A 93-year-old woman is admitted to a conventional hospital ward for an acute respiratory infection. The patient has type 2 diabetes mellitus of approximately 15 years evolution and has no other associated comorbidities, except for progressive dependence due to senescence and a previous hospitalization for pneumonia 6 months ago. She is currently in an assisted-living residence. A recent laboratory test revealed an HbA1c level of 7.8%, with a serum creatinine level of 1.3mg/dl (MDRD, 45ml/min). Her standard treatment consists of 5mg of glibenclamide a day and 850mg of metformin every 12h. What regimen should we follow once she is hospitalized? Does she require any change in her treatment at discharge?
Una mujer de 93 años de edad ingresa en una planta de hospitalización convencional por una infección respiratoria aguda. La paciente tiene diabetes mellitus tipo 2 de unos 15 años de evolución y no presenta otras comorbilidades asociadas, salvo progresiva dependencia por senescencia y un ingreso hospitalario previo por neumonía hace 6 meses; actualmente vive en una residencia asistida. En un análisis reciente tenía una HbA1c de 7,8%, con una creatinina sérica de 1,3mg/dl (MDRD: 45ml/min). Su tratamiento habitual consistía en glibenclamida 5mg al día y metformina 850mg cada 12h. ¿Qué pauta debemos seguir una vez hospitalizada? ¿Precisa de alguna modificación de su tratamiento al alta?
The prevalence of diabetes mellitus (DM) is increasing due to a growing population, with age being one of the main factors driving this increase. According to the United Nations (UN), people over 60 years of age comprise the elderly population; however, this definition does not take into account the discrepancy between biological and chronological ages. Other variables, such as functional state, comorbidities and frailty, are needed for a proper definition of the elderly.1 The UN definition, however, has been used in the design of numerous studies on elderly populations and diabetes.
The 1993 UN Human Development Report estimated that by 2030 the number of people older than 64 years with DM will reach 82 million in developing countries and more than 48 million in developed countries.2
Using the 2001 UN definition for elderly populations (in its report on the health of disabled patients), we defined the elderly population in our community as those who are living in nursing homes or residing in centers (both public and private). These homes and centers are mainly, though not exclusively, designed to provide long-term health care and represent a change in the patient's usual environment.
The prevalence of DM in the elderly population is difficult to determine, given that there are no established screening protocols, and that this is a heterogeneous population in terms of intellectual, social and physical abilities. A number of studies from the 1990s found a prevalence that varied between 10% and 26%.3–10
In Spain, the estimated prevalence of diabetes in the population over 65 years of age is 15%, reaching 20% in those over 80.11,12 The Di@bet.es study found a prevalence of known and unknown diabetes of 21.7% and 14.3%, respectively, in a group of 61–75-year-old patients, with similar figures (21.9% and 17.4%) for those over 76.13 However, this study excluded subjects who were institutionalized. Therefore, currently in Spain there are no clear data on the prevalence of DM in institutionalized patients.
Which DM detection test should be used for the institutionalized elderly?For institutionalized elderly patients with unknown DM, an oral glucose tolerance test (OGTT) of 75g of glucose can detect up to a 30% prevalence of carbohydrate intolerance and a 14–15% prevalence of DM.5,6
However, the main problem is in detecting these disorders, given that the reference test (the OGTT) has little adherence in this population type. The use of the fasting plasma glucose (FPG) test as a screening tool in this population has also not been validated. Aspray et al.7 analyzed the use of FPG levels >110mg/dL and postprandial plasma glucose (PPG) levels >200mg/dL 2h after intake as a screening for DM in a cohort of elderly patients, 11.4% of whom were known to have diabetes. The authors observed that FPG levels achieved a sensitivity of 71%, with a negative predictive value (NPV) of 97%; PPG levels achieved a sensitivity of 43% and an NPV of 95%. With this method, the prevalence of undiagnosed DM was 8%, rising to an overall diagnosis of 20%, with a higher prevalence of DM in elderly individuals with cognitive impairment. The authors argued that the use of FPG levels alone in this population type would result in a large number of undiagnosed patients, given that thin, elderly patients with diabetes have an insulin release disorder as a response to intake, and therefore an isolated FPG disorder would be infrequent. The authors therefore recommend conducting both tests for screening, particularly for frail elderly individuals in whom performing OGTT is not possible.7
As with FPG, the use of HbA1c levels in elderly populations can underestimate the prevalence of DM. In the study by Hayes et al.,8 the rate of undiagnosed DM cases was 28% and 37% for the HbA1c levels >6.5% criteria and the OGTT as diagnostic test, respectively. The DECODE study (Diabetes Epidemiology Collaborative analysis Of Diagnostic criteria in Europe) observed that PPG levels increase linearly with age, while FPG levels do not. The addition of PPG measurement is therefore the main reason for the increased prevalence of unknown DM for this age group.9
The harmful effects of hypoglycemiaIn elderly patients with diabetes, hypoglycemic symptoms are less marked and less specific, with more episodes than can go undetected and less time between adrenergic and neuroglycopenic symptoms than occurs with younger people.14
In general, the risk factors for hypoglycemia in elderly patients with diabetes are frailty, advanced age, polymedication, recent hospitalization, malnutrition, use of insulin or medium to long-acting sulfonylureas, acute intercurrent illness, prior asymptomatic hypoglycemia and an abnormal counterregulatory response.14–19
A European baseline study found rates of severe hypoglycemia of 0.4 and 1.5 per 100 patients/year in subjects treated with oral hypoglycemic agents and insulin, respectively.15 Rates of severe hypoglycemia (per 100 patients/year) in the elderly reached 1.23 (95% CI 1.08–1.38) in sulfonylurea users and 2.76 (95% CI 2.47–3.06) among insulin users.16
Data on institutionalized patients are scarce. In these patients, the hypoglycemia is multifactorial and is not only due to excessive glycemic control but also on associated comorbidities, polypharmacy, weight loss and loss of apetite.19
The consequences of hypoglycemia in this type of population are a higher incidence of stroke, cognitive impairment, acute coronary syndrome, severe tachyarrhythmia, falls and hip fractures.20 Elderly patients also remain hospitalized for longer periods (2.8 days longer if there is hypoglycemia when compared with patients without hypoglycemia). Every additional day with hypoglycemia is associated with an 85.3% increase in the odds ratio of hospital death (p=.009) and a 65.8% increase (p<.001) for death within a year from the time of discharge.21
Both the ADVANCE study22 and the ACCORD cohort23 had no significant reduction in mortality or cardiovascular events in the intensive blood glucose control groups in elderly patients with type 2 DM and high cardiovascular risk. Several subanalyses were subsequently conducted to assess whether the cause of this increase in mortality could have been related to hypoglycemic episodes.24,25 McCoy et al.26 observed that the 5-year mortality of patients with diabetes and severe hypoglycemia was higher than in those with mild episodes or without hypoglycemia (23.7% vs. 13%, p=.01). The authors concluded that hypoglycemia in and of itself could be a marker of frailty and vulnerability.
A significant association was found in elderly patients with type 2 DM between severe hypoglycemic episodes (requiring hospitalization) and dementia, especially in those with more than 1 hypoglycemia episode.27
Objectives of glycemic controlThe United Kingdom Prospective Diabetes Study28 demonstrated the importance of glycemic control for reducing the risk of microvascular complications and cardiovascular disease in patients with recently diagnosed type 2 diabetes. This benefit requires good long-term glycemic control, although it is doubtful that this control has any benefits for frail elderly patients or those with a limited life expectancy.
The Veterans Affairs Diabetes Trial29 recruited 1791 military veterans with type 2 DM diagnosed approximately 11 years ago, poor metabolic control (HbA1c levels of 8.4%) and high cardiovascular risk. The participants were randomly assigned to either an intensive blood sugar control group or a standard treatment group and were underwent follow-up at 5.6 years. No differences were found in cardiovascular events (HR, 1.07; 95% CI 0.81–1.42; p=.62) or in microvascular complications, except for the progression of microalbuminuria and an increase in hypoglycemia in the intensive group.
Therefore, strict metabolic control in elderly patients with high cardiovascular risk and type 2 DM diagnosed several years previously does not appear to reduce cardiovascular mortality rates and can increase morbidity. The general objectives for glycemic control should therefore be personalized according to the patient's life expectancy and functional dependence (Table 1), controlling for other cardiovascular risk factors. Thus, a 5-year life expectancy would be an acceptable cut-off point to identify elderly patients who would not benefit from more intensive management.30
Diabetes treatment objectives for elderly institutionalized patients.
Functional class | Objectives | Comments |
Functionally independent | HbA1c: 7–7.5%.Maintain an adequate physical condition.Constant intake of carbohydrates. Avoid soft drinks, high amounts of sugar and fruit juices. | Patients who live autonomously, who experience no significant difficulties with daily life tasks and who require no care givers and minimal support. They have more than 1 comorbidity (such as diabetes) that can have an effect. |
Functionally dependent(A) Frailty(B) Dementia | HbA1c: 7–8.5%Stimulate water intake.Intake with a high protein and energy content.HbA1c: 7–8.5%Identify difficulties with intake and provide a meal schedule to ensure no interference from disturbances. | Patients who have lost the capacity to perform daily life activities (bathing, dressing, personal hygiene). The category of frailty is a combination of fatigue, weight loss, significant movement restriction, reduced strength and propensity for falls. The category of dementia includes patients with cognitive disorders that cause significant memory problems, disorientation and personality changes and patients who are incapable of caring for themselves. |
Final stage of life | Prevent symptomatic hyperglycemia (>270mg/dL).Minimize hypoglycemia.Individualize enteral and parenteral nutrition. | Those with a severe medical condition or cancer and a life expectancy of less than a year. |
Source: Adapted from the International Diabetes Federation.1
Although these studies did not include institutionalized patients, it can be extrapolated that this population group should not be overtreated and that the objectives should be to improve the patient's wellbeing and quality of life and reduce the risk of hypoglycemia.
In general, an HbA1c level of 7–7.5% has been suggested as an objective for patients with more than 1 affected system (and with no other comorbidities), with 7–8% for functionally dependent elderly patients and up to 8.5% for those with cognitive impairment and/or who are frail.1,31
The objective for patients with terminal illness is to prevent hypoglycemia and symptomatic hyperglycemia, maintaining preprandial glucose levels between 160mg/dL and 270mg/dL, accounting for other factors such as cachexia, vomiting and the use of diabetogenic drugs such as corticoids.1,32
TreatmentTreatment for elderly patients with diabetes needs to be personalized according to their comorbidities, functional dependence and life expectancy and the patient's and/or care providers abilities.1,33 It should also be taken into account that an elderly patient with diabetes diagnosed several years previously and with high cardiovascular risk is not the same as an elderly patient with a recent onset of diabetes. From a pathophysiological point of view, drug treatment (Tables 2 and 3) for the latter case should be aimed at controlling PPG levels.34 It is therefore important to start treatment with low doses that are gradually increased. In many cases, reaching the maximum dose is not necessary and inadvisable due to the increased risk of adverse effects with no improvement in the drug's efficacy.
General recommendations for the treatment of diabetes in elderly institutionalized patients.
Functional class | Objectives | Comments |
Functionally independent | 1st line: metformin (if there are no contraindications).2nd line: sulfonylurea (if metformin is not tolerated or is contraindicated).DDP-4: Consider adding as a second drug in 1st line treatment.Glinides: postprandial hyperglycemia control and erratic eating habits. Attention to interactions.Basal insulin (NPH, glargine, detemir). Add to 2nd or 3rd line. | Assess premixed if HbA1c levels >8.5% |
Functionally dependent(A) Frailty(B) Dementia | Avoid or discontinue any agents that cause nausea, gastrointestinal issues and/or excessive weight loss (GLP-1, metformin).Insulin can have anabolic benefits. | Use drugs with low risk of hypoglycemia.Simple insulin regimens.Educate family members or care givers in identifying hypoglycemia. |
Final stage of life | Assess discontinuation of unnecessary treatments and minimize doses. Consider withdrawing all therapies, including insulin, in the terminal stage. | Same precautions as in the dependent group. |
Source: Adapted from the International Diabetes Federation.1
Summary of the main characteristics of hypoglycemic agents used in frail elderly patients.
Group | Risk of hypoglycemia | Indications | Limiting effects/constraints | Renal excretion | Others |
Biguanides (metformin) | Low | Overweight | AnorexiaWeight loss | Yes.Contraindicated CCr <30mL/min.Annual renal function monitoringFrail or >80 years, urinary CCr measurement. | Risk of lactic acidosis in acute circumstances (comorbidities) and/or diagnostic tests. |
Sulfonylureas | High | BMI 22–25 | HypoglycemiaWeight gain | YesAvoid long half-life (glyburide, glimepiride and glibenclamide) | Drug interactions (salicylates, dicoumarinic drugs, sulfonamides, fibrates, allopurinol, beta blockers). |
Glinides (repaglinide) | Low | Postprandial hyperglycemia.Erratic eating habits. | HypoglycemiaWeight increaseAdministration 3 times a day | Renal excretion <10%.Hepatic metabolism | Rapid absorption.1% reduction in HbA1c levels. |
Thiazolidinediones (pioglitazone) | Low | Increased peripheral insulin sensitivity. | Rarely used due to frequent comorbidities in elderly patients. | Sodium and water retention, heart failure and fractures | Precautions due to warnings of increased bladder cancer |
Alpha-glucosidase inhibitors (acarbose, miglitol) | Intermediate | Diet rich in carbohydratesLow efficacy | Gastrointestinal effects | If other drugs are contraindicated or not tolerated | |
Dipeptidyl peptidase-4 (DPP-4) inhibitors | Low | Effective in monotherapy | Edemas in the legs.Headache.Nasopharyngitis | Yes.Moderate renal failure (50% dose reduction) Severe renal failure (individualize the compound).Linagliptin does not require dosage adjustment. | Rare drug interactions.Combined with metformin and/or sulfonylureas and/or insulin. |
Glucagon-like peptide-1 (GLP-1) analogs | Low | Little experience with use in elderly patients | Gastrointestinal effects (nausea, vomiting).Anorexia | Very limited use in elderly patients, especially those who are frail and institutionalized. | |
Insulin | Dose-dependent | Extended use | Weight gainInjectable, Limitations on self-administration due to sensory and cognitive deficits.Hypoglycemia.Monitoring. | YesDosage reduction if clearance is reduced | Addition of a single 0.1–0.2 UI/kg dose of basal insulin analog (glargine, detemir) to 1 or 2 hypoglycemic agents |
Biguanides (metformin) have a low risk of causing hypoglycemia but can reduce caloric intake and cause weight loss in frail elderly patients. At least 1 annual renal function checkup should be conducted (especially to determine creatinine clearance in urine35), and situations that increase the risk of lactic acidosis should be avoided.
Sulfonylureas increase the risk of hypoglycemia in elderly patients due to age-related reduced elimination and renal failure, which can lead to an increase in the half-life of these drugs. Therefore, drugs with a long half-life such as glyburide, glimepiride and glibenclamide should be avoided, while those with a shorter half-life, such as gliclazide, could be indicated for elderly patients with a body mass index between 22 and 25. For frail institutionalized elderly patients, the risk of hypoglycemia related to the use of these drugs is even higher and is a marker of a poor prognosis for this population group, with a 23% survival rate at 1 year of the hypoglycemic episode.36
Glinides (repaglinide) are faster acting (with a shorter half-life) and act mainly on PPG levels and, therefore, pose a lower risk of hypoglycemia. They may be considered a useful drug for elderly patients with erratic eating habits.37
Thiazolidinediones (pioglitazone) increase peripheral insulin sensitivity. They are safe drugs in terms of hypoglycemic risk, but the adverse effects of water and sodium retention, heart failure and fractures reduce its usefulness in elderly patients.38
The use of alpha-glucosidase inhibitors (acarbose, miglitol) is limited due to their low efficacy and the relative frequency of adverse gastrointestinal effects (flatulence, diarrhea).
Dipeptidyl peptidase-4 (DPP-4) inhibitors are well-tolerated drugs that do not present a high risk of hypoglycemia, do not affect weight and have limited drug interactions. They reduce FPG and PPG levels and inhibit glucagon secretion. They are effective in monotherapy and can be added to treatment with metformin and/or sulfonylureas. Although there is still a need for data on their long-term safety and efficacy as well as more studies with elderly populations, these drugs are safe in the presence of renal failure, with dose adjustments where necessary.37,39–42
There is very limited experience in the use of glucagon-like peptide-1 analogs in elderly patients. Although these drugs present barely any risk of hypoglycemia, their adverse gastrointestinal effects, appetite reduction and weight loss limit their use in elderly patients, especially those who are frail and institutionalized.43
Insulin is the most effective treatment for reducing glycemia, although elderly patients and particularly those who are frail and cognitively impaired require insulin therapy at low, progressive doses and should not be prescribed complex treatments. Although they have not been shown to be superior to neutral protamine Hagedorn (NPH) insulin, long-acting insulin analogs pose a lower risk of hypoglycemia, especially nocturnal hypoglycemia.44–46
Nutritional considerationsOver time, the body undergoes various changes, such as a loss of lean mass and a reduction in exercise and baseline metabolism, which results in a 20–30% reduction in energy requirements. The most reliable indicator of a deficient nutritional state in elderly patients is therefore a change in body weight; all involuntary weight loss that exceeds 5% in less than 6 months should be assessed.47
In institutionalized elderly patients, malnutrition can occur due to a lack of adequate food options and unnecessary dietary restrictions. The use of specialized diets for patients with diabetes does not appear to be any better than normal unrestricted diets and does not improve glycemic control.48 A number of studies suggest that special diets and therapies are associated with malnutrition in institutionalized elderly patients.10 Therefore, it is recommended that residents be served a regular menu with consistent amounts of carbohydrates in their meals and that antidiabetic treatments be intensified if necessary.49,50
Clinical guidelinesSpecific guidelines for the treatment of institutionalized elderly patients with diabetes are scarce. In 2011, Diabetes UK published a clinical practice guide for elderly patients with diabetics living in care homes (Table 4).51 The International Diabetes Federation recently wrote a complete guide for the treatment of elderly patients with type 2 DM, with a chapter dedicated to institutionalized elderly patients.1 Another guide of particular interest published by Diabetes UK covers the issue of treating diabetes in terminally ill patients.32 A consensus was recently published in Spain that covers the various aspects of diabetes treatment in elderly patients.52
Important clinical aspects to consider when treating elderly institutionalized patients.
Objectives | Measures |
Nutritional evaluation and dietary strategy. | Early recognition of intake difficulties, malnutrition and hypoalbuminemia. All patients who enter a care home should undergo a nutritional assessment at the start. |
Screening for diabetes. | HbA1c (>6.5%) and FPG (>126mg/dL) levels or the combination of FPG (>126mg/dL) and PPG (>200mg/dL) levels. |
Individual treatment and follow-up plan for each resident. | Objectives agreed upon by the patient, their family, the primary care doctor or usual specialist and the residence staff (objectives include metabolic control and annual check-ups). |
Detailed annual review. Review of dietary plan and main care objectives. | Clinical history and pharmacological treatment assessment, nutritional and functional assessment and visual acuity (ocular fundus), renal function and glucose control assessment. |
Individualized glucose control, at all times prioritizing the well-being of the elderly patient. | Prevent hypoglycemia, FPG >126mg/dL and <153mg/dL. Hyperglycemia (random blood sugar >165mg/dL) should be controlled to avoid osmotic diuresis, dehydration and microvascular complications.The proper range for HbA1c levels is 7–8%. |
Hypoglycemia is a significant consequence of treatment. | It is essential to identify residents at greatest risk: patients overtreated with antidiabetic agents or insulin, malnourished patients, patients with kidney disease, polymedicated patients and advanced age patients (>80 years). |
Annual review of foot ulcer risk and access to a podiatrist, if necessary. | – |
Screening for retinopathy. | Staff training on the importance of maintaining the visual health of residents with diabetes. |
In regard to treatment, both excessive and poor metabolic control should be avoided.The use of insulin requires trained staff and nurses experienced in diabetes. | Personalize according to glucose control objectives, renal function, comorbidities, side effects and risk of hypoglycemia.Long-term insulin analogs, insulin detemir or insulin glargine have a lower incidence of hypoglycemia compared with NPH insulin, which can be an advantage for frail elderly patients. |
Palliative care at the end of life. | Unnecessary glucose tests and complex insulin regimens are not recommended.In this stage of life, trying to maintain strict glycemic control can reduce the patient's quality of life and well-being.The risk of hypoglycemia increases in cases of weight loss, cancer and liver or kidney disease. |
Pain tends to be underdiagnosed and therefore untreated. | The sources of pain for elderly patients with diabetes include peripheral neuropathy, neuropathic arthropathy, foot ulcerations and peripheral arterial disease. It is important that the staff of the residence and healthcare personnel detect these conditions and treat them properly. |
Proper training on the potential complications of diabetes and on the treatment, nutrition and care of elderly patients by healthcare personnel of the residence and the staff. | Also related to the care of these elderly patients during periods of intercurrent illness and effective and suitable referrals to hospital for acute illness. |
Source: Adapted from Sinclair AJ, on behalf of the Task and Finish Group of Diabetes UK. Diabetes UK.51
Abbreviations: FPG fasting blood glucose; PPG postprandial glucose 2h after intake.
The scope of many of the studies assessed might not include elderly populations living in medium to long-term care homes and hospitals. However, this type of population might also be included in many of their conclusions. In other words, the concept of institutionalization encompasses a heterogeneous group of elderly patients, and the currently available data do not allow for comparing groups from different institutions or patients receiving assistance in their social or family homes. However, arguably the most influential factors for treatment are the patient's cognitive state, the presence of geriatric syndromes and the patient's social environment.
Comments on the caseIn regard to the case presented, the patient was treated with oral antidiabetic medications during her hospital stay and was started on low-dose basal-bolus insulin therapy. Following the resolution of the acute illness, glibenclamide was withdrawn, the metformin dosage was reduced and the patient was started on dipeptidyl peptidase-4 (DPP-4) adjusted to renal function.
ConclusionsElderly institutionalized patients with diabetes are a specific population requiring metabolic control, for whom careful personalization of the objectives and a risk-benefit analysis of the drug therapy are necessary. Proper patient education on diabetes for patients, families and care givers is essential, with special care for institutionalized patients and those with other risk factors in addition to hypoglycemia.
All elderly patients entering residences should undergo an overall nutritional assessment and be given a personalized strategy for glycemic control. Early recognition is necessary for any difficulties with food intake, malnutrition, foot and pressure ulcers, hypoglycemia, physical activities and limitations, visual acuity, and in general, the overall functional capacity.
Conflicts of interestThe authors declare that they have no conflicts of interest.
Please cite this article as: Cano Megias M, Guisado Vasco P. Diabetes mellitus tipo 2 en el paciente anciano institucionalizado. Rev Clin Esp. 2014;214:521–528.