D iabetes mellitus is a chronic illness

P O S I T I O N S T A T E M E N T Standards of Medial Care in Diabetesd2013 AMERICAN DIABETES ASSOCIATION D iabetes mellitus is a hroni illness that requires ontinuing medial are and ongoing patient self-management eduation and support to prevent aute ompliations and to redue the risk of long-term ompliations. Diabetes are is omplex and requires multifatorial risk redution strategies beyond glyemi ontrol. A large body of evidene exists that supports a range of interventions to improve diabetes outomes. These standards of are are intended to provide liniians, patients, researhers, payers, and other interested individuals with the omponents of diabetes are, general treatment goals, and tools to evaluate the quality of are. Although individual preferenes, omorbidities, and other patient fators may require modifiation of goals, targets that are desirable for most patients with diabetes are provided. Speifially titled setions of the standards address hildren with diabetes, pregnant women, and people with prediabetes. These standards are not intended to prelude linial judgment or more extensive evaluation and management of the patient by other speialists as needed. For more detailed information about management of diabetes, refer to referenes (1 3). The reommendations inluded are sreening, diagnosti, and therapeuti ations that are known or believed to favorably affet health outomes of patients with diabetes. A large number of these interventions have been shown to be osteffetive (4). A grading system (Table 1), developed by the Amerian Diabetes Assoiation (ADA) and modeled after existing methods, was utilized to larify and odify the evidene that forms the basis for the reommendations. The level of evidene that supports eah reommendation is listed after eah reommendation using the letters A, B, C, or E. These standards of are are revised annually by the ADA s multidisiplinary Professional Pratie Committee, inorporating new evidene. For the urrent revision, ommittee members systematially searhed Medline for human studies related to eah subsetion and published sine 1 January 2011. Reommendations (bulleted at the beginning of eah subsetion and also listed in the Exeutive Summary: Standards of Medial Care in Diabetesd2013 ) were revised based on new evidene or, in some ases, to larify the prior reommendation or math the strength of the wording to the strength of the evidene. A table linking the hanges in reommendations to new evidene an be reviewed at http://professional.diabetes. org/cpr. As is the ase for all position statements, these standards of are were reviewed and approved by the Exeutive Committee of ADA s Board of Diretors, whih inludes health are professionals, sientists, and lay people. Feedbak from the larger linial ommunity was valuable for the 2013 revision of the standards. Readers who wish to omment on the Standards of Medial Care in Diabetesd2013 are invitedtodosoathttp://professional. diabetes.org/cpr. Members of the Professional Pratie Committee dislose all potential finanial onflits of interest with industry. These dislosures were disussed at the onset of the standards revision meeting. Members of the ommittee, their employer, and their dislosed onflits of interest are listed in the Professional Pratie Committee for the 2013 Clinial Pratie Reommendations table (see p. S109). The ADA funds development of the standards and all its position statements out of its general revenues and Originally approved 1988. Most reent review/revision Otober 2012. DOI: 10.2337/d13-S011 2013 by the Amerian Diabetes Assoiation. Readers may use this artile as long as the work is properly ited, the use is eduational and not for profit, and the work is not altered. See http://reativeommons.org/ lienses/by-n-nd/3.0/ for details. does not use industry support for these purposes. I. CLASSIFICATION AND DIAGNOSIS A. Classifiation The lassifiation of diabetes inludes four linial lasses: Type 1 diabetes (results from b-ell destrution, usually leading to absolute insulin defiieny) Type 2 diabetes (results from a progressive insulin seretory defet on the bakground of insulin resistane) Other speifi types of diabetes due to other auses, e.g., geneti defets in b-ell funtion, geneti defets in insulin ation, diseases of the exorine panreas (suh as ysti fibrosis), and drug- or hemial-indued (suh as in the treatment of HIV/AIDS or after organ transplantation) Gestational diabetes mellitus (GDM) (diabetes diagnosed during pregnany that is not learly overt diabetes) Some patients annot be learly lassified as type 1 or type 2 diabeti. Clinial presentation and disease progression vary onsiderably in both types of diabetes. Oasionally, patients who otherwise have type 2 diabetes may present with ketoaidosis. Similarly, patients with type 1 diabetes may have a late onset and slow (but relentless) progression of disease despite having features of autoimmune disease. Suh diffiultiesindiagnosismay our in hildren, adolesents, and adults. The true diagnosis may beome more obvious over time. B. Diagnosis of diabetes For deades, the diagnosis of diabetes was based on plasma gluose riteria, either the fasting plasma gluose (FPG) or the 2-h value in the 75-g oral gluose tolerane test (OGTT) (5). In 2009, an International Expert Committee that inluded representatives of the ADA, the International Diabetes are.diabetesjournals.org DIABETES CARE, VOLUME 36, SUPPLEMENT 1, JANUARY 2013 S11

Table 1dADA evidene grading system for linial pratie reommendations Level of evidene A B C E Federation (IDF), and the European Assoiation for the Study of Diabetes (EASD) reommended the use of the A1C test to diagnose diabetes, with a threshold of $6.5% (6), and the ADA adopted this riterion in 2010 (5). The diagnosti test should be performed using a method that is ertified by the NGSP and standardized or traeable to the Diabetes Control and Compliations Trial (DCCT) referene assay. Although point-of-are (POC) A1C assays may be NGSP ertified, profiieny testing is not mandated for performing the test, so use of these assays for diagnosti purposes ould be problemati. Epidemiologial datasets show a similar relationship for A1C to the risk of retinopathy as has been shown for the orresponding FPG and 2-h PG thresholds. The A1C has several advantages to the FPG and OGTT, inluding greater onveniene (sine fasting is not required), evidene to suggest greater preanalytial stability, and less day-to-day perturbations during periods of stress and illness. These advantages must be balaned by greater ost, the limited availability of A1C testing in ertain regions of the developing world, and the inomplete orrelation between A1C and average gluose in ertain individuals. In addition, HbA 1 levels may vary with patients rae/ethniity (7,8). Some Desription Clear evidene from well-onduted, generalizable RCTs that are adequately powered, inluding: Evidene from a well-onduted multienter trial Evidene from a meta-analysis that inorporated quality ratings in the analysis Compelling nonexperimental evidene, i.e., all or none rule developed by the Centre for Evidene-Based Mediine at the University of Oxford Supportive evidene from well-onduted RCTs that are adequately powered, inluding: Evidene from a well-onduted trial at one or more institutions Evidene from a meta-analysis that inorporated quality ratings in the analysis Supportive evidene from well-onduted ohort studies Evidene from a well-onduted prospetive ohort study or registry Evidene from a well-onduted meta-analysis of ohort studies Supportive evidene from a well-onduted ase-ontrol study Supportive evidene from poorly ontrolled or unontrolled studies Evidene from randomized linial trials with one or more major or three or more minor methodologial flaws that ould invalidate the results Evidene from observational studies with high potential for bias (suh as ase series with omparison with historial ontrols) Evidene from ase series or ase reports Confliting evidene with the weight of evidene supporting the reommendation Expert onsensus or linial experiene have posited that glyation rates differ by rae (with, for example, Afrian Amerians having higher rates of glyation), but this is ontroversial. A reent epidemiologial study found that, when mathed for FPG, Afrian Amerians (with and without diabetes) indeed had higher A1C than whites, but also had higher levels of frutosamine and glyated albumin and lower levels of 1,5 anhydrogluitol, suggesting that their glyemi burden (partiularly postprandially) may be higher (9). Epidemiologial studies forming the framework for reommending use of the A1C to diagnose diabetes have all been in adult populations. Whether the ut point would be the same to diagnose hildren or adolesents with type 2 diabetes is an area of unertainty (3,10). A1C inaurately reflets glyemia with ertain anemias and hemoglobinopathies. For patients with an abnormal hemoglobin but normal red ell turnover, suh as sikle ell trait, an A1C assay without interferene from abnormal hemoglobins should be used (an updated list is available at www. ngsp.org/interf.asp). For onditions with abnormal red ell turnover, suh as pregnany, reent blood loss or transfusion, or some anemias, the diagnosis of diabetes must employ gluose riteria exlusively. The established gluose riteria for the diagnosis of diabetes (FPG and 2-h PG) remain valid as well (Table 2). Just as there is less than 100% onordane between the FPG and 2-h PG tests, there is no perfet onordane between A1C and either gluose-based test. Analyses of the National Health and Nutrition Examination Survey (NHANES) data indiate that, assuming universal sreening of the undiagnosed, the A1C ut point of $6.5% identifies one-third fewer ases of undiagnosed diabetes than a fasting gluose ut point of $126 mg/dl (7.0 mmol/l) (11), and numerous studies have onfirmed that at these ut points the 2-h OGTT value diagnoses more sreened people with diabetes (12). However, in pratie, a large portion of the diabeti population remains unaware of its ondition. Thus, the lower sensitivity of A1C at the designated ut point may well be offset by the test s greater pratiality, and wider appliation of a more onvenient test (A1C) may atually inrease the number of diagnoses made. As with most diagnosti tests, a test result diagnosti of diabetes should be repeated to rule out laboratory error, unless the diagnosis is lear on linial grounds, suh as a patient with a hyperglyemi risis or lassi symptoms of hyperglyemia and a random plasma gluose $200 mg/dl. It is preferable that the same test be repeated for onfirmation, sine there will be a greater likelihood of onurrene in this ase. For example, if the A1C is 7.0% and a repeat result is 6.8%, the diagnosis of diabetes is onfirmed. However, if two different tests (suh as A1C and FPG) are both above the diagnosti thresholds, the diagnosis of diabetes is also onfirmed. On the other hand, if two different tests are available in an individual and the results are disordant, the test whose result is above the diagnosti ut point should be repeated, and the diagnosis is made based on the onfirmed test. That is, if a patient meets the diabetes riterion of the A1C (two results $6.5%) but not the FPG (,126 mg/ dl or 7.0 mmol/l), or vie versa, that person should be onsidered to have diabetes. Sine there is preanalytial and analytial variability of all the tests, it is also possible that when a test whose result was above the diagnosti threshold is repeated, the seond value will be below the diagnosti ut point. This is least likely for A1C, somewhat more likely for FPG, and most likely for the 2-h PG. Barring a laboratory error, suh patients are likely to have test results near the margins of the threshold for a diagnosis. The health are professional might opt to S12 DIABETES CARE, VOLUME 36, SUPPLEMENT 1, JANUARY 2013 are.diabetesjournals.org

follow the patient losely and repeat the testing in 3 6 months. The urrent diagnosti riteria for diabetes are summarized in Table 2. C. Categories of inreased risk for diabetes (prediabetes) In 1997 and 2003, the Expert Committee on Diagnosis and Classifiation of Diabetes Mellitus (13,14) reognized an intermediate group of individuals whose gluose levels, although not meeting riteria for diabetes, are nevertheless too high to be onsidered normal. These persons were defined as having impaired fasting gluose (IFG) (FPG levels 100 mg/dl [5.6 mmol/l] to 125 mg/dl [6.9 mmol/l]) or impaired gluose tolerane (IGT) (2-h values in the OGTT of 140 mg/dl [7.8 mmol/l] to 199 mg/dl [11.0 mmol/l]). It should be noted that the World Health Organization (WHO) and a number of other diabetes organizations define the utoff for IFG at 110 mg/dl (6.1 mmol/l). Individuals with IFG and/or IGT have been referred to as having prediabetes, indiating the relatively high risk for the future development of diabetes. IFG and IGT should not be viewed as linial entities in their own right but rather risk fators for diabetes as well as ardiovasular disease (CVD). IFG and IGT are assoiated with obesity (espeially abdominal or viseral obesity), dyslipidemia with high triglyerides and/or low HDL holesterol, and hypertension. As is the ase with the gluose measures, several prospetive studies that Table 2dCriteria for the diagnosis of diabetes A1C $6.5%. The test should be performed in a laboratory using a method that is NGSP ertified and standardized to the DCCT assay.* OR FPG $126 mg/dl (7.0 mmol/l). Fasting is defined as no alori intake for at least 8 h.* OR 2-h plasma gluose $200 mg/dl (11.1 mmol/l) during an OGTT. The test should be performed as desribed by the WHO, using a gluose load ontaining the equivalent of 75 g anhydrous gluose dissolved in water.* OR In a patient with lassi symptoms of hyperglyemia or hyperglyemi risis, arandomplasmagluose$200 mg/dl (11.1 mmol/l). *In the absene of unequivoal hyperglyemia, result should be onfirmed by repeat testing. used A1C to predit the progression to diabetes demonstrated a strong, ontinuous assoiation between A1C and subsequent diabetes. In a systemati review of 44,203 individuals from 16 ohort studies with a follow-up interval averaging 5.6 years (range 2.8 12 years), those with an A1C between 5.5 and 6.0% had a substantially inreased risk of diabetes with 5-year inidenes ranging from 9 to 25%. An A1C range of 6.0 6.5% had a 5-year risk of developing diabetes between 25 to 50% and relative risk (RR) 20 times higher ompared with an A1C of 5.0% (15). In a ommunitybased study of blak and white adults without diabetes, baseline A1C was a stronger preditor of subsequent diabetes and ardiovasular events than was fasting gluose (16). Other analyses suggest that an A1C of 5.7% is assoiated with diabetes risk similar to that in the highrisk partiipants in the Diabetes Prevention Program (DPP) (17). Hene, it is reasonable to onsider an A1C range of 5.7 6.4% as identifying individuals with prediabetes. As is the ase for individuals found to have IFG and IGT, individuals with an A1C of 5.7 6.4% should be informed of their inreased risk for diabetes as well as CVD and ounseled about effetive strategies to lower their risks (see Setion IV). As with gluose measurements, the ontinuum of risk is urvilinear, so that as A1C rises, the risk of diabetes rises disproportionately (15). Aordingly, interventions should be most intensive and follow-up partiularly vigilant for those with A1Cs above 6.0%, who should be onsidered to be at very high risk. Table 3 summarizes the ategories of prediabetes. II. TESTING FOR DIABETES IN ASYMPTOMATIC PATIENTS Reommendations Testing to detet type 2 diabetes and prediabetes in asymptomati people should be onsidered in adults of any age who are overweight or obese (BMI $25 kg/m 2 ) and who have one or more additional risk fators for diabetes (Table 4). In those without these risk fators, testing should begin at age 45. (B) If tests are normal, repeat testing at least at 3-year intervals is reasonable. (E) To test for diabetes or prediabetes, the A1C, FPG, or 75-g 2-h OGTT are appropriate. (B) In those identified with prediabetes, identify and, if appropriate, treat other CVD risk fators. (B) Position Statement Table 3dCategories of inreased risk for diabetes (prediabetes)* FPG 100 mg/dl (5.6 mmol/l) to 125 mg/dl (6.9 mmol/l) (IFG) OR 2-h plasma gluose in the 75-g OGTT 140 mg/dl (7.8 mmol/l) to 199 mg/dl (11.0 mmol/l) (IGT) OR A1C 5.7 6.4% *For all three tests, risk is ontinuous, extending below the lower limit of the range and beoming disproportionately greater at higher ends of the range. For many illnesses, there is a major distintion between sreening and diagnosti testing. However, for diabetes, the same tests would be used for sreening as for diagnosis. Diabetes may be identified anywhere along a spetrum of linial senarios ranging from a seemingly low-risk individual who happens to have gluose testing, to a higher-risk individual whom the provider tests beause of high suspiion of diabetes, to the symptomati patient. The disussion herein is primarily framed as testing for diabetes in those without symptoms. The same assays used for testing for diabetes will also detet individuals with prediabetes. A. Testing for type 2 diabetes and risk of future diabetes in adults Prediabetes and diabetes meet established riteria for onditions in whih early detetion is appropriate. Both onditions are ommon, inreasing in prevalene, and impose signifiantpublihealth burdens. There is a long presymptomati phase before the diagnosis of type 2 diabetes is usually made. Relatively simple tests are available to detet prelinial disease. Additionally, the duration of glyemi burden is a strong preditor of adverse outomes, and effetive interventions exist to prevent progression of prediabetes to diabetes (see Setion IV) and to redue risk of ompliations of diabetes (see Setion VI). Type 2 diabetes is frequently not diagnosed until ompliations appear, and approximately one-fourth of all people with diabetes in the U.S. may be undiagnosed. The effetiveness of early identifiation of prediabetes and diabetes through mass testing of asymptomati individuals has not been proven definitively, and rigorous trials to provide suh proof are unlikely to our. In a large randomized ontrolled trial (RCT) in Europe, general pratie patients between the ages of 40 69 years were sreened for diabetes and are.diabetesjournals.org DIABETES CARE, VOLUME 36, SUPPLEMENT 1, JANUARY 2013 S13

then randomly assigned by pratie to routine are of diabetes or intensive treatment of multiple risk fators. After 5.3 years of follow-up, CVD risk fators were modestly but signifiantly more improved with intensive treatment. Inidene of first CVD event and mortality rates were not signifiantly different between groups (18). This study would seem to add support for early treatment of sreen-deteted diabetes, as risk fator ontrol was exellent even in the routine treatment arm and both groups had lower event rates than predited. The absene of a ontrol unsreened arm limits the ability to definitely prove that sreening impats outomes. Mathematial modeling studies suggest that sreening independent of risk fators beginning at age 30 years or age 45 years is highly ost-effetive (,$11,000 per quality-adjusted lifeyear gained) (19). Reommendations for testing for diabetes in asymptomati, undiagnosed adults are listed in Table 4. Testing should be onsidered in adults of any age with BMI $25 kg/m 2 and one or more of the known risk fators for diabetes. In addition to the listed risk fators, ertain mediations, suh as gluoortioids and antipsyhotis (20), are known to inrease the risk of type 2 diabetes. There is ompelling evidene that lower BMI ut points suggest diabetes risk in some raial and ethni groups. In a large multiethni ohort study, for an equivalent inidene rate of diabetes onferred by a BMI of 30 kg/m 2 in whites, the BMI utoff value was 24 kg/m 2 in South Asians, 25 kg/m 2 in Chinese, and 26 kg/m 2 in Afrian Amerians (21). Disparities in sreening rates, not explainable by insurane status, are highlighted by evidene that despite muh higher prevalene of type 2 diabetes, non-cauasians in an insured population are no more likely than Cauasians to be sreened for diabetes (22). Beause ageisamajorriskfatorfordiabetes,testing of those without other risk fators should begin no later than age 45 years. The A1C, FPG, or the 2-h OGTT are appropriate for testing. It should be noted that the tests do not neessarily detet diabetes in the same individuals. The effiay of interventions for primary prevention of type 2 diabetes (23 29) has primarily been demonstrated among individuals with IGT, not for individuals with isolated IFG or for individuals with speifi A1C levels. The appropriate interval between testsisnotknown(30).therationale for the 3-year interval is that false negatives will be repeated before substantial time elapses, and there is little likelihood that an individual will develop signifiant ompliations of diabetes within 3 years of a negative test result. In the modeling study, repeat sreening every 3 or 5 years was ost-effetive (19). Beause of the need for follow-up and disussion of abnormal results, testing should be arried out within the health are setting. Community sreening outside a health are setting is not reommended beause people with positive tests may not Table 4dCriteria for testing for diabetes in asymptomati adult individuals 1. Testing should be onsidered in all adults who are overweight (BMI $25 kg/m 2 *) and have additional risk fators: physial inativity first-degree relative with diabetes high-risk rae/ethniity (e.g., Afrian Amerian, Latino, Native Amerian, Asian Amerian, Paifi Islander) women who delivered a baby weighing.9 lb or were diagnosed with GDM hypertension ($140/90 mmhg or on therapy for hypertension) HDL holesterol level,35 mg/dl (0.90 mmol/l) and/or a triglyeride level.250 mg/dl (2.82 mmol/l) women with polyysti ovary syndrome A1C $5.7%, IGT, or IFG on previous testing other linial onditions assoiated with insulin resistane (e.g., severe obesity, aanthosis nigrians) history of CVD 2. In the absene of the above riteria, testing for diabetes should begin at age 45 years. 3. If results are normal, testing should be repeated at least at 3-year intervals, with onsideration of more frequent testing depending on initial results (e.g., those with prediabetes should be tested yearly) and risk status. *At-risk BMI may be lower in some ethni groups. seek, or have aess to, appropriate follow-up testing and are. Conversely, there may be failure to ensure appropriate repeat testing for individuals who test negative. Community sreening may also be poorly targeted; i. e.,itmayfailtoreahthegroupsmostatrisk and inappropriately test those at low risk (the worried well) or even those already diagnosed. B. Sreening for type 2 diabetes in hildren Reommendations Testing to detet type 2 diabetes and prediabetesshouldbeonsideredinhildren and adolesents who are overweight and who have two or more additional risk fators for diabetes (Table 5). (E) The inidene of type 2 diabetes in adolesents has inreased dramatially in the last deade, espeially in minority populations (31), although the disease remains rare in the general pediatri population (32). Consistent with reommendations for adults, hildren and youth at inreased risk for the presene or the development of type 2 diabetes should be tested within the health are setting (33). The reommendations of the ADA onsensus statement Type 2 Diabetes in Children and Adolesents, with some modifiations, are summarized in Table 5. C. Sreening for type 1 diabetes Reommendations Consider referring relatives of those with type 1 diabetes for antibody testing for risk assessment in the setting of a linial researh study. (E) Generally, people with type 1 diabetes present with aute symptoms of diabetes and markedly elevated blood gluose levels, and some ases are diagnosed with life-threatening ketoaidosis. Evidene from several studies suggests that measurement of islet autoantibodies in relatives of those with type 1 diabetes identifies individuals who are at risk for developing type 1 diabetes. Suh testing, oupled with eduation about symptoms of diabetes and follow-up in an observational linial study, may allow earlier identifiation of onset of type 1 diabetes and lessen presentation with ketoaidosis at time of diagnosis. This testing may be appropriate in those who have relatives with type 1 diabetes, in the ontext of S14 DIABETES CARE, VOLUME 36, SUPPLEMENT 1, JANUARY 2013 are.diabetesjournals.org

Table 5dTesting for type 2 diabetes in asymptomati hildren* Criteria Overweight (BMI.85th perentile for age and sex, weight for height.85th perentile, or weight.120% of ideal for height) Plus any two of the following risk fators: Family history of type 2 diabetes in first- or seond-degree relative Rae/ethniity (Native Amerian, Afrian Amerian, Latino, Asian Amerian, Paifi Islander) Signs of insulin resistane or onditions assoiated with insulin resistane (aanthosis nigrians, hypertension, dyslipidemia, polyysti ovary syndrome, or small-for-gestationalage birth weight) Maternal history of diabetes or GDM during the hild s gestation Age of initiation: age 10 years or at onset of puberty, if puberty ours at a younger age Frequeny: every 3 years *Persons aged 18 years and younger. linial researh studies (see, for example, http://www.diabetestrialnet.org). However, widespread linial testing of asymptomati low-risk individuals annot urrently be reommended, as it would identify very few individuals in the general population who are at risk. Individuals who sreen positive should be ounseled about their risk of developing diabetes and symptoms of diabetes, followed losely to prevent development of diabeti ketoaidosis, and informed about linial trials. Clinial studies are being onduted to test various methods of preventing type 1 diabetes in those with evidene of autoimmunity. Some interventions have demonstrated modest effiay in slowing b-ell loss early in type 1 diabetes (34,35), and further researh is needed to determine whether they may be effetive in preventing type 1 diabetes. III. DETECTION AND DIAGNOSIS OF GDM Reommendations Sreen for undiagnosed type 2 diabetes at the first prenatal visit in those with risk fators, using standard diagnosti riteria. (B) In pregnant women not previously known to have diabetes, sreen for GDM at 24 28 weeks of gestation, usinga75-g2-hogttandthediagnosti ut points in Table 6. (B) Sreen women with GDM for persistent diabetes at 6 12 weeks postpartum, using the OGTT and nonpregnany diagnosti riteria. (E) Women with a history of GDM should have lifelong sreening for the development of diabetes or prediabetes at least every 3 years. (B) Women with a history of GDM found to have prediabetes should reeive lifestyle interventions or metformin to prevent diabetes. (A) For many years, GDM was defined as any degree of gluose intolerane with onset or first reognition during pregnany (13), whether or not the ondition persisted after pregnany, and not exluding the possibility that unreognized gluose intolerane may have antedated or begun onomitantly with the pregnany. This definition failitated a uniform strategy for detetion and lassifiation of GDM, but its limitations were reognized for many years. As the ongoing epidemi of obesity and diabetes has led to more type 2 diabetes in women of hildbearing age, the number of pregnant women with undiagnosed type 2 diabetes has inreased (36). Beause of this, it is reasonable to Table 6dSreening for and diagnosis of GDM Perform a 75-g OGTT, with plasma gluose measurement fasting and at 1 and 2 h, at 24 28 weeks of gestation in women not previously diagnosed with overt diabetes. The OGTT should be performed in the morning after an overnight fast of at least 8 h. The diagnosis of GDM is made when any of the following plasma gluose values are exeeded: Fasting: $92 mg/dl (5.1 mmol/l) 1h:$180 mg/dl (10.0 mmol/l) 2h:$153 mg/dl (8.5 mmol/l) sreen women with risk fators for type 2 diabetes (Table 4) for diabetes at their initial prenatal visit, using standard diagnosti riteria (Table 2). Women with diabetes found at this visit should reeive a diagnosis of overt, not gestational, diabetes. GDM arries risks for the mother and neonate. The Hyperglyemia and AdversePregnanyOutome(HAPO)study (37), a large-sale (;25,000 pregnant women) multinational epidemiologial study, demonstrated that risk of adverse maternal, fetal, and neonatal outomes ontinuously inreased as a funtion of maternal glyemia at 24 28 weeks, even within ranges previously onsidered normal for pregnany. For most ompliations, there was no threshold for risk. These results have led to areful reonsideration of the diagnosti riteria for GDM. After deliberations in 2008 2009, the International Assoiation of Diabetes and Pregnany Study Groups (IADPSG), an international onsensus groupwithrepresentatives from multiple obstetrial and diabetes organizations, inluding ADA, developed revised reommendations for diagnosing GDM. The group reommended that all women not known to have prior diabetes undergo a 75-g OGTT at 24 28 weeks of gestation. Additionally, the group developed diagnosti ut points for the fasting, 1-h, and 2-h plasma gluose measurements that onveyed an odds ratio for adverse outomes of at least 1.75 ompared with women with the mean gluose levels in the HAPO study. Current sreening and diagnosti strategies, based on the IADPSG statement (38), are outlined in Table 6. These new riteria will signifiantly inrease the prevalene of GDM, primarily beause only one abnormal value, not two, is suffiient to make the diagnosis. The ADA reognizes the antiipated signifiant inrease in the inidene of GDM diagnosed by these riteria and is sensitive to onerns about the medialization of pregnanies previously ategorized as normal. These diagnosti riteria hanges are being made in the ontext of worrisome worldwide inreases in obesity and diabetes rates, with the intent of optimizing gestational outomes for women and their babies. Admittedly, there are few data from randomized linial trials regarding therapeuti interventions in women who will now be diagnosed with GDM based on only one blood gluose value above the are.diabetesjournals.org DIABETES CARE, VOLUME 36, SUPPLEMENT 1, JANUARY 2013 S15

speified ut points (in ontrast to the older riteria that stipulated at least two abnormal values). However, there is emerging observational and retrospetive evidene that women diagnosed with the new riteria (even if they would not have been diagnosed with older riteria) have inreased rates of poor pregnany outomes similar to those of women with GDM by prior riteria (39,40). Expeted benefits to these pregnanies and offspring are inferred from intervention trials that foused on women with more mild hyperglyemia than identified using older GDM diagnosti riteria and that found modest benefits (41,42). The frequeny of followup and blood gluose monitoring for these women is not yet lear, but likely to be less intensive than for women diagnosed by the older riteria. It is important to note that 80 90% of women in both of the mild GDM studies (whose gluose values overlapped with the thresholds reommended herein) ould be managed with lifestyle therapy alone. The Amerian College of Obstetriians and Gyneologists announed in 2011 that they ontinue to reommend use of prior diagnosti riteria for GDM (43). Several other ountries have adopted the new riteria, and a report from the WHO on this topi is pending at the time of publiation of these standards. The National Institutes of Health is planning to hold a onsensus development onferene on this topi in 2013. Beause some ases of GDM may represent pre-existing undiagnosed type 2 diabetes, women with a history of GDM should be sreened for diabetes 6 12 weeks postpartum, using nonpregnant OGTT riteria. Beause of their prepartum treatment for hyperglyemia, use of the A1C for diagnosis of persistent diabetes at the postpartum visit is not reommended (44). Women with a history of GDM have a greatly inreased subsequent risk for diabetes (45) and should befollowedupwithsubsequentsreening for the development of diabetes or prediabetes, as outlined in Setion II. Lifestyle interventions or metformin should be offered to women with a history of GDM who develop prediabetes, as disussed in Setion IV. In the prospetive Nurses Health Study II, risk of subsequent diabetes after a history of GDM was signifiantly lower in women who followed healthy eating patterns. Adjusting for BMI moderately, but not ompletely, attenuated this assoiation (46). IV. PREVENTION/DELAY OF TYPE 2 DIABETES Reommendations Patients with IGT (A), IFG (E), or an A1C of 5.7 6.4% (E) should be referred to an effetive ongoing support program targeting weight loss of 7% of body weight and inreasing physial ativity to at least 150 min/week of moderate ativity suh as walking. Follow-up ounseling appears to be important for suess. (B) Based on the ost-effetiveness of diabetes prevention, suh programs should be overed by third-party payers. (B) Metformin therapy for prevention of type 2 diabetes may be onsidered in those with IGT (A), IFG (E), or an A1C of 5.7 6.4% (E), espeially for those with BMI.35 kg/m 2, aged,60 years, and women with prior GDM. (A) At least annual monitoring for the development of diabetes in those with prediabetes is suggested. (E) Sreening for and treatment of modifiable risk fators for CVD is suggested. (B) RCTs have shown that individuals at high risk for developing type 2 diabetes (those with IFG, IGT, or both) an signifiantly derease the rate of onset of diabetes with partiular interventions (23 29). These inlude intensive lifestyle modifiation programs that have been shown to be very effetive (;58% redution after 3 years) and use of the pharmaologial agents metformin, a-gluosidase inhibitors, orlistat, and thiazolidinediones, eah of whih has been shown to derease inident diabetes to various degrees. Follow-up of all three large studies of lifestyle intervention has shown sustained redution in the rate of onversion to type 2 diabetes, with 43% redution at 20 years in the Da Qing study (47), 43% redution at 7 years in the Finnish Diabetes Prevention Study (DPS) (48), and 34% redution at 10 years in the U.S. Diabetes Prevention Program Outomes Study (DPPOS) (49). A osteffetiveness model suggested that lifestyle interventions as delivered in the DPP are ost-effetive (50), and atual ost data from the DPP and DPPOS onfirm that lifestyle interventions are highly ost-effetive (51). Group delivery of the DPP intervention in ommunity settings has the potential to be signifiantly less expensive while still ahieving similar weight loss (52). Based on the results of linial trials and the known risks of progression of prediabetes to diabetes, persons with an A1C of 5.7 6.4%, IGT, or IFG should be ounseled on lifestyle hanges with goals similar to those of the DPP (7% weight loss and moderate physial ativity of at least 150 min/week). Regarding drug therapy for diabetes prevention, metformin has a strong evidene base and demonstrated long-term safety (53). For other drugs, issues of ost, side effets, and lak of persistene of effet in some studies (54) require onsideration. Metformin was less effetive than lifestyle modifiation in the DPP and DPPOS, but may be ost-saving over a 10-year period (51). It was as effetive as lifestyle modifiation in partiipants with a BMI of at least 35 kg/m 2, but not signifiantly better than plaebo than those over age 60 years (23). In women in the DPP with a history of GDM, metformin and intensive lifestyle modifiation led to an equivalent 50% redution in the risk of diabetes (55). Metformin therefore might reasonably be reommended for very high-risk individuals (those with a history of GDM, the very obese, and/or those with more severe or progressive hyperglyemia). People with prediabetes often have other ardiovasular risk fators, suh as obesity, hypertension, and dyslipidemia. Assessing and treating these risk fators is an important aspet of reduing ardiometaboli risk. In the DPP and DPPOS, ardiovasular event rates have been very low, perhaps due to appropriate management of ardiovasular risk fators in all arms of the study (56). V. DIABETES CARE A. Initial evaluation A omplete medial evaluation should be performed to lassify the diabetes, detet the presene of diabetes ompliations, review previous treatment and risk fator ontrol in patients with established diabetes, assist in formulating a management plan, and provide a basis for ontinuing are. Laboratory tests appropriate to the evaluation of eah patient s medial ondition should be performed. A fous on the omponents of omprehensive are (Table 7) will assist the health are team to ensure optimal management of the patient with diabetes. B. Management People with diabetes should reeive medial are from a team that may inlude physiians, nurse pratitioners, physiian s assistants, nurses, dietitians, pharmaists, and mental health professionals with S16 DIABETES CARE, VOLUME 36, SUPPLEMENT 1, JANUARY 2013 are.diabetesjournals.org

Table 7dComponents of the omprehensive diabetes evaluation Medial history Age and harateristis of onset of diabetes (e.g., DKA, asymptomati laboratory finding) Eating patterns, physial ativity habits, nutritional status, and weight history; growth and development in hildren and adolesents Diabetes eduation history Review of previous treatment regimens and response to therapy (A1C reords) Current treatment of diabetes, inluding mediations, mediation adherene and barriers thereto, meal plan, physial ativity patterns, and readiness for behavior hange Results of gluose monitoring and patient s use of data DKA frequeny, severity, and ause Hypoglyemi episodes Hypoglyemia awareness Any severe hypoglyemia: frequeny and ause History of diabetes-related ompliations Mirovasular: retinopathy, nephropathy, neuropathy (sensory, inluding history of foot lesions; autonomi, inluding sexual dysfuntion and gastroparesis) Marovasular: CHD, erebrovasular disease, and PAD Other: psyhosoial problems*, dental disease* Physial examination Height, weight, BMI Blood pressure determination, inluding orthostati measurements when indiated Fundosopi examination* Thyroid palpation Skin examination (for aanthosis nigrians and insulin injetion sites) Comprehensive foot examination Inspetion Palpation of dorsalis pedis and posterior tibial pulses Presene/absene of patellar and Ahilles reflexes Determination of proprioeption, vibration, and monofilament sensation Laboratory evaluation A1C, if results not available within past 2 3 months If not performed/available within past year Fasting lipid profile, inluding total, LDL and HDL holesterol and triglyerides Liver funtion tests Test for urine albumin exretion with spot urine albumin-to-reatinine ratio Serum reatinine and alulated GFR TSH in type 1 diabetes, dyslipidemia or women over age 50 years Referrals Eye are professional for annual dilated eye exam Family planning for women of reprodutive age Registered dietitian for MNT DSME Dentist for omprehensive periodontal examination Mental health professional, if needed *See appropriate referrals for these ategories. expertise and a speial interest in diabetes. It is essential in this ollaborative and integrated team approah that individuals with diabetes assume an ative role in their are. The management plan should be formulated as a ollaborative therapeuti alliane among the patient and family, the physiian, and other members of the health are team. A variety of strategies and tehniques should be used to provide adequate eduation and development of problem-solving skills in the various aspets of diabetes management. Implementation of the management plan requires that the goals and treatment plan are individualized and take patient preferenes into aount. The management plan should reognize diabetes selfmanagement eduation (DSME) and ongoing diabetes support as an integral omponent of are. In developing the plan, onsideration should be given to the patient s age, shool or work shedule and onditions, physial ativity, eating patterns, soial situation and ultural fators, and presene of ompliations of diabetes or other medial onditions. C. Glyemi ontrol 1. Assessment of glyemi ontrol Two primary tehniques are available for health providers and patients to assess the effetiveness of the management plan on glyemi ontrol: patient self-monitoring of blood gluose (SMBG) or interstitial gluose, and A1C. a. Gluose monitoring Reommendations Patients on multiple-dose insulin (MDI) or insulin pump therapy should do SMBG at least prior to meals and snaks, oasionally postprandially, at bedtime, prior to exerise, when they suspet low blood gluose, after treating low blood gluose until they are normoglyemi, and prior to ritial tasks suh as driving. (B) When presribed as part of a broader eduational ontext, SMBG results may be helpful to guide treatment deisions and/or patient self-management for patients using less frequent insulin injetions or noninsulin therapies. (E) When presribing SMBG, ensure that patients reeive ongoing instrution and regular evaluation of SMBG tehnique and SMBG results, as well as their ability to use SMBG data to adjust therapy. (E) Continuous gluose monitoring (CGM) in onjuntion with intensive insulin regimens an be a useful tool to lower A1C in seleted adults (aged $25 years) with type 1 diabetes. (A) Although the evidene for A1C lowering is less strong in hildren, teens, and younger adults, CGM may be helpful in these groups. Suess orrelates with adherene to ongoing use of the devie. (C) CGM may be a supplemental tool to SMBG in those with hypoglyemia unawareness and/or frequent hypoglyemi episodes. (E) Major linial trials of insulin-treated patients that demonstrated the benefits of intensive glyemi ontrol on diabetes ompliations have inluded SMBG as part of multifatorial interventions, suggesting that SMBG is a omponent of effetive therapy. SMBG allows patients to evaluate their individual response to are.diabetesjournals.org DIABETES CARE, VOLUME 36, SUPPLEMENT 1, JANUARY 2013 S17

therapy and assess whether glyemi targets are being ahieved. Results of SMBG an be useful in preventing hypoglyemia and adjusting mediations (partiularly prandial insulin doses), medial nutrition therapy (MNT), and physial ativity. The frequeny and timing of SMBG should be ditated by the partiular needs and goals of the patient. SMBG is espeially important for patients treated with insulin to monitor for and prevent asymptomati hypoglyemia and hyperglyemia. Most patients with type 1 diabetes and others on intensive insulin regimens (MDI or insulin pump therapy) should do SMBG at least prior to meals and snaks, oasionally postprandially, at bedtime, prior to exerise, when they suspet low blood gluose, after treating low blood gluose until they are normoglyemi, and prior to ritial tasks suh as driving. For many patients, this will require testing 6 8 times daily, although individual needs may be greater. Although there are few rigorous studies, a database study of almost 27,000 hildren and adolesents with type 1 diabetes showed that, after adjustment for multiple onfounders, inreased daily frequeny of SMBG was signifiantly assoiated with lower A1C (20.2% per additional test per day, leveling off at five tests per day) and with fewer aute ompliations (57). The optimal frequeny of SMBG for patients on nonintensive regimens, suh as those with type 2 diabetes on basal insulin, is not known, although a number of studies have used fasting SMBG for patient or provider titration of the basal insulin dose. The evidene base for SMBG for patients with type 2 diabetes on noninsulin therapy is somewhat mixed. Several randomized trials have alled into question the linial utility and ost-effetiveness of routine SMBG in non insulin-treated patients (58 60). A reent meta-analysis suggested that SMBG redued A1C by 0.25% at 6 months (61), while a Cohrane review onluded that the overall effet of SMBG in suh patients is small up to 6 months after initiation and subsides after 12 months (62). Beause the auray of SMBG is instrument and user dependent (63), it is important to evaluate eah patient s monitoring tehnique, both initially and at regular intervals thereafter. Optimal use of SMBG requires proper review and interpretation of the data, both by the patient and provider. Among patients who heked their blood gluose at least one daily, many reported taking no ation when results were high or low (64). In one study of insulin-naïve patients with suboptimal initial glyemi ontrol, use of strutured SMBG (a paper tool to ollet and interpret 7-point SMBG profiles over 3 days at least quarterly) redued A1C by 0.3% more than in an ative ontrol group (65). Patients should be taught how to use SMBG data to adjust food intake, exerise, or pharmaologial therapy to ahieve speifi goals, and the ongoing need for and frequeny of SMBG should be re-evaluated at eah routine visit. Real-time CGM through the measurement of interstitial gluose (whih orrelates well with plasma gluose) is available. These sensors require alibration with SMBG, and the latter are still reommended for making aute treatment deisions. CGM devies have alarms for hypo- and hyperglyemi exursions. A 26-week randomized trial of 322 type 1 diabeti patients showed that adults aged $25 years using intensive insulin therapy and CGM experiened a 0.5% redution in A1C (from ;7.6 to 7.1%) ompared with usual intensive insulin therapy with SMBG (66). Sensor use in hildren, teens, and adults to age 24 years did not result in signifiant A1C lowering, and there was no signifiant differene in hypoglyemia in any group. Importantly, the greatest preditor of A1C lowering in this study for all age-groups was frequeny of sensor use, whih was lower in younger age-groups. In a smaller RCT of 129 adults and hildren with baseline A1C,7.0%, outomes ombining A1C and hypoglyemia favored the group utilizing CGM, suggesting that CGM is also benefiial for individuals with type 1 diabetes who have already ahieved exellent ontrol (67). A trial omparing CGM plus insulin pump to SMBG plus multiple injetions of insulin in adults and hildren with type 1 diabetes showed signifiantly greater improvements in A1C with sensoraugmented pump therapy (68,69), but this trial did not isolate the effet of CGM itself. Overall, meta-analyses suggest that ompared with SMBG, CGM lowers A1C by ;0.26% (70). Altogether, these data suggest that, in appropriately seleted patients who are motivated to wear it most of the time, CGM redues A1C. The tehnology may be partiularly useful in those with hypoglyemia unawareness and/or frequent episodes of hypoglyemia, although studies as yet have not shown signifiant redutions in severe hypoglyemia (70). CGM forms the underpinning for the development of pumps that suspend insulin delivery when hypoglyemia is developing and for the burgeoning work on artifiial panreas systems. b. A1C Reommendations Perform the A1C test at least two times a year in patients who are meeting treatment goals (and who have stable glyemi ontrol). (E) Perform the A1C test quarterly in patients whose therapy has hanged or who are not meeting glyemi goals. (E) Use of POC testing for A1C provides the opportunity for more timely treatment hanges. (E) Beause A1C is thought to reflet average glyemia over several months (63) and has strong preditive value for diabetes ompliations (71,72), A1C testing should be performed routinely in all patients with diabetes, at initial assessment and then as part of ontinuing are. Measurement approximately every 3 months determines whether patient s glyemi targets have been reahed and maintained. For any individual patient, the frequeny of A1C testing should be dependent on the linial situation, the treatment regimen used, and the judgment of the liniian. Some patients with stable glyemia well within target may do well with testing only twie per year, while unstable or highly intensively managed patients (e.g., pregnant type 1 diabeti women) may be tested more frequently than every 3 months. The availability of the A1C result at the time that the patient is seen (POC testing) has been reported in small studies to result in inreased intensifiation of therapy and improvement in glyemi ontrol (73,74). However, two reent systemati reviews and meta-analyses found no signifiant differene in A1C between POC and laboratory A1C usage (75,76). The A1C test is subjet to ertain limitations. Conditions that affet erythroyte turnover (hemolysis, blood loss) and hemoglobin variants must be onsidered, partiularly when the A1C result does not orrelate with the patient s linial situation (63). In addition, A1C does not provide a measure of glyemi variability or hypoglyemia. For patients prone to glyemi variability (espeially type 1 diabeti patients or type 2 diabeti patients with severe insulin defiieny), glyemi ontrol is best judged by the ombination of results of self-monitoring and the A1C. The A1C may also serve as a hek on the auray of the patient s meter (or the patient s reported SMBG S18 DIABETES CARE, VOLUME 36, SUPPLEMENT 1, JANUARY 2013 are.diabetesjournals.org

results) and the adequay of the SMBG testing shedule. Table 8 ontains the orrelation between A1C levels and mean plasma gluose levels based on data from the international A1C-Derived Average Gluose (ADAG) trial utilizing frequent SMBG and CGM in 507 adults (83% Cauasian) with type 1, type 2, and no diabetes (77). The ADA and the Amerian Assoiation for Clinial Chemistry have determined that the orrelation (r 5 0.92) is strong enough to justify reporting both an A1C result and an estimated average gluose (eag) result when a liniian orders the A1C test. The table in pre- 2009 versions of the Standards of Medial Care in Diabetes desribing the orrelation between A1C and mean gluose was derived from relatively sparse data (one 7-point profile over 1 day per A1C reading) in the primarily Cauasian type 1 diabeti partiipants in the DCCT (78). Cliniians should note that the numbers in the table are now different, as they are based on ;2,800 readings per A1C in the ADAG trial. In the ADAG trial, there were no signifiant differenes among raial and ethni groups in the regression lines between A1C and mean gluose, although there was a trend toward a differene between Afrian/Afrian Amerian partiipants and Cauasian ones. A small study omparing A1C to CGM data in type 1 diabeti hildren found a highly statistially signifiant orrelation between A1C and mean blood gluose, although the orrelation (r 5 0.7) was signifiantly lower than in the ADAG trial (79). Whether Table 8dCorrelation of A1C with average gluose A1C (%) Mean plasma gluose mg/dl mmol/l 6 126 7.0 7 154 8.6 8 183 10.2 9 212 11.8 10 240 13.4 11 269 14.9 12 298 16.5 These estimates are based on ADAG data of ;2,700 gluose measurements over 3 months per A1C measurement in 507 adults with type 1, type 2, and no diabetes. The orrelation between A1C and average gluose was 0.92 (ref. 77). A alulator for onverting A1C results into eag, in either mg/dl or mmol/l, is available at http://professional.diabetes.org/eag. there are signifiant differenes in how A1C relates to average gluose in hildren or in Afrian Amerian patients is an area for further study. For the time being, the question has not led to different reommendations about testing A1C or to different interpretations of the linial meaning of given levels of A1C in those populations. For patients in whom A1C/eAG and measured blood gluose appear disrepant, liniians should onsider the possibilities of hemoglobinopathy or altered red ell turnover, and the options of more frequent and/or different timing of SMBG or use of CGM. Other measures of hroni glyemia suh as frutosamine are available, but their linkage to average gluose and their prognosti signifiane are not as lear as is the ase for A1C. 2. Glyemi goals in adults Reommendations Lowering A1C to below or around 7% has been shown to redue mirovasular ompliations of diabetes and if implemented soon after the diagnosis of diabetes is assoiated with long-term redution in marovasular disease. Therefore, a reasonable A1C goal for many nonpregnant adults is,7%. (B) Providers might reasonably suggest more stringent A1C goals (suh as,6.5%) for seleted individual patients, if this an be ahieved without signifiant hypoglyemia or other adverse effets of treatment. Appropriate patients might inlude those with short duration of diabetes, long life expetany, and no signifiant CVD. (C) Less stringent A1C goals (suh as,8%) may be appropriate for patients with a history of severe hypoglyemia, limited life expetany, advaned mirovasular or marovasular ompliations, extensive omorbid onditions, and those with long-standing diabetes in whom the general goal is diffiult to attain despite DSME, appropriate gluose monitoring, and effetive doses of multiple gluose-lowering agents inluding insulin. (B) Hyperglyemia defines diabetes, and glyemi ontrol is fundamental to the management of diabetes. The DCCT (71), a prospetive RCT of intensive versus standard glyemi ontrol in patients with relatively reently diagnosed type 1 diabetes, showed definitively that improved glyemi ontrol is assoiated with signifiantly dereased rates of Position Statement mirovasular (retinopathy and nephropathy) and neuropathi ompliations. Follow-up of the DCCT ohorts in the Epidemiology of Diabetes Interventions and Compliations (EDIC) study (80,81) demonstrated persistene of these mirovasular benefits in previously intensively treated subjets, even though their glyemi ontrol approximated that of previous standard arm subjets during follow-up. The Kumamoto Study (82) and UK Prospetive Diabetes Study (UKPDS) (83,84) onfirmed that intensive glyemi ontrol was assoiated with signifiantly dereased rates of mirovasular and neuropathi ompliations in patients with type 2 diabetes. Long-term follow-up of the UKPDS ohorts showed persistene of the effet of early glyemi ontrol on most mirovasular ompliations (85). Subsequent trials in patients with more long-standing type 2 diabetes, designed primarily to look at the role of intensive glyemi ontrol on ardiovasular outomes, also onfirmed a benefit, although more modest, on onset or progression of mirovasular ompliations. The Veterans Affairs Diabetes Trial (VADT) showed signifiant redutions in albuminuria with intensive (ahieved median A1C 6.9%) ompared with standard glyemi ontrol, but no differene in retinopathy and neuropathy (86,87). The Ation in Diabetes and Vasular Disease: Preterax and Diamiron MR Controlled Evaluation (ADVANCE) study of intensive versus standard glyemi ontrol in type 2 diabetes found a statistially signifiant redution in albuminuria, but not in neuropathy or retinopathy, with an A1C target of,6.5% (ahieved median A1C 6.3%) ompared with standard therapy ahieving a median A1C of 7.0% (88). Analyses from the Ation to Control Cardiovasular Risk in Diabetes (ACCORD) trial have shown lower rates of onset or progression of early-stage mirovasular ompliations in the intensive glyemi ontrol arm ompared with the standard arm (89,90). Epidemiologial analyses of the DCCT and UKPDS (71,72) demonstrate a urvilinear relationship between A1C and mirovasular ompliations. Suh analyses suggest that, on a population level, the greatest number of ompliations will be averted by taking patients from very poor ontrol to fair or good ontrol. These analyses also suggest that further lowering of A1C from 7 to 6% is assoiated with further redution in the risk of mirovasular ompliations, albeit the absolute risk are.diabetesjournals.org DIABETES CARE, VOLUME 36, SUPPLEMENT 1, JANUARY 2013 S19