Creative Touch Permanent Makeup and Aesthetics

The timing isn’t random.

Human appetite follows a circadian rhythm, with its lowest point in the morning and a progressive increase throughout the day. ^[1] This pattern is governed by the body’s master clock, a cluster of neurons in the hypothalamus (an area at the base of the brain) called the suprachiasmatic nucleus (SCN). The SCN coordinates biological timing across virtually every organ and tissue in the body. Approximately 15% of all human genes are expressed rhythmically across the day, ^[1] and hunger is among the most strongly patterned of all biological drives.

Cortisol, the body’s primary alertness and glucose-mobilising hormone, follows a well-characterised daily arc, established across decades of chronobiological research. It peaks in the early morning, providing the energy and focus needed to start the day, then gradually declines. By early to mid-afternoon, it hits a natural trough. This is physiologically normal, not a malfunction. Many cultures have historically built afternoon rest periods around exactly this biological reality.

In a metabolically healthy individual eating whole, unprocessed foods, this cortisol dip is barely noticeable. Fat oxidation maintains relatively stable energy, the fuel-switching system operates cleanly, and the dip passes without drama.

But when a high-glycaemic lunch is added to this circadian pattern, something different happens. The circadian data from McHill & Butler (2024) place the peak of appetite at approximately 20:00, with a trough around early morning. Through the afternoon, appetite is on an ascending slope, not yet at its peak. That ascending circadian drive creates the backdrop. What turns it into a crisis is the postprandial glucose crash from lunch, typically arriving three to five hours after eating, landing on exactly that rising afternoon appetite curve. Both factors converge: cortisol is at its daily trough, appetite is climbing, and blood glucose is crashing. The adrenal response kicks in, ghrelin rises, and the brain signals something that feels a lot like emergency: find glucose, now.

That’s why it’s 4pm. Not the circadian peak, but the worst possible moment for the metabolic crash to arrive.

This is where the science becomes genuinely striking, because what happens in the brain four hours after a high-glycaemic meal isn’t just hunger. It’s measurably different from normal appetite.

In a randomised, blinded, crossover trial, Lennerz et al. (2013) had 12 overweight and obese men consume two test meals on separate occasions. ^[2] Both meals were identical in calories, macronutrient composition, ingredient sources, and, critically, palatability, confirmed by taste ratings that showed no significant difference between the two (5.5 vs 5.3 cm on a visual analogue scale; p = 0.7). The only difference was the glycaemic index. The high-GI meal had a calculated GI of 84; the low-GI meal, 37. (Neuroimaging data were available from 11 of the 12 participants; one participant’s fMRI data were lost due to a data-storage error. Blood glucose, insulin, and hunger data were analysed from all 12.)

At four hours, the researchers measured two things: blood glucose levels and brain activity via functional MRI.

In the 11 participants with usable neuroimaging data, the results were precise. Blood glucose was significantly lower after the high-GI meal (4.7 ± 0.14 vs 5.3 ± 0.16 mmol/L; p = 0.005). Hunger ratings were significantly higher (p = 0.04). And brain activity was significantly greater in a specific, prespecified region: the right nucleus accumbens, with an 8.2% relative difference in cerebral blood flow (p = 0.0006, Bonferroni corrected), spreading to the broader striatum and olfactory areas.

The nucleus accumbens is the brain’s primary craving and reward centre. It’s the same region that becomes activated in substance use disorders. And critically, its activation in this study was related to glucose and insulin levels, not to how the food tasted. Palatability was not significantly associated with nucleus accumbens activity (p = 0.56). The food was the same. The brain response was not.

What this means in practice is important. The craving you feel at 4pm is not about the biscuits in the kitchen being irresistible. It may be about what happened in your bloodstream three hours ago, as your lunch was digested. The drive is metabolic, not sensory.

This is the argument that Professor David Ludwig of Harvard Medical School makes explicitly in a 2026 perspective paper in PLOS Medicine: that the concept of “hyperpalatability” as the mechanism linking ultra-processed foods to addictive eating is circular reasoning. Hyper-palatable foods are defined as foods that drive overconsumption, but saying they drive overconsumption because they’re hyperpalatable explains nothing. The more precise explanation, grounded in the Lennerz data and others, is post-ingestive and metabolic: processed carbohydrates disrupt the food matrix that would normally slow digestion, accelerating glucose absorption, producing a sharper insulin spike, and setting up the four-hour crash that activates reward circuitry. ^[3]

Many of the foods most commonly associated with binge eating, such as biscuits, crackers, crisps, and white bread, are relatively bland. The drive to eat them compulsively is not about their taste. It’s about their glycaemic fingerprint hours earlier.

Here’s the question the earlier section raises: if blood glucose is low four hours after lunch, why doesn’t the body simply switch to burning its vast stores of fat? Why the sudden emergency signal?

The answer lies in an elegant cellular “fuel-switching” system discovered in 1963 by Sir Philip Randle and now known as the Randle Cycle. In a metabolically flexible person, the body smoothly shifts from glucose to fat burning between meals. But when a high-GI lunch triggers a sharp insulin spike, the switch gets temporarily jammed.

Insulin tells the cell to make a molecule called malonyl-CoA, which locks the gate that lets fatty acids enter the mitochondria for burning (the gate is called CPT-1). While that gate stays closed, the cell is caught between fuels: glucose is low, yet fat can’t be accessed. The result is the exact 4 pm crisis you feel. For the full cellular details and how this pattern affects skin ageing through chronic insulin signalling, see our Free Fatty Acids and Lipid Synthesis knowledgebase pages.

This jammed switch is especially common with the moderate-fat + high-refined-carbohydrate combination found in ultra-processed foods; the very profile that creates the metabolic impasse Hue & Taegtmeyer described in their 2009 review. ^[4]

A bag of crisps is the perfect example. Fat and processed carbohydrate, together, are precisely engineered into the profile that blocks fat oxidation while delivering the glucose spike that sets up the four-hour crash.

The cellular energy crisis extends to the brain’s own hunger-control centres, and this is where the picture becomes particularly important for understanding why the 4pm craving can feel genuinely unmanageable rather than merely uncomfortable.

The hypothalamus contains two primary populations of glucose-sensing neurons. POMC neurons are satiety neurons: when glucose levels are adequate, they activate and suppress appetite. NPY/ AgRP neurons are the counterpart: when glucose falls, they activate and drive food seeking, raising hunger to restore blood glucose levels. ^[5]

Under normal circumstances, this system operates precisely. Blood glucose falls, hunger rises, the individual eats, glucose is restored, satiety is signalled. Clean, efficient, appropriately calibrated.

But in high-fat, high-sugar dietary conditions, hypothalamic glucose sensing becomes impaired. Diet-induced obesity has been shown to reduce GLUT2 expression in the hypothalamus and cause inappropriate AMPK activation, disrupting the sensitivity of these glucose-sensing neurons. ^[5] POMC neuron glucose responsiveness is impaired in rodent models of high-fat-diet-induced obesity, where mitochondrial dysfunction (particularly in surrounding microglia) disrupts normal glucose sensing. ^[5] The precision of the system degrades.

Dr Ben Bikman, a professor of cell biology and physiology, describes this state using a striking image from Coleridge’s Rime of the Ancient Mariner: “Water, water, everywhere, nor any drop to drink.” The insulin-resistant brain is like the Ancient Mariner surrounded by undrinkable salt water. ^[7] Blood glucose may be present in the circulation. Fat stores may contain months of energy. And yet the brain’s hunger centres are signalling deprivation, because they cannot access either fuel effectively. The glucose-sensing neurons are impaired, the fat oxidation pathway is blocked at CPT-1, and ketone production is suppressed by chronically elevated insulin, preventing hepatic fat oxidation.

The sailor has no fresh water, and the ship’s water ration is locked away.

This is the precise cellular scenario that Ludwig (2026) describes as the “central conundrum” of obesity: why people on calorie-restricted diets experience what physiologically resembles starvation, despite carrying months of stored energy. ^[3] The problem is not the energy. It is the access.

The difference between a lunch that sets up the 4pm crisis and one that doesn’t isn’t always dramatic, and it’s not about eating less. Based on the Lennerz et al. (2013) study design, two meals of identical calories, identical macronutrient ratios, and indistinguishable taste can produce fundamentally different metabolic and neurological consequences four hours later.

The test meals in the Lennerz study used liquid form and focused on overweight and obese men aged 18–35, which limits direct generalisation to all populations and to solid food (Lennerz et al., 2013). ^[2] These are genuine limitations worth acknowledging. But the mechanistic logic, corroborated by the metabolic framework (Ludwig, 2026) and the Randle Cycle biochemistry (Hue & Taegtmeyer, 2009), holds: it is the rate of carbohydrate digestion, shaped by the food matrix, that determines whether the late postprandial period is physiologically calm or neurologically urgent.

The 4pm craving is diagnostic information. It’s the body reporting on what happened at breakfast and lunch.

If the craving is predictable, arriving reliably at the same time each day, and if it specifically pulls toward high-GI foods rather than general hunger, it’s worth taking seriously as a signal rather than treating it as a test of character. The hypothalamic AgRP neurons don’t fire because someone lacks discipline. They fire because glucose has dropped below a threshold and the fuel-switching alternative has been blocked.

There is something worth knowing about the circadian layer here, too. The repeated glucose crash and snacking cycle driven by high-GI ultra-processed foods doesn’t just cause problems in isolation, it distributes eating events across the full day in a way that flattens the normal high-amplitude rhythm of appetite. ^[1] That rhythm should be lowest in the morning and appropriately elevated in the afternoon; when constant grazing disrupts it, the appetite regulation system becomes less responsive to its own timing signals. Chronically disrupted circadian eating patterns are associated with increased risk of obesity, type 2 diabetes, and cardiovascular disease. ^[1]

The 4pm craving, experienced repeatedly, may therefore be indicating not just a single metabolic event, but a pattern of circadian and metabolic dysregulation that runs across the whole day.

This section carries the most important caveat in the article: there is no single intervention that resolves the 4pm craving, because the craving has multiple contributing causes. What the evidence does support is addressing the root causes rather than managing the symptoms.

Reduce the glycaemic load of lunch. This is the most directly evidenced intervention, following from Lennerz et al. (2013) directly. The goal is not to remove carbohydrate but to preserve the food matrix: intact grains, legumes, vegetables, and proteins that slow digestion and blunt the early postprandial glucose and insulin spike. The mechanism is well established. The four-hour brain response depends on the early postprandial glucose and insulin trajectory.

Avoid the fat plus refined carbohydrate combination. Not fat, and not carbohydrate in isolation, but the UPF combination that simultaneously delivers the malonyl-CoA block and the glucose that causes it. This is the metabolically specific harm identified by the Randle Cycle analysis (Hue & Taegtmeyer, 2009). ^[4]

Consider eating timing. Concentrating caloric intake earlier in the active phase of the day, and avoiding grazing across the full 24 hours, supports circadian resonance and has clinical evidence of metabolic benefit in time-restricted eating research. ^[1] The evidence that morning-weighted caloric intake specifically reduces afternoon cravings is indirect rather than direct; this is a reasonable mechanistic inference rather than a confirmed finding, and should be understood as such.

Brief movement after lunch. Post-meal walking activates AMPK in skeletal muscle, which phosphorylates and inactivates acetyl-CoA carboxylase, reducing malonyl-CoA and reopening CPT-1. ^[4] In practical terms, this means fat oxidation becomes accessible sooner after a meal, potentially reducing the duration of CPT-1 inhibition and improving the transition at hour four. The evidence for this, specifically reducing the 4pm craving, is mechanistically supported but not directly tested in human trials to our knowledge. Worth noting, not overstating.

If you’re on GLP-1 medication, or have recently stopped. Tirzepatide and semaglutide suppress appetite through GLP-1 and GIP pathways, which can mask the 4pm craving signal without addressing its metabolic origin. When medication is reduced or stopped, the underlying metabolic inflexibility remains. The dietary framing above, particularly the food matrix and the fat-plus-sugar combination, remains relevant (see our article Life After Mounjaro: Our Practical Guide to Maintaining Weight Loss for a fuller treatment of the post-medication transition). The goal post-medication is to restore the fuel-switching capacity that GLP-1 agonists partially compensate for, not simply to sustain weight through reduced appetite.

The 4pm craving, understood this way, is not a nuisance or a personal weakness. It is the body’s reporting system working as designed, telling you that a cellular-level fuel crisis is occurring, that the fuel-switching mechanism is under stress, and that the metabolic architecture of your eating is placing a recurring load on your brain’s reward circuitry.

The Lennerz data points toward one further implication. Wang et al. (2001) found striatal dopamine D2 receptor availability to be lower in obese individuals than in lean controls, a cross-sectional association that, as Lennerz and colleagues explicitly noted, could not establish causal direction. ^[6] Whether chronic overconsumption reduces receptor availability or whether pre-existing low availability predisposes to overeating remains unresolved. What the Lennerz data add is a plausible mechanism: each high-GI meal may provide a repeated activation stimulus to the nucleus accumbens, and if recurrent activation contributes to receptor downregulation over time, the link between the daily 4pm craving and longer-term reward-system recalibration becomes neurologically coherent, even if not yet confirmed in prospective human trials. This is the mechanism underlying what the Yale Food Addiction Scale 2.0 measures at the symptomatic level: a pattern of compulsive, craving-driven eating that has neurological as well as behavioural dimensions.

The craving isn’t permanent. Metabolic flexibility, the clean switching between fuel sources that prevents the hour-four crisis, is responsive to dietary change. The respiratory quotient measures fuel-use patterns directly. (A respiratory quotient near 0.7 indicates fat oxidation predominating; near 1.0, glucose (it shifts measurably with dietary modification).) The system is recoverable.

If your 4pm pattern feels less like ordinary hunger and more like a compulsion, including eating past fullness, guilt, or a sense of loss of control, our Yale Food Addiction Scale 2.0 questionnaire offers a validated framework for understanding what might be happening at the behavioural level, alongside the metabolic explanation in this article.

The Ancient Mariner eventually found fresh water; not by chance, and wiser for the voyage, with a clearer eye for the currents that had nearly undone him. So too can our biology, once we understand the hidden tides driving that 4 pm craving.