Diabetes Screening and Prevention

Table II.

Test	Advantages	Limitations	Recommendations
Urine glucose	Does not require a blood sample; can be tested in fasting, random or postprandial state; rapid processing time; inexpensive	Unable to measure glucose above the renal threshold; variable renal threshold; affected by fluid intake and urine concentration; not fully quantitative	Not recommended
Random blood glucose	Easy to obtain; no fasting required; inexpensive	Requires prompt processing (<2 hours) thus the potential for error; point measurement can be affected by several factors (short-term lifestyle changes, time since prior meal, etc.)	Not recommended
Fasting plasma glucose	Relatively cheap and simple; single plasma glucose level measured; highly correlated with presence of complications	Requires the patient to fast overnight (at least 8 hours), potential for processing error; point measurement can be affected by short-term lifestyle changes, risks of phlebotomy	Can be used
75-g oral glucose tolerance test	Current gold standard for the diagnosis of diabetes; most sensitive test for impaired glucose tolerance	Requires 8-hour fast, lengthy and requires commitment of nurse staff, overall test-retest reproducibility lower than with other tests	Can be used
Glycated hemoglobin (HbA1c)	Stable marker of long-term glycemic level; no fasting required, can be completed at any time; not affected by short-term lifestyle changes; requires only venous blood or a point-of-care testing capillary sample, lower intraindividual variability (<2%) than fasting plasma glucose	Value may vary with assay method used; possible non-glycemic causes of error such as hemoglobinopathies and anemia; may be insensitive for detection of impaired fasting glucose or impaired glucose tolerance; high cost (more expensive than glucose testing); limited availability in some areas of the world	Can be used
50-g glucose challenge test	No fasting required; not influenced by the time of the day	Cumbersome to administer; test-retest reproducibility unclear	Not recommended
Capillary blood glucose measurement (point-of-care testing)	Simple; cheap; no phlebotomy required	Imprecision may be high; not standardized	Not recommended
Fructosamine	Marker of glycemic level over several weeks; may be an alternative to HbA1c in case of hemoglobinopathy	Can be influenced by serum protein or albumin levels; no clear association between the concentration and chronic complications of diabetes; performance may be limited	Not recommended

Urine glucose – Glycosuria is a poor screening instrument for diabetes. The sensitivity of urine glucose testing ranges from 16% to 64%, and the positive predictive value from 11%-37%.

Random blood glucose (RBG) – The use of RBG as a screening tool is somewhat limited by its low performance. At a cut-off of ≥6.9 mmol/l, the specificity of RBG is about 93% and the sensitivity 41%. For identification of prediabetes, the specificity has been estimated at ~ 94% and the sensitivity at 23%. An expert panel has recommended a RBG cut-off of ≥7.2 mmol/l (sensitivity of 63% and specificity of 87%).

Fasting plasma glucose (FPG) – FPG screening has modest sensitivity as a screen for hyperglycemia. A threshold of ≥ 7mmol/l detects only 55.7% of people with diabetes, based on diagnosis by an OGTT, with 100% specificity. A cut-off for FPG of > 6.1 mmol/l may be optimal with a somewhat lower sensitivity of 85.2% but a specificity of 88.5%. FPG may not be sensitive for the diagnosis of IGT as compared to the OGTT; a threshold of > 5.6 mmol/l only detects 28.9% of IGT cases whereas OGTT would identify 87.4% of cases.

Glycated hemoglobin (HbA_1c) – An HbA_1c value between 5.8% and 6.3% has a 6%-95% sensitivity and 97%-98% specificity as a screen for diabetes. 6.1%-6.2% is optimal in terms of sensitivity. HbA_1c ≥6.1% would have a 63.2% sensitivity and 97.4% specificity for screening for type 2 diabetes; this value corresponds most closely with a two-hour post-prandial glucose concentration of 11.1 mmol/1, and would include about 41% of non-diabetic subjects and 21% of subjects with IGT.

For detecting prediabetes, HbA_1c ≥6.1% has at best a sensitivity of 50% for the detection of IGT (47); HbA_1c ≥5.8% will only detect about 30% of people with prediabetes (50), HbA_1c of ≥5.7% has a sensitivity of 59.4% and a specificity of 73.9%, HbA_1c of ≥5.6 % may be optimal for prediabetes detection with an area under the receiver-operating-characteristic curve (AUC) of 0.63 to 0.71.

In combination with either RBG or FPG, HbA_1c may add value in identifying the subgroups of individuals that need to undergo an OGTT.

Effectiveness of screening for hyperglycemia

Impact of screening for hyperglycemia on morbidity and mortality

Ideally, randomized trials of screening comparing people offered and those not offered screening would provide the highest level of evidence on the effect of screening for diabetes on morbidity and mortality. However, the results of such a trial have not yet been published; the ongoing ADDITION-Cambridge study may provide an answer to the question. The ADDITION trial, which compared intensive treatment vs. routine care, demonstrated benefits from early treatment of people with screen-detected diabetes.

Observational studies have shown a likely benefit of screening on morbidity and mortality. A 10-year retrospective matched case-control study found that screen-detected diabetes was associated with a 13% (95% CI: -62, 98) relative reduction in microvascular outcomes compared with routine diagnosis. Another study showed that over 12 years compared with age- and sex-matched controls without diabetes (P<0.005), people with diabetes detected by glycosuria lost 1.96 years of life and people with conventionally diagnosed diabetes, 3.42 years. Unfortunately, these observational studies may be plagued by biases and confounding that limit their validity. In addition, they either focused on microvascular outcomes or conducted glycosuria screening, which has a very limited performance (see above).

There is also no direct trial evidence of benefit in health outcomes from screening for prediabetes. Nonetheless, intensive programs of lifestyle modification (diet, exercise, and behavior), and metformin do reduce the incidence of diabetes among screen-detected people in the US Diabetes Prevention Program, with collateral benefits on cardiovascular risk factors.

Psychosocial impact of screening for hyperglycemia

Observational studies have found limited or no psychological effect of screening on newly-detected people with type 2 diabetes. No effect has been found on (1) screen-detected patients’ perceived health status and well-being after notification of the results, (2) anxiety levels in screen-detected patients, and (3) short-term (2 weeks) or long-term (6-12 months) adverse or positive effects on quality of life after diagnosis. This suggests that diabetes screening has little or no labeling effect. However, observational data is susceptible to misinterpretation because of biases, and it mainly focuses on people with screen-detected diabetes, ignoring those who screened negative but may still be affected psychologically.

Trial data also suggest no adverse psycho-social effect of screening for diabetes. In both the short- and long-term, the anxiety level, illness perception and self-rated health of participants invited to screening (with or without diabetes at screening) do not differ from that of those not invited, whether measured immediately after the test, at 6 weeks, at 3-6 months or at 12-15 months. Furthermore, negative screening test results do not promote false reassurance (expressed as lower perceived risk, lower intentions for health-related behavioral change, or higher self-rated health), or negatively affect behaviors (smoking, alcohol consumption, dietary intake, or physical activity), at least in the first year after screening.

The psychological effects of receiving a diagnosis of prediabetes largely remain unclear. However, there are suggestions that participation in a diabetes prevention program would not be associated with higher levels of anxiety, depression or overall psychological distress.

Cost-effectiveness of screening for hyperglycemia

A systematic review of economic studies of diabetes screening indicates that (i) screening for hyperglycemia would be cost-effective, (ii) targeted screening of high-risk groups would be more cost-effective than universal screening and (iii) treatment is the key determinant of cost-effectiveness. Screening for IGT followed by lifestyle modification or treatment with metformin was also cost-effective, with the lifestyle having a better cost-effectiveness ratio.

Detection and treatment of people with prediabetes appeared to be much more cost-effective than detection and treatment limited to individuals with both IGT and IFG. Screening for diabetes and IGT, followed by lifestyle intervention in persons with IGT, would be more cost-effective than screening for type 2 diabetes alone or screening for type 2 diabetes and IGT followed by metformin therapy.

A recent US-based simulation of nine screening strategies found that screening for undiagnosed diabetes would be most cost-effective if started between the ages of 30 and 45 years and repeated every 3-5 years. Another US-based comparison of five opportunistic screening strategies (plasma or capillary RBG, HbA_1c, plasma or capillary GCT and subsequent OGTT) for both prediabetes and type 2 diabetes found that such an approach is cost saving from a health care perspective and is cost-neutral to society.

Screening intervals

The frequency with which screening for hyperglycemia would be most cost-effective is not known, as there are no compelling data from which to make this determination. An optimal interval between screening rounds would be one at which the prevalence of undiagnosed cases reaches the prevalence of such cases at the previous screening, and the cost-effectiveness is the same for each screening. A recent US-based simulation suggested that targeted screening for type 2 diabetes would be most cost-effective if repeated every 3-5 years. Based on expert opinions, professional organizations have generally favored a 3-year interval.

Management or treatment of screen-detected prediabetes or undiagnosed diabetes

Therapies for undiagnosed diabetes

Multifactorial intervention for screen-detected diabetes – The ADDITION study compared intensive treatment to standard therapy in exclusively screen-detected patients. The ADDITION intensive intervention was modeled after the regimen used in the Steno-2 trial, which showed that intensified multifactorial intervention (lifestyle modification and multidrug therapy) aimed at controlling cardiovascular risk factors is more cost-effective (with a reduction in macrovascular and microvascular complications and all-cause mortality) than standard therapy among people with longstanding type 2 diabetes and microalbuminuria.

In ADDITION, intensive multifactorial treatment (targeting several cardiovascular risk factors) of screen-detected people over five years provided a non-significant 17% reduction in a composite cardiovascular primary endpoint compared with routine care, but greater improvements in cardiovascular risk factors, reductions in cardiovascular death (12%), nonfatal myocardial infarction (30%), and revascularization (21%). These results may partly be explained by improvements in the quality of diabetes care during the trial period, a phenomenon that is likely to happen in any screening trial. Nonetheless, ADDITION showed some macrovascular benefits of early treatment for screen-detected diabetes.

The long term follow-up of the ADDITION study will provide a more definite answer on the effect of early treatment in people with screen-detected diabetes. While waiting for more direct evidence for treatment among screen-detected patients, data from intervention studies comparing the effects of individual treatment to lower blood glucose, blood pressure and serum cholesterol in conventionally diagnosed diabetic populations (ideally newly diagnosed) can inform the practice of early treatment.

Glycemic control – The United Kingdom Prospective Diabetes Study (UKPDS), which looked at newly-diagnosed type 2 diabetes patients, provides estimates of the likely effect of glycemic control during the lead-time. The 10-year post-trial monitoring results showed that tight glycemic control reduced myocardial infarction by 15% (P=0.01) and 33% (P=0.005) in the sulfonylurea-insulin and metformin groups, respectively, compared with the control group. However, there were no reductions in stroke, heart failure, angina and all-cause mortality associated with tight glycemic control.

Lipid-lowering therapy – There is clear evidence that lowering lipids in people with diabetes is beneficial. A meta-analysis of 12 trials using either statins and/or fibric acid derivatives showed that lipid-lowering therapy is equally efficacious in people with and without diabetes for primary and secondary cardiovascular disease prevention. In both primary and secondary primary prevention, the use of lipid-lowering drugs was associated with a significant 21% reduction in major coronary events in patients with diabetes, respectively.

Except for the Collaborative Atorvastatin Diabetes Study (CARDS), the trials included in the aforementioned meta-analysis were not specifically aimed at people with diabetes and only demonstrated these results in post-hoc analyses of sub-populations with diabetes. CARDS, a primary prevention trial conducted exclusively in patients with type 2 diabetes, showed a 37% (95% CI: 17, 52) relative reduction in cardiovascular events. This reduction was independent of the baseline level of cholesterol. Similar findings were reported in the Heart Protection Study (HPS) demonstrating a 22% (95% CI 13, 30) relative reduction in cardiovascular events in the sub-population with diabetes.

Antihypertensive treatment – In UKPDS, a study of people with newly diagnosed diabetes, tight blood pressure control compared with less tight control significantly reduced macrovascular and microvascular risks in patients with newly diagnosed diabetes. This indicates the potential beneficial effects of blood-pressure-lowering therapy among screen-detected patients with undiagnosed diabetes.

Aspirin therapy – Aspirin has been shown to be effective for secondary prevention of cardiovascular disease in people with diabetes. A meta-analysis confirmed that aspirin allocation yielded greater absolute reductions in serious vascular events (6·7% vs. 8·2% per year, P<0.0001), total stroke (2·08% vs. 2·54% per year, P=0.002) and coronary events (4·3% vs. 5·3% per year, P<0.0001). When studied for primary prevention in individuals with diabetes, aspirin was found be several meta-analyses not to reduce the risk of cardiovascular events among people with diabetes. As a result, the ADA and the American Heart Association (AHA) do not recommend the blanket use of aspirin in all patients with diabetes, but rather that the use of aspirin should be guided by an assessment of their absolute global cardiovascular risk.

Lifestyle modification – Compared to routine diabetes education and support, intensive lifestyle modification has been shown to independently improve cardiovascular risk factors in people with type 2 diabetes over a 4-year period in the Look AHEAD (Action for Health in Diabetes) study. Participants had a greater percentage of weight loss compared to routine diabetes support and education participants (–6.15% vs –0.88%; P< 0.001) and greater improvements in fitness (12.74% vs 1.96%; P< 0.001), HbA_1c level (–0.36% vs –0.09%; P < 0.001), systolic (–5.33 vs –2.97 mm Hg; P< 0.001) and diastolic (–2.92 vs –2.48 mm Hg; P = 0.01) blood pressure, and levels of high-density lipoprotein cholesterol (3.67 vs 1.97 mg/dL; P< 0.001) and triglycerides (–25.56 vs. –19.75 mg/dL; P<0.001). However, the independent effect of intensive lifestyle modification on cardiovascular events remains unknown.

Therapies for prediabetes

Screen-detected prediabetes can be managed with lifestyle intervention and/or pharmacotherapy. There is robust trial evidence supporting the efficacy of each of these therapies. Of the two, lifestyle intervention appears to be more effective.

Lifestyle intervention (dietary and physical activity recommendations to achieve weight loss) has been shown to reduce the progression from IGT to diabetes by 30%-60% in several randomized controlled trials. Lifestyle interventions have been efficacious across a broad range of ethnic groups, and the benefits of lifestyle modification in reducing diabetes incidence persisting several years after discontinuation of the active intervention. In three landmark studies, sustained relative reductions (34% to 43%) in the incidence of diabetes were evident 3 to 14 years after the termination of the active lifestyle intervention.

Lifestyle modification has also been shown to improve cardiovascular risk factors (blood pressure, serum cholesterol, and serum triglycerides) in people with prediabetes; however, there was no significant decline in hard cardiovascular outcomes. Lifestyle intervention also has beneficial effects on microvascular outcomes; 14 years after the active phase of the Da Qing study, the lifestyle intervention group exhibited a 47% lower incidence of severe retinopathy than the control group. Delivery Group delivery of lifestyle intervention in community settings has the potential to be significantly less expensive while still achieving weight loss similar to that observed with individually-based interventions used in landmark trials.

Several drugs have been shown to prevent the progression of impaired glucose tolerance to type 2 diabetes; these include metformin, troglitazone, rosiglitazone, arcabose, orlistat, voglibose, and valsartan. However, for most drugs, concerns regarding their costs, side effects, and the lack of persistence of efficacy limit their use, including acarbose despite its collateral positive effects on cardiovascular events (49% reduction) and hypertension (34% decreased incidence). The drug with the best safety profile and efficacy is metformin, and it has therefore been recommended for preventing diabetes in patients with a body mass index ≥35 kg/m², and in younger patients (less than 60 years of age). A study in the British Medical Journal found that whereas lifestyle interventions benefitted patients at any level of risk for diabetes, metformin only benefitted the highest quartile, but did so significantly (number needed to treat of 4.6).

Recommendations for screening

Recommendations of various US professional organizations

The currently available evidence does not support universal screening; consequently, most professional organizations advocate a selective and opportunistic approach in high-risk populations.

The US Preventive Services Task Force (USPSTF) recommends screening for undiagnosed type 2 diabetes in adults with a sustained blood pressure ≥135/80 mmHg. The American Diabetes Association (ADA) recommends opportunistic screening of adults of any age with a body mass index ≥25 kg/m² and additional risk factors, which include physical inactivity, a first-degree relative with diabetes, high-risk race/ethnicity (e.g., African American, Latino, Native American, Asian American, Pacific Islander), women who delivered a baby weighing 9 lb or were diagnosed with gestational diabetes mellitus, hypertension (≥140/90 mmHg or on therapy for hypertension), an HDL-cholesterol level <35 mg/dl (0.90 mmol/l) and/or a triglyceride level >250 mg/dl (2.82 mmol/l), women with polycystic ovarian syndrome, HbA_1c≥5.7%, IGT, or IFG on previous testing, other clinical conditions associated with insulin resistance (e.g., severe obesity, acanthosis nigricans), and a history of cardiovascular disease. In adults with normal weight and without any risk factors, screening can begin at age 45. Both the USPTF and the ADA recommend using either a fasting plasma glucose or HbA_1c for testing every three years. The ADA further recommends the consideration of more frequent testing depending on initial results and risk.

The American Academy of Family Physicians (AAFP) recommends screening for type 2 diabetes in adults with hypertension and hyperlipidemia. The AAFP found insufficient evidence to recommend for or against adults who are at low risk for coronary vascular disease.

For prediabetes, the ADA and USPSTF recommendations are similar to that for undiagnosed diabetes, but with the use of OGTT as a screening test in addition to FPG or HbA_1c. The ADA recommends that patients detected with prediabetes or an HbA_1c of 5.7%–6.4% be referred to an effective ongoing support program targeting weight loss of 7% and increasing physical activity to at least 150 min/week of moderate activity, such as walking. For those at very-high-risk of progression to type 2 diabetes (presence of both IGT and IFG or of multiple risk factors), especially if there is progression of hyperglycemia [e.g., HbA_1c ≥6.0%] despite lifestyle interventions), the ADA advises additional use of metformin therapy. Furthermore, monitoring for the development of prediabetes should be performed every year.

The Endocrine Society recommends screening patients for prediabetes for metabolic syndrome, as defined by the American Heart Association/National Heart, Lung, and Blood Institute, at least every 3 years. At least 3 components in any patient puts him/her at high risk for metabolic syndrome. It also recommends that patients with a previous history of diabetes be screened for diabetes every 1 to 2 years using FPG or OGTT. Also, all patients at metabolic risk should undergo global risk assessment to determine their 10-year risk for coronary heart disease or cerebral vascular disease using the Framingham Risk Evaluation algorithm before starting preventive therapy.

The American Association of Clinical Endocrinologists (AACE) recommends testing for prediabetes among all subjects with high-risk factors for developing diabetes using the 2-hour OGTT or FPG.

The Indian Health Service recommends screening for prediabetes in high-risk patients using FPG or 2-hour OGTT to screen high-risk patients.

Of all the recommendations made by US professional organizations, the USPSTF and ADA recommendations were developed after the most rigorous process of systematic review of the available evidence, and thus offer a solid basis for clinical practice. However, the USPSTF guidelines are more restrictive in terms of the target population; consequently, these may be less appropriate for use in clinical practice.

Summary and recommendations for clinical practice

Reasons to screen for prediabetes and diabetes include the increasingly high prevalence of these conditions worldwide, identifiable pre-diabetic and long latent diabetic phases, and the availability of suitable, reliable, high performance and acceptable tests that can aid in the detection of hyperglycemic conditions. Although no trial assessing the effectiveness of screening for hyperglycemia on health outcomes (morbidity and mortality) has been published, testing for diabetes does not seem to be associated with adverse psychosocial consequences. However, the effect of labeling someone as “prediabetic” remains unknown.

There are accepted and cost-effective treatments for recognized prediabetes, which include lifestyle modification and metformin; however, the effect of these therapies on cardiovascular outcomes remains unknown. There also seems to be an incremental benefit of intensive multifactorial therapy over current standards in people with screen-detected diabetes, which will need confirmation with additional follow-up. Furthermore, indirect evidence from studies of people with recently conventionally-diagnosed diabetes indicates potential benefits of early treatment for diabetes. Economic models strongly support the undertaking of targeted screening for both prediabetes and diabetes.

Based on the overall available evidence, it appears to be justifiable to undertake opportunistic screening in an organized manner for prediabetes and undiagnosed diabetes among asymptomatic high-risk people every 3 years using the ADA criteria, which are based on a robust derivation process. People detected as having diabetes should be treated using a multifactorial approach aiming to control blood glucose and other cardiovascular risk factors using existing cardioprotective cost-effective drugs as well as lifestyle modification. People detected with prediabetes should primarily be managed by lifestyle modification, with the addition of metformin for those at highest risk.

Copyright © 2017, 2013 Decision Support in Medicine, LLC. All rights reserved.

No sponsor or advertiser has participated in, approved or paid for the content provided by Decision Support in Medicine LLC. The Licensed Content is the property of and copyrighted by DSM.