Insulin Therapy in People WithType 2 Diabetes: Opportunitiesand Challenges?Diabetes Care 2014;37:1499–1508 | DOI: 10.2337/dc13-2743

Given the continued interest in defining the optimal management of individualswith type 2 diabetes, the Editor of Diabetes Care convened a working party ofdiabetes specialists to examine this topic in the context of insulin therapy. Thiswas prompted by recent new evidence on the use of insulin in such people. Thegroup was aware of evidence that the benefits of insulin therapy are still usuallyoffered late, and thus the aim of the discussion was how to define the optimaltiming and basis for decisions regarding insulin and to apply these concepts inpractice. It was noted that recent evidence had built upon that of the previousdecades, together confirming the benefits and safety of insulin therapy, albeitwith concerns about the potential for hypoglycemia and gain in body weight.Insulin offers a unique ability to control hyperglycemia, being used from the timeof diagnosis in some circumstances, when metabolic control is disturbed by medicalillness, procedures, or therapy, as well as in the longer term in ambulatory care. Forthose previously starting insulin, various other forms of therapy can be added later,which offer complementary effects appropriate to individual needs. Here we reviewcurrent evidence and circumstances in which insulin can be used, consider individ-ualized choices of alternatives and combination regimens, and offer some guidanceon personalized targets and tactics for glycemic control in type 2 diabetes.

The ultimate goal of diabetes management is prevention of long-term complica-tions. An important means to this end is improvement andmaintenance of glycemiccontrol over time. Unfortunately, this is not a simple task due to the progressivenature of the disease, which requires timely optimization of treatment, leading in amajority of cases to insulin therapy. Various forces oppose and thus delay startinginsulin, and the lag between the time insulin is needed and the time it is used hasbeen described as due to “clinical inertia” (1). Shah et al. (2) have reported that lessthan one-half of patients with high HbA1c levels have their treatment optimizedeven when specialists manage their condition. However, in that study, specialistsweremore active in prescribing insulin than primary care physicians. Nonetheless, inall areas of clinical practice, use of insulin tends to be delayed and irreversiblecomplications can already be present by the time it is started. In a multinationalsurvey involving.66,000 diabetic patients, average HbA1c at the time of beginninginsulin was 80 mmol/mol (9.5%) and ;90% of the participants already had somekind of complication (3). Various concerns serve as barriers to starting insulin, andoften it is a physician rather than the patient who decides to postpone insulintherapy (4).The specific point at which insulin therapy should begin can be difficult to define

for an individual person, and universal guidance has proved elusive. Type 2 diabetes is

1Newcastle University, Newcastle upon Tyne, U.K.2OregonHealth&ScienceUniversity, Portland, OR3Pennington Biomedical Research Center, Louisi-ana State University System, Baton Rouge, LA4FRCPath, Aston University, Birmingham, U.K.5Justus Liebeg University, Giessen, Germany6Department of Clinical and Experimental Med-icine, University of Pisa School of Medicine, Pisa,Italy7Mount Sinai Medical School, New York, NY8Rambam Technion Hospital, Haifa, Israel9Department of Medicine I, Rudolfstiftung Hos-pital, Vienna, Austria10Antwerp University Hospital, Antwerp, Belgium11Diabetes Unit, Department of Internal Medi-cine, Hadassah Hebrew University Hospital, Jer-usalem, Israel

Corresponding author: William T. Cefalu, william.cefalu@pbrc.edu.

Received 23 November 2013 and accepted 18February 2014.

© 2014 by the American Diabetes Association.See http://creativecommons.org/licenses/by-nc-nd/3.0/ for details.

Philip Home,1 Matthew Riddle,2

William T. Cefalu,3 Clifford J. Bailey,4

Reinhard G. Bretzel,5 Stefano del Prato,6

Derek Leroith,7,8 Guntram Schernthaner,9

Luc van Gaal,10 and Itamar Raz11

Diabetes Care Volume 37, June 2014 1499

PER

SPEC

TIVES

INCARE

characterized by progressive b-cell(b-cell) failure, but the natural historyof b-cell decline is variable and assess-ment of b-cell function is difficult. Be-yond the problem of assessing theneed for insulin, exogenous insulin haspotential effects that frequently worryboth people with diabetes and healthcare professionals (4). These include hy-poglycemia and weight gain. Fear of in-jections themselves and various negativeconnotations of insulin therapy, such asadvanced disease and personal failure,also pose significant hurdles for somepeople despite modern injection devices.However, suggestion of a need to in-crease self-monitoring, and thus fingerpricking, to support optimization of insu-lin have ameliorated injection problemsto some degree. Finally, educational sup-port for starting injections and adjustingthedosage of insulin is not easily availableto all people.

GROWING EVIDENCE FOR INSULINTHERAPY

As noted above, a commonly acceptedview is that type 2 diabetes developswhen insulin secretion can no longercompensate for the underlying meta-bolic disturbance. As secretory capacityprogressively declines with time (5), it isunderstood that most people with type2 diabetes will eventually require insulintherapy. Increasing use of therapies toprotect against cardiovascular disease isextending the life of people with diabe-tes (6), and consequently more peoplewill come to need insulin therapy. Diag-nosis at a younger age will also extendthe time of active treatment of diabetes.The effect of use of insulin in type 2 di-abetes from the time of diagnosis hasbeen evaluated in clinical trials, notablythe UK Prospective Diabetes Study(UKPDS) and Outcome Reduction WithInitial Glargine Intervention (ORIGIN)(7–9). UKPDS showed that early andcontinued glucose control can reducemicrovascular complications and, inthe long-term, improve cardiovascularprognosis (7,10). The beneficial effectof insulin therapy is further supportedby studies in type 1 diabetes where itis apparent that if insulin therapy isused effectively to induce early glycemiccontrol, both micro- and macrovascularprotection is achieved (11). Although itis acknowledged that achieving HbA1c,53 mmol/mol (7.0%) is a difficult

task, improvement of glycemic controlwith insulin is associated with improvedpatient well-being even if the HbA1c tar-get is not achieved (12) (Table 1).

Short-term studies comparing insu-lins show that a high percentage of peo-ple with established type 2 diabetes notwell controlled with oral therapies canachieve blood glucose control to target,and without high rates of hypoglycemia(13–15). Further, with optimization ofdosage and timing, such control can bemaintained for least 3 years, with hypo-glycemia rates constant at;10% of peo-ple with an event in a year (16).Observational studies show that in rou-tine practice, HbA1c of ;57 mmol/mol(7.4%) is achievedwith little hypoglycemiain diverse countries around the world(3,17). Most recently, the ORIGIN trialintroduced insulin therapy in peoplewith relatively short-duration type 2 di-abetes and high cardiovascular risk (9).With continuation of prior oral therapyand systematic titration of basal insulin(glargine), fasting glucose was kept atnormal levels (,5.3 mmol/L) and me-dian HbA1c #45 mmol/mol (6.3%) for.6 years. Systematically intensifiedoral therapy, the comparator regimenin ORIGIN, also maintained very goodcontrol during this time. This studyshowed that despite maintaining very

good glycemic control, insulin use wasassociated with a weight gain of a mod-est 2.1 kgmore than on the oral regimen,the rate of hypoglycemia requiring assis-tance was modest (1% per year; ,10%per year for nonsevere confirmed hypo-glycemia), and there was no evidenceof increased risk of malignancy or otherserious adverse events. Adding the expe-rience from ORIGIN to previous observa-tions, particularly UKPDS, it can beargued that the evidence base forinsulin is better than that for any otherglucose-lowering agents, except perhapsmetformin.

Even so, it must be acknowledgedthat although the safety and benefit ofmoderate doses of insulin in the long-term are clear (7,9,10), the use of high-dose insulin therapy, especially withmealtime injections, for people withobesity and severe insulin resistancemight still be associated with adverseoutcomes. Also, the evidence basedoes not satisfactorily address the an-swers to common questions that physi-cians ask themselves, such as whether itis time to start insulin or, alternatively,whether other therapies should beconsidered.

Another issue is what insulin andwhat insulin regimen should be chosen,perhaps bearing in mind the future need

Table 1—Summary of the evidence base for starting insulin in type 2 diabetes

Evidence type Evidence findings

Strong observational and randomized clinical trial evidenceInsulin secretory capacity deteriorates with time (5)Insulin improves glycemic control in trials and in routine clinical

practice (3,9,13,14,16)Improved glucose control improves HRQoL (12)LADA phenotype is associated with early need for insulin therapy (28)

Randomized clinical trial evidence of variable qualityOutcomes of acute illness are improved if glycemic control is better (61)Long-term medical outcomes are improved by better glycemic

control (7,10,62)Glycemic control to ,53 mmol/mol (,7.0%) HbA1c is difficult to

achieve and maintain with insulin (3,16)Insulin is successful in combination with oral agents and

GLP-1RAs (49,63)

General knowledge and expert experienceInsulin treats and prevents ketoacidosis (64)Severe hyperglycemia predisposes to infection (53)Physician hesitancy in starting insulin therapy is a main barrier to

insulin use (4)Patient preferences and views of injected therapies vary markedly (4)Insulin therapy can be tailored rapidly to changes in need during acute

illnessInsulin has potential powerful anabolic effects (wound healing, etc.)

HRQoL, health-related quality of life.

1500 Insulin Therapy in People With Type 2 Diabetes Diabetes Care Volume 37, June 2014

of optimization associated with the pro-gression of the condition, and consider-ing potentially high-risk subgroups forwhich specific tactics might be chosen.Basal insulins offer simplicity in injectionfrequency and ease of dose titration butmay not be adequate to address post-prandial excursions (15,18). This is par-ticularly true with further loss of isletb-cell function, whence a mealtime +basal regimen will be needed, some-times with mealtime insulin introducedmeal by meal (16,19). Premixed insulins,started once or twice daily and occa-sionally increased to thrice daily, canoffer a simpler approach where meal-time insulin is needed. However, premixinsulins lack flexibility for dose adjust-ment and limit calibration of doses be-tween the mealtime and basal needs.They also give a poorer basal insulin pro-file at night and more risk of daytimehypoglycemia compared with basalalone if mealtime insulin is not needed(20).

WHEN SHOULD INSULIN THERAPYBE STARTED?

Indications for insulin therapy and whento begin it are poorly defined in guide-lines and still subject to individual judg-ment based on a wide range of opinion(21,22). Personal beliefs and experi-ence, familiarity with the use of the dif-ferent insulin preparations and deliverysystems, individual preference, patientneedle phobia, concern about chronichyperinsulinemia, risk of hypoglycemia,and difficulties in controlling bodyweight are some of the many consider-ations regarding insulin therapy (4,23).Each one of these factors can beweighted differently between doctorsand between people with diabetes.The expert group proposed that oneway to rationalize the approach to insu-lin treatment could be to consider someclinical scenarios. These could be as fol-lows: 1) the time of diagnosis or earlythereafter; 2) in the presence of otheremerging medical conditions; and 3) inthe course of routine ambulatory diabe-tes management.

Time of Diagnosis or Early ThereafterIn specific circumstances, insulin may beconsidered a potential first line therapy(Table 2 and Fig. 1) (24). Presentation ofdiabetes with ketonuria, with or withoutmarked symptoms, should raise the

suspicion of type 1 diabetes, althoughthe possibility of accelerated lipolysisand ketogenesis due to starvation alsodeserves consideration. Ketoacidosis canbepresent at the timeof diagnosis of type2 diabetes, especially in the presence ofanother metabolic stress (i.e., myocardialinfarction or infection or use of antiretro-viral therapy). In such situations, immedi-ate use of insulin is recommended andsometimes mandatory.

However, more common is the pre-sentation of marked hyperglycemiawith or without notable symptoms orketonuria or advanced comorbidities(Fig. 1). Some guidelines suggest insulinis indicated for HbA1c .69 mmol/mol(.8.5%) at diagnosis (24). This approachis not well substantiated by evidence,but it clearly can help by reducing glu-cose toxicity. Some evidence suggestsshort-term insulin therapy might helppreserve b-cell mass, but this is proba-bly still too speculative to influence clin-ical practice decisions (25). Moresubstantial is evidence that restorationof nearly normal glucose levels withshort-term administration of insulin at

the time of diagnosis can allow recoveryof b-cell function and improvement of in-sulin sensitivity sufficient to induce a clini-cal remission, during which lifestyle effortseither alone or with oral therapy canmain-tain control for some time (26,27).

Insulin is usually needed when diabe-tes is diagnosed in the context of anacute medical event causing acute meta-bolic deterioration, or in case of surgery orany other invasive procedure, temporarilyor for longer term (Table 2 and Fig. 1).Some oral agents may be contraindicatedin these situations, and insulin may beused to provide rapid and reliable controlof the metabolic disturbance.

Lack of response or, even worse, fur-ther deterioration of glycemic control inspite of lifestyle modification and/orglucose-lowering agents should promptconsideration of the diagnosis of latentautoimmune diabetes in adults (LADA)(28). Measurement of glutamic acid de-carboxylase (GAD) antibodies can con-firm the diagnosis, but usually theclinical phenotype of low BMI, low waistcircumference, little or no dyslipidemia(high triglycerides/low HDL cholesterol),

Table 2—“Algorithm” for starting/shifting insulin therapy

1. Is there an acute need?c Ketones (in absence of starvation)/ketoacidosis/dehydrationc Marked hyperglycemia (without carbohydrate/calorie abuse)c Acute medical event with actual/potential decompensationc Marked hyperglycemia with uncertain near-future environment (foreign travel)→Strong immediate needdpersuasive advice may be needed

2. Within ;2 years of diagnosis?c LADA phenotype or secondary pancreatic phenotypec Presence of GAD antibodies/low BMI/waist circumference/no dyslipidemiac Deterioration in glucose control in 6–24 months despite uptitration of multiple oral agenttherapy

c Concomitant disease: pancreatitis, hepatic cirrhosis, chronic steroid therapy, relapsinginflammatory disease, antirejection therapy, other

→Discuss that insulin therapy will be inevitable or make medical self-management easierand saferdfirm recommendation to start now

3. Taking one to four other glucose-lowering therapies and not to target (HbA1c.58 mmol/mol[.7.5%])?c No other explanation for change in glucose controlc Progression of hyperglycemia despite oral agent uptitration or neglected glucose control:c on three or more oral agents/GLP-1RA with HbA1c above 58 mmol/mol (7.5%)c on two or more or more oral agents/GLP-1RA with deterioration of .8 mmol/mol(0.7%) HbA1c since last seen without explanation

c on one or more oral agent(s) with deterioration.11 mmol/mol (1.0%) HbA1c since lastseen without explanation

c Previous patient education given and personal lifestyle input unlikely to improvec Patient preference for injectable natural hormone to drug→Discuss longer-term benefits of blood glucose control to target and efficacy (andtolerability issues) of insulino Consider alternative of a GLP-1RA (in particular if obese and happy to tolerate GIsymptoms)

o With caution if hypoglycemia experienced on sulfonylureas

care.diabetesjournals.org Home and Associates 1501

and lack of family history of diabetes issufficient to raise suspicion. In these pa-tients, insulin therapy can rapidly re-store glycemic control and possiblyslow the decline of insulin secretoryfunction (28).

Presence of Other Nonacute MedicalConditionsPeople with a LADA-like phenotype(lean, relatively young, and no familyhistory of diabetes) but without GADantibodies could have secondary pan-creatic diabetes due to acute or chronicpancreatitis or, more rarely, pancreaticcancer (Table 2). Other conditions po-tentially associated with the need forearly insulin therapy include hepaticcirrhosis (29). In this setting, mealtimeinsulin is often needed to control post-prandial hyperglycemia. People usingglucocorticoid therapy, especially whentaken intermittently at high dosage,

quite often benefit from the power andflexibility of insulin therapy. Starting in-sulin can also ease blood glucosemanage-ment during use of immune-suppressanttherapy or antiretroviral therapies.

Ambulatory Diabetes ManagementThe more typical question is when toadvise starting insulin for people takingtwo or three (or sometimes one or four)oral agents. Critical questions at thispoint are the level of glycemic control,rate of progression of hyperglycemia,available alternatives, and personalpreferences (23). These will be modu-lated by concerns about and perhapspast experience with hypoglycemia(e.g., from prior use of a sulfonylurea),morbid obesity, concomitant disease,and life expectancy, all of which may in-fluence the choice of therapy and per-sonal glycemic targets (23).

Earlier and more rapid progression ofhyperglycemia, perhaps evidenced by amore rapid need for multiple oralagents, suggests greater need for insulin(Table 2, item 3 and Fig. 1). For an in-dividual who appears eligible for a “typ-ical” glycemic target, such as HbA1c,53mmol/mol (,7.0%), it is reasonable tostart thinking about insulin if HbA1c is.58 mmol/mol (7.5%). If HbA1c is .64mmol/mol (8.0%), an additional oralagent will generally not be effective inreaching the target level unless lifestyleeducation has previously been ne-glected or adherence to lifestyle mea-sures might improve, and thus insulinbecomes a more attractive option.Because oral agents have limited effi-cacy, if glucose control is deterioratingrapidly, they may not be able to over-take the rate of deterioration, and inthat case, insulin therapy may bepreferable.

Individual preference may be to delayinsulin by starting another treatment.When this is done, it is desirable to ad-vise the patient that the plan will bereevaluated in 4–6 months. If glycemiccontrol is then unacceptable, insulinshould be recommended, usually withcontinuation of some of the othertherapies.

The stepped algorithm for use of di-abetes therapies with time does seem toget ever more complex, as the range ofoptions increases with newer classes ofglucose-lowering medications (24). Oneadvantage of insulin therapy is often for-gotten here, namely that since it isinevitable for the majority of individu-als, starting it early, when people areonly on one or two oral agents, consid-erably simplifies further management.Essentially recurrent decisions and as-sessments over which other class or in-dividual medication to use are avoidedfor the most part, along with the prob-lems ofmonitoring success and eventual“failure” of each of those other agents.Since however many medications are li-censed in at least some major marketsfor combination with insulin therapy,starting insulin does not preclude the ad-dition of dipeptidyl peptidase-4 (DPP-4)inhibitors, glucagon-like peptide-1 re-ceptor agonists (GLP-1RAs), thiazolidine-diones, or sodium/glucose cotransporter2 (SGLT2) blockers, should a specific in-dication become clear while insulin isbeing used.

Figure 1—Summary of the use of insulin and other glucose-lowering approaches over the courseof time in a person with type 2 diabetes. ause of insulin may be temporary. BP, blood pressure;OGLDs, oral glucose-lowering drugs; TG, triglyceride.

1502 Insulin Therapy in People With Type 2 Diabetes Diabetes Care Volume 37, June 2014

CONSIDERATION OF OTHEROPTIONS

The decision to start insulin therapy willnot usually be taken in isolation. Indeedrather than “should insulin be started?”,the most appropriate question is usually“which of a series of options is mostappropriate for this person already ontreatment but needing better blood glu-cose control?” In the above discussion,we consider situations in which insulinmight be started, but in only a few (e.g.,ketoacidosis or other acute metabolicdisturbance or prospect thereof) will in-sulin be mandatory, although in manycases (pancreatic diabetes, suspectedLADA, or intermittent steroid therapy),it can also be the simplest approach tomanage in the medium-term (Fig. 1).

Further Lifestyle Advice and OralAgentsAt the time of diagnosis, an oral agentcan have quite prominent glucose-lowering effects, especially when initi-ated along with lifestyle instruction/modification. The rapid action of the sul-fonylureas can help achieve a rapid im-provement of glycemic control (30), andthe slower onset of action by metformincan similarly lead to an 11–22mmol/mol(1.0–2.0%) reduction of HbA1c, assistedby reversal of glucose toxicity. As notedabove, guidelines generally recommendinsulin instead if HbA1c is very high,and this would be particularly true iflifestyle already seemed good or therewas suspicion of LADA or pancreatic di-abetes (Fig. 1).In continuing ambulatory care, with

established oral therapy, appropriatelifestyle advice will most often alreadyhave been given, even if full applica-tion of that advice was problematic tothe individual. In these circumstances,delaying insulin or other effective ther-apy while expecting significant furtherchange is unwise. However, when thepatient has had limited or no prior edu-cation and instruction, further lifestylecounseling in conjunction with even amoderately effective additional therapycan achieve good results.Otherwise, all oral agents have lim-

ited efficacy in the setting of failure ofone or more prior oral therapies. Thereduction in HbA1c 12 months aftersupplementing prior oral therapywith a sulfonylurea, metformin, or athiazolidinedione is no more than 5–11

mmol/mol (0.5–1.0%) from a baseline of;64 mmol/mol (8.0%), so very few peo-ple can be expected to get to targetby adding these agents alone (31,32).DPP-4 inhibitors are probably even lesseffective and SGLT2 blockers no moreeffective. At higher HbA1c, failure to at-tain target levels will be inevitable with-out other change. In all cases, reassessmentof glycemic control after 4–6 months isdesirable.

GLP-1RAsAn appropriate option to be consideredin the ambulatory situation where insu-lin is also a possible choice is a GLP-1RA.This does not apply to other circumstan-ces where insulin might be started, suchas more extreme hyperglycemia, otheracute medical illness, or for LADA phe-notype or secondary pancreatic diabe-tes. Advantages compared with insulinmight include lack of need for dose titra-tion after the first few weeks, less fre-quent self-monitoring of glucose, lowerrisk of hypoglycemia, and a favorable ef-fect on body weight (33). Disadvantagesare nausea and sometimes vomiting andother gastrointestinal side effects, as wellas uncertainty over increased risk of pan-creatitis and indeed other long-term ef-fects. The efficacy of these agents is welldocumented even when compared withbasal insulin treatment, although the in-sulin doses used in such studies oftenseem suboptimal (34,35). It is noteworthythat despite frequent reluctance to beginany injected therapy, in both clinical stud-ies and routine practice, insulin is morelikely to be continued after initiation thanGLP-1RAs.

The GLP-1RA advantage of not beingassociated with weight gain and very of-ten leading to weight loss is of particularrelevance given the frequency of obesityamong type 2 patients. Weight gain as-sociated with insulin therapy ranges be-tween 0 and 5 kg in different studies,depending on the baseline HbA1c andthe circumstances in which insulin isstarted (3,7,16). In studies in routinecare, the gain of weight is greatest inpeople who are less obese (36). Al-though weight gain associated with in-sulin therapy is widely seen as anargument against using insulin, clear ev-idence for negative medical outcomesof weight gain is lacking.

Nonetheless, the favorable effect ofGLP-1RAs on weight is attractive to

providers and patients alike. Preliminarydata suggest that this effect is evenmore apparent for people who haveBMI .30.0 kg/m2 (37,38). Therefore,for an obese person with diabetes notmaintaining control targets on lifestyleplus oral agents, the use of a GLP1-RAbefore introducing insulin therapy maybe logical. Obese people on treatmentfor diabetes are rarely able to fullyimplement a desirable change oflifestyle, but on occasion, effectiveweight-reducing therapy can provide astimulus to further change. However, ingeneral, people with higher HbA1c levelswill have less b-cell function, which isneeded for GLP-1RA therapy to be fullyeffective in lowering glucose. Therefore,although glucose reductions are greaterwith the longer-acting GLP-1RAs (lira-glutide and once-weekly exenatide)than with oral agents, it is unreasonableto expect targets to be attained in a ma-jority of people from a baseline HbA1c of.69 mmol/mol (8.5%) unless furtherlifestyle change is made (33).

Combination therapy of insulin andGLP-1RAs appears effective and maygain the glucose-lowering advantagesof both while controlling body weightand reducing risk of hypoglycemia (39).Preliminary evidence suggests that oneroute to injectable therapy might in-volve starting the two in combination,even if this means limited titration ofinsulin doses initially (40). Meanwhile,the addition of either to the other maybe clinically advantageous.

Bariatric SurgeryBariatric surgery has consistently dem-onstrated significant improvement inblood glucose control compared withstandard or intensive medical therapyfor people with diabetes, patients withBMI .35.0 kg/m2, and sometimes theless obese (41–43). Efficacy has beendemonstrated across a wide range ofage and diabetes duration, comorbidities,or need for current glucose-loweringtherapies including insulin. Weightloss after bariatric surgery can be verymarked (average of 25 kg or more) andthe need for glucose-lowering andblood pressure–lowering medicationsreduced. However, these proceduresappear to be less effective in diabeteswhen BMI is ,35.0 kg/m2, when theduration of diabetes is .4 years, orwhen fasting C-peptide levels are

care.diabetesjournals.org Home and Associates 1503

,0.96 nmol/L (,2.9 ng/mL) (44,45).Older people appear likely to havegreater risks associated with these sur-gical procedures (46,47).This suggests that bariatric proce-

dures deserve consideration for obesebut otherwise vigorous people withtype 2 diabetes who have disappointingresponses to usual glucose-loweringtherapies. That bariatric surgery is of-fered to obese people without diabetestells us that people with diabetes mightbenefit if obesity itself is a sufficientindication. For obese people with dia-betes already taking insulin, whosecontinuing weight gain appears to bepreventing attainment of glucose con-trol, bariatric surgery may be an optionworth considering. However, due to lackof long-term follow-up of outcomes,bariatric surgery does not at the presenttime seem appropriate as an alternativeto insulin and other validated therapiesfor hyperglycemia but rather as a sup-plement to them in controlling obesity(48).

A GUIDE FOR THE DECISION TOSTART INSULIN

Most clinicians do not think algorithmi-cally when managing clinical conditionsand advising people with diabetes. Theyoften prefer to follow a patient-ledagenda, and individuals will highlighttheir problems and preferences for sol-utions in quite different ways. As aresult, a simple algorithm for startinginsulin is not feasible. However, it is pos-sible to provide a checklist thatmay helpto guide the clinician-patient interactionto ensure that decisions occur in a logi-cal way, and importantly without miss-ing the opportunity to obtain relevantinformation (Table 2).The first consideration might be to

assess whether an acute need is pres-ent, although usually that will be obvi-ous. At diagnosis, other referral, orwhen admitted for whatever reason asan in-patient, the presence or absenceof marked hyperglycemia, weight loss,ketones, ketoacidosis, or dehydrationmust be ascertained. If marked hyper-glycemia alone is present, is there anacute precipitating factor, and if not, isthere any prospect that glycemic controlcan be restored by lifestyle change? Isthe patient in a risky or uncertain envi-ronment? In these cases, there may be astrong, immediate need, and persuasive

advice to consider insulin may be appro-priate (Fig. 1).

Then, consider whether the LADAphenotype, pancreatic diabetes, or he-patic cirrhosis may be present. GAD an-tibody levels may be requested or beavailable. Deterioration in glucose con-trol over the next 6–24 months despiteuptitration of oral agents is perhaps themost important indicator. That shouldlead to a discussion that insulin therapyis inevitable; in time, it will make man-agement easier and safer.

In longer-term diabetes manage-ment, has hyperglycemia progressedrapidly or glucose control been ne-glected recently? Here the discussion in-cludes alternatives and the likelyscenarios for such therapies. This needsdiscussion of the effects and tolerabilityof insulin, including hypoglycemia, risksof weight gain, perhaps together withthe eventual inevitability of the needfor insulin. Whether the introductionof insulin will affect an individual’s abil-ity to drive or maintain other key lifeactivities, especially related to occupa-tion, should be explored. Also consideralternatives, including GLP-1RAs, andtheir attractive features and problems.It is important not to present insulin asindicative of failure or a punishment.

At any stage, the preferred optionshould be negotiated, with in mostcases a contingency planning option ifpersonal targets are not met in a periodofmonths. A further issue here concernscontinuing of oral agent therapy, ormore specifically sulfonylureas, as met-formin is usually continued. If a regimenincluding mealtime insulin is chosen (in-cluding premixes), sulfonylureas areusually stopped to avoid compoundingthe hypoglycemia risk, but there is lesscertainty when beginning basal insulinalone where at least temporarily glu-cose control will deteriorate withoutrapid insulin dose titration (49). If hypo-glycemia does occur on the combinationof insulin and a sulfonylurea, clinical ex-perience is that the sulfonylurea shouldbe stopped rather than insulin doses re-duced. Thiazolidinediones are also oftenstopped when starting insulin as theedema and weight gain risks are com-pounded, but may be worth continuingin the very obese insulin-insensitive in-dividual. A DPP-4 inhibitor can be con-tinued, although perhaps to littlebenefit.

FURTHER OPTIMIZATION ANDPERSONALIZATION OF INSULINTHERAPY

After systematic application of lifestyle,oral agent, and injectable therapies fortype 2 diabetes, some patients remainunable to achieve or maintain sufficientglycemic control to avoid the onset orprogression of glycemia-related compli-cations. Specifically, even after adoptionof initial lifestyle changes, trying ade-quate courses of oral glucose-loweringagents, starting and optimizing insulintherapy, and consideration of GLP-1RAs, a significant proportion of peoplehave HbA1c levels .53 mmol/mol(7.0%). In the population enrolled inthe Action to Control CardiovascularRisk in Diabetes (ACCORD) trial, whichhad both 10 years average duration ofdiabetes and high cardiovascular risk,about 25% of individuals assigned to in-tensive glucose-lowering therapy wereunable to maintain HbA1c levels ,52mmol/mol (7.0%) (50). The further ob-servation that this subgroup, in whichinsulin therapy was expected to be op-timized, had both higher risk of hypogly-cemia and higher risk of all-causemortality has focused attention on theclinical dilemmas posed by such people.A general conclusion is that such peopleneed further consideration of the aimsin personalizing their therapy regimens,with the goal of balancing the gains andlosses in risk and quality of life frommedications. The difficulty in translatingthis conclusion into practical clinicalguidance is well recognized and was re-flected by discussions at this expert fo-rum. Several questions were addressedto focus these discussions.

How Can This Category of Patients BeDefined and Identified?An analysis of data from ACCORD soughtto identify baseline characteristics thatwere associated with excess mortalityaccompanying intensive as comparedwith standard treatment. Just three clin-ical predictors emerged: baseline levelsof HbA1c .69 mmol/mol (.8.5%), anyprior history of neuropathy, and currentaspirin use (51). It is possible the studywas too short to have the power to iden-tify other predictors. Poor glycemic con-trol at entry to the study seems likely tohave indicated preexisting difficultywith glycemic management, and this initself may have been associated with

1504 Insulin Therapy in People With Type 2 Diabetes Diabetes Care Volume 37, June 2014

other factors driven out by the multivar-iable analysis, such as comorbidities.The other “historical” findings (neurop-athy and aspirin) imply the presence ofalready known or suspected complica-tions of diabetes, microvascular andmacrovascular. In addition, a subse-quent analysis of on-treatment experi-ence in ACCORD indicated that inability toreduce HbA1c by at least by 5 mmol/mol(0.5% units) from the baseline level withan intensive treatment regimen corre-lated with risk of excess mortality (50).Together these observations suggestthat high medical risk (among peoplealready known to have high cardio-vascular risk) accompanying intensivetherapy can be identified quite early,either by finding poor glycemic controlbefore treatment optimization is at-tempted or by observing a disappoint-ing response to the first efforts toimprove treatment. However, as notedabove, many people in poor glucosecontrol have a very good response toinsulin therapy, so it is important todistinguish difficulty in obtaining con-trol from neglect of good control.Stated otherwise, people who have lit-tle success at first are likely to havedifficulty later and are at high risk ofdiabetes-associated and comorbid ad-verse outcomes.Where response to initial insulin ther-

apy is poor, or is initially satisfactory butthen deteriorates with time due to pro-gressive islet b-cell failure, the multi-tude of insulin types do offer variousways to improve clinical results. How-ever, where basal insulin alone is beingused, a first action is to ensure the dosehas been titrated to usual fasting plasmaglucose targets. Use of more extendedself-monitoring can then establish thepattern of daytime and nocturnal glu-cose levels, allowing an informed deci-sion as to the addition of one or moremealtime insulin doses. If simplicity butless flexibility is judged desirable, aswitch to premix insulin may be pre-ferred. Coprescription of a GLP-1RA(see GLP-1RAs) is another option.

How Should Glycemic Targets BeAltered?The simplest and most often discussedmodification of usual therapy for peoplethought to be at high risk is to modifythe HbA1c target range. Instead of seek-ing HbA1c ,53 mmol/mol (,7.0%),

higher-risk individuals are commonlyadvised to aim for between 53 and64 mmol/mol (7.0 and 8.0%) (23). Al-though this was the target range forthe standard treatment arm in ACCORD,evidence arguing that this range is al-ways the most appropriate remains lim-ited. For example, individuals withlimited life expectancy and havingHbA1c of 75–86 mmol/mol (9.0–10.0%)probably have little to gain from stren-uous efforts to reduce this value to,64mmol/mol (8.0%). Similarly, people withserious medical illnesses requiring com-plex treatments, such as those withcancer undergoing treatment with cyto-toxic regimens, might not need to seekeven a relaxed HbA1c target range (23).However, very high HbA1c levels (per-haps .86 mmol/mol [.10.0%]) are as-sociated with an acute risk of increasedinfection and vascular thrombosis asnoted above (52,53), as well as tired-ness, weight loss, and inconvenientpolyuria.

How Can Therapeutic Approaches BeRevised?Even when individuals with conditionsor circumstances allowing exemptionfrom specific glycemic targets are re-moved from discussion, a sizable groupof people who have no apparent reasonnot to attainHbA1c in the53–64mmol/mol(7.0 to 8.0%) range remains. Insulin ther-apy is often said to be unlimited in itscapacity to lower glucose levels, but inpractice, even very high prescribeddoses sometimes yield results thatfall short of expectations (54). The un-derlying causes of failure of usual treat-ments are undoubtedly numerous, and to

understand them calls for further effortto identify the personal characteristics ofeach person that may prove relevant(Table 3). In many cases, progressiveobesity, as a marker for high calorie in-take and insulin resistance, identifies ametabolic challenge that resists successeven when ample insulin is delivered totissues. Other medical conditions maybe important. Examples include unrec-ognized Cushing syndrome or a geneticor acquired disorder of extreme insulinresistance.

For some people, psychological fac-tors may interfere with adherence tothe regimen or lead to very poor deci-sions on the timing and dosage of insu-lin. Obtaining accurate informationabout actual use of insulin and othermedications, independent of what hasbeen prescribed, can be very challeng-ing. For others, environmental pres-sures, including financial constraints,family or work-related conflict, or socialisolation, may prove to be centralfactors.

Only occasionally will addition of anew medication or change of dosageof existing ones then solve the problemat hand, but consideration of addition ofGLP-1RAs, DPP-4 inhibitors, pioglita-zone, or SGLT2 blockers is appropriatein some cases (licensed indications varyby market) (39,55,56). In general, allthese further lower glucose in theshort-term, but this effect may be coun-tered by insulin dose reduction withtime. Medications such as GLP-1RAsand SGLT2 blockers can reduce the ef-fects of insulin onweight gain. Pioglitazonecan sometimes have dramatic effects onvery high insulin dose requirements, but

Table 3—Dealing with problems when on insulin therapy

1. Failure to improve glucose control after adequate dose titrationc Presence of comorbidities, prior control difficulties, occurrence of hypoglycemiadhigher-risk person for amelioration of glucose control targets?

c Adherence and acceptance of insulin issues?

2. Excessive weight gainc Implies high calorie intakedreview lifestyle issues and consider a trial of adding GLP1-RAor SGLT2 inhibitor therapy

3. High insulin dose requirement in the markedly obese personc Consider trial of adding a thiazolidinedione

4. Severe recurring hypoglycemic eventsc Nocturnaldconsider using long-acting insulin analog, timing of basal insulin injection,possibility of hepatic cirrhosis or steroid therapy with need for predominant meal-timeinsulin therapy

c Daytimedconsider contribution of meal-time insulin (reduce or move to basal-onlyregimen) or lifestyle issues (missed meals or alcohol)

care.diabetesjournals.org Home and Associates 1505

at the expense of further weight gain.These combinations can be very expen-sive or of unproven safety and so cur-rently seem appropriate for a minorityof insulin takers (57).When evidence or suspicion of re-

peated hypoglycemia is present, achange of insulin type or distributionmight be beneficial. For people usingNPH as basal insulin, the long-acting in-sulin analogs (glargine, detemir, and de-gludec) offer evidence-based reductionin risk of nocturnal hypoglycemia(13,14,58,59). For people with daytimehypoglycemia when using sulfonylureaor mealtime insulin together with basalinsulin, consideration can be given toreducing the dosage of these agents orstopping them entirely while continuingbasal insulin. Substituting a shorter-acting GLP-1RA for mealtime insulin isanother option (39). Appropriate life-style adjustments to limit hypoglycemiacaused by insulin may also be helpful.Exploring individual factors and their

interactions lies at the center of person-alized treatment and poses a more dif-ficult challenge in this setting than at thetime of starting standard treatments foran unselected population of patients(60). Personalized use of insulin andother therapies by people who have al-ready demonstrated little success with amore generic approach requires time,expertise, and motivation.

SUMMARY AND CONCLUSIONS

The group’s discussion noted that insu-lin has been used longer than any othermedication for treating diabetes, andthe body of evidence supporting a favor-able balance of benefits to risks of its usecontinues to grow. Some large, long-term clinical trials, including the UKPDS,Diabetes Control and ComplicationsTrial/Epidemiology of Diabetes Interven-tions and Complications (DCCT/EDIC),and ORIGIN, have shown that both mi-crovascular and cardiovascular benefitscan be obtained. Initiation of insulintherapy is appropriate at various stagesof diabetes, depending on the clinicalcircumstances of each case. Use as thefirst form of pharmacotherapy is neededwhen the diagnosis of diabetes is madeat a time of acute illness or severe met-abolic decompensation, or when a formof diabetes characterized by marked in-sulin deficiency is present. Later, whenlifestyle intervention and oral therapies

are no longer fully successful, insulin isusually the most desirable first injectedtherapy, but GLP-1RAs should be consid-ered in some cases because of theirunique ability to limit weight gain andreduce glucose with little risk of hypo-glycemia. Some large interventionalstudies, notably ACCORD, suggest thatsubgroups with less potential benefitand higher risk of hypoglycemia andother complications of insulin use shouldbe identified and managed differently.Although evidence remains incomplete,a history of established complications ofdiabetes, long-term difficulty with glyce-mic control, and a poor response to re-cent efforts to intensify control all arguefor adjustment of either the HbA1c targetrange or therapeutic tactics, or both.Further information on the use of com-binations of therapies, especially includ-ing the newer agents such as GLP-1RAs,DPP-4 inhibitors, and SGLT2 blockers, forthis problematic subgroup of patients isneeded. Finally, this discussion clearlydemonstrates that despite its longhistory, much remains to be learnedabout the best ways to use insulin ther-apy, which continues to be centrally im-portant in the management of type 2diabetes.

Duality of Interest. P.H. has received (orinstitutions with which he is associated havereceived) funding for his educational, advisory,and research activities from AstraZeneca/BMS Col-laboration, Eli Lilly and Company, GlaxoSmithKline,Janssen/Johnson & Johnson, Merck Sharp &Dohme, Novo Nordisk, Roche Diagnostics, RochePharmaceuticals, Sanofi, and Takeda. M.R. hasreceived honoraria for consulting and/or speak-ing from Eli Lilly and Company, Elcelyx, Sanofi,and Valeritas and research grant supportthrough Oregon Health & Science Universityfrom Amylin/Bristol-Myers Squibb/AstraZeneca,Eli Lilly and Company, and Sanofi. This duality ofinterest has been reviewed and managed byOregon Health & Science University. W.T.C. hasserved as principal investigator for research stud-ies awarded to his institution from MannKind,AstraZeneca, GlaxoSmithKline, Bristol-MyersSquibb, and Sanofi and served as a consultantfor Intarcia. C.J.B. has undertaken ad hoc consul-tancy for Bristol-Myers Squibb, AstraZeneca,Merck Sharp & Dohme, Novo Nordisk, Sanofi,Janssen, Eli Lilly and Company, Roche, andTakeda; delivered continuing medical educationprograms sponsored by Bristol-Myers Squibb,AstraZeneca, GlaxoSmithKline, Merck Serono,Merck Sharp & Dohme, Eli Lilly and Company,and Boehringer Ingelheim; and received travelor accommodation reimbursement from Astra-Zeneca and Bristol-Myers Squibb. R.G.B. has re-ceived funding for his research activities from

Merck Sharp & Dohme and Sanofi. He was a con-sultant to Merck Sharp & Dohme and Sanofi. Heserved on the advisory boards of AstraZeneca/BMS Collaboration, GlaxoSmithKline, MerckSharp & Dohme, Novo Nordisk, and Sanofi, andwas on the speakers’ bureau for Bayer, Eli Lillyand Company, Merck Sharp & Dohme, Novartis,Novo Nordisk, Roche, and Sanofi. S.d.P. hasserved on the following advisory panels:AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, Eli Lilly and Company, GI DynamicsInc., GlaxoSmithKline, Hanmi Pharmaceutical, In-tarcia Therapeutics Inc., Janssen Pharmaceuticals,Merck Sharp & Dohme, Novartis, Novo Nordisk,Roche Diagnostics, Sanofi, and Takeda. He re-ceived research support from Bristol-MyersSquibb, Merck Sharp & Dohme, Novartis, andNovoNordisk. D.L. has consulted for Sanofi,Merck,Janssen, and AstraZeneca/BMS. G.S. served on theadvisory board of Amgen, AstraZeneca, BoehringerIngelheim, Bristol-Myers Squibb, Eli Lilly andCompany, Janssen/Johnson & Johnson, MerckSharp & Dohme, Novartis, Novo Nordisk, Poxel,Sanofi, and Takeda and was on the speaker’sbureau for AstraZeneca, Boehringer Ingelheim,Bristol-Myers Squibb, Eli Lilly and Company,Janssen/Johnson & Johnson, Merck Sharp &Dohme, Novo Nordisk, Sanofi, and Takeda.L.v.G. is/has been a member of the advisoryboard and speakers’ bureau of AstraZeneca/BMS, Boehringer Ingelheim, Eli Lilly and Com-pany, Janssen/Johnson & Johnson, Merck Sharp& Dohme, Novartis, Novo Nordisk, and Sanofi(period 2011–2013). He also received grant sup-port from the European Union (Hepadip & Re-solve Consortium) and National Research Funds,Belgium. I.R. served on the advisory board ofNovo Nordisk, AstraZeneca/BMS, Sanofi, MerckSharp &Dohme, Eli Lilly and Company, andMed-scape. He was a consultant to AstraZeneca/BMS,Insuline, and Andromeda Biotech Ltd. and wason the speakers’ bureau for Novo Nordisk,AstraZeneca/BMS, Sanofi, Merck Sharp &Dohme, Eli Lilly and Company, and Novartis.No other potential conflicts of interest relevantto this article were reported.Author Contributions. All authors contrib-uted to the ideas and discussions underpinningthe manuscript and to writing the manuscriptand review of revisions.

References1. Brown JB, Nichols GA, Perry A. The burden oftreatment failure in type 2 diabetes. DiabetesCare 2004;27:1535–15402. Shah BR, Hux JE, Laupacis A,MdcmBZ, AustinPC, vanWalraven C. Diabetic patients with priorspecialist care have better glycaemic controlthan those with prior primary care. J Eval ClinPract 2005;11:568–5753. Home P, Naggar NE, Khamseh M, et al. Anobservational non-interventional study of peo-ple with diabetes beginning or changed to in-sulin analogue therapy in non-Westerncountries: the A1chieve study. Diabetes ResClin Pract 2011;94:352–3634. Peyrot M, Rubin RR, Lauritzen T, et al.; In-ternational DAWN Advisory Panel. Resistanceto insulin therapy among patients and pro-viders: results of the cross-national DiabetesAttitudes, Wishes, and Needs (DAWN) study.Diabetes Care 2005;28:2673–2679

1506 Insulin Therapy in People With Type 2 Diabetes Diabetes Care Volume 37, June 2014

5. U.K. Prospective Diabetes Study Group. U.K.prospective diabetes study 16. Overview of6 years’ therapy of type II diabetes: a progressivedisease. Diabetes 1995;44:1249–12586. Lutgers HL, Gerrits EG, Sluiter WJ, et al. Lifeexpectancy in a large cohort of type 2 diabetespatients treated in primary care (ZODIAC-10).PLoS ONE 2009;4:e68177. UK Prospective Diabetes Study (UKPDS)Group. Intensive blood-glucose control with sul-phonylureas or insulin compared with conven-tional treatment and risk of complications inpatients with type 2 diabetes (UKPDS 33). Lan-cet 1998;352:837–8538. UK Prospective Diabetes Study (UKPDS)Group. Effect of intensive blood-glucose controlwith metformin on complications in overweightpatients with type 2 diabetes (UKPDS 34). Lan-cet 1998;352:854–8659. Gerstein HC, Bosch J, Dagenais GR, et al.;ORIGIN Trial Investigators. Basal insulin andcardiovascular and other outcomes in dysgly-cemia. N Engl J Med 2012;367:319–32810. Holman RR, Paul SK, Bethel MA, MatthewsDR, Neil HAW. 10-year follow-up of intensiveglucose control in type 2 diabetes. N EnglJ Med 2008;359:1577–158911. Nathan DM, Cleary PA, Backlund JY, et al.;Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions andComplications (DCCT/EDIC) Study ResearchGroup. Intensive diabetes treatment and car-diovascular disease in patients with type 1 di-abetes. N Engl J Med 2005;353:2643–265312. Shah S, Zilov A,MalekR, Soewondo P, BechO,Litwak L. Improvements in quality of life associ-ated with insulin analogue therapies in peoplewith type 2 diabetes: results from the A1chieveobservational study. Diabetes Res Clin Pract 2011;94:364–37013. Riddle MC, Rosenstock J, Gerich J; InsulinGlargine 4002 Study Investigators. The treat-to-target trial: randomized addition of glargineor human NPH insulin to oral therapy of type 2diabetic patients. Diabetes Care 2003;26:3080–308614. Hermansen K, Davies M, Derezinski T,Martinez Ravn G, Clauson P, Home P; on behalfof the Levemir Treat-to-Target Study Group. A26-week, randomized, parallel, treat-to-targettrial comparing insulin detemir with NPH insulinas add-on therapy to oral glucose-loweringdrugs in insulin-naive people with type 2 diabe-tes. Diabetes Care 2006;29:1269–127415. Yki-Jarvinen H, Kauppinen-Makelin R,Tiikkainen M, et al. Insulin glargine or NPH com-bined with metformin in type 2 diabetes: theLANMET study. Diabetologia 2006;49:442–45116. Holman RR, Farmer AJ, Davies MJ, et al.; 4-TStudy Group. Three-year efficacy of complex in-sulin regimens in type 2 diabetes. N Engl J Med2009;361:1736–174717. Home P, Dain M-P, Freemantle N, et al.Four-year evolution of insulin regimens, glycae-mic control, hypoglycaemia and body weightafter starting insulin therapy in type 2 diabetesacross three continents (Abstract). Diabetologia2013;56(Suppl. 1):S41018. Dashora UK, Sibal L, Ashwell SG, Home PD.Insulin glargine in combination with nateglinidein people with type 2 diabetes: a randomized

placebo-controlled trial. Diabet Med 2007;24:344–34919. Riddle MC, Rosenstock J, Vlajnic A, Gao L.Randomized, 1-year comparison of three waysto initiate and advance insulin for type 2 diabe-tes: twice-daily premixed insulin versus basalinsulin with either basal-plus one prandial insu-lin or basal-bolus up to three prandial injections.Diabetes Obes Metab 2014;16:396–40220. Holman RR, Thorne KI, Farmer AJ, et al.; 4-TStudy Group. Addition of biphasic, prandial, orbasal insulin to oral therapy in type 2 diabetes. NEngl J Med 2007;357:1716–173021. International Diabetes Federation ClinicalGuidelines Task Force. Global guidelines fortype 2 diabetes. Available from http://www.idf.org/ global-guideline-type-2-diabetes-2012.Accessed 5 December 201222. Inzucchi SE, Bergenstal RM, Buse JB, et al.;American Diabetes Association (ADA); Euro-pean Association for the Study of Diabetes(EASD). Management of hyperglycemia in type 2diabetes: a patient-centered approach: positionstatement of the American Diabetes Association(ADA) and the EuropeanAssociation for the Studyof Diabetes (EASD). Diabetes Care 2012;35:1364–137923. Raz I, Riddle MC, Rosenstock J, et al. Per-sonalized management of hyperglycemia intype 2 diabetes: reflections from a DiabetesCare Editors’ Expert Forum. Diabetes Care2013;36:1779–178824. Canadian Diabetes Association. 2013 Clini-cal Practice Guidelines, Pharmacologic manage-ment of type 2 diabetes. Can J Diabetes 2013;37:S61–S6825. Retnakaran R, Zinman B. Short-term inten-sified insulin treatment in type 2 diabetes: long-term effects on b-cell function. Diabetes ObesMetab 2012;14(Suppl. 3):161–16626. Hu Y, Li L, Xu Y, et al. Short-term intensivetherapy in newly diagnosed type 2 diabetes par-tially restores both insulin sensitivity and b-cellfunction in subjects with long-term remission.Diabetes Care 2011;34:1848–185327. Chen HS, Wu TE, Jap TS, Hsiao LC, Lee SH,Lin HD. Beneficial effects of insulin on glycemiccontrol and b-cell function in newly diagnosedtype 2 diabetes with severe hyperglycemia aftershort-term intensive insulin therapy. DiabetesCare 2008;31:1927–193228. Pozzilli P, DiMario U. Autoimmune diabetesnot requiring insulin at diagnosis (latent auto-immune diabetes of the adult): definition, char-acterization, and potential prevention. DiabetesCare 2001;24:1460–146729. Kruszynska YT, Harry DS, Bergman RN,McIntyre N. Insulin sensitivity, insulin secretionand glucose effectiveness in diabetic and non-diabetic cirrhotic patients. Diabetologia 1993;36:121–12830. Kahn SE, Haffner SM, Heise MA, et al.;ADOPT Study Group. Glycemic durability of ro-siglitazone, metformin, or glyburide monother-apy. N Engl J Med 2006;355:2427–244331. Phung OJ, Scholle JM, Talwar M, ColemanCI. Effect of noninsulin antidiabetic drugs addedto metformin therapy on glycemic control,weight gain, and hypoglycemia in type 2 diabe-tes. JAMA 2010;303:1410–141832. Bolen S, Feldman L, Vassy J, et al. System-atic review: comparative effectiveness and

safety of oral medications for type 2 diabetesmellitus. Ann Intern Med 2007;147:386–39933. Diamant M, Van Gaal L, Stranks S, et al.Once weekly exenatide compared with insulinglargine titrated to target in patients with type 2diabetes (DURATION-3): an open-label rando-mised trial. Lancet 2010;375:2234–224334. Heine RJ, Van Gaal LF, Johns D, Mihm MJ,Widel MH, Brodows RG; GWAA Study Group.Exenatide versus insulin glargine in patientswith suboptimally controlled type 2 diabetes:a randomized trial. Ann Intern Med 2005;143:559–56935. Nauck MA, Duran S, Kim D, et al. A compar-ison of twice-daily exenatide and biphasic insu-lin aspart in patients with type 2 diabetes whowere suboptimally controlled with sulfonylureaand metformin: a non-inferiority study. Diabe-tologia 2007;50:259–26736. Home P,Malek R, Prusty V, Latif ZA, HaddadJ. Impact of insulin detemir on weight change inrelation to baseline BMI: observations from theA1chieve study (Abstract). Poster 963-P pre-sented at 73rd Scientific Sessions of the Amer-ican Diabetes Association, 21–25 June 2013,Chicago, IL37. Vilsbøll T, Christensen M, Junker AE, KnopFK, Gluud LL. Effects of glucagon-like peptide-1receptor agonists on weight loss: systematic re-view and meta-analyses of randomised con-trolled trials. BMJ 2012;344:d777138. Barnett AH. The role of GLP-1 mimetics andbasal insulin analogues in type 2 diabetes melli-tus: guidance from studies of liraglutide. Diabe-tes Obes Metab 2012;14:304–31439. Buse JB, Bergenstal RM, Glass LC, et al. Useof twice-daily exenatide in Basal insulin-treatedpatients with type 2 diabetes: a randomized,controlled trial. Ann Intern Med 2011;154:103–11240. Buse JB, Gough SC, Woo VC, et al.IDegLira, a novel fixed ratio combination of in-sulin degludec and liraglutide, is efficacious andsafe in subjects with type 2 diabetes: a large,randomized phase 3 trial (Abstract). Diabetes2013;62(Suppl. 1):6541. Buchwald H, Estok R, Fahrbach K, et al.Weight and type 2 diabetes after bariatric sur-gery: systematic review and meta-analysis. Am JMed 2009;122:248–256, e542. Reis CE, Alvarez-Leite JI, Bressan J, AlfenasRC. Role of bariatric-metabolic surgery in thetreatment of obese type 2 diabetes with bodymass index ,35 kg/m2: a literature review. Di-abetes Technol Ther 2012;14:365–37243. Schauer PR, Kashyap SR, Wolski K, et al.Bariatric surgery versus intensive medical ther-apy in obese patients with diabetes. N Engl JMed 2012;366:1567–157644. Scopinaro N, Adami GF, Papadia FS, et al.The effects of biliopancreatic diversion on type 2diabetes mellitus in patients with mild obesity(BMI 30-35 kg/m2) and simple overweight (BMI25-30 kg/m2): a prospective controlled study.Obes Surg 2011;21:880–88845. Scopinaro N, Adami GF, Papadia FS, et al.Effects of biliopanceratic diversion on type 2diabetes in patients with BMI 25 to 35. AnnSurg 2011;253:699–70346. Oreopoulos A, Kalantar-Zadeh K, SharmaAM, Fonarow GC. The obesity paradox inthe elderly: potential mechanisms and clinical

care.diabetesjournals.org Home and Associates 1507

implications. Clin Geriatr Med 2009;25:643–659, viii47. Ramanan B, Gupta PK, Gupta H, Fang X,Forse RA. Development and validation of a bari-atric surgery mortality risk calculator. J Am CollSurg 2012;214:892–90048. Van Gaal LF, De Block CE. Bariatric surgeryto treat type 2 diabetes: what is the recent ev-idence? Curr Opin Endocrinol Diabetes Obes2012;19:352–35849. National Institute for Clinical Excellence.Type 2 diabetes: the management of type 2 di-abetes. Available from www.nice.org.uk/cg66.Accessed 31 December 201350. Riddle MC, Ambrosius WT, Brillon DJ, et al.;Action to Control Cardiovascular Risk in Diabe-tes Investigators. Epidemiologic relationshipsbetween A1C and all-cause mortality during amedian 3.4-year follow-up of glycemic treat-ment in the ACCORD trial. Diabetes Care 2010;33:983–99051. Calles-Escandon J, Lovato LC, Simons-MortonDG, et al. Effect of intensive compared with stan-dard glycemia treatment strategies on mortalityby baseline subgroup characteristics: the Action toControl Cardiovascular Risk in Diabetes (ACCORD)trial. Diabetes Care 2010;33:721–72752. Juhan I, Vague P, Buonocore M, Moulin JP,Jouve R, Vialettes B. Abnormalities of erythro-cyte deformability and platelet aggregation in

insulin-dependent diabetics corrected by insu-lin in vivo and in vitro. Lancet 1982;1:535–53753. Pozzilli P, Leslie RDG. Infections, immunity,and diabetes. In International Textbook ofDiabetes Mellitus. 3rd ed. DeFronzo R,Ferrannini E, Keen H, Zimmet P, Eds. Colchester,John Wiley, 2004, p. 1729–173954. Wilding JPH, Norwood P, T’joen C, BastienA, List JF, Fiedorek FT. A study of dapagliflozin inpatients with type 2 diabetes receiving highdoses of insulin plus insulin sensitizers: applica-bility of a novel insulin-independent treatment.Diabetes Care 2009;32:1656–166255. Bailey CJ. Renal glucose reabsorption inhib-itors to treat diabetes. Trends Pharmacol Sci2011;32:63–7156. Charbonnel B, Schweizer A, Dejager S. Com-bination therapy with DPP-4 inhibitors and in-sulin in patients with type 2 diabetes mellitus:what is the evidence? Hosp Pract (1995) 2013;41:93–10757. Aljabri K, Kozak SE, Thompson DM. Additionof pioglitazone or bedtime insulin to maximaldoses of sulfonylurea and metformin in type 2diabetes patients with poor glucose control:a prospective, randomized trial. Am J Med2004;116:230–23558. Lee P, Chang A, Blaum C, Vlajnic A, Gao L,Halter J. Comparison of safety and efficacy ofinsulin glargine and neutral protamine

hagedorn insulin in older adults with type 2 di-abetes mellitus: results from a pooled analysis. JAm Geriatr Soc 2012;60:51–5959. Ratner RE, Gough SC, Mathieu C, et al. Hy-poglycaemia risk with insulin degludec comparedwith insulin glargine in type 2 and type 1 diabetes:a pre-planned meta-analysis of phase 3 trials.Diabetes Obes Metab 2013;15:175–18460. Riddle MC, Karl DM. Individualizing targetsand tactics for high-risk patients with type 2 di-abetes: practical lessons from ACCORD andother cardiovascular trials. Diabetes Care2012;35:2100–210761. Van den Berghe G, Wilmer A, Hermans G,et al. Intensive insulin therapy in the medicalICU. N Engl J Med 2006;354:449–46162. Ray KK, Seshasai SRK, Wijesuriya S, et al.Effect of intensive control of glucose on cardio-vascular outcomes and death in patients with di-abetes mellitus: a meta-analysis of randomisedcontrolled trials. Lancet 2009;373:1765–177263. Janka HU, Plewe G, RiddleMC, Kliebe-FrischC, Schweitzer MA, Yki-Jarvinen H. Comparisonof basal insulin added to oral agents versustwice-daily premixed insulin as initial insulintherapy for type 2 diabetes. Diabetes Care2005;28:254–25964. Alberti KG, Hockaday TD, Turner RC. Smalldoses of intramuscular insulin in the treatmentof diabetic “coma.” Lancet 1973;2:515–522

1508 Insulin Therapy in People With Type 2 Diabetes Diabetes Care Volume 37, June 2014