Screening for type 2 diabetes mellitus and prediabetes

INTRODUCTION — 

Diabetes is a major cause of early illness and death worldwide. Globally, diabetes affects more than 500 million people, and its prevalence continues to rise. (See "Type 2 diabetes mellitus: Prevalence and risk factors", section on 'Prevalence'.)

This topic discusses the evidence and recommendations related to screening asymptomatic adults for type 2 diabetes mellitus and prediabetes. The evaluation of individuals with signs and symptoms of hyperglycemia (eg, polydipsia, polyuria, blurred vision, paresthesias, or unexplained weight loss) and screening for type 1 diabetes mellitus and gestational diabetes are discussed separately. The prevalence, risk factors, and prevention of type 2 diabetes are also discussed separately:

●(See "Gestational diabetes mellitus: Screening, diagnosis, and prevention".)

●(See "Clinical presentation, diagnosis, and initial evaluation of diabetes mellitus in adults".)

●(See "Type 2 diabetes mellitus: Prevalence and risk factors".)

●(See "Prevention of type 2 diabetes mellitus".)

●(See "Type 1 diabetes mellitus: Disease prediction and screening".)

BENEFITS AND HARMS OF SCREENING — 

The decision about whether to screen for a specific condition depends on whether the benefits of screening and treatment outweigh potential harms [1-3]. Several key factors inform decisions about screening, including screening for diabetes. These include characteristics of the disease (ie, is it common and/or deadly), available screening tests (ie, are tests accurate, acceptable, feasible, safe, and cost effective), and its treatment (ie, are safe and effective treatments available, and does early treatment improve outcomes). These components are summarized below and described more generally elsewhere. (See "Evidence-based approach to prevention".)

Burden of disease — Diabetes is common, and type 2 diabetes mellitus constitutes over 90 percent of cases worldwide [4]. In 2021, approximately 15 percent of all adults in the United States had diabetes, and 3.4 percent were undiagnosed [5]. The prevalence of diabetes in the United States and worldwide continues to rise, with the highest rates of increase occurring in low- and middle-income countries [4-6]. (See "Type 2 diabetes mellitus: Prevalence and risk factors".)

Clinical outcomes of screening — We do not have direct evidence from clinical trials to demonstrate that diabetes screening reduces mortality or cardiovascular outcomes. By contrast, the results of randomized trials of early intervention in patients with newly diagnosed diabetes provide indirect evidence that diabetes screening might improve clinical outcomes. (See 'Earlier treatment initiation' below.)

●No mortality benefit with diabetes screening

•Trials of diabetes screening – Trials of diabetes screening interventions have not demonstrated reductions in clinical outcomes. In a 2021 systematic review that included two large trials of diabetes screening interventions (n = 25,120), screening did not reduce all-cause, cardiovascular, or diabetes-related mortality at 10-year follow-up [7]. As an example, in one of the trials, rates of all-cause mortality at a mean follow-up of 9.6 years were similar in the screened and control groups (hazard ratio 1.06, 95% CI 0.90-1.25) [8]. Results from both trials were imprecise and included the possibility of mortality benefit and harm [7].

•Intervention trials in those with screen-detected prediabetes or diabetes – Similarly, trials of intensive interventions in individuals with screen-detected diabetes or prediabetes have not found significant differences in mortality or cardiovascular disease (CVD) events at up to 10 years of follow-up. In a well-designed, randomized trial of 3234 participants with prediabetes who were randomized to intensive lifestyle changes, metformin, or placebo, no significant differences in all-cause mortality or cardiovascular events were observed at 10-year follow-up. A second trial in China reported differences in all-cause and CVD mortality at 23 years, although the strength of evidence was limited by a moderate risk of bias and concerns about generalizability to other populations [7,9,10].

●Potential explanations – The observed lack of effect of diabetes screening on mortality is likely due to the insufficient duration of most trials; the ethical and practical challenges of conducting such trials; the low rates of diagnosed diabetes in some screening trials; and improvements in the management of other cardiovascular risk factors, such as hyperlipidemia and hypertension. Many individuals identified via population-based screening are likely early in the progression from prediabetes to diabetes. Death rates are low in such individuals, and demonstrating a mortality benefit with screening would require large numbers of participants and lengths of follow-up that exceed 10 years. Most randomized trials in those with prediabetes included few CVD events or deaths, and follow-up lengths were less than six years [7].

Potential benefits

Improved detection — Screening for diabetes effectively detects diabetes in asymptomatic individuals (table 1) [11]. Diabetes screening can also identify individuals with prediabetes (table 2), in whom early intervention may prevent or delay the onset of diabetes or increase the intensity of other CVD risk factor reduction efforts. (See 'Delayed diabetes onset' below.)

●Diabetes – Hyperglycemia has a cumulative, exponential impact on the risk of diabetes complications. The natural history of diabetes consists of a long asymptomatic period characterized by insulin resistance, impaired insulin secretion, and hyperglycemia [12-14]. Even in asymptomatic individuals, hyperglycemia increases the risk of microvascular and macrovascular complications. Approximately 10 to 30 percent of individuals exhibit micro- and/or macrovascular complications at the time of diabetes diagnosis [15-19], suggesting that earlier diagnosis and intervention could provide further benefit for a significant proportion of people.

The prevalence of undiagnosed diabetes varies depending on the population studied and the diagnostic criteria that are used. In the United States, estimated percentages of undiagnosed diabetes range from approximately 10 to 20 percent of total diabetes cases [5,20]. The prevalence of undiagnosed diabetes is higher in individuals who lack access to health care or health insurance, have overweight or obesity, or are older or Asian American [20]. Globally, some studies report that up to 45 percent of individuals are unaware that they have diabetes [4,6].

In the United States, the proportion of undiagnosed diabetes cases has declined since the 1990s, which is likely due, in part, to increased diabetes screening [15,20]. In a study of United States adults that used data from the National Health and Nutrition Examination Survey (NHANES), the proportion of cases of diabetes that were undiagnosed fell from 33 to 18 percent between 1988 to 1994 and 2017 to 2020 [20].

●Prediabetes – Hyperglycemia exists along a continuum. The term prediabetes describes states of intermediate hyperglycemia that likely reflect abnormal carbohydrate metabolism but do not meet diagnostic thresholds for diabetes. In prediabetes, the risk of progression to diabetes is moderately increased and the risk of early microvascular complications is mildly increased. The prevalence of prediabetes depends on which diagnostic thresholds are used. As an example, in an analysis of data from the NHANES, the prevalence of prediabetes in the United States ranged from 4.3 to 44 percent, depending on how prediabetes was defined [21].

Delayed diabetes onset — In persons with prediabetes, both intensive lifestyle interventions and pharmacotherapy with metformin or other agents can prevent or delay the onset of diabetes [21-27]. (See "Prevention of type 2 diabetes mellitus".)

Detection of prediabetes may improve other cardiovascular risk factors. In adults with prediabetes identified via screening, intensive lifestyle interventions and pharmacotherapy can improve intermediate health outcomes, such as blood pressure, weight, and body mass index [7].

Earlier treatment initiation — Treatment at the time of diabetes diagnosis improves clinical outcomes, including reducing microvascular complications, cardiovascular events, and all-cause and diabetes mortality.

●Mortality – In a systematic review of five randomized trials of participants with clinically diagnosed type 2 diabetes, intensive glucose control reduced the risk of all-cause (relative risk [RR] 0.87, 95% CI 0.79-0.96) and diabetes-related mortality (RR 0.83, 95% CI 0.73-0.96) at 20-year follow-up, compared with conventional treatment [7]. Intensive glucose control also reduced all-cause mortality at 10-year follow-up in the subset of participants with overweight or obesity (RR 0.64, 95% CI 0.45-0.91) [7].

●Microvascular and macrovascular complications – In individuals with established type 2 diabetes, early, intensive treatment of hyperglycemia reduces the incidence and progression of microvascular disease (eg, retinopathy, nephropathy, and neuropathy) [28]. Intensive glycemic management and risk factor modification also reduce the risk of cardiovascular events in individuals with type 2 diabetes [29-31].

The effect of glycemia on diabetes-related vascular complications is reviewed in detail separately, as is the approach to modifying risk factors for macrovascular disease in people with type 2 diabetes. (See "Glycemic management and vascular complications in type 2 diabetes mellitus" and "Overview of general medical care in nonpregnant adults with diabetes mellitus", section on 'Reducing the risk of macrovascular disease'.)

Potential harms — Potential harms of screening include those resulting from screening itself and those caused by early treatment of diabetes and prediabetes.

●Limited harms of screening – Limited evidence suggests minimal harms from diabetes screening. In a systematic review of three trials, results from the two larger trials did not reveal higher rates of anxiety between screened and unscreened groups, and no trial found significant differences in measures of depression, worry, or self-reported health [7]. Low response rates limit the certainty of these conclusions.

The main potential harms include:

•Overdiagnosis – Potential overdiagnosis can stem from false-positive test results and the uncertain prognostic implications of a diagnosis of prediabetes, especially in older adults.

-False-positive tests – False-positive results can occur, particularly when one-time screening is not confirmed by subsequent testing. (See 'Test characteristics' below and 'Interpretation' below.)

Substantial proportions of individuals who meet diagnostic criteria for prediabetes "regress" to normoglycemia on follow-up testing. As an example, in a meta-analysis of 47 prospective cohort studies, 17 to 42 percent of participants who met the criteria for prediabetes at baseline regressed to normoglycemia after up to 11 years of follow-up [32]. This was similar to the proportion of individuals with a baseline glycated hemoglobin (A1C) of 5.7 to 6.4 percent who developed type 2 diabetes (14 to 31 percent) within the same time frame. This is one reason to perform confirmatory testing to clearly establish the diagnosis of diabetes. (See 'Test interpretation and follow-up' below.)

-Prognosis of prediabetes in older adults – The prognostic implications of prediabetes are unclear in older (>70 years of age) individuals. Most older patients who meet the criteria for prediabetes do not progress to diabetes, and rates of progression are not demonstrably higher among those with prediabetes than those with normoglycemia. In a prospective cohort study of 3412 older adults (ages 71 to 90 years), among those with a baseline A1C of 5.7 to 6.4 percent, 9 percent progressed to diabetes, 13 percent regressed to normoglycemia, and 19 percent died after a mean follow-up of 6.5 years [33]. Participants with baseline A1Cs of <5.7 percent had similar outcomes. A cohort study of older Swedish adults (age ≥60 years) reported similar results [34].

•Other potential harms – Other potential harms of screening include false reassurance, stigma, the labeling effects of diagnosis, and overtreatment [35,36].

Normal screening test results could falsely reassure individuals who have an elevated lifetime risk of developing diabetes due to genetic factors, sedentary lifestyles, or overweight and obesity. Given that screening tests for diabetes are only moderately sensitive, significant proportions of those screened will have false-negative test results [37]. (See 'Test characteristics' below.)

Stigma, labeling effects, and other unintended consequences of screening interventions are discussed separately. (See "Evidence-based approach to prevention", section on 'Unintended consequences of screening'.)

●Harms of treatment – Although early treatment of diabetes or prediabetes does not appear to cause significant harms [7], the burdens and risks of adverse effects of early treatment may outweigh the benefits in some older adults, especially those with frailty or other comorbidities that limit life expectancy or who are at low risk to progress to diabetes or develop diabetes complications [38,39]. Older adults may also be at higher risk of adverse effects of glucose-lowering treatments, particularly hypoglycemia.

Therapeutic intervention trials in adults with screen-detected or newly diagnosed type 2 diabetes have not shown significant differences in rates of serious hypoglycemia, serious adverse events, or mortality, compared with usual care [7]. Low event rates for most adverse outcomes limit the certainty of these findings.

Among adults with prediabetes, a systematic review of 13 pharmacologic and 8 lifestyle interventions found similar rates of hypoglycemia (pharmacologic interventions) and musculoskeletal injury (lifestyle interventions), compared with usual care [7]. Studies of medication-related side effects reported mixed results, with six trials reporting higher rates of study withdrawal due to medication-related side effects in the intervention groups and six reporting comparable withdrawal rates.

Cost-effectiveness of screening — Cost-utility analyses suggest that diabetes screening is cost effective. In a cost-utility model, screening and standard diabetes treatment at diagnosis reduced the incidence of myocardial infarction and microvascular complications and increased quality-adjusted life years (QALYs) added over 50 years, compared with no screening [40]. The most cost-effective strategies performed screening every three to five years, starting between the ages of 30 and 45 years. The reliability of these results was limited by model assumptions of perfect performance and compliance with screening. A second cost-utility analysis of diabetes screening and intervention (lifestyle or pharmacologic) reported similar results [41].

Screening targeted to high-risk individuals may be more cost effective than universal screening [42]. In one model, targeted screening of those with hypertension at ages 55 to 75 years was cost effective compared with no screening (estimated cost per QALY USD $34,375 in 1997) [43]. This falls within the generally accepted threshold for a screening intervention (ie, USD $50,000 to 100,000). By contrast, a model utilizing universal screening was not cost effective. (See "A short primer on cost-effectiveness analysis", section on 'Interpretation'.)

WHOM TO SCREEN

Targeted approach to screening — We suggest screening for type 2 diabetes in adults ages 35 to 70 years who are at increased risk of developing diabetes and/or have cardiovascular risk factors. This includes individuals with any of the following (table 3):

●Overweight or obesity

●History of gestational diabetes

●Polycystic ovary syndrome

●Metabolic syndrome

●Human immunodeficiency virus (HIV) disease

●First-degree relative with type 2 diabetes

●Hypertension

●Dyslipidemia

●Established cardiovascular disease (CVD)

●Chronic use of medications that can cause diabetes (eg, second-generation antipsychotics, chronic glucocorticoid treatment, immune checkpoint inhibitors, HIV antiretrovirals)

Our approach focuses on those individuals in whom screening and early intervention are more likely to reduce clinically important outcomes, such as cardiovascular events and mortality. It also balances the benefits of early diabetes treatment with the potential harms of overdiagnosis and lack of direct evidence that diabetes screening reduces mortality. (See 'Benefits and harms of screening' above.)

Since individuals with cardiovascular risk factors have a higher absolute risk of cardiovascular events, they may be more likely to benefit from intensive risk reduction strategies than those without CVD risk factors, including strategies that target glycemic lowering. As an example, in a systematic review of five trials in adults with newly diagnosed type 2 diabetes, intensive glucose management reduced all-cause mortality at 10-year follow-up in the subset of participants with overweight or obesity (relative risk 0.64, 95% CI 0.45-0.91) but not the overall population [7]. (See 'Earlier treatment initiation' above.)

The American Diabetes Association suggests an alternative approach that includes screening all adults ages 35 years and older as well as adults under age 35 years who are at increased risk of diabetes (table 4) [44]. Guidelines from the United States Preventive Services Task Force recommend screening adults aged 35 to 70 with overweight or obesity [45].

Risk assessment — Clinical evaluation or risk assessment tools can identify individuals at elevated risk of diabetes. Common risk factors for diabetes are shown in the table (table 3) and are discussed separately. (See "Type 2 diabetes mellitus: Prevalence and risk factors".)

Online risk assessment tools include the American Diabetes Association's risk test and the Know Your Risk score in the United Kingdom [46,47]. Their sensitivities and specificities for identifying type 2 diabetes are approximately 80 and 70 percent, respectively [48-52]. Few studies have compared the validity of risk assessment tools at the population level [53,54]. (See "Type 2 diabetes mellitus: Prevalence and risk factors", section on 'Prediction models'.)

We typically perform diabetes risk assessment in the broader context of cardiovascular risk assessment. Cardiovascular risk assessment is discussed separately. (See "Atherosclerotic cardiovascular disease risk assessment for primary prevention in adults" and "Overview of primary prevention of cardiovascular disease in adults".)

When to stop screening — We typically stop screening in individuals over age 70 years because the benefits of screening are less clear and the likelihood of overdiagnosis is higher. However, we individualize our approach to screening among patients in this age group who have multiple cardiovascular risk factors and an estimated life expectancy that exceeds 10 years.

Available evidence suggests that reductions in diabetes-related mortality are unlikely in the initial 10 years after screening (see 'Clinical outcomes of screening' above). Additionally, in older patients, diabetes screening does not reliably identify a population of individuals that is at high risk of developing diabetes. (See 'Potential harms' above.)

Stopping screening at age 70 years is consistent with 2021 recommendations from the United States Preventive Services Task Force [45]. By contrast, screening recommendations from the American Diabetes Association do not include an upper age limit [44].

SCREENING TESTS

Preferred strategies — We use glycated hemoglobin (A1C) to screen for diabetes because it correlates best with rates of incident diabetes, diabetes complications, and all-cause mortality (table 1). Alternative testing strategies include adding fasting plasma glucose (FPG) to A1C testing or FPG testing alone. Both FPG and A1C assays accurately identify prediabetes and diabetes in asymptomatic adults and are inexpensive and easy to perform [44,55]. (See 'Test characteristics' below.)

Simultaneous testing with FPG and A1C may facilitate the diagnosis of diabetes because it does not require a second patient visit to confirm abnormal test results [56]. This strategy may also identify individuals with prediabetes who are at highest risk for progression to diabetes. In a prospective cohort of adults aged 48 to 68 years without diabetes, participants with both A1C and FPG in the prediabetes range had a higher risk of incident diabetes at six years than those with isolated elevated A1C or FPG (15 versus 4.9 and 3.5 percent, respectively) [57].

Measurements should ideally take place via laboratory testing of whole blood samples for A1C and plasma samples for glucose. Point-of-care (POC) testing is an acceptable alternative if laboratory testing is not available or if the devices are consistently calibrated and approved by the governing regulatory agency (eg, US Food and Drug Administration in the United States) [44].

In the United States, POC A1C testing devices do not require federal certification of test accuracy, standardization, and performance, and POC testing using fingerstick samples has lower accuracy and reproducibility than laboratory A1C tests [58,59]. In a meta-analysis of 61 studies that evaluated 13 POC devices for A1C testing, significant bias and variability existed both within and among devices [60]. (See "Measurements of chronic glycemia in diabetes mellitus", section on 'Point-of-care testing'.)

Test characteristics

Reference standard for diabetes tests — Evaluations of test accuracy of FPG or A1C typically use a two-hour oral glucose tolerance test (OGTT) as a reference standard. (See 'Other tests' below.)

Glycated hemoglobin (A1C) — A1C offers the advantages of acceptability, predictive ability, and test accuracy.

●Acceptability – A1C testing is convenient for patients and clinicians because it does not require fasting and testing can occur at any time of the day. Additionally, in individuals with an elevated A1C, test results can guide treatment decisions.

●Accuracy – The sensitivity and specificity of A1C for type 2 diabetes screening vary according to the reference standards used for comparison, thresholds for defining diabetes and prediabetes, and the population tested (table 5). A1C is a measure of chronic hyperglycemia that is well standardized across laboratories and is largely unaffected by a patient's recent changes in caloric intake or glucose metabolism [61].

A1C has moderate sensitivity and high specificity for diagnosing type 2 diabetes compared with OGTT criteria (table 5). In a meta-analysis of nine studies, the pooled sensitivity and specificity of A1C (≥6.5 percent [≥48 mmol/mol]) were 68 and 96 percent, respectively, using an OGTT reference standard (two-hour blood glucose >200 mg/dL [11.1 mmol/L]) [37]. Sensitivity and specificity may be lower for people over the age of 65.

A1C is less sensitive but still moderately specific for diagnosing prediabetes. In a meta-analysis of 46 studies, the pooled sensitivity and specificity of A1C were 49 and 79 percent, respectively, compared with the reference standard of a two-hour OGTT [62]. These results should be interpreted with caution because the studies used different diagnostic thresholds for prediabetes and high levels of between-study variability existed.

●Predictive ability – Elevations in A1C correlate strongly with diabetes-related clinical outcomes and the development of type 2 diabetes. The presence of diabetic retinopathy correlates better with A1C ≥6.5 percent than with FPG or OGTT criteria. Similarly, A1C-based definitions of prediabetes in a large, population-based cohort better predicted rates of incident diabetes, chronic kidney disease, cardiovascular disease, and all-cause mortality than did FPG- or OGTT-based definitions [21,63].

●Disadvantages – A1C testing is more expensive than glucose testing and can be difficult to interpret in patients with abnormal red cell turnover (table 6). A1C estimates chronic glucose exposure over time by measuring glucose bound to hemoglobin in red blood cells. Glucose gradually accumulates over the lifespan of the red blood cell, which is approximately 120 days. Consequently, conditions that accelerate red cell turnover, such as hemolysis, can falsely lower A1C results (table 7). A1C test parameters and alternatives for measuring chronic glycemia are discussed separately. (See 'Discordant test results' below and "Measurements of chronic glycemia in diabetes mellitus", section on 'Glycated hemoglobin (A1C)'.)

Fasting plasma glucose (FPG) — FPG is a less expensive, widely available alternative to A1C testing (table 1).

●Advantages and disadvantages – FPG is preferentially used in patients with abnormal red blood cell turnover [44]. However, FPG testing requires fasting, which is inconvenient and may adversely affect adherence. Multiple patient factors can influence FPG test reproducibility, including diurnal variations in plasma glucose levels, the duration of fasting, recent exercise, and acute stress (table 6). Glycolysis in vitro can falsely lower glucose test results if samples are improperly stored or not promptly processed [61].

●Accuracy – Similar to A1C, the sensitivity and specificity of FPG for type 2 diabetes screening vary according to the reference standards used for comparison, thresholds used to define diabetes and prediabetes, and the population tested.

FPG is moderately sensitive and highly specific for diagnosing type 2 diabetes when compared with OGTT (table 5). In a meta-analysis of nine studies, the pooled sensitivity and specificity of FPG (≥126 mg/dL [7 mmol/L]) were 56 and 98 percent, respectively, using an OGTT reference standard (two-hour blood glucose >200 mg/dL [11.1 mmol/L]) [37].

FPG is not sensitive for diagnosing prediabetes, but it is highly specific. In a meta-analysis of 19 studies, the pooled sensitivity and specificity of FPG were 25 and 94 percent, respectively, compared with the reference standard of an abnormal two-hour OGTT [62]. However, most included studies defined prediabetes (or glucose intolerance) using older diagnostic thresholds (FPG 110 to 125 mg/dL [6.0 to 6.9 mmol/L]); sensitivity of FPG would likely be higher when using current diagnostic criteria from the American Diabetes Association (FPG 100 to 125 mg/dL [5.6 to 6.9 mmol/L]).

Other tests

●Two-hour plasma glucose during an OGTT – OGTT is considered the gold standard for diabetes diagnosis (table 1) [44]. However, due to limitations, including inconvenience, cost, and variable reproducibility, we do not use OGTT to screen for type 2 diabetes in most individuals [64]. OGTT consists of a 75-gram glucose load that is given the morning after an overnight fast. Elevated blood glucose levels two hours after the glucose load indicate diabetes (glucose ≥200 mg/dL [≥11.1 mmol/L]) or prediabetes (glucose 140 to 199 mg/dL [7.8 to 11 mmol/L]). The roles of OGTT in the diagnosis of gestational and cystic fibrosis-related diabetes are reviewed separately. (See "Gestational diabetes mellitus: Screening, diagnosis, and prevention", section on 'Screening for GDM at 24 to 28 weeks' and "Cystic fibrosis-related diabetes mellitus", section on 'Surveillance'.)

●Continuous glucose monitoring – Insufficient evidence exists to support the use of continuous glucose monitoring to screen for type 2 diabetes [44].

●Urine glucose – We do not use urine glucose testing for type 2 diabetes screening, because it lacks sensitivity and specificity [65]. Defects in renal tubular function, such as type 2 (proximal) renal tubular acidosis, can cause glucosuria, so positive test results require confirmatory testing [66]. (See "Urinalysis in the diagnosis of kidney disease" and "Etiology and diagnosis of distal (type 1) and proximal (type 2) renal tubular acidosis".)

TEST INTERPRETATION AND FOLLOW-UP — 

Initial screening results guide follow-up and subsequent management.

Interpretation — Test results from diabetes screening are categorized as normal, prediabetes (or intermediate hyperglycemia), and diabetes mellitus. These categories were developed by an expert committee on the diagnosis and classification of diabetes mellitus from the American Diabetes Association [44,55]. Other organizations use slightly different diagnostic criteria [4,62].

●Normal glycemia: A1C <5.7 percent (<38.8 mmol/mol) or FPG value <100 mg/dL (5.6 mmol/L) – These test results are normal. If both tests were performed and the results are discordant, repeat testing is warranted. (See 'Discordant test results' below.)

●Prediabetes: A1C 5.7 to 6.4 percent (38.8 to 46.4 mmol/mol) or FPG 100 to 125 mg/dL (5.6 to 6.9 mmol/L) – These test results indicate prediabetes (table 2). We repeat glycated hemoglobin (A1C) or fasting plasma glucose (FPG) testing, or both, to confirm a diagnosis of prediabetes. (Related Lab Interpretation Monograph(s): "High glycated hemoglobin (A1C) in nonpregnant adults".)

The category of prediabetes encompasses a heterogeneous group of individuals whose risk of developing diabetes and diabetes-associated clinical outcomes varies considerably [32,33]. Intervention strategies in these patients are discussed separately. (See "Prevention of type 2 diabetes mellitus".)

●Diabetes: A1C ≥6.5 percent (≥47.5 mmol/mol) or FPG ≥126 mg/dL (7 mmol/L) – These test results are consistent with diabetes (table 1) [44]. (Related Lab Interpretation Monograph(s): "High glycated hemoglobin (A1C) in nonpregnant adults".)

Individuals who meet diagnostic criteria for diabetes by A1C or FPG measurements require confirmatory testing to establish a diagnosis of diabetes. If only one screening test was initially performed (eg, either FPG or A1C), confirmatory testing should take place on a subsequent day using the same test. If two screening tests were initially performed (eg, both FPG and A1C), confirmatory testing is not necessary if both test results are above their diagnostic thresholds (ie, are concordant) [56]. (See "Clinical presentation, diagnosis, and initial evaluation of diabetes mellitus in adults", section on 'Diagnostic criteria'.)

Discordant test results require additional evaluation. (See 'Discordant test results' below.)

Management of individuals with diabetes and prediabetes is discussed separately. (See "Overview of general medical care in nonpregnant adults with diabetes mellitus" and "Initial management of hyperglycemia in adults with type 2 diabetes mellitus".)

Discordant test results — If two different tests are discordant, the test that is diagnostic of diabetes should be repeated to confirm the diagnosis [44]. Clinicians should also investigate the potential reasons for discordant test results (table 6). An approach to discordant test results and factors that cause inaccurate FPG or A1C measurements are discussed separately. (See 'Screening tests' above and "Measurements of chronic glycemia in diabetes mellitus", section on 'Unexpected or discordant values'.)

When to rescreen — Intervals for rescreening depend on the patient's initial screening test results and baseline risk of developing diabetes.

●Normal results – We typically rescreen individuals with normal test results every three years because this interval may minimize the burden of false-positive results and the potential impact of false-negative results [44]. This includes patients with a history of gestational diabetes or other diabetes risk factors whose initial screening tests are normal.

●Borderline results (prediabetes) – Individuals who meet the criteria for prediabetes should undergo repeat screening at one year. Management and ongoing monitoring of these individuals are discussed separately. (See "Prevention of type 2 diabetes mellitus".)

These rescreening intervals are consistent with standards of care from the American Diabetes Association [44].

SOCIETY GUIDELINE LINKS — 

Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Diabetes mellitus in adults".)

INFORMATION FOR PATIENTS — 

UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or email these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topics (see "Patient education: Type 1 diabetes (The Basics)" and "Patient education: Type 2 diabetes (The Basics)" and "Patient education: Hemoglobin A1C tests (The Basics)" and "Patient education: Lowering your risk of prediabetes and type 2 diabetes (The Basics)")

●Beyond the Basics topics (see "Patient education: Type 1 diabetes: Overview (Beyond the Basics)" and "Patient education: Type 2 diabetes: Overview (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●Benefits and harms of screening – Although a direct benefit of diabetes screening on mortality has not been established, screening for type 2 diabetes improves detection of diabetes and, when coupled with intensive interventions, can delay the onset of diabetes and enhance control of cardiovascular risk factors, such as hypertension. Treatment at the time of diabetes diagnosis reduces cardiovascular events, rates of microvascular complications, and all-cause and diabetes mortality. (See 'Benefits and harms of screening' above.)

Potential harms of screening include overdiagnosis and false reassurance. Potential harms of treatment include risks of overtreatment and potential adverse effects from medications, especially in frail, older adults. (See 'Potential harms' above.)

●Whom to screen – For adults ages 35 to 70 years who are at increased risk for diabetes and/or have cardiovascular risk factors, we suggest screening for type 2 diabetes (Grade 2C). This includes individuals with overweight and obesity, prior gestational diabetes, polycystic ovary syndrome, HIV disease, or a first-degree relative with type 2 diabetes as well as those with hypertension, dyslipidemia, or established cardiovascular disease (table 3). This approach balances the benefits of early diabetes treatment with the potential harms of overdiagnosis and the lack of clear evidence that diabetes screening reduces mortality. (See 'Whom to screen' above.)

●Screening tests – We use glycated hemoglobin (A1C) for diabetes screening because it correlates best with rates of future diabetes and diabetes complications. Alternative strategies include using fasting plasma glucose (FPG) with or without simultaneous A1C testing. (See 'Preferred strategies' above and 'Test characteristics' above.)

Both A1C and FPG are relatively inexpensive and convenient. Both tests are moderately sensitive and highly specific for diagnosing diabetes but have lower sensitivities and specificities for prediabetes (table 5). (See 'Glycated hemoglobin (A1C)' above and 'Fasting plasma glucose (FPG)' above.)

●Interpreting results

•A1C <5.7 percent (<38.8 mmol/mol) or FPG value <100 mg/dL (5.6 mmol/L) – These test results are normal. If the two tests were performed simultaneously and one is above these thresholds, repeat testing is warranted to clarify discordant test results. (See 'Discordant test results' above.)

•A1C 5.7 to 6.4 percent (38.8 to 46.4 mmol/mol) or FPG 100 to 125 mg/dL (5.6 to 6.9 mmol/L) – These test results indicate prediabetes (table 2).

•A1C of ≥6.5 percent (≥47.5 mmol/mol) or FPG of ≥126 mg/dL (7 mmol/L) – These test results are consistent with diabetes (table 1). (Related Lab Interpretation Monograph(s): "High glycated hemoglobin (A1C) in nonpregnant adults".)

In individuals with abnormal initial tests, we repeat testing to confirm the diagnosis. In individuals with unexpected or discordant test results, we investigate potential causes of abnormal tests (table 6) and repeat the test whose result met the diagnostic criteria for diabetes or prediabetes. (See 'Interpretation' above and 'Discordant test results' above.)

●Frequency of rescreening – We rescreen individuals with normal test results at three-year intervals and those with prediabetes at one year. (See 'When to rescreen' above.)

●Management of patients with diabetes or prediabetes – Individuals who meet diagnostic criteria for diabetes (table 1) and prediabetes (table 2) should receive intensive lifestyle counseling. Management of these patients is reviewed separately. (See "Overview of general medical care in nonpregnant adults with diabetes mellitus" and "Initial management of hyperglycemia in adults with type 2 diabetes mellitus" and "Prevention of type 2 diabetes mellitus".)

ACKNOWLEDGMENT — 

The UpToDate editorial staff acknowledges David McCulloch, MD, who contributed to earlier versions of this topic review.