationship with food has crossed into territory that needs professional help. If you have episodes of eating large amounts with a sense of loss of control, if you eat in secret, if there's intense guilt or compensatory behavior afterward, that isn't solved by a list of seven habits and it deserves evaluation by a mental health professional or an eating-disorder specialist. Asking for that help isn't a failure; it's the correct step.

It also isn't for you if you have type 1 or type 2 diabetes under treatment and you're thinking about large carbohydrate changes on your own. There, adjustments have to be coordinated with whoever manages your medication, because the risk of hypoglycemia is real.

It isn't for you if you're looking for permission to eliminate entire food groups. The evidence in this article points to free sugars and sugary drinks, not to a war on carbohydrates.

And it isn't for you if you already take in under 25 grams of free sugars a day, eat protein at breakfast, and don't drink your calories. If that's your situation and the weight isn't moving, this isn't your variable, and pushing harder here wastes time you could spend evaluating sleep, hormones or labs.

It is for you if you recognized yourself in the nine o'clock scene, have been treating it as a character flaw, and nobody had explained that it has mechanics — and that the mechanics can be disarmed.

See the programs

GLP-1 programs from $88.99 per month. Care available in English and Spanish, 100% telemedicine.

Frequently asked questions

How much sugar can I eat per day?

The umbrella review published in BMJ in 2023, which pooled 73 meta-analyses, recommends reducing free or added sugar consumption to below 25 g per day — roughly 6 teaspoons — and limiting sugar-sweetened beverages to less than one serving per week. That refers to added and free sugars, not whole fruit.

Why do I crave sweets only at night?

Timing matters biologically, not just as habit. A review in Proceedings of the Nutrition Society documented that a late chronotype is associated with greater evening energy intake, worse diet quality, and more cravings, disinhibition and binge eating. On top of that sit the day's accumulated triggers: low protein, glucose swings and cue exposure.

Does an afternoon snack help or make things worse?

It depends on the snack. In a crossover trial with 39 adults, a low-sugar afternoon snack reduced evening dessert snacking by about 20% compared with no snack and compared with a granola bar of the same calories but 16 g of sugar. Satiety rose with both snacks; the effect on evening dessert came only from the low-sugar one.

Does skipping breakfast help me eat less?

In a 7-day-per-pattern crossover study with 13 adults, eating a high-protein breakfast lowered hunger, desire to eat and ghrelin, and reduced evening intake of high-carbohydrate and high-fat foods compared with skipping breakfast. Evening sugar intake tended to fall without reaching significance. It's a small study, but it points the opposite way from intuition.

Are diet drinks worse than regular soda?

The available evidence doesn't support that claim. A review in Advances in Nutrition notes there is no clear mechanism by which artificially sweetened beverages would cause obesity, that the epidemiological studies are highly inconsistent, and that reverse causality bias is an important problem. By contrast, the evidence on sugary drinks and weight gain is more consistent.

How much weight can I lose if I quit soda?

We can't tell you that. What the 2013 meta-analysis measured is that each additional daily serving of a sugary drink was associated with 0.22 kg of weight gain per year in adults, and that adding them in trials produced 0.85 kg of gain. Those are cumulative effects measured as group averages. Individual results vary and are not guaranteed.

Does this apply if I'm already in a weight-management treatment?

The large trials of weight-management medications studied the drug as an adjunct to a lifestyle intervention, not in place of one. Also, physical hunger and craving are different mechanisms, and many people notice the second one persists. Talk to your provider about your case; eligibility is determined by medical evaluation.

Do you offer care in Spanish?

Yes. All of our content and our care are available in English and Spanish.

Sources

• Huang Y, Chen Z, Chen B, et al. Dietary sugar consumption and health: umbrella review. BMJ. 2023;381:e071609. — DOI · via PubMed
• Malik VS, Pan A, Willett WC, Hu FB. Sugar-sweetened beverages and weight gain in children and adults: a systematic review and meta-analysis. Am J Clin Nutr. 2013;98(4):1084-1102. — DOI · via PubMed
• Beaulieu K, Finlayson G, Quist JS. Chronotypical influence on eating behaviour and appetite control. Proc Nutr Soc. 2026;85(1):6-12. — DOI · via PubMed
• Reister EJ, Leidy HJ. An Afternoon Hummus Snack Affects Diet Quality, Appetite, and Glycemic Control in Healthy Adults. J Nutr. 2020;150(8):2214-2222. — DOI · via PubMed
• Gwin JA, Leidy HJ. Breakfast Consumption Augments Appetite, Eating Behavior, and Exploratory Markers of Sleep Quality Compared with Skipping Breakfast in Healthy Young Adults. Curr Dev Nutr. 2018;2(11):nzy074. — DOI · via PubMed
• Pereira MA. Sugar-sweetened and artificially-sweetened beverages in relation to obesity risk. Adv Nutr. 2014;5(6):797-808. — DOI · via PubMed

This article is informational and is not medical advice. SCTS1 does not prescribe medications. All medical care, prescriptions and treatment plans are provided by licensed physicians through our partner platform. Eligibility is determined by medical evaluation. Compounded medications contain the same class of active ingredient but are not FDA-approved finished pharmaceutical products. Individual results vary and are not guaranteed. If you have episodes of loss of control around eating, or if you live with diabetes, consult a health professional before making major dietary changes.