Researchers from more than 2 dozen large-scale epidemiological studies have reported a link between type 2 diabetes mellitus and cognitive impairment, cognitive decline, and dementia.1-5 Cognitive dysfunction is now considered by many researchers and clinicians to be an important late-life consequence of type 2 diabetes mellitus.5 Increasing attention has turned to whether clinical factors that predispose individuals to diabetes may confer increased risk of late-life cognitive dysfunction.
This article provides a focused discussion of 3 key diabetes-related risk factors that affect cognitive aging: obesity, insulin resistance, and the metabolic syndrome. Indeed, emerging data suggest that these conditions, which often precede the clinical diagnosis of diabetes, may be significantly associated with various forms of cognitive impairment, including Alzheimer disease (AD)—independent of diabetes or of the complications that often accompany that disease.
The obesity epidemic is one of the most alarming public health concerns of our age—in large part because of its potential to lead to the development of diabetes.6,7 Given the rapid rise of obesity among young people, the late-life consequences of long-term obesity are garnering increased attention: recent findings indicate that midlife obesity and overweight may be associated with heightened risk of dementia, including AD, and vascular dementia.8
Results from the Cardiovascular Risk Factors, Aging, and Dementia study found that midlife obesity (body mass index [BMI] greater than 30 kg/m2) is related to dementia and AD.9 In a 21-year, follow-up, late-life cognitive assessment of 1449 individuals aged 65 to 79 years, researchers found significant 2-fold adjusted relative risk (RR) of dementia and AD. That risk was only mildly attenuated by further adjustment for midlife blood pressure, total cholesterol level, and smoking. In another long-term prospective study, Whitmer and colleagues10 found that midlife central obesity (as measured by sagittal abdominal diameter) was associated with a nearly 3-fold increased RR of dementia among 6583 adults 36 years later, even after adjusting for total BMI.
Notably, a dynamic relationship between obesity and cognitive aging has been described in which midlife obesity and increased adiposity appear related to risk of dementia, while late-life underweight and weight loss have also been found to be associated with increased dementia risk. Among 2798 adults in the Cardiovascular Health Study (mean age, 74.7 years) followed up for 5.4 years, the RR of dementia was significantly higher (RR, 1.4) among persons with midlife obesity (BMI, greater than 30) than among normal weight persons (BMI, 20 to 25). The pattern was reversed for late-life BMI: in late-life underweight persons (BMI, less than 20), the RR of dementia was significantly higher (RR, 1.6) than in normal weight persons.11
Evidence of this seeming paradox has been observed in other large-scale, long-term prospective studies.12 It appears that while midlife overweight and obesity are related to increased long-term risk of dementia, late-life underweight and weight loss may actually be early symptoms or markers of incipient dementia—rather than true short-term risk factors.
Obesity is a major risk factor for insulin resistance. Insulin resistance is typically accompanied by persistent elevations in blood insulin levels—a result of the reduced efficiency of cellular insulin uptake and utilization—and may have deleterious effects on cognitive aging. Insulin resistance is strongly associated with cerebral microvascular and macrovascular damage and may contribute to cognitive decline and vascular dementia. In addition, insulin resistance frequently accompanies elevations of inflammatory markers, such as C-reactive protein (CRP) and interleukin-6 (IL-6).13 Sustained increases in levels of inflammatory response compounds have been widely implicated in the development of vascular disease, but their involvement in the development of AD pathology has also been suggested.14
Furthermore, there may be cognitive aging effects of chronic hyperinsulinemia that are not mediated by vascular disease or injury. Insulin may directly affect levels of amyloid-β peptide (Aβ)—the primary component of neuritic plaques, a central element of AD pathology—which represents an alternative and intriguing pathway by which hyperinsulinemia may adversely affect brain health.15 Findings on the insulin-degrading enzyme provide a possible explanation for how hyperinsulinemia could lead to elevated levels of Aβ. Insulin-degrading enzyme is the major enzyme responsible for the breakdown of insulin in the body.16,17 It is also the first protease demonstrated to degrade Aβ; in laboratory studies, overexpression of insulin-degrading enzyme markedly reduced levels of both extracellular and intracellular Aβ.18-20 However, insulin-degrading enzyme binds more readily to insulin relative to other substrates, and insulin acts as a competitive inhibitor of Aβ degradation.16 Thus, the chronic hyperinsulinemia in insulin resistance may potentially interfere with Aβ clearance.21,22 Possible evidence for this in humans was provided in a report from Kulstad and colleagues.23 Patients with AD were found to have reduced insulin clearance and elevations in insulin-provoked plasma Aβ levels.
Nevertheless, an important distinction must be made between the effects of acute or temporary increases in insulin levels in the setting of normal metabolic function and those effects associated with insulin levels that are chronically elevated.21,24 An acute rise in insulin in response to glucose is a part of normal metabolism, but chronically high levels of blood insulin usually indicate insu-lin resistance. Finally, in addition to potential consequences of vascular damage and increased Aβ, Craft22 proposed that chronic peripheral hyperinsulinemia may actually induce a relative insulin-deprived state in the brain, which impairs normal glucose metabolism by neurons.
An emerging epidemiological literature has addressed the independent relationship of insulin resistance and chronic hyperinsulinemia (ie, in the absence of diabetes) to the risk of cognitive decline and dementia. Overall, it appears that the biologically compelling hypothesized relationship of insulin resistance to adverse brain health outcomes is being demonstrated in prospective, large-scale epidemiological studies. One example comes from the Washington Heights-Inwood Columbia Aging Project. Luchsinger and colleagues25 identified a doubling of risk for AD (RR, 2.3; 95% confidence interval, 1.5 - 3.6), with higher versus lower plasma fasting insulin levels among 531 elderly adults without diabetes.
Our research group recently followed up earlier cross-sectional investigations that identified significant associations between midlife elevated insulin secretion and risk of late-life cognitive impairment among men and women from 2 large-scale epidemiological cohorts.26,27 In 2 prospective analyses involving random samples of approximately 1200 women in the Nurses’ Health Study, elevated midlife fasting insulin and C-peptide (a measure of insulin secretion) levels were both significantly associated with more rapid decline at late life on 3 serial assessments of general cognitive function and episodic memory conducted over 4 years.28,29 Among participants in a random sample of 1353 men in the Physicians’ Health Study II, both higher fasting insulin and higher C-peptide levels were significantly related to greater average decline in general cognitive performance on 2 assessments conducted an average of 2 years apart (range, 1.5 to 4 years).30
Finally, Rnnemaa and coworkers31 reported that among 2322 men of the Uppsala cohort, midlife insulin resistance and impaired insulin response were significantly related to increased risk of AD and dementia more than 30 years later.
The metabolic syndrome combines aspects of both obesity and insulin resistance, as well as their likely consequences. The metabolic syndrome has been variably defined but typically involves core features of obesity (particularly central adiposity), dyslipidemia, hypertension, and evidence of insulin resistance or impaired glucose tolerance. The most commonly used criteria are from the National Cholesterol Education Program: Adult Treatment Panel III (NCEP-ATP III).32 Although the literature is still relatively scant, several recent reports suggest that the metabolic syndrome may be related to increased risk of cognitive impairment and dementia.
In a prospective study of 4895 women (mean age, 66 years) followed up for 4 years, the metabolic syndrome was associated with a nearly 70% increased adjusted RR of clinically significant cognitive impairment, including mild cognitive impairment and dementia.33 Moreover, because the NCEP-ATP III criteria do not preclude the presence of clinical diabetes, researchers performed analyses of data gathered from patients who met criteria for metabolic syndrome but who did not have diabetes, and the increased risk remained unchanged. The metabolic syndrome has been associated with a significant 3-fold increased RR of AD specifically, regardless of whether patients had diabetes.34,35
There may be important interactions on cognitive outcomes that involve the metabolic syndrome and the increased systemic inflammation that often accompanies insulin resistance. In a prospective study involving more than 2600 black and white community-dwelling elderly adults, the metabolic syndrome was associated with a 20% increased risk of cognitive decline36; however, the increased likelihood of cognitive dysfunction was strongly modified by the presence of high blood levels of inflammatory markers (CRP and IL-6), which is consistent with the potential role of inflammation in the relationship between insulin resistance and cognitive decline.
Although results have been inconsistent, markers of inflammation have been associated with risk of cognitive decline and dementia. In a cross-sectional study of CRP and late-life cognitive function among 447 persons aged 40 to 85 years (mean, 63 years), Wersching and colleagues37 observed that elevated CRP levels were associated with worse executive function and brain MRI measures of white matter integrity, including in the frontal lobe. By contrast, Women’s Health Study researchers found that plasma CRP levels were not significantly associated with cross-sectional cognitive performance among 4231 women aged 60 to 90 years who were tested 2 to 3 years after blood draw.38 Similarly, Dik and colleagues39 did not find a relationship between CRP or IL-6 and 3-year cognitive decline among 1284 adults aged 62 to 85 years at baseline.
However, in a 25-year follow-up of the Honolulu-Asia Aging Study, men who had higher levels of high-sensitivity CRP earlier in life had an increased RR of dementia later in life.40 Compared with men in the lowest quartile of CRP, men in the upper 3 quartiles had a 3-fold significantly increased RR of all dementias combined, as well as for AD and vascular dementia. Thus, it is possible that—as with midlife obesity—long-term rather than short-term exposure to high levels of inflammation may be most relevant to the risk of late-life cognitive decline and dementia.
Finally, there may be additional connections of metabolic syndrome to late-life cognitive dysfunction. For example, obesity, insulin resistance, and diabetes have been associated with depression, and inflammation may play a mechanistic role in the development of depression.41-45 A relationship between depressed mood and psychological distress and dementia has also been found (depression may not be merely an early manifestation of dementia).46 Indeed, depression and diabetes-related risk factors may have complex interactions that affect late-life cognition, which warrant further study.47,48
Metabolic risk factors have substantial implications for public health. Obesity is increasing among younger people, including children, and insulin resistance is increasingly prevalent in young and middle-aged adults. The Table provides an overview of increased hyperinsulinemia prevalence from 1984 through 2002.49 Although advancing age is still the most prominent risk factor for cognitive decline and dementia, the escalating prevalence of obesity, insulin resistance (including frank diabetes), and attendant metabolic problems could create a scenario in which the burden of late-life cognitive disorders will actually be far greater than that which would be anticipated by virtue of the age shift alone.
Fortunately, these metabolic risk factors are readily modifiable. Regular physical activity, healthy diet, and weight reduction/control can decrease the risk of obesity and metabolic problems. Physical activity promotes insulin sensitivity, reduces hyperinsulinemia, and lowers levels of inflammatory markers.50,51 Dietary replacement of high glycemic index carbohydrates (eg, refined sugars, products made with white flour) with minimally refined cereals and whole grains reduces the risk of insulin resistance and diabetes.52 Finally, obesity provokes insulin resistance; thus, a strategy of maintaining a healthy weight and avoiding weight gain during adulthood is considered “the cornerstone of diabetes prevention.”6
A potential adjunct to these prevention strategies is the development of pharmacotherapies that target diabetes-related biology and offer either enhancement or protection of late-life cognitive function. Some preliminary work shows promise in this regard. For example, Watson and colleagues53 reported on the potential role of a class of diabetes drugs called peroxisome proliferator-activated receptor-γ (PPAR-γ) agonists. These drugs enhance insulin sensitivity and appear to reduce blood levels of inflammatory markers, such as CRP and IL-6, even in persons without diabetes.54
At the end of a double-blind, placebo-controlled, 6-month study of 30 patients with mild AD or amnestic mild cognitive impairment, those who received rosiglitazone (a PPAR-γ agonist) demonstrated better performance on a task of delayed recall than placebo recipients.53 However, the recent alarm about the adverse cardiovascular profile of rosiglitazone may limit exploration of such alternative uses for PPAR-γ agonists, and researchers may turn their attention to the development of other insulin-sensitizing interventions.55
As exploration of pharmacological options continues, clinical psychiatrists can adopt a proactive role in addressing the numerous nonpharmacological means of reducing the likelihood of adverse cognitive aging effects of diabetes-related factors with their patients. Specifically, psychiatrists can aim to incorporate focused counseling on pertinent life-style modifications into their work:
• Recommendations for regular physical activity sufficient to break a sweat (30 minutes a day, 3 to 5 times a week)
• Referral to nutrition consultation and/or weight management programs for patients who are overweight or obese
• Diligent monitoring of metabolic factors (eg, levels of fasting glucose, lipids, hemoglobin A1c) for patients at high risk for obesity or metabolic syndrome
In summary, a growing body of research indicates that diabetes-related risk factors pose potent threats to healthy cognitive aging, even in the absence of actual diabetes. Psychiatrists can play their part in dementia prevention by monitoring closely the cognitive function of their older patients who already have diabetes, and encouraging patients of all ages to address modifiable risk factors for obesity, insulin resistance, and diabetes.
References1. Ott A, Stolk RP, van Harskamp F, et al. Diabetes mellitus and the risk of dementia: the Rotterdam Study. Neurology. 1999;53:1937-1942.