Fasting Glucose: Normal Lab Range vs. Functional Optimal Range

Fasting glucose (also called fasting blood sugar or fasting plasma glucose, FPG) measures the amount of glucose in your blood after 8 to 12 hours without food. It is distinct from HbA1c (a 2-to-3 month average), the oral glucose tolerance test (a measured response to a glucose load), and continuous glucose monitoring (a real-time trace of glucose across the day). This article covers the standard laboratory reference range and the narrower target some preventive-medicine clinicians use, and where the evidence for each comes from.

The direct answer: Most U.S. labs and the American Diabetes Association (ADA) treat any fasting glucose under 100 mg/dL as "normal," but that cutoff was set to identify people at high risk of progressing to diabetes, not to define the level associated with the lowest long-term cardiometabolic risk. Some clinicians and cohort studies use a narrower window, roughly 75 to 89 mg/dL, as more consistent with low measured insulin resistance and lower cardiovascular event rates, while values from 90 to 99 mg/dL, though technically "normal," have been associated with higher future risk in some cohorts. Whether that association changes what an individual patient should do depends on additional data, particularly fasting insulin and HbA1c, not on the fasting glucose number alone.

The useful question for most readers is not "is my fasting glucose normal?" but "where does my number sit inside the normal range, and does it need a second data point before I decide it's fine?"

What fasting glucose actually measures

Fasting glucose reflects your liver's glucose output overnight and your tissues' baseline sensitivity to insulin. During an overnight fast, insulin drops and the liver releases glucose from glycogen stores to keep the brain and red blood cells supplied. In someone with normal insulin signaling, counter-regulatory hormones (glucagon, cortisol, growth hormone) keep this output within a narrow band. When the liver overproduces glucose, or tissues resist insulin's signal to take glucose up, fasting values drift upward.

Fasting glucose is a single time point. It does not show how your body handles a carbohydrate load after a meal. A 2-hour oral glucose tolerance test (OGTT), or continuous glucose monitoring (CGM) across normal eating, can reveal post-meal glucose spikes in people whose fasting value looks unremarkable. A 2026 narrative review on CGM as a cardiovascular and longevity screening tool discusses this gap directly, arguing that fasting and HbA1c values can miss glycemic variability that CGM captures, though the authors frame this as hypothesis-generating rather than established practice (PubMed, 2026). That distinction matters for anyone whose fasting number looks fine but who wants a fuller picture of glucose handling.

How the standard "normal" range was set

The ADA classifies fasting plasma glucose into three tiers: normal (under 100 mg/dL), impaired fasting glucose or prediabetes (100 to 125 mg/dL), and diabetes (126 mg/dL or higher, confirmed on a second occasion) (ADA Standards of Care, Diabetes Care). These cutoffs come from population-level analyses correlating fasting glucose with diagnostic thresholds on the OGTT and with progression to diabetes, and they are designed to maximize screening sensitivity for disease, not to describe an individually optimal metabolic state.

The World Health Organization uses a different impaired-fasting-glucose cutoff (110 mg/dL rather than 100 mg/dL) for its intermediate-hyperglycemia category, which is a source of real disagreement between international bodies, not a settled number (WHO, definition and diagnosis of diabetes mellitus). When two major bodies use different cutoffs for the same physiological category, that is itself evidence that 100 mg/dL is a threshold chosen for a specific screening purpose, not a natural boundary of metabolic health.

Standard lab reference ranges are also built statistically: labs sample an "apparently healthy" population and report the range that captures the central 95% of results. If a large share of that reference population already has undiagnosed insulin resistance, hypertension, or dyslipidemia, the reported "normal" range can include people who are not, in a broader cardiometabolic sense, healthy. This is a structural limitation of any lab reference range, not specific to glucose, and it is the main reason some clinicians look at where a number falls inside the range rather than only whether it clears the cutoff.

Why "normal" and "optimal" are not the same question

The debate concentrates in the 90 to 99 mg/dL zone. Some prospective cohort studies have reported that cardiovascular risk rises in a graded, continuous fashion with fasting glucose, without a sharp jump at the 100 mg/dL diagnostic line, and a widely cited pooled analysis of cardiometabolic risk factors from multiple cohorts supports this pattern in principle (PubMed). We flag the exact relative-risk figures sometimes quoted for specific glucose thresholds (for example, precise percentage increases at 100 mg/dL versus 90 mg/dL) as requiring direct verification against the primary paper before they are used in patient-facing material; the graded-risk pattern is better supported than any single point estimate.

Similarly, the Diabetes Prevention Program (DPP) enrolled adults with impaired fasting glucose or impaired glucose tolerance and found that intensive lifestyle intervention (structured physical activity and modest weight loss) reduced progression to type 2 diabetes substantially more than metformin did over the study period (PubMed). The DPP is strong trial evidence that intervening earlier in the glycemic continuum works. Specific year-by-year progression rates for the 95 to 109 mg/dL subgroup, and any attributed quotations from trial investigators, should be verified against the original New England Journal of Medicine report before being restated as fact; we have removed an unverifiable quotation that appeared in an earlier draft of this material rather than risk misattribution.

Some professional consensus statements on insulin resistance argue that fasting glucose in the high-normal range, especially alongside elevated fasting insulin or central adiposity, warrants proactive evaluation rather than reassurance. This represents a professional society's judgment rather than a diagnostic threshold every clinician uses.

Reading your number: a decision framework, not a verdict

The table below is a functional-medicine-informed way to think about a single fasting glucose result. It is not a diagnostic standard, and no single row should be used to self-diagnose or self-treat. It exists to help you decide whether one more test (fasting insulin, HbA1c) is worth requesting before you either dismiss or over-react to a number.

The framework's core limitation: cohort associations between high-normal glucose and long-term risk are population-level findings. They tell you that people with numbers in a certain range have, on average, somewhat higher risk over years of follow-up. They do not tell you what your individual risk is, and they should prompt a conversation and possibly one more test, not a diagnosis.

What a high fasting glucose means

A fasting glucose at or above 100 mg/dL indicates that hepatic glucose output and peripheral glucose uptake are no longer balancing as they should overnight, consistent with insulin resistance, reduced insulin secretion, or both. Prediabetes affects a large share of U.S. adults, and the CDC's national surveillance data are the authoritative current source for prevalence figures; check the CDC's data page directly for the most recent numbers rather than relying on a fixed percentage, since prevalence estimates are updated periodically (CDC, National Diabetes Statistics, accessed 2026).

High-normal and prediabetic glucose values commonly cluster with other findings: elevated triglycerides, low HDL cholesterol, higher waist circumference, and elevated blood pressure. This clustering is the basis of the metabolic syndrome diagnosis, in which fasting glucose of 100 mg/dL or higher is one of five components used by AHA/NHLBI scientific statement criteria (PubMed).

Contributors to a higher fasting reading, in rough order of how directly they act on the liver or insulin sensitivity, include poor or short sleep, chronic stress via cortisol-driven glucose release, sedentary behavior, visceral fat, certain medications (corticosteroids, thiazide diuretics, some antipsychotics), and the dawn phenomenon, a normal early-morning cortisol and growth hormone rise that can push glucose higher than it would be at midnight. Sleep's effect on glucose regulation is documented in the sleep-and-metabolism literature, though exact milligram-per-deciliter effect sizes attributed to a specific number of hours of sleep loss vary by study population and should be treated as approximate (PubMed).

What a low fasting glucose means

Fasting glucose below 70 mg/dL meets the Endocrine Society's definition of hypoglycemia; levels below roughly 54 mg/dL are considered clinically significant and require prompt treatment (Endocrine Society guideline, PubMed). In someone not taking insulin or a sulfonylurea, true fasting hypoglycemia is uncommon and should be investigated rather than dismissed; causes include insulinoma, adrenal insufficiency, severe liver disease, sepsis, and medication error.

Reactive hypoglycemia, which occurs 2 to 4 hours after eating, is more common than fasting hypoglycemia and is not detected by a fasting test at all. The clinical standard for diagnosing a hypoglycemic disorder is Whipple's triad: a measured low glucose, symptoms consistent with hypoglycemia at that time, and resolution of symptoms once glucose is raised. A low number on a lab report, by itself, without symptoms, is not a diagnosis.

Anyone on insulin or a sulfonylurea whose fasting readings run low should contact the prescribing clinician about dose adjustment rather than changing the medication independently. This is a situation where self-management carries real risk.

What actually moves fasting glucose

Lifestyle change is the best-supported first-line approach for glucose in the prediabetic and high-normal range. The DPP trial found that structured physical activity combined with modest weight loss reduced progression to diabetes considerably more than metformin alone over the study's follow-up period (PubMed; 10-year follow-up).

Resistance training has plausible mechanistic support (increased skeletal muscle GLUT4 transporter density) and has been studied in people with prediabetes and type 2 diabetes; systematic reviews comparing resistance and aerobic exercise exist in the literature and are worth reviewing directly for the specific effect size before quoting a precise number, since meta-analytic estimates vary by which trials are pooled.

Sleep extension has been studied in a randomized trial context for its effect on glucose and energy intake; the direction of effect (more sleep, better glucose regulation) is consistent with the broader sleep-metabolism literature, though the exact magnitude reported for any single trial should be checked against the original publication before it is used as a specific promise to a reader (PubMed).

Dietary approaches with reasonable evidence behind them include reducing refined carbohydrate at the evening meal (the meal before an overnight fast disproportionately affects the next morning's reading), time-restricted eating with a 12-to-14-hour overnight fast, and higher soluble fiber intake to blunt post-meal glucose excursions that can carry into the next fasting value (time-restricted eating trial, PubMed).

Pharmacologic options are used when lifestyle change alone is insufficient or when prediabetes/diabetes is already diagnosed. Metformin remains the most studied first-line agent for prediabetes and reduced diabetes incidence over long-term follow-up in the DPP outcomes study (PubMed). GLP-1/GIP receptor agonists (semaglutide, tirzepatide) are FDA-approved for type 2 diabetes and produce substantially larger fasting glucose reductions in trial populations with established diabetes; the SURPASS-1 trial studied tirzepatide in treatment-naive type 2 diabetes (PubMed). These agents are not indicated or approved for someone whose fasting glucose is simply in the high-normal range without a diabetes diagnosis, and using them off-label for that purpose is a decision that belongs with a prescribing clinician, not a self-directed choice based on a lab number.

Separately, a retrospective study on vitamin D supplementation and glycemic control in people living with diabetes exists in the literature; if a reader is specifically asking whether vitamin D affects their glucose, that paper is the closest evidence anchor available here, though it is an observational/retrospective design and should not be read as proof of a causal effect (PubMed).

Using fasting glucose alongside other markers

Fasting glucose is one data point in a 24-hour cycle. Pairing it with additional markers changes what a single number can tell you.

HbA1c reflects average glucose over roughly 90 days. The ADA defines normal as under 5.7%, prediabetes as 5.7% to 6.4%, and diabetes as 6.5% or higher (ADA Standards of Care). A fasting glucose in the high-80s or low-90s with an HbA1c already near 5.7% suggests post-meal glucose excursions may be contributing more than the fasting number alone reveals.

Fasting insulin is the marker functional-medicine clinicians most often add. Standard lab reference ranges for fasting insulin are wide; there is no single universally agreed "optimal" cutoff in guideline literature, though clinicians commonly look for insulin at the low end of the reference range as reassuring. A normal fasting glucose maintained alongside a high fasting insulin suggests the body may be compensating for insulin resistance rather than being free of it.

HOMA-IR combines fasting glucose and fasting insulin into a single index (glucose in mg/dL multiplied by insulin in µIU/mL, divided by 405), originally described by Matthews and colleagues in 1985 (PubMed). Lower values indicate greater insulin sensitivity; higher values are associated with insulin resistance and metabolic syndrome, though HOMA-IR is a research and clinical-adjunct tool rather than a diagnostic criterion in ADA or WHO guidelines.

Oral glucose tolerance test (OGTT) remains the reference standard for diagnosing impaired glucose tolerance: a 2-hour post-75g-glucose reading under 140 mg/dL is normal, 140 to 199 mg/dL is impaired glucose tolerance, and 200 mg/dL or higher, confirmed, is diabetes (ADA Standards of Care).

When and how often to test

The USPSTF recommends screening for prediabetes and type 2 diabetes in adults aged 35 to 70 who have overweight or obesity, a grade B recommendation (USPSTF). The ADA also recommends screening adults with a BMI of 25 or higher (23 or higher for Asian Americans) plus one additional risk factor, or any adult aged 45 and older regardless of BMI.

If you are tracking the effect of a lifestyle change or medication, rechecking fasting glucose alongside HbA1c and, if relevant, fasting insulin every few months gives more actionable information than a single value, since day-to-day fasting glucose can vary meaningfully with sleep, hydration, stress, and the time of the blood draw. Draw the sample in the morning after an 8-to-12-hour fast, water only; dehydration can concentrate plasma and shift the reading slightly higher.

What is established, what is plausible, and what is not established

Established: the ADA's three-tier classification (normal, prediabetes, diabetes) and its numeric cutoffs are the current diagnostic standard in U.S. clinical practice. The USPSTF screening recommendation for adults 35 to 70 with overweight or obesity is a current, graded guideline recommendation. Metformin and lifestyle intervention are proven, guideline-supported approaches to reducing progression from prediabetes to diabetes.

Plausible but not settled: that a narrower "optimal" range (roughly 75 to 89 mg/dL) meaningfully outperforms the standard reference range as a predictor of individual long-term outcomes. This idea is supported by cohort associations and professional consensus statements pointing to graded risk below the diagnostic cutoff, but it has not been adopted as a diagnostic standard by the ADA or WHO, and no randomized trial has tested "treating to 89 mg/dL" as an intervention target against usual care.

Not established: that any single fasting glucose value in the 90 to 99 mg/dL range, by itself, requires medication or aggressive intervention. The evidence supports requesting additional data (fasting insulin, HbA1c) in that zone rather than treating the number alone as actionable.

If your fasting glucose is in the gray zone, urgent care is not the right setting for follow-up; a routine visit with your primary care clinician to discuss fasting insulin and HbA1c is the appropriate next step. Urgent or emergency evaluation is appropriate for symptomatic hypoglycemia (confusion, loss of consciousness, seizure) or for very high glucose readings accompanied by symptoms of diabetic ketoacidosis (rapid breathing, fruity breath odor, vomiting, altered mental status).