Since 2012,⁹ treatment algorithms for the management of patients with T2DM have followed the approach to patient-centered care established by the Committee on Quality of Health Care in America: “providing care that is respectful of and responsive to individual patient preferences, needs, and values and ensuring that patient values guide all clinical decisions.”¹¹ Diabetes management includes the setting of individualized glucose targets to achieve HbA_1c levels as close to normal as possible in patients who are most likely to benefit from good glycemic control (ie, those without clinical evidence of macrovascular complications) while also minimizing the possible risks of hypoglycemia. It also includes relaxing targets in patients with limited life expectancy, in patients with existing diabetes complications or longer duration of disease, and in those for whom there is greater concern about the development of hypoglycemia.^10,12 

When prescribing insulin therapy, the astute osteopathic physician implements the most appropriate form on the basis of the insulin type, dose, and delivery device used.^13,14 For most patients whose HbA_1c levels are not at goal, the simplest first step is to start insulin therapy with a single injection of a long-acting basal insulin analog.⁹ Basal insulin suppresses hepatic glucose production overnight and between meals. It constitutes approximately 50% of the daily insulin needs of an individual.⁹ Once-daily use of a basal insulin analog (insulin glargine or insulin detemir) offers the advantages of simple dosing and ease of titration (ie, patients can learn to manage their T2DM with limited training).^15-18 This treatment is highly effective in improving glycemic control in patients who no longer respond to combination oral antidiabetic therapy.¹³ The goal of basal insulin analog therapy is to improve fasting blood glucose levels. Typical starting doses are 10 to 20 U of insulin glargine or insulin detemir given once daily (or, as an alternative, 0.2 U/kg).⁹ Neutral protamine Hagedorn (NPH) insulin can be a more economical option, but physicians should be aware that NPH insulin is more of an intermediate-acting insulin and that it therefore must be dosed 2 to 3 times per day to serve as a basal insulin. Physicians should also be aware that NPH insulin is associated with a greater risk of hypoglycemia, especially nocturnal hypoglycemia, than are basal insulin analogs.¹⁹ 

Oral antidiabetic agents are usually continued when insulin is started, unless there are specific contraindications or substantial risks of hypoglycemia (in some cases, the dose of sulfonylureas may be decreased or discontinued).⁹ This continuance of therapy helps patients avoid the loss of further glycemic control during the transition to insulin. The insulin dose should be titrated on the basis of a fasting blood glucose target. The American Diabetes Association recommends a goal of less than 130 mg/dL.²⁰ A patient should expect an approximately 0.5% decrease in the HbA_1c level for each insulin dose increment of 0.1 U/kg per day. Basal insulins can be self-titrated up to either a target fasting blood glucose level or an approximate dose of 0.5 U/kg per day. At higher doses, the improvement in the decrease in the HbA_1c level is less substantial, and the risk of hypoglycemia increases.²¹ If, after sufficient time, the HbA_1c level still has not reached goal with the use of 0.5 U of basal insulin per kilogram per day, then attention should be focused on mealtime or prandial glucose excursions. If excursions are present, consider adding an agent to target postprandial hyperglycemia, which is the likely cause of persistent hyperglycemia. This can be confirmed by having the patient check his or her glucose level 2 hours after the meal.²² A number of options exist for the treatment of prandial hyperglycemia, including prandial insulin (eg, a fast-acting insulin or rapid-acting insulin analog, such as insulin lispro, insulin aspart, or insulin glulisine) or an incretin-based therapy (eg, a dipeptidyl peptidase-4 [DPP-4] inhibitor [such as sitagliptin, saxagliptin, or linagliptin] or a glucagon-like peptide-1 [GLP-1] receptor agonist [such as exenatide (administered twice per day), liraglutide (administered once per day), or exenatide (administered weekly)]). Exenatide extended release is not currently approved by the US Food and Drug Administration for use with insulin therapy. 

Hypoglycemia has always been the rate-limiting step in achieving perfect glycemic control for patients with T2DM. Hypoglycemia can be dangerous, and assessing patients for this condition is critically important. For patients, a fear of hypoglycemia can often have a considerable negative impact on diabetes management, metabolic control, and subsequent health outcomes.²³ Not performing this assessment may result in patients taking such actions as engaging in “defensive” eating or omitting insulin doses to preclude hypoglycemia, thus thwarting the best efforts of the physician to help patients achieve glycemic control. This behavior becomes particularly dangerous if the physician is titrating insulin regimens. 

Failure to address even mild hypoglycemia and glycemic control early in the course of T2DM may compromise the success of treatment in the long term.²⁴ Fear of hypoglycemia certainly affects the willingness of the patient to accept insulin therapy, so it is vital to choose regimens carefully and to address this topic with patients. Some hypoglycemia may occur in association with insulin therapy, even when the most skilled physicians and knowledgeable patients are involved in its use. Thus, identifying hypoglycemia and providing prompt, appropriate treatment are paramount. The body may respond to extremely low nocturnal blood glucose levels by rebounding with high blood glucose levels in the morning (referred to as the Somogyi effect). This rebound could be incorrectly identified as fasting hyperglycemia. Both the patient and the physician should routinely review blood glucose patterns. Patients with type 1 diabetes mellitus, as well as those with T2DM of long duration, may be at risk of hypoglycemia. With increasing age, the potential reaction time between awareness and onset of symptoms is decreased, contributing to an increased risk for asymptomatic hypoglycemia and greater susceptibility to cognitive impairment.^25,26 Recurrent, unrecognized hypoglycemia can occur even in patients with T2DM who have well-controlled glycemia.²⁷ Asymptomatic hypoglycemia and nocturnal hypoglycemia can interfere with the ability of patients to recognize subsequent hypoglycemia, and they can also limit patients' ability to take appropriate action, thereby exacerbating the situation. This outcome, known as hypoglycemic unawareness, is a marker of a high risk of severe hypoglycemia. 

Furthermore, high glucose variability (blood glucose excursions occurring throughout the day) may be a predictor of diabetic complications, independent of HbA_1c levels, in patients with T2DM.²⁸ As a result, because patients fear hypoglycemia, they may be reluctant to follow or adjust their insulin regimens when needed. This lack of action may result in chronic hyperglycemia, oxidative stress, and the risk of long-term complications.²⁹ Better daily control of blood glucose excursions, especially in the postprandial period, may reduce glucose variability in patients with T2DM³⁰ and may also reduce the risk of complications. 

For patients who require intensification of insulin therapy,³¹ strategies may include sequentially adding prandial insulin doses and then switching to premixed insulin (which provides basal and prandial insulin in a single injection, albeit in a fixed-dose ratio) or full basal-bolus dosing (which involves multiple daily injections of insulin).^9,21,32 Patients who have severe insulin resistance (eg, those who require more than 200 U of insulin)³³ may benefit from receiving concentrated insulin formulations (eg, 500 U/mL, as opposed to 100 U/mL for most insulins).³³ Providing very large doses of less concentrated insulin may result in difficulties with reliable insulin absorption because of simple volume problems. 

Edith, a 74-year-old African American woman (weight, 100 kg), presents with substantial osteoarthritis (OA) of the hip that requires total hip replacement. Her surgeon has reported that he will not operate until Edith's diabetes is better controlled. Edith has T2DM of 10 years' duration. Although her T2DM initially had been controlled through lifestyle changes and metformin therapy, she is currently taking 1000 mg of metformin twice daily and 10 mg of glipizide twice daily. Her current HbA_1c level is 10.8%. Edith checks her fasting blood glucose level every morning. She has a history of OA and hypertension and has undergone percutaneous transluminal coronary angioplasty. She reports wanting to be more active. 

Edith has uncontrolled T2DM and is likely to have comorbidities. Her renal function should be checked. (Older patients and those with elevated HbA_1c levels are at greater risk for evidence of renal impairment,³⁵ which may influence the approach to antihyperglycemic therapy.³⁶) Edith's serum creatinine level is 1.2 mg/dL; however, her glomerular filtration rate is 65 mL/min/1.73 m². 

Metformin should be avoided if Edith's serum creatinine level is 1.4 mg/dL or higher (a serum creatinine level of ≥1.5 mg/dL is the cutoff for men). Because of the risk of hypoglycemia, sulfonylureas should be used cautiously in patients with renal impairment. If a sulfonylurea is used, glipizide is preferred because of its route of excretion. If insulin is chosen—which is likely because Edith's blood glucose level needs to be lowered expeditiously before her operation—more modest dosing and the use of insulin analogs are warranted to minimize the risks of nocturnal hypoglycemia. Whereas prescribing information for the basal insulin, insulin glargine, indicates that it may be given either in the morning or evening, insulin detemir indications are for administration in the evening only. In clinical practice, both insulin analogs can be given at any time of day. If nocturnal hypoglycemia is the major concern, a basal insulin analog can be administered in the morning; this may be an off-label use, but it will allow the physician and the patient to become comfortable with the dosing and titration of the insulin. In setting a treatment goal for Edith, it is noted that her current HbA_1c level is 10.8%. Insulin is very effective at lowering blood glucose levels, with its dose limited only by hypoglycemia. However, that is a strong caveat to consider, especially in an older patient. Increased age and impaired renal function are considered risk factors for hypoglycemia.³⁷ 

The diabetes guidelines from the US Department of Veterans Affairs and the US Department of Defense (VA/DoD) were updated in 2010. As with other guidelines, the VA/DoD guidelines do not distinguish treatment goals by age group. However, they do provide some guidance that may be applicable to this case study: “for the patient with longer duration diabetes (more than 10 years) or with comorbid conditions and who requires a combination medication regimen including insulin should have an A1C target of <8%.”^38(p46) If following these guidelines, physicians may set one goal for the preoperative period and another goal for the postoperative period, depending on how successful the operation is and when more time might be available to titrate insulin therapy. 

Patients can safely and effectively self-titrate basal insulin by using one of several physician-directed or patient-driven treat-to-target titration algorithms (Table 1).^39-41 Prescribing information suggests a starting dose of 10 U of a basal insulin analog. This dose may frustrate patients with T2DM, who may be insulin resistant and overweight. For patients with T2DM, it may make more sense to refer to the guidelines of the American Diabetes Association, which recommend starting basal insulin therapy at a dose of 0.2 U/kg.⁹ Self-titration empowers patients to be involved in their therapy, allows for more rapid adjustments of insulin than do visits to the physician's office, and reduces the burden on the physician. The insulin dose should be titrated on the basis of the fasting blood glucose level noted before breakfast. The target is a fasting blood glucose level below 130 mg/dL.²⁰ Continuation of oral antidiabetic drugs is typical unless there is a specific contraindication. These drugs can be reduced or eliminated once glucose control is improving, but it is always better to add to therapy when the patient has hyperglycemia. 

Edith starts receiving 20 U of a basal insulin analog in the morning. She is allowed to self-titrate but is advised to call the physician's office if she experiences any signs or symptoms of hypoglycemia. Her blood glucose data logs show improvements not only in her fasting glucose levels but also in her postprandial glucose levels (Table 2). 

In this case study, Edith undergoes the operation and rehabilitation and then resumes walking. When she returns to the physician's office, her HbA_1c level is 6.4%. Edith now checks her glucose levels only in the morning and whenever she feels like her blood glucose levels are low (Table 3), which sometimes occurs when she is “out and about.” She carries food with her, “just in case.” If Edith experiences “low blood sugar,” she sometimes decides to skip her insulin dose the next day. Her current antihyperglycemic medication regimen consists of 1000 mg of metformin twice daily and 10 mg of glipizide twice daily; she also takes 62 U of basal insulin analog once daily. 

Evidence of documented hypoglycemia exists, as do signs of increased glucose variability in the aforementioned patterns. The basal insulin analogs have stable profiles, and patients who are receiving stable doses should have stable glucose profiles. Edith, however, is experiencing a lot of glucose variability in her morning blood glucose level. Her blood glucose levels may be decreasing during the night and then rebounding with hyperglycemia in the morning (the Somogyi effect). Edith's lifestyle is now limited by hypoglycemia; she is experiencing symptoms, eating defensively, skipping insulin doses, and possibly experiencing hypoglycemia without being aware of it. Table 4⁴² summarizes the signs and symptoms of hypoglycemia, whereas Table 5⁴² shows the physiologic responses to hypoglycemia. At Edith's age, and with her lifestyle-limiting symptoms, the most important step is to address her hypoglycemia. 

It is very important to educate patients about the signs and symptoms of hypoglycemia. The body has multiple defenses with which to prevent hypoglycemia, but age and a history of hypoglycemia can affect the body's hypoglycemic threshold for symptoms and when it can respond. The first line of defense is suppression of insulin production; the second is increased glucagon secretion. These 2 steps prevent most occurrences of hypoglycemia in people without diabetes or in individuals with T2DM of limited duration. If the blood glucose levels continue to drop after suppression of insulin and secretion of glucagon, then release of epinephrine, cortisol, and growth hormone will follow. At even lower blood glucose levels, defensive eating is observed as a response. 

Patients with type 1 diabetes mellitus are missing the 2 aforementioned lines of defense, which is why they are more prone to hypoglycemia than are patients with T2DM. However, by the time patients with T2DM require insulin, they are physiologically similar to patients with type 1 diabetes mellitus. These patients with T2DM of longer duration are at greater risk for hypoglycemia than are younger patients with T2DM (Figure 1).^25,26 

Concerns about hypoglycemia are discussed with Edith, and it is agreed that the glipizide dose will be reduced to 10 mg given once daily, and the basal insulin analog dose is reduced to 52 U. She is asked to send in her blood glucose readings 2 weeks after making this medication adjustment. At that time, her morning blood glucose level is more stable, and the hypoglycemia has stopped. Edith says that she notices that she needs to eat snacks a lot less often. 

Six months later, Edith's fasting blood glucose levels are at goal (Table 6), but her HbA_1c level is elevated, suggesting that postprandial hyperglycemia may be the cause. Increasing the sulfonylurea dose will not correct the situation and may even increase the risk of hypoglycemia. It is time to consider intensifying the insulin therapy. 

Several options are available for introducing prandial correction insulin, including adding a single injection of prandial insulin to the largest meal or the meal with the highest postmeal glucose level (a basal “plus” strategy), providing basal-bolus insulin (in multiple daily injections), or using premixed insulin (which provides basal and prandial insulin in a single injection). The basal “plus” strategy can be used to transition to the use of basal-bolus insulin (Figure 2). When the patient's HbA_1c level is not at goal after 1 injection of basal insulin analog, options include adding a rapid-acting insulin analog before meals. If this is done, then the physician should subtract 10% of the basal dose and add that amount to the largest meal,⁴³ or switch to a premixed insulin analog by dividing the total basal dose in half and administering it twice daily before breakfast and dinner.^32,44 Ideally, the patient should have a regular, stable, daily treatment schedule to minimize the risk of hypoglycemia resulting from these fixed-dose products. Concern surrounds the low levels of blood glucose noted at 3 to 5 hours after injection, when there is overlap in the action of the short-acting and long-acting insulins. People who have erratic daily schedules and mealtimes or people who skip meals would not be good candidates for premixed insulin regimens. Regarding premixed insulin, it is also important to differentiate between the human and the analog formulations available. The pharmacokinetics are remarkably different because the short-acting insulin in the human formulation is regular insulin, which has a very different pattern of action than the rapid-acting analog insulins (insulin aspart and insulin lispro) in the analog formulations. 

Given Edith's history, a single dose of prandial insulin is added before her largest meal; she is asked to check preprandial and postprandial glucose levels at dinner (her biggest meal) and to check her morning glucose levels less often. We reduce her basal analog insulin dose and discuss her sending in monthly readings to help with titration of her therapy. With her age and need for insulin at mealtime, it appears that Edith is no longer responding to the sulfonylurea, so it will be removed as part of her titration. One month later, she is taking 46 U of basal insulin and 6 U of a rapid-acting insulin analog at dinner. We then stop the glipizide and increase her insulin dose at mealtime to 12 U. After another month, we reduce her basal insulin because she has further increased her activity level. With increased activity, and with less snacking caused by defensive eating, Edith has lost 10 lbs and is feeling good about her progress (Table 7). Edith is taking metformin (1000 mg twice daily), a basal insulin analog (40 U) in the morning, and a rapid-acting insulin analog (12 U) before dinner. Edith is so pleased with her results that she refers her son, William. 

William is a 52-year-old man who works as an accountant. He has had diabetes mellitus for 15 years, and his current HbA_1c level is 9.6%. William takes 80 U of basal insulin analog twice daily. He also takes 10 U of rapid-acting insulin at meals, when he remembers to do so; he takes it, on average, twice daily. His blood glucose levels are not low. William also has hypertension, dyslipidemia, and obstructive sleep apnea (OSA). 

Localized predominantly in the southern states and Appalachia, the “Diabetes Belt” is an area of the United States where the incidence of diabetes is 11.7%, compared with the 8.7% incidence in the rest of the country (Figure 3).⁴⁵ Nearly one-third of the increased risk of diabetes among individuals in the Diabetes Belt is associated with a sedentary lifestyle and obesity.⁴⁶ There also are more African Americans with diabetes in the Diabetes Belt than in other areas of the United States. William shares some of the characteristics of individuals from the Diabetes Belt (Figure 4).⁴⁶ He also has OSA, which is another common comorbidity of T2DM. Primary care providers tend to underdiagnose OSA in patients with T2DM,⁴⁷ so William's OSA is something to consider. 

Physicians should always remember to confirm treatment adherence in patients who experience poor glycemic control while ostensibly receiving very high doses of insulin therapy. Inspection of injection sites and review of injection technique are also important, because injection-related problems have been known to affect glycemic control. Many patients will develop changes in the subcutaneous tissue, including lipohypertrophy or lipoatrophy. This scarring or dimpling of the skin will also affect insulin absorption. Check to make sure that insulin is not expired or denatured. In William's case, he had many vials of “used” insulin in his house; many vials were not dated when he started to use them, and some had clearly expired. In addition, it should be noted that large doses of insulin may produce a depot effect (albeit less so with analog insulin) and may be associated with unpredictable absorption.⁴⁸ It is more likely that William was not actually absorbing the full dose of insulin that he was injecting. 

Some patients with T2DM are profoundly insulin resistant and require very large doses of insulin for glycemic control. Large volumes of subcutaneous conventional U-100 insulin can cause discomfort at the injection site, resulting in poor patient adherence to insulin therapy. Furthermore, the depot effect may change insulin absorption as mentioned above. A therapeutic option that clinicians should be aware of is the use of U-500 insulin (Humulin R U-500; Eli Lilly and Company, Indianapolis, Indiana), which substantially reduces the volume of insulin to be injected.⁴⁹ Obese patients with uncontrolled T2DM who are severely insulin resistant can achieve satisfactory glycemic control with the use of U-500 regular insulin administered either by subcutaneous injection or via insulin pump.^33,50,51 Candidates for U-500 insulin include patients who have immune-mediated insulin resistance, lipoatrophic diabetes, antibodies against the insulin receptor, genetic abnormalities of the insulin receptor, or obesity.⁵² U-500 insulin is also very helpful for people who need more than 300 U of insulin per day and still have uncontrolled hyperglycemia. U-500 is a human regular insulin, but its time-action profile is not the same as that of the usual U-100 formulation. It has the same peak as U-100 regular insulin, but its duration is more like that of NPH insulin (approximately 12 h). Given these characteristics, U-500 works as both basal and prandial (bolus) insulin, so patients who use it not only take less insulin by volume, but they also typically need only one type of insulin. The use of U-500 insulin has also been associated with increased patient satisfaction and cost savings.³⁴ 

U-500 insulin is 5 times as potent as U-100 insulin. When converting U-100 insulin dosing to U-500 dosing, if the total daily insulin dose is less than 200 U, first divide by 5 and then split to twice-daily dosing (with 60% of the total daily dose given before the morning meal and 40% given before the evening meal). If the total daily dose is more than 200 U, first divide by 5 and then use thrice-daily dosing (at breakfast, lunch, and bedtime). In William's case, he takes a total of 180 U per day; because 180 U÷5=36 U, William can take 22 U before breakfast and 14 U before dinner. This allows William to obtain insulin coverage with only 2 injections. This dosing provides a less physiologic profile than what William was previously taking, but he shows increased adherence with this treatment and improves his glycemic control. 

Insulin remains our most effective agent for lowering blood glucose levels. Insulin formulations continue to be developed and refined, and they will likely continue to be essential in the management of T2DM. Physicians should feel comfortable initiating insulin therapy at any stage of T2DM, and they should be able to teach patients to self-titrate their insulin. Insulin therapy should be intensified in a patient-centric fashion, using the method that best fits each patient's profile, including adding prandial insulin, using full basal-bolus therapy, switching to premixed insulin therapy, or using more-concentrated insulin formulations. It is beneficial for patients to titrate insulin themselves, using parameters that let them know what dose or events signal the need to stop titration. Glucose variability and hypoglycemia are unexpected and warrant a call to the physician. By increasing patient levels of comfort with regard to insulin therapy, improved rates of glycemic control can be achieved.