Challenges to Intervention in Older Adults
Predicting Outcomes and Competing Risk
Traditional mathematical models for assessing risk for diabetes complications are based on the epidemiology and natural history of diabetes, as well as on the transitions of patients across their health status. Such models aid in the assessment of competing risk by highlighting the interactions between risks for diabetes complications and nondiabetes events. Clinical trials have demonstrated that the cardiovascular or mortality benefit of glycemic or blood pressure control become apparent only after 5 to 10 years following treatment initiation.[2,3] Although most older diabetic patients have 5 or more years of remaining life expectancy to benefit from interventions, remaining life expectancy is important for diabetes care decisions for those older patients near the end of life. A simulation model, based in part on a mortality risk index encompassing age, comorbidities, and functional status, indicates that the expected benefit of glycemic control declines as the levels of morbidity and functional impairment increase. Studies of nationally representative samples of older patients have found three naturally occurring clusters of patients based on comorbidity, with significant differences in mortality rates.[3,48] Thus, functional status and level of comorbidity are important factors in assessing risk. Models of diabetes complications would benefit from epidemiological data of older patients, including data on geriatric complications, dementia, or other conditions now known to be associated with diabetes.
The complexity of the older population with diabetes is increased by several sources of heterogeneity. One is general health status. In one study that stratified participants by health status and age, the proportion of those considered healthy was lower among participants aged older than 75 years, compared with those aged 51–64 years (Fig. 5). However, even among participants aged 51–64 years, 19% were expected to have difficulties in self-management and an additional 7% were expected to receive limited benefit from diabetes interventions because of the severity of their comorbidities. Conversely, 53% of diabetic patients aged older than 75 years were relatively healthy. In addition, compared with nondiabetic patients, diabetic participants in the Health and Retirement Study have much higher rates of comorbidities, such as dementia or heart disease, beginning in middle age. However, the relative impact of diabetes decreases with age. How to manage older diabetic patients with severe physical impairments is not clear. Diabetes management overall relies on patient preferences, as it occurs often in the absence of evidence or consensus regarding appropriateness and quality of interventions.
Heterogeneity in health status among patients with diabetes, based on data from the Health and Retirement Study of people over age 50 (48). People with known diabetes were assigned to one of four mutually exclusive categories: a Very Healthy group with no comorbidities; a healthy Intermediate group with comorbidities constrained to osteoarthritis and hypertension, and with no functional impairments; a group for whom intensive diabetes management would be Difficult to Implement due to multiple comorbidities and/or any one of the following: mild cognitive impairment, poor vision, two or more minor functional impairments; and a group with Uncertain Benefit from intensive diabetes management due to having the poorest health status, with one or more of the following: moderate-to-severe cognitive impairment, two or more major functional dependencies, and/or residence in a long-term nursing facility. As the Health and Retirement Study is a U.S. population-based survey, the y-axis estimates the total number of people in the U.S. over age 50 with diabetes in each category.
Racial and ethnic differences in subclinical disease, disease burden, and access to care also contribute to heterogeneity.[50,51] A1C levels are higher among black, Asian, and American Indian patients with diabetes. CVD and CHF incidence are higher among black patients compared with white individuals, but prevalence of coronary calcifications is lower among black, Hispanic, and Asian American individuals, and intima-media thickness is lower among Chinese individuals. CAD burden is lower among black and Hispanic patients, but black patients fare worse after surgery, angioplasty, or coronary revascularizations.[50,51] Although all racial groups in the Get With the Guidelines-Stroke program improve in response to interventions such as thrombolysis, deep vein thrombosis prophylaxis, smoking cessation, antithrombotic medications, atrial fibrillation and anticoagulation, and lipid therapy, black patients are less likely than white patients to receive these interventions.
Biological differences, such as family history and host genetics, are another source of heterogeneity. The offspring of centenarians have an exceptionally longer health span and are healthier and less likely to have diabetes or CVD than age-matched control subjects. Likewise, risk for diabetes and Alzheimer disease is lower for individuals with a parent who lived longer than 80 years. Among individuals with exceptional longevity, enriched genotypes include variants of FOXO3A, insulin growth factor receptor, CETP, and apoC3. Other biological differences may include those in pathways underlying diabetes complications or in age-related changes in fat, muscle, or cognition.
Aging Biology and Pathophysiology: A Possible Paradox
Restoring insulin sensitivity is a valid approach to treating diabetes, but the risks associated with insulin resistance are not always clear. Centenarians appear to have some insulin resistance compared with their offspring and with individuals who do not live as long, suggesting that insulin resistance is not necessarily negative. In nematodes, downregulation of the insulin receptor–like daf-2 confers insulin resistance but prolongs life. In rat studies, insulin sensitivity improves, even on ad libitum diets, when visceral fat is removed, and the animals live longer. However, a recent summary of studies of insulin-sensitivity models and longevity suggests that enhanced insulin sensitivity is neither necessary nor sufficient for enhanced longevity in mammals. It is possible then that age-related changes in metabolic regulation, vascular function, and other processes are accompanied by protective repair or adaptive responses. In that case, attempts to reverse processes that decline with age, such as insulin resistance, might provide no benefit or even worsen outcomes.
Diabetes. 2014;63(8):2578-2589. © 2014 American Diabetes Association, Inc.