Blood Sugar Ranges for People Without Diabetes

In an adult without diabetes, fasting glucose sits between roughly 70 and 99 mg/dL. Glucose two hours after a 75 g oral glucose load stays under 140 mg/dL. A continuous glucose monitor worn for a couple of weeks shows a 24-hour mean near 99 mg/dL with about 96% of readings between 70 and 140 mg/dL ¹. Post-meal peaks in healthy people usually land in the 110-140 mg/dL band and return toward baseline within two hours, though a fast-absorbing meal can push a perfectly normal person past 140 for a few minutes. Four hours after eating, you should be back at fasting values.

Those are the numbers. The more useful question is what the spread around them looks like, and that is where most of the ranking pages stop.

What the clinical thresholds are, and what they are for

Three tests define the standard categories. They are diagnostic instruments, and applying them to yourself is not a diagnosis. A diagnosis requires a clinician, with repeat testing on a separate day.

• Fasting plasma glucose (8+ hours, venous draw): under 100 mg/dL normal, 100-125 impaired fasting glucose, 126 or above in the diabetes range.
• 75 g OGTT, 2-hour value: under 140 normal, 140-199 impaired glucose tolerance, 200 or above in the diabetes range.
• HbA1c: under 5.7% normal, 5.7-6.4% prediabetes range, 6.5% or above in the diabetes range.

The cut points are population thresholds chosen against retinopathy risk. They are not physiological boundaries. A fasting glucose of 99 and one of 101 are the same measurement with different labels attached. Fingerstick meters carry roughly ±15% error at these levels, and a single venous fasting glucose has a within-person coefficient of variation of about 5-7% day to day, so one value near a cut point tells you very little.

HbA1c has its own failure mode: it is a function of both mean glucose and red cell lifespan. Anemia, hemoglobin variants, recent blood donation, and high erythrocyte turnover all shift it independently of glucose. CGM catches dysglycemia that HbA1c and fasting glucose miss. In a Spanish cohort without diagnosed diabetes, continuous monitoring was more sensitive than either static test at detecting abnormal glucose patterns ².

What CGM data from non-diabetic people look like

The reference set worth knowing is the multicenter study by Shah and colleagues: 153 participants without diabetes, ages 7 to 80, wearing a blinded Dexcom G6 for up to 10 days. Mean 24-hour sensor glucose was about 99 mg/dL, median time in 70-140 mg/dL was roughly 96% of the day, and time above 140 was a small single-digit percentage ¹. Earlier reference work using older sensors reached similar conclusions: healthy adults spend the large majority of the day inside a narrow band, and glycemic variability is low ³⁴.

The distribution has a long right tail that most summaries hide. Brief excursions above 140 are common in healthy people, especially after a high-glycemic meal or in the hour after waking. Exercise, meal composition, and meal timing all move the trace in ways that make a single spike uninterpretable ⁵. Large-scale profiling of thousands of non-diabetic individuals shows that CGM metrics vary systematically with age, BMI, and sex, so comparing your mean glucose to a single population number without that context overstates what you know ⁶.

Useful targets for a non-diabetic adult, from that reference data:

• Mean glucose: 90-105 mg/dL over 14 days.
• Coefficient of variation (SD / mean): typically 15-20% in people without diabetes. Above about 25% is worth looking at.
• Time in 70-140: 90%+ .
• Time above 180: near zero.
• Nocturnal mean: usually flat and within a few mg/dL of fasting.

Getting the data out and analyzing it

Both major sensors give you a CSV.

Dexcom: Clarity web account, Export, choose the date range. The output is a CSV with Timestamp (YYYY-MM-DDThh:mm:ss), Event Type, Glucose Value (mg/dL), and a few event columns. Readings are every 5 minutes. High and low readings are clamped as the strings High and Low rather than numbers, which will silently turn your glucose column into an object dtype if you do not handle them.

Abbott: LibreView, Export data, gives a CSV with historic (15-minute) and scan/real-time rows distinguished by a Record Type column. Filter to type 0 for the historic series or you will double-count.

A minimal pipeline:

import pandas as pd
df = pd.read_csv("clarity.csv", skiprows=range(1,11))
df = df[df["Event Type"] == "EGV"]
g = pd.to_numeric(df["Glucose Value (mg/dL)"], errors="coerce")
t = pd.to_datetime(df["Timestamp (YYYY-MM-DDThh:mm:ss)"])
s = pd.Series(g.values, index=t).sort_index()
s = s.resample("5min").mean().interpolate(limit=2)   # gaps >10 min stay NaN

mean = s.mean()
cv   = s.std() / mean * 100
tir  = ((s >= 70) & (s <= 140)).mean() * 100

Do not interpolate across long gaps. A sensor that fell off overnight will otherwise fill eight hours of flat line at your daytime average and drag every metric.

For meal responses, compute incremental AUC over a 2-hour window with baseline set to the mean of the 15 minutes before the first bite, using trapezoidal integration (numpy.trapz on a 5-minute grid). Peak height and time-to-return-to-baseline are more reproducible than iAUC when meals are small.

Sensor error and the artifacts you will see

CGM measures interstitial fluid. It does not measure blood. Two consequences:

Lag. Interstitial glucose trails plasma by roughly 5-10 minutes, more when glucose is changing fast. Your true post-meal peak is usually earlier and slightly higher than the sensor shows.

Accuracy. Current sensors run around 9-10% MARD against a reference method, and error is not uniform across the range. In the flat 70-110 band a 10% error is 8-11 mg/dL, which is the entire width of the distinction people try to draw between “excellent” and “fine” ⁷. Two sensors worn simultaneously on the same person disagree by clinically visible amounts ⁴.

Common artifacts:

• Compression lows. Sleeping on the sensor produces a 40-60 mg/dL drop that recovers within 20-30 minutes of rolling over. If your only lows are between 1 and 5 a.m. and V-shaped, the cause is mechanical. It is not physiological.
• Day-1 noise. The first 12-24 hours after insertion are the least accurate part of wear. Consider discarding day 1 from summary statistics.
• Ascorbic acid interference. High-dose vitamin C can falsely elevate readings on some Abbott sensors. Log supplements.
• End-of-wear drift. The last day often reads low.

For anything you intend to act on, confirm with a venous draw. There is still no consensus on what non-diabetic CGM patterns mean for long-term risk, and the evidence base for using these sensors in people without diabetes is early ⁸⁹.

A meal test you can run on yourself

Pick one meal and eat it three times on three separate mornings after a 10-hour fast, at the same hour. Allow no exercise in the preceding 12 hours and none for 3 hours after. Log the first bite time to the minute. Compare peak height and the 120-minute value across the three trials. The replicate spread tells you your own measurement noise, which is typically larger than people expect. Without that, a single 155 mg/dL peak means nothing.

Confounders that will wreck a comparison if uncontrolled: sleep duration the night before, prior-day exercise (which lowers responses for up to 48 hours), caffeine, meal order, and time of day ⁵.

When a high reading is an emergency

Glucose above 250 mg/dL together with vomiting, abdominal pain, or deep or rapid breathing is a medical emergency. The same is true when it comes with fruity breath odor, confusion, or drowsiness. Go to an emergency department. The same applies to any reading above 300 in someone who feels unwell, and to essentially any reading above 400-500, whatever the symptoms. Do not try to interpret it yourself and do not wait for a repeat measurement. A number that high in someone who has never had a diabetes diagnosis needs a clinician the same day regardless of how well they feel, and confirmation with a lab draw, since meters and sensors can both fail high.

Questions people also ask

How high does a non-diabetic’s blood sugar go after a meal? Usually a peak of 110-140 mg/dL at 30-60 minutes, back near baseline by two hours. Excursions somewhat above 140 after a large or fast-absorbing meal occur in healthy people, but they are brief and uncommon in aggregate. Reference CGM data put time above 140 at a few percent of the day ¹.

What should blood sugar be four hours after eating? Back at fasting level, roughly 70-99 mg/dL. Persistent elevation at four hours is not a normal pattern and is worth a lab fasting glucose and HbA1c ordered by a clinician.

What is the ideal blood sugar for someone without diabetes? There is no single ideal. The reference pattern is a 24-hour mean near 99 mg/dL, low variability (CV around 15-20%), and most of the day between 70 and 140 ¹³. Chasing a flatter trace than that has no outcome evidence behind it.

Should I go to the ER if my blood sugar is over 250 or over 300? With any of the symptoms listed above, yes, immediately. Without symptoms, an unexplained reading that high still needs same-day contact with a clinician and a confirmatory lab test.

What are alarming signs that blood sugar is too high? The signs are excessive thirst and urination, blurred vision, and unexplained weight loss. They also include nausea or vomiting, fruity-smelling breath, rapid deep breathing, and confusion. Those last four are emergency signs.

Does a normal HbA1c rule out a glucose problem? No. HbA1c is a time-averaged proxy and can be normal in people whose CGM traces show frequent excursions, which is exactly what continuous monitoring picks up over static tests ²¹⁰.

Footnotes

1. Viral N Shah, Stephanie N DuBose, Zoey Li, et al. Continuous Glucose Monitoring Profiles in Healthy Nondiabetic Participants: A Multicenter Prospective Study. The Journal of Clinical Endocrinology & Metabolism, 2019. https://doi.org/10.1210/jc.2018-02763 ↩ ↩² ↩³ ↩⁴

2. Santiago Rodriguez-Segade, Javier Rodriguez, Félix Camiña, et al. Continuous glucose monitoring is more sensitive than HbA1c and fasting glucose in detecting dysglycaemia in a Spanish population without diabetes. Diabetes Research and Clinical Practice, 2018. https://doi.org/10.1016/j.diabres.2018.05.026 ↩ ↩²

3. Nathan R. Hill, Nick S. Oliver, Pratik Choudhary, et al. Normal Reference Range for Mean Tissue Glucose and Glycemic Variability Derived from Continuous Glucose Monitoring for Subjects Without Diabetes in Different Ethnic Groups. Diabetes Technology & Therapeutics, 2011. https://doi.org/10.1089/dia.2010.0247 ↩ ↩²

4. Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group. Variation of Interstitial Glucose Measurements Assessed by Continuous Glucose Monitors in Healthy, Nondiabetic Individuals. Diabetes Care, 2010. https://doi.org/10.2337/dc09-1971 ↩ ↩²

5. Stephanie N. DuBose, Zoey Li, Jennifer L. Sherr, et al. Effect of Exercise and Meals on Continuous Glucose Monitor Data in Healthy Individuals Without Diabetes. Journal of Diabetes Science and Technology, 2020. https://doi.org/10.1177/1932296820905904 ↩ ↩²

6. Ayya Keshet, Smadar Shilo, Anastasia Godneva, et al. CGMap: Characterizing continuous glucose monitor data in thousands of non-diabetic individuals. Cell Metabolism, 2023. https://doi.org/10.1016/j.cmet.2023.04.002 ↩

7. David Rodbard. Continuous Glucose Monitoring: A Review of Successes, Challenges, and Opportunities. Diabetes Technology & Therapeutics, 2016. https://doi.org/10.1089/dia.2015.0417 ↩

8. Filip Wilczek, Jan Gerrit van der Stouwe, Gloria Petrasch, et al. Non-Invasive Continuous Glucose Monitoring in Patients Without Diabetes: Use in Cardiovascular Prevention—A Systematic Review. Sensors, 2025. https://doi.org/10.3390/s25010187 ↩

9. David C. Klonoff, Kevin T. Nguyen, Nicole Y. Xu, et al. Use of Continuous Glucose Monitors by People Without Diabetes: An Idea Whose Time Has Come?. Journal of Diabetes Science and Technology, 2022. https://doi.org/10.1177/19322968221110830 ↩

10. R. P. Radermecker, A. Sultan, C. Piot, et al. Continuous glucose monitoring as a tool to identify hyperglycaemia in non‐diabetic patients with acute coronary syndromes. Diabetic Medicine, 2009. https://doi.org/10.1111/j.1464-5491.2008.02643.x ↩