What Do Blood Sugar Tests Measure? — Three Numbers, and the Lines Each Organization Drew

One Line for Blood Sugar on the Health Checkup Sheet

In the blood test section of a health checkup report sheet, there is a single blood sugar line. It is a number measured using blood drawn after fasting overnight. When a label like 「Caution」 is attached next to it, people usually recall a single sentence they heard somewhere. They say you need to be careful if it goes over 100.

When you go to the hospital, you sometimes hear one more number. This time, it is a number written as a percentage, with an explanation that it shows the past few months. Usually, no one tells you whether the two numbers are telling the same story, or which one you should trust if one is fine while the other is high.

This article takes the health checkup report sheet from South Korea as its starting point. South Korea's general health checkup includes a blood sugar test measured while fasting (PMID 35610963). The line of the figure that divides 「High」 comes from documents issued by the American Diabetes Association and international organizations, and whom to measure and when comes from recommendations in South Korea and the United States. The judgment phrases written on the Korean checkup sheet and the figure lines of the Korean society are not covered in this article.

The previous article, "What Is Diabetes? — The Letter That Never Comes, the Letter That Isn't Read", covered a disease where blood sugar stays high for a long time, and the task of reading that blood sugar as a number was passed on to this article. If we write down the answer in advance, it is like this. Blood sugar tests are broadly three, and the three measure different things. The line for 「High」 is an agreement that organizations drew after looking at the risk of complications, so it has changed and differs slightly from organization to organization.

Three Tests, Three Questions

Blood sugar is the amount of sugar dissolved in the blood. Glucose is sugar created as carbohydrates in rice, bread, and fruit are digested, in a form that cells can burn for fuel right away. The sugar measured by a blood sugar test is this glucose.

Diabetes is a condition where blood sugar stays higher than the normal range for a long time. To diagnose diabetes and distinguish the stage right before it, doctors use blood sugar after fasting, blood sugar after drinking sugar water, and traces of sugar left on proteins in the blood (PMID 42162460, PMID 25612821).

The first test is that line on the health checkup sheet. Fasting blood sugar is the blood sugar measured before breakfast after not eating anything all night. As we saw in the earlier article on blood sugar, "How Blood Sugar Stays Steady — A Letter to Store, a Letter to Release", this value is the result balanced between the glucose the liver released overnight and the glucose the body used. Fasting blood sugar snapshots a single moment in the body after a night has passed.

The second test purposely puts sugar into the body and observes the reaction. The oral glucose tolerance test is a test where, after fasting overnight, a person drinks water with 75 grams of glucose dissolved in it, and their blood sugar is measured again two hours later. This two-hour value is one of the criteria distinguishing diabetes and the stage right before it (PMID 25612821, PMID 32497744).

Insulin is a hormone released by the pancreas when blood sugar rises, telling the liver, muscles, and fat to "take up and store the glucose in the blood." Compared to a letter, the oral glucose tolerance test is a test to see how much the letter telling the body to "store" the sugar that came in all at once cleared it away over two hours, and it looks at a different breakdown than fasting blood sugar. The condition where fasting blood sugar is high and the condition where blood sugar does not drop well after drinking sugar can have different causes inside the body (PMID 40676018).

The third test reads traces left on proteins in the blood. Hemoglobin is a protein filling up red blood cells, which the blood uses to carry oxygen. Glucose in the blood naturally sticks to this hemoglobin without the help of enzymes (PMID 21975967). Glycated hemoglobin is the proportion of hemoglobin inside red blood cells that glucose has stuck to. That is why it is written as a percentage on the test report.

Glycated hemoglobin is a number that stands in for the average blood sugar over the past 8 to 12 weeks, or two to three months (PMID 21975967). Separately from this, red blood cells live for an average of about 120 days, and the sticking of glucose to hemoglobin happens throughout that lifespan (PMID 16974013). Compared to a letter, a red blood cell is a mail carrier circulating through the blood for nearly four months, and the more days blood sugar stays high, the more sugar gets stained on its bag. This number is a mark stained while the mail carrier is working, so if something happens to the mail carrier, the mark changes too.

Fasting blood sugar looks at a single moment after a night, the oral glucose tolerance test looks at the ability to clear away incoming sugar, and glycated hemoglobin looks at the average of the past two to three months.

The Same Person, Different Verdicts

When the same person is measured by two tests, the answers often disagree. When about 35,000 Koreans with no history of a diabetes diagnosis were examined together using fasting blood glucose and glycated hemoglobin, among the 1,491 newly diagnosed with diabetes, fewer than half had both tests cross the line for diabetes (44.9%). The rest had only fasting blood glucose, or only glycated hemoglobin, cross the line (PMID 20970868).

The people who disagreed also differed in their characteristics. Those caught only by glycated hemoglobin were older, more likely to be female, and had lower hemoglobin concentrations. Those caught only by fasting blood glucose had higher blood pressure and fasting insulin levels (PMID 20970868). The authors of this study wrote that further research is needed to determine which of the two criteria better predicts the risk of vascular complications that develop over time (PMID 20970868).

Expanding the view to the world, the picture is similar. In an analysis of 117 population surveys, among the diabetes cases newly discovered during the surveys, 29% to 39% of people had both numbers high, depending on the region, and the rest had only one side high (PMID 37946056). In most low- and middle-income regions, people with high glycated hemoglobin alone were more common. In such regions, using only fasting blood glucose can delay diagnosis and lead to diabetes being undercounted in reality (PMID 37946056).

When an oral glucose tolerance test is set as the standard, the nature of glycated hemoglobin is revealed. In an analysis combining multiple studies, when previously undiagnosed people were screened using an oral glucose tolerance test, about half of the people confirmed to have diabetes by the oral glucose tolerance test were found by the glycated hemoglobin line of 6.5%. Instead, almost everyone who was not diabetic by the oral glucose tolerance test was also below that line for glycated hemoglobin (PMID 33216783). The authors concluded that at the current cutoff, glycated hemoglobin misses many people and misdiagnoses few, and wrote that if the goal is to screen asymptomatic people in advance, lowering the cutoffs for both glycated hemoglobin and fasting blood glucose together could be considered (PMID 33216783).

Fasting blood glucose and glycated hemoglobin pick out different people among the same populations as having diabetes (PMID 37946056). Which one better predicts complications has different answers depending on the evidence. The earlier Korean study wrote in 2011 that further research is needed to decide this answer between fasting blood glucose and glycated hemoglobin (PMID 20970868). In a 2024 position statement, the International Diabetes Federation argues that blood glucose measured one hour after drinking a sugar drink predicts diabetes, complications, and death better than fasting blood glucose, glycated hemoglobin, and the two-hour value (PMID 38458916).

The three tests frequently yield conflicting verdicts for the same person, and which one better predicts complications has different answers depending on the evidence.

When the Lifespan of the Mail Carrier Changes

How many days a red blood cell lives, and how much glucose enters it, vary slightly from person to person whether they have diabetes or not (PMID 20144305). Because glycated hemoglobin is a number that leans on red blood cells, when properties like the lifespan of red blood cells change, it deviates from the actual blood sugar (PMID 37306444).

The deviation sometimes arises because of disease. When kidney function is severely impaired, in anemia where red blood cells break down quickly, with variants that have a different hemoglobin structure, and in cases like iron deficiency, glycated hemoglobin becomes hard to trust (PMID 42779420). Conditions like anemia, pregnancy, and liver disease also shake this number (PMID 33389957). One review covering a single hemoglobin variant reported that the variant likely reduces the lifespan of red blood cells and lowers the accuracy of glycated hemoglobin, and recommended using other tests like fasting blood sugar together in such people (PMID 32923278).

A more common case is individual differences unrelated to disease. Just the fact that red blood cells differ slightly within the normal range changes the relationship between glycated hemoglobin and average blood sugar. That difference can cause a person's blood sugar exposure to be estimated as higher or lower than it actually is, to an extent that leads to the risk of excessive treatment or insufficient treatment (PMID 37306444).

The mismatch sometimes changes diagnosis. In a case report of two people, one person's glycated hemoglobin crossed the line for diabetes, but an oral glucose tolerance test done again was normal. Another person, who was living with HIV, had a normal glycated hemoglobin for several years, and around the time it just began to rise slightly, diabetes was confirmed on an oral glucose tolerance test (PMID 32525987). The authors who reported the cases wrote that when the numbers mismatch in a person at high risk, doing an oral glucose tolerance test is strictly necessary even if it has limitations (PMID 32525987).

Because of these weaknesses, other glycated protein tests that are not affected by the lifespan of red blood cells are also used (PMID 28393586). Because glycated hemoglobin is a number that borrows red blood cells to read blood sugar, when the red blood cells change, the number also deviates from blood sugar.

Who Drew the Line?

The fasting blood sugar line of 126 mg/dL that divides diabetes is a value introduced as a "new criterion" by a 1998 review article. That new criterion lowered the fasting blood sugar line to this value and organized the diagnostic conditions so that fasting blood sugar must cross the line twice (PMID 9803200). The review article introducing that criterion noted that this change was a method to diagnose people at risk of complications due to high blood sugar more easily and reliably (PMID 9803200).

A complication is a disease that subsequently develops in another part of the body because of a certain disease. Long-term high blood sugar is related to complications in small and large blood vessels, as well as nerves (PMID 25612821). The blood sugar line is a value determined based on this risk. Because it is a value decided by humans, it can be moved.

According to a review article organizing the diagnostic criteria currently in use, diabetes is diagnosed by at least one of the following: a fasting blood sugar of 126 mg/dL or higher, a blood sugar two hours into an oral glucose tolerance test of 200 mg/dL or higher, or a glycated hemoglobin of 6.5% or higher (PMID 25612821). There are also sections with names below that.

Prediabetes is a section where blood sugar is higher than normal but has not reached the diabetes line. This section is considered a preceding stage on the road to type 2 diabetes (PMID 40676018). An American Diabetes Association position statement defines this section using three numbers—fasting blood sugar, an oral glucose tolerance test, and glycated hemoglobin—and sets 5.7 to 6.4% as this section for glycated hemoglobin (PMID 25612821).

Impaired fasting glucose is a state where fasting blood sugar is between the normal line and the diabetes line. Under the American Diabetes Association's line, it is 100 to 125 mg/dL (PMID 25612821). It is a number like 「100」 that people who receive a health checkup report hear.

Impaired glucose tolerance is a state where blood sugar two hours into an oral glucose tolerance test is between the normal line and the diabetes line. Under the same criteria, it is 140 to 199 mg/dL. And there can be discrepancies among the three tests (PMID 25612821).

The lines have been moved. The fasting blood sugar line dividing prediabetes has also been changed to lower values (PMID 27457639). Whether the fasting blood sugar line that determines impaired fasting glucose and dividing prediabetes by glycated hemoglobin are appropriate is still under debate (PMID 25612821). Various definitions and lines have emerged for prediabetes, and there is no single one agreed upon by everyone (PMID 21281840). The American Diabetes Association criteria and the World Health Organization criteria pick out different people at risk of progressing to diabetes (PMID 32497744).

There are also proposals to add new lines. The 2024 position statement from the International Diabetes Federation seen earlier proposed adding blood sugar one hour after drinking sugar to the diagnostic criteria (PMID 38458916). Another review article covering the one-hour value considers that value promising, but believes it needs further confirmation with data including cost analysis (PMID 32497744).

Diagnosis does not end with a single number. The 1998 criteria conditioned that fasting blood sugar must cross the line twice (PMID 9803200), and the International Diabetes Federation position statement also wrote that people whose one-hour value crossed the diabetes line should be retested to confirm the diagnosis (PMID 38458916).

Who to measure and when also differ by organization. A screening test is a test intended to find people who have a disease in advance among people without symptoms. In a 2022 position statement, the Korean Diabetes Association recommended diabetes screening for adults aged 35 and older even without risk factors, and for all adults aged 19 and older if they have even a single risk factor, and added abdominal obesity seen through waist circumference to the risk factors (PMID 36455530). The United States Preventive Services Task Force recommends screening for adults aged 35 to 70 who are overweight or obese (PMID 34427594).

The line of "high" is an agreement drawn by organizations looking at complication risk, has changed, and differs by organization.

What Prediabetes Means

34.5% of American adults meet the criteria for prediabetes (PMID 34427594). That is about one in three people. There are two converging lines of evidence regarding what this label given to so many people actually means.

Type 2 diabetes is a condition where insulin resistance is compounded by a decline in beta cell function, meaning the body cannot release enough insulin for its needs, which causes blood sugar to rise. In the terms of the previous primer, it is a diabetes where letters are not read well, and sending more cannot keep up. The malfunction where the body responds less to insulin and the malfunction in insulin secretion are already present in prediabetes (PMID 37328664).

One direction is risk. People with prediabetes have a high likelihood of developing type 2 diabetes, as well as a high risk of complications in blood vessels and various other systems (PMID 40676018). Whether this risk is identical across the different subtypes is viewed differently by various review articles. Some reviews indicate that the increased risk appears similar regardless of which criteria are used to define it (PMID 31766914), whereas other reviews suggest that impaired fasting glucose and impaired glucose tolerance may have different risks for causes, progression, and adverse outcomes (PMID 40676018). The risk of progressing to diabetes or cardiovascular disease also varies depending on other factors such as sex, weight, and race (PMID 31766914).

The other direction is that not everyone goes all the way down that path. Not everyone with prediabetes ends up with diabetes. However, the same paper adds that finding out the risk is nonetheless increased remains meaningful (PMID 37328664). Estimates of the risk of going from prediabetes to diabetes vary widely across studies, which may be because the methods used to define prediabetes differ (PMID 34427594).

Data from older adults demonstrate both directions together. In a Swedish study following 2,575 individuals aged 60 and older for up to 12 years, 918 had prediabetes defined by glycated hemoglobin. Among them, 22% returned to normal blood sugar, 13% developed diabetes, and 23% passed away (PMID 31165572). The authors concluded that the majority of older adults with prediabetes either stayed where they were or returned to normal, and only about a third developed diabetes or passed away (PMID 31165572).

The authors of a study following 2,027 older adults in the UK with a mean age of 70 concluded that while prediabetes was common, progression to diabetes was rare, and returning to normal or passing away occurred more frequently (PMID 35134845). The numbers from the same study show the other side as well. The rate at which people with prediabetes defined by glycated hemoglobin developed diabetes was about five times that of people who were normal at the start, and among people with prediabetes, slightly more developed diabetes than passed away (PMID 35134845).

There is also evidence that as people get older, the risk of going from prediabetes to diabetes decreases, and a diagnosis of type 2 diabetes in older adults does not lead to outcomes as adverse as those seen in younger groups (PMID 37452769). Because current definitions do not account for age, there is debate over how heavily to treat prediabetes in older adults (PMID 37452769).

There is also evidence broadening the path back. In an analysis combining 44 randomized trials involving 14,742 people with prediabetes, lifestyle interventions to lose weight brought 11 more people out of 100 back to normal blood sugar over a median of 24 months, and kept 8 fewer people from progressing to type 2 diabetes (PMID 39222689). However, other reviews note that the evidence for prediabetes management comes primarily from people who are overweight or obese and have impaired glucose tolerance, and it remains unclear whether there is a benefit for people with impaired fasting glucose or normal weight (PMID 40676018).

There is debate even about the label itself. In 2016, the American Diabetes Association journal Diabetes Care set up pro and con pages addressing the question "Is There a Problem with the Term Prediabetes?". The author on the opposing side argues that the term obscures the differences among the three distinct paths defined by fasting glucose, two-hour glucose, and glycated hemoglobin, making it difficult to interpret interventions and outcomes. They also point out that too many people meet these criteria, far exceeding the capacity of the healthcare system to address them individually, and that there is no evidence that interventions provide a benefit outside of the impaired glucose tolerance path (PMID 27457638).

The author on the supporting side was the editor-in-chief of the journal at the time. He believes that the lowered cutoff still identifies individuals at high risk, draws attention to states with elevated cardiovascular disease risk, and provides an opportunity to intervene earlier. He therefore argues that the current cutoff is appropriate and provides useful information for reducing the burden of diabetes in the future (PMID 27457639). An introductory article for these pages noted that there appeared to be no consensus on the definition set by the current cutoff (PMID 27457639).

Prediabetes is a range with an increased risk of progressing to diabetes, and not everyone goes all the way down that path. Two viewpoints still clash over whether this label is useful.

A Record That Links the Day

The previous article on blood sugar looked at the ups and downs of a day through records kept by healthy people wearing sensors on their bodies. Continuous glucose monitoring is a method where a sensor attached to the body measures blood sugar every few minutes to record it continuously throughout the day (PMID 33310176). The sensor measures glucose in the fluid filling the spaces between cells instead of blood, and gathers values at 1 to 5 minute intervals (PMID 36493795). Right now, it is mainly used to watch the blood sugar of people with diabetes (PMID 41276244).

This record sees what glycated hemoglobin misses. Glycated hemoglobin cannot show short-term ups and downs (PMID 35094087), and it does not tell you how much blood sugar fluctuates or the risk of low blood sugar (PMID 34126129). Low blood sugar is a state where blood sugar has fallen below the normal range.

Time in range is the proportion of the day that blood sugar stays within a set target range. International consensus sets that range at 70 to 180 mg/dL for most people with diabetes (PMID 34126129). This proportion moves in strong alignment with glycated hemoglobin, and is also connected to the risk of complications that arise after a long time (PMID 34126129).

However, most studies showing that connection have looked at a single sliced point in time or looked back at past records, and data directly showing whether making this number better actually reduces complications are still sparse (PMID 41208627). In South Korea, a section covering continuous glucose monitoring was newly added to the Korean Diabetes Association's 2021 clinical practice guidelines (PMID 34352984).

Continuous glucose monitoring is a tool to see the flow of the day, and it answers a different question from the three numbers on a health checkup sheet.

Holding a Single Health Checkup Report

A person has received their health checkup results. A number slightly above 100 is written in the fasting blood sugar box. We read this single line with what we have seen so far.

This number is a single moment of the body after passing a night, that morning. This single measured value is within the range (100–125 mg/dL) set by the American Diabetes Association for impaired fasting glucose (PMID 25612821). The lower line of that range has been changed to lower values over time (PMID 27457639), and the criteria of other organizations pick out slightly different people (PMID 32497744).

When hemoglobin A1c is measured together at the hospital, that number shows the average of the past two to three months by borrowing red blood cells (PMID 21975967). The two numbers can point in the same direction or they can cross. In Korean data, more than half of people newly diagnosed with diabetes crossed the line on only one test (PMID 20970868). If this person has anemia, the hemoglobin A1c side can waver (PMID 33389957).

When the numbers cross, medical staff may measure again or add tests that ask different questions, such as the oral glucose tolerance test (PMID 32525987). Reference documents that confirm a diagnosis include a process of remeasuring to verify (PMID 9803200, PMID 38458916).

If the result is prediabetes, that name means it is a section where the risk of progressing to diabetes has increased (PMID 40676018), and it also means that not everyone goes all the way down that path. Even so, finding out about the increased risk is still meaningful (PMID 37328664).

A single line on a health checkup report is a single moment of that morning, and a judgment is made by placing different numbers and remeasured numbers together.

Where to Look Next

This article builds on the normal operation drawn in "How Blood Sugar Stays Steady — A Letter to Store, a Letter to Release" and the breakdown drawn in "What Is Diabetes? — The Letter That Never Comes, the Letter That Isn't Read". How to measure how well a test finds a disease and how often it points in the wrong direction is covered in "What a Test Actually Tells You — An Introduction to Sensitivity, Specificity, and Pretest Probability". The other lines on the health checkup sheet are read by "What Cholesterol Is and Why We Measure It — Four Lines on a Blood Test, and the Packages They Name" and "What Blood Pressure Measures — Two Numbers, and an Answer That Changes With Where You Measure". An article covering the relationship between weight and diabetes has not yet been released.

Some questions still have divided answers. One is which test better predicts complications that develop a long time later. A 2011 Korean study noted that more research is needed between fasting blood sugar and glycated hemoglobin (PMID 20970868). The International Diabetes Federation argues that blood sugar one hour later is better (PMID 38458916), while another review paper considers that one-hour value as something that needs further confirmation (PMID 32497744).

Source Papers

• "Current Status of the National Health Screening Programs in South Korea." (2022) — PMID 35610963 — Includes tests such as fasting blood glucose in the general health screening of the Korean National Health Screening Program, and presents tasks for improvement.
• "[Diabetes mellitus-definition, classification, diagnosis, screening and prevention (Update 2026)]." (2026) — PMID 42162460 — Summarizes diagnostic criteria using fasting and oral glucose tolerance test blood glucose and HbA1c, and screens for risk groups.
• "Definitions (and Current Controversies) of Diabetes and Prediabetes." (2016) — PMID 25612821 — Diagnostic criteria for diabetes (fasting ≥126, 2-hour ≥200, HbA1c ≥6.5%, etc.), ADA prediabetes categories (fasting 100–125 = impaired fasting glucose, 2-hour 140–199 = impaired glucose tolerance, HbA1c 5.7–6.4%), discrepancies between tests, and the cutoff for impaired fasting glucose and the appropriateness of HbA1c remain debated.
• "Review of methods for detecting glycemic disorders." (2020) — PMID 32497744 — Determines impaired fasting glucose and impaired glucose tolerance using a 75g oral glucose tolerance test, ADA and WHO criteria select different populations with no international consensus, 1-hour blood glucose is promising but needs confirmation with additional data (including cost analysis).
• "Prediabetes." (2025) — PMID 40676018 — Prediabetes is the stage preceding type 2, increasing the risk of type 2 and vascular and non-vascular complications, impaired fasting glucose and impaired glucose tolerance may differ in pathophysiology and natural history, and evidence for intervention largely comes from overweight and obese impaired glucose tolerance while it is unclear in impaired fasting glucose and normal weight.
• "Genetic determinants of variability in glycated hemoglobin (HbA(1c)) in humans: review of recent progress and prospects for use in diabetes care." (2011) — PMID 21975967 — HbA1c is the ratio of glucose non-enzymatically bound to hemoglobin, serving as a surrogate indicator of average blood glucose over the past 8 to 12 weeks (average red blood cell lifespan).
• "Glycated haemoglobin (HbA1c) monitoring." (2006) — PMID 16974013 — Glycation occurs throughout the 120-day average lifespan of red blood cells, and in situations that shorten their lifespan, HbA1c does not fully reflect current control.
• "Discordance between fasting glucose-based and hemoglobin A1c-based diagnosis of diabetes mellitus in Koreans." (2011) — PMID 20970868 — Among 35,624 non-diabetic Koreans and 1,491 newly diagnosed, fasting glucose alone accounted for 31.6%, HbA1c alone 23.5%, and both 44.9%; the HbA1c group was older, female, and had lower hemoglobin, and further research is needed on which criterion better predicts vascular complications.
• "Global variation in diabetes diagnosis and prevalence based on fasting glucose and hemoglobin A1c." (2023) — PMID 37946056 — Across 117 population surveys, 29–39% of diabetes found through screening showed high levels on both tests, and isolated HbA1c elevation was more common in low- and middle-income regions, meaning using only fasting glucose delays diagnosis and underestimates prevalence.
• "Diagnostic accuracy of tests for type 2 diabetes and prediabetes: A systematic review and meta-analysis." (2020) — PMID 33216783 — 37 studies based on oral glucose tolerance tests, sensitivity of HbA1c 6.5% was 50% and specificity 97.3%, concluding that lowering the cutoff may be considered if the goal is screening.
• "Biologic variability in plasma glucose, hemoglobin A1c, and advanced glycation end products associated with diabetes complications." (2009) — PMID 20144305 — Normal individual differences exist in red blood cell glucose distribution and red blood cell lifespan regardless of diabetes status.
• "Personalized Glycated Hemoglobin in Diabetes Management: Closing the Gap with Glucose Management Indicator." (2023) — PMID 37306444 — Changes in properties such as red blood cell lifespan distort HbA1c, and more commonly, individual differences within the normal range overestimate or underestimate blood glucose exposure, risking over- or undertreatment.
• "Glycated albumin: An old yet new biomarker in people with diabetes and the potential for novel applications." (2026) — PMID 42779420 — HbA1c becomes unreliable when red blood cell lifespan or glycation is disrupted, such as in renal failure, hemolytic anemia, hemoglobin variants, or iron deficiency.
• "Time in Range from Continuous Glucose Monitoring: A Novel Metric for Glycemic Control." (2020) — PMID 33389957 — HbA1c is easily interfered with by anemia, pregnancy, liver disease, etc., and continuous glucose monitoring core indicators include time in target range.
• "The Sickle Effect: The Silent Titan Affecting Glycated Hemoglobin Reliability." (2020) — PMID 32923278 — The possibility that sickle cell trait affects HbA1c accuracy via decreased red blood cell lifespan, recommending parallel use with other diagnostic tools.
• "Pitfalls of HbA1c in the Diagnosis of Diabetes." (2020) — PMID 32525987 — 2 case reports (risk of overdiagnosis and underdiagnosis with HbA1c), noting that oral glucose tolerance tests are essential when values are discordant in high-risk groups despite limitations.
• "Evaluation of glycated albumin (GA) and GA/HbA1c ratio for diagnosis of diabetes and glycemic control: A comprehensive review." (2017) — PMID 28393586 — HbA1c is not suitable for all patients, and glycated albumin, which is unaffected by red blood cell lifespan, is considered an alternative marker.
• "Diagnosis and classification of diabetes mellitus: new criteria." (1998) — PMID 9803200 — Simplified diagnostic criteria to two abnormal fasting blood glucose tests and lowered the fasting blood glucose cutoff to 126 mg/dL, diagnosing individuals at risk of complications more easily and reliably.
• ""Prediabetes": Are There Problems With This Label? No, We Need Heightened Awareness of This Condition!" (2016) — PMID 27457639 — The pro side in a Diabetes Care pro-con debate, noting the fasting blood glucose cutoff has been lowered, the lower cutoff still finds high-risk groups and provides opportunities for early intervention, though a consensus on definition seems lacking.
• ""Prediabetes": Are There Problems With This Label? Yes, the Label Creates Further Problems!" (2016) — PMID 27457638 — The con side of the same debate, arguing that the label obscures differences among the three subcategories, burdens the healthcare system beyond capacity, and lacks evidence of intervention benefits outside of impaired glucose tolerance.
• "Treatment recommendations for prediabetes." (2011) — PMID 21281840 — There are multiple definitions and diagnostic criteria for prediabetes with no universal consensus.
• "International Diabetes Federation Position Statement on the 1-hour post-load plasma glucose for the diagnosis of intermediate hyperglycaemia and type 2 diabetes." (2024) — PMID 38458916 — Reports that 1-hour blood glucose is superior to fasting, HbA1c, and 2-hour values in predicting diabetes, complications, and mortality, proposing its addition to diagnostic criteria, with values exceeding the diabetes threshold confirmed by retesting.
• "Screening for Prediabetes and Diabetes in Korean Nonpregnant Adults: A Position Statement of the Korean Diabetes Association, 2022." (2022) — PMID 36455530 — Recommends screening for those aged 35 and older regardless of risk factors, and 19 and older if risk factors are present, adding waist circumference-based abdominal obesity to risk factors.
• "Screening for Prediabetes and Type 2 Diabetes: US Preventive Services Task Force Recommendation Statement." (2021) — PMID 34427594 — 34.5% of US adults meet prediabetes criteria, progression risk estimates vary widely due to definition differences, and recommends screening for overweight and obese adults aged 35–70 (B).
• "[Prediabetes as a therapeutic challenge in internal medicine]." (2023) — PMID 37328664 — Insulin resistance and secretory dysfunction are already present in the prediabetic stage, and not everyone with prediabetes inevitably progresses to diabetes.
• "Risk and management of pre-diabetes." (2019) — PMID 31766914 — Prediabetes increases the risk of type 2 and cardiovascular disease, and risks appear similar regardless of definition criteria but vary by sex, body mass index, and race.
• "Natural history of prediabetes in older adults from a population-based longitudinal study." (2019) — PMID 31165572 — 2,575 Swedes aged 60 and older, among 918 with HbA1c prediabetes, over 12 years 22% reverted to normal, 13% developed diabetes, and 23% died, with most remaining unchanged or returning to normal.
• "Risk of progression to diabetes and mortality in older people with prediabetes: The English longitudinal study on ageing." (2022) — PMID 35134845 — 2,027 individuals with an average age of 70.6 years, where prediabetes is common but progression to diabetes is rare, and return to normal or death is more frequent.
• "The evolving continuum of dysglycaemia: Non-diabetic hyperglycaemia in older adults." (2023) — PMID 37452769 — Progression risk decreases with age, type 2 diagnosis in the elderly is not associated with adverse outcomes as severe as in younger groups, and definitions are debated for not adjusting for age.
• "Efficacy of lifestyle weight loss interventions on regression to normoglycemia and progression to type 2 diabetes in individuals with prediabetes: a systematic review and pairwise and dose-response meta-analyses." (2024) — PMID 39222689 — 44 RCTs with 14,742 participants, median 24 months, per 100 people 11 more returned to normal and 8 fewer progressed to type 2 (moderate certainty of evidence).
• "Glucose measurement in body fluids: A ready reckoner for clinicians." (2021) — PMID 33310176 — Continuous glucose monitoring measures interstitial fluid glucose and is reported as a standard indicator.
• "Continuous glucose monitoring and metrics for clinical trials: an international consensus statement." (2023) — PMID 36493795 — Continuous glucose monitoring devices collect data at 1 to 5 minute intervals, and HbA1c is a measure of average blood glucose over the past 3 months.
• "Continuous glucose monitoring in high-risk individuals." (2026) — PMID 41276244 — Continuous glucose monitoring is a primary tool for blood glucose monitoring in diabetic patients, and accuracy is affected by various factors.
• "Complications of Diabetes and Metrics of Glycemic Management Derived From Continuous Glucose Monitoring." (2022) — PMID 35094087 — HbA1c fails to show short-term fluctuations, and higher time in the target range is associated with a lower risk of various complications.
• "Time-in-range for monitoring glucose control: Is it time for a change?" (2021) — PMID 34126129 — HbA1c lacks information on glycemic variability and hypoglycemia risk, defines time in target range, with most international targets at 70–180 mg/dL >70%, showing a strong correlation with HbA1c and association with complication risk.
• "Continuous glucose monitoring and microvascular complications in diabetes: Bridging glycemic metrics with clinical outcomes." (2026) — PMID 41208627 — Low time in target range is associated with microvascular complications, but mostly from cross-sectional and retrospective studies, and direct evidence that improving metrics leads to reduced complications is limited.
• "2021 Clinical Practice Guidelines for Diabetes Mellitus of the Korean Diabetes Association." (2021) — PMID 34352984 — The 7th edition guidelines of the Korean Diabetes Association, establishing a new section on continuous glucose monitoring.