The definition of “elderly” varies, with some studies defining the elderly population as ≥60 years of age. Administrative guidelines frequently classify people >65 years of age as elderly. Although there is no uniformly agreed-upon definition of elderly, it is generally accepted that this is a concept that reflects an age continuum starting sometime after age 65 and is characterized by a slow, progressive impairment in function that continues until the end of life (1).
As noted in the Definition, Classification and Diagnosis of Diabetes, Prediabetes and Metabolic Syndrome chapter (p. S8), glycated hemoglobin (A1C) can be used as 1 of the diagnostic tests for type 2 diabetes in adults. Unfortunately, normal aging is associated with a progressive increase in A1C, and there is a significant discordance between fasting plasma glucose–based and A1C-based diagnosis of diabetes in this age group, a difference that is accentuated by race and gender (2). Pending further studies to define the role of A1C in the diagnosis of diabetes in the elderly, other screening tests may need to be considered in some patients. Screening for diabetes may be warranted in select individuals. In the absence of positive intervention studies on morbidity or mortality in this population, the decision about screening for diabetes should be made on an individual basis.
Reducing the Risk of Developing Diabetes
Lifestyle interventions are effective in reducing the risk of developing diabetes in elderly people at high risk for the development of the disease (3). Acarbose (4), rosiglitazone (5) and pioglitazone (6) also are effective in preventing diabetes in elderly people at high risk. Metformin may not be effective (3).
As interdisciplinary interventions, especially those that have been specifically designed for this age group, have been shown to improve glycemic control in elderly individuals with diabetes, these people should be referred to a diabetes healthcare team (7–9). Pay-for-performance programs improve a number of quality indicators in this age group (10,11). Telemedicine case management and web-based interventions can improve glycemic control, lipids, blood pressure (BP), psychosocial well-being and physical activity; reduce hypoglycemia and ethnic disparities in care; and allow for detection and remediation of medically urgent situations, as well as reduce hospitalizations (12–21). A pharmaceutical care program can significantly improve medication compliance, as well as the control of diabetes and its associated risk factors (22).
The same glycemic targets apply to otherwise healthy elderly as to younger people with diabetes. In older patients with diabetes of several years' duration and established complications, intensive control reduces the risk of microvascular events but does not reduce macrovascular events or mortality (23–25). However, better glycemic control appears to be associated with less disability and better function (26,27). It is known that postprandial glucose values are a better predictor of outcome in elderly patients with diabetes than A1C or preprandial glucose values. Recently, it has been demonstrated that older patients with type 2 diabetes who have survived an acute myocardial infarction may have a lower risk for a subsequent cardiovascular (CV) event with targeting of postprandial vs. fasting/preprandial glycemia (28). In patients with equivalent glycemic control, greater variability of glucose values is associated with worse cognition (29).
Unfortunately, aging is a risk factor for severe hypoglycemia with efforts to intensify therapy (30). Asymptomatic hypoglycemia, as assessed by continuous glucose monitoring, is frequent in this population (31). This increased risk of hypoglycemia appears to be due to an age-related reduction in glucagon secretion, impaired awareness of hypoglycemic warning symptoms and altered psychomotor performance, which prevents the patient from taking steps to treat hypoglycemia (32,33). Episodes of severe hypoglycemia may increase the risk of dementia (34), although this is controversial. Cognitive dysfunction in elderly subjects has been identified as a significant risk factor for the development of severe hypoglycemia (35,36). Therefore, the most important issue to address when attempting to achieve treatment guidelines in elderly patients is to prevent hypoglycemia as much as possible, even if that means that higher glycemic targets must be used.
“Frailty” is a widely used term associated with aging that denotes a multidimensional syndrome that gives rise to increased vulnerability. Frailty may have a biological basis and appears to be a distinct clinical syndrome. Many definitions of frailty have been proposed. The most commonly applied definition (Fried's Frailty Phenotype) suggests that a patient is frail when 3 or more of the following criteria are present: unintentional weight loss (>10 pounds in the past year), self-reported exhaustion, weakness (grip strength), slow walking speed and low physical activity (37). Progressive frailty has been associated with reduced function and increased mortality, and older patients with diabetes are more likely to be frail (38). When frailty occurs, it is a better predictor of complications and death in elderly patients with diabetes than is chronological age or burden of comorbidity (39). The Clinical Frailty Scale, developed by Rockwood et al (40), has demonstrated validity as a 7-point frailty scale that has since been modified to a 9-point frailty scale from 1 (very fit) to 9 (terminally ill), which can help to determine which subjects are frail (41) ( Figure 1 ). In people with multiple comorbidities, a high level of functional dependency and limited life expectancy (i.e. frail patients), decision analysis suggests that the benefit of intensive control is likely to be minimal (42). From a clinical perspective, the decision to offer more or less stringent glycemic control should be based on the degree of frailty. Patients with moderate or more advanced frailty ( Figure 1 ) have a reduced life expectancy and should not undergo stringent glycemic control. When attempts are made to improve glycemic control in these patients, there are fewer episodes of significant hyperglycemia but also more episodes of severe hypoglycemia (43).
Nutrition and physical activity
Nutrition education programs can improve metabolic control in ambulatory older people with diabetes (44). Amino acid supplementation may improve glycemic control and insulin sensitivity in these patients, although this is controversial (45,46). Physical training programs can be successfully implemented in older people with diabetes, although comorbid conditions may prevent aerobic physical training in many patients, and increased activity levels may be difficult to sustain. Prior to instituting an exercise program, elderly subjects should be carefully evaluated for underlying CV or musculoskeletal problems that may preclude such programs. Aerobic exercise improves arterial stiffness and baroreflex sensitivity, both surrogate markers of increased CV morbidity and mortality (47,48). While the effects of aerobic exercise programs on glucose and lipid metabolism are inconsistent (49–51), resistance training has been shown to result in modest improvements in glycemic control, as well as improvements in strength, body composition, and mobility (52–56). Exercise programs may reduce the risk of falls and improve balance in patients with neuropathy (57,58). However, it appears difficult to maintain these lifestyle changes outside of a supervised setting (59).
Oral antihyperglycemic agents
In lean elderly people with type 2 diabetes, the principal metabolic defect is impairment in glucose-induced insulin secretion (60). Therefore, initial therapy for these individuals should involve agents that stimulate insulin secretion. In obese elderly people with type 2 diabetes, the principal metabolic defect is resistance to insulin-mediated glucose disposal, with insulin secretion being relatively preserved (61–63). Initial therapy for obese older people with diabetes should involve agents that improve insulin resistance. There have been no randomized trials of metformin in the elderly, although clinical experience suggests it is an effective agent. Metformin may reduce the risk of cancer in elderly patients with diabetes (64,65). Alpha-glucosidase inhibitors are modestly effective in older people with diabetes, but a substantial percentage of individuals cannot tolerate them because of gastrointestinal side effects (66–69). Thiazolidinediones are effective agents but are associated with an increased incidence of edema and congestive heart failure (CHF) in older people (70–73). Rosiglitazone, but not pioglitazone, may increase the risk of CV events and death (74–77). These agents also increase the risk of fractures in women (77,78). When used as monotherapy, they are less likely to fail than metformin or glyburide (73). Interestingly, drugs that increase insulin sensitivity, such as thiazolidinediones and metformin, may attenuate the progressive loss in muscle mass that occurs in older people with diabetes and contributes to frailty (79).
Sulphonylureas should be used with caution because the risk of severe or fatal hypoglycemia increases exponentially with age (80,81) and appears to be higher with glyburide (82–84). Gliclazide and glimepiride are preferred over glyburide in the elderly because they are associated with a lower frequency of hypoglycemia and CV events (85–90). A long-acting formulation of gliclazide resulted in equivalent glycemic control and the same frequency of hypoglycemic events as regular gliclazide in the elderly (87), and appears to result in a lower frequency of hypoglycemic events than glimepiride (88). Meglitinides (repaglinide and nateglinide) are associated with a lower frequency of hypoglycemia in the elderly compared to glyburide (91–93) and would be preferred in individuals with irregular eating habits.
Dipeptidyl peptidase (DPP)-4 inhibitors (linagliptin, saxagliptin and sitagliptin) are similarly effective in young and old patients, cause minimal hypoglycemia when used alone and do not result in weight gain (94–97). The efficacy of liraglutide with respect to A1C and weight is independent of age and is well tolerated in the elderly with a low risk of hypoglycemia (98).
Insulin regimens in the elderly should be individualized and selected to promote patient safety. The clock drawing test can be used to predict which elderly subjects are likely to have problems with insulin therapy (99). In elderly people, the use of premixed insulins as an alternative to mixing insulins (100) and prefilled insulin pens as an alternative to conventional syringes (101,102) minimizes dose errors and may improve glycemic control. Premixed insulin analogues can be administered after meals (103–105) and may be associated with better control than basal insulins, but at the expense of more hypoglycemia and greater weight gain (106). Basal-bolus regimens may be associated with greater improvements in glycemic control, health status and mood than twice-daily injections of long-acting insulin (107), although premixed insulin analogues can result in equivalent glycemic control to basal-bolus regimens (108). In older people with poorly controlled type 2 diabetes requiring insulin, both continuous subcutaneous insulin infusion and basal-bolus regimens can result in excellent glycemic control with reduced glycemic variability, as well as good safety and patient satisfaction (109,110). One study demonstrated equivalent glycemic control in older people treated with either twice-daily insulin injections or a combination of a single injection of NPH insulin with an oral antihyperglycemic agent (111). The addition of glargine to oral agents results in improved control and a reduced frequency of hypoglycemia when compared to escalation of oral agents (112). Both detemir and glargine resulted in a reduced rate of hypoglycemia when compared to 30/70 insulin or NPH (113,114). Finally, elderly patients with diabetes are at increased risk for falls and fractures, and insulin therapy increases this risk, although the mechanism for this effect is unclear (115).
Prevention and Treatment of Complications
Treatment of isolated systolic hypertension or combined systolic and diastolic hypertension in elderly people with diabetes is associated with a significant reduction in CV morbidity and mortality and microvascular events. Also, the number needed to treat (NNT) reduces with increasing age (116–120). Treatment of isolated systolic hypertension may also preserve renal function in elderly people with diabetes (121). Several different classes of antihypertensive agents have been shown to be effective in reducing the risk of CV events and end stage renal disease, including thiazide-like diuretics, long-acting calcium channel blockers, angiotensin-converting enzyme (ACE) inhibitors, and angiotensin II receptor blockers (116–126). Any of these agents is a reasonable first choice (122–124). Although the calcium channel blocker amlodipine may be associated with an increased risk of CHF (124), the combination of ACE inhibitor and amlodipine appears to reduce CV events more than the combination of an ACE inhibitor and hydrochlorothiazide (127). Cardioselective beta blockers and alpha-adrenergic blockers are less likely to reduce CV risk than the above agents (122–125). ACE inhibitors may be particularly valuable for people with diabetes and ≥1 other CV risk factor (128). More intensive control of BP (systolic<140 vs. <120) does not improve outcomes and results in more side effects (129). As a result, there has been discussion about altering the systolic BP target for the elderly to 140 mm Hg; however, the Canadian Hypertension Education Panel (CHEP), in collaboration with the Canadian Diabetes Association, have maintained the target BP of <130/80 mm Hg in diabetes. There has been a significant improvement in the last decade in the number of older people treated for hypertension, and therapies being used are more consistent with current clinical practice guidelines (130).
The treatment of dyslipidemia with statins for both primary and secondary prevention of CV events has been shown in most, although not all, studies to significantly reduce CV morbidity and mortality in older people with diabetes (131–139). The data on the use of fibrates in this patient population are equivocal (140,141), although they may reduce albuminuria and slow glomerular filtration rate loss (142).
Type 5 phosphodiesterase inhibitors appear to be effective for the treatment of erectile dysfunction in carefully selected elderly people with diabetes (143–145).
Depression is common in elderly patients with diabetes, and a systematic approach to the treatment of this illness not only improves quality of life but reduces mortality (146).
Diabetes in Nursing Homes
Diabetes is often undiagnosed in nursing home patients (147–150). The prevalence of diabetes is high in institutions, and individuals frequently have established macro- and microvascular complications, as well as substantial comorbidity (150–153). Antipsychotic drug use is a risk factor for the development of diabetes in patients in institutions (154). In observational studies, the degree of glycemic control varies widely between different centres (147,152), adherence to clinical practice guidelines is poor and insulin sliding scales are used frequently despite lack of evidence for their effectiveness (150). Undernutrition is a major problem in people with diabetes living in nursing homes (152).
There are very few intervention studies on diabetes in nursing homes. The short-term substitution of a regular diet or a standard nutritional formula for a “diabetic diet” or a diabetic nutritional formula did not modify the level of glycemic control (150,155–157). For selected nursing home residents with type 2 diabetes, substitution of regular insulin by lispro insulin (bolus analogue) may improve glycemic control and A1C levels with a reduced number of hypoglycemic episodes (158).