Dawn phenomenon: why your blood sugar is higher in the morning

Share
Summarize with AI

If your meter or continuous glucose monitor (CGM, the sensor worn on the arm) reads higher at 7 a.m. than it did at bedtime, and you ate nothing in between, your liver put that glucose there. From around 4 a.m. your body starts getting ready to wake, and part of that is a hormone signal telling the liver to release more sugar into the blood. Doctors first described the morning rise in people with type 1 diabetes in the early 1980s 5, then found a smaller version of it in healthy volunteers 2. Whether it shows on your meter depends on how well your insulin keeps up.

Why your blood sugar rises while you sleep

Overnight you are fasting, so the glucose in your blood comes from your liver. In the early hours, the hormones that prepare you to wake start to climb. In a 1984 study that sampled eight healthy volunteers every half hour from 1 to 9 a.m., growth hormone came in pulses between 1 and 4:30 a.m., and cortisol, adrenaline and noradrenaline rose from their night-time lows between 4 and 6:30 a.m. 2. After 5:30 a.m. the liver's glucose output rose, blood glucose rose, and insulin rose with it 2. Glucagon, often listed as a cause, did not change 2. The rise itself causes no symptoms; you find it on a meter, a CGM or a fasting lab result.

How a quiet night becomes a higher morning number

1
Growth hormone pulses, 1 to 4:30 a.m.
Pulses in early sleep make the liver and muscles less responsive to insulin a few hours later 7.
2
Cortisol and adrenaline climb, 4 to 6:30 a.m.
The wake-up hormones rise from their overnight low 2.
3
The liver releases more glucose
From about 5:30 a.m. the liver makes and releases more glucose into the blood 2.
4
Insulin decides what you see
A healthy pancreas raises insulin to match, so the number barely moves. With diabetes or insulin resistance, insulin falls behind and the rise shows on the meter.
Timings from Bolli and colleagues, who sampled healthy volunteers through the night 2. The growth hormone role comes from Campbell and colleagues: in six people with type 1 diabetes, preventing the night-time growth hormone surges stopped the dawn rise, and replacing them brought it back 7.

These results indicate that a dawn-like phenomenon, initiated by an increase in glucose production, occurs in nondiabetic individuals.Bolli GB et al., Diabetes, 1984 2

The same paper proposed that the early-morning rise seen in type 1 and type 2 diabetes is an exaggerated version of that normal daily rhythm 2. So the dawn phenomenon is a normal morning rise in glucose output; type 2 diabetes, prediabetes and insulin resistance make it big enough to see.

Overnight glucose with normal insulin and with type 2 diabetes
Normal insulinType 2 diabetes
Glucose, mg/dL10 p.m.Midnight2 a.m.3 a.m.4 a.m.5 a.m.6 a.m.7 a.m.8 a.m.Lowest point
Illustrative, not data from one person. The shape follows the studies: in healthy volunteers glucose and insulin rise together after about 5:30 a.m., so the climb stays small 2; in type 2 diabetes insulin falls behind, and this curve shows a rise above the 20 mg/dL cut-off 1.

Does the dawn phenomenon happen if you don't have diabetes?

Yes, in a small way. Bolli's healthy volunteers had a measurable dawn rise in both glucose and insulin, so the rise is not limited to diabetes 2. With normal insulin the extra glucose is matched, so a CGM on a healthy person often shows only a gentle slope up toward waking. That slope is normal.

In a 2019 study of 153 healthy people wearing a blinded sensor (Shah and colleagues), glucose sat between 70 and 140 mg/dL for 96% of the day, averaged about 99 mg/dL, and went above 140 for about 30 minutes a day.

CGM for non-diabetics: what normal looks like, from Shah VN et al., J Clin Endocrinol Metab, 2019 9

Blinded means the volunteers could not see their readings. A gentle overnight climb that settles in the 80s or 90s on waking needs no action. A waking value that sits at 100 mg/dL or more most mornings is a reason to ask your doctor about a lab fasting glucose and an A1C, because a CGM cannot diagnose prediabetes. The CGM guide covers what else a sensor can and cannot tell you.

Who gets a noticeable dawn rise

Most people with diabetes. In a 1984 study that kept glucose steady overnight with an insulin drip, insulin needs rose by at least 50% for an hour and a half after 6 a.m. in 77% of people with type 2 diabetes and 75% with type 1 8. In type 2 diabetes the rise was larger in people who slept badly 10. Prediabetes and insulin resistance are likely to show it too, because insulin is slower to keep up, though these groups have been studied less. It is not the same every night: in five people with type 1 diabetes studied four times each, it appeared on 17 of the 20 nights 8, so judge it on several mornings or a week of CGM data, not one.

How big a rise counts as a problem

Researchers measure the dawn phenomenon from the night's lowest reading, often somewhere between 2 and 4 a.m., to the last reading before breakfast. In a 2013 study of 248 people with type 2 diabetes who were not on insulin and wore a CGM, Monnier and colleagues used a rise above 20 mg/dL (1.1 mmol/L) as the cut-off 1.

16 mg/dLmedian rise from the night's lowest reading to breakfast in type 2 diabetes248 people not on insulin, wearing a CGM; the rise added about 12 mg/dL to their 24-hour average glucose 1

The size of the rise tells you how strong your dawn phenomenon is. Your waking number on its own tells you where you stand. Here are the waking glucose bands most US labs and clinics use, in both units:

Waking or fasting glucosemg/dLmmol/LWhat it means
LowUnder 70Under 3.9Low for anyone on insulin or a sulfonylurea; check for a night-time low 6
Normal fasting70 to 993.9 to 5.5Normal on a lab test 6
Prediabetes range100 to 1255.6 to 6.9Prediabetes on a lab test; worth repeating or pairing with an A1C 6
Diabetes range126 or more7.0 or moreDiabetes if a second test agrees 6
Usual target with diabetes80 to 1304.4 to 7.2The ADA's usual pre-meal goal; your own target may differ 6
CalculatorGlucose converterMeter or app in mmol/L? Convert any waking reading to mg/dL, and back.Work it out

Example: a waking glucose of 104 mg/dLmg/dL

104 mg/dLA little high · same as 5.8 mmol/L
Low40–70Normal70–100Prediabetes range100–126Diabetes range126–160
One morning reading of 104 mg/dL is not a diagnosis. Note the time it was taken and read it with an A1C; if it was a lab test, ask your doctor whether to repeat it.
Cut-offs from the ADA Standards of Care 2025: prediabetes 100 to 125 mg/dL, diabetes 126 mg/dL or more confirmed on a second test 6. Pages for single readings: fasting glucose 105 and fasting glucose 110.

Dawn phenomenon or Somogyi effect? How to tell

The Somogyi effect is the idea that a low blood sugar in the night triggers a burst of stress hormones and a rebound high by morning. It sounds plausible, and the researchers who tested it note that it is still firmly embedded in clinical practice 3. If you were advised to lower your evening insulin for this reason, bring overnight readings or CGM data to your diabetes team before changing anything. Their CGM study followed 89 people with type 1 diabetes and counted a night low as a sensor reading under 63 mg/dL (3.5 mmol/L). The fingerstick glucose on waking was typically 99 mg/dL (5.5 mmol/L) after nights with a low and 261 mg/dL (14.5 mmol/L) after nights without one 3. Rebound highs above 180 mg/dL (10 mmol/L) after a night low happened only twice, both probably after the low had been treated 3.

Our data indicate that, in clinical practice, the Somogyi effect is rare.

Choudhary P et al., Diabetic Medicine, 2013 3

When the waking fingerstick was below 90 mg/dL (5 mmol/L), there had been a low during the night on 94% of nights 3. On insulin, a low-normal morning number deserves attention too, because it can mean a low you slept through; mention it to your care team.

To see which pattern is yours, check your glucose at bedtime, between 2 and 3 a.m. and on waking, on two or three nights with your usual dinner and doses, or read the same three points off a CGM trace. Then compare them with the table.

PatternBedtimeAround 3 a.m.WakingWhat it points to
Dawn phenomenonIn rangeIn range, the night's lowest readingMore than 20 mg/dL above the 3 a.m. readingWake-up hormones outpacing insulin 1
Somogyi rebound (rare; on insulin or a sulfonylurea)In rangeLow, under 70 mg/dL 6HighA night-time low followed by a rebound 3
Hidden night low (on insulin or a sulfonylurea)In rangeLowLow-normal, 90 mg/dL or underA low you slept through; tell your care team 3
Medication wearing off or late dinnerHigh, or risingAlready highHighLong-acting insulin or tablets running out overnight, or a late, carb-heavy meal still being absorbed

Why your morning lab result may read higher

A fasting glucose drawn at 8 a.m. includes the dawn rise. In the US NHANES III survey, 12,882 adults without known diabetes were randomly assigned to a morning or an afternoon visit, and fasted before both 4. Fasting glucose averaged 97.4 mg/dL in the morning group and 92.4 mg/dL in the afternoon group, and half as many afternoon-tested people reached the 126 mg/dL diabetes cut-off 4. The two groups had the same A1C, so the gap came from the time of day. The authors worked out that an afternoon cut-off of 114 mg/dL found the same share of diabetes as 126 in the morning, and that using 126 in the afternoon would miss about half of undiagnosed cases 4.

Note the draw time on each report and compare morning with morning; the blood test prep guide covers fasting and timing. Treat a single fasting value of 101 to 105 mg/dL as a reason to ask your doctor about a repeat test. Lean on A1C for the long view, because the time of the draw does not change it 4; A1C vs fasting glucose explains how the two tests divide the work. If the flag on your report says high, the high fasting glucose page walks through the next steps.

Does the dawn phenomenon matter for long-term health?

In diabetes, yes. In Monnier's study, the dawn phenomenon raised A1C by about 0.4% (4.3 mmol/mol) and 24-hour average glucose by 12.4 mg/dL. It was there to a similar degree in people on diet alone, on tablets that help insulin work (such as metformin), and on tablets that push the pancreas to release more insulin (such as glipizide), and the authors concluded that none of the diabetes tablets then available removed it 1. Those tablets still lower glucose across the day, so this is no reason to stop one; that decision belongs with your prescriber. For someone sitting at an A1C of 7.2% against a 7% goal, that 0.4% matters.

Without diabetes, a small morning rise with a normal fasting glucose and a normal A1C is physiology, and no study shows harm from it. The signal worth watching is a waking number that drifts up over months while A1C creeps toward 5.7%, which is how prediabetes often first shows.

How to stop or lower the dawn phenomenon

The evidence is uneven, so here are the options graded by how well they have been tested against the dawn rise itself.

OptionStudied inEvidenceWhat the studies found
Background (basal) insulin that keeps working through the early morning hours (a pump, or the timing and type of long-acting insulin), set with your diabetes teamPeople on insulin, type 1 and type 2 diabetesStrongestBackground insulin that does not wane before dawn, from a pump or a long-acting insulin, is the approach 30 years of research supports 5; the night-time growth hormone surges explain why the need rises then 7
Weight loss and better insulin sensitivity overallType 2 diabetes, prediabetes, insulin resistanceSomeLowers glucose across the day; trials aimed at the dawn rise alone are scarce
Diabetes tablets such as metforminType 2 diabetes, on prescriptionSomeLower glucose across the day, but the dawn rise was as large on tablets as on diet alone 1; do not stop one without your prescriber
Better sleepType 2 diabetes (small studies)EarlyPoor sleep went with a larger dawn rise in 81 people 10; a prescription sleep medicine (suvorexant) cut the early-morning surge from about 28 to 18 mg/dL after 3 days of treatment in an open-label study of 18 people with type 2 diabetes and insomnia, with no comparison group 11
An earlier, lighter dinner and a walk after itGeneral glucose studies, not the dawn riseModest, indirectLowers the evening rise and the starting point for the night; not tested as a dawn treatment
Bedtime apple cider vinegar11 adults with well-controlled type 2 diabetesWeakWaking glucose a few percent lower 12; it can slow stomach emptying, so ask your care team first if you have gastroparesis or take insulin
MagnesiumNo dawn studiesNone for the dawn riseNo trial against the dawn phenomenon; a blood test shows whether your level is low (magnesium deficiency)

Good sleep habits cost nothing, though the one trial that shrank the dawn rise used a sleep medicine. The deep sleep guide covers what tends to help.

Low carb, keto and fasting: why your morning number can creep up

Some people on very low-carb or keto diets, and people doing longer fasts, report waking glucose a little higher than they expect, sometimes in the high 90s or low 100s, while their A1C and fasting insulin stay low. One proposed explanation, sometimes called physiological insulin resistance, is that after a long stretch with little carbohydrate, muscles take up less glucose to save it for the brain, so the same liver output reads a little higher. No controlled trial has measured this as a dawn effect, so read one waking number alongside A1C and fasting insulin rather than alone. The keto diet guide covers what else the diet does to your labs.

Which blood tests tell you more

The dawn phenomenon has no blood test of its own. Overnight fingersticks or a CGM show it; lab tests show whether it sits on top of a bigger problem. Fasting insulin and HOMA-IR (a score made from your fasting glucose and insulin) can point to insulin resistance before fasting glucose rises, though neither is a standardized diagnostic test. A morning cortisol test does not diagnose the dawn phenomenon, because the cortisol rise behind it is normal; doctors usually look into cortisol only when there are other signs of Cushing's syndrome, such as easy bruising or a rounder face. Growth hormone tests do not help here either.

CalculatorHOMA-IR calculatorHave fasting glucose and fasting insulin from the same draw? Turn them into a HOMA-IR score.Work it out

For how each of these fits on a standard panel, see blood sugar tests, explained.

When to talk to a doctor

  • Your waking glucose is 126 mg/dL (7.0 mmol/L) or more on more than one morning.
  • You are treated for diabetes and your waking readings sit above the target you agreed, most days for two weeks or more.
  • You see readings under 70 mg/dL (3.9 mmol/L) at night or on waking, or you wake with sweats, a headache or vivid dreams. See hypoglycemia.
  • Morning highs come with thirst, peeing a lot, blurred vision or weight loss you did not plan.
  • A new morning pattern starts after a change in medicine, including steroids.
Free, no accountRead your blood testGot a fasting glucose, A1C or insulin result? Read every line of your report in plain words, free.Read my results
Free, no accountDrop in your lab report and see every marker explained in plain words, placed on the body, against your lab's range. It's read on your device: nothing is uploaded and nothing is kept.
The distinctions
  1. 01
    A normal dawn rise vs a dawn phenomenon

    The same liver rise happens in everyone; with normal insulin it is matched and barely shows 2. When insulin lags and the rise from the night's lowest reading to breakfast is above 20 mg/dL (1.1 mmol/L), researchers call it a dawn phenomenon 1.

  2. 02
    Dawn phenomenon vs Somogyi effect

    The dawn phenomenon follows a normal night; the Somogyi effect is a rebound after a night low and is rare. On insulin, a low-normal waking number, 90 mg/dL or under, is the better clue to a hidden night low 3.

  3. 03
    Morning vs afternoon fasting glucose

    Fasting glucose averaged 97.4 mg/dL in morning draws and 92.4 mg/dL in afternoon draws in US adults without known diabetes, and an afternoon cut-off of 114 mg/dL matched 126 in the morning 4; A1C does not depend on the time of day.

Citations
  1. Monnier L, Colette C, Dejager S, Owens D. Magnitude of the dawn phenomenon and its impact on the overall glucose exposure in type 2 diabetes: is this of concern? Diabetes Care 2013;36(12):4057-4062.
  2. Bolli GB, De Feo P, De Cosmo S, et al. Demonstration of a dawn phenomenon in normal human volunteers. Diabetes 1984;33(12):1150-1153.
  3. Choudhary P, Davies C, Emery CJ, Heller SR. Do high fasting glucose levels suggest nocturnal hypoglycaemia? The Somogyi effect, more fiction than fact? Diabet Med 2013;30(8):914-917.
  4. Troisi RJ, Cowie CC, Harris MI. Diurnal variation in fasting plasma glucose: implications for diagnosis of diabetes in patients examined in the afternoon. JAMA 2000;284(24):3157-3159.
  5. Porcellati F, Lucidi P, Bolli GB, Fanelli CG. Thirty years of research on the dawn phenomenon: lessons to optimize blood glucose control in diabetes. Diabetes Care 2013;36(12):3860-3862.
  6. American Diabetes Association Professional Practice Committee. Standards of Care in Diabetes 2025 (sections 2, Diagnosis and classification of diabetes, and 6, Glycemic goals and hypoglycemia). Diabetes Care 2025;48(Suppl 1).
  7. Campbell PJ, Bolli GB, Cryer PE, Gerich JE. Pathogenesis of the dawn phenomenon in patients with insulin-dependent diabetes mellitus. Accelerated glucose production and impaired glucose utilization due to nocturnal surges in growth hormone secretion. N Engl J Med 1985;312(23):1473-1479.
  8. Bolli GB, Gerich JE. The "dawn phenomenon": a common occurrence in both non-insulin-dependent and insulin-dependent diabetes mellitus. N Engl J Med 1984;310(12):746-750.
  9. Shah VN, DuBose SN, Li Z, et al. Continuous glucose monitoring profiles in healthy nondiabetic participants: a multicenter prospective study. J Clin Endocrinol Metab 2019;104(10):4356-4364.
  10. Huang Y, Wang H, Li Y, Tao X, Sun J. Poor sleep quality is associated with dawn phenomenon and impaired circadian clock gene expression in subjects with type 2 diabetes mellitus. Int J Endocrinol 2017;2017:4578973.
  11. Toi N, Inaba M, Kurajoh M, et al. Improvement of glycemic control by treatment for insomnia with suvorexant in type 2 diabetes mellitus. J Clin Transl Endocrinol 2019;15:37-44.
  12. White AM, Johnston CS. Vinegar ingestion at bedtime moderates waking glucose concentrations in adults with well-controlled type 2 diabetes. Diabetes Care 2007;30(11):2814-2815.

Educational context only, not medical advice or a diagnosis. Always discuss your results with a clinician.