The dawn phenomenon — why your fasting glucose spikes

The dawn phenomenon is the predictable, hormonally driven rise in fasting glucose that occurs between 4 AM and 8 AM — with no food, no insulin skipped, no illness to explain it. You go to bed with a glucose of 110 mg/dL, wake up at 130 mg/dL, and assume you did something wrong. You did not. The cause is a coordinated release of counter-regulatory hormones — cortisol, growth hormone, glucagon, and epinephrine — that begin secreting in the small hours to prepare the body for waking. These hormones signal the liver to release stored glucose (hepatic glucose output) and simultaneously reduce peripheral insulin sensitivity. In people without diabetes, the pancreas simply produces more insulin to handle the surge. In Type 1 and Type 2 diabetes, that compensatory response is absent or blunted, so fasting glucose climbs. Per studies published in Diabetes Care, up to 54% of people with Type 2 diabetes experience clinically significant dawn phenomenon (defined as a morning rise of ≥20 mg/dL). Identifying it — and distinguishing it from the Somogyi effect, which has opposite treatment implications — is one of the highest-value things a person with diabetes can do for their A1C.

The hormonal cascade that drives early-morning glucose

The dawn phenomenon is not a blood sugar problem that starts in the blood. It starts in the brain and the adrenal glands, and the liver is where the glucose damage actually happens.

Between 2 AM and 4 AM, the hypothalamus begins increasing pulsatile secretion of growth hormone releasing hormone (GHRH). This triggers the pituitary to release growth hormone (GH) in bursts. GH is directly counter-regulatory: it promotes lipolysis, antagonises insulin action in peripheral tissues, and drives the liver to upregulate gluconeogenesis — the de novo synthesis of glucose from amino acids and glycerol.¹

Cortisol follows roughly 30–60 minutes later, peaking between 6 AM and 8 AM in the typical circadian pattern. Cortisol increases hepatic glucose output by inducing key gluconeogenic enzymes (PEPCK and G6Pase), promotes protein catabolism in muscle (releasing amino acids for gluconeogenesis), and directly reduces insulin receptor signalling at the cellular level. The cortisol-driven reduction in insulin sensitivity is measurable — insulin requirements in people with Type 1 diabetes are demonstrably higher per gram of carbohydrate at 7 AM than at noon.²

Glucagon, secreted by pancreatic alpha cells, adds a third layer. Glucagon directly stimulates hepatic glycogenolysis — the breakdown of stored glycogen into glucose — flooding the portal circulation with glucose before you eat a single bite. Epinephrine (adrenaline) completes the picture by further stimulating glycogenolysis and suppressing insulin secretion from beta cells.¹

The net result: the liver is producing and releasing glucose at an elevated rate, peripheral tissues are less responsive to whatever insulin is present, and the beta cells (compromised in Type 2 diabetes) or absent (Type 1 diabetes) cannot increase output to compensate. Fasting glucose rises accordingly. The entire sequence is teleologically sensible — the body is preparing itself for the metabolic demands of waking — but it is physiologically inconvenient for anyone whose glucose regulation is already impaired.

The magnitude varies substantially between individuals. Studies using continuous glucose monitoring (CGM) in people with Type 2 diabetes show dawn phenomenon magnitudes ranging from less than 10 mg/dL (subclinical) to over 50 mg/dL in individuals with poorly controlled basal glucose or high cortisol reactivity.² Stress, poor sleep, and alcohol the previous evening can amplify the response. Consistent sleep schedules and low-stress evenings tend to attenuate it.

How to confirm you have the dawn phenomenon — not something else

Not every elevated fasting glucose in the morning is the dawn phenomenon. Three other explanations must be ruled out before you attribute your morning highs to the hormonal cascade described above.

Under-basal dosing is the most common. If your long-acting insulin (Type 1) or basal metabolic function (Type 2) is insufficient to cover overnight glucose, your fasting glucose will be high regardless of the dawn phenomenon. Under-basal dosing produces a flat-to-gradually-rising CGM trace through the night, not the characteristic smooth upswing that begins around 3–4 AM. The distinguishing test: check glucose at 2 AM. If it is already elevated (above 130 mg/dL), you likely have insufficient basal coverage, not a pure dawn phenomenon.

Rebound hyperglycaemia after nocturnal hypoglycaemia (historically called the Somogyi effect — discussed in the next section) produces a different morning high, triggered by a different mechanism.

Simple fasting hyperglycaemia — elevated glucose that has been high since dinner — is distinguishable by checking glucose at 10 PM and comparing it to the morning reading. If glucose is already high at bedtime and stays elevated, the problem is post-dinner management, not the dawn phenomenon.

To confirm the dawn phenomenon specifically:²

1. Log fasting glucose every morning for 7–14 consecutive days, at the same time, before eating, before any correction. You need a trend, not a single data point.
2. If you have a CGM, review the overnight trace. The dawn phenomenon produces a characteristic shape: glucose is stable or slightly drifting from 10 PM to 2–3 AM, then rises smoothly and continuously through to 6–8 AM. The rise is gradual, not sudden.
3. Check glucose at 2–3 AM on at least two nights. If glucose is normal (under 100 mg/dL) at 3 AM and elevated by 7 AM, the dawn phenomenon is confirmed.

The ADA Standards of Care 2024 acknowledge CGM as the most informative tool for identifying nocturnal glucose patterns, particularly for people who cannot perform multiple overnight fingerstick checks.³

Dawn phenomenon vs Somogyi effect — the test that separates them

The Somogyi effect, described by endocrinologist Michael Somogyi in 1959, proposes that nocturnal hypoglycaemia triggers a counter-regulatory hormone surge that produces morning hyperglycaemia — a rebound. For decades, it was a widely taught explanation for high morning glucose readings in people on insulin.

The problem is that the Somogyi effect, in its classical form, appears to be rare. Multiple CGM-based studies have examined overnight glucose trajectories in people with Type 1 and Type 2 diabetes who presented with high fasting readings. The overwhelming majority showed a smooth upward curve (consistent with the dawn phenomenon), not a nocturnal hypoglycaemic dip followed by a rebound.¹ A 2003 study in Diabetes Care using CGM found that less than 1 in 5 overnight high glucose readings in Type 1 diabetes was preceded by detectable hypoglycaemia — the rest were dawn phenomenon or under-basal dosing.²

This matters clinically because the treatments are opposite. If you assume Somogyi and reduce your bedtime basal insulin to prevent the alleged nocturnal low, you will worsen the actual dawn phenomenon by leaving the morning hours even more underinsulinised. Conversely, if you have true nocturnal hypoglycaemia and increase basal insulin, you risk a dangerous overnight low.

The separating test is simple: check blood glucose at 2–3 AM on two or three consecutive nights. Record the result alongside the morning reading:

• 2–3 AM glucose is low (under 70 mg/dL) + morning glucose is high: Somogyi effect pattern (rare). Discuss bedtime snack or basal reduction with your care team.
• 2–3 AM glucose is normal or high + morning glucose is higher still: Dawn phenomenon or under-basal dosing. Treatment is more basal coverage, not less.

CGM makes this test automatic — the overnight trace shows the full picture without waking up to measure.

Dietary strategies that blunt the morning glucose rise

The hormonal cascade of the dawn phenomenon cannot be switched off, but its glycaemic impact can be moderated. Dietary interventions work by reducing the substrate available for hepatic glucose production and by blunting the post-meal spike that amplifies the morning rise.

A lower-carbohydrate evening meal. Hepatic glycogen stores filled by a high-carbohydrate dinner give the liver more substrate for overnight glycogenolysis. A dinner with 30–45 g of carbohydrate (rather than 60–80 g) reduces the glycogen pool without causing hypoglycaemia, and observational data in Type 2 diabetes shows measurable reductions in fasting glucose over 4–8 weeks on lower-carbohydrate evening meals.³

Avoiding high-glycaemic bedtime snacks. A bowl of cereal, crackers, or sweetened yogurt at 9 PM adds carbohydrate that peaks in the bloodstream around 11 PM, is partially stored as glycogen, and becomes substrate for early-morning hepatic output. Replacing a starchy bedtime snack with a protein-and-fat option — 1 oz of cheese and a few almonds, for example — preserves satiety without loading the hepatic glycogen pool. The carbohydrate load from this swap is under 3 g, versus 20–30 g from a small bowl of cereal.

A post-dinner walk. Even 15 minutes of brisk walking after the evening meal meaningfully reduces post-prandial glucose by accelerating glucose uptake in skeletal muscle through a non-insulin-dependent GLUT4 pathway. Studies in people with Type 2 diabetes show that a 15-minute walk within 30 minutes of dinner reduces post-meal glucose area under the curve by approximately 12%, and may modestly reduce fasting glucose the following morning by limiting glycogen restorage.⁴

Timing matters for morning meals. The dawn phenomenon peaks between 6 and 9 AM in most people. A breakfast eaten after 9 AM (if feasible) encounters a partial normalization of insulin sensitivity and may produce a smaller post-prandial spike than the same meal eaten at 7 AM. This is not always practical, but it is a legitimate modification for shift workers or those with flexible schedules.

Medication and insulin adjustments for the dawn phenomenon

For people on insulin therapy, the dawn phenomenon has well-established pharmacological solutions. For people on oral agents for Type 2 diabetes, the options are more limited but meaningful.

Basal insulin timing (Type 1 and insulin-using Type 2). Long-acting insulin analogues (glargine, detemir, degludec) reach effective tissue concentrations 2–4 hours after injection and maintain them for 20–24 hours. When injected in the morning, their action profile may be waning by 3–4 AM — precisely when the dawn phenomenon begins. Shifting basal insulin injection to bedtime (10–11 PM) ensures that peak tissue concentrations align with the 3–6 AM window when hepatic glucose output is surging. Clinical data consistently shows that bedtime glargine reduces fasting glucose more effectively than morning glargine in people with Type 2 diabetes.³

Continuous subcutaneous insulin infusion (insulin pump). Pump therapy allows programmable basal rates by time of day. The standard solution for the dawn phenomenon is a programmed basal rate increase starting at 3 AM — for example, increasing from 0.8 units/hour to 1.2 units/hour between 3 and 7 AM, then returning to baseline. This is the most precise pharmacological intervention available and is guided by CGM data. Pump users who are not using this feature and experiencing dawn phenomenon should raise it with their diabetes technology team.

Metformin extended-release (Type 2 diabetes). Metformin’s primary mechanism is hepatic glucose output suppression. Extended-release metformin taken with the evening meal has a pharmacodynamic profile that delivers maximal hepatic suppression overnight — directly targeting the driver of the dawn phenomenon. Per ADA Standards of Care 2024, metformin remains the preferred first-line agent for Type 2 diabetes partly because of this overnight hepatic mechanism.³

GLP-1 receptor agonists. Drugs like semaglutide (Ozempic/Wegovy) and liraglutide suppress glucagon secretion and slow gastric emptying. Glucagon suppression directly reduces hepatic glucose output. Published data on GLP-1 agonists in Type 2 diabetes shows significant reductions in fasting glucose, with a proportion of that reduction attributable to overnight glucagon suppression affecting the dawn phenomenon.³

Using a food and glucose log to quantify your personal curve

Population statistics describe the average dawn phenomenon. Your dawn phenomenon has a personal magnitude, timing, and set of triggers that only your own data can reveal. Building a structured log turns a vague problem into a solvable one.

A useful 7-day log captures: dinner composition (total carbohydrate, time of last meal), bedtime glucose, 3 AM glucose (if checking), fasting glucose at wake-up, and any unusual stressors or sleep disruptions. After one week, patterns become visible. If every morning after a dinner with 70+ g carbohydrate shows a rise of 25+ mg/dL, but dinners under 45 g carb show rises of 10–15 mg/dL, you have a specific dietary lever. If the 3 AM glucose is reliably normal but the 7 AM glucose is reliably elevated by 20–30 mg/dL, you have confirmed the dawn phenomenon mechanism rather than under-basal dosing.

CalEye’s meal log captures the dinner composition side of this — total carbohydrate, macro breakdown, and meal timing — and when integrated with CGM data, surfaces correlations between evening dietary choices and overnight glucose trajectories. You do not need a CGM to start this analysis; fingerstick checks at bedtime, 3 AM, and fasting provide the same diagnostic information, just more laboriously. The goal is to move from “my morning glucose is just high sometimes” to “my morning glucose rises by approximately 22 mg/dL on nights when my dinner carbohydrate exceeds 55 g, and this is addressable.”

The dawn phenomenon is not a malfunction. It is a normal circadian physiology problem that becomes clinically significant in diabetes. It is also, for most people, a controllable one — with the right data, the right timing adjustments, and the right conversation with your care team.

References

1. Bolli GB, De Feo P, De Cosmo S, et al. “Demonstration of a Dawn Phenomenon in Normal Human Volunteers.” Diabetes 33, no. 12 (1984): 1150–1153.

2. Rybicka M, Krysiak R, Okopień B. “The Dawn Phenomenon and the Somogyi Effect — Two Phenomena of Morning Hyperglycaemia.” Endokrynologia Polska 62, no. 3 (2011): 276–284.

3. American Diabetes Association Professional Practice Committee. “Facilitating Positive Health Behaviors and Well-being to Improve Health Outcomes: Standards of Care in Diabetes — 2024.” Diabetes Care 47, Supplement 1 (2024): S77–S110.

4. Colberg SR, Sigal RJ, Yardley JE, et al. “Physical Activity/Exercise and Diabetes: A Position Statement of the American Diabetes Association.” Diabetes Care 39, no. 11 (2016): 2065–2079.

5. Porcellati F, Rossetti P, Busciantella NR, et al. “Comparison of Pharmacokinetics and Dynamics of the Long-Acting Insulin Analogues Glargine and Detemir at Steady State in Type 1 Diabetes.” Diabetes Care 30, no. 10 (2007): 2447–2452.