Cuidadomdicoendiabetes 2009-091110224108-phpapp01


Published on

Published in: Health & Medicine
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Cuidadomdicoendiabetes 2009-091110224108-phpapp01

  1. 1. Standards of Medical Care in Diabetes—2009AMERICAN DIABETES ASSOCIATIONDiabetes is a chronic illness that re-quires continuing medical careand patient self-management ed-ucation to prevent acute complicationsand to reduce the risk of long-termcomplications. Diabetes care is complexand requires that many issues, beyondglycemic control, be addressed. A largebody of evidence exists that supports arange of interventions to improve dia-betes outcomes.These standards of care are in-tended to provide clinicians, patients,researchers, payors, and other inter-ested individuals with the componentsof diabetes care, treatment goals, andtools to evaluate the quality of care.While individual preferences, comor-bidities, and other patient factors mayrequire modification of goals, targetsthat are desirable for most patients withdiabetes are provided. These standardsare not intended to preclude more ex-tensive evaluation and management ofthe patient by other specialists asneeded. For more detailed information,refer to references 1–3.The recommendations included arescreening, diagnostic, and therapeuticactions that are known or believed tofavorably affect health outcomes of pa-tients with diabetes. A grading system(Table 1), developed by the AmericanDiabetes Association (ADA) and mod-eled after existing methods, was utilizedto clarify and codify the evidence thatforms the basis for the recommenda-tions. The level of evidence that sup-ports each recommendation is listedafter each recommendation using theletters A, B, C, or E.I. CLASSIFICATION ANDDIAGNOSISA. ClassificationIn 1997, ADA issued new diagnostic andclassification criteria (4); in 2003, modi-fications were made regarding the diagno-sis of impaired fasting glucose (5). Theclassification of diabetes includes fourclinical classes:● type 1 diabetes (results from ␤-cell de-struction, usually leading to absoluteinsulin deficiency)● type 2 diabetes (results from a progres-sive insulin secretory defect on thebackground of insulin resistance)● other specific types of diabetes due toother causes, e.g., genetic defects in␤-cell function, genetic defects in insu-lin action, diseases of the exocrine pan-creas (such as cystic fibrosis), and drug-or chemical-induced (such as in thetreatment of AIDS or after organ trans-plantation)● gestational diabetes mellitus (GDM)(diabetes diagnosed during pregnancy)Some patients cannot be clearly classified astype 1 or type 2 diabetes. Clinical presenta-tion and disease progression vary consider-ably in both types of diabetes. Occasionally,patientswhootherwisehavetype2diabetesmay present with ketoacidosis. Similarly,patients with type 1 may have a late onsetand slow (but relentless) progression of dis-ease despite having features of autoimmunedisease. Such difficulties in diagnosis mayoccur in children, adolescents, and adults.The true diagnosis may become more obvi-ous over time.B. Diagnosis of diabetesCurrent criteria for the diagnosis of diabetesinnonpregnantadultsareshowninTable2.Three ways to diagnose diabetes are recom-mended at the time of this statement, andeach must be confirmed on a subsequentday unless unequivocal symptoms of hy-perglycemia are present. Although the 75-goral glucose tolerance test (OGTT) is moresensitive and modestly more specific thanthe fasting plasma glucose (FPG) to diag-nose diabetes, it is poorly reproducible anddifficult to perform in practice. Because ofease of use, acceptability to patients, andlower cost, the FPG has been the preferreddiagnostic test. Though FPG is less sensitivethan the OGTT, the vast majority of peoplewho do not meet diagnostic criteria for dia-betes by FPG but would by OGTT will havean A1C value well under 7.0% (6).Though the OGTT is not recom-mended for routine clinical use, it may beuseful for further evaluation of patients inwhom diabetes is still strongly suspectedbut who have normal FPG or IFG (seeSection I.C).The use of the A1C for the diagnosisof diabetes has previously not been rec-ommended due to lack of global stan-dardization and uncertainty aboutdiagnostic thresholds. However, with aworld-wide move toward a standardizedassay and with increasing observationalevidence about the prognostic signifi-cance of A1C, an Expert Committee onthe Diagnosis of Diabetes was convenedin 2008. This joint committee of ADA, theEuropean Association for the Study of Di-abetes, and the International DiabetesFederation will likely recommend that theA1C become the preferred diagnostic testfor diabetes. Diagnostic cut-points are be-ing discussed at the time of publication ofthis statement. Updated recommenda-tions will be published in Diabetes Careand will be available at Diagnosis of pre-diabetesHyperglycemia not sufficient to meet thediagnostic criteria for diabetes is catego-● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●The recommendations in this article are based on the evidence reviewed in the following publication:Standards of Care for Diabetes (Technical Review). Diabetes Care 17:1514–1522, 1994.Originally approved 1988. Most recent review/revision October 2008.DOI: 10.2337/dc09-S013© 2009 by the American Diabetes Association. Readers may use this article as long as the work is properlycited, the use is educational and not for profit, and the work is not altered. See for details.P O S I T I O N S T A T E M E N TDIABETES CARE, VOLUME 32, SUPPLEMENT 1, JANUARY 2009 S13
  2. 2. rized as either impaired fasting glucose(IFG) or impaired glucose tolerance(IGT), depending on whether it is identi-fied through the FPG or the OGTT:● IFG ϭ FPG 100 mg/dl (5.6 mmol/l) to125 mg/dl (6.9 mmol/l)● IGT ϭ 2-h plasma glucose 140 mg/dl(7.8 mmol/l) to 199 mg/dl (11.0mmol/l)IFG and IGT have been officially termed“pre-diabetes.” Both categories of pre-diabetes are risk factors for future diabetesand for cardiovascular disease (CVD) (7).II. TESTING FOR PRE-DIABETES AND DIABETESIN ASYMPTOMATICPATIENTSRecommendations● Testing to detect pre-diabetes and type2 diabetes in asymptomatic peopleshould be considered in adults of anyage who are overweight or obese (BMIՆ25 kg/m2) and who have one or moreadditional risk factors for diabetes (Ta-ble 3). In those without these risk fac-tors, testing should begin at age 45years. (B)● If tests are normal, repeat testing shouldbe carried out at least at 3-year inter-vals. (E)● To test for pre-diabetes or diabetes, anFPG test or 2-h OGTT (75-g glucoseload) or both are appropriate. (B)● An OGTT may be considered in pa-tients with IFG to better define the riskof diabetes. (E)● In those identified with pre-diabetes,identify and, if appropriate, treat otherCVD risk factors. (B)For many illnesses, there is a major dis-tinction between screening and diagnos-tic testing. However, for diabetes, thesame tests would be used for “screening”as for diagnosis. Type 2 diabetes has along asymptomatic phase and significantclinical risk markers. Diabetes may beidentified anywhere along a spectrum ofclinical scenarios ranging from a seem-ingly low-risk individual who happens tohave glucose testing, to a higher-risk in-dividual whom the provider tests becauseof high suspicion of diabetes, to the symp-tomatic patient. The discussion herein isprimarily framed as testing for diabetes inthose without symptoms. Testing for dia-betes will also detect individuals with pre-diabetes.A. Testing for pre-diabetes and type2 diabetes in adultsType 2 diabetes is frequently not diag-nosed until complications appear, andapproximately one-third of all peoplewith diabetes may be undiagnosed. Al-though the effectiveness of early identifi-cation of pre-diabetes and diabetesthrough mass testing of asymptomatic in-dividuals has not been definitively proven(and rigorous trials to provide such proofare unlikely to occur), pre-diabetes anddiabetes meet established criteria for con-ditions in which early detection is appro-priate. Both conditions are common,increasing in prevalence, and impose sig-nificant public health burdens. There is along presymptomatic phase before the di-agnosis of type 2 diabetes is usually made.Relatively simple tests are available to de-tect preclinical disease (8). Additionally,the duration of glycemic burden is astrong predictor of adverse outcomes,and effective interventions exist to pre-vent progression of pre-diabetes to diabe-tes (see Section IV) and to reduce risk ofcomplications of diabetes (see SectionVI).Recommendations for testing for pre-Table 1—ADA evidence grading system for clinical practice recommendationsLevel ofevidence DescriptionA Clear evidence from well-conducted, generalizable, randomized controlledtrials that are adequately powered, including:● Evidence from a well-conducted multicenter trial● Evidence from a meta-analysis that incorporated quality ratings in theanalysisCompelling nonexperimental evidence, i.e., “all or none” rule developedby the Centre for Evidence-Based Medicine at OxfordSupportive evidence from well-conducted randomized controlled trialsthat are adequately powered, including:● Evidence from a well-conducted trial at one or more institutions● Evidence from a meta-analysis that incorporated quality ratings in theanalysisB Supportive evidence from well-conducted cohort studies, including:● Evidence from a well-conducted prospective cohort study or registry● Evidence from a well-conducted meta-analysis of cohort studiesSupportive evidence from a well-conducted case-control studyC Supportive evidence from poorly controlled or uncontrolled studies● Evidence from randomized clinical trials with one or more major orthree or more minor methodological flaws that could invalidate theresults● Evidence from observational studies with high potential for bias (suchas case series with comparison to historical controls)● Evidence from case series or case reportsConflicting evidence with the weight of evidence supporting therecommendationE Expert consensus or clinical experienceTable 2—Criteria for the diagnosis of diabetes1. FPG Ն126 mg/dl (7.0 mmol/l). Fasting is defined as no caloric intake for atleast 8 h.*OR2. Symptoms of hyperglycemia and a casual (random) plasma glucose Ն200mg/dl (11.1 mmol/l). Casual (random) is defined as any time of day withoutregard to time since last meal. The classic symptoms of hyperglycemiainclude polyuria, polydipsia, and unexplained weight loss.OR3. 2-h plasma glucose Ն200 mg/dl (11.1 mmol/l) during an OGTT. The testshould be performed as described by the World Health Organization usinga glucose load containing the equivalent of 75-g anhydrous glucosedissolved in water.**In the absence of unequivocal hyperglycemia, these criteria should be confirmed by repeat testing on adifferent day (5).Standards of Medical CareS14 DIABETES CARE, VOLUME 32, SUPPLEMENT 1, JANUARY 2009
  3. 3. diabetes and diabetes in asymptomatic,undiagnosed adults are listed in Table 3.Testing should be considered in adults ofany age with BMI Ն25 kg/m2and one ormore risk factors for diabetes. Because ageis a major risk factor for diabetes, testingof those without other risk factors shouldbegin no later than age 45 years.Either FPG testing or the 2-h OGTT isappropriate for testing. The 2-h OGTTidentifies people with either IFG or IGT,and thus, more pre-diabetic people at in-creased risk for the development of dia-betes and CVD. It should be noted thatthe two tests do not necessarily detect thesame pre-diabetic individuals (9). The ef-ficacy of interventions for primary pre-vention of type 2 diabetes (10–16) hasprimarily been demonstrated among in-dividuals with IGT, not individuals withIFG (who do not also have IGT). As notedin the diagnosis section (Section I.B), theFPG test is more convenient, more repro-ducible, less costly, and easier to admin-ister than the 2-h OGTT (4,5). An OGTTmay be useful in patients with IFG to bet-ter define the risk of diabetes.The appropriate interval betweentests is not known (17). The rationale forthe 3-year interval is that false-negativeswill be repeated before substantial timeelapses, and there is little likelihood thatan individual will develop significantcomplications of diabetes within 3 yearsof a negative test result.Because of the need for follow-up anddiscussion of abnormal results, testingshould be carried out within the healthcare setting. Community screening out-side a health care setting is not recom-mended because people with positivetests may not seek, or have access to, ap-propriate follow-up testing and care.Conversely, there may be failure to ensureappropriate repeat testing for individualswho test negative. Community screeningmay also be poorly targeted, i.e., it mayfail to reach the groups most at risk andinappropriately test those at low risk (theworried well) or even those already diag-nosed (18,19).B. Testing for type 2 diabetes inchildrenThe incidence of type 2 diabetes in ado-lescents has increased dramatically in thelast decade, especially in minority popu-lations (20), although the disease remainsrare in the general adolescent population(21). Consistent with recommendationsfor adults, children and youth at in-creased risk for the presence or the devel-opment of type 2 diabetes should betested within the health care setting (22).The recommendations of the ADA con-sensus statement on type 2 diabetes inchildren and youth, with some modifica-tions, are summarized in Table 4.C. Screening for type 1 diabetesGenerally, people with type 1 diabetespresent with acute symptoms of diabetesand markedly elevated blood glucose lev-els, and most cases are diagnosed soonafter the onset of hyperglycemia. How-ever, evidence from type 1 preventionstudies suggests that measurement of isletautoantibodies identifies individuals whoare at risk for developing type 1 diabetes.Such testing may be appropriate in high-risk individuals, such as those with priortransient hyperglycemia or those who haverelatives with type 1 diabetes, in the contextof clinical research studies (see, for exam-ple, clinical testing of asymptom-atic low-risk individuals cannot currentlybe recommended, as it would identify veryfew individuals in the general populationwhoareatrisk.Individualswhoscreenpos-itive should be counseled about their risk ofdeveloping diabetes. Clinical studies are be-ing conducted to test various methods ofpreventing type 1 diabetes, or reversingearly type 1 diabetes, in those with evidenceof autoimmunity.III. DETECTION ANDDIAGNOSIS OF GDMRecommendations● Screen for GDM using risk factor anal-ysis and, if appropriate, use of anOGTT. (C)● Women with GDM should be screenedfor diabetes 6–12 weeks postpartumand should be followed up with subse-quent screening for the development ofdiabetes or pre-diabetes. (E)GDM is defined as any degree of glucoseintolerance with onset or first recognitionduring pregnancy (4). Although mostcases resolve with delivery, the definitionapplies whether or not the condition per-sists after pregnancy and does not excludethe possibility that unrecognized glucoseintolerance may have antedated or begunconcomitantly with the pregnancy. Ap-Table 3—Criteria for testing for pre-diabetes and diabetes in asymptomatic adult individuals1. Testing should be considered in all adults who are overweight (BMI Ն25 kg/m2*) andhave additional risk factors:● physical inactivity● first-degree relative with diabetes● members of a high-risk ethnic population (e.g., African American, Latino, NativeAmerican, Asian American, Pacific Islander)● women who delivered a baby weighing Ͼ9 lb or were diagnosed with GDM● hypertension (Ն140/90 mmHg or on therapy for hypertension)● HDL cholesterol level Ͻ35 mg/dl (0.90 mmol/l) and/or a triglyceride level Ͼ250mg/dl (2.82 mmol/l)● women with polycystic ovarian syndrome (PCOS)● IGT or IFG on previous testing● other clinical conditions associated with insulin resistance (e.g., severe obesity,acanthosis nigricans)● history of CVD2. In the absence of the above criteria, testing for pre-diabetes and diabetes should beginat age 45 years3. If results are normal, testing should be repeated at least at 3-year intervals, withconsideration of more frequent testing depending on initial results and risk status.*At-risk BMI may be lower in some ethnic groups.Table 4—Testing for type 2 diabetes inasymptomatic childrenCriteria:● Overweight (BMI Ͼ85th percentile forage and sex, weight for height Ͼ85thpercentile, or weight Ͼ120% of ideal forheight)Plus any two of the following risk factors:● Family history of type 2 diabetes in first-or second-degree relative● Race/ethnicity (Native American, AfricanAmerican, Latino, Asian American, PacificIslander)● Signs of insulin resistance or conditionsassociated with insulin resistance(acanthosis nigricans, hypertension,dyslipidemia, PCOS, or small-for-gestational-age birthweight)● Maternal history of diabetes or GDMduring the child’s gestationAge of initiation: age 10 years or at onset ofpuberty, if puberty occurs at a youngerageFrequency: every 3 yearsTest: FPG preferredPosition StatementDIABETES CARE, VOLUME 32, SUPPLEMENT 1, JANUARY 2009 S15
  4. 4. proximately 7% of all pregnancies (rang-ing from 1 to 14% depending on thepopulation studied and the diagnostictests employed) are complicated by GDM,resulting in more than 200,000 casesannually.Because of the risks of GDM to themother and neonate, screening and di-agnosis are warranted. The screeningand diagnostic strategies, based on the2004 ADA position statement on gesta-tional diabetes mellitus (23), are outlined inTable 5.Results of the Hyperglycemia and Ad-verse Pregnancy Outcomes study (24), alarge-scale (including ϳ25,000 pregnantwomen) multinational epidemiologicstudy, demonstrated that risk of adversematernal, fetal, and neonatal outcomescontinuously increased as a function ofmaternal glycemia at 24–28 weeks, evenwithin ranges previously considered nor-mal for pregnancy. For most complica-tions, there was no threshold for risk.These results have led to careful reconsid-eration of the diagnostic criteria for GDM.An international group representing mul-tiple obstetrical and diabetes organiza-tions, including ADA, is currentlyworking on consensus toward 1) a world-wide standard for which diagnostic test touse for GDM and 2) rational diagnosticcut points.Because women with a history ofGDM have a greatly increased subse-quent risk for diabetes (25), they shouldbe screened for diabetes 6–12 weekspostpartum, using nonpregnant OGTTcriteria, and should be followed up withsubsequent screening for the develop-ment of diabetes or pre-diabetes, as out-lined in Section II. For information onthe National Diabetes Education Pro-gram (NDEP) campaign to prevent type2 diabetes in women with GDM, go PREVENTION/DELAYOF TYPE 2 DIABETESRecommendations● Patients with IGT (A) or IFG (E) shouldbe referred to an effective ongoing sup-port program for weight loss of 5–10%of body weight and for increasing phys-ical activity to at least 150 min per weekof moderate activity such as walking.● Follow-up counseling appears to be im-portant for success. (B)● Based on potential cost savings of dia-betes prevention, such counselingshould be covered by third-party pay-ors. (E)● In addition to lifestyle counseling, met-formin may be considered in those whoare at very high risk for developing di-abetes (combined IFG and IGT plusother risk factors such as A1C Ͼ6%,hypertension, low HDL cholesterol, el-evated triglycerides, or family history ofdiabetes in a first-degree relative) andwho are obese and under 60 years ofage. (E)● Monitoring for the development of di-abetes in those with pre-diabetesshould be performed every year. (E)Randomized controlled trials have shownthat individuals at high risk for develop-ing diabetes (those with IFG, IGT, orboth) can be given interventions that sig-nificantly decrease the rate of onset of di-abetes (10 –16). These interventionsinclude intensive lifestyle modificationprograms that have been shown to be veryeffective (Ն58% reduction after 3 years)and use of the pharmacologic agents met-formin, acarbose, orlistat, and thiazo-lidinediones (TZDs), each of which hasbeen shown to decrease incident diabetesto various degrees. A summary of majordiabetes prevention trials is shown in Ta-ble 6.Two studies of lifestyle interventionhave shown persistent reduction in therate of conversion to type 2 diabetes with3 (26) to 14 years (27) of postinterventionfollow-up.Based on the results of clinical trialsand the known risks of progression ofpre-diabetes to diabetes, an ADA Consen-sus Development Panel (7) concludedthat persons with pre-diabetes (IGTand/or IFG) should be counseled on life-style changes with goals similar to those ofthe Diabetes Prevention Program (DPP)(5–10% weight loss and moderate physi-cal activity of ϳ30 min per day). Regard-ing the more difficult issue of drugTable 5—Screening for and diagnosis of GDMCarry out GDM risk assessment at the first prenatal visit.Women at very high risk for GDM should be screened for diabetes as soon as possible afterthe confirmation of pregnancy. Criteria for very high risk are:● severe obesity● prior history of GDM or delivery of large-for-gestational-age infant● presence of glycosuria● diagnosis of PCOS● strong family history of type 2 diabetesScreening/diagnosis at this stage of pregnancy should use standard diagnostic testing (Table 2).All women of greater than low risk of GDM, including those above not found to havediabetes early in pregnancy, should undergo GDM testing at 24–28 weeks of gestation.Low risk status, which does not require GDM screening, is defined as women with ALL ofthe following characteristics:● age Ͻ25 years● weight normal before pregnancy● member of an ethnic group with a low prevalence of diabetes● no known diabetes in first-degree relatives● no history of abnormal glucose tolerance● no history of poor obstetrical outcomeTwo approaches may be followed for GDM screening at 24–28 weeks:1. Two-step approach:A. Perform initial screening by measuring plasma or serum glucose 1 h after a 50-g oralglucose load. A glucose threshold after 50-g load of Ն140 mg/dl identifies ϳ80% ofwomen with GDM, while the sensitivity is further increased to ϳ90% by a threshold ofՆ130 mg/dl.B. Perform a diagnostic 100-g OGTT on a separate day in women who exceed thechosen threshold on 50-g screening.2. One-step approach (may be preferred in clinics with high prevalence of GDM): Performa diagnostic 100-g OGTT in all women to be tested at 24–28 weeks.The 100-g OGTT should be performed in the morning after an overnight fast of at least 8 h.To make a diagnosis of GDM, at least two of the following plasma glucose values must be found:Fasting: Ն95 mg/dl1 h: Ն180 mg/dl2 h: Ն155 mg/dl3 h: Ն140 mg/dlStandards of Medical CareS16 DIABETES CARE, VOLUME 32, SUPPLEMENT 1, JANUARY 2009
  5. 5. therapy for diabetes prevention, the con-sensus panel felt that metformin shouldbe the only drug considered for use indiabetes prevention. For other drugs, theissues of cost, side effects, and lack of per-sistence of effect in some studies led thepanel to not recommend their use for di-abetes prevention. Metformin use wasrecommended only for very-high-risk in-dividuals (those with combined IGT andIFG who are obese and under 60 years ofage with at least one other risk factor fordiabetes). In addition, the panel high-lighted the evidence that in the DPP, met-formin was most effective compared tolifestyle in those with BMI of at least 35kg/m2and those under age 60 years.V. DIABETES CAREA. Initial evaluationA complete medical evaluation should beperformed to classify the diabetes, detectthe presence of diabetes complications,review previous treatment and glycemiccontrol in patients with established diabe-tes, assist in formulating a managementplan, and provide a basis for continuingcare. Laboratory tests appropriate to theevaluation of each patient’s medical con-dition should be performed. A focus onthe components of comprehensive care(Table 7) will assist the health care team toensure optimal management of the pa-tient with diabetes.B. ManagementPeople with diabetes should receive med-ical care from a physician-coordinatedteam. Such teams may include, but arenot limited to, physicians, nurse practitio-ners, physician’s assistants, nurses, dieti-tians, pharmacists, and mental healthprofessionals with expertise and a specialinterest in diabetes. It is essential in thiscollaborative and integrated team ap-proach that individuals with diabetes as-sume an active role in their care.The management plan should be for-mulated as an individualized therapeuticalliance among the patient and family, thephysician, and other members of thehealth care team. A variety of strategiesand techniques should be used to provideadequate education and development ofproblem-solving skills in the various as-pects of diabetes management. Imple-mentation of the management planrequires that each aspect is understoodand agreed on by the patient and the careproviders and that the goals and treat-ment plan are reasonable. Any planshould recognize diabetes self-manage-ment education (DSME) as an integralcomponent of care. In developing theplan, consideration should be given to thepatient’s age, school or work scheduleand conditions, physical activity, eatingpatterns, social situation and personality,cultural factors, and presence of compli-cations of diabetes or other medical con-ditions.C. Glycemic control1. Assessment of glycemic controlTwo primary techniques are available forhealth providers and patients to assess theeffectiveness of the management plan onglycemic control: patient self-monitoringof blood glucose (SMBG) or of interstitialglucose and measurement of A1C.a. Glucose monitoringRecommendations● SMBG should be carried out three ormore times daily for patients using mul-tiple insulin injections or insulin pumptherapy. (A)● For patients using less frequent insulininjections, noninsulin therapies, ormedical nutrition therapy (MNT) andphysical activity alone, SMBG may beuseful as a guide to the success of ther-apy. (E)● To achieve postprandial glucose tar-gets, postprandial SMBG may be appro-priate. (E)Table 6—Therapies proven effective in diabetes prevention trialsStudy (ref.) n PopulationMeanage(years)Duration(years)Intervention(daily dose)Conversion incontrol subjects(%/year) Relative riskLifestyleFinnish DPS (11) 522 IGT, BMI Ն25 kg/m255 3.2 Individualdiet/exercise6 0.42 (0.30–0.70)DPP (10) 2,161* IGT, BMI Ն24 kg/m2,FPG Ͼ5.3 mmol/l51 3 Individualdiet/exercise10 0.42 (0.34–0.52)Da Qing (12) 259* IGT (randomizedgroups)45 6 Group diet/exercise 16 0.62 (0.44–0.86)Toranomon study (28) 458 IGT (men), BMI ϭ 24kg/m255 4 Individualdiet/exercise2 0.33 (0.10–1.0)†Indian DPP (16) 269* IGT 46 2.5 Individualdiet/exercise22 0.71 (0.63–0.79)MedicationsDPP (10) 2,155* IGT, BMI Ͼ24 kg/m2,FPG Ͼ5.3 mmol/l51 2.8 Metformin (1,700 mg) 10 0.69 (0.57–0.83)Indian DPP (16) 269* IGT 46 2.5 Metformin (500 mg) 22 0.74 (0.65–0.81)STOP NIDDM (14) 1,419 IGT, FPG Ͼ5.6 mmol/l 54 3.2 Acarbose (300 mg) 13 0.75 (0.63–0.90)XENDOS (29) 3,277 BMI Ͼ30 kg/m243 4 Orlistat (360 mg) 2 0.63 (0.46–0.86)DREAM (15) 5,269 IGT or IFG 55 3.0 Rosiglitazone (8 mg) 9 0.40 (0.35–0.46)*Number of participants in the indicated comparisons, not necessarily in the entire study. †Calculated from information in the article. DPP, Diabetes PreventionProgram; DREAM, Diabetes REduction Assessment with ramipril and rosiglitazone Medication; DPS, Diabetes Prevention Study; STOP NIDDM, Study to PreventNon-Insulin Dependent Diabetes; XENDOS, Xenical in the prevention of Diabetes in Obese Subjects. This table has been reprinted with permission (30) with somemodification.Position StatementDIABETES CARE, VOLUME 32, SUPPLEMENT 1, JANUARY 2009 S17
  6. 6. ● When prescribing SMBG, ensure thatpatients receive initial instruction in,and routine follow-up evaluation of,SMBG technique and their ability to usedata to adjust therapy. (E)● Continuous glucose monitoring (CGM)in conjunction with intensive insulinregimens can be a useful tool to lowerA1C in selected adults (age Ն25 years)with type 1 diabetes (A).● Although the evidence for A1C loweringis less strong in children, teens, andyounger adults, CGM may be helpful inthese groups. Success correlates with ad-herence to ongoing use of the device. (C)● CGM may be a supplemental tool toSMBG in those with hypoglycemia un-awareness and/or frequent hypoglyce-mic episodes. (E)The ADA’s consensus and position state-ments on SMBG provide a comprehensivereview of the subject (31,32). Major clin-ical trials of insulin-treated patients thatdemonstrated the benefits of intensiveglycemic control on diabetes complica-tions have included SMBG as part of mul-tifactorial interventions, suggesting thatSMBG is a component of effective ther-apy. SMBG allows patients to evaluatetheir individual response to therapy andassess whether glycemic targets are beingachieved. Results of SMBG can be usefulin preventing hypoglycemia and adjust-ing medications (particularly prandial in-sulin doses), MNT, and physical activity.The frequency and timing of SMBGshould be dictated by the particular needsand goals of the patients. SMBG is espe-cially important for patients treated withinsulin to monitor for and prevent asymp-tomatic hypoglycemia and hyperglyce-mia. For most patients with type 1diabetes and pregnant women taking in-sulin, SMBG is recommended three ormore times daily. For this population, sig-nificantly more frequent testing may berequired to reach A1C targets safely with-out hypoglycemia. The optimal frequencyand timing of SMBG for patients with type2 diabetes on noninsulin therapy is un-clear. A meta-analysis of SMBG in non–insulin-treated patients with type 2diabetes concluded that some regimen ofSMBG was associated with a reduction inA1C of Ն0.4%. However, many of thestudies in this analysis also included pa-tient education with diet and exercisecounseling and, in some cases, pharma-cologic intervention, making it difficult toassess the contribution of SMBG alone toimproved control (33). Several recent tri-als have called into question the clinicalutility and cost-effectiveness of routineSMBG in non–insulin-treated patients(34–36).BecausetheaccuracyofSMBGisinstru-ment and user dependent (37), it is impor-tant to evaluate each patient’s monitoringtechnique, both initially and at regular in-tervals thereafter. In addition, optimal useof SMBG requires proper interpretation ofthe data. Patients should be taught how touse the data to adjust food intake, exercise,or pharmacological therapy to achieve spe-cific glycemic goals, and these skills shouldbe reevaluated periodically.CGM through the measurement of in-terstitial glucose (which correlates wellwith plasma glucose) is available. Thesesensors require calibration with SMBG,and the latter are still recommended formaking acute treatment decisions. CGMdevices also have alarms for hypo- andhyperglycemic excursions. Small studiesin selected patients with type 1 diabeteshave suggested that CGM use reduces thetime spent in hypo- and hyperglycemicranges and may modestly improve glyce-mic control. A larger 26-week random-ized trial of 322 type 1 patients showedthat adults age 25 years and older usingTable 7—Components of the comprehensive diabetes evaluationMedical history● age and characteristics of onset of diabetes (e.g., DKA, asymptomatic laboratory finding)● eating patterns, physical activity habits, nutritional status, and weight history; growthand development in children and adolescents● diabetes education history● review of previous treatment regimens and response to therapy (A1C records)● current treatment of diabetes, including medications, meal plan, physical activitypatterns, and results of glucose monitoring and patient’s use of data● DKA frequency, severity, and cause● hypoglycemic episodes● hypoglycemia awareness● any severe hypoglycemia: frequency and cause● history of diabetes-related complications● microvascular: retinopathy, nephropathy, neuropathy (sensory, including history offoot lesions; autonomic, including sexual dysfunction and gastroparesis)● macrovascular: CHD, cerebrovascular disease, PAD● other: psychosocial problems,* dental disease*Physical examination● height, weight, BMI● blood pressure determination, including orthostatic measurements when indicated● fundoscopic examination*● thyroid palpation● skin examination (for acanthosis nigricans and insulin injection sites)● comprehensive foot examination:● inspection● palpation of dorsalis pedis and posterior tibial pulses● presence/absence of patellar and Achilles reflexes● determination of proprioception, vibration, and monofilament sensationLaboratory evaluation● A1C, if results not available within past 2–3 monthsIf not performed/available within past year:● fasting lipid profile, including total, LDL- and HDL-cholesterol and triglycerides● liver function tests● test for urine albumin excretion with spot urine albumin/creatinine ratio● serum creatinine and calculated GFR● thyroid-stimulating hormone in type 1 diabetes, dyslipidemia or women over age 50Referrals● annual dilated eye exam● family planning for women of reproductive age● registered dietitian for MNT● diabetes self-management education● dental examination● mental Health professional, if needed*See appropriate referrals for these categories.Standards of Medical CareS18 DIABETES CARE, VOLUME 32, SUPPLEMENT 1, JANUARY 2009
  7. 7. intensive insulin therapy and CGM expe-rienced a 0.5% reduction in A1C (fromϳ7.6 to 7.1%) compared with usual in-tensive insulin therapy with SMBG (38).Sensor use in children, teens, and adultsto age 24 years did not result in significantA1C lowering, and there was no signifi-cant difference in hypoglycemia in anygroup. Importantly, the greatest predictorof A1C lowering in this study for all age-groups was frequency of sensor use,which was lower in younger age-groups.Although CGM is an evolving technology,emerging data suggest that, in appropri-ately selected patients who are motivatedto wear it most of the time, it may offerbenefit. CGM may be particularly usefulin those with hypoglycemia unawarenessand/or frequent episodes of hypoglyce-mia, and studies in this area are ongoing.b. A1CRecommendations● Perform the A1C test at least two timesa year in patients who are meeting treat-ment goals (and who have stable glyce-mic control). (E)● Perform the A1C test quarterly in pa-tients whose therapy has changed orwho are not meeting glycemic goals. (E)● Use of point-of-care testing for A1C al-lows for timely decisions on therapychanges, when needed. (E)Because A1C is thought to reflect averageglycemia over several months (37), andhas strong predictive value for diabetescomplications (10,39), A1C testingshould be performed routinely in all pa-tients with diabetes at initial assessmentand then as part of continuing care. Mea-surement approximately every 3 monthsdetermines whether a patient’s glycemictargets have been reached and main-tained. For any individual patient, the fre-quency of A1C testing should bedependent on the clinical situation, thetreatment regimen used, and the judg-ment of the clinician. Some patients withstable glycemia well within target may dowell with testing only twice per year,while unstable or highly intensively man-aged patients (e.g., pregnant type 1women) may be tested more frequentlythan every 3 months. The availability ofthe A1C result at the time that the patientis seen (point-of-care testing) has been re-ported to result in increased intensifica-tion of therapy and improvement inglycemic control (40,41).The A1C test is subject to certain lim-itations. Conditions that affect erythro-cyte turnover (hemolysis, blood loss) andhemoglobin variants must be considered,particularly when the A1C result does notcorrelate with the patient’s clinical situa-tion (37). In addition, A1C does not pro-vide a measure of glycemic variability orhypoglycemia. For patients prone to gly-cemic variability (especially type 1 pa-tients, or type 2 patients with severeinsulin deficiency), glycemic control isbest judged by the combination of resultsof SMBG testing and the A1C. The A1Cmay also serve as a check on the accuracyof the patient’s meter (or the patient’s re-ported SMBG results) and the adequacy ofthe SMBG testing schedule.Table 8 contains the correlation be-tween A1C levels and mean plasma glu-cose levels based on data from theinternational A1C-Derived Average Glu-cose (ADAG) trial utilizing frequentSMBG and continuous glucose monitor-ing in 507 adults (83% Caucasian) withtype 1, type 2, and no diabetes (49) TheADA and American Association of Clini-cal Chemists have determined that thecorrelation (r ϭ 0.92) is strong enough tojustify reporting both an A1C result andan estimated average glucose (eAG) resultwhen a clinician orders the A1C test. Thetable in previous versions of the Stan-dards of Medical Care in Diabetes describ-ing the correlation between A1C andmean glucose was derived from relativelysparse data (one seven-point profile over1 day per A1C reading) in the primarilyCaucasian type 1 participants in the Dia-betes Control and Complications Trial(DCCT) trial (43). Clinicians should notethat the numbers in the table are now dif-ferent, as they are based on ϳ2,800 read-ings per A1C in the ADAG trial.In the ADAG study, there were no sig-nificant differences among racial and eth-nic groups in the regression lines betweenA1C and mean glucose, although therewas a trend toward a difference betweenAfrican/African-American and Caucasianparticipants’ regression lines that mighthave been significant had more African/African-American participants been stud-ied. A recent study comparing A1C toCGM data in 48 type 1 children found ahighly statistically significant correlationbetween A1C and mean blood glucose,although the correlation (r ϭ 0.7) was sig-nificantly lower than in the ADAG trial(44). Whether there are significant differ-ences in how A1C relates to average glu-cose in children or in African-Americanpatients is an area for further study. Forthe time being, the question has not led todifferent recommendations about testingA1C or to different interpretations of theclinical meaning of given levels of A1C inthose populations.For patients in whom A1C/eAG andmeasured blood glucose appear discrep-ant, clinicians should consider the possi-bilities of hemoglobinopathy or alteredred cell turnover and the options of morefrequent and/or different timing of SMBGor use of CGM. Other measures of chronicglycemia such as fructosamine are avail-able, but their linkage to average glucoseand their prognostic significance are notas clear as is the case for A1C.2. Glycemic goals in adults● Lowering A1C to below or around 7%has been shown to reduce microvascu-lar and neuropathic complications oftype 1 and type 2 diabetes. Therefore,for microvascular disease prevention,the A1C goal for nonpregnant adults ingeneral is Ͻ7%. (A)● In type 1 and type 2 diabetes, random-ized controlled trials of intensive versusstandard glycemic control have notshown a significant reduction in CVDoutcomes during the randomized por-tion of the trials. Long-term follow-upof the DCCT and UK Prospective Dia-betes Study (UKPDS) cohorts suggeststhat treatment to A1C targets below oraround 7% in the years soon after thediagnosis of diabetes is associated withlong-term reduction in risk of macro-vascular disease. Until more evidencebecomes available, the general goal ofϽ7% appears reasonable for manyTable 8—Correlation of A1C with averageglucoseA1C (%)Mean plasma glucosemg/dl mmol/l6 126 7.07 154 8.68 183 10.29 212 11.810 240 13.411 269 14.912 298 16.5Estimates based on ADAG data of ϳ2,700 glucosemeasurements over 3 months per A1C measure-ment in 507 adults with type 1, type 2, and no dia-betes. Correlation between A1C and averageglucose: 0.92 (42). A calculator for converting A1Cresults into eAG, in either mg/dl or mmol/l, is avail-able at StatementDIABETES CARE, VOLUME 32, SUPPLEMENT 1, JANUARY 2009 S19
  8. 8. adults for macrovascular risk reduc-tion. (B)● Subgroup analyses of clinical trials suchas the DCCT and UKPDS and the mi-crovascular evidence from the Action inDiabetes and Vascular Disease: Preteraxand Diamicron MR Controlled Evalua-tion (ADVANCE) trial suggest a smallbut incremental benefit in microvascu-lar outcomes with A1C values closer tonormal. Therefore, for selected individ-ual patients, providers might reason-ably suggest even lower A1C goals thanthe general goal of Ͻ7%, if this can beachieved without significant hypogly-cemia or other adverse effects of treat-ment. Such patients might includethose with short duration of diabetes,long life expectancy, and no significantCVD. (B)● Conversely, less stringent A1C goalsthan the general goal of Ͻ7% may beappropriate for patients with a historyof severe hypoglycemia, limited life ex-pectancy, advanced microvascular ormacrovascular complications, exten-sive comorbid conditions, and thosewith longstanding diabetes in whomthe general goal is difficult to attain de-spite DSME, appropriate glucose mon-itoring, and effective doses of multipleglucose-lowering agents including in-sulin. (C)Glycemic control is fundamental to themanagement of diabetes. The DCCT, aprospective, randomized, controlled trialof intensive versus standard glycemiccontrol in patients with relatively recentlydiagnosed type 1 diabetes, showed defin-itively that improved glycemic control isassociated with significantly decreasedrates of microvascular (retinopathy andnephropathy) as well as neuropathiccomplications (45). Follow-up of theDCCT cohorts in the Epidemiology of Di-abetes Interventions and Complications(EDIC) study has shown persistence ofthis effect in previously intensivelytreated subjects, even though their glyce-mic control has been equivalent to that ofprevious standard arm subjects duringfollow-up (46,47).In type 2 diabetes, the Kumamotostudy (48) and the UKPDS (49,50) dem-onstrated significant reductions in micro-vascular and neuropathic complicationswith intensive therapy. Similar to theDCCT-EDIC findings, long-term fol-low-up of the UKPDS cohort has recentlydemonstrated a “legacy effect” of early in-tensive glycemic control on long-termrates of microvascular complications,even with loss of glycemic separation be-tween the intensive and standard cohortsafter the end of the randomized con-trolled (51).In each of these large randomizedprospective clinical trials, treatment regi-mens that reduced average A1C to Ն7%(Ն1% above the upper limits of normal)were associated with fewer long-term mi-crovascular complications; however, in-tensive control was found to increase therisk of severe hypoglycemia, most notablyin the DCCT, and led to weight gain(39,52).Epidemiological analyses of theDCCT and UKPDS (39,45) demonstrate acurvilinear relationship between A1C andmicrovascular complications. Such anal-yses suggest that, on a population level,the greatest number of complications willbe averted by taking patients from verypoor control to fair or good control. Theseanalyses also suggest that further loweringof A1C from 7 to 6% is associated withfurther reduction in the risk of microvas-cular complications, albeit the absoluterisk reductions become much smaller.Given the substantially increased risk ofhypoglycemia (particularly in those withtype 1 diabetes) and the relatively muchgreater effort required to achieve near-normoglycemia, the risks of lower targetsmay outweigh the potential benefits onmicrovascular complications on a popu-lation level. However, selected individualpatients, especially those with little co-morbidity and long life expectancy (whomay reap the benefits of further loweringof glycemia below 7%) may, at patientand provider judgment, adopt glycemictargets as close to normal as possible aslong as significant hypoglycemia does notbecome a barrier.Whereas many epidemiologic studiesand meta-analyses (53,54) have clearlyshown a direct relationship between A1Cand CVD, the potential of intensive glyce-mic control to reduce CVD has been lessclearly defined. In the DCCT, there was atrend toward lower risk of CVD eventswith intensive control (risk reduction41%, 95% CI 10–68%), but the numberof events was small. However, 9-yearpost-DCCT follow-up of the cohort hasshown that participants previously ran-domized to the intensive arm had a 42%reduction (P ϭ 0.02) in CVD outcomesand a 57% reduction (P ϭ 0.02) in therisk of nonfatal myocardial infarction(MI), stroke, or CVD death comparedwith those previously in the standard arm(55).The UKPDS trial of type 2 diabetesobserved a 16% reduction in cardiovascu-lar complications (combined fatal or non-fatal MI and sudden death) in theintensive glycemic control arm, althoughthis difference was not statistically signif-icant (P ϭ 0.052), and there was no sug-gestion of benefit on other CVD outcomessuch as stroke. In an epidemiologic anal-ysis of the study cohort, a continuous as-sociation was observed, such that forevery percentage point lower median on-study A1C (e.g., 8 to 7%) there was a sta-tistically significant 18% reduction inCVD events, again with no glycemicthreshold. A recent report of 10 years offollow-up of the UKPDS cohort describes,for the participants originally randomizedto intensive glycemic control comparedwith those randomized to conventionalglycemic control, long-term reductions inMI (15% with sulfonylurea or insulin asinitial pharmacotherapy, 33% with met-formin as initial pharmacotherapy, bothstatistically significant) and in all-causemortality (13 and 27%, respectively, bothstatistically significant) (51).Because of ongoing uncertainty re-garding whether intensive glycemic con-trol can reduce the increased risk of CVDevents in people with type 2 diabetes, sev-eral large long-term trials were launchedin the past decade to compare the effectsof intensive versus standard glycemiccontrol on CVD outcomes in relativelyhigh-risk participants with establishedtype 2 diabetes.The Action to Control CardiovascularRisk in Diabetes (ACCORD) study ran-domized 10,251 participants with eitherhistory of a CVD event (ages 40–79 years)or significant CVD risk (ages 55–79) to astrategy of intensive glycemic control (tar-get A1C Ͻ6.0%) or standard glycemiccontrol (A1C target 7.0–7.9%). Investiga-tors used multiple glycemic medicationsin both arms. ACCORD participants wereon average 62 years old and had a meanduration of diabetes of 10 years, with 35%already treated with insulin at baseline.From a baseline median A1C of 8.1%, theintensive arm reached a median A1C of6.4% within 12 months of randomiza-tion, while the standard group reached amedian A1C of 7.5%. Other risk factorswere treated aggressively and equally inboth groups. The intensive glycemic con-trol group had more use of insulin in com-bination with multiple oral agents,significantly more weight gain, and moreStandards of Medical CareS20 DIABETES CARE, VOLUME 32, SUPPLEMENT 1, JANUARY 2009
  9. 9. episodes of severe hypoglycemia than thestandard group.In February 2008, the glycemic con-trol study of ACCORD was halted on therecommendation of the study’s data safetymonitoring board due to the finding of anincreased rate of mortality in the intensivearm compared with the standard arm(1.41%/year vs. 1.14%/year; HR 1.22[95% CI 1.01–1.46]), with a similar in-crease in cardiovascular deaths. The pri-mary outcome of ACCORD (MI, stroke,or cardiovascular death) was lower in theintensive glycemic control group due to areduction in nonfatal MI, although thisfinding was not statistically significantwhen the study was terminated (HR 0.90[95% CI 0.78–1.04]; P ϭ 0.16) (56).Exploratory analyses of the mortalityfindings of ACCORD (evaluating vari-ables including weight gain, use of anyspecific drug or drug combination, andhypoglycemia) were reportedly unable toidentify an explanation for the excessmortality in the intensive arm. Prespeci-fied subset analyses showed that partici-pants with no previous CVD event andthose who had a baseline A1C Ͻ8% had astatistically significant reduction in theprimary CVD outcome.The ADVANCE study randomized11,140 participants to a strategy of inten-sive glycemic control (with primary ther-apy being the sulfonylurea gliclizide andadditional medications as needed toachieve a target A1C of Յ6.5%) or to stan-dard therapy (in which any medicationbut gliclizide could be used and the gly-cemic target was according to “localguidelines”). ADVANCE participants(who had to be at least 55 years of agewith either known vascular disease or atleast one other vascular risk factor) wereslightly older and of similar high CVD riskas those in ACCORD. However, they hadan average duration of diabetes 2 yearsshorter, lower baseline A1C (median7.2%), and almost no use of insulin atenrollment. The median A1C levelsachieved in the intensive and standardarms were 6.3 and 7.0%, respectively,and maximal separation between thearms took several years to achieve. Use ofother drugs that favorably impact CVDrisk (aspirin, statins, ACE inhibitors) waslower in ADVANCE than in the ACCORDor Veterans Affairs Diabetes Trial (VADT).The primary outcome of ADVANCEwas a combination of microvascularevents (nephropathy and retinopathy)and major adverse cardiovascular events(MI, stroke, and cardiovascular death).Intensive glycemic control significantlyreduced the primary endpoint (HR 0.90[95% CI 0.82–0.98]; P ϭ 0.01), althoughthis was due to a significant reduction inthe microvascular outcome (0.86 [0.77–0.97], P ϭ 0.01), primarily developmentof macroalbuminuria, with no significantreduction in the macrovascular outcome(0.94 [0.84–1.06]; P ϭ 0.32). There wasno difference in overall or cardiovascularmortality between the intensive and thestandard glycemic control arms (57).The VADT randomized 1,791 partic-ipants with type 2 diabetes uncontrolledon insulin or maximal dose oral agents(median entry A1C 9.4%) to a strategy ofintensive glycemic control (goal A1CϽ6.0%) or standard glycemic control,with a planned A1C separation of at least1.5%. Medication treatment algorithmswere used to achieve the specified glyce-mic goals, with a goal of using similarmedications in both groups. Median A1Clevels of 6.9 and 8.4% were achieved inthe intensive and standard arms, respec-tively, within the first year of the study.Other CVD risk factors were treated ag-gressively and equally in both groups.The primary outcome of the VADTwas a composite of CVD events (MI,stroke, cardiovascular death, revascular-ization, hospitalization for heart failure,and amputation for ischemia). During amean 6-year follow-up period, the cumu-lative primary outcome was nonsignifi-cantly lower in the intensive arm (HR0.87 [95% CI 0.73–1.04]; P ϭ 0.12).There were more CVD deaths in the in-tensive arm than in the standard arm (40vs. 33; sudden deaths 11 vs. 4), but thedifference was not statistically significant.Post hoc subgroup analyses suggestedthat duration of diabetes interacted withrandomization such that participantswith duration of diabetes less than about12 years appeared to have a CVD benefitof intensive glycemic control while thosewith longer duration of disease beforestudy entry had a neutral or even adverseeffect of intensive glycemic control. Otherexploratory analyses suggested that se-vere hypoglycemia within the past 90days was a strong predictor of the primaryoutcome and of CVD mortality (58).The cause of the excess deaths in theintensive glycemic control arm of AC-CORD compared with the standard armhas been difficult to pinpoint. By design ofthe trial, randomization to the intensivearm was associated with or led to manydownstream effects, such as higher ratesof severe hypoglycemia; more frequentuse of insulin, TZDs, other drugs, anddrug combinations; and greater weightgain. Such factors may be associated sta-tistically with the higher mortality rate inthe intensive arm but may not be caus-ative. It is biologically plausible that se-vere hypoglycemia could increase the riskof cardiovascular death in participantswith high underlying CVD risk. Otherplausible mechanisms for the increase inmortality in ACCORD include weightgain, unmeasured drug effects or interac-tions, or the overall “intensity” of the AC-CORD intervention (use of multiple oralglucose-lowering drugs along with multi-ple doses of insulin, frequent therapy ad-justments to push A1C and self-monitored blood glucose to very lowtargets, and an intense effort to aggres-sively reduce A1C by ϳ2% in participantsentering the trial with advanced diabetesand multiple comorbidities).Since the ADVANCE trial did notshow any increase in mortality in the in-tensive glycemic control arm, examiningthe differences between ADVANCE andACCORD supports additional hypothe-ses. ADVANCE participants on averageappeared to have earlier or less advanceddiabetes, with shorter duration by 2–3years and lower A1C at entry despite verylittle use of insulin at baseline. A1C wasalso lowered less and more gradually inthe ADVANCE trial, and there was no sig-nificant weight gain with intensiveglycemic therapy. Although severe hypo-glycemia was defined somewhat differ-ently in the three trials, it appears that thisoccurred in fewer than 3% of intensivelytreated ADVANCE participants for theentire study duration (median 5 years)compared with ϳ16% of intensivelytreated subjects in ACCORD and 21% inVADT.It is likely that the increase in mortal-ity in ACCORD was related to the overalltreatment strategies for intensifying glyce-mic control in the study population, notthe achieved A1C per se. The ADVANCEstudy achieved a median A1C in its inten-sive arm similar to that in the ACCORDstudy, with no increased mortality haz-ard. Thus, the ACCORD mortality find-ings do not imply that patients with type 2diabetes who can easily achieve or main-tain low A1C levels with lifestyle modifi-cations with or without pharmacotherapyare at risk and need to “raise” their A1C.The three trials compared treatmentsto A1C levels in the “flatter” part of theobservational glycemia-CVD risk curves(median A1C of 6.4–6.9% in the inten-Position StatementDIABETES CARE, VOLUME 32, SUPPLEMENT 1, JANUARY 2009 S21
  10. 10. sive arms compared with 7.0–8.4% inthe standard arms). Importantly, their re-sults should not be extrapolated to implythat there would be no cardiovascularbenefit of glucose lowering from verypoor control (e.g., A1C Ͼ9%) to goodcontrol (e.g., A1C Ͻ7%).All three trials were carried out in par-ticipants with established diabetes (meanduration 8–11 years) and either knownCVD or multiple risk factors suggestingthe presence of established atherosclero-sis. Subset analyses of the three trials sug-gested a significant benefit of intensiveglycemic control on CVD in participantswith shorter duration of diabetes, lowerA1C at entry, and/or or absence of knownCVD. The DCCT-EDIC study and thelong-term follow-up of the UKPDS cohortboth suggest that intensive glycemic con-trol initiated soon after diagnosis of dia-betes in patients with a lower level of CVDrisk may impart long-term protectionfrom CVD events. As is the case with mi-crovascular complications, it may be thatglycemic control plays a greater role be-fore macrovascular disease is well devel-oped and minimal or no role when it isadvanced.The benefits of intensive glycemiccontrol on microvascular and neuro-pathic complications are well establishedfor both type 1 and type 2 diabetes. TheADVANCE trial has added to that evi-dence base by demonstrating a significantreduction in the risk of new or worseningalbuminuria when A1C was lowered to6.3% compared with standard glycemiccontrol achieving an A1C of 7.0%. Thelack of significant reduction in CVDevents with intensive glycemic control inACCORD, ADVANCE, and VADT shouldnot lead clinicians to abandon the generaltarget of an A1C Ͻ7.0% and thereby dis-count the benefit of good control on whatare serious and debilitating microvascularcomplications.The evidence for a cardiovascularbenefit of intensive glycemic control pri-marily rests on long-term follow-up ofstudy cohorts treated early in the courseof type 1 and type 2 diabetes and subsetanalyses of ACCORD, ADVANCE, andVADT. Conversely, the mortality findingsin ACCORD suggest that the potentialrisks of very intensive glycemic controlmay outweigh its benefits in some pa-tients, such as those with very long dura-tion of diabetes, known history of severehypoglycemia, advanced atherosclerosis,and advanced age/frailty. Certainly, pro-viders should be vigilant in preventing se-vere hypoglycemia in patients withadvanced disease and should not aggres-sively attempt to achieve near-normalA1C levels in patients in whom such atarget cannot be reasonably easily andsafely achieved.Recommended glycemic goals fornonpregnant adults are shown in Table 9.The recommendations are based on thosefor A1C, with listed blood glucose levelsthat appear to correlate with achievementof an A1C of Ͻ7%. The issue of pre- ver-sus postprandial SMBG targets is complex(59). Elevated postchallenge (2-h OGTT)glucose values have been associated withincreased cardiovascular risk indepen-dent of FPG in some epidemiologicalstudies. In diabetic subjects, some surro-gate measures of vascular pathology, suchas endothelial dysfunction, are negativelyaffected by postprandial hyperglycemia(60). It is clear that postprandial hyper-glycemia, like preprandial hyperglyce-mia, contributes to elevated A1C levels,with its relative contribution being higherat A1C levels that are closer to 7%. How-ever, outcome studies have clearly shownA1C to be the primary predictor of com-plications, and landmark glycemic con-trol trials such as the DCCT and UKPDSrelied overwhelmingly on preprandialSMBG. Additionally, a randomized con-trolled trial presented at the 68th Scien-tific Sessions of the American DiabetesAssociation in June 2008 found no CVDbenefit of insulin regimens targeting post-prandial glucose compared with thosetargeting preprandial glucose. A reason-able recommendation for postprandialtesting and targets is that for individualswho have premeal glucose values withintarget but have A1C values above target,monitoring postprandial plasma glucose(PPG) 1–2 h after the start of the meal andtreatment aimed at reducing PPG valuesto Ͻ180 mg/dl may help lower A1C.As noted above, less stringent treat-ment goals may be appropriate for adultswith limited life expectancies or advancedvascular disease. Glycemic goals for chil-dren are provided in Section VII.A.1.a.Severe or frequent hypoglycemia is an ab-solute indication for the modification oftreatment regimens, including settinghigher glycemic goals.Regarding goals for glycemic controlfor women with GDM, recommendationsfrom the Fifth International Workshop-Conference on Gestational Diabetes Mel-litus (61) were to target the followingmaternal capillary glucose concentra-tions:● preprandial: Յ95 mg/dl (5.3 mmol/l)and either● 1-h postmeal: Յ140 mg/dl (7.8 mmol/l)or● 2-h postmeal: Յ120 mg/dl (6.7 mmol/l)For women with preexisting type 1 ortype 2 diabetes who become pregnant, arecent consensus statement (62) recom-mended the following as optimal glyce-mic goals, if they can be achieved withoutexcessive hypoglycemia:● premeal, bedtime, and overnight glu-cose 60–99 mg/dlTable 9—Summary of glycemic recommendations for non-pregnant adults with diabetesA1C Ͻ7.0%*Preprandial capillary plasma glucose 70–130 mg/dl (3.9–7.2 mmol/l)Peak postprandial capillary plasma glucose Ͻ180 mg/dl (Ͻ10.0 mmol/l)Key concepts in setting glycemic goals:● A1C is the primary target for glycemic control.● Goals should be individualized based on:● duration of diabetes● age/life expectancy● comorbid conditions● known CVD or advanced microvascularcomplications● hypoglycemia unawareness● individual patient considerations● More or less stringent glycemic goals may beappropriate for individual patients.● Postprandial glucose may be targeted if A1C goals arenot met despite reaching preprandial glucose goals.*Referenced to a nondiabetic range of 4.0–6.0% using a DCCT-based assay. Postprandial glucose measure-ments should be made 1–2 h after the beginning of the meal, generally peak levels in patients with diabetes.Standards of Medical CareS22 DIABETES CARE, VOLUME 32, SUPPLEMENT 1, JANUARY 2009
  11. 11. ● peak postprandial glucose 100–129mg/dl● A1C Ͻ6.0%3. Approach to treatmenta. Therapy for type 1 diabetes. TheDCCT clearly showed that intensive insu-lin therapy (three or more injections perday of insulin or continuous subcutane-ous insulin infusion (CSII, or insulinpump therapy) was a key part of im-proved glycemia and better outcomes(45). At the time of the study, therapy wascarried out with short- and intermediate-acting human insulins. Despite better mi-crovascular outcomes, intensive insulintherapy was associated with a marked in-crease in severe hypoglycemia (62 epi-sodes per 100 patient-years of therapy).Since the time of the DCCT, a number ofrapid-acting and long-acting insulin ana-logs have been developed. These analogswere designed to be more “physiological”in their pharmacokinetics and pharmaco-dynamics and are associated with less hy-poglycemia with equal A1C lowering intype 1 diabetes (63,64).Therefore, recommended therapy fortype 1 diabetes consists of the followingcomponents: 1) use of multiple dose in-sulin injections (3–4 injections per day ofbasal and prandial insulin) or CSII ther-apy; 2) matching of prandial insulin tocarbohydrate intake, premeal blood glu-cose, and anticipated activity; and 3) formany patients (especially if hypoglycemiais a problem), use of insulin analogs.There are excellent reviews available thatguide the initiation and management ofinsulin therapy to achieve desired glyce-mic goals (3,63,65).b. Therapy for type 2 diabetes. The ADAand the European Association for theStudy of Diabetes published a consensusstatement on the approach to manage-ment of hyperglycemia in individualswith type 2 diabetes (66) and recentlypublished an update (67). Highlights ofthis approach are: intervention at the timeof diagnosis with metformin in combina-tion with lifestyle changes (MNT andexercise) and continuing timely augmen-tation of therapy with additional agents(including early initiation of insulin ther-apy) as a means of achieving and main-taining recommended levels of glycemiccontrol (i.e., A1C Ͻ7% for most patients).The overall objective is to achieve andmaintain glycemic control and to changeinterventions when therapeutic goals arenot being met.The algorithm took into account theevidence for A1C-lowering of the individ-ual interventions, their additive effects,and their expense. The precise drugs usedand their exact sequence may not be asimportant as achieving and maintainingglycemic targets safely. Medications notincluded in the consensus algorithm, ow-ing to less glucose-lowering effectiveness,limited clinical data, and/or relative ex-pense, still may be appropriate choices inindividual patients to achieve glycemicgoals. Initiation of insulin at time of diagno-sis is recommended for individuals present-ing with weight loss or other severehyperglycemic symptoms or signs. For a listof currently approved diabetes medica-tions, see MNTGeneral recommendations● Individuals who have pre-diabetes ordiabetes should receive individualizedMNT as needed to achieve treatmentgoals, preferably provided by a regis-tered dietitian familiar with the compo-nents of diabetes MNT. (B)● MNT should be covered by insuranceand other payors. (E)Energy balance, overweight, andobesity● In overweight and obese insulin-resistant individuals, modest weightloss has been shown to reduce insulinresistance. Thus, weight loss is recom-mended for all overweight or obese in-dividuals who have or are at risk fordiabetes. (A)● For weight loss, either low-carbohy-drate or low-fat calorie restricted dietsmay be effective in the short-term (upto 1 year). (A)● For patients on low-carbohydrate diets,monitor lipid profiles, renal function,and protein intake (in those with ne-phropathy) and adjust hypoglycemictherapy as needed. (E)● Physical activity and behavior modifica-tion are important components of weightloss programs and are most helpful inmaintenance of weight loss. (B)Primary prevention of diabetes● Among individuals at high risk for de-veloping type 2 diabetes, structuredprograms that emphasize lifestylechanges that include moderate weightloss (7% body weight) and regularphysical activity (150 min/week), withdietary strategies including reducedcalories and reduced intake of dietaryfat, can reduce the risk for developingdiabetes and are therefore recom-mended. (A)● Individualsathighriskfortype2diabetesshould be encouraged to achieve the U.S.Department of Agriculture recommenda-tion for dietary fiber (14 g fiber/1,000kcal) and foods containing whole grains(one-half of grain intake). (B)Dietary fat intake in diabetesmanagement● Saturated fat intake should be Ͻ7% oftotal calories. (A)● Intakeoftransfatshouldbeminimized.(B)Carbohydrate intake in diabetesmanagement● Monitoring carbohydrate, whether bycarbohydrate counting, exchanges, orexperience-based estimation, remains akey strategy in achieving glycemic con-trol. (A)● For individuals with diabetes, the use ofthe glycemic index and glycemic loadmay provide a modest additional bene-fit for glycemic control over that ob-served when total carbohydrate isconsidered alone. (B)Other nutrition recommendations● Sugar alcohols and nonnutritive sweet-eners are safe when consumed withinthe acceptable daily intake levels estab-lished by the Food and Drug Adminis-tration (FDA). (A)● If adults with diabetes choose to use alco-hol, daily intake should be limited to amoderate amount (one drink per day orless for adult women and two drinks perday or less for adult men). (E)● Routine supplementation with antioxi-dants, such as vitamins E and C andcarotene, is not advised because of lackof evidence of efficacy and concern re-lated to long-term safety. (A)● Benefit from chromium supplementa-tion in people with diabetes or obesityhas not been conclusively demon-strated and, therefore, cannot be rec-ommended. (E)MNT is an integral component of diabetesprevention, management, and self-management education. In addition to itsrole in preventing and controlling diabe-tes, ADA recognizes the importance ofnutrition as an essential component of anoverall healthy lifestyle. A full review ofthe evidence regarding nutrition in pre-Position StatementDIABETES CARE, VOLUME 32, SUPPLEMENT 1, JANUARY 2009 S23
  12. 12. venting and controlling diabetes and itscomplications and additional nutrition-related recommendations can be found inthe ADA position statement “NutritionRecommendations and Interventions forDiabetes,” published in 2007 and up-dated for 2008 (68). Achieving nutrition-related goals requires a coordinated teameffort that includes the active involvementof the person with pre-diabetes or diabe-tes. Because of the complexity of nutritionissues, it is recommended that a registereddietitian who is knowledgeable andskilled in implementing nutrition therapyinto diabetes management and educationbe the team member who provides MNT.Clinical trials/outcome studies ofMNT have reported decreases in A1C at3–6 months ranging from 0.25 to 2.9%with higher reductions seen in type 2diabetes of shorter duration. Multiplestudies have demonstrated sustained im-provements in A1C at 12 months andlonger when a registered dietician pro-vided follow-up visits ranging frommonthly to three sessions per year (69–76). Meta-analyses of studies in nondia-betic, free-living subjects report that MNTreduces LDL cholesterol by 15–25 mg/dl(77) or can lower LDL cholesterol by upto 16% (78), while clinical trials support arole for lifestyle modification in treatinghypertension (78,79).Because of the effects of obesity oninsulin resistance, weight loss is an im-portant therapeutic objective for over-weight or obese individuals with pre-diabetes or diabetes (80). Short-termstudies have demonstrated that moderateweight loss (5% of body weight) in sub-jects with type 2 diabetes is associatedwith decreased insulin resistance, im-proved measures of glycemia and lipemia,and reduced blood pressure (81); longer-term studies (52 weeks) showed mixedeffects on A1C in adults with type 2 dia-betes (82–85), and results were con-founded by pharmacologic weight losstherapy. A systematic review of 80 weightloss studies of Ն1 year duration demon-strated that moderate weight lossachieved through diet alone, diet and ex-ercise, and meal replacements can beachieved and maintained over the longterm (4.8–8% weight loss at 12 months)(86). The multifactorial intensive lifestyleintervention employed in the DPP, whichincluded reduced intake of fat and calo-ries, led to weight loss averaging 7% at 6months and maintenance of 5% weightloss at 3 years, associated with a 58% re-duction in incidence of type 2 diabetes(10). Look AHEAD (Action for Health inDiabetes) is a large clinical trial designedto determine whether long-term weightloss will improve glycemia and preventcardiovascular events in subjects withtype 2 diabetes. One-year results of theintensive lifestyle intervention in this trialshow an average of 8.6% weight loss, sig-nificant reduction of A1C, and reductionin several CVD risk factors (87). Whencompleted, the Look AHEAD trial shouldprovide insight into the effects of long-term weight loss on important clinicaloutcomes.The optimal macronutrient distribu-tion of weight loss diets has not been es-tablished. Although low-fat diets havetraditionally been promoted for weightloss, several randomized controlled trialsfound that subjects on low-carbohydratediets (Ͻ130 g/day of carbohydrate) lostmore weight at 6 months than subjects onlow-fat diets (88,89); however, at 1 year,the difference in weight loss between thelow-carbohydrate and low-fat diets wasnot significant, and weight loss was mod-est with both diets. Another study of over-weight women randomized to one of fourdiets showed significantly more weightloss at 12 months with the Atkins low-carbohydrate diet than with higher-carbohydrate diets (90). Changes inserum triglyceride and HDL cholesterolwere more favorable with the low-carbohydrate diets. In one study, thosesubjects with type 2 diabetes demon-strated a greater decrease in A1C with alow-carbohydrate diet than with a low-fatdiet (89). A recent meta-analysis showedthat at 6 months, low-carbohydrate dietswere associated with greater improve-ments in triglyceride and HDL cholesterolconcentrations than low-fat diets; how-ever, LDL cholesterol was significantlyhigher on the low-carbohydrate diets(91). In a 2-year dietary interventionstudy, Mediterranean and low-carbohy-drate diets were found to be effective andsafe alternatives to a low-fat diet forweight reduction in moderately obeseparticipants (85).The recommended dietary allowancefor digestible carbohydrate is 130 g/dayand is based on providing adequate glu-cose as the required fuel for the centralnervous system without reliance on glu-cose production from ingested protein orfat. Although brain fuel needs can be meton lower carbohydrate diets, long-termmetabolic effects of very-low-carbohy-drate diets are unclear, and such dietseliminate many foods that are importantsources of energy, fiber, vitamins, andminerals that are important in dietary pal-atability (92).Although numerous studies have at-tempted to identify the optimal mix ofmacronutrients for meal plans of peoplewith diabetes, it is unlikely that one suchcombination of macronutrients exists.The best mix of carbohydrate, protein,and fat appears to vary depending onindividual circumstances. For those indi-viduals seeking guidance as to macronu-trient distribution in healthy adults, theDietary Reference Intake (DRI) systemmay be helpful (92). It must be clearlyrecognized that regardless of the macro-nutrient mix, total caloric intake must beappropriate for the weight managementgoal. Further, individualization of the ma-cronutrient composition will depend onthe metabolic status of the patient (e.g.,lipid profile, renal function) and/or foodpreferences. Individuals who choose toconsume plant-based diets that are wellplanned and nutritionally adequate (i.e.,vegetarian) may continue, as this can bedone without being deleterious to meta-bolic control (93,94).The primary goal with respect to di-etary fat in individuals with diabetes is tolimit saturated fatty acids, trans fatty ac-ids, and cholesterol intake so as to reducerisk for CVD. Saturated and trans fatty ac-ids are the principal dietary determinantsof plasma LDL cholesterol. There is a lackof evidence on the effects of specific fattyacids on people with diabetes, so the rec-ommended goals are consistent withthose for individuals with CVD (78,95).The FDA has approved five nonnutri-tive sweeteners for use in the U.S.: acesul-fame potassium, aspartame, neotame,saccharin, and sucralose. Before being al-lowed on the market, all underwent rig-orous scrutiny and were shown to be safewhen consumed by the public, includingpeople with diabetes and women duringpregnancy. Reduced calorie sweetenersapproved by the FDA include sugar alco-hols (polyols) such as erythritol, isomalt,lactitol, maltitol, mannitol, sorbitol, xyli-tol, tagatose, and hydrogenated starch hy-drolysates. The use of sugar alcoholsappears to be safe; however, they maycause diarrhea, especially in children.Reimbursement for MNTMNT, when delivered by a registered di-etitian according to nutrition practiceguidelines, is reimbursed as part of theMedicare program as overseen by theCenters for Medicare and Medicaid Ser-Standards of Medical CareS24 DIABETES CARE, VOLUME 32, SUPPLEMENT 1, JANUARY 2009
  13. 13. vices (CMS) ( Bariatric surgeryRecommendations● Bariatric surgery should be consideredfor adults with BMI Ն35 kg/m2andtype 2 diabetes, especially if the diabe-tes is difficult to control with lifestyleand pharmacologic therapy. (B)● Patients with type 2 diabetes who haveundergone bariatric surgery need life-long lifestyle support and medicalmonitoring. (E)● Although small trials have shown glyce-micbenefitofbariatricsurgeryinpatientswith type 2 diabetes and BMI of 30–35kg/m2, there is currently insufficient evi-dencetogenerallyrecommendsurgeryinpatientswithBMIϽ35kg/m2outsideofaresearch protocol. (E)● The long-term benefits, cost-effectiveness, and risks of bariatric sur-gery in individuals with type 2 diabetesshould be studied in well-designed ran-domized controlled trials with optimalmedical and lifestyle therapy as thecomparator. (E)Gastric reduction surgery, either gastricbanding or procedures that involve by-passing or transposing sections of thesmall intestine, when part of a compre-hensive team approach, can be an effec-tive weight loss treatment for severeobesity, and national guidelines supportits consideration for people with type 2diabetes who have BMI at or exceeding 35kg/m2. Bariatric surgery has been shownto lead to near or complete normalizationof glycemia in ϳ55–95% of patients withtype 2 diabetes, depending on the surgicalprocedure. A meta-analysis of studies ofbariatric surgery reported that 78% ofindividuals with type 2 diabetes had com-plete “resolution” of diabetes (normaliza-tion of blood glucose levels in the absenceof medications), and that the resolutionrates were sustained in studies that hadfollow-up exceeding 2 years (96). Resolu-tion rates are lowest with procedures thatonly constrict the stomach and higherwith those that bypass portions of thesmall intestine. Additionally, there is in-creasing evidence that intestinal bypassprocedures may have glycemic effects thatare independent of, and additive to, theireffects on weight.A recent randomized controlled trialcompared adjustable gastric banding to“best available” medical and lifestyletherapy in subjects with type 2 diabetesdiagnosed less than 2 years before ran-domization and BMI 30–40 kg/m2(97).In this trial, 73% of surgically treated pa-tients achieved “remission” of their diabe-tes, compared with 13% of those treatedmedically. The latter group lost only 1.7%of body weight, suggesting that their ther-apy was not optimal. Overall, the trial had60 subjects, and only 13 had a BMI under35 kg/m2, making it difficult to generalizethese results widely to diabetic patientswho are less severely obese or with longerduration of diabetes.Bariatric surgery is costly in the shortterm and has some risks. Rates of morbid-ity and mortality directly related to thesurgery have been reduced considerablyin recent years, with 30-day mortalityrates now 0.28%, similar to those of lapa-roscopic cholecystectomy (98). Longer-term concerns include vitamin andmineral deficiencies, osteoporosis, andrare but often severe hypoglycemia frominsulin hypersecretion. Cohort studies at-tempting to match subjects suggest thatthe procedure may reduce longer-termmortality rates (99), and it is reasonable topostulate that there may be recouping ofcosts over the long run. However, studiesof the mechanisms of glycemic improve-ment, long-term benefits and risks, andcost-effectiveness of bariatric surgery inindividuals with type 2 diabetes will re-quire well-designed randomized clinicaltrials, with optimal medical and lifestyletherapy of diabetes and cardiovascularrisk factors as the comparitor.F. DSMERecommendations● People with diabetes should receiveDSME according to national standardswhen their diabetes is diagnosed and asneeded thereafter. (B)● Self-management behavior change isthe key outcome of DSME and shouldbe measured and monitored as part ofcare. (E)● DSME should address psychosocial is-sues, since emotional well-being isstrongly associated with positive diabe-tes outcomes. (C)● DSME should be reimbursed by third-party payors. (E)DSME is an essential element of diabetescare (100–106), and National Standardsfor DSME (107) are based on evidence forits benefits. Education helps people withdiabetes initiate effective self-care whenthey are first diagnosed. Ongoing DSMEalso helps people with diabetes maintaineffective self-management as their diabe-tes presents new challenges and treatmentadvances become available. DSME helpspatients optimize metabolic control, pre-vent and manage complications, andmaximize quality of life in a cost-effectivemanner (108).Since the 1990s, there has been a shiftfrom a didactic approach, with DSME fo-cusing on providing information, to askill-based approach that focuses onhelping those with diabetes make in-formed self-management choices. Care ofdiabetes has shifted to an approach that ismore patient centered and that places theperson with diabetes, and joint decision-making with heath care professionals, atthe center of the care model. Patient-centered care is respectful of and respon-sive to individual patient preferences,needs, and values and ensures that patientvalues guide all decision making (109).Evidence for the benefits of DSMESeveral studies have found that DSME isassociated with improved diabetesknowledge and improved self-care behav-ior (101), improved clinical outcomessuch as lower A1C (102,103,105,106,110), lower self-reported weight (101),and improved quality of life (104). Betteroutcomes were reported for DSME inter-ventions that were longer and includedfollow-up support (101), that were tai-lored to individual needs and preferences(100), and that addressed psychosocial is-sues (100,101,105). Both individual andgroup approaches have been found effec-tive ((111,112). There is increasing evi-dence for the role of a community healthworker in delivering diabetes educationin addition to the core team (113).National standards for DSMEADA-recognized DSME programs havestaff that must be certified diabetes edu-cators or have recent experience in diabe-tes education and management. Thecurriculum of ADA-recognized DSMEprograms must cover all nine areas of di-abetes management, with the assessedneeds of the individual determiningwhich areas are addressed. The ADA Ed-ucation Recognition Program (ERP) is amechanism to ensure diabetes educationprograms meet the national standards andprovide quality diabetes care.Position StatementDIABETES CARE, VOLUME 32, SUPPLEMENT 1, JANUARY 2009 S25
  14. 14. Reimbursement for DSMEDSME, when provided by a program thatmeets ADA ERP standards, is reimbursedas part of the Medicare program as over-seen by the Centers for Medicare andMedicaid Services (CMS) ( Physical activityRecommendations● People with diabetes should be advisedto perform at least 150 min/week ofmoderate-intensity aerobic physical ac-tivity (50 –70% of maximum heartrate). (A)● In the absence of contraindications,people with type 2 diabetes should beencouraged to perform resistance train-ing three times per week. (A)ADA technical reviews on exercise in pa-tients with diabetes have summarized thevalue of exercise in the diabetes manage-ment plan (114,115). Regular exercisehas been shown to improve blood glucosecontrol, reduce cardiovascular risk fac-tors, contribute to weight loss, and im-prove well being. Furthermore, regularexercise may prevent type 2 diabetes inhigh-risk individuals (10–12). Struc-tured exercise interventions of at least 8weeks’ duration have been shown tolower A1C by an average of 0.66% in peo-ple with type 2 diabetes, even with nosignificant change in BMI (116). Higherlevels of exercise intensity are associatedwith greater improvements in A1C and infitness (117).Frequency and type of exerciseThe U.S. Surgeon General’s report (118)recommended that most adults accumu-late at least 30 min of moderate-intensityactivity on most, ideally all, days of theweek. The studies included in the meta-analysis of effects of exercise interventionson glycemic control (116) had a meannumber of sessions per week of 3.4, witha mean of 49 min per session. The DPPlifestyle intervention, which included 150min per week of moderate intensity exer-cise, had a beneficial effect on glycemia inthose with pre-diabetes. Therefore, itseems reasonable to recommend ϳ150min of exercise per week for people withdiabetes.Resistance exercise improves insulinsensitivity to about the same extent as aer-obic exercise (119). Clinical trials haveprovided strong evidence for the A1C-lowering value of resistance training inolder adults with type 2 diabetes(120,121) and for an additive benefit ofcombined aerobic and resistance exercisein adults with type 2 diabetes (122).Evaluation of the diabetic patientbefore recommending an exerciseprogramPrior guidelines suggested that before rec-ommending a program of physical activ-ity, the provider should assess patientswith multiple cardiovascular risk factorsfor coronary artery disease (CAD). As dis-cussed more fully in Section VI.A.5, thearea of screening asymptomatic diabeticpatients for CAD remains unclear, and arecent ADA consensus statement on thisissue concluded that routine screening isnot recommended (123). Providersshould use clinical judgment in this area.Certainly, high-risk patients should beencouraged to start with short periods oflow-intensity exercise and increase the in-tensity and duration slowly.Providers should assess patients forconditions that might contraindicate cer-tain types of exercise or predispose to in-jury, such as uncontrolled hypertension,severe autonomic neuropathy, severe pe-ripheral neuropathy or history of foot le-sions, and advanced retinopathy. Thepatient’s age and previous physical activ-ity level should be considered.Exercise in the presence ofnonoptimal glycemic controlHyperglycemia. When people with type1 diabetes are deprived of insulin for12–48 h and are ketotic, exercise canworsen hyperglycemia and ketosis (124);therefore, vigorous activity should beavoided in the presence of ketosis. How-ever, it is not necessary to postpone exer-cise based simply on hyperglycemia,provided the patient feels well and urineand/or blood ketones are negative.Hypoglycemia. In individuals taking in-sulin and/or insulin secretagogues, phys-ical activity can cause hypoglycemia ifmedication dose or carbohydrate con-sumption is not altered. For individualson these therapies, added carbohydrateshould be ingested if pre-exercise glucoselevels are Ͻ100 mg/dl (5.6 mmol/l)(125,126). Hypoglycemia is rare in dia-betic individuals who are not treated withinsulin or insulin secretagogues, and nopreventive measures for hypoglycemiaare usually advised in these cases.Exercise in the presence of specificlong-term complications of diabetesRetinopathy. In the presence of prolifer-ative diabetic retinopathy (PDR) or severenonproliferative diabetic retinopathy(NPDR), vigorous aerobic or resistanceexercise may be contraindicated becauseof the risk of triggering vitreous hemor-rhage or retinal detachment (127).Peripheral neuropathy. Decreased painsensation in the extremities results in in-creased risk of skin breakdown and infec-tion and of Charcot joint destruction.Therefore, in the presence of severe pe-ripheral neuropathy, it may be best to en-courage non–weight-bearing activitiessuch as swimming, bicycling, or arm ex-ercises (128,129).Autonomic neuropathy. Autonomicneuropathy can increase the risk of exer-cise-induced injury or adverse eventsthrough decreased cardiac responsivenessto exercise, postural hypotension, impairedthermoregulation, impaired night visiondue to impaired papillary reaction, and un-predictable carbohydrate delivery from gas-troparesis predisposing to hypoglycemia(128). Autonomic neuropathy is alsostrongly associated with CVD in peoplewith diabetes (130,131). People with dia-betic autonomic neuropathy should un-dergo cardiac investigation beforebeginning physical activity more intensethan that to which they are accustomed.Albuminuria and nephropathy. Physicalactivity can acutely increase urinary pro-tein excretion. However, there is no evi-dence that vigorous exercise increases therate of progression of diabetic kidney dis-ease and likely no need for any specificexercise restrictions for people with dia-betic kidney disease (132).H. Psychosocial assessment and careRecommendations● Assessment of psychological and socialsituation should be included as an on-going part of the medical managementof diabetes. (E)● Psychosocial screening and follow-upshould include, but is not limited to,attitudes about the illness, expectationsfor medical management and out-comes, affect/mood, general and diabe-tes-related quality of life, resources(financial, social, and emotional), andpsychiatric history. (E)● Screen for psychosocial problems suchas depression, anxiety, eating disor-ders, and cognitive impairment whenStandards of Medical CareS26 DIABETES CARE, VOLUME 32, SUPPLEMENT 1, JANUARY 2009
  15. 15. adherence to the medical regimen ispoor. (E)Psychological and social problems can im-pair the individual’s (133–138) or family’s(139) ability to carry out diabetes care tasksand therefore compromise health status.There are opportunities for the clinician toassess psychosocial status in a timely andefficient manner so that referral for appro-priate services can be accomplished.Key opportunities for screening ofpsychosocial status occur at diagnosis,during regularly scheduled managementvisits, during hospitalizations, at discov-ery of complications, or when problemswith glucose control, quality of life, or ad-herence are identified (140). Patients arelikely to exhibit psychological vulnerabil-ity at diagnosis and when their medicalstatus changes, i.e., the end of the honey-moon period, when the need for intensi-fied treatment is evident, and whencomplications are discovered (135,137).Issues known to impact self-management and health outcomes in-clude but are not limited to attitudesabout the illness, expectations for medicalmanagement and outcomes, affect/mood,general and diabetes-related quality oflife, resources (financial, social, and emo-tional) (136), and psychiatric history(137,140,141). Screening tools are avail-able for a number of these areas (142).Indications for referral to a mental healthspecialist familiar with diabetes manage-ment may include gross noncompliancewith medical regimen (by self or others)(141), depression with the possibility ofself-harm (134,143), debilitating anxiety(alone or with depression), indications ofan eating disorder (144), or cognitivefunctioning that significantly impairsjudgment (143). It is preferable to incor-porate psychological assessment andtreatment into routine care rather thanwaiting for identification of a specificproblem or deterioration in psychologicalstatus (142). Although the clinician maynot feel qualified to treat psychologicalproblems, utilizing the patient-providerrelationship as a foundation for furthertreatment can increase the likelihood thatthe patient will accept referral for otherservices. It is important to establish thatemotional well-being is part of diabetesmanagement (140).I. When treatment goals are not metFor a variety of reasons, some people withdiabetes and their health care providersdo not achieve the desired goals of treat-ment (Table 9). Re-thinking the treatmentregimen may require assessment of barri-ers to adherence including income, edu-cational attainment, and competingdemands, including those related to fam-ily responsibilities and family dynamics.Other strategies may include culturallyappropriate and enhanced DSME, co-management with a diabetes team, refer-ral to a medical social worker forassistance with insurance coverage orchange in pharmacological therapy. Initi-ation of or increase in SMBG, utilizationof continuous glucose monitoring, fre-quent contact with the patient, or referralto an endocrinologist may be useful.J. Intercurrent illnessThe stress of illness, trauma, and/or sur-gery frequently aggravates glycemic con-trol and may precipitate diabeticketoacidosis (DKA) or nonketotic hyper-osmolar state, life-threatening conditionsthat require immediate medical care toprevent complications and death (145).Any condition leading to deterioration inglycemic control necessitates more fre-quent monitoring of blood glucose and(in ketosis-prone patients) urine or bloodketones. Marked hyperglycemia requirestemporary adjustment of the treatmentprogram and, if accompanied by ketosis,vomiting, or alteration in level of con-sciousness, immediate interaction withthe diabetes care team. The patient treatedwith noninsulin therapies or MNT alonemay temporarily require insulin. Ade-quate fluid and caloric intake must be as-sured. Infection or dehydration are morelikely to necessitate hospitalization of theperson with diabetes than the personwithout diabetes.The hospitalized patient should betreated by a physician with expertise in themanagement of diabetes. For further infor-mation on management of patients with hy-perglycemia in the hospital, see SectionVIII.A. For further information on manage-ment of DKA or nonketotic hyperosmolarstate,refertotheADApositionstatementonhyperglycemic crises (145).K. HypoglycemiaRecommendations● Glucose (15–20 g) is the preferredtreatment for the conscious individualwith hypoglycemia, although any formof carbohydrate that contains glucosemay be used. If SMBG 15 min aftertreatment shows continued hypoglyce-mia, the treatment should be repeated.Once SMBG glucose returns to normal,the individual should consume a mealor snack to prevent recurrence of hypo-glycemia. (E)● Glucagon should be prescribed for allindividuals at significant risk of severehypoglycemia, and caregivers or familymembers of these individuals should beinstructed in its administration. Gluca-gon administration is not limited tohealth care professionals. (E)● Individuals with hypoglycemia un-awareness or one or more episodes ofsevere hypoglycemia should be advisedto raise their glycemic targets to strictlyavoid further hypoglycemia for at leastseveral weeks to partially reverse hypo-glycemia unawareness and reduce riskof future episodes. (B)Hypoglycemia is the leading limiting fac-tor in the glycemic management of type 1and insulin-treated type 2 diabetes (146).Treatment of hypoglycemia (plasma glu-cose Ͻ70 mg/dl) requires ingestion ofglucose- or carbohydrate-containingfoods. The acute glycemic response cor-relates better with the glucose contentthan with the carbohydrate content of thefood. Although pure glucose is the pre-ferred treatment, any form of carbohy-drate that contains glucose will raiseblood glucose. Added fat may retard andthen prolong the acute glycemic response(147). Ongoing activity of insulin or in-sulin secretagogues may lead to recur-rence of hypoglycemia unless furtherfood is ingested after recovery.Severe hypoglycemia (where the indi-vidual requires the assistance of anotherperson and cannot be treated with oralcarbohydrate due to confusion or uncon-sciousness) should be treated using emer-gency glucagon kits, which require aprescription. Those in close contact with,or having custodial care of, people withhypoglycemia-prone diabetes (familymembers, roommates, school personnel,child care providers, correctional institu-tion staff, or coworkers) should be in-structed in use of such kits. An individualdoes not need to be a health care profes-sional to safely administer glucagon. Careshould be taken to ensure that unexpiredglucagon kits are available.Prevention of hypoglycemia is a crit-ical component of diabetes management.Teaching people with diabetes to balanceinsulin use, carbohydrate intake, and ex-ercise is a necessary but not always suffi-cient strategy. In type 1 diabetes andseverely insulin-deficient type 2 diabetes,Position StatementDIABETES CARE, VOLUME 32, SUPPLEMENT 1, JANUARY 2009 S27